University of Southern California Dissertations and Theses

The extent and effectiveness of nursing home regulation in the 50 states

(USC Thesis Other)

The extent and effectiveness of nursing home regulation in the 50 states

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE EXTENT AND EFFECTIVENESS OF NURSING HOME REGULATION IN THE 50 STATES Copyright 2004 by Christopher Michael Kelly A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (GERONTOLOGY) August 2004 Christopher Michael Kelly R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UMI Number: 3145215 Copyright 2004 by Kelly, Christopher Michael All rights reserved. INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 3145215 Copyright 2004 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Dedication This dissertation is dedicated with love to my grandparents, Conrad and Eugenia Zalewski and John and Ruth Kelly. Each of you inspired me to pursue a career as a gerontologist. You have been with me in spirit every step of the way. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. iii Acknowledgments I would first like to express my appreciation to Dr. Phoebe Liebig, the chair of my committee, who always had faith that this was within me. It has been a long and ultimately rewarding journey and I am better for the experience. My sincere thanks to the members of my committee, Dr. Eileen Crimmins and Dr. Mike Nichol, each of whom challenged me to find new strengths. Dr. Charlene Harrington, Dr. Clive Thomas, Dr. Peter Schmidt, Eric Anderson, Rob Kennison, and Jeff Hyde all generously provided their time and their insights to me throughout this endeavor. I would also like to thank Dr. Jon Pynoos, who has been both a friend and a mentor and Maria Henke, who has always found a home for me in the Division of Policy and Services Research. I have been blessed with many friends who have shared the good and bad times with me during my years at USC. I would especially like to recognize the following: Elizabeth Arlotti, Eric Badura, John Bauer, Marianne Case, Danielle Colayco, Aaron Hagedom, Brian Hajek, Russ Kava, Matt Leal, Dominic Leslie, Michele Maines, T.J. McCallum, Fr. Bill Messenger, Huy Nguyen, Betty Oswald, Alex Rosales, Dory Sabata, George Shannon, Leslie Vitin, and Sarah Winningham. Thank you for the inspiration each of you has uniquely given me. Most importantly, I would like to thank the members of my family, who have been in my comer every day: my parents Michael and Dianne Kelly, my sister Joanna Kelly Meysenburg and my brother Patrick Kelly. The six years and hundreds of phone calls were worth it. Thank you for your love, patience, and belief in me. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. iv Table of Contents Dedication ii Acknowledgments iii List of Tables vi List of Figures xiii Abstract xiv Preface xvi Chapter 1: Introduction 1 Nursing Homes 1 The Development of Nursing Home Regulation 9 Present Study 15 Chapter 2: Literature Review 21 Regulation and Regulatory Federalism 22 State Nursing Home Regulation 31 Understanding State Regulatory Policy: Theoretical Models 43 Chapter 3: Methods 63 Sample 64 Outcome Measures 65 External and Internal Determinants 70 Statistical Analysis 135 Chapter 4: Results 141 The Extent and Effectiveness of State Nursing Home Regulation 142 Factors Predicting State Nursing Home Regulation 162 Theoretical Models and HLM Analysis 202 Summary of Factors Predicting State Nursing Home Regulation 211 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. V Chapter 5: Discussion 214 The Extent and Effectiveness of State Nursing Home Regulation 215 Factors Predicting State Nursing Home Regulation 220 Theoretical Models 232 Dissertation Summary 237 References 240 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. vi List of Tables Table 1: CMS Scope and Severity Grid for Nursing Home Deficiencies 13 Table 2: Independent Variables Predicting State Nursing Home Regulation 71 Table 3: Economic Development and Proportion of Oldest-Old 74 Table 4: Correlations among Economic Development Variables 76 Table 5: Auxiliary Regressions among Economic Development Variables 77 Table 6: Dominant Political Subcultures in the 50 States 80 Table 7: CMS Expenditures for State NH Survey/Certification 82 Table 8: Number of Nursing Home Beds in the 50 States 83 Table 9: CMS Expenditures for State NH Inspections, per NH Bed 85 Table 10: Correlations among External Determinants 87 Table 11: Total Expenditures for NH Survey/Certification; State Proportion 90 Table 12: Total Expenditures for State NH Survey/Certification, per NH Bed 92 Table 13: Number of State NH Surveying Agency Employees 95 Table 14: Number of NH Beds/FTE State Agency Employee 96 Table 15: Number of Years of Experience of State NH Surveyors 98 Table 16: Correlations among State Surveying Agency Variables 100 Table 17: Political Control of State Legislatures 103 Table 18: Legislative Professionalism 105 Table 19: Aging and Long-Term Care Committees in State Legislatures 108 Table 20: Correlations among Legislative Variables 109 Table 21: Influence of Nursing Home Associations and Senior Citizens 112 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. vii List of Tables (Continued) Table 22: Interest Group State Offices/Offices in State Capitals 114 Table 23: Correlations among Interest Group Variables 115 Table 24: Political Party of State Governors 117 Table 25: Governors’ Institutional Powers 119 Table 26: Governors’ References to Nursing Homes in SOS Addresses 122 Table 27: Correlations among Gubernatorial Variables 123 Table 28: Correlations among Internal Determinants 126 Table 29: Correlations among Independent Variables 130 Table 30: Independent Variables Predicting State Nursing Home Regulation 134 (Revised) Table 31: Number of State NH Deficiency Citations per 100 NH Beds 144 Table 32: Percentage of State NH Deficiency Citations G and Above 147 Table 33: Correlations among Dependent Variables, Extent 148 Table 34: Repetitive Scheduling of Annual State NH Inspections 151 Table 35: Number/Severity of State Nursing Home Deficiency Citations 153 Table 36: Average Number of Days for Deficiency Resolution by the State 155 Table 37: Correlations among Dependent Variables, Deficiency Resolution 158 Table 38: Correlations among Dependent Variables, Effectiveness 159 Table 39: Correlations among Dependent Variables 160 Table 40a: Correlations among External Determinants and Outcome Measures, 163 Year 1 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. viii List of Tables (Continued) Table 40b: Correlations among External Determinants and Outcome Measures, 163 Year 2 Table 40c: Correlations among External Determinants and Outcome Measures, 163 Year 3 Table 41a: Multiple Regression of Volume of State Deficiency Citations on 165 External Determinants Table 41b: Multiple Regression of Change in Volume of State Citations on 165 External Determinants Table 42a: Multiple Regression of Severity of State Deficiency Citations on 166 External Determinants Table 42b: Multiple Regression of Change in Severity of State Citations on 166 External Determinants Table 43 a: Multiple Regression of Repetitive State Inspection Scheduling on 168 External Determinants Table 43b: Multiple Regression of Change in Repetitive Inspection Scheduling 168 on External Determinants Table 44a: Multiple Regression of State Resolution of C-G Deficiencies on 169 External Determinants Table 44b: Multiple Regression of Change in C-G Deficiency Resolution on 169 External Determinants Table 45a: Multiple Regression of State Resolution of H-L Deficiencies on 169 External Determinants Table 45b: Multiple Regression of Change in H-L Deficiency Resolution on 170 External Determinants Table 46a: Correlations among Agency Determinants and Outcome Measures, 171 Year 1 Table 46b: Correlations among Agency Determinants and Outcome Measures, 171 Year 2 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. ix List of Tables (Continued) Table 46c: Correlations among Agency Determinants and Outcome Measures, 171 Year 3 Table 47a: Multiple Regression of Volume of State Deficiency Citations on 173 Agency Determinants Table 47b: Multiple Regression of Change in Volume of State Citations on 173 Agency Determinants Table 48a: Multiple Regression of Severity of State Deficiency Citations on 174 Agency Determinants Table 48b: Multiple Regression of Change in Severity of State Citations on 174 Agency Determinants Table 49a: Multiple Regression of Repetitive State Inspection Scheduling on 176 Agency Determinants Table 49b: Multiple Regression of Change in Repetitive Inspection Scheduling 176 on Agency Determinants Table 50a: Multiple Regression of State Resolution of C-G Deficiencies on 177 Agency Determinants Table 50b: Multiple Regression of Change in C-G Deficiency Resolution on 177 Agency Determinants Table 51a: Multiple Regression of State Resolution of H-L Deficiencies on 178 Agency Determinants Table 51b: Multiple Regression of Change in H-L Deficiency Resolution on 178 Agency Determinants Table 52a: Correlations among Legislative Determinants and Outcome 179 Measures, Year 1 Table 52b: Correlations among Legislative Determinants and Outcome 179 Measures, Year 2 Table 52c: Correlations among Legislative Determinants and Outcome 179 Measures, Year 3 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. X List of Tables (Continued) Table 53a: Multiple Regression of Volume of State Deficiency Citations on 181 Legislative Determinants Table 53b: Multiple Regression of Change in Volume of State Citations on 181 Legislative Determinants Table 54a: Multiple Regression of Severity of State Deficiency Citations on 182 Legislative Determinants Table 54b: Multiple Regression of Change in Severity of State Citations on 182 Legislative Determinants Table 55a: Multiple Regression of Repetitive State Inspection Scheduling on 184 Legislative Determinants Table 55b: Multiple Regression of Change in Repetitive Inspection Scheduling 184 on Legislative Determinants Table 56a: Multiple Regression of State Resolution of C-G Deficiencies on 185 Legislative Determinants Table 56b: Multiple Regression of Change in C-G Deficiency Resolution on 185 Legislative Determinants Table 57a: Multiple Regression of State Resolution of H-L Deficiencies on 186 Legislative Determinants Table 57b: Multiple Regression of Change in H-L Deficiency Resolution on 186 Legislative Determinants Table 58a: Correlations among Interest Group Determinants and Outcome 187 Measures, Year 1 Table 58b: Correlations among Interest Group Determinants and Outcome 187 Measures, Year 2 Table 58c: Correlations among Interest Group Determinants and Outcome 188 Measures, Year 3 Table 59a: Multiple Regression of Volume of State Deficiency Citations on 189 Interest Group Determinants R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. xi List of Tables (Continued) Table 59b: Multiple Regression of Change in Volume of State Citations on 189 Interest Group Determinants Table 60a: Multiple Regression of Severity of State Deficiency Citations on 190 Interest Group Determinants Table 60b: Multiple Regression of Change in Severity of State Citations on 190 Interest Group Determinants Table 61a: Multiple Regression of Repetitive State Inspection Scheduling on 192 Interest Group Determinants Table 61b: Multiple Regression of Change in Repetitive Inspection Scheduling 192 on Interest Group Determinants Table 62a: Multiple Regression of State Resolution of C-G Deficiencies on 193 Interest Group Determinants Table 62b: Multiple Regression of Change in C-G Deficiency Resolution on 193 Interest Group Determinants Table 63a: Multiple Regression of State Resolution of H-L Deficiencies on 194 Interest Group Determinants Table 63b: Multiple Regression of Change in H-L Deficiency Resolution on 194 Interest Group Determinants Table 64a: Correlations among Gubernatorial Determinants and Outcome 195 Measures, Year 1 Table 64b: Correlations among Gubernatorial Determinants and Outcome 195 Measures, Year 2 Table 64c: Correlations among Gubernatorial Determinants and Outcome 196 Measures, Year 3 Table 65a: Multiple Regression of Volume of State Deficiency Citations on 197 Gubernatorial Determinants Table 65b: Multiple Regression of Change in Volume of State Citations on 197 Gubernatorial Determinants R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. xii List of Tables (Continued) Table 66a: Multiple Regression of Severity of State Deficiency Citations on Gubernatorial Determinants Table 66b: Multiple Regression of Change in Severity of State Citations on Gubernatorial Determinants Table 67a: Multiple Regression of Repetitive State Inspection Scheduling on Gubernatorial Determinants Table 67b: Multiple Regression of Change in Repetitive Inspection Scheduling on Gubernatorial Determinants Table 68a: Multiple Regression of State Resolution of C-G Deficiencies on Gubernatorial Determinants Table 68b: Multiple Regression of Change in C-G Deficiency Resolution on Gubernatorial Determinants Table 69a: Multiple Regression of State Resolution of H-L Deficiencies on Gubernatorial Determinants Table 69b: Multiple Regression of Change in H-L Deficiency Resolution on Gubernatorial Determinants Table 70: Summary of Multiple Regression Analyses: Outcome Measures and their Determinants Table 71: HLM Estimates of Volume of State Deficiency Citations Table 72: HLM Estimates of Severity of State Deficiency Citations Table 73: HLM Estimates of Repetitive Scheduling of Annual State Inspections Table 74: HLM Estimates of Resolution of C-G Deficiencies Table 75: HLM Estimates of Resolution of H-L Deficiencies Table 76: Summary of HLM Analyses: Outcome Measures and their Determinants 198 198 200 200 201 201 202 202 203 205 206 208 209 211 2 1 2 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. xiii List of Figures Figure 1: Dye’s Policy Formation Model 46 Figure 2: Dominant Political Subcultures in the 50 States 49 Figure 3: Sharkansky’s Adaptation of Elazar’s Political Culture Model 50 Figure 4: The Advocacy Coalition Framework (ACF) 54 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. xiv Abstract This dissertation assessed the extent and the effectiveness of state nursing home regulation by analyzing all the deficiency citations issued by the 50 state surveying agencies during a recent three-year period (1999-2002). Extent was assessed as the volume and severity of deficiency citations. Citation volume was stable over the three-year period, with California having the most citations per 100 nursing home beds and Wisconsin the fewest. However, the severity of citations steadily declined from 1999 to 2002; Oregon had the highest proportion of actual harm deficiencies and California the lowest. Effectiveness was measured by the success of the states in implementing federal mandates to: avoid scheduling a nursing home’s annual inspection during the same month two years in a row and 2) resolve non-actual harm deficiencies within 60 days and actual harm deficiencies within 30 days. Georgia was the most successful state in avoiding repetitive scheduling of state inspections; Iowa was the least successful. Alaska resolved non-actual harm deficiencies in the shortest period of time, while Connecticut took the longest time. Maine was the fastest in correcting actual harm problems; Alaska the slowest. The impact of external state characteristics and factors related to key actors in state nursing home policy (state surveying agency, legislature, interest groups, and governor) on state nursing home regulation was then tested. Multiple regression and hierarchical linear modeling (HLM) indicated that external and gubernatorial determinants predict the extent of state nursing home regulation, while agency and R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. legislative determinants predict its effectiveness. HLM further revealed that theoretical models integrating variables external and internal to the state nursing home policy subsystem are useful in explaining interstate variation in state nursing home regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. xvi Preface This dissertation is a product of more than ten years of personal and professional experience with nursing home care in the United States. During this time, I have witnessed the wide range of quality that exists among these facilities. I have learned of persistent abuse and neglect of frail residents in a number of nursing homes. However, I have also visited nursing homes that provide excellent services and a pleasant living environment. I have worked alongside nursing home staff members personally committed to preserving the health and dignity of the residents in their care. Moreover, I have come to understand the continued necessity of nursing homes for the growing numbers of Americans who require round-the-clock skilled nursing care. The roots of this particular research can be traced to a specific event that occurred when I worked as an activities coordinator in a nursing home in northern California. One morning I asked the administrator the reason for the additional staffing I had noticed that week, and I was told that the nursing home was preparing for the state’s annual “surprise” inspection. Indeed, inspectors from the state surveying agency arrived later that day. It occurred to me then that the purpose of a “surprise” inspection was defeated when the timing of the survey could reliably be predicted by the inspected. I learned that day that something was amiss with the regulation of nursing homes and I wanted to understand why. In this dissertation, I hope to help answer this question. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 Chapter 1 Introduction The purpose of this dissertation is to examine the extent and effectiveness of nursing home (NH) regulation in the 50 states and to identify factors that predict interstate variation in its extent and effectiveness. The study of state enforcement is important because it is the states that are responsible for ensuring that NHs meet national standards for quality of care (Centers for Medicare and Medicaid Services (CMS), 2003a). The consequence of interstate variation in this enforcement is that NH residents in the United States receive unequal protection from abuse and neglect. The first part of the introduction chronicles the evolution of NH care in the United States from private board-and-care homes to an industry dominated by large corporations. The second part describes the current state ofNHs in the United States: demographic characteristics, the primary means of payment and the composition of NH staffing. The third part describes the development of state NH regulation, the growing body of research into its extent and effectiveness, and the place of the present study in this literature. Nursing Homes The nursing home (NH), the foundation of long-term care in the United States, is defined by the U.S. Department of Health and Human Services (USDHHS) as a facility with three or more beds that routinely provides nursing services. A NH R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 can be a freestanding facility or a separate unit of a larger entity, such as a hospital or a continuing care retirement community (USDHHS, 2002). According to the U.S. General Accounting Office (USGAO), approximately 17,000 NHs currently provide care for more than 1.7 million persons (USGAO, 2003) at an average annual cost of more than $46,000 per resident (USDHHS, 2002). The nation spends more than $90 billion a year on NH care; 60 percent of this cost is paid by public funds through the Medicare and Medicaid programs (Walshe & Harrington, 2002). History of Nursing Homes The modem NH is a product of the major age-based entitlement programs of fh the 20 century. The Social Security Act, passed in 1935 as part of the “New Deal” program of President Franklin D. Roosevelt, improved the standard of living for older Americans. Social Security provided cash assistance to the elderly and disabled, giving older Americans the means to live their last years in dignity (Steuerle & Bakija, 1994). In particular, Social Security reduced the reliance of the elderly on the “outdoor relief’ provided by local government and private charity (Liebig, 1998). However, the closure of public almshouses left a housing demand for the frail elderly who depended on these institutions for shelter. This demand was partially filled by private board-and-care homes, largely established as a source of extra income for homeowners during the 1930s and 1940s (Vladeck, 1980). The private R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 board-and-care homes of the Great Depression era are today regarded as the first NHs in the United States (Kelly & Liebig, 2003). After World War II, these mom- and-pop operations developed into larger institutions. National laws, such as the Hill-Burton Act of 1946, established building standards, bringing further change to NHs. New facilities were subsequently constructed based on the hospital model. The passage of Medicare and Medicaid in 1965 during the “Great Society” era of President Lyndon B. Johnson further transformed the NH industry. Medicare was created to provide health insurance for older Americans, but coverage was primarily limited to acute care, not to long-term care (LTC). However, Medicaid, which was established to provide health coverage to the poor, contained a provision permitting states to enroll persons who became medically needy due to high medical costs, such as high LTC expenses. Because of this “sleeper” policy, Medicaid became the largest purchaser of LTC services in the United States and NHs emerged as the primary recipients of Medicaid LTC payments (Liebig, 1998). The NH population soared following the passage of Medicaid. The number of facilities tripled between 1960 and 1980, and the NH population grew to over 1.6 million during this period (Vladeck, 1980). As NHs grew, the care provided in these facilities evolved, becoming more diversified, specialized and medicalized (Redfoot, 2003). One example is the development of Special Care Units (SCUs), now located in over 20 percent of all NHs, mostly devoted to the care of Alzheimer’s patients. SCUs have specific admission and discharge criteria, specially trained staff and R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 personal care strategies and a secured area, such as a designated floor or wing of a building (Kelly & Liebig, 2003). Despite innovations such as SCUs, NH growth slowed during the 1980s, and NH utilization rates have declined over the past two decades. Explanations for these trends include the development of assisted living facilities, increased Medicaid reimbursement for home-and-community-based services, and expanded LTC insurance coverage for non-institutional care. In addition, many states have passed Certificate of Need (CON) laws designed to control the supply of NH beds. Nursing Homes Today Today, more than 1.7 million Americans live in NHs at any one time. Most NH residents (90 percent) are seniors age 65 years or over; 46 percent of residents are age 85 years and older (USDHHS, 2002). Roughly 10 percent of the NH population is younger than age 65; these are primarily persons who suffer from mental illness or other central nervous system problems (Kelly & Liebig, 2003). The majority of NH residents are white (88.7 percent); 8.9 percent are black (Bishop, 1999). Minority groups are underrepresented in NHs, for reasons including poverty and cultural preference. For many minority families, the only financial option for LTC is a NH that exclusively treats Medicaid patients. In addition, few NHs serve populations such as Asians and Native Americans (Aeschleman, 2000). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 Almost three-quarters of NH residents are female, which reflects the higher rates of chronic illness, longer life expectancy, lower marital rates and lower income of older women (Harrington Meyer, 2001). More than half of the NH population is widowed (USDHHS, 2002). More than three-quarters of NH residents require assistance with bathing and dressing and nearly half receive help with eating and using the toilet (USDHHS, 2002). Finally, the NH market is dominated by large corporations. Nearly two- thirds of NHs are in the for-profit sector, with 56 percent owned or operated by chains (USDHHS, 2002). Voluntary and nonprofit organizations and the government (largely through the Veterans Administration) operate the remainder (USDHHS, 2002). The small, private “mom-and-pop” operations are now a part of the past. Financing Nursing Home Care Medicare Medicare, a federal program created in 1965 under Title XVIII of the Social Security Act, provides health insurance to older Americans. However, Medicare was modeled on the employer-based health insurance plans of its day and, as such, was designed to reimburse acute care providers, such as hospitals and physicians. Title XVIII made no provision for therapies such as prescription drugs or LTC. Today, Medicare reimbursement of LTC providers such as NHs is primarily limited to short- R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 term rehabilitative care for patients following a hospital stay of three or more days (Kelly & Liebig, 2003). Medicare currently pays for 14 percent of NH care in the nation (CMS, 2003b; Walshe & Harrington, 2002). Medicare pays for up to 100 days of post-acute care in a NH following a hospital stay of three or more days. Prior to 1999, NHs billed Medicare retrospectively for each service; this encouraged NHs to provide Medicare residents with as much care as possible. However, under the Balanced Budget Act of 1997 (BBA), NHs assign a fixed daily rate prospectively for every incoming Medicare patient based on the patient’s Resource Utilization Group (RUG). The RUG anticipates the cost of medication, supplies, rehabilitative therapies and other patient needs; it gives NHs an incentive to ration care (Kane, 1998). The BBA was later modified under the Balanced Budget Refinement Act of 1999 (BBRA) to allow a gradual phase-in of the RUG (Angelelli, Mor, Intrator, Feng, & Zinn, 2003). Medicaid Medicaid, established under Title XIX of the Social Security Act to provide need-based health insurance, is funded jointly by the federal government and the states. However, discretion over the patient care to be reimbursed under Medicaid is left to the states. Medicaid is the largest public payer of NH care in the United States. Medicaid spent $39.6 billion in 2000 and currently covers 46 percent of all NH care in the country (CMS, 2003b; Walshe & Harrington, 2002). Medicaid pays R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 the NH costs for an indefinite period for residents who are eligible under their state’s “medically needy” (MN) option. Persons may qualify as MN immediately or may become Medicaid-eligible by “spending down” their income to their state’s MN income level (CMS, 2003b). Medicaid systems vary from state to state. The states vary in their Medicaid payment policies, such as whether services such as physical therapy and prescription drugs are included in daily rates. Medicaid per diem rates for NHs also vary across the country; in some states Medicaid NH rates are only 70 to 80 percent of private pay rates. This may create an incentive for NHs to admit private pay patients rather than Medicaid patients (Kelly & Liebig, 2003). Finally, state Medicaid programs vary in their use of managed care. Arizona’s entire LTC system is capitated and Texas and Minnesota have recently implemented managed care LTC programs (Burwell, 2001). However, managed care has not had a large impact on Medicaid NH reimbursement in most states (Kelly & Liebig, 2003). Long-Term-Care Insurance The remaining 40 percent of NH care is paid through private sources, mostly out-of-pocket, by patients and their families (Kelly & Liebig, 2003). An alternative to public programs and out-of-pocket spending for NH payment is long-term-care insurance (LTCI). In the mid 1980s, LTCI barely existed; but by 1999, more than 6 million policies had been sold. Two different models of LTCI exist. Most insurers R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 follow the indemnity model, covering only eligible expenses for specified services. Under indemnity policies, the insured person purchases a particular amount of daily services in a NH or other residential care setting. By contrast, the disability model does not finance a prescribed set of services. Instead, disability plans provide a cash benefit or voucher directly to insured persons to spend as they see fit (Stone, 2001). LTCI is expected to expand, due to an increase of the numbers of employers and insurers entering the market, and the growing interest in LTCI among aging baby boomers (Cohen, Miller, & Weinrobe, 2001). Nursing Home Staffing More than 1.5 million full-time employees provide care to NH residents in the United States. Almost two-thirds of these employees are nursing staff providing direct services, such as personal care, to residents. These include registered nurses (RNs), licensed practical nurses (LPNs), certified nursing assistants (CNAs) and orderlies. CNAs and orderlies account for 64 percent of all nursing staff. A much smaller percentage of NH employees (six percent) provide medical and therapeutic services to residents; this category includes dentists, physical therapists, speech pathologists, podiatrists and social workers (USDHHS, 2002). The physician has a limited role in daily NH care. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 The Development of Nursing Home Regulation Quality of Care Problems in Nursing Homes The quality of NH care in the United States has been a public policy concern for more than 30 years (Walshe & Harrington, 2002). During the 1970s, reports in the mainstream media described NHs as “houses of death”, “concentration camps” and “warehouses of the dying” (Vladeck, 1980: 3). In 1980, Vladeck reported that thousands of facilities in every state had substandard sanitation, staffing and patient care. A 1986 report by the Institute of Medicine (IOM) documented the widespread abuse and neglect of NH residents and recommended the strengthening of national NH standards. Following publication of the IOM report, the U.S. Congress enacted reforms of federal and state NH regulation, as part of the Omnibus Budget Reconciliation Act of 1987 (OBRA 87). In spite of these reform efforts, NH quality-of-care problems have persisted. The proportion of NHs with serious quality problems remains unacceptably high (USGAO, 2003). A 2003 study by the USDHHS, Office of Inspector General (OIG) reported at least one instance of substandard care in 78 percent of the nation’s NHs. Inspections of NHs continue to describe evidence of untrained staff, inadequate therapies, poor food and unsafe living conditions (USGAO, 2000, 2003). No other segment of the health care industry has been documented to have such poor quality of care as the nation’s NHs (Harrington, 1999). However, the focus has increasingly R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 10 shifted to the public officials responsible for detecting and correcting NH problems (USGAO, 2000, 2003; Walshe & Harrington, 2002; USDHHS, 2003). The Intergovernmental Regulation of Nursing Homes Today, the responsibility for correcting NH problems is shared by the federal government and the states. However, this intergovernmental policy is a relatively recent arrangement. Prior to 1971, NH quality assurance was left entirely to state governments. During the 1950s, most states established licensing requirements for NHs and state health departments began conducting NH inspections (Vladeck, 1980). In 1965, following the passage of Medicare and Medicaid, federal law established “conditions of participation” for health care facilities seeking reimbursement under these programs. However, enforcement of these conditions was left to the states (Kane, 1995). Active federal participation in NH policy began after a visit by President Richard M. Nixon to a southern Illinois NH in 1971 (Netting, Huber, Paton, & Kautz, 1995). Aware of media reports of abuse and neglect of residents, the President asked government official Arthur Flemming what could be done to improve the nation’s NHs (Coleman, 1991). This discussion led to President Nixon’s Eight Point Initiative to improve NH care. As a part of this initiative, federal inspectors from the U.S. Department of Health, Education and Welfare (HEW) began to “validate” state NH inspections. Federal oversight was assumed by R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 the newly created Health Care Financing Administration (HCFA) in 1977 (Vladeck, 1980). Congress introduced sweeping changes to federal and state NH regulation under OBRA 87. NHs were required to provide comprehensive assessments of residents at admission, and periodically thereafter, to ensure that residents receive an appropriate level of care (Kelly & Liebig, 2003). OBRA 87 also clarified the roles of the states and the federal government in NH regulation. State NH surveying agencies were required to inspect every facility in a state every nine to 15 months and to investigate all quality of care complaints. HCFA was required to provide federal oversight by conducting its own inspections of five percent of the NHs in every state (USGAO, 1999). In 1995, Congress enacted current national NH regulatory policy. The states are required to specify the scope and severity of NH deficiencies, with sanctions ranging from fines to facility closure (USGAO, 1999a). HCFA, which was renamed the Centers for Medicare and Medicaid Services (CMS) in 2001, continues to provide oversight of this process. The Current Nursing Home Survey and Certification Process Under federal law, the states are required to conduct an annual survey of every NH that receives Medicare or Medicaid reimbursement. An annual survey involves a team of inspectors from the state surveying agency spending several days R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 12 in a facility in order to assess its compliance with federal requirements. The state inspection team determines whether the NH meets national minimum standards of providing adequate quality care, such as preventing avoidable pressure sores, weight loss or accidents (Mullan & Harrington, 2001; USGAO, 2003). State surveyors issue a deficiency citation for every instance in which a NH is noncompliant with federal regulations. Deficiencies are classified in one of 12 categories according to their scope and severity (see Table 1). The states are required to report every NH deficiency citation to the federal government by entering this information into the Center for Medicare and Medicaid Services’ (CMS) On-line Survey, Certification and Reporting (OSCAR) system (USGAO, 2003). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 13 Table 1: CMS Scope and Severity Grid for Nursing Home Deficiencies Scope Severity Isolated Pattern Widespread Required Optional Sanction Sanction Actual or potential for death/serious injury J K L Group 3 Group 1 or 2 Other actual harm G H I Group 2 Group 1 Potential for more than minimal harm D E F Group 1 for categories D and E; group 2 for category F Group 2 for categories D and E; group 1 for category F Potential for minimal harm (substantial compliance) A B C None None Group 1 sanctions are directed plan of correction, directed in-service training, and/or state monitoring. Group 2 sanctions are denial of payment for new admissions or all individuals and/or civil monetary penalties of $50 to $3,000 per day of noncompliance. Group 3 sanctions are temporary management, termination, and/or civil monetary penalties of $3,050 to $10,000 per day of noncompliance. Source: USGAO (1999b) The Study of State Nursing Home Regulation Previous Research A growing body of literature describes the extent of state enforcement of national standards of NH quality (Harrington & Carrillo, 1999; Grabowski, 2001; Mullan & Harrington, 2001; Walshe, 2001). These studies suggest that while overall NH deficiencies have declined since 1990, continued quality-of-care problems persist. For example, Harrington and Carrillo (1999) reported that of seven R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 14 deficiencies considered indicators of poor patient care, only one (restraint use) had declined significantly between 1991 and 1997. This literature has also identified the states in which NH enforcement activity has been the most and the least extensive. These results reveal wide interstate variation in the extent of state regulatory activity. For example, in 1997, Nevada had the highest average number of deficiencies per NH (14.3), while New Mexico (1.8) reported the lowest (Harrington & Carrillo, 1999). However, Wiener (2003) doubts that NH quality varies from state to state as much as these statistics suggest. In contrast, there has been relatively little research into the effectiveness of state NH regulation and limitations exist in the available data on state surveying agencies (USGAO, 2000; Walshe & Harrington, 2002). First, the quality of state NH regulation is often inferred from the extent of its deficiency citations; however, it is difficult to discern whether this information reflects on NHs or on the state agency inspecting them (Wiener, 2003). Second, while different measures of state surveying agencies have been suggested, such as effectiveness of state implementation of federal NH initiatives (USGAO, 2000), these measures have not been collected in all 50 states. Consequently, the extent and effectiveness of state NH regulation across the country are not well-understood. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 15 Present Study This dissertation is designed to fill this gap in the existing literature by examining the extent and the effectiveness of NH regulation by the 50 state surveying agencies. This level of analysis is critical because it is the states that are responsible for ensuring that NHs meet national standards for quality of care (CMS, 2003c). In this study, state NH deficiency citations issued over a three-year period (1999-2002) were collected and outcome measures for the extent and effectiveness of state enforcement were created. The extent of NH regulation was assessed through the volume and the severity of state NH deficiency citations, while effectiveness was measured as the success of the state surveying agencies in implementing recent federal initiatives designed to improve state NH regulation. This study then evaluated these outcome measures through linear modeling. First, multiple regression models were used to determine which factors predicted the extent and effectiveness of state NH regulation in a particular year, as well as change in extent and effectiveness from year to year. Factors related to the states’ external environment, its state surveying agency, the state legislature, interest groups (representing the NH industry and older adults) and the governor were tested. Factors found in the multiple regression models to have significant effects on extent and effectiveness were then tested using hierarchical linear modeling (HLM). HLM facilitated the selection of the single model in this study that best explains each outcome measure over the three-year period of observation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 16 Research Questions Upon completing these research objectives, this dissertation answers the following questions: 1. What is the extent of NH regulation (as measured by the volume and severity of NH deficiency citations) in the 50 states? 2. What is the effectiveness of NH regulation (as measured by the scheduling of annual inspections and the timeliness of deficiency resolution) in the 50 states? 3. Which determinants predict the extent and effectiveness of state NH regulation during each year of the study? 4. Which determinants predict change in the extent and effectiveness of state NH regulation from year to year in the study? 5. Which theoretical model best explains each measure of extent and effectiveness in this study? 6. What are the implications of these findings for further research? Contributions The current research represents several advances to the study of state NH regulation. Previous studies reported either the number of NH citations in the 50 states, but not their severity (Harrington & Carrillo, 1999; USDHHS, 2003). This R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 17 may falsely identify states in which the majority of NH deficiencies were minor as having the most serious NH problems. Other studies have reported the severity of NH citations, but only in a small sample of states (USGAO, 2000, 2003), which limits their ability to establish national patterns in the extent of state NH enforcement across the country. In this dissertation, the analysis of the volume and severity of deficiency citations issued by all 50 state surveying agencies addresses the shortcomings of previous studies of the extent of state NH regulation. This dissertation also enhances the understanding of the effectiveness of state NH regulatory activity. Previous studies of state implementation of national NH policy (Harrington & Carrillo, 1999; Mullan & Harrington, 2001) reported the extent of state issuance of particular quality-of-care deficiencies prioritized by the CMS. This approach yields valuable insight into interstate differences in NH quality, but is limited in illuminating the performance of state surveying agencies. It is difficult to determine whether a high frequency of state NH deficiency citations reflects poor NH care in a state or vigorous NH enforcement by a state. The present study addresses this dilemma by constructing outcome measures for the effectiveness of state regulatory activity as well as its extent. Effectiveness is operationalized as the success with which states fulfill two objectives of NH regulation described by the USGAO (2000): 1) to avoid repetition in scheduling annual NH inspections and 2) to ensure timely resolution of quality-of-care deficiencies. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 18 Further, this dissertation advances the literature through its use of theoretical models to explain interstate variation in state NH regulation. Studies of environmental protection (Gormley, 1986) and occupational safety enforcement (Scholz, Twombly, & Headrick, 1991) have concluded that the extent and effectiveness of a state regulatory policy are determined by factors both external and internal to the state policy subsystem. This dissertation applies this approach to the study of state NH regulation. It employs multiple regression to identify important factors, both external and internal to the state NH policy subsystem, that account for differences among states in the extent and the effectiveness of NH enforcement. Using HLM, this study then proposes a single model to predict each outcome measure of extent and effectiveness of state NH regulation. The fact state NH enforcement varies is well-understood; this dissertation helps to explain why interstate variations exist in what was designed to be a uniform national NH policy. In summary, this dissertation enhances the work of both gerontologists and public administrators interested in regulatory policy efforts in the 50 states. It has immediate applications for those concerned with the condition of NH care in the states, particularly individuals and organizations that advocate on behalf of the vulnerable older adults who constitute the majority of the nation’s NH population. This research also has lasting implications for the study of regulation in the United States, as it demonstrates one of the results of the administration of a national policy R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 19 by the states. This dissertation reveals that the CMS’s goal of consistent protection of NH residents by the states is largely unmet. Organization This dissertation is organized into five chapters. Chapter 2, the Literature Review provides a more detailed explanation of the role of the states in administering national regulatory policy and establishes state NH regulation as the focus of this dissertation research. This chapter addresses the relevant literature which identifies the external and internal determinants that predict regulatory outputs and anticipates the variables that are important in shaping state NH regulation. The Methods chapter (Chapter 3) presents the research objectives of this dissertation. This chapter explains how NH regulatory activity in the 50 states was measured. It identifies the independent variables expected to predict state NH policy outputs. Finally, the Methods chapter proposes the use of multiple regression to evaluate the relationships between these variables and the extent and effectiveness of state NH regulation and of HLM to create a single model for each outcome measure. In the Results (Chapter 4), the regulatory activity of state NH surveying agencies during three recent annual inspection periods is reviewed. This chapter describes the extent and effectiveness of state NH enforcement in each year of the study and the change in these outcome measures from year to year. It then identifies factors revealed by multiple regression to impact extent and effectiveness. This R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 20 chapter concludes by proposing a single theoretical model for each outcome measure, using HLM. Chapter 5, the Discussion, focuses on the extent and the effectiveness of state NH regulation and the individual determinants and theoretical models that explain why this regulation varies across the states. This chapter considers how these results contribute to the fields of aging policy and comparative state policy. The discussion concludes by reflecting on how well this dissertation achieved its objectives and where future research should be directed. Chapter Summary This introductory chapter reviewed the state of NH care in the United States. It identified the NH population, the means of payment for NH residents, current problems with NH care and the state surveying agencies that are responsible for ensuring NH quality. The introduction then established that the purpose of this dissertation research is to focus on the state NH surveying agencies responsible for implementing national NH policy. This study is designed to demonstrate the differences that exist among the states in their NH regulatory activities and to identify the key factors that account for this interstate variation. This research is expected to advance the understanding of how factors unique to each state determine its NH regulatory activity, and to enhance the efforts of those who advocate on behalf of the nation’s NH residents. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 21 Chapter 2 Literature Review The literature review comprises three sections. The first section introduces regulatory policy as an important public policy area. It describes the birth of the federal bureaucracy in the United States during the 1930s, and the emergence of regulatory federalism in the 1960s, an era that continues to this day. Two aspects of regulatory federalism are critical to the present study: the administration of a national policy by the states and interstate variation in regulatory policy. The second section of the literature review introduces the subject of this investigation. Nursing home (NH) policy is an example of regulatory federalism and problems in current regulation stem from state administration of a national regulatory policy. Federal NH standards are enforced by state surveying agencies, which vary in the extent and the effectiveness of their regulatory activities. This pattern has resulted in a national NH policy that varies in its implementation by the states. The third section proposes the use of theoretical models to understand interstate variation in NH regulation. Two types of models are used in the comparative state policy literature: 1) external determinants models reflecting a state’s socioeconomic and political environment and 2) internal determinants models related to the key actors involved in a specific policy. A model integrating external and internal determinants is explored in this dissertation as a means of incorporating the full range of measures that influence state NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 22 Regulation and Regulatory Federalism Regulation is a distinct area of public policy in that it involves the efforts of government to control the effects of the marketplace. It is also unique in relegating responsibility not to elected officials, but to bureaucrats. In the United States, this bureaucracy exists at federal, state and local levels. The modem federal bureaucracy emerged in the 1930s, when the national government under President Franklin Roosevelt assumed new responsibilities in regulating the private sector (Gormley, 1986). However, since the 1960s, the national government has shifted much of its administrative duty to the states, a relationship described by Kincaid (2002) as regulatory federalism. Devolution of regulatory policy continues to this day, with growing responsibilities for state bureaucracies. Two consequences of this trend are: 1) state agencies enforce national policy in many regulatory areas and 2) in many regulatory areas, the implementation of national policy varies across the country due to differences in the enforcement activities of state agencies. The Study of Regulation State regulatory policy is an important area of study for three reasons. First, regulation is unique in its use of public authority to constrain the behavior of private actors and protect citizens from the vagaries of the marketplace (Breyer, 1982). In the United States, the expanding federal role in ensuring the health and well-being of Americans has resulted in a substantial bureaucracy at the national, state and local R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 23 levels (Meier, 2000; Elling, 2003). For example, current NH regulation grew out of public concern for the well-being of older persons in those facilities. Second, regulatory policy in all areas is distinct due to the authority yielded to the bureaucrats responsible for its implementation (Gerber & Teske, 2000). However, in regulatory areas described by Gormley (1987) as low in public awareness and in complexity, bureaucrats can become the key decision-makers. Gormley (1983: 9) identified NH policy as an issue area in which citizens and politicians are infrequent participants, relegating the primary responsibility for enforcement to street-level bureaucrats. Third, while the federal bureaucracy is a familiar topic in public policy, state and local bureaucracies are less well understood, despite their growing prominence. Since the 1960s, the federal government has delegated more responsibilities to states. Consequently, the need for regulatory policy research has become most acute at the state level (Coggbum & Schneider, 2003). This research focuses on a state-level bureaucracy that is particularly important, because in the area of NH regulation, the federal government has ceded the day-to-day enforcement of national policy to the states. The Rise of the Federal Bureaucracy in the United States The modem bureaucracy is a creation of the 20th Century. Max Weber believed a rational-legal authority with a body of generalized rules, independent of the traditional privileged group, was necessary to protect the political and economic R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 sectors of social life within a jurisdiction (Parsons, 1947). The rise of the so-called welfare state brought challenges that transcended existing government structures. Weber proposed that an authority constituted of politically neutral career civil servants was the most expedient way of conducting the state’s business while protecting individual rights (Parsons, 1947). As governments became more complex, the public sectors created by many nations began to bear an uncanny resemblance to Weber’s model (Scholz, Twombly, & Headrick, 1991). In the United States, the bureaucracy as envisioned by Weber began to take shape during the 1930s. The administration of Franklin Delano Roosevelt (1933- 1945) created numerous federal social programs designed both to stimulate the economy and protect the well-being of Americans during the Great Depression. The “New Deal” marked the rise of the power of the federal government and heralded a new era in American intergovernmental relations. The federal government assumed responsibility in fields that were formerly left to state and local governments and the private sector (Beer, 1977). For example, under the Social Security Act (1935), the federal government provided income security for retired workers and, by 1939, for their spouses and children. With every New Deal program, the size of the federal bureaucracy and its influence on everyday American life grew. The federal bureaucracy matured after World War II. In the United States, “government by administrators” began to be superseded by “government by professionals”, a core of officials with scientific and professional training (Beer, 1977). The “professional state” reached its apex during the 1960s in terms of the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 5 scope o f its functions, with the passage of programs such as Medicare and Medicaid. President Lyndon Johnson’s “Great Society” also signaled a rise in the responsibilities of state and local government and a shift from a cooperative relationship between the federal government and state and local governments to a more coercive one. Kincaid (2002) describes this change as the beginning of the era of regulatory federalism. The Era of Regulatory Federalism Prior to the Johnson Administration, the working arrangement between the federal government and the states was defined by a mutual agreement for each to be independent and preeminent within certain spheres. Federal and state governments were considered coequal and autonomous (Kincaid, 2002). However, during the Great Society era, the boundary between federal and state interests became blurred. Federal officials began to see state and local governments as tools by which to implement national policy. The increased federalization of the states was due to several factors, including national trends such as reapportionment, the spread of television, and the proliferation of primary elections that broke down regional differences and much of the traditional power of state governments (Kincaid, 2002). Another factor favoring federalization was the improvement of state government. By the 1960s, state governments had begun to reinvent themselves, becoming more professional and more efficient (Gerber & Teske, 2000). Governors assumed more institutional R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 6 powers and state legislatures became more professional (Gormley, 1987). Further, states began to make substantial improvements to their bureaucracies through merit- based hiring and increased pay (Nice, 1987). Additionally, uncertainty at the federal level often gave rise to the suspicion that the states might do a better job if left to their own devices. By the 1960s, there was increased justification for the federal government to entrust the states with carrying out national policy (Gormley, 1987). However, the most important development during the Great Society era favoring federalization of the states was the growth of large federal grant-in-aid programs. New programs such as Head Start and Medicaid transformed state (as well as local and private) agencies into implementers of federal programs (Beer, 1977; Milward & Francisco, 1983; Gerber & Teske, 2000). Despite this “devolution revolution” (Hanson, 2003), Washington has remained the locus of power in this intergovernmental relationship (Kincaid, 2002). Typically, the federal government sets minimum national standards for state agencies to follow and assumes jurisdiction when the states fail to meet these standards, a type of regulatory relationship described as partial preemption (Hanson, 2003). Since the late 1960s, the federal-state relationship has become less coequal, and more vertical, with the national government at the top of this hierarchy (Milward & Francisco, 1983). The modem era has witnessed partial preemption in many areas previously assigned to state government (Thompson & Scicchitano, 1987). Regulatory federalism has resulted in two important developments. First, national policy in areas such as environmental protection and occupational safety is largely R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 27 achieved through the administration by the states. Second, variations in the implementation of national policy have emerged, due to differences among the states. State bureaucrats increasingly determine whether a national regulatory policy achieves its objectives. State Administration Today, the federal government contracts with state agencies to enforce federal regulations in such areas as environmental protection, occupational safety and Medicaid reimbursement (Milward & Francisco, 1983). In each case, a federal agency sets broad parameters for state agencies to follow, giving the states day-to- day authority over program implementation (Gerber & Teske, 2000). Congress established the Environmental Protection Agency (EPA) over 30 years ago to centralize authority over environmental protection. Since then, the EPA has delegated considerable responsibility to the states in implementing legislation such as the Clean Air Act and the Clean Water Act. The states have become active partners in environmental protection: running federal programs, increasing their own funding, and developing innovative approaches to solving problems (Brown, 2002). Occupational safety is a second example of administration of national regulatory policy by the states. The Occupational Safety and Health Administration (OSHA) is responsible for budgeting and rule-making. However, it is at the state and local levels where most regulatory activity takes place (Scholz, et al., 1991). Each state runs its own workers’ compensation program to insure against workplace R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 8 injuries, and these programs impose far more restrictions on local businesses than does OSHA (Eshbaugh-Soha & Meier, 2003). Finally, while Medicaid is jointly financed by the federal government and the states, it is administered by the states. States determine eligibility and benefit guidelines and are given discretion to broaden their coverage for their vulnerable populations beyond the minimum federal requirements. For example, Medicaid funding can be used to support state clinics and hospitals that serve low-income and uninsured populations (Rowland, 2002). Interstate Variation Interstate variations are witnessed across federal regulatory policy. Kincaid (2002) suggests these differences have little relevance for regulatory outputs because the states have almost no authority over decision-making. However, the dominant view in the regulatory policy literature is that state agencies retain their unique identities and respond in different ways to federal mandates (Thompson & Scicchitano, 1987; Gerber & Teske, 2000). With everyday enforcement left to state agencies, a wide range in performance is possible. The state agencies can be either effective implementing agents of national regulatory policy or barriers to its implementation (Coggbum & Schneider, 2003; Elling, 2003). Differences exist in state implementation of national environmental policy. The sources of funding for environmental protection vary; some states rely solely on user fees, while others use money from the state general fund. Environmental R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 9 statutes also differ; for example, vehicle emissions standards are much stiffer in California than elsewhere. Finally, the extent of enforcement varies, with several states (such as New York) issuing disproportionately more actions than others (Brown, 2002). Occupational safety is another federal-state program in which the vigor of regulatory activity varies from state to state. This is a product both of the wide discretion granted state and local officials, and of the problem of interpreting national regulations under a diversity of local conditions, leading to considerable variation in the way in which different inspectors and agencies enforce the same law. More active enforcement is identified in states such as New York in which organized labor is influential (Scholz, et al., 1991). Finally, states vary in Medicaid eligibility guidelines for their children, adult and elderly populations, as well as in the services Medicaid is allowed to finance (Schneider, 1993). States also differ in daily Medicaid rates for NHs (Swan, Bhagavatula, Algotar, Seirawan, Clemena, & Harrington, 2001) and in their use of waivers that enable Medicaid funds to be directed away from NHs and towards long term care (LTC) alternatives such as assisted living facilities and home and community based services. Finally, the amount of state Medicaid funding allocated to NH regulation varies, with some states exceeding the minimum level mandated by the federal government (Walshe & Harrington, 2002). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 30 Regulatory Federalism Today State implementation of national policy remains an important topic in public policy. In fact, following the terrorist attacks of September 11, 2001, the demands being made of state and local governments have increased. Under the Homeland Security Act, states are expected to shoulder a greater share of the financial burden in keeping the nation’s boundaries secure. However, while the federal government has increased its demands, the resources allocated to the states to accomplish these tasks have decreased, forcing elected officials to choose between cutting programs and raising taxes (Kincaid, 2002). Across many regulatory areas, state agencies have more to do and must make do with less. Further, as the previously described examples illustrate, the states have assumed broader policy discretion in recent years. A consequence of the devolution of decision-making authority from the national government to the states is that interstate variations in regulatory policy increasingly matter. It is largely up to the states to ensure that national guidelines are enforced. For this reason, consistency in state regulation has become a priority. Unfortunately, interstate differences in enforcement of regulatory policy indicate that Americans receive unequal protection due to their state of residence. Today, the success of a national policy is often determined by the state in which it is implemented. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 31 State Nursing Home Regulation Building on studies of intergovernmental policies such as environmental protection, occupational safety and Medicaid reimbursement, this dissertation explores another area of intergovernmental regulation, that of NHs. A federal agency, the Center for Medicare and Medicaid Services (CMS1 ) is responsible for national NH policy, but it is the state surveying agencies that actually carry out CMS mandates. Thus the states are responsible for certifying that NHs meet national standards in order to receive Medicare and Medicaid payment. The Origins of NH Regulation Initially, NH regulation was a task left entirely to the states. The modem NH is a 20th Century creation with its roots in the almshouse and the hospital. Prior to the 1930s, poor and frail elders not living with their families were primarily left to local governments and private charity for support, and housed in public almshouses (Vladeck, 1980). The passage of the Social Security Act in 1935 provided income security for older Americans, but a need remained for frail elders who required supportive housing. This need was initially met by board-and-care homes established by private individuals in search of extra income during the Great Depression. The modem NH industry grew from these “mom-and-pop operations” (Kelly & Liebig, 2003). 1 The Health Care Financing Administration (HCFA) was renamed the Center for Medicare and Medicaid Services in January 2001. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 32 During the 1940s and 1950s, as the population of frail elders grew, the small board-and-care homes were gradually replaced by larger institutions. With the growing size of NHs came greater scrutiny. The 1946 Hill-Burton Act established national building standards for NHs, as well as other health facilities. These standards led new NHs to be built on the hospital model, a design that persists to the present day. The regulation of health and safety standards also began in earnest during the 1950s. However, unlike NH building standards, there were no nationally prescribed benchmarks for NH health and safety at that time. State health inspectors were responsible for ensuring NH quality. The federal government had no role in NH regulation. Medicare, Medicaid and the Growth of Nursing Home Regulation The number of NHs tripled between 1960 and 1980 (Vladeck, 1980), largely due to the passage of Medicare and Medicaid in 1965. Medicare provided health insurance for the aged, but its reimbursement was limited to acute care. While Medicare covered short-term rehabilitation in NHs, it made no provision for LTC. However, because Medicaid reimbursement was allowed for the “medically needy”, this program became the de facto payer of LTC, and the NH emerged as the primary location of LTC (Kane, 1995). The Medicare and Medicaid legislation of 1965 established federal guidelines for all health providers receiving payment under these programs, including NHs (CMS, 2003b). Prior to 1971, enforcing these guidelines in NHs was left to the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 33 states, which were given complete autonomy in conducting inspections (Vladeck, 1980). However, alarmed by the conditions he found during a 1971 visit to a southern Illinois NH, President Nixon determined that state enforcement alone was insufficient in protecting residents from abuse and neglect. Nixon’s Eight-Point Initiative in 1971 gave the federal government a role in NH regulation (Netting, Huber, Paton, & Kautz, 1995). Among its mandates, Nixon’s Initiative required federal oversight over state NH inspections. In 1972, inspectors from the Department of Health, Education and Welfare (HEW) began resurveying roughly 10 percent of the NHs visited by each state survey agency (Vladeck, 1980). Following the 1977 reorganization of the federal government, the newly-created Health Care Financing Administration (HCFA), a division within the Department of Health and Human Services (DHHS), assumed this task (Vladeck, 1980). Omnibus Budget Reconciliation Act of 1987 Persistent reports during the 1970s and 1980s chronicled quality-of-care NH deficiencies. The Institute of Medicine (IOM) reported widespread problems, including untrained staff, inadequate health care, and unsanitary and unsafe conditions and recommended the strengthening of national standards (USGAO, 1999a). Congress achieved this through the Nursing Home Reform Act, part of the Omnibus Budget Reconciliation Act of 1987 (OBRA 87), gradually implemented between 1987 and 1995 (Walshe, 2001). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 4 OBRA 87 radically changed the existing NH inspection process. The legislation required comprehensive assessments of all residents at the time of admission and periodically thereafter. It mandated the use by NH inspectors of quality indicators related to the outcome of NH care, instead of the process of care. OBRA 87 emphasized a patient-oriented survey process and required inspectors to interview residents, rather than merely review medical records (Walshe & Harrington, 2002). Additionally, OBRA 87 specified the roles of state and federal inspectors. Under the terms of their contracts with HCFA, state NH survey agencies were required to ensure that NHs provide quality care to their residents by conducting annual inspections of all federally-certified NHs, investigating all NH complaints, and reporting their findings to HCFA (USGAO, 1999a; Walshe, 2001). When problems were found, the states were required to take enforcement actions to certify that the NHs were compliant with federal standards and thus eligible to receive federal payments under Medicare and Medicaid (USGAO, 1999a). OBRA 87 required federal oversight of this process, and this oversight was achieved through the resurveying of five percent of the NHs in each state by the CMS regional offices (USGAO, 1999a). However, this oversight task was designed to evaluate the performance of the state survey agencies, not of the NHs themselves. Thus, the responsibility for ensuring that the nation’s 17,000 NHs provide quality care for their 1.6 million residents fell directly to the state surveying agencies (Walshe & Harrington, 2002). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 35 Recent CMS Mandates In 1998, President Clinton issued a new set of initiatives designed to strengthen the rigor with which states conduct their annual NH inspections. First, states were instructed to avoid, if possible, scheduling a NH survey during the same month in consecutive years (USGAO, 2000). In some states, the annual inspection of a particular NH occurred during the same month, year after year. This enabled NHs to prepare for their “surprise” inspections and hide potential problems from state surveyors. For example, NHs could overstaff on the days the state inspection was anticipated, creating a false impression of the NH’s daily operation. By scheduling a NH’s annual inspection in a different month each year, a state prevents a facility from being able to predict the arrival of state inspectors; this allows the inspectors to detect deficiencies that the NH might otherwise have concealed (USGAO, 2000). Second, CMS offered new guidance to state survey agencies on resolving NH deficiencies. Previously, CMS allowed a “grace period” for NHs to correct their problems before disciplinary action was taken, even for serious repeated deficiencies. Also, CMS allowed states to accept the NH’s assertion that it had returned to compliance, rather than confirm the resolution of serious deficiencies through an on-site “revisit” (USGAO, 2000). However, the 1998 initiatives put more teeth into enforcement. CMS now instructs states to deny a grace period for any NH cited with deficiencies causing actual harm or the potential of death or serious injury to residents in two consecutive inspections or complaint investigations. State R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 6 surveying agencies are required to perform revisits to ensure that these NHs cited had resolved their serious deficiencies (USGAO, 2003). Current State NH Regulation Under OBRA 87, CMS contracts with state surveying agencies to conduct annual inspections of every Medicare and Medicaid-certified NH, and to conduct all investigations when quality-of-care complaints are made (USGAO, 1999a-c). State NH surveyors are required to follow federal regulations for inspections specified by CMS in its State Operations Manual (Walshe & Harrington, 2002). During an annual NH inspection or complaint investigation, state surveyors evaluate NHs based on 175 CMS-selected measures encompassing structural, procedural and outcome measures of quality (Grabowski, 2001). The surveying agency issues a deficiency citation for each incidence in which the NH fails to comply with these federal requirements. Since 1995, these deficiency citations have been rated according to the scope (whether deficiencies are isolated, constitute a pattern, or are widespread) and severity (minor impact, minimal impact, negative impact, and immediate jeopardy to resident heath and safety) using a grid devised by CMS (see Table 1). State survey agencies are responsible for taking enforcement actions for every NH deficiency found during an annual inspection or complaint investigation. State inspectors give NHs a “grace period” of 30 to 60 days to correct non life- threatening deficiencies (USGAO, 2000). At the end of the grace period, an R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 7 uncorrected deficiency results in penalties ranging from monetary fines to Medicare and Medicaid decertification, which is the ultimate sanction. The most serious NH deficiencies, those that can cause death or serious injury to residents, require the state to close the NH immediately. However, few NHs are terminated; most of these are reinstated and continue to have problems (Walshe & Harrington, 2002). The federal government is required to pay for 100 percent of Medicare survey and certification expenses, and for 75 percent of Medicaid survey and certification costs. The states are required to pay the remaining 25 percent of Medicaid costs for NH inspection, but can elect to exceed this amount. Total survey and certification costs in 2000 were $135.9 million under Medicare and $147.8 million under Medicaid (Walshe & Harrington, 2002). Problems with Current State NH Regulation State Regulatory Activity and NH Quality The “success” of state NH regulation is a controversial topic. Some studies suggest that the NH quality has improved in response to OBRA 87 and the efforts of state inspectors, citing fewer NH deficiencies for the use of physical and chemical restraints (Harrington & Carrillo, 1999), urinary incontinence and catheterization (Walshe, 2001), and feeding tube use (Grabowski, 2001). The IOM (2001) described decreased regulatory activity as an indication that “regulation has brought some improvements in the quality of NH care” (Walshe & Harrington, 2002: 475). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 8 In contrast, others believe that overall NH quality remains poor. The USGAO (1998,1999a, 2000) reported that one-fourth of the nation’s NHs, and nearly one-third of California NHs, were cited for deficiencies that cause actual harm or the potential for death and serious injury. The USGAO also concluded that state and federal inspectors had not been effective in ensuring that deficiencies are corrected and stay corrected. The USGAO (2000, 2003) has argued that the number of serious citations reflects not only poor NH quality, but also an ineffective regulatory system in which these problems are allowed to persist. However, conclusions made on the basis of NH citations are problematic. There are multiple interpretations of the frequency of regulatory activity. A high number of citations may reflect poor NH care; however, it may also indicate a highly effective regulatory system in which the state survey agency is doing its job: identifying poorly performing NHs. Conversely, a low number of NH deficiencies may mean good NH care, but it also may reflect an ineffective regulatory system, in which state inspectors overlook quality-of-care problems, enabling poor performers to go unpunished. For this reason, the quality of NH care and the effectiveness of NH regulation are increasingly recognized as separate issues. Recent studies have become reluctant to interpret trends in deficiency citations as indicators of state survey agency effectiveness. Walshe and Harrington (2002) attributed the 44 percent decline in the total number of NH deficiencies during the 1990s to deteriorating state inspections rather than improving NHs. Wiener doubted that the wide interstate range in R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 9 regulatory activity was realistic; “on its face, it is difficult to believe that quality of care actually varies that much” (2003: 20). The NH regulatory process has flaws. However, using the outputs of this process, the state deficiency citations, without consideration of the larger context in which each state survey agency operates, may lead to false conclusions. In order to evaluate the quality of state NH regulation, the performance of the state surveying agency must be disentangled from that of the NHs the agency inspects (Walshe, 2001). This dissertation proposes that two standards should be used to evaluate the quality of state NH regulation. These are: 1) the extent of the states in their issuance of deficiency citations and sanctions, and 2) the effectiveness of the state surveying agencies in implementing CMS mandates. Extent o f State NH Regulation A coherent national NH policy depends on consistency among state surveying agencies. However, interstate differences persist in the volume of the deficiency citations they issue (Wood, 2002), undermining the reliability of national inspection data. In 1998, there were 14.2 deficiencies per NH in Nevada, but only 1.9 in New Jersey, an 8:1 ratio that Wiener (2003) described as unrealistic. States also vary in the severity of the citations they issue. In 1999, actual harm deficiencies were found in 10.5 percent of NHs in Maine, but in 58 percent of NHs in Washington, a gap that belies the comparable quality of NH care in these two states R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 0 (USGAO, 2000). These findings indicate that the states apply different standards when citing deficiencies (Walshe & Harrington, 2002). States also vary in their use of intermediate sanctions, such as temporary management, denial of new Medicare and Medicaid admissions or the imposition of civil monetary penalties. This irregularity means that many violations go uncorrected (Wood, 2002). Although states infrequently terminate a NH from the Medicare and Medicaid programs, wide interstate variation exists here as well. During a recent eight-year period, rates of involuntary termination ranged from zero in thirteen states to 12.5 percent ofNHs in Nevada (Angelelli, et al., 2003). Some of these differences in citations and sanctions can be attributed to social, economic, and political differences among the states (Mullan & Harrington, 2001). However, Walshe and Harrington (2002) believe that interstate variation in citations and sanctions mainly derive from significant differences among the state survey agencies in their behavior, performance and attitudes towards regulation. Despite the efforts of CMS to maintain consistent national standards for the detection and correction of NH deficiencies (USGAO, 2000), variations among state surveying agencies in their performance of these standards suggest that uniformity in state enforcement has not been achieved. The nation’s NH residents receive unequal protection as a result. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 41 Effectiveness o f State Nursing Home Regulation The 1998 CMS Initiatives require states to: 1) avoid repetitive scheduling of annual inspections (i.e., scheduling a particular NH’s annual inspection during the same month in consecutive years) and 2) resolve all actual harm deficiencies within 60 days and all non-actual harm deficiencies within 90 days. However, the states fall short of fulfilling these requirements. In a six-state study, the USGAO found that the scheduling of annual inspections remained repetitive and that chronic NH problems were unresolved (USGAO, 2000). First, the USGAO reported that the timing of annual NH surveys continued to be predictable, allowing facilities undue preparation time for “surprise” state inspections. The USGAO (2000) found that between 29 and 56 percent of NH surveys in the 1999-2000 annual inspection period occurred during the same month as in the previous year. For example, over half of the surveys in Tennessee were conducted within 15 days of the date of the previous annual inspection. This predictability may cause the states to miss serious problems. One indication of this is the number of new deficiencies found in federal resurveys of NHs previously inspected by the states (Mullan & Harrington, 2001). Second, serious deficiencies cited by state surveyors have not been resolved in accordance with CMS mandates; quality-of-care problems recurred in the required follow-up inspections (Angelelli, et al., 2003). “Repeat offenders” have been allowed to remain open, in spite of federal requirements that NHs in persistent noncompliance be closed. In September 2000, CMS reported that a little more than R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 2 half of the historically poor-performing NHs targeted for semiannual surveys had actually received this enhanced monitoring (USGAO, 2000). The IOM 2001 report recommended a greater focus on dealing with persistently poorly performing NHs (Walshe & Harrington, 2002). Effective regulation achieves its statutory objectives (Sabatier & Mazmanian, 1983). However, several state NH surveying agencies have failed to carry out the objectives of the 1998 CMS initiatives regarding annual inspections and complaint investigations (USGAO, 2000, 2003). Regulatory federalism, in terms of the implementation of national NH policy by the states, is not working. “At present, NH regulation exhibits few, if any, of the features of responsive regulation” (Walshe, 2001: 135). Summary of State NH Regulation NH regulation is a responsibility shared by the federal government and the states (USGAO, 2000; CMS, 2002), an example of regulatory federalism (O’Toole, 2000; Kincaid, 2002). The federal government, through CMS, is responsible for setting national NH policy. However, the states have the responsibility for the day- to-day implementation of this policy (USGAO, 2000; CMS, 2002). This gives the states a central role in NH policy and an opportunity to become “laboratories of innovation” (Nathan, 1990). According to reports made by the USGAO and others, state NH regulation has instead revealed two weaknesses. First, the states vary in the extent to which R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 3 they enforce national NH policy. Second, many states are ineffective in meeting some of the most important federal standards prescribed by CMS (Harrington, et al., 1999; USGAO, 2000,2003; Walshe, 2001). This dissertation research is intended to help explain the factors that shape the extent and effectiveness of this state regulatory policy. Understanding State Regulatory Policy: Theoretical Models Interstate variations in regulatory policy have received more attention in recent years due to the increased devolution of responsibilities to the states (Gerber & Teske, 2000). However, interstate differences have been a critical issue throughout American history. Sectionalism was legitimated by the U. S. Constitution (Shelley, 1988), which decentralized authority over broad areas of domestic policy. The states have followed distinct patterns in their politics and policies ever since. Sectional variations in the practice of slavery led to the Civil War, and the states continued to pursue civil rights at their own pace for the next century. In short, interstate variation in the implementation of national policy continues the historical trend of the states marching to the beat of their own drums. The same forces that have accounted for such wide interstate variations in other areas of politics and policy are those that shape state regulatory policy. Current study of comparative state policy traces its roots to the work of V.O. Key, whose seminal work Southern Politics in State and Nation (1949) examined the distinctive political environment of the southern region of the United States. Key concluded R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. that southern politics was burdened by the South’s “special problems”: its poverty, its agrarian economy, and especially its troubled racial heritage. More recent authors have also recognized that a state’s policies are shaped by forces unique to that state, studying the effects of a state’s socioeconomic and political characteristics on its policy outputs (Jennings, 1979). Two types of models dominate the current discussion of interstate variations in regulatory policy in the comparative state policy literature. External Determinants models propose that a state’s socioeconomic and/or political environment shapes its policy outputs (Blomquist, 1999). Internal Determinants models maintain that the key actors concerned with a particular issue (such as the regulatory agency, state legislature, interest groups, and the governor) are responsible for generating policy (Sabatier & Jenkins-Smith, 1999). External Determinants External determinants models share the assumption that a state’s specific socioeconomic and/or political circumstances shape its ultimate policy outcomes. These models typically include measures such as economic development, demographic composition, and political culture. An additional political variable is federal support, particularly in areas of intergovernmental regulation, in which a national policy is implemented by the states. In many areas of intergovernmental regulation, the U.S. Congress appropriates money for the regulatory activities of state bureaucrats. The federal funding of state NH surveying activities through the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 5 Medicare and Medicaid programs is an example of this support (USGAO, 1999a; Walshe & Harrington, 2002). Economic Development Dye (1966) believed a state’s economic development shapes its political institutions and policy outputs (see Figure 1). The more developed a state’s economy, the greater are both the needs and the means for policy implementation. Dye (1966) described economic development as encompassing four components: urbanization, industrialization, education and wealth. Urbanization increases the social problems that require government action. Industrialization also necessitates a public response, to protect both workers and consumers from private sector abuses. Education expands economic opportunity, but also demands public oversight of its standards. Finally, the wealth of its citizens provides a state government with the necessary tax revenue to perform these functions (Dye, 1966). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 6 Figure 1: Dye’s Policy Formation Model Y Socioeconomic Variables Source: Dye (1966) In most areas of regulation, all four measures are identified as predictors of state regulatory activity (Gerber & Teske, 2000). For example, Sigelman and Smith (1980) reported that urbanized states with better-educated and wealthier residents enacted tougher consumer protection laws. Thompson and Scicchitano (1987) found that industrialized, high-income states were more successful in implementing federal occupational safety standards. Industrialization and income also predicted greater state activity in air pollution control (Ringquist, 1993) and groundwater regulation (Blomquist, 1991). However, in studies of state health care regulation, the picture is somewhat different. Three of Dye’s components of economic development (urbanization, education and wealth) are frequently identified as determinants of state regulatory activity in health care; one (industrialization) is not. Lynk (1981) considered Political System Characteristics Policy Outcomes R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 47 urbanized, high-income states as those most likely to impose control on physician costs. Begun, Crowe, and Feldman (1981) reported stronger regulation of optometrists in urbanized states. Graddy and Nichol (1990) studied the educational level of a state’s residents as a possible predictor of the state’s performance in licensing nurses, reasoning that educated consumers are more likely to be knowledgeable of the health care system. Schneider (1993) found that states with higher per capita income spend more money on the administration of their Medicaid programs. Similarly, of Dye’s four components of economic development, three have been predicted to favor tougher state NH regulation: urbanization, education and wealth (Harrington, Swan, Nyman, & Carrillo, 1997; Grabowski 2001). The absence of industrialization in these studies suggests that while it is an important factor in regulatory policies that affect a state’s manufacturing interests (such as occupational safety and environmental protection); industrialization appears to have little impact on health care. Demographic Composition Models used in comparative state policy analysis often include a number of demographic characteristics such as race/ethnicity, gender and age (Van Meter & Van Horn, 1975; Sigelman, 1976). The purpose of including these measures extends beyond their statistical utility as control variables. The density of certain R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 8 demographic interests is thought to have real influence on specific areas of state policy. Jennings (1979) concluded that the greater a state’s proportion of low-income persons, the more generous was its welfare policy. As the proportion o f vulnerable subpopulations in a state increases, state policies towards these groups become more favorable (Wright, Erikson, & Mclver, 1987). Lammers and Klingman (1984) and Browne (1987) detected a greater presence of aging issues on the legislative agendas of states with older populations. Political Culture A prominent school of thought posits that a state’s political environment shapes its policy outputs. Daniel Elazar’s discussion of political culture is particularly influential. Elazar (1972) described three political subcultures in the United States, each with a distinct philosophy on the role of government. Some states are dominated by a moralistic subculture (M) driven to create the “good society”, others by an individualistic subculture (I) dominated by economic interests. Finally, a third group of states is identified as dominated by a traditionalistic subculture (T) determined to preserve the status quo (see Figure 2). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 9 Figure 2: Dominant Political Subcultures in the 50 States Sources: Sharkansky (1969); Elazar (1984) Sharkansky (1969) modified Elazar’s model by creating a linear scale for political culture, with purely moralistic states (1) and traditionalistic states (9) at the extremes, and purely individualistic states in the middle (see Figure 3). Several comparative state policy studies have included political culture as a variable by using this scale (Ritt, 1974; Savage, 1981; Boeckelman, 1991). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 0 Figure 3: Sharkansky’s Adaptation of Elazar’s Political Culture Model M MT MI IM I IT TI TM T 1 2 3 4 5 6 7 8 9 Sharkansky designates the following scale values for prevailing political culture(s) in a state. 1: Moralistic culture 2: Dominant moralistic culture; traditionalistic subculture 3: Dominant moralistic culture; individualistic subculture 4: Dominant individualistic culture; moralistic subculture 5: Individualistic culture 6: Dominant individualistic culture; traditionahstic subculture 7: Dominant traditionalistic culture; individuabstic subculture 8: Dominant traditionalistic culture; moralistic subculture 9: Traditionalistic culture Source: Sharkansky (1969) Several researchers have studied the impact of political culture on a state’s politics and policies. Erikson, Wright, and Mclver (1987) and Paddock (1997) found strong connections between political culture and partisanship; Lammers and Klingman (1984) concluded that moralistic states were more favorably disposed towards aging policy. Walker (1969) and Kirst, et al. (1984) discovered differences between the three subcultures in policy innovation; states with a dominant moralistic subculture emerged as regional leaders. Several authors have also concluded that political culture of a state impacts the activities of its regulatory agencies (Shaffitz, 1974; Miller, 1991; Morgan & Watson, 1991); moralistic states had the most frequent issuance of citations and sanctions across several regulatory areas (Jacoby & Schneider, 2001). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 51 Federal Financial Support Another external determinant of state policy is the funding provided by the federal government. Federal financial support to state agencies expanded during the 1960s, largely due to new programs such as Medicaid; state agencies rely heavily on appropriations from the U.S. Congress (Gormley, 1987; Wulf, 2002). This funding usually comes with strings attached; state agencies typically find federal aid accompanied by detailed rules regarding its use (Brudney & Hebert, 1987). State agencies that receive federal aid may also bear consequences for poor performance. The federal government has a range of financial sanctions, from fines to termination of federal aid, at its disposal to use against states that fail to comply with national standards (Van Meter & Van Horn, 1975; Meier, 2000). However, federal aid is both a boon and a bane for state agencies. Federal funding can be used as an incentive to reward good performance. For example, CMS authorizes increased funding for states that meet national goals for timely completion of annual inspections (CMS, 2003b). Further, higher levels of federal funding decrease the dependence of state agencies on state funding, which may also make the state agency less susceptible to influence from the legislature, governor and interest groups (Brudney & Hebert, 1987). One illustration is state NH regulation; the federal government provides over 75 percent of the funding for the state survey and certification process, through the Medicare and Medicaid programs (Walshe & Harrington, 2002; CMS, 2003b). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 2 Summary of External Determinants Models The external environment of a state plays an important role in the formation and implementation of its policies. Economic development, demographic characteristics, political culture and federal support all act as catalysts and constraints on states; the unique socioeconomic and political milieu of a state provides a useful context for its regulatory policies. However, this approach has its critics. Mazmanian and Sabatier (1981) concluded that external determinants models explain a small proportion of interstate differences in policy outputs. Further, Blomquist (1999) found that the emphasis on the state environment ignores the role of human agency. In external determinants models, the players in the policy process do not act in creating state policy. Instead, the players are acted upon, by forces beyond their control. While external determinants models convey the context in which regulatory policy is made, they fail to explain the process itself (Blomquist, 1999). An alternative explanation may be found in internal determinants models. Internal Determinants Internal determinants models propose that state policy is the product of public and private actors (Sabatier & Jenkins-Smith, 1999: 119). Early conceptions of the policy process described a closed “iron triangle” of bureaucratic agencies, legislative committees, and interest groups (Kingdon, 1995). However, later observers agreed that the open, expanded network suggested by Heclo’s issue system model (1978) more closely reflects the resource exchanges among the hundreds of actors that are R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 3 involved in the policy process at any given time (Milward & Francisco, 1983; Kirst, Meister, & Rowley, 1984; Brudney & Hebert, 1987). Paul Sabatier and Hank Jenkins-Smith (1993) have clarified this concept. They propose that for each specific area of regulation, a unique Advocacy Coalition Framework (ACF) exists, composed of a finite number of public and private actors (see Figure 4). These actors share a general set of beliefs about an issue, and use a variety of legal and political instruments to translate their beliefs into public policy over a period of time, usually a decade or more. Sabatier and Jenkins-Smith (1999) studied actors that are regular participants in an ACF, such as the regulatory agency, state legislature and interest groups, as well as irregular actors, such as the governor. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 4 Figure 4: The Advocacy Coalition Framework (ACF) POLICY SUBSYSTEM Coalition A Policy Brokers a.Beliefs b.Resources Coalition B , a. Beliefs b.Resources Coalition A Strategy Coalition B Strategy Policy Decisions Institutional Rules, Resource Allocations and Appointments Policy Outputs Policy Impacts Constraints Subsystem Resources Actors EXTERNAL (SYSTEM) EVENTS 1 . Changes in socio-economic condition 2. Changes in governing coalition 3. Policy decisions STABLE PARAMETERS 1. Basic attributes of the problem area 2. Basic distribution of natural resources 3. Values and social structure 4. Constitutional structure (rules) Source: Sabatier & Jenkins-Smith (1999) Agency State agencies have missions that make them the cornerstone of an ACF devoted to a particular policy (Sabatier & Jenkins-Smith, 1993), such as NH regulation. However, state regulatory agencies rely on the political support of other ACF members in order to survive and grow (Brudney & Hebert, 1987). State R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 5 agencies have become increasingly important during an era in which the national government has increasingly devolved regulatory duties to the states (Holzer, 2002; Elling, 2003). Today, state agencies enforce national policy in areas such as emergency management (Hembree, 2002), public welfare (Van Lare, 2002), environmental protection (Brown, 2002), higher education governance (Coble & Watts, 2002) and NH regulation (USGAO, 1999a). The financial and human resources available to the state agency are critical factors in determining the success of its implementation of national policy. A state agency with more abundant financial resources, relative to its regulatory task, is more likely to be effective in meeting federal standards than is a state agency with inadequate funding (Mazmanian & Sabatier, 1989; Lester & Bowman, 1989; Ringquist, 1993). The number of surveyors employed by a state agency is also considered a reliable predictor of effective enforcement of federal regulations. State agencies with greater human resources have a higher degree of specialization and expertise (Brudney & Hebert, 1987) and are more successful in rapidly implementing federal mandates (Lester & Bowman, 1989; Ringquist, 1993). Legislature The state legislature is an important component of any regulatory policy ACF. Legislative awareness of a particular issue is important in determining the success of the policy (Sigelman & Smith, 1980). Gormley (1987) argued that the U.S. Congress is the true principal of state agencies, since state activities are usually R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 56 dictated by federal grants and guidelines. However, much of the discretion over state agency funding has shifted to state legislatures (Brudney & Hebert, 1987; Rosenthal & Jones, 2002; Hamm & Moncrief, 2003). For example, legislatures can choose to appropriate funds for NH inspections in addition to the state’s required Medicaid match (Walshe & Harrington, 2002). State legislatures vary in terms of political control. The control of a state legislature by either of the two major parties is likely to affect its regulatory policy decisions (Teske, 1990; Hwang & Gray, 1991; Radcliff & Saiz, 1998). In areas such as telecommunications regulation (Teske, 1990), environmental protection (Wood, 1991) and occupational safety (Scholz, et al., 1991), Republican legislators have favored the interests of businesses by seeking to reduce regulatory activity, while Democratic legislators have been more supportive of expanding regulation. State NH surveyors have identified the legislatures as major constraints on the scope of their activities (Walshe & Harrington, 2002); partisan composition may play a role. Currently, state legislatures are evenly split between Republican and Democratic control (Hamm & Moncrief, 2003). States with a high degree of legislative professionalism are considered more responsive to regulatory policy, because legislators who view their office as a career are more likely to invest themselves in the workings of government (Squire, 1992). State agencies can expect increased vigilance from career legislators (Brudney & Hebert, 1987), which is expected to lead to more effective regulation (Sigelman & Smith, 1980; Gerber & Teske, 2000). Wide disparity exists among the states in three R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 57 measures of legislative professionalism defined by Squire (1992, 2000): legislative salary, length of the legislative session, and size of legislative staff (Hamm & Moncrief, 2003). Finally, the presence o f a standing committee in the state legislature devoted to a particular policy area may indicate that the domain is considered a state priority, one with a sustained place on the legislative agenda (Lammers & Klingman, 1984; Gilligan & Krehbiel, 1987). A standing committee also provides a forum for all parties affected by the policy to be heard. Interest Groups Two types of interest groups are present in regulatory policy ACFs: groups representing the regulated industry and those representing consumers. National interest groups vary greatly across states, and assessing the power of industry and consumer groups in affecting regulatory policy in the states is problematic. Size and financial resources are unreliable predictors o f an interest group’s impact on regulatory activities (Day, 1990; Thomas & Hrebenar, 2003). Members are often attracted to a group for reasons other than its policy positions (Day, 1990); the AARP is an example (Lammers & Klingman, 1984; Liebig, 1992). Further, it is difficult to determine the proportion of a group’s financial resources directed towards lobbying activities (Thomas & Hrebenar, 2003). In determining the power of an interest group to shape regulatory policy, size does not always matter. Researchers instead have preferred qualitative assessments R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 8 of state interest group strength, drawing largely on the relevant literature and on interviews with key policymakers (Morehouse, 1981; Thomas & Hrebenar, 2003). Hrebenar and Thomas (1993,2003) have assessed interest group influence in the 50 states, and have identified the most influential interest group coalitions in each state. Their study is considered the gold standard for comparing the strength of different interests at the state level (Gray & Lowery, 1999). Another measure of the capacity of an interest group to affect regulatory policy is its access to decision makers. In assessing state regulation, it is important to study lobbying activity in the state capital, where decisions are made that affect a wide range of interest groups (Day, 1990; Thomas & Hrebenar, 2003). It is in the state capital where lobbyists primarily organize to support or block state policymaking (Day, 1990; Liebig, 1992; Thomas & Hrebenar, 2003). For this reason, the presence or absence of an interest group office in the state capital is a useful predictor of the ability of the group to affect state regulatory policy. Governor State governors are irregular members of regulatory policy ACFs (Sabatier & Jenkins-Smith, 1999); they generally do not care about regulation (Gerber & Teske, 2000). However, state agencies report that governors have considerable influence over bureaucratic decision-making when they decide to participate in regulatory policy. The governor can be crucial in mobilizing political resources (Brudney & R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 9 Hebert, 1987). One area in which the impact of governors is increasingly felt is NH regulation (Polivka, 2003). One factor believed to predict gubernatorial influence over regulatory policy is the political party o f the governor. Democratic governors are expected to pursue agendas that expand regulatory activities, Republican governors to curtail them (Wood, 1991; Schneider, 1993). State Medicaid spending on regulation (Schneider, 1993) and state environmental agency enforcement (Wood, 1991) were shown to be significantly greater under Democratic governors. Barrilleaux (1999) argues that the institutional powers o f the governor also favor regulatory activity. A governor’s institutional powers over areas such as appointments and the budget are determined by the state constitution, state statures, and the results of voter referenda (Beyle, 2003). Barrilleaux (1999) contends that the competitiveness of state politics forces gubernatorial candidates to promote policies that offer widespread benefits to voters, such as consumer protection. The greater the institutional power of the office, the greater is the ability of a governor to support these policies once in office (Barrilleaux, 1999). As Beyle (2003) demonstrates, the states differ sharply in the institutional powers of their governors. Finally, the support of a governor for a particular regulatory policy may be discerned from the governors ’ state-of-the-state addresses, which provides a public record of the governor’s agenda (Sabato, 1983; DiLeo, 1997). By tracking the topics in these speeches over several years, shifting gubernatorial priorities towards issues such as NH regulation can be identified. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 0 Summary of Internal Determinants Models Internal determinants models identify the key actors in the policymaking process. The ACF simplifies this process by recognizing that, while hundreds of actors are involved in regulatory policy formation and implementation, a finite number play a role in a particular policy domain. State regulatory agencies, legislators and interest groups are regular players in regulatory policy ACFs. Irregular players include state governors. According to the ACF, factors related to these actors either encourage or discourage regulatory activity. The ACF illuminates the “black box” through which policy change occurs, identifying the key players, the resource exchanges between them, and their collective effect on policy outcomes. Additionally, the ACF is clear and simple (Schlager, 1995; Sato, 1999). However, this model is also criticized for failing to explain the reasons for the actions of these key players (Schlager, 1995). Despite its success in addressing the policy process, the ACF provides little understanding of the context in which this process occurs. An Integrated Model External and internal determinants models have strengths and weaknesses. External determinants models describe the connection between a state’s socioeconomic and political environment and its policy outputs, but do not recognize either the policy subsystem through which this environment is processed or the roles of the actors who participate in the policy process. Internal determinants models identify the policy subsystem and its regular and irregular players, but disregard the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 61 effects o f the state’s environment on these actors, and on their outputs. In order to understand the policy process, external and internal environments both need to be considered (Browne, 1987; Van Meter & Van Horn, 1975). The strengths and weaknesses of external and internal determinants models suggest that a model integrating the two may be better in order to incorporate the full range o f measures that determine state regulatory policy. Several precedents exist in the comparative state policy literature. Integrated models combining environmental measures with the effects of actors within a particular policy subsystem have been used in studies of state environmental protection (Gormley, 1986), occupational safety enforcement (Scholz, et al., 1991) and reflect the realistic premise that factors inside and outside state policy subsystems impact policy outputs and account for interstate variations in regulation (Gormley, 1986; Scholz, et al., 1991). Applying the Integrated Model: State NH Regulation The 50 state surveying agencies implement national NH policy, and it is the role of the states in NH regulation that is the focus of this investigation. However, state NH regulation is often ineffective; many state surveying agencies fail to follow federal requirements. In addition, state agencies are inconsistent in the extent of their enforcement activity (Walshe & Harrington, 2002; USGAO, 2003). Two types of factors are crucial in understanding interstate variation in the extent and effectiveness of state NH regulation. The first are factors external to the state NH policy ACF (economic development, aged proportion, political culture and R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 2 federal support). The second are factors related to actors within this policy ACF (state NH surveying agencies, state legislatures, key interest groups and governors). This study proposes a model integrating these factors for each outcome measure of extent and effectiveness. Multiple regression is first used to identify factors that influence state NH regulation. At the conclusion of these analyses, a theoretical model is proposed for the five outcome measures. These five models are then tested using hierarchical linear modeling (HLM). The methodology ofthis process is described in the following chapter. Chapter Summary This literature review first described the growing importance of the states in implementing national regulatory policy and chronicled the evolution of NH regulation as an example of regulatory federalism. Problems in national NH policy are due to problems in its administration by the state surveying agencies, whose enforcement often deviates from national guidelines. This chapter then reviewed the use of theoretical models in the comparative state policy literature to predict state policy formation. Models related to external state characteristics and to specific actors in a state policy subsystem were discussed, and a model integrating these external and internal determinants was recommended. Finally, an integrated model encompassing external and internal determinants was proposed as a means to explain interstate variations in the extent and effectiveness of state NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 63 Chapter 3 Methods This chapter comprises three sections. The first section describes the data collected for this study and the outcome measures that were created from these data. This study examines NH regulation in the 50 states during three annual inspection periods (1999-2000,2000-2001,2001-2002). Two sets of outcomes measures are analyzed in this dissertation: the extent and the effectiveness of state NH regulation. Extent is reflected in the volume and the severity o f deficiency citations state agencies issue to NHs. Effectiveness is measured by the success of state implementation of federal mandates concerning the scheduling of annual state inspections and the timely resolution of NH deficiencies. The second section identifies the factors that were expected to predict the extent and effectiveness of state NH regulation. In this study, characteristics of a state’s socioeconomic and political environment and factors related to the major players in a state NH policy subsystem (the state surveying agency, the state legislature, key interest groups and the governor) were believed to influence state enforcement. The hypothetical relationships between these determinants and the outcome measures in this study are then elaborated. The third section of the methods chapter describes the two types of linear modeling employed in this dissertation. First, multiple regression was used to identify the determinants that predicted the extent and effectiveness of state NH regulation. Five sets of variables were evaluated: external determinants, agency R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 4 determinants, legislative determinants, interest group determinants and gubernatorial determinants. For each of the five outcome measures in this study, a single theoretical model was then proposed, comprising the variables found to be predictors. A single pooled analysis for each model was then completed, using Hierarchical Linear Modeling (HLM). Sample Because this study analyzes all 50 states, the sampling bias that would result from using a smaller group of the states has been avoided (Kish, 1965). For this reason, the external validity is excellent (Kaskie, 1998). The District of Columbia is excluded from this study because two key elements of the NH policy ACF, the state legislature and the governor, are not present in the nation’s capital. Territories such as Puerto Rico and the U.S. Virgin Islands are excluded because too few facilities exist in these places for NH regulation to be evaluated. During the three years of observation, there were seven federally certified NHs in Puerto Rico, and one in the U.S. Virgin Islands (CMS, 2003a). This dissertation examines state NH enforcement in the 50 states during three recent annual inspection periods, identified in this study as Year 1 (June 1, 1999- May 31, 2000), Year 2 (June 1, 2000-May 31, 2001), and Year 3 (June 1, 2001-May 31, 2002). There were over 1.7 million NH beds in the 50 states during the three years of the study (CMS, 2003a). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 65 The enforcement activities of interest are the deficiency citations state surveying agencies issue to NHs that fail to comply with Medicare and Medicaid requirements (USGAO, 2003). State agencies issue a citation for each quality-of- care problem found during a NH inspection; these citations are ranked according to their scope and severity, and reported to CMS (USGAO, 1999c). The data used in this study were obtained from the CMS Nursing Home Compare website in January 2003. This is the same information recorded in the On-Line Survey, Certification, and Reporting System (OSCAR), the database that contains the complete inspection results of all federally certified NHs in the United States (USGAO, 2003). Outcome Measures Extent of State NH Regulatory Activity A recurring problem in intergovernmental regulation is inconsistency among state agencies in their regulatory activities. Thompson and Scicchitano (1987) found that interstate variations in the number of occupational safety citations were due more to inconsistent state agency practices than to differences in the regulated industries from state to state. Similarly, observers have attributed the broad interstate disparity in NH deficiency citations more to divergent practices of state surveyors than to differences in NH quality across the nation (Walshe & Harrington, 2002; Wiener, 2003). In this study, consistency in state NH regulation is assessed by comparing the volume o f deficiency citations (number of citations per 100 NH beds) and the severity o f deficiency citations in the 50 states. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 66 Volume o f Deficiency Citations CMS contracts with state surveying agencies to perform annual inspections of every Medicare and Medicaid-certified NH in the United States. The agencies are responsible for citing every instance in which a NH is deficient in meeting national standards, as described by CMS in the State Operations Manual (SOM). The agencies then report these deficiency citations to CMS by recording this survey information in the OSCAR database (CMS, 2003a). Consistent national NH enforcement is a CMS objective; however, wide variations among states in deficiency citations are believed to be an indication that state surveyors apply different standards (Walshe & Harrington, 2002; USGAO, 2003). The volume o f deficiency citations is computed as the number of state deficiency citations per 100 NH beds. The number of state citations in each annual inspection period was obtained from the NHCompare database, while the number of NH beds in each state during this three-year period was obtained from the U. S. Centers for Disease Control (CDC). The denominator (100 NH beds) is useful for interstate comparisons of citation volume because it reflects the size of a typical NH in the United States (Grabowski, 2001; CDC, 2003). Severity o f Deficiency Citations Since 1995, federal and state NH surveyors have used the Scope and Severity Grid developed by CMS to rate NH deficiencies (see Table 1). This grid places a deficiency in one of twelve categories, labeled “A” through “L”. The most serious R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 7 category (L) is for a widespread deficiency that causes actual harm or the potential for death or serious injury to residents; the least serious category (A) is for an isolated deficiency that results in no actual harm and has the potential only for minimal harm. The severity o f deficiency citations in the United States is also obtained from NHCompare, which contains the scope and severity category of every NH deficiency. In this study, deficiency severity, not scope, is of interest, because it is the former that determines both the issuance and the extent of state sanctions. There are no required sanctions for potential for minimal harm deficiencies (categories A to C). Sanctions are required for potential for more than minimal harm deficiencies (D to F); these include a directed plan of correction, directed in-service training, and/or state monitoring. States may punish NHs with other actual harm deficiencies (G to I) with denial of Medicare and Medicaid payment for new admissions and/or civil monetary penalties of $50 to $3,000 for each day of noncompliance. Finally, NHs with actual or potential for death/serious injury (J to L) deficiencies receive the most severe sanctions, which include temporary management, termination, and/or civil monetary penalties of $3,050 to $10,000 per day (USGAO, 1999c). The percentages of NH deficiencies in the four severity categories in the 50 states are compared. The higher a state’s percentage of citations in the two highest severity categories, the more extensive its NH regulation is considered, as these categories require the strictest penalties (USGAO, 2003). The USGAO (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 8 concluded that several states continue to overlook and/or underestimate serious deficiencies. Effectiveness of State Implementation of Federal Mandates Comparative state policy researchers have evaluated the success of state agencies in implementing intergovernmental policies. In studies of state environmental protection policy, state agencies that effectively implemented federal mandates for air pollution control (Lester & Bowman, 1989) and hazardous waste removal (Ringquist, 1993) were identified as successful. As discussed in the previous chapter, the effectiveness of state NH surveying agencies is measured in this dissertation by their success in implementing two recent CMS initiatives, to: 1) avoid repetition in the scheduling of annual state inspections by scheduling a particular NH’s survey during different months in consecutive years, and 2) resolve deficiencies found during annual inspections and complaint investigations by the end of a 30 to 60 day “grace period”. These outcome measures are now described in greater detail. Repetitive Scheduling o f Annual State Inspections CMS requires state survey agencies to inspect every federally certified NH during a 12-month period (USGAO, 2003). hi its 1998 Initiative, CMS instructed states to avoid scheduling a NH’s annual inspection during the same month as the one in which the NH’s previous annual inspection occurred. However, the USGAO R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 69 (2000) reported that predictability of annual surveys continues to be a problem. This enables NHs to prepare for “surprise” state inspections, which possibly allows them to conceal serious quality-of-care problems. In this dissertation, repetitive scheduling of annual state inspections is determined by two measures: 1) the percentage of a state’s NHs inspected during the same month in Year 1 and Year 2, and 2) the percentage of a state’s NHs inspected during the same month in Year 2 and Year 3. The higher the percentage of a state’s NHs inspected during the same month in consecutive years, the less effective is the state in implementing the CMS mandate to avoid repetitive scheduling. Timely Resolution o f Deficiencies Current CMS regulations authorize states to give NHs with deficiencies above the C level and at or below the G level a 30 to 60 day grace period to correct these problems. NHs that resolve their deficiencies by the end of this grace period are considered to be in compliance with national standards, and no further enforcement action is taken by the state. NHs with deficiencies at the H level and above are given no grace period to correct their deficiencies (USGAO, 2000, 2003). However, despite these clear guidelines, the states remain slow in resolving NH deficiencies (Walshe & Harrington, 2002; USGAO, 2003). Deficiency resolution by a state is operationalized in this study as: 1) the average number of days a state takes to resolve C to G level deficiencies during Year 1, Year 2 and Year 3, and 2) the average number of days a state takes to resolve R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 0 deficiencies at the H level and above during Year 1, Year 2 and Year 3. A state that resolved the former within 60 days and the latter within 30 days was in accordance with CMS regulations. The fewer days it takes a state to correct NH problems, the more effective is its implementation of this federal mandate. Section Summary The collection of NH deficiency citations by state surveying agencies yielded valuable insight into the extent and effectiveness of state NH regulatory activity. The volume and severity of NH deficiency citations by the different states illustrated the extent to which national NH standards are enforced. The adherence of state agencies to federal mandates concerning survey scheduling and deficiency resolution indicated the effectiveness of each state in implementing national NH policy. External and Internal Determinants In the literature review, it was proposed that a set of external determinants (including economic development, proportion of the oldest-old, political culture and the federal support) and four sets of internal determinants (factors related to actors in the state NH policy ACF: the state NH surveying agency, state legislature, interest groups (both industry and consumer) and the governor) predict the extent and the effectiveness of state NH regulation. Table 2 reviews the independent variables introduced in the previous chapter. In the following section, these five sets of determinants are described in detail and five sets of hypotheses are presented. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 71 Table 2; Independent Variables Predicting State Nursing Home Regulation Variable Previous Studies Description Source Economic Development: 1) Urbanization 2) Education 3) Income Dye, 1966; Thompson & Scicchitano, 1987; Blomquist, 1991; Boeckelman, 1991; Harrington, et al., 1997; Radcliff & Saiz, 1998; Grabowski, 2001 1) % of state population living in urban areas 2) Median family income 3) % of state population with bachelors degree or higher U.S. Census Bureau, 2003 Proportion of Oldest- Old Kaskie, 1998; Grabowski, 2001; Harrington, Mullan & Carrillo, 2004 % of state population aged 85+ U.S. Census Bureau, 2003 Political Culture Sharkansky, 1969; Elazar, 1972, 1984; Radcliff & Saiz, 1998 Sharkansky’s 1 to 9 linear scale ofElazar’s political culture Elazar, 1984; Gray & Hanson, 2003 Federal Support of State Nursing Home Inspectors Lammers & Klingman, 1984; Gormley, 1987; Walshe & Harrington, 2002 Medicare & Medicaid $ (for NH inspections)/NH bed CMS, 2003c State Proportion of Total State Surveying Agency Funding Lester & Bowman, 1989; Ringquist, 1993; Walshe & Harrington, 2002 % of total state agency S (for NH inspections) provided by states CMS, 2003c Financial Resources Available to State Surveying Agencies Lester & Bowman, 1989; Ringquist, 1993; Walshe & Harrington, 2002 Total state agency S (for NH inspections)/NH bed CMS, 2003c Human Resources Available to State Walshe & Harrington, 2002; USDHHS, 2003; 1) FTE state inspectors/NH bed 2) Average number of years of USGAO, 2003; USDHHS, Surveying Agencies USGAO, 2003 experience of state surveyors 2003 Political Control of Teske, 1990; Hwang & Democratic, both houses = 1; CSG, 2002 Legislature Gray, 1991; Radcliff & Saiz, 1998 Mixed control = 0; Republican, both houses = -1 Legislative Professionalism Sigelman & Smith, 1980; Lammers & Klingman, 1984; Squire, 1992, 2000; Kaskie, 1998 Index: 1) compensation for legislators, 2) length of legislative session, 3) size of legislative staff Squire, 2000; Gray & Hanson, 2003 Aging/LTC Committee in State Legislature Lammers & Klingman, 1984, Browne, 1987 Aging/LTC committee in state legislature? Yes = 1; No = 0 NCSL, 2003 Nursing Home Interest Group Influence Morehouse, 1981; Thomas & Hrebenar, 2003 Influence of nursing homes. Very effective = 2; Somewhat effective = 1; Not effective = 0 Thomas & Hrebenar, 2003 Nursing Home Interest Group Access Day, 1990; Liebig, 1992; Thomas & Hrebenar, 2003 AHCA, AAHSA offices in state capital? Yes = 1; No = 0 AHCA, 2004; AAHSA, 2004 Aging Advocacy Group Influence Morehouse, 1981; Thomas & Hrebenar, 2003 Influence of seniors. Very effective = 2; Somewhat effective = 1; Not Effective = 0. Thomas & Hrebenar, 2003 Aging Advocacy Group Access Day, 1990; Liebig, 1992; Thomas & Hrebenar, 2003 AARP, Alzheimer’s Association offices in state capital city? Yes = l;No = 0 AARP, 2004; AA, 2004 Political Party of Governor Thompson & Scicchitano, 1987; Schneider, 1993 Democratic = 1; Republican = 0 NGA, 2003 Institutional Power of Barrilleaux, 1999; Beyle, Index: 1) elected state officials; Gray & Governors 2003; 2) tenure potential; 3) appointment power; 4) budget power; 5) veto power; 6) party control Hanson, 2003 Governor State of the Sabato, 1983; DiLeo, 1997 Nursing home references in NGA, 2003 State Address SOS? Yes = 1; No = 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 72 Procedure In this dissertation, the selection of the determinants included in HLM models for the extent and effectiveness of state NH regulation occurred in three steps. First, measurements for the variables described in the literature review (see Table 2) were constructed and the distributions of these variables across the 50 states were analyzed. Second, multicollinearity among these factors was tested. Variables that were strongly correlated with other factors were removed. Reducing the number of variables increased statistical degrees of freedom (Bohmstedt & Knoke, 1994; Kaskie, 1998). Third, the effects of the remaining variables on extent and effectiveness were evaluated through multiple regression. The determinants evaluated in this study are described in the following section. External Determinants External determinants are characteristics of a state’s environment that impact both the NH policy ACF and NH regulatory outputs produced by this subsystem. In the literature review, economic development, proportion o f the oldest-old, political culture and federal support were proposed as factors that predict the extent and effectiveness of state NH regulation. Economic Development Studies of state regulation have discussed the link between economic development and policy outcomes described by Dye (1966). In studies of R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 73 occupational safety (Thompson & Scicchitano, 1987) and environmental protection (Blomquist, 1991; Ringquist, 1993), researchers have found that in more economically developed states, regulatory policy is a higher priority and enforcement activity is more frequent. In state health care regulation, three components of economic development described by Dye (1966) are frequently used: urbanization, education and income (Begun, et al., 1981; Lynk, 1981; Graddy & Nichol, 1990; Schneider, 1993). These factors are included in the present study (see Table 2). A positive relationship between a state’s economic characteristics and its NH regulation is expected for three reasons. First, higher levels of urbanization, education and income in a state lead to larger numbers of state residents who are able to afford NH care, placing a higher priority on state NH regulation (Harrington, et al., 1997; Grabowski, 2001). Second, a more prosperous state has greater financial resources available to support NH regulatory activity (Thompson & Scicchitano, 1987; Blomquist, 1991). Third, an urban, educated and high-income population is more likely to be knowledgeable and supportive of health care regulation (Begun, et al., 1981; Lynk, 1981; Graddy & Nichol, 1990). Information on economic development was obtained from the 2000 U.S. Census (United States Census Bureau, 2003). Table 3 provides a comparison of the states on these measures. It was predicted that greater urbanization, higher education, and greater median household income in a state lead to more extensive and effective state NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 4 Table 3: Economic Development and Proportion of Oldest-Old State Urbanization (%) Bachelors Degree (%) Median Household Income ($) 85+ (%) Alabama 55.4 19.0 34,135 1.5 Alaska 65.7 24.7 51,571 0.4 Arizona 88.2 23.5 40,558 1.3 Arkansas 52.4 16.7 32,182 1.7 California 94.5 26.6 47,493 1.3 Colorado 84.5 32.7 47,203 1 .1 Connecticut 87.7 31.4 53,935 1.9 Delaware 80.0 25.0 47,381 1.3 Florida 89.3 22.3 38,819 2.1 Georgia 71.7 24.3 42,433 1.1 Hawaii 91.6 26.2 49,820 1.4 Idaho 66.4 21.7 37,572 1.4 Illinois 87.8 26.1 46,590 1.5 Indiana 70.8 19.4 41,567 1.5 Iowa 61.1 21.2 39,469 2.2 Kansas 71.4 25.8 40,624 1.9 Kentucky 55.7 17.1 33,672 1.4 Louisiana 72.7 18.7 32,566 1.3 Maine 40.2 22.9 37,240 1.8 Maryland 86.1 31.4 52,868 1.3 Massachusetts 91.4 33.2 50,502 1.8 Michigan 74.7 21.8 44,667 1.4 Minnesota 70.9 27.4 47,111 1.7 Mississippi 48.8 16.9 31,330 1.5 Missouri 69.4 21.6 37,934 1.8 Montana 54.0 24.4 33,024 1.7 Nebraska 69.7 23.7 39,250 2.0 Nevada 91.6 18.2 50,849 0.9 New Hampshire 59.2 28.7 49,467 1.5 New Jersey 94.3 29.8 55,146 1.6 New Mexico 75.0 23.5 34,133 1.3 New York 87.5 27.4 43,393 1.6 North Carolina 60.2 22.5 39,184 1.3 North Dakota 55.8 22.0 34,604 2.3 Ohio 77.3 21.1 40,956 1.6 Oklahoma 65.3 20.3 33,400 1.7 Oregon 78.7 25.1 40,916 1.7 Pennsylvania 77.0 22.4 40,106 1.9 Rhode Island 90.9 25.6 42,090 2.0 South Carolina 60.5 20.4 37,082 1.3 South Dakota 51.9 21.5 35,282 2.1 Tennessee 63.6 19.6 36,360 1.4 Texas 82.5 23.2 39,927 1 .1 Utah 88.3 26.1 51,022 1.0 Vermont 38.2 29.4 40,856 1.6 Virginia 73.0 29.5 46,677 1.2 Washington 82.0 27.7 45,776 1.4 West Virginia 46.1 14.8 29,696 1.8 Wisconsin 68.3 22.4 52,911 1.8 Wyoming 65.2 21.9 37,892 1.4 Mean 71.7 23.8 41,784 1.5 SD 14.9 4.3 6,771 0.4 Source: U.S. Census Bureau (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Urbanization Hwang and Gray (1991) and Boeckelman (1991) calculated urbanization as the percentage of a state’s residents living in a metropolitan area. In the 2000 U.S. Census, 80.2 percent of Americans reported living in a metropolitan area (U.S. Census Bureau, 2003). California was the most urban state, with 94.5 percent of its citizens living in urban areas; West Virginia (46.1 percent) was the least urban state (see Table 3). Income Income is frequently measured as median household income (Sigelman & Smith, 1980; Stonecash & Hayes, 1981; Erikson, Wright, & Mclver, 1987; Greenberg & Amer, 1989). The median household income in the United States in 2000 was $41,994 (U.S. Census Bureau, 2003). Among the 50 states, New Jersey ($55,146) had the highest median household income and West Virginia ($29,696) had the lowest (see Table 3). Education Education is assessed as the percentage of the adult population (persons aged 25 years and older) that has attained a bachelors degree or higher (Boeckelman, 1991). In the United States, 24.4 percent of the adult population had a bachelors degree in 2000 (U.S. Census Bureau, 2003). Massachusetts (33.2 percent of adults R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 6 with bachelors degrees) was the state with the highest level of education among residents; West Virginia (14.8 percent) was the state with the lowest (see Table 3). Summary o f Economic Development Variables As Table 4 demonstrates, the three variables, urbanization, income and education were significantly correlated with each other, supporting the belief of Dye (1966) that these variables represent components of the same underlying concept. This suggests that a single factor score could be constructed from these three measures (Kim & Mueller, 1978), which would result in a more parsimonious regression model (Kaskie, 1998). _______ Table 4: Correlations among Economic Development Variables Variable 1. Urbanization 2. Education 3. Income_____ N= 50; "p < .01 However, auxiliary regressions among the three variables suggested another way to improve the parsimony of the model. Urbanization and education explained nearly two-thirds of the variance in median household income (see Table 5). This indicates that differences among the states in wealth are largely driven by urbanization and education. Because of this robust finding, the wealth variable was removed from the rest of the study. .494 .650 .723* R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 77 Table 5; Auxiliary Regressions among Economic Development Variables Variable R R2 F ( 1. Urbanization .651 .423 17.26**** 2. Education .723 .523 25.76**** 3. Income .797 .635 40.96**** N=50; "* > < .001 The choice of combining the remaining two variables into a single factor score was rejected. Although urbanization and education are related, each is believed to have a unique impact on state NH regulation. Hypothesis la: Greater urbanization will lead to more extensive and effective state NH regulation. Hypothesis lb: Higher level o f education attained (proportion o f state residents with a bachelors degree or higher) will lead to more extensive and effective state NH regulation. Proportion of Oldest-Old Researchers have described the relationship between a state’s demographic composition and its policy outputs. Several studies have concluded that policies on behalf vulnerable subpopulation are strongest in states in which the concentration of citizens receiving these benefits is the greatest. Jennings (1979) and Wright, Erikson, and Mclver (1987) found that the higher the proportion of a state’s population living in poverty, the more generous was the state’s welfare policy. A similar trend in state regulatory policy has been reported. Thompson and Scicchitano (1987) found that regulation protecting workers was strongest in states with the highest rates of unionization. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 8 Similarly, aging policy is expected to be strongest in the states with the highest proportion of older persons. The policy priorities of older persons are distinct from those of younger cohorts (Lake, 1983; Lammers & Klingman, 1984), and are bigger agenda items in states with older populations (Browne, 1987). NH regulation is a particular priority for older persons. More than 90 percent of NH residents in the United States are age 65 and older; nearly half are age 85 and older (CDC, 2003). The larger the aged proportion in a state, the greater is the political constituency for state NH enforcement (Harrington, Mullan, & Carrillo, 2004). NH regulation is predicted to be strongest in the states with the highest proportion o f oldest-old (age 85 and older) persons, because this population is the most likely to receive NH services. Information on the oldest-old population in the United States in 2000 is obtained from the U.S. Census Bureau (2003). Nationally, over 4 million Americans were age 85 and older in 2000, 1.5 percent of the total U.S. population. Among the states, North Dakota (2.3 percent) had the highest proportion of oldest-old persons; Alaska (0.4 percent) had the lowest (see Table 3). Hypothesis lc: A greater proportion o f oldest-old (persons age 85 and older) will lead to more extensive and effective state NH regulation. Political Culture Several studies in the comparative state policy literature have described the connection between political culture and regulatory policy. Miller (1991) and Morgan and Watson (1991) found that political culture predicted regulatory R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 9 expenditures; moralistic states commit more tax revenue to regulatory agencies than do traditionalistic states. Radcliff and Saiz (1998) and Jacoby and Schneider (2001) concluded that political culture is associated with differences in regulatory priorities; moralistic states emphasize particularized benefits (such as public health, education and welfare), while traditionalistic states pursue collective goals (such as housing, economic development, and police protection). State aging policy is also shaped by political culture (Lammers & Klingman, 1984), and this is believed to extend to NH regulation. Harrington and Carrillo (1999) reported that some states do not carry out NH enforcement vigorously, and suggested that the political culture of a state may be a factor. In this dissertation, states with a moralistic subculture, with the goal of creating the “good society”, are expected to be most supportive of NH regulation. Traditionalistic states, driven to protect the status quo, are predicted to be least supportive. Finally, in individualistic states, dominated by the marketplace, support for NH regulation is expected to fall between these two extremes. Elazar’s 1984 description of the dominant political subcultures in each state has been cited in recent studies (Radcliff & Saiz, 1998; Jacoby & Schneider, 2001; Gray, 2003), and is displayed in Table 6. Political culture is operationalized as an interval variable, using the 1 to 9 linear scale created by Sharkansky (1969). Purely moralistic states are coded as 1, purely traditionalistic states as 9, and purely individualistic states as 5; the other values on this scale represent states in which different combinations of these cultures are present (see Figure 3). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 0 Table 6: Dominant Political Subcultures in the 50 States State Dominant Political Subculture Alabama Traditionalistic Alaska Individualistic Arizona Traditionalistic/Individualistic Arkansas Traditionalistic California Moralistic/Individualistic Colorado Moralistic Connecticut Individualistic/Moralistic Delaware Individualistic Florida Traditionalistic/Moralistic Georgia Traditionalistic Hawaii Individualistic/Traditionalistic Idaho Moralistic/Individualistic Illinois Individualistic Indiana Individualistic Iowa Moralistic/Individualistic Kansas Moralistic/Individualistic Kentucky Traditionalistic/Moralistic Louisiana Traditionalistic Maine Moralistic Maryland Individualistic Massachusetts Individualistic/Moralistic Michigan Moralistic Minnesota Moralistic Mississippi Traditionalistic Missouri Individualistic/Traditionalistic Montana Moralistic/Traditionalistic Nebraska Individualistic/Moralistic Nevada Individualistic New Hampshire Moralistic/Individualistic New Jersey Individualistic New Mexico Traditionalistic/Moralistic New York Individualistic/Moralistic North Carolina Traditionalistic/Individualistic North Dakota Moralistic Ohio Individualistic Oklahoma Traditionalistic/Moralistic Oregon Moralistic Pennsylvania Individualistic Rhode Island Individualistic/Moralistic South Carolina Traditionalistic South Dakota Moralistic/Individualistic Tennessee Traditionalistic Texas Traditionalistic/Moralistic Utah Moralistic Vermont Moralistic Virginia Traditionalistic Washington Moralistic/Individualistic West Virginia Traditionalistic/Moralistic Wisconsin Moralistic Wyoming Individualistic/Moralistic Source: Elazar (1984) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 81 Hypothesis Id: A lower political culture score (stronger moralistic tendencies) will lead to more extensive and effective state NH regulation. Federal Support CMS allocates funding to state NH surveying agencies for annual inspections and complaint investigations. CMS pays 100 percent of the survey and certification costs for NHs reimbursed under the Medicare program, and 75 percent of Medicaid survey and certification costs; the states are required to pay 25 percent of Medicaid costs (Walshe & Harrington, 2002; USGAO, 2003). CMS funding is based on the previous year’s amount, not on the number of NHs or NH beds reimbursed under Medicare and Medicaid (CMS, 2003b). Consequently, many state agencies report that they receive less than their fair share of federal support, particularly agencies in states with a large number of NH beds (Walshe & Harrington, 2002; USGAO, 2003). In this study, federal support of state NH surveying agencies was measured as total CMS survey funding per NH bed during each annual survey period. Total CMS survey funding was the sum of the Medicare and Medicaid allocations by the federal government to each state for NH inspection activities during each year (see Table 7). This information is contained in the CMS-435, the end-of-the-year expenditure report each state NH surveying agency annually submits to the federal government (CMS, 2003c). The number of NH beds in a state illustrates the size of the state surveying agency’s regulatory task. The number of NH beds in each state was obtained from the CDC (see Table 8). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 82 Table 7: CMS Expenditures for State NH Survey/Certification State 1999-2000 2000-2001 2001-2002 CMS Total ($) CMS Total (S) CMS Total ($) Alabama 3,633,526 4,185,742 4,427,760 Alaska 430,676 433,901 472,969 Arizona 1,841,341 2,263,166 2,473,578 Arkansas 4,153,628 4,189,109 5,771,553 California 30,819,148 33,609,315 36,742,547 Colorado 4,238,876 4,359,587 4,843,443 Connecticut 4,933,451 5,594,793 6,333,255 Delaware 836,097 1,026,826 770,024 Florida 8,847,484 8,864,241 8,476,814 Georgia 4,284,656 4,202,186 4,593,455 Hawaii 784,030 802,622 735,943 Idaho 1,730,504 1,686,613 1,620,094 Illinois 14,899,179 14,920,549 15,750,140 Indiana 6,174,259 7,136,476 7,839,744 Iowa 3,000,903 3,289,402 4,143,955 Kansas 4,232,968 3,833,614 4,097,070 Kentucky 2,806,904 3,371,600 4,012,041 Louisiana 4,758,500 5,481,293 5,754,025 Maine 1,627,886 1,877,380 1,867,357 Maryland 2,205,098 3,139,903 3,795,461 Massachusetts 7,255,070 7,958,871 8,862,466 Michigan 7,110,797 9,110,870 10,027,999 Minnesota 6,312,573 7,337,376 7,245,975 Mississippi 2,061,890 2,187,610 1,694,362 Missouri 7,374,383 8,242,421 9,422,426 Montana 1,231,539 1,320,110 1,648,645 Nebraska 2,345,501 2,231,859 2,773,350 Nevada 1,158,786 1,208,976 1,702,332 New Hampshire 904,095 1,093,562 1,006,085 New Jersey 5,054,682 5,551,447 5,934,373 New Mexico 1,448,055 2,435,350 2,196,707 New York 11,196,736 13,425,873 18,195,750 North Carolina 6,076,044 7,394,652 7,891,245 North Dakota 1,314,403 1,284,441 1,264,999 Ohio 15,284,439 17,102,419 17,715,150 Oklahoma 3,235,213 3,534,093 4,353,984 Oregon 3,220,314 3,127,030 3,304,043 Pennsylvania 9,797,394 9,937,128 10,359,264 Rhode Island 1,591,517 1,891,641 1,898,401 South Carolina 2,098,975 1,902,300 2,266,951 South Dakota 1,103,665 1,159,338 1,203,078 Tennessee 3,746,473 3,987,794 4,380,350 Texas 34,918,799 39,866,985 47,370,748 Utah 1,330,563 1,399,088 1,580,295 Vermont 550,905 663,929 668,141 Virginia 2,465,422 2,994,936 3,392,433 Washington 5,356,136 6,224,741 7,398,570 West Virginia 1,379,648 1,566,942 1,644,739 Wisconsin 8,154,076 8,021,519 9,308,518 Wyoming 538,565 820,991 972,157 Mean 5,237,115 5,785,052 6,441,751 SD 6,671,284 7,425,984 8,512,528 Source: CMS (2003c) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 83 Table 8: Number of Nursing Home Beds in the 50 States State 1999 2000 2001 Alabama 25,204 25,248 25,797 Alaska 818 821 882 Arizona 18,005 17,458 16,836 Arkansas 25,575 25,715 25,061 California 132,962 131,762 129,928 Colorado 20,265 20,240 20,119 Connecticut 32,386 32,433 31,721 Delaware 5,081 4,906 4,736 Florida 83,129 83,365 84,110 Georgia 39,774 39,817 39,806 Hawaii 3,858 4,006 4,040 Idaho 6,277 6,181 6,368 Illinois 111,026 110,766 108,287 Indiana 62,235 56,762 56,861 Iowa 37,494 37,034 36,944 Kansas 27,844 27,067 26,735 Kentucky 25,431 25,341 25,482 Louisiana 39,110 39,430 38,861 Maine 8,393 8,248 8,002 Maryland 30,137 31,495 30,507 Massachusetts 57,409 56,030 54,514 Michigan 51,104 50,696 49,535 Minnesota 44,611 42,149 40,836 Mississippi 17,096 17,068 17,454 Missouri 55,020 54,829 54,882 Montana 7,672 7,667 7,594 Nebraska 18,150 17,877 17,369 Nevada 5,196 5,547 5,073 New Hampshire 7,906 7,837 7,883 New Jersey 51,138 52,195 52,463 New Mexico 7,328 7,289 7,263 New York 118,656 120,514 121,592 North Carolina 40,730 41,376 42,194 North Dakota 7,049 6,954 6,757 Ohio 104,817 105,038 103,974 Oklahoma 34,611 33,903 32,776 Oregon 13,776 13,500 12,977 Pennsylvania 96,248 95,063 94,833 Rhode Island 10,391 10,271 10,183 South Carolina 17,875 18,102 18,185 South Dakota 7,938 7,844 7,568 Tennessee 39,275 38,593 38,923 Texas 125,904 125,052 123,350 Utah 7,451 7,651 7,683 Vermont 3,760 3,743 3,636 Virginia 30,160 30,595 31,102 Washington 26,264 25,905 24,983 West Virginia 11,219 11,413 11,373 Wisconsin 47,286 46,395 45,652 Wyoming 3,163 3,119 3,098 Mean 36,084 35,846 35,536 SD 34,839 34,687 34,431 Source: CDC (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 8 4 Alaska, the state with the fewest NH beds (see Table 8), received the most CMS funding per NH bed in each of the three years of observation (see Table 9), followed by Texas (Years 1 and 3) and New Mexico (Year 2). Federal support of the state agencies varied widely. Alaska received more than five times the federal support of the states that received the least CMS funding per NH bed each year. These were Maryland (Year 1), Iowa (Year 2), and Mississippi (Year 3), respectively (see Table 9). Since the majority of state agency funding comes from CMS (Walshe & Harrington, 2002; US GAO, 2003), the size of this federal funding is crucial to the agency’s activities. Hypothesis le: Greater federal support (total CMS funding per NH bed) will lead to more extensive and effective state NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 85 Table 9: CMS Expenditures for State NH Inspections, per NH Bed State 1999-2000 ($/bed) 2000-2001 ($/bed) 2001-2002 ^ e d ) Alabama 144 166 172 Alaska 526 529 536 Arizona 102 130 147 Arkansas 162 163 230 California 232 255 283 Colorado 209 215 241 Connecticut 152 173 200 Delaware 165 209 163 Florida 106 106 101 Georgia 108 106 115 Hawaii 203 200 182 Idaho 276 273 254 Illinois 134 135 145 Indiana 99 126 138 Iowa 80 89 112 Kansas 152 142 153 Kentucky 110 133 157 Louisiana 122 139 148 Maine 194 228 233 Maryland 73 100 124 Massachusetts 126 142 163 Michigan 139 180 202 Minnesota 142 174 177 Mississippi 121 128 97 Missouri 134 150 172 Montana 161 172 217 Nebraska 129 125 160 Nevada 223 218 336 New Hampshire 114 140 128 New Jersey 99 106 113 New Mexico 198 334 302 New York 94 111 150 North Carolina 149 179 187 North Dakota 186 185 187 Ohio 146 163 170 Oklahoma 93 104 133 Oregon 234 232 255 Pennsylvania 102 105 109 Rhode Island 153 184 186 South Carolina 117 105 125 South Dakota 139 148 159 Tennessee 95 103 113 Texas 277 319 383 Utah 179 183 206 Vermont 147 177 184 Virginia 82 98 109 Washington 204 240 296 West Virginia 123 137 145 Wisconsin 172 173 204 Wyoming 170 263 314 Mean 156 174 190 SD 72 77 82 Sources: CMS (2003); CDC (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 86 Summary of External Determinants Correlations among the external determinants revealed four significant relationships. First, urbanization was positively correlated with education. However, urbanization and education are believed to have unique effects on state NH regulation and were retained as separate variables in the regression models (see above). Second, education was negatively correlated with political culture (see Table 10); states with a high percentage of college graduates were likely to be states on the low (moralistic) end of Sharkansky’s political culture scale (see Figure 3). As was the case with the previous pair of variables, education and political culture were thought to have separate effects on state NH regulation, reflecting the debate as to whether economic (Dye, 1966) or political (Elazar, 1972) characteristics of a state predict policy outcomes. Third, there was a strong negative correlation between proportion of oldest- old and federal support in each of the three years of observation; states with smaller percentages of residents aged 85 and over received more federal dollars per NH bed for inspections (see Table 10). This is illustrated in Alaska, which has the smallest proportion of oldest-old residents (see Table 3) and the smallest number of NH beds of any state (see Table 8), yet receives the most Medicare and Medicaid funding per bed in the country (see Table 9). Fourth, federal support in Years 1, 2, and 3 were strongly correlated (see Table 10). This was expected, as federal expenditures on NH regulation are R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 87 primarily based on the previous year’s allocations (CMS, 2003b). In this study, because separate regressions were performed for each year of observation, strong correlations between federal support in years 1, 2 and 3 were not a problem. There was no possibility of multiple years being included in the same model, and no risk of multicollinearity. Table 10: Correlations among External Determinants Variable 1 2 3 4 5a 5b 5c 1. Urbanization — .494” -.217 -.049 .039 .033 .079 2. Education — -.058 -.436 .048 .066 .038 3. Proportion of Oldest-Old 4. Political Culture — -.265 -.480** -.241 -.499” -.224 -.512 -.238 5a. Federal Support (Year 1) 5b. Federal Support (Year 2) 5c. Federal Support (Year 3) .946 .913” .946** N= 50; p < .01 In summary, five external determinants previously described as possible determinants of state NH regulation were retained for use in the regression models to follow. These were: urbanization, education, proportion of oldest-old, political culture and federal support. One determinant (income) was removed from the study. Internal Determinants Sabatier and Jenkins-Smith (1999) describe the policy subsystem that forms around a particular policy as an advocacy coalition framework (ACF). Internal determinants are factors related to the key players within this ACF that are expected to affect the policy outputs that emerge. In this dissertation, characteristics of four R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 88 sets o f actors- the state NH surveying agency, state legislature, interest groups (NH industry groups and aging advocacy groups) and the governor- are expected to predict the extent and effectiveness of state NH regulation. State Surveying Agency State surveying agencies are the central actors in state NH regulatory ACFs; they are responsible for ensuring that NHs meet national Medicare and Medicaid standards. State agencies conduct annual inspections, investigate complaints, enforce sanctions and report their findings to CMS (USGAO, 2000; Walshe, 2001; Wiener, 2003). Two characteristics of state surveying agencies were studied in this dissertation as possible predictors of the extent and effectiveness of NH regulation: financial resources and human resources. Financial Resources The scope of state regulation largely depends on the funding available to the agency relevant state agency (Lester & Bowman, 1989; Mazmanian & Sabatier, 1989; Ringquist, 1993; Elling, 2003). Walshe and Harrington (2002) suggested a connection exists between the fiscal capacity of a state NH surveying agency and its regulatory activities. Furthermore, in their survey of agencies in the 50 states and the District of Columbia, 37 (or 72.5 percent) agencies reported that they had insufficient funding to meet national NH inspection requirements (Walshe & Harrington, 2002). In this study, state agencies with greater financial resources R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 89 were expected to provide more extensive and more effective NH regulation. Financial resources were assessed through two variables: the percentage o f total state survey agency funding provided by the state and total state survey agency funding per NH bed. Proportion o f Total State Surveying Agency Funding Provided by the State The federal government, which pays all Medicare survey costs and 75 percent of Medicaid costs, provides the majority of state survey agency funding. The states are required to pay the remaining 25 percent of Medicaid costs in matching funds (USGAO, 2000; Walshe & Harrington, 2002). However, the states can raise additional funds for NH regulation from other sources, such as NH licensing fees (USGAO, 2000; Walshe & Harrington, 2002). For example, during the years of observation in this study, three states (Illinois, Michigan and Ohio) used revenues raised through state NH licensing fees, as well as state Medicaid funds, to pay for state NH inspections (CMS, 2003c). The proportion of state agency funding provided by the states themselves was obtained from the annual CMS-435 expenditure reports submitted to CMS. Ohio (Year 1) and Illinois (Years 2 and 3), which used state NH licensing fees in addition to Medicaid funds, shouldered the highest percentage of state survey costs. Alaska (Year 1), Wyoming (Year 2) and Montana (Year 3) paid the smallest percentages of their own NH surveying agency costs (see Table 11). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 0 Table 11: Total Expenditures for NH Survey/Certification; State Proportion State 1999-2000 Total ($) State % 2000-2001 Total ($) State % 2001-2002 Total ($) State % Alabama 4,144,742 12.33 4,793,099 12.67 5,071,775 12.70 Alaska 480,318 10.34 487,982 11.08 537,488 12.00 Arizona 2,068,566 10.98 2,547,835 11.17 2,785,726 11.21 Arkansas 5,018,787 17.24 5,054,984 17.13 6,678,440 13.58 California 35,776,622 13.86 39,033,466 13.90 42,703,391 13.96 Colorado 4,963,657 14.60 5,084,412 14.26 5,627,352 13.93 Connecticut 5,984,447 17.56 6,784,640 17.54 7,490,646 15.45 Delaware 994,141 15.90 1,288,237 20.29 896,155 14.07 Florida 10,413,039 15.03 10,486,046 15.47 9,984,504 15.10 Georgia 5,057,042 15.27 4,944,850 15.02 5,420,738 15.26 Hawaii 904,653 13.33 925,778 13.30 847,683 13.18 Idaho 2,354,512 26.50 2,191,498 23.04 2,045,896 20.81 Illinois 17,545,856 15.08 22,599,724 33.98 23,767,074 33.73 Indiana 7,303,493 15.46 8,411,459 15.16 9,157,102 14.39 Iowa 3,914,474 23.34 4,281,753 23.18 5,350,724 22.55 Kansas 5,092,983 16.89 4,591,864 16.51 4,862,591 15.74 Kentucky 3,221,528 12.87 3,863,015 12.72 4,719,985 15.00 Louisiana 5,831,251 18.40 6,661,960 17.72 6,977,100 17.53 Maine 1,879,792 13.40 2,159,159 13.05 2,138,638 12.68 Maryland 2,517,505 12.41 3,611,571 13.06 4,388,805 13.52 Massachusetts 8,628,931 15.92 9,186,896 13.37 10,205,731 13.16 Michigan 8,203,865 13.32 12,481„052 27.00 12,230,527 18.01 Minnesota 7,408,335 14.79 8,570,116 14.38 8,493,340 14.69 Mississippi 2,517,862 18.11 2,634,362 16.96 2,117,254 19.97 Missouri 8,634,558 14.59 9,630,116 14.41 10,987,947 14.25 Montana 1,389,673 11.38 1,471,852 10.31 1,807,353 8.78 Nebraska 2,886,701 18.75 2,759,885 19.13 3,374,692 17.82 Nevada 1,343,106 13.72 1,390,114 13.03 1,956,958 13.01 New Hampshire 1,038,462 12.94 1,258,274 13.09 1,174,420 14.33 New Jersey 6,690,227 24.45 6,350,780 12.59 6,795,028 12.67 New Mexico 1,682,493 13.93 2,776,570 12.29 2,614,296 15.97 New York 14,286,065 21.62 15,790,270 14.97 21,174,215 14.07 North Carolina 7,146,930 14.98 8,673,231 14.74 9,144,749 13.71 North Dakota 1,522,105 13.65 1,485,262 13.52 1,467,802 13.82 Ohio 21,170,290 27.80 22,394,305 23.63 22,435,974 21.04 Oklahoma 4,082,305 20.75 4,459,441 20.75 5,225,860 16.68 Oregon 3,797,717 15.20 3,600,556 13.15 3,798,100 13.01 Pennsylvania 11,512,289 14.90 11,617,988 14.47 12,011,315 13.75 Rhode Island 1,792,494 11.21 2,128,486 11.13 2,167,657 12.42 South Carolina 2,511,098 16.41 2,210,156 13.93 2,695,354 15.89 South Dakota 1,295,650 14.82 1,361,930 14.88 1,407,751 14.54 Tennessee 4,488,287 16.53 4,776,848 16.52 5,264,603 16.80 Texas 42,272,847 17.40 47,252,704 15.63 56,891,284 16.94 Utah 1,563,854 14.92 1,650,849 15.25 1,862,057 15.13 Vermont 624,601 11.80 771,819 13.98 757,446 11.79 Virginia 2,833,731 13.00 3,443,934 13.04 3,923,250 13.53 Washington 6,315,775 15.19 7,394,404 15.82 8,674,056 14.70 West Virginia 1,694,386 18.58 1,913,000 18.09 1,918,934 14.29 Wisconsin 9,886,179 17.52 9,655,602 16.92 11,276,646 17.45 Wyoming 612,337 12.05 910,675 9.85 1,075,112 9.58 Mean 6,306,011 15.82 6,996,096 15.74 7,727,630 15.16 SD Source: CMS (2003c) 8,049,228 3.79 8,976,571 4.41 10,000,000 3.78 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 91 Hypothesis 2a: A greater proportion o f total state surveying agency funding provided by the state will lead to more extensive and effective state NH regulation. Total State Surveying Agency Funding per NH Bed In this study, total state survey agency funding in each year was obtained from the CMS-435 reports submitted by the states for Years 1-3 of the present study. The CMS-435 details the total amount of money spent in each state on NH survey and certification activities, which includes federal payments under Medicare, federal and state payments under Medicaid, and any additional funding provided by the state (CMS, 2003c). The number of NH beds in each state was obtained from the CDC. Alaska, the state with the fewest NH beds, had the highest total state agency funding per bed each year, with an average of $609 per NH bed in Year 3 (see Table 12). Idaho and New Mexico, two other states with few NH beds, ranked second in Years 1 and 2, respectively. However, Texas, the state with the second-largest number of NH beds (see Table 7) had the second-highest total state agency funding, an average of $461 per NH bed, in Year 3 (see Table 12). Maryland ($83 per bed in Year 1), Virginia ($113 per bed in Year 2), and Florida ($119 per bed in Year 3) had the lowest total state agency funding per NH bed, respectively, during the three years of observation. Virginia, Maryland and Mississippi had the second lowest total state agency funding in Years 1, 2, and 3, respectively (see Table 12). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 2 Table 12: Total Expenditures for State NH Survey/Certification, per NH Bed State 1999-2000 ($/bed) 2000-2001 (S/bed) 2001-2002 ($/bed) Alabama 164 190 197 Alaska 587 594 609 Arizona 115 146 165 Arkansas 196 197 266 California 269 296 329 Colorado 245 251 280 Connecticut 185 209 236 Delaware 196 263 189 Florida 125 126 119 Georgia 127 124 136 Hawaii 234 231 210 Idaho 375 354 321 Illinois 158 204 219 Indiana 117 148 161 Iowa 104 116 145 Kansas 183 170 182 Kentucky 127 152 185 Louisiana 149 169 180 Maine 224 262 267 Maryland 83 115 144 Massachusetts 150 164 187 Michigan 161 246 247 Minnesota 166 203 208 Mississippi 147 154 121 Missouri 157 176 200 Montana 181 192 238 Nebraska 159 154 194 Nevada 258 251 386 New Hampshire 131 161 149 New Jersey 131 122 130 New Mexico 230 381 360 New York 120 131 174 North Carolina 175 210 217 North Dakota 216 214 217 Ohio 202 213 216 Oklahoma 118 132 159 Oregon 276 262 293 Pennsylvania 120 122 127 Rhode Island 172 207 213 South Carolina 140 122 148 South Dakota 163 173 186 Tennessee 114 124 135 Texas 336 378 461 Utah 210 216 242 Vermont 166 206 208 Virginia 94 113 126 Washington 240 285 347 West Virginia 151 168 169 Wisconsin 209 208 247 Wyoming 194 292 347 Mean 185 206 224 SD 83 87 94 Sources: CMS (2003); CDC (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 93 Hypothesis 2b: Greater total surveying agency funding (per NH bed) will lead to more extensive and effective state NH regulation. Human Resources Regulatory policy also depends on an agency’s human resources (Mazmanian & Sabatier, 1989; Elling, 2003). Lester and Bowman (1989) and Ringquist (1993) concluded that the larger an agency staff, the greater was the state’s ability to rapidly implement federal requirements. Scholz, et al. (1991) found that the experience of agency personnel was directly related to the quantity and quality of regulatory outputs. The quantity and quality of an agency’s human resources were also expected to play a role in state NH regulation. In this study, human resources were assessed through: 1) workload and 2) experience. These measures were derived from responses of the 50 state NH surveying agency directors to a questionnaire conducted by the United States Department of Health and Human Services, Office of the Inspector General (OIG) in April and May 2002 (USDHHS, 2003). This dissertation utilized responses of state directors to two OIG questionnaire items: 1) number o f full-time equivalent (FTE) employees in the state surveying agency and 2) average number o f years o f experience o f state survey agency employees. Although the OIG questionnaire was conducted during the 2001-2002 annual inspection period (Year 3 of the present study), the responses of state agency directors to the two questionnaire items of interest were applied to Year 1 and Year 2 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 94 in the present study as well. This procedure was appropriate because state agency staffing levels over the three-year period were generally stable, due to hiring limitations imposed in most states (USGAO, 2003; USDHHS, 2003). Workload o f State NH Surveyors State NH surveying agencies are frequently described as understaffed (USGAO, 2003; USDHHS, 2003). In 2002, only seven states reported that they had sufficient staffing to carry out federal requirements (Walshe & Harrington, 2002). In this study, the number o f state NH surveying agency employees was obtained from the responses of state agency directors to the 2002 OIG questionnaire (USDHHS, 2003). An average of 80 full-time equivalent (FTE) employees (SD = 91) worked at state agencies during the 2001-2002 inspection period. Texas (449) reported the most FTE employees, Alaska (six) the fewest (see Table 13). However, comparisons of state agency capacity are incomplete unless the regulatory task each state undertakes is also considered. Because wide differences exist in the number of NH beds overseen by each state, the workload o f state NH surveying agency employees was selected as a more appropriate measure of capacity. Workload was calculated as the number ofN H beds per state agency FTE employee. During Year 3, Colorado had the heaviest average workload (915 NH beds per FTE employee), while Alaska (147) had the lightest burden (see Table 14). Heavy workload is an indication of an understaffed state agency, one ill-equipped to properly implement federal policy (Lester & Bowman, 1989; Ringquist, 1993). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 5 Table 13: Number of State NH Surveying Agency Employees State Number of FTE Employees Alabama 47 Alaska 6 Arizona 37 Arkansas 65 California 402 Colorado 22 Connecticut 45 Delaware 15 Florida 168 Georgia 66 Hawaii 12 Idaho 15 Illinois 226 Indiana 106 Iowa 61 Kansas 58 Kentucky 76 Louisiana 99 Maine 24 Maryland 51 Massachusetts 82 Michigan 136 Minnesota 45 Mississippi 46 Missouri 212 Montana 24 Nebraska 23 Nevada 10 New Hampshire 16 New Jersey 80 New Mexico 31 New York 225 North Carolina 110 North Dakota 22 Ohio 154 Oklahoma 90 Oregon 63 Pennsylvania 136 Rhode Island 17 South Carolina 20 South Dakota 19 Tennessee 79 Texas 449 Utah 16 Vermont 13 Virginia 45 Washington 62 West Virginia 22 Wisconsin 121 Wyoming 9 Mean 80 SD Source: USDHHS (2003) 91 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 96 Table 14; Number NH Beds/FTE State Agency Employee State Workload 1999-2000 Workload 2000-2001 Workload 2001-2002 Alabama 536 537 549 Alaska 136 137 147 Arizona 487 472 455 Arkansas 393 396 386 California 331 328 323 Colorado 921 920 915 Connecticut 720 721 705 Delaware 339 327 316 Florida 495 496 501 Georgia 603 603 603 Hawaii 322 334 337 Idaho 418 412 425 Illinois 491 490 479 Indiana 587 535 536 Iowa 615 607 606 Kansas 480 467 461 Kentucky 335 333 335 Louisiana 395 398 393 Maine 350 344 333 Maryland 591 618 598 Massachusetts 700 683 665 Michigan 376 373 364 Minnesota 991 937 907 Mississippi 372 371 379 Missouri 260 259 259 Montana 320 319 316 Nebraska 789 777 755 Nevada 520 555 507 New Hampshire 494 490 493 New Jersey 639 652 656 New Mexico 236 235 234 New York 527 536 540 North Carolina 370 376 384 North Dakota 320 316 307 Ohio 681 682 675 Oklahoma 385 377 364 Oregon 219 214 206 Pennsylvania 708 699 697 Rhode Island 611 604 599 South Carolina 894 905 909 South Dakota 418 413 398 Tennessee 497 489 493 Texas 280 279 275 Utah 466 478 480 Vermont 289 288 280 Virginia 670 680 691 Washington 424 418 403 West Virginia 510 519 517 Wisconsin 391 383 377 Wyoming 351 347 344 Total 485 483 478 SB 185 183 181 Sources: CDC (2003); USDHHS (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 97 Hypothesis 2c: A lighter workload for surveying agency employees will lead to more extensive and effective state NH regulation. Experience o f State NH Surveyors Several states have reported problems in attracting and retaining qualified NH surveyors (Walshe & Harrington, 2002). In addition to staffing shortages, this turnover in state agencies has resulted in a lack of experienced personnel. In its 2002 survey of state agencies, the USGAO found a large number of inexperienced surveyors; 11 states reported that over half of their NH surveyors had two years on the job or less (USGAO, 2002). The absence of veteran NH surveyors in the states has dire consequences. Inexperienced state NH surveyors are more likely to miss minor quality-of-care problems that, left unchecked, can result in serious injury or death for NH residents (Walshe & Harrington, 2002). The average number of years of experience of state NH surveyors was asked of the 50 state survey agency directors in the 2002 OIG questionnaire; 41 state directors provided a response to this item (USDHHS, 2003). State NH surveying agency directors reported that NH surveyors had an average of 6.5 years of experience (SD = 4.0). State surveyors in Ohio (15.0 years) were the most experienced, while North Dakota (1.0 year) had the least experienced state NH surveyors (see Table 15). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 98 Table 15: Number of Years of Experience of State NH Surveyors State Average Number of Years of Experience Alabama 5 Alaska 9 Arizona 5 Arkansas 5 California 3 Colorado 5 Connecticut 13 Delaware 5 Florida 3 Georgia N/A* Hawaii 7 Idaho 3 Illinois 12 Indiana 4 Iowa 7 Kansas 6 Kentucky 2 Louisiana 2 Maine 7 Maryland N/A* Massachusetts 7 Michigan 5 Minnesota 8 Mississippi N/A* Missouri N/A* Montana N/A* Nebraska 9 Nevada 5 New Hampshire 8 New Jersey 13 New Mexico 4 New York 9 North Carolina 5 North Dakota 1 Ohio 15 Oklahoma 14 Oregon 4 Pennsylvania 1 1 Rhode Island N/A* South Carolina 6 South Dakota N/A* Tennessee 2 Texas 8 Utah N/A* Vermont 8 Virginia 4 Washington N/A* West Virginia 9 Wisconsin 9 Wyoming 2 Mean 6.5 SD 3.5 State survey agency director did not respond to this item in the OIG questionnaire Source: USDHHS (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 9 Hypothesis 2d: More experienced NH surveyors will lead to more extensive and effective state NH regulation. Summary of State Surveying Agency Variables Correlation analyses among the state surveying agency variables revealed two significant relationships. First, a strong positive correlation existed between state proportion of total state surveying agency funding and average number of years of experience of state agency employees in Year 1 (see Table 16). This suggests that, at least in the first year of observation, states that paid a higher proportion of their own surveying costs were also more likely to be states that retained experienced NH surveyors. Second, a strong negative correlation existed between total state surveying agency funding per NH bed and workload of state surveyors (see Table 16). This finding suggests that an inverse relationship existed between a state agency’s financial resources and agency workload. The less money that was available to a state agency, the smaller was its staff relative to its regulatory task, which resulted in a greater workload for NH surveyors. To reduce the multicollinearity risk between these two highly correlated variables, one (workload of state NH surveyors) was removed from the analysis. This was a logical approach, because the hiring of new NH surveyors to reduce agency workload is largely dictated by the availability of funding (Walshe & Harrington, 2002; USGAO, 2003; USDHHS, 2003). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 0 0 The experience of state NH surveyors was unrelated to the other state surveying agency variables (see Table 16), and was consequently included in the regression models. In short, the correlation analyses of the four state surveying agency variables resulted in the retention of three (percentage of total state surveying agency funding provided by states, total state survey agency funding per NH bed and experience of state NH surveyors) and the removal of one (workload of state NH surveyors). Table 16: Correlations among State Surveying Agency Variables Variable la lb lc 2a 2b 2c 3a 3b 3c 4 la. State % — .514*' .478** -.063 -.176 -.193 .240 .243 .253 .429** total agency funding (Yl) lb. State % .880** -.085 -.072 -.128 .077 .072 .069 .344* total agency funding (Y2) lc. State % -.067 -.069 -.112 .114 .112 .115 .301 total agency funding (Y3) 2a. Total — .938** .904** -.419" -.418*’ -.419*' -.035 state agency funding/NH bed (Yl) 2b. Total .943** -.490** -.496** -.499** -.085 state agency funding/NH bed (Y 2) 2c. Total -.438** -.436** -.448** -.072 state agency funding/NH bed (Y3) 3 a. — .997** .995*’ .274 Workload (Yl) 3b. .998** .283 Workload (Y2) 3c. .277 Workload (Y3) 4. Surveyor . . . experience N = 50; > < .05; "p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 101 State Legislature State legislatures are regular actors in policy ACFs (Sabatier & Jenkins- Smith, 1999), and play an important role in determining state NH policy (Walshe & Harrington, 2002; Polivka, Salmon, Hyer, Johnson, & Hedgecock, 2003). Legislative support is an important factor in the success of a state’s NH regulatory activity, and this support varies from state to state (Walshe & Harrington, 2002; Polivka, et al., 2003). Three measures of the state legislature are believed to predict stronger legislative support for state NH regulatory activity. These are: political control of the legislature, legislative professionalism and the presence o f an aging or long term care committee in the state legislature. Political Control A robust linkage has been demonstrated between political control of the state legislature and state regulatory policy, with a Democratic majority encouraging more vigorous enforcement and a Republican-controlled legislature discouraging it (Teske, 1990; Scholz, et al., 1991). Party control is predicted to have a similar effect on state NH policy, with a Democratic majority expected to favor state NH regulatory activity, and a Republican majority expected to oppose it. However, the Republican- controlled Florida State Senate passed Senate Bill 1202, which requires the state to revoke the license of any NH with two severe deficiencies (level I and above) within a 30-month period. This suggests that legislatures in “maverick” states occasionally defy political expectations (Lammers, 1989; Polivka, et al., 2003). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 0 2 In this study, the linkage between political control of the legislature and NH regulatory activity during each of the three annual NH inspection periods (see Table 17) was tested. States with Democratic control of both houses during an annual NH inspection period1 are coded as 1, states with Republican control of both houses were coded as -1, and states in which legislative control is split (each party controls one house) are coded as 0. Democrats controlled both chambers in 21 states in 1999,19 in 2000 and 16 in 2001. Republicans had majorities in the two houses in 17 states in both 1999 and 2000 and in 18 states in 2001. Eleven states had split legislatures (each party controlling one chamber) in 1999, 13 in 2000 and 15 in 2001 (National Conference of State Legislatures, 2003). In one state (Nebraska) the legislature is nonpartisan; no value was assigned to Nebraska for this variable. 1 For inspection periods that occur during an election year, the pre-election composition of a legislative house is used, because the majority of an annual inspection period (June 1-May 31) occurs before a post-election legislature begins its session (in January of the following year). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 0 3 Table 17: Political Control of State Legislatures — - -- -■ -- £2______________________ State 1999-2000 2000-2001 2001-2002 Alabama DEMOCRATIC DEMOCRATIC DEMOCRATIC Alaska REPUBLICAN REPUBLICAN REPUBLICAN Arizona REPUBLICAN MIXED MIXED Arkansas DEMOCRATIC DEMOCRATIC DEMOCRATIC California DEMOCRATIC DEMOCRATIC DEMOCRATIC Colorado REPUBLICAN MIXED MIXED Connecticut DEMOCRATIC DEMOCRATIC DEMOCRATIC Delaware MIXED MIXED MIXED Florida REPUBLICAN REPUBLICAN REPUBLICAN Georgia DEMOCRATIC DEMOCRATIC DEMOCRATIC Hawaii DEMOCRATIC DEMOCRATIC DEMOCRATIC Idaho REPUBLICAN REPUBLICAN REPUBLICAN Illinois MIXED MIXED MIXED Indiana MIXED MIXED MIXED Iowa REPUBLICAN REPUBLICAN REPUBLICAN Kansas REPUBLICAN REPUBLICAN REPUBLICAN Kentucky MIXED MIXED MIXED Louisiana DEMOCRATIC DEMOCRATIC DEMOCRATIC Maine DEMOCRATIC MIXED DEMOCRATIC Maryland DEMOCRATIC DEMOCRATIC DEMOCRATIC Massachusetts DEMOCRATIC DEMOCRATIC DEMOCRATIC Michigan REPUBLICAN REPUBLICAN REPUBLICAN Minnesota MIXED MIXED MIXED Mississippi DEMOCRATIC DEMOCRATIC DEMOCRATIC Missouri DEMOCRATIC MIXED MIXED Montana REPUBLICAN REPUBLICAN REPUBLICAN Nebraska N/A2 N/A N/A Nevada MIXED MIXED MIXED New Hampshire MIXED REPUBLICAN REPUBLICAN New Jersey REPUBLICAN REPUBLICAN MIXED New Mexico DEMOCRATIC DEMOCRATIC DEMOCRATIC New York MIXED MIXED MIXED North Carolina DEMOCRATIC DEMOCRATIC DEMOCRATIC North Dakota REPUBLICAN REPUBLICAN REPUBLICAN Ohio REPUBLICAN REPUBLICAN REPUBLICAN Oklahoma DEMOCRATIC DEMOCRATIC DEMOCRATIC Oregon REPUBLICAN REPUBLICAN REPUBLICAN Pennsylvania MIXED REPUBLICAN REPUBLICAN Rhode Island DEMOCRATIC DEMOCRATIC DEMOCRATIC South Carolina MIXED REPUBLICAN REPUBLICAN South Dakota REPUBLICAN REPUBLICAN REPUBLICAN Tennessee DEMOCRATIC DEMOCRATIC DEMOCRATIC Texas MIXED MIXED MIXED Utah REPUBLICAN REPUBLICAN REPUBLICAN Vermont DEMOCRATIC MIXED MIXED Virginia REPUBLICAN REPUBLICAN REPUBLICAN Washington MIXED MIXED DEMOCRATIC West Virginia DEMOCRATIC DEMOCRATIC DEMOCRATIC Wisconsin MIXED MIXED MIXED Wyoming REPUBLICAN REPUBLICAN REPUBLICAN Source: National Conference of State Legislatures (2003) 2 Nebraska’s state legislature is nonpartisan (Council of State Governments 2002). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 104 Hypothesis 3a: Democratic control o f the legislature will lead to more extensive and effective state NH regulation. Legislative Professionalism The greater the professionalism of a state legislature, the more effective and extensive its regulatory policy is expected to be. Legislators who look upon their office as a profession are believed to be more engaged in all aspects of government, and are considered more likely to insinuate themselves into the activities of state regulatory agencies (Squire, 1992a; Gerber & Teske, 2000). Sigelman and Smith (1980) found that higher legislative professionalism led to tighter consumer regulation. In addition, Lammers and Klingman (1984) suggested that a more professional state legislature favored the development of aging policy. Thus, it seems reasonable to expect a positive relationship between legislative professionalism and regulatory policy affecting older persons. This study uses Squire’s 2000 index to examine the connection between legislative professionalism and NH regulation. The index includes: 1) legislative salary, computed as base salary plus unvouchered expenditures for those legislators outside the capital city area; 2) length o f legislative session, averaged over the biennial period and including regular and special sessions and 3) staffing, rounded to the nearest ten employees (Squire, 2000). This index has two advantages. First, it incorporates multiple measures of professionalism, providing a parsimonious representation of this concept (Kaskie, 1998). Second, instead of comparing state R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 105 legislatures with each other, Squire’s index uses the U.S. Congress as a common baseline (Hamm & Moncrief, 2003), which simplifies national comparisons. The state legislatures averaged .183 (SD = .122) in Squire’s 200 index; the average legislature was 18.3 percent as professional as the U.S. Congress. The state legislature of California (.571) was considered the most professional; New Hampshire’s (.034) was the least professional (see Table 18). Table 18: Legislative Professionalism State Compensation3 Calendar4 Permanent Staff5 Professionalism5 Alabama 30,600 120 320 .067 Alaska 43,700 133 240 .232 Arizona 24,000 110 470 .185 Arkansas 12,800 45 290 .104 California 114,700 257 2,510 .571 Colorado 30,000 121 210 .172 Connecticut 28,000 122 450 .178 Delaware 33,400 170 60 .151 Florida 27,900 60 260 .249 Georgia 23,900 61 510 .107 Hawaii 32,000 108 260 .252 Idaho 23,600 79 60 .110 Illinois 67,300 135 970 .236 Indiana 27,300 138 180 .106 Iowa 29,800 105 180 .151 Kansas 21,900 134 120 .109 Kentucky 12,900 50 320 .087 Louisiana 23,800 68 420 .144 Maine 10,800 152 130 .098 Maryland 31,500 91 510 .189 Massachusetts 50,100 362 N/A .332 Michigan 79,600 337 1,360 .516 Minnesota 34,400 105 640 .179 Mississippi 17,700 92 130 .127 Missouri 40,800 133 480 .198 3 Compensation is computed as base salary plus unvouchered expenditures for legislators outside the capital city area, as reported by the Council of State Governments (2002) and does not include expenditures requiring vouchers. For states with biennial legislatures, or those with a mandated short and long session every other year, the compensation figures are averaged (Gray & Hanson, 2003). 4 Figures are averaged over the biennial period and include regular session and special sessions (Gray & Hanson, 2003). 5 Permanent staff is rounded to the nearest ten (Gray & Hanson, 2003). 6 Squire’s index is based on comparisons to the U.S. Congress for legislative salary, session length and staff (Squire, 2000). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 106 Table 18: Legislative Professionalism (Continued) State Compensation Calendar Permanent Staff Professionalism Montana 7,000 75 50 .073 Nebraska 12,000 119 200 .172 Nevada 7,800 60 170 .171 New Hampshire 100 312 140 .034 New Jersey 49,000 361 1,460 .320 New Mexico 0 48 50 .053 New York 79,400 198 3,460 .515 North Carolina 31,600 173 170 .149 North Dakota 9,400 51 30 .058 Ohio 51,700 359 550 .315 Oklahoma 45,200 72 260 .188 Oregon 23,900 104 240 .152 Pennsylvania 61,900 346 2,680 .283 Rhode Island 11,200 258 220 .113 South Carolina 10,400 166 270 .135 South Dakota 10,100 66 60 .065 Tennessee 21,200 136 210 .117 Texas 15,900 70 1,960 .215 Utah 12,000 45 110 .067 Vermont 23,300 145 40 .117 Virginia 24,300 55 470 .150 Washington 38,900 84 540 .198 West Virginia 22,900 71 160 .116 Wisconsin 73,800 363 690 .459 Wyoming 3,800 46 20 .057 Sources: Squire (2000); Gray and Hanson (2003) Hypothesis 3b: Higher legislative professionalism will lead to more extensive and effective state NH regulation. Aging or Long-Term-Care Committee in State Legislature A standing committee within a legislature devoted to a policy area signals a sustained awareness of this issue (Gilligan & Krehbiel, 1987). In regulatory policy, a standing committee represents not only a continued place for the issue on the legislative agenda; it also provides a venue for groups and individuals affected by regulatory activities to be heard. Lammers and Klingman (1984) identified aging and long-term-care (LTC) committees in the state legislatures as forums that offer R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 107 older persons opportunities to discuss aging policy. States with these committees are expected to be more responsive to NH issues. Information on aging or long-term committees was obtained from the National Conference on State Legislatures (NCSL) and official websites of legislative houses in the fifty states. States with aging or LTC committees are coded as 1; states without as 0. The presence of a standing aging or long-term care (LTC) committee in state legislatures is not widespread (National Conference on State Legislatures, 2003). Sixteen states have an aging or LTC committee in one or both of their legislative houses; 34 do not (see Table 19). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 108 Table 19: Aging and Long-Term Care Committees in State Legislatures State Committee? Name(s) Alabama NO Alaska NO Arizona NO Arkansas YES House Aging Subcommittee; Senate Nursing Home Subcommittee California YES House Committee on Aging /LTC; Senate Subcommittee on Aging & LTC Colorado NO Connecticut YES House Select Committee on Aging; Senate Select Committee on Aging Delaware NO Florida YES House Cmte. on Nursing Homes; Senate Cmte. on Health, Aging & LTC Georgia YES House Human Relations/Aging Committee Hawaii NO Idaho NO Illinois YES House Aging Committee Indiana NO Iowa NO Kansas NO Kentucky NO Louisiana NO Maine NO Maryland NO Massachusetts YES Joint Committee on Human Services & Elder Affairs Michigan YES House Sr. Health/Security/Retirement Cmte.; Senate Sr. Citizen/VA Cmte. Minnesota NO Mississippi NO Missouri YES House Sr. Sec. Cmte.; Senate Aging/Families/ Public Health Cmte. Montana NO Nebraska NO Nevada NO New Hampshire YES House Health, Human Services & Elder Affairs Committee New Jersey NO New Mexico NO New York YES House Aging Committee; Senate Aging Committee North Carolina YES House Aging Committee; Senate Pensions/Retirement/Aging Committee North Dakota NO Ohio YES House Hum. Svcs./Aging Cmte.; Senate Health/Hum. Svcs./Aging Cmte. Oklahoma NO Oregon NO Pennsylvania YES House Aging & Older Adult Services Cmte.; Senate Aging & Youth Cmte. Rhode Island NO South Carolina NO South Dakota NO Tennessee NO Texas NO Utah NO Vermont NO Virginia NO Washington YES Senate Health & LTC Committee West Virginia NO Wisconsin YES Health, Children, Families, Aging & LTC Committee Wyoming NO Source: National Conference on State Legislatures (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 109 Hypothesis 3c: The presence o f an aging or LTC committee in the state legislature will lead to more extensive and effective state NH regulation. Summary of Legislative Variables Correlation analysis revealed that a more professional state legislature was more likely to have an aging or long term care committee (see Table 20). This was not surprising, as one hallmark of a professional legislature is a highly developed and specialized committee structure (Hamm & Moncrief, 2003). However, due to the specific role of an aging or long term care committee in raising awareness of NH policy, this variable and legislative professionalism were retained in the present study, despite their strong correlation. Political control was unrelated to legislative professionalism and presence of an aging or long term care committee in all three years under observation (see Table 20). All three legislative variables were included in the remainder of this analysis. Table 20: Correlations among Legislative Variables Variable 1 lb lc 2 3 la. Political Control — .890** .873*’ -.025 .170 (Year 1) lb. Political Control .959** .019 .086 (Year 2) lc. Political Control .032 .086 (Year 3) 2. Legislative Professionalism — .538** 3. Aging/LTC Committee in State ------ Legislature N — 50; p<.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 0 Interest Groups Interest groups are regular players in state NH policy. On one side are industry groups, led by the American Health Care Association (AHCA), which represents for-profit NHs and the American Association of Homes and Services for the Aging (AAHSA), representing not-for-profit NHs. On the other side are aging advocacy groups such as AARP and the Alzheimer’s Association (AA), two of the most influential groups representing older adults at the state level (Liebig, 1992). Using size or financial resources to assess the effect of interest groups on state NH policy is problematic. First, the number of AHCA members often belies the clout of the NH industry, particularly in a state in which a few chains controls a large proportion of NH beds (Grabowski, 2001; Walshe, 2001). Second, despite its large size (over 35 million members), AARP has had an inconsistent impact on state aging policy (Lammers & Klingman, 1984; Liebig, 1992). Third, financial resources are an imperfect measure because the proportion of a group’s revenue spent on state lobbying activities is often unclear (Thomas & Hrebenar, 2003). Instead of size or financial resources, this study uses two other variables to measure the effect of NH industry groups and aging advocacy groups on state NH regulation: the influence of these interest groups and their access to decision makers. Influence of Nursing Home Industry Groups and Aging Advocacy Groups The Hrebenar-Thomas study has assessed the influence of interest group in the 50 states on five occasions, most recently in 2002 (Thomas & Hrebenar, 2003). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. I l l For each state, Hrebenar and Thomas chose a political scientist (preferably an interest group specialist) to interview state policymakers and political observers. Respondents were asked to identify the most important interests in the state over the previous five-year period. These interests were then ranked as “very effective”, “somewhat effective”, and “less/not effective”, according to the frequency with which these interests were identified as influential (1993b, 2003). The state-level influence of two interest group coalitions is relevant to this study: NH associations and senior citizens. The 2003 Hrebenar-Thomas assessment of interest group strength in the fifty states was used to construct two variables: influence ofN H industry groups and influence o f senior citizen groups. In each state, hospital/NH associations or senior citizen groups described in the 2003 Hrebenar- Thomas assessment as “very influential” were coded as 2, “somewhat influential” groups as 1, and interests described as “less/not influential” as 0. NH industry groups have more influence than senior citizen groups. In 17 states, nursing home associations are described as “very influential”. In only two states (Iowa and New Mexico) are senior citizens considered as influential. In addition, in three of the four states with the highest proportion of oldest-old (North Dakota, South Dakota, and Florida), senior citizens are described as “less/not influential” (see Table 21). NH industry and aging advocacy groups are expected to have opposite effects on enforcement; state agencies attract opposition from the industries they regulate and support from the consumers they protect (Brudney & Hebert, 1987). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 112 Table 21: Influence of Nursing Home Associations and Senior Citizens State NH: NH: NH: SC: SC: SC: Very Somewhat Less/Not Very Somewhat Less/Not Influential Influential Influential Influential Influential Influential Alabama * * Alaska * * Arizona * Arkansas * * California * * Colorado * * Connecticut * * Delaware * * Florida * * Georgia * * Hawaii * * Idaho * * Illinois * * Indiana * * Iowa * * Kansas * * Kentucky * * Louisiana * * Maine * * Maryland * * Massachusetts * * Michigan * * Minnesota * * Mississippi * * Missouri * * Montana * # Nebraska * * Nevada * * New Hampshire * * New Jersey * * New Mexico * * New York * * North Carolina * * North Dakota * * Ohio * * Oklahoma * * Oregon * * Pennsylvania * * Rhode Island * * South Carolina * * South Dakota * * Tennessee * * Texas * * Utah * * Vermont * * Virginia * * Washington * * West Virginia * * Wisconsin * * Wyoming * * Source: Thomas and Hrebenar (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 113 Hypothesis 4a: An influential NH industry will lead to less extensive and effective state N H regulation, while influential senior citizens will lead to more extensive and effective regulation. Access of Nursing Home Industry Groups and Aging Advocacy Groups The access of an interest group coalition to decision makers is another indication of its ability to advance its agenda (Day, 1990; Liebig, 1992). In state policy, lobbying activity in the state capital is important, because major policy and regulatory decisions are made in the legislature and in state agencies (Thomas & Hrebenar, 2003). In this study, access of the NH industry, and of NH residents and their families, is assessed by the presence of interest group offices in the state capital. The locations of state offices of four interest groups (AHCA, AAHSA, AARP, and the AA) are obtained from their official websites. Two variables are created from this information: NH industry groups ’ access to decision makers and aging advocacy groups ’ access to decision makers. In this study, two interest groups offices in the state capital represented the highest level of access for an interest group coalition and were coded as 2. One interest group office reflected moderate access and was coded as 1. Zero interest group offices in the state capital indicated minimal access and this was coded as 0. In 17 states, both AHCA and AAHSA have offices in the capital city, while in 24 states AARP and the AA each have capital city offices (see Table 22). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 4 Table 22: Interest Group State Offices/Offices in State Capitals State AARP AA Aging Total AHCA AAHSA NH Total Alabama Y/Y Y/N 2/1 Y/Y Y/N 2/1 Alaska Y/N N/N 1/0 Y/Y N/N 1/1 Arizona Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Arkansas Y/Y Y/Y 2/2 Y/Y N/N 1/1 California Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Colorado Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Connecticut Y/Y Y/Y 2/2 Y/N Y/N 2/0 Delaware Y/N Y/N 2/0 Y/N Y/N 2/0 Florida Y/Y Y/N 2/1 Y/Y Y/Y 2/2 Georgia Y/Y Y/Y 2/2 Y/N Y/Y 2/1 Hawaii Y/Y Y/Y 2/2 Y/Y N/N 1/1 Idaho Y/N Y/Y 2/1 Y/Y Y/N 2/1 Illinois Y/Y Y/Y 2/2 Y/Y Y/N 2/1 Indiana Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Iowa Y/Y Y/Y 2/2 Y/N Y/N 2/0 Kansas Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Kentucky Y/N Y/N 2/0 Y/N Y/N 2/0 Louisiana Y/Y Y/N 2/1 N/N Y/Y 1/1 Maine Y/N Y/N 2/0 Y/Y Y/N 2/1 Maryland Y/N Y/N 2/0 Y/N Y/N 2/0 Massachusetts Y/Y Y/N 2/1 Y/N Y/Y 2/1 Michigan Y/Y Y/N 2/1 Y/Y Y/Y 2/2 Minnesota Y/Y Y/N 2/1 Y/N Y/Y 2/1 Mississippi Y/Y Y/Y 2/2 Y/N Y/N 2/0 Missouri Y/N Y/N 2/0 Y/Y Y/Y 2/2 Montana Y/Y Y/N 2/1 N/N Y/Y 1/1 Nebraska Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Nevada Y/N Y/N 2/0 Y/N N/N 1/0 New Hampshire Y/N Y/Y 2/1 Y/Y Y/N 2/1 New Jersey Y/N Y/N 2/0 Y/N Y/N 2/0 New Mexico Y/Y Y/N 2/1 Y/N N/N 1/0 New York Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 North Carolina Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 North Dakota Y/Y Y/N 2/1 Y/Y N/N 1/1 Ohio Y/Y Y/Y 2/2 Y/N Y/Y 2/1 Oklahoma Y/N Y/Y 2/1 N/N Y/N 1/0 Oregon Y/N Y/Y 2/1 Y/N Y/N 2/0 Pennsylvania Y/Y Y/Y 2/2 Y/Y Y/N 2/1 Rhode Island Y/Y Y/Y 2/2 Y/N Y/Y 2/1 South Carolina Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 South Dakota Y/N Y/N 2/0 Y/N Y/N 2/0 Tennessee Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Texas Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Utah Y/N Y/Y 2/1 Y/Y N/N 1 /1 Vermont Y/Y Y/Y 2/2 Y/Y Y/N 2/1 Virginia Y/Y Y/N 2/1 Y/Y Y/Y 2/2 Washington Y/N Y/N 2/0 Y/N Y/N 2/0 West Virginia Y/Y Y/Y 2/2 Y/Y N/N 1/1 Wisconsin Y/Y Y/Y 2/2 Y/Y Y/Y 2/2 Wyoming Y/Y N/N 1/1 Y/Y Y/Y 2/2 Total 50/35 48/29 98/64 47/30 42/24 89/54 Sources: AARP (2004); Alzheimer’s Association (2004); AHCA (2004); AAHSA (2004) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 115 Hypothesis 4b: Greater NH industry group access to decision makers will lead to less extensive and effective state NH regulation; greater aging advocacy group access to decision makers will lead to more extensive and effective regulation. Summary of Interest Group Variables Correlations among the interest group variables revealed one strong relationship. The presence of the NH industry and that of aging advocacy organizations in the state capital city were strongly correlated with one another (see Table 23). This finding suggests that not only do NH and senior citizens groups both pursue access to state policymakers, each also seeks to counteract the influence of the other, as Kingdon (1995) has described. Because NH industry and aging advocacy groups are each believed to impact state NH regulation, the influence and access of both actors were included in the regression analyses. Table 23: Correlations among Interest Group Variables Variable 1 2 3 4 1 . Nursing Home Industry Group Influence .129 -.066 .086 2. Aging Advocacy Group Influence 3. Nursing Home Industry Group Access 4. Aging Advocacy Group Access -.151 .067 a s r * * .465 N = 50; "p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 116 Governor The governor has not been considered a frequent player in NH regulatory policy (Walshe & Harrington, 2002). However, governors have recently taken a more active role in NH reform (Polivka, et al., 2003), and the demand for governors to do more on behalf of seniors is mounting (Beyle, 2002). In this study, three factors are considered important indicators of gubernatorial influence on state NH regulation: political party, the institutional power of the governor, and the governor’ s state-of-the-state address. Political Party of the Governor Democratic governors are associated with more regulation, Republican governors with less (Wood, 1991; Schneider, 1993). Similarly, Democratic governors are expected to be more supportive of NH regulation than Republican governors (Harrington, et al., 2004). However, as the recommendation of Governor Jeb Bush (R, FL) for tougher licensure requirements indicates (Polivka, et al., 2003), a Republican governor may not always lead to industry-friendly NH policy. In this dissertation, a Democratic governor was coded as 1, and a Republican governor was coded as 0. There were 18 Democratic governors in 1999, 19 in 2000 and 20 in 2001. Republican governors numbered 30, 29 and 28 during the same years (see Table 24). In two states (Maine and Minnesota), the governor was neither a Democrat nor a Republican during this three-year period (National Governors Association, 2003); no value was assigned to these two states for this variable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 117 Table 24: Political Party of State Governors State 1999-2000 2000-2001 2001-2002 Alabama Siegelman (D) Siegelman (D) Siegelman (D) Alaska Knowles (D) Knowles (D) Knowles (D) Arizona Hull (R) Hull (R) Hull (R) Arkansas Huckabee (R) Huckabee (R) Huckabee (R) California Davis (D) Davis (D) Davis (D) Colorado Owens (R) Owens (R) Owens (R) Connecticut Rowland (R) Rowland (R) Rowland (R) Delaware Carper (D) Carper (D) Minner (D) Florida Bush (R) Bush (R) Bush (R) Georgia Barnes (D) Barnes (D) Barnes (D) Hawaii Cayetano (D) Cayetano (D) Cayetano (D) Idaho Kempthome (R) Kempthome (R) Kempthome (R) Illinois Ryan (R) Ryan (R) Ryan (R) Indiana O’Bannon (D) O’Bannon (D) O’Bannon (D) Iowa Vilsack (D) Vilsack (D) Vilsack (D) Kansas Vilsack (D) Vilsack (D) Vilsack (D) Kentucky Patton (D) Patton (D) Patton (D) Louisiana Foster: (R) Foster: (R) Foster: (R) Maine King (I) King (I) King (I) Maryland Glendening (D) Glendening (D) Glendening (D) Massachusetts Cellucci (R) Cellucci (R) Cellucci (R) Michigan Engler (R) Engler (R) Engler (R) Minnesota Ventura (Reform) Ventura (Reform) Ventura (Reform) Mississippi Fordice (R) Musgrove (D) Musgrove (D) Missouri Carnahan (D) Carnahan (D) Holden (D) Montana Racicot (R) Racicot (R) Martz (R) Nebraska Johanns(R) Johanns(R) Johanns(R) Nevada Guinn (R) Guinn (R) Guinn (R) New Hampshire Shaheen (D) Shaheen (D) Shaheen (D) New Jersey Whitman (R) Whitman (R) Whitman (R) New Mexico Johnson (R) Johnson (R) Johnson (R) New York Pataki (R) Pataki (R) Pataki (R) North Carolina Hunt (D) Hunt (D) Easley (D) North Dakota Schafer (R) Schafer (R) Hoeven (R) Ohio Taft (R) Taft (R) Taft (R) Oklahoma Keating (R) Keating (R) Keating (R) Oregon Kitzhaber (D) Kitzhaber (D) Kitzhaber (D) Pennsylvania Ridge (R) Ridge (R) Ridge (R) Rhode Island Almond (R) Almond (R) Almond (R) South Carolina Hodges (D) Hodges (D) Hodges (D) South Dakota Janklow (R) Janklow (R) Janklow (R) Tennessee Sundquist (R) Sundquist (R) Sundquist (R) Texas Bush (R) Bush (R) Perry (R) Utah Leavitt (R) Leavitt (R) Leavitt (R) Vermont Dean (D) Dean (D) Dean (D) Virginia Gilmore (R) Gilmore (R) Gilmore (R) Washington Locke(D) Locke(D) Locke(D) West Virginia Underwood (R) Underwood (R) Wise (D) Wisconsin Thompson (R) Thompson (R) Thompson (R) Wyoming Geringer (R) Geringer (R) Geringer (R) Sources: National Governors Association (2003) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 118 Hypothesis 5a: Democratic governors will lead to more extensive and effective state NH regulation. Institutional Power of the Governor Barrilleaux (1999) posits that a powerful governor predicts active regulation because of the governor’s need to deliver programs that the public demands. This is expected to be especially true for state NH regulation. First, programs for the aging enjoy popular support (Lammers & Klingman, 1984; Day, 1990). Second, reports of abuse and neglect of NH residents have increased public awareness of NH problems (Walshe & Harrington, 2002; Polivka, et al., 2003). Third, NH policy will continue to grow in importance as more Americans require long-term care (Kane, 1995; Walshe, 2001). Due to the growing public demand for NH reform (Walshe & Harrington, 2002), it is predicted that the greater the institutional power of a state’s governor, the more extensive and effective its NH regulation. This study uses the 2002 update of Beyle’s index of governors’ institutional power. Similar to Squire’s index of legislative professionalism, this scale uses the national government as its baseline; the institutional powers of the 50 state governors are compared with those of the President on six measures. A five-point scale for each measure is used; the overall average score across these measures is then computed (Beyle, 2003). The overall average score of the 50 state governors in the 2002 Beyle index was 3.5 (SD - .4). Governors in Illinois, New York and Utah had the most institutional power; the Alabama governor had the least (see Table 25). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 119 Table 25: Governors’ Institutional Powers >7 T D § A t> T > 1 0 TTtdTTI n n il'2 TV State SEP7 Tps AP5 BP1 0 VP1 1 PC1 2 Total1 3 GIP1 4 Alabama 1.0 4.0 2.0 3 4 2 16.0 2.7 Alaska 5.0 4.0 3.0 3 5 4 24.0 4.0 Arizona 2.0 4.0 2.5 3 5 1 17.5 2.9 Arkansas 2.5 4.0 2.5 3 4 2 18.0 3.0 California 1.0 4.0 3.5 3 5 4 20.5 3.4 Colorado 4.0 4.0 3.0 3 5 4 23.0 3.8 Connecticut 4.0 5.0 3.5 3 5 2 22.5 3.8 Delaware 2.5 4.0 3.5 3 5 3 21.0 3.5 Florida 3.0 4.0 1.5 3 5 4 20.5 3.4 Georgia 1.0 4.0 1.0 3 5 3 17.0 2.8 7 SEP: Separately elected executive branch officials: 5 = only governor/lieutenant governor team elected; 4.5 = governor or governor/lieutenant governor team, with one other elected official; 4 = governor/elected governor team with some process officials (attorney general, secretary of state, treasurer, auditor) elected; 3 = governor/lieutenant governor team with process officials and some major and minor party officials elected; 2.5 = governor (no team) with six or fewer officials elected, but none are major policy officials; 2 = governor (no team) with six or fewer officials elected, including one major party official; 1.5 = governor (no team) with six or fewer officials elected, but two are major party officials; 1 = governor (no team) with seven or more process and several minor officials elected. 8 TP: Tenure potential o f governors: 5 = four-year term, no restraint on reelection; 4.5 = four-year term, only three terms permitted; 4 = four-year term, only two terms permitted; 3 = four-year term, no consecutive election permitted; 2 = two-year term, no restraint on reelection; 1 = two-year term, only two terms permitted. 9 AP: Governor’ s appointment powers in six major functional areas (corrections, K-12 education, health, highways/transportation, public utilities regulation, and welfare); the six individual office scores are totaled and then averaged and rounded to the nearest .5 for the state score: 5 = governor appoints, no other approval needed; 4 = governor appoints, a board, council or legislature approves; 3 = someone else appoints, governor approves or shares appointment; 2 = someone else appoints, governor and others approve; 1 = someone else appoints, no approval or confirmation needed. BP: Governor’ s budget power: 5 = governor has full responsibility, legislature may not increase budget; 4 = governor has full responsibility, legislature can increase by special majority vote or subject to item veto; 3 = governor has full responsibility, legislature has unlimited power to change budget; 2 = governor shares responsibility, legislature has unlimited power to change budget; 1 = governor shares responsibility with other elected official, legislature has unlimited power to change budget. 1 0 BP: Governor's budget power: 5 = governor has full responsibility, legislature may not increase executive budget; 4 = governor has full responsibility, legislature can increase by special majority vote or subject to item veto; 3 = governor has full responsibility, legislature has unlimited power to change executive budget; 2 = governor shares responsibility, legislature has unlimited power to change executive budget; 1 = governor shares responsibility with other elected official, legislature has unlimited power to change executive budget. VP: Governor’ s veto power: 5 = governor has item veto and a special majority vote of the legislature is needed to override a veto (three-fifths of legislators elected or two-thirds of legislators present); 4 = has item veto with a majority of the legislators elected needed to override; 3 = has item veto with only a majority of the legislators present needed to override; 2 = no item veto, with a special legislative authority needed to override a regular veto; 1 = no item veto, only a simple legislative majority needed to override a regular veto. 1 1 VP: Governor’ s veto power: 5 = governor has item veto and a special majority vote of the legislature is needed to override a veto (3/5 of legislators elected or 2/3 of legislators present); 4 = has item veto with a majority of the legislators elected needed to override; 3 = has item veto with only a majority of the legislators present needed to override; 2 = no item veto, with a special legislative authority needed to override a regular veto; 1 = no item veto, only a simple legislative majority needed to override a regular veto. 1 2 PC: Gubernatorial party control: 5 = governor’s party (gp) has a substantial majority (75 percent or more) in both houses of the legislature; 4 = gp has a simple majority in both houses (under 75 percent) or a substantial majority in one house and a simple majority in the other; 3 = split party control in the legislature or a nonpartisan legislature; 2 = gp has a simple minority (25 percent or more) in both houses or a simple minority in one and a substantial minority (under 25 percent) in the other; 1 = gp has a substantial minority in both houses. 1 3 Sum of the scores of the six individual indices 1 4 Total divided by 6 to keep five-point scale R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 120 Table 25: Governors’ Institutional Powers (Continued) State SEP TP AP BP VP PC Total GIP Hawaii 5.0 4.0 3.0 3 5 2 22.0 3.7 Idaho 2.0 4.0 2.0 3 5 5 21.0 3.5 Illinois 4.0 5.0 3.5 3 5 4 24.5 4.1 Indiana 3.0 4.0 3.5 3 2 3 18.5 3.1 Iowa 3.0 5.0 3.5 3 5 2 21.5 3.6 Kansas 3.0 4.0 3.5 3 5 2 20.5 3.4 Kentucky 3.0 4.0 3.5 3 4 3 20.5 3.4 Louisiana 1.0 4,0 3.5 3 5 2 18.5 3.1 Maine 5.0 4.0 4.5 3 2 4 22.5 3.8 Maryland 4.0 4.0 3.0 5 5 2 23.0 3.8 Massachusetts 4.0 4.0 3.5 3 5 1 20.5 3.4 Michigan 4.0 4.0 3.5 3 5 2 21.5 3.6 Minnesota 4.0 5.0 4.0 3 5 3 24.0 4.0 Mississippi 1.5 4.0 3.0 3 5 4 20.5 3.4 Missouri 2.5 4.0 2.5 3 5 2 19.0 3.2 Montana 3.0 4.0 2.5 3 5 4 21.5 3.6 Nebraska 3.0 4.0 3.0 4 5 3 22.0 3.7 Nevada 2.5 4.0 3.5 3 2 3 18.0 3.0 New Hampshire 5.0 2.0 3.0 3 2 4 19.0 3.2 New Jersey 5.0 4.0 3.5 3 5 3 23.5 3.9 New Mexico 3.0 4.0 3.0 3 5 4 22.0 3.7 New York 4.0 5.0 3.5 4 5 3 24.5 4.1 North Carolina 1.0 4.0 3.5 3 2 3 16.5 2.8 North Dakota 3.0 5.0 3.5 3 5 4 23.5 3.9 Ohio 4.0 4.0 3.5 3 5 4 23.5 3.9 Oklahoma 1.0 4.0 1.5 3 5 4 18.5 3.1 Oregon 1.5 4.0 3.0 3 5 3 19.5 3.3 Pennsylvania 4.0 4.0 4.0 3 5 2 22.0 3.7 Rhode Island 2.5 4.0 4.0 3 2 1 16.5 2.8 South Carolina 1.0 4.0 2.5 2 5 4 18.5 3.1 South Dakota 3.0 4.0 3.5 3 5 4 22.5 3.8 Tennessee 4.5 4.0 4.0 3 4 4 23.5 3.9 Texas 1.0 5.0 1.5 2 5 4 18.5 3.1 Utah 4.0 4.5 3.0 3 5 5 24.5 4.1 Vermont 2.5 2.0 5.0 3 2 2 16.5 2.8 Virginia 2.5 3.0 3.5 3 5 2 19.0 3.2 Washington 1.0 4.0 3.0 3 5 3 19.0 3.2 West Virginia 2.0 4.0 4.0 5 5 4 24.0 4.0 Wisconsin 3.0 5.0 3.5 3 5 2 21.5 3.6 Wyoming 2.0 4.0 3.5 3 5 2 19.5 3.3 Source: Beyle (2003) Hypothesis 5b: A governor with greater institutional power will lead to more extensive and effective state NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 121 Governors’ State of the State Addresses DiLeo (1997) measured gubernatorial support for redistributive policies by identifying whether these agenda items were present in a governor’s state-of-the- state (SOS) address. Though not required in every state, most governors give an SOS address every year at the beginning of the legislative session, or every other year, in states in which the legislature meets biennially (National Governors Association (NGA), 2003). DiLeo analyzed the 32 states in which SOS were delivered every year from 1991 through 1995. A similar methodology can be used to detect whether NH policy is a priority for a governor. Stronger NH regulation is expected in those states in which NH policy appeared in the text of the governor’s most recent SOS, thereby signaling this issue is a priority. In this study, the text of every SOS address given by governors from 1999 through 2001 was reviewed. These addresses were obtained from the NGA, the Pew Center on the States and official state websites. For years in which a state’s governor gave no SOS1 5 , the previous year’s speech is substituted (see Table 26). In each state, for each year, an SOS that includes NH policy is coded as 1. An SOS that does not include NH policy is coded as 0. 1 5 State-of-the-state addresses were not given in the following states: Kentucky (1999), Arkansas (2000), Minnesota (2000), Montana (2000), Nevada (2000), North Carolina (2000), Texas (2000). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 122 Table 26; Governors’ References to Nursing Homes in SOS Addresses State 1999 2000 2001 Alabama Siegelman (D): NO NO YES Alaska Knowles (D): NO NO NO Arizona Hull (R): NO NO NO Arkansas Huckabee (R): YES N/A* YES California Davis (D): NO YES NO Colorado Owens (R): NO NO NO Connecticut Rowland (R): NO NO NO Delaware Carper (D): NO NO Minner (D): NO Florida Bush (R): YES YES YES Georgia Barnes (D): NO NO NO Hawaii Cayetano (D): NO NO YES Idaho Kempthome (R): NO NO NO Illinois Ryan (R): NO YES NO Indiana O’Bannon (D): NO NO NO Iowa Vilsack (D): YES YES NO Kansas Graves (R): NO YES NO Kentucky Patton (D): NO NO NO Louisiana Foster: (R): NO NO YES Maine King (I): NO NO NO Maryland Glendening (D): NO NO NO Massachusetts Cellucci (R): NO NO NO Michigan Engler (R): NO NO YES Minnesota Ventura (Reform): YES N/A* NO Mississippi Fordice (R): NO Musgrove (D): NO NO Missouri Carnahan (D): NO YES Holden (D): NO Montana Racicot (R): NO N/A* Martz (R): NO Nebraska Johanns (R): NO NO NO Nevada Guinn (R): NO N/A* NO New Hampshire Shaheen (D): NO NO NO New Jersey Whitman (R): NO YES NO New Mexico Johnson (R): NO NO NO New York Pataki (R) NO NO YES North Carolina Hunt (D): NO N/A* Easley (D): NO North Dakota Schafer (R): YES NO Hoeven (R): NO Ohio Taft (R): NO NO NO Oklahoma Keating (R): NO NO YES Oregon Kitzhaber (D): NO NO NO Pennsylvania Ridge (R): NO NO YES Rhode Island Almond (R): YES NO NO South Carolina Hodges (D): NO NO NO South Dakota Janklow (R): YES NO YES Tennessee Sundquist (R): NO YES NO Texas Bush (R): NO N/A* Perry (R): YES Utah Leavitt (R): NO NO NO Vermont Dean (D): NO YES YES Virginia Gilmore (R): YES YES NO Washington Locke (D): NO YES NO West Virginia Underwood (R): NO YES Wise (D): YES Wisconsin Thompson (R): NO NO NO Wyoming Geringer (R): NO NO NO Sources: National Governors Association (2003) *State-of-the-state address not given; previous year’s address used R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 123 Hypothesis 5 c: The presence ofNH policy in the governor’ s SOS will lead to more extensive and effective state NH regulation. Summary of Gubernatorial Variables There were few significant relationships among the gubernatorial variables. In Year 1 (1999-2000), a modest negative relationship existed between institutional power and political party, indicating that powerful governors were more likely to be Republican during that year. In Year 3 (2001-2002), there was a moderate positive correlation between institutional power and reference to NH policy in the SOS address, suggesting that strong governors were attuned to this issue that year. Political party was not indicative of the governor’s likelihood to refer to NHs in the SOS address (see Table 27). Correlation analyses indicated there was little risk of multicollinearity in the effects of the three gubernatorial variables on NH regulation; all three remained in the regression models. Table 27: Correlations among Gubernatorial Variables Variable la lb lc 2 3a 3b 3c la. Political Party (Year 1) — .957** .842” -.332* .012 -.182 .018 lb. Political Party (Year 2) — .880” -.268* .082 -.110 .142 lc. Political Party (Year 3) — -.214 .192 -.090 .234 2. Institutional Power — .068 -.107 .283* 3 a. SOS Address (Year 1) — .037 .238 3b. SOS Address (Year 2) . . . .106 3c. SOS Address (Year 3) . . . N — 50; p < .05; p < M R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 124 Summary of Internal Determinants Correlation analyses revealed three relationships between governors and state surveying agencies (see Table 28). First, the governor’s reference to NH policy in the SOS address in Year 3 strongly corresponded with the state’s proportion of total survey agency funding during that year. Second, a modest positive correlation existed between the governor’s institutional power and state proportion of total survey agency funding in Year 3. Third, gubernatorial institutional power positively corresponded with state surveyor experience; states with powerful governors were more successful in retaining personnel. These correlations may be interpreted as indications of mild gubernatorial support for state surveying agency activities, particularly among powerful governors. One modest correlation between gubernatorial and legislative variables was revealed (see Table 28). Republican control of the state legislature corresponded with the governor’s institutional power during Year 1. This may be related to the fact that the number of Republican governors was highest that year (see Table 24); control by the governor’s party of the state legislature is an element of gubernatorial power (Beyle, 2003) (see Table 25). In any event, the implication of this relationship for NH regulation is unclear. Finally, one relationship was detected between legislative and interest group variables (see Table 28). Legislative professionalism was positively correlated with NH industry access to decision makers in the state capital. In contrast, aging advocacy group presence in the state capital was unrelated to legislative R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 125 professionalism. This is a potentially important finding, as it suggests that the NH industry has more access than aging advocacy groups to decision-makers in the states with longer legislative sessions and greater legislative resources. The correlation analyses among the internal determinants provided no justification for further reconfiguring of these variables. The results indicate that the actors within the state NH policy ACF (the state NH surveying agency, state legislature, NH industry groups, aging advocacy groups and the governor) are independent of one another, each institution expected to have its own impact on the extent and effectiveness of state NH regulation, and supports the use in this dissertation of separate multiple regression models for each actor. In summary, 13 of the 14 internal determinants previously described as possible determinants of state NH regulation were retained for the regression analyses. Three variables related to the state surveying agency, three to the state legislature variables, two to NH industry, two to aging advocacy groups and one to the governor were included in regression models for the extent and effectiveness of state NH regulation for each year of observation. One variable (workload of state NH surveyors) was removed from the study. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 126 Table 28: Correlations among Internal Determinants Variable la lb lc 2a 2b 2c 3 4a 4b 4c la. State .514" .478** -.063 -.176 -.193 .429" -.124 -.097 -.059 % Agency Fund Y1 lb. State — » * o oo oo -.085 -.072 -.128 .344* -.074 -.047 -.077 % Agency Fund Y2 lc. State — -.067 -.069 -.112 .301 -.030 -.012 -.046 % Agency Fund Y3 2a. Total — .938’* .904" -.035 -.207 -.164 -.150 $/bed Y1 2b. Total — .943" -.085 -.138 -.092 -.078 $/bed Y2 2c. Total — -.072 -.147 -.087 -.068 S/bed Y3 3. Surv. — .060 -.032 .030 Exper. 4a. PC — .890*’ .873" Leg. Y1 4b. PC .959" Leg. Y2 4c. PC Leg. Y3 5. Leg. Prof. 6. Aging/ LTC cmte. 7. NH Influence 8. Aging Influence 9. NH. Access 10. Aging Access 11a. PP Gov. Y1 lib. PP Gov. Y2 11c. PP Gov. Y3 12. Inst. Power Governor 13a. SOS Gov. Y1 13b. SOS Gov. Y2 13c. SOS Gov. Y3 N = 50; ’ p < .05; *><.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 127 Table 28: Correlations among Internal Determinants (Continued) Variable 5 6 7 8 9 10 11a lib 11c la. State .224 .111 .208 .109 -.128 .206 -.245 -.213 -.158 % Agency Fund Y1 lb. State .252 .290* .104 .044 -.039 .236 -.178 -.155 -.222 % Agency Fund Y2 lc. State .146 .217 .206 .084 -.066 .277 -.111 -.120 -.181 % Agency Fund Y3 2a. Total .056 -.123 -.018 -.087 -.027 -.192 .062 .044 -.062 $/bed Y1 2b. Total .037 -.111 -.028 .029 -.046 -.235 .059 .041 -.083 $/bed Y2 2c. Total .071 -.076 .034 -.027 -.012 -.227 -.014 -.039 -.165 $/bed Y3 3. Surv. .282 .276 .113 -.081 -.253 .187 -.168 -.127 .002 Exper. 4a. PC -.025 .170 -.061 .203 -.128 .168 .285 .331* .224 Leg. Y1 4b. PC .019 .086 -.119 .211 -.124 .238 .143 .194 .047 Leg. Y2 4c. PC .032 .086 -.091 .172 -.191 .116 .154 .204 .100 Leg. Y3 5. Leg. — .538** -.040 .026 .282* .157 -.106 -.128 -.058 Prof. 6. Aging/ — .036 .069 .207 .195 .000 -.030 -.030 LTC cmte. 7. NH — .129 -.066 .086 -.074 -.074 .028 Influence 8. Aging — -.151 .067 .173 .149 .082 Influence 9. NH. — .465** -.192 -.194 -.204 Access 10. Aging — -.147 -.107 -.157 Access 11a. PP — .957’* .842” Gov. Y1 1 lb. PP — .880” Gov. Y2 11c. PP — Gov. Y3 12. Inst. Power Governor 13a. SOS Gov. Y 1 13b. SOS Gov. Y2 13c. SOS Gov. Y3 N = 50; 'p < .05; "p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 128 Table 28: Correlations among Internal Determinants (Continued) Variable 12 13a 13b 13c la. State % Agency Fund Y1 .253 .105 .047 .479** lb. State % Agency Fund Y2 .261 .190 .119 .108 lc. State % Agency Fund Y3 .275 .180 -.061 .138 2a. Total $/bed Y1 .139 -.182 -.112 -.220 2b. Total $/bed Y2 .120 -.196 -.110 -.262 2c. Total S/bed Y3 .055 -.168 -.102 -.273 3. Surv. Exper. .342’ .029 .121 .282 4a. PC Leg. Y1 -.295’ .023 .187 -.016 4b. PC Leg. Y2 -.258 -.023 .141 .017 4c. PC Leg. Y3 -.222 .038 .095 .072 5. Leg. Prof. .238 .094 .171 .174 6. Aging/ LTC cmte. -.024 .145 .082 -.086 7. NH Influence .114 .094 .042 .074 8. Aging Influence -.081 .031 -.121 .217 9. NH. Access -.115 .109 .057 -.208 10. Aging Access -.104 .062 .139 .138 11a. PP Gov. Y1 -.332 .012 -.182 .018 1 lb. PP Gov. Y2 -.268 .082 -.110 .142 11c. PP Gov. Y3 -.214 .192 -.090 .234 12. Inst. Power Governor .068 -.107 .283’ 13a. SOS Gov. Y1 13b. SOS Gov. Y2 13c. SOS Gov. Y3 .037 .238 .106 N = 50; > < .05; "p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 129 Summary of Independent Variables Before examining the extent and effectiveness of state NH regulation, collinearity among the 18 independent variables was assessed (see Table 29). A strong positive correlation existed between political culture and party control of the state legislature. This meant that legislative chambers in states with a traditionalistic culture were more likely to have Democratic majorities. This was largely due to Democratic control of both the Assembly and the Senate in several states in the Deep South, where the dominant political culture is traditionalistic (see Figure 2). Traditionalistic culture also corresponded to states in which governors had less institutional power. Correlation analyses also revealed strong negative relationships between the proportion of the oldest-old and total state surveying agency funding (see Table 29). This suggests that surveying agency resources are limited in the states in which the need for NH services is most immediate. Finally, correlation analyses among the independent variables revealed significant positive relationships between federal support and total state agency funding (see Table 29). Because federal Medicare and Medicaid allocations to the states represent the majority of state surveying agency funding, inclusion of federal support and total funding in regression models was considered redundant and a risk for multicollinearity. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 130 Table 29: Correlations among Independent Variables Variable 1 2 3 4 5a 5b 5c 1. Urbanization . . . .494** -.217 -.049 .043 .036 .078 2. Education . . . -.058 -.436 .045 .068 .036 3. % Oldest-Old . . . -.265 -.478"* -.501" -.512** 4. Political Culture — -.240 -.220 -.237 5a. Fed $ Y1 . . . .942** .913** 5b. Fed$Y2 . . . .945** 5c. Fed $ Y3 . . . 6a. state % total agency funding Y1 6b. state % total agency funding Y2 6c. state % total agency funding Y3_____ 7 a. Total $/bed Y1 7b. Total $/bedY2 7c. Total $/bedY3 8. Experience_________ 9a. Political Control Legislature Y1 9b. Political Control Legislature Y2 9c. Political Control Legislature Y3 10. Leg. Prof._________ 11. Aging Cmte._______ 12. NH Industry Group Influence_______ 13. Aging Advocacy Group Influence_______ 14. NH Industry Group Access_________ 15. Aging Advocacy Group Access_________ 16a. Party Gov. Y1 16b. Party Gov. Y2 16c. Party Gov. Y3_____ 17. Gov. Inst. Power 18a. Gov. SOS Y1 18b. Gov. SOS Y2 18c. Gov. SOS Y3 N = 50; ><.05; ’><.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 131 Table 29; Correlations among Independent Variables (Continued) Variable 6c 7a 7b 7c 8 9a 9b 1. Urbanization .074 .047 .045 .096 .131 -.131 .054 2. Education -.152 .031 .051 .028 .229 -.101 -.070 3. % Oldest-Old .100 .."“461""" -.490" -.509” .138 -.002 -.094 4. Political Culture .066 -.244 -.240 -.237 -.107 ....7399".... .443" 5a. Fed $ Y1 -.122 .992*’ .941" .914" -.060 -.189 -.147 5b. Fed $ Y2 -.177 .929*’ .990" .942" -.109 -.115 -.072 5c. Fed $ Y3 -.213 .896" .933" .994" -.101 -.143 -.087 6a. state % total .478" -.063 -.176 -.193 .429" -.124 -.097 agency funding Y1 6b. state % total .880** -.085 -.072 -.128 .344* -.074 -.047 agency funding Y2 6c. state % total -.067 -.069 -.112 .301 -.030 -.012 agency funding Y3 7a. Total $/bed Y! — .938" .904" -.035 -.207 -.164 7b. Total $/bedY2 — .943" -.085 -.138 -.092 7c. Total $^edY3 — -.072 -.147 -.087 8. Experience — .060 -.032 9a. Political Control — .890" Legislature Y1 9b. Political Control Legislature Y2 9c. Political Control Legislature Y3_________ 10. Leg. Prof.________ _ 11. Aging Cmte._______ 12. NH Industry Group Influence 13. Aging Advocacy Group Influence_______ 14. NH Industry Group Access ______ 15. Aging Advocacy Group Access 16a. Party Gov. Y1 16b. Party Gov. Y2 16c. Party Gov. Y3_____ 17. Gov. Inst. Power 18a. Gov. SOS Y1 18b. Gov. SOS Y2 18c. Gov. SOS Y3 N = 50; > < .05; ’> < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 132 Table 29: Correlations among Independent Variables (Continued) Variable 9c 10 11 12 13 14 15 1. Urbanization .056 .498" .242 .004 .040 .050 .020 2. Education -.017 .234 .160 -.168 -.040 .001 -.058 3. % Oldest-Old -.073 -.029 .131 .000 .142 -.067 .170 4. Political Culture .386** -.184 -.012 .195 .055 .054 .154 5a. Fed $ Y1 -.135 .047 -.137 -.033 -.083 -.021 -.212 5b. Fed $ Y2 -.056 .007 -.145 -.026 .043 -.057 -.263 5c. Fed $ Y3 -.065 .051 -.101 .012 -.032 -.009 -.252 6a. state % total -.059 .224 .111 .208 .109 -.128 .206 agency funding Y 1 6b. state % total -.077 .252 .290* .104 .044 -.039 .236 agency funding Y2 6c. state % total -.046 .146 .217 .206 .084 -.066 .277 agency funding Y3 7a. Total S/bed Y1 -.150 .056 -.123 -.018 -.087 -.027 -.192 7b. Total $/bed Y2 -.078 .037 -.111 -.028 .029 -.046 -.235 7c. Total $/bed Y3 -.068 .071 -.076 .034 -.027 -.012 -.227 8. Experience .030 .282 .276 .113 -.081 -.253 .187 9a. Political Control .873" -.025 .170 -.061 .203 -.128 .168 Legislature Y1 9b. Political Control .959** .019 .086 -.119 .211 -.124 .238 Legislature Y2 9c. Political Control .032 .086 -.091 .172 -.191 .116 Legislature Y3 10. Leg. Prof. . . . .538** -.040 .026 .282’ .157 11. Aging Cmte. — .036 .069 .207 .195 12. NH Industry — .129 -.066 .086 Group Influence 13. Aging Advocacy Group Influence — -.151 .067 14. NH Industry — A S - r * * .465 Group Access 15. Aging Advocacy Group Access -- 16a. Party Gov. Y1 16b. Party Gov. Y2 16c. Party Gov. Y3 17. Gov. Inst. Power 18a. Gov. SOS Y 1 18b. Gov. SOS Y2 18c. Gov. SOS Y3 N = 50; ’ p < .05; p < M R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 133 Table 29: Correlations among Independent Variables (Continued) Variable 16a 16b 16c 17 18a 18b 18c 1. Urbanization -.219 -.295* -.210 .127 -.029 -.146 -.093 2. Education .003 -.086 .018 .197 .059 -273 -.138 3. % Oldest-Old -.241 -.211 -.185 .122 .226 .179 .196 4. Political Culture .109 .140 .201 -.402" .049 .201 .092 5a. Fed $ Y1 .092 .072 -.035 .125 -.181 -.109 -.244 5b. Fed $ Y2 .084 .063 -.052 .084 -.212 -.114 -.262 5c. Fed $ Y3 .006 -.017 -.137 .029 -.181 -.098 -.275 6a. state % total -.245 -.213 -.158 .253 .105 .047 .479*’ agency funding Y1 6b. state % total -.178 -.155 -.222 .261 .190 .119 .108 agency funding Y2 6c. state % total -.111 -.120 -.181 .275 .180 -.061 .138 agency funding Y3 7a. Total $/bedYl .062 .044 -.062 .139 -.182 -.112 -.220 7b. Total $/bed Y2 .059 .041 -.083 .120 -.196 -.110 -.262 7c. Total $/bedY3 -.014 -.039 -.165 .055 -.168 -.102 -.273 8. Experience -.168 -.127 .002 .342’ .029 .121 .282 9a. Political Control .285 .331* .224 -.295* .023 .187 -.016 Legislature Y1 9b. Political Control .143 .194 .047 -.258 -.023 .141 .017 Legislature Y2 9c. Political Control .154 .204 .100 -.222 .038 .095 .072 Legislature Y3 10. Leg. Prof. -.106 -.128 -.058 .238 .094 .171 .174 11. Aging Cmte. .000 -.030 -.030 -.024 .145 .082 -.086 12. NH Industry -.074 -.074 .028 .114 .094 .042 .074 Group Influence 13. Aging Advocacy Group Influence .173 .149 .082 -.081 .031 -.121 .217 14. NH Industry -.192 -.194 -.204 -.115 .109 .057 -.208 Group Access 15. Aging Advocacy Group Access -.147 -.107 -.157 -.104 .062 .139 .138 16a. Party Gov. Y1 — .957** .842" -.332 .012 -.182 .018 16b. Party Gov. Y2 — .880” -.268 .082 -.110 .142 16c. Party Gov. Y3 — -.214 .192 -.090 .234 17. Gov. Inst. Power — .068 -.107 .283’ 18a. Gov. SOS Y1 — .037 .238 18b. Gov. SOS Y2 — .106 18c. Gov. SOS Y3 — V=50; ><.05; ’><.01 Federal support (the less comprehensive of the two measures) was removed from the study (as was Hypothesis le). This resulted in a total of 17 independent variables to be used in the balance of this study (see Table 30). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 134 Table 30: Independent Variables Predicting State Nursing Home Regulation ______________________________(Revised) ___________ Variable Previous Studies Description Source Economic Development: 1) Urbanization 2) Education Dye, 1966; Thompson & Scicchitano, 1987; Blomquist, 1991; Boeckelman, 1991; Harrington, etal., 1997; Radcliff & Saiz, 1998; Grabowski, 2001 1) % of state population living in urban areas 2) Median family income U.S. Census Bureau, 2003 Proportion of Oldest-Old Kaskie, 1998; Grabowski, 2001; Harrington, Mullan & Carrillo, 2004 % of state population aged 85+ U.S. Census Bureau, 2003 Political Culture Sharkansky, 1969; Elazar, 1972,1984; Radcliff & Saiz, 1998 Sharkansky’s 1 to 9 linear scale of Elazar’s political culture Elazar, 1984; Gray & Hanson, 2003 State Proportion of Total State Surveying Agency Funding Lester & Bowman, 1989; Ringquist, 1993; Walshe & Harrington, 2002 % of total state agency $ (for NH inspections) provided by states CMS, 2003c Financial Resources Lester & Bowman, 1989; Total state agency $ (for NH CMS, 2003c Available to State Surveying Agencies Ringquist, 1993; Walshe & Harrington, 2002 inspections)/NH bed Human Resources Available to State Surveying Agencies Walshe & Harrington, 2002; USDHHS, 2003; USGAO, 2003 Average number of years of experience of state surveyors USGAO, 2003; USDHHS, 2003 Political Control of Teske, 1990; Hwang & Democratic, both houses = 1; CSG, 2002 Legislature Gray, 1991; Radcliff & Saiz, 1998 Mixed control = 0; Republican, both houses = -1 Legislative Professionalism Sigelman & Smith, 1980; Lammers & Klingman, 1984; Squire, 1992, 2000; Kaskie, 1998 Index: 1) compensation for legislators, 2) length of legislative session, 3) size of legislative staff Squire, 2000; Gray & Hanson, 2003 Aging/LTC Committee in State Legislature Lammers & Klingman, 1984, Browne, 1987 Aging/LTC committee in state legislature? Yes = 1; No = 0 NCSL, 2003 Nursing Home Interest Group Influence Morehouse, 1981; Thomas & Hrebenar, 2003 Influence of nursing homes. Very effective = 2; Somewhat effective = 1; Not effective = 0 Thomas & Hrebenar, 2003 Nursing Home Interest Group Access Day, 1990; Liebig, 1992; Thomas & Hrebenar, 2003 AHCA, AAHSA offices in state capital? Yes = 1; No = 0 AHCA, 2004; AAHSA, 2004 Aging Advocacy Group Influence Morehouse, 1981; Hrebenar & Thomas, Influence of seniors. Very effective = 2; Somewhat Thomas & Hrebenar, 1987, 1992, 1993; effective = 1; Not Effective = 0. 2003 Aging Advocacy Group Access Day, 1990; Liebig, 1992; Thomas & Hrebenar, 2003 AARP, Alzheimer’s Association offices in state capital city? Yes = 1; No = 0 AARP, 2004; AA, 2004 Political Party of Governor Thompson & Scicchitano, 1987; Schneider, 1993 Democratic = 1; Republican = 0 NGA, 2003 Institutional Power of Governors Barrilleaux, 1999; Beyle, 2003; Index: 1) elected state officials; 2) tenure potential; 3) appointment power; 4) budget power; 5) veto power; 6) party control Gray & Hanson, 2003 Governor State of the State Address Sabato, 1983; DiLeo, 1997 Nursing home references in SOS? Yes = 1; No = 0 NGA, 2003 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 135 Statistical Analysis In this dissertation, bivariate statistics and linear regression were used to test the effects of external and internal determinants on the extent and effectiveness of state NH regulation. Bivariate correlations and siginificance tests explored the relationship between each determinant and outcome measure of extent and effectiveness. Multiple regression identified the variables that predicted: 1) extent and effectiveness of state NH regulation in each annual inspection period and 2) change in these outcome measures over the three years of observation. Five sets of determinants (external, agency, legislative, interest group and gubernatorial) were examined. At the conclusion of the multiple regression analyses, a theoretical model was proposed for the five outcome measures, comprising the determinants found to be predictors of each. HLM was then used to test these five theoretical models. Bivariate Statistics Bivariate correlations tested the presence of relationships between the independent and dependent variables during each year of observation. Independent variables included external determinants (socioeconomic and political characteristics of the state) and internal determinants (factors related to key actors in the state NH policy ACF). Dependent variables represented the extent (volume and severity of deficiency citations) and effectiveness (repetitive scheduling of annual state NH inspections and timeliness of deficiency resolution by the states) of state NH regulation. Significance tests revealed the strength of relationships between R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 136 independent and dependent variables. Strong correlation between a pair of variables resulted in the rejection of the null hypothesis that no relationship existed between a determinant and an outcome (Bohmstedt & Knoke, 1994). Multiple Regression Multiple regression allows the effects of several variables of interest to be assessed while controlling for each of the other factors (Schroeder, Sjoquist, & Stephan 1986; Borhnstedt & Knoke, 1994; Allison, 1999). Comparative state policy analyses of environmental protection (Gormley, 1986), occupational safety (Scholz, et al., 1991) and legislation for persons with dementia (Kaskie, 1998) have used multiple regression to integrate factors external and internal to policy subsystems. In the analysis of state NH regulation, statistical control of rival factors is necessary because random assignment of subjects (NHs) to treatment groups (the states that inspect them) is impossible. By rendering the 50 states equivalent on key external and internal determinants, potentially confounding effects of these variables are controlled. In the present study, multiple regression enables the variables that predict a regulatory outcome in a given year, and change in this outcome from year to year, to be identified. Two sets of dependent variables are observed in this dissertation. The first set of outcome measures (volume and severity of deficiency citations) represents the extent of state NH regulatory activity. The second set (repetitive scheduling of annual state inspections and deficiency resolution) reflect the effectiveness of state R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 137 implementation of national NH policy. For each outcome measure, five sets of factors were tested: one comprising external state factors, the other four related to important actors in the state NH policy ACF. The general equations for multiple regression are as follows: External Determinants Y = a + bi (urbanization) + b2(education) + 1 > 3 (proportion of oldest-old) + t)4(political culture) + e Agency Determinants Y =a + bi(state percentage of total state NH surveying agency funding) + b2(total state NH surveying agency funding) + Inexperience of state NH surveyors) + e Legislative Determinants Y = a + bi (political control of legislature) + b2(legislative professionalism) + b3(aging/LTC committee in state legislature)+ e Interest Groups Determinants Y = a + bi(NH industry influence) + b2(NH industry group access) + Imaging advocacy group influence) + b^aging advocacy group access) + e Gubernatorial Determinants Y = a + bi (political party of governor) + b2(institutional power of governor) + b3(governor’s SOS address) + e where Y is the outcome measure; a is the least squares estimate of the intercept; bi, b2, ... b4 are least squares estimates of the regression coefficients of the internal determinants, and e is a residual effect (Jaccard, et al., 1990). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 138 HLM Hierarchical linear modeling (HLM) is a regression technique that is particularly useful for analyses in which measurements are repeatedly gathered on an individual or organization over a period of time (Bryk & Raudenbush, 1992; Osborne, 2000). HLM is frequently used in studies of state educational policy, since much of the data in this issue area consist of repeated observations nested within individual students (Willms, 1999; Osborne, 2000). HLM has also been applied to other state policies that utilize longitudinal data, such as Corazzini’s study of changes to state Medicaid home health care programs (2003). HLM is ideal in the present study of state NH regulation, which involves the collection of repeated measures from individual states. By pooling these observations into a single analysis, HLM provides a parsimonious means of evaluating theoretical models for extent and effectiveness. For each dependent variable, a theoretical (conditional with time) model, encompassing the outcome in each year as well as change in the outcome from year to year, was evaluated by comparing its fit with the fit of two other models: 1) intercept (unconditional) and 2) intercept plus slope (unconditional with time). The general equations for these three models are as follows: Unconditional In the unconditional model is single-level; the outcome measure is a function of the intercept, such as: Y = / 3o+R R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 139 where Y is the outcome measure, f t is the least squares estimate of the intercept and R is a residual effect. Unconditional with Time In the unconditional with time model, the outcome measure is a linear combination of time, plus the intercept, such as: Y = f t + f t + R where Y is the outcome measure, f t is the slope of the intercept, ft is the slope of time and R is a residual effect. Conditional with Time The conditional with time model is multi-level. On the first level, the outcome measure is a linear combination of time, plus the intercept, such as: Y = f t + ft + R where Y is the outcome measure, f t is the slope of the intercept, ft is the slope of time and R is a residual effect. On the second level, the slope of the intercept and the slope of time become dependent variables predicted by determinants in the theoretical model, such as: f t = Too + 7oi X i + ... + Y ok X k + R ft = Yio +Y11X1+ ... + YikXk+R where Y o o and Y 10 are intercepts and Y 01 and Y i 1 represent slopes predicting f t and ft from variables beginning with Xi (Bryk & Raudenbush, 1992; Osborne, 2000). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 140 Section Summary The construction and analysis of independent variables in this section provided a description of the factors believed to impact the extent and effectiveness of state NH regulation. After examining these variables, 17 were selected for subsequent analyses. Linear regression was used to examine the effects of factors external and internal to the state NH policy subsystem on the extent and effectiveness of state NH regulation and to propose and test theoretical models for each outcome measure. Chapter Summary The methods chapter of this dissertation first described the sample used in this study, and how it was collected. Second, this chapter described the outcome measures for the extent and effectiveness of state NH regulation used in this dissertation. Third, the external and internal determinants that were expected to predict these outcome measures were identified. Fourth, this chapter proposed the use of bivariate statistics and multiple regression to identify factors that predict the extent and effectiveness of state NH regulation. Finally, this chapter proposed the construction of a model integrating external and internal determinants for each outcome measure and the testing of these models through HLM. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 141 Chapter 4 Results The results of this dissertation are reported in two sections. Each section corresponds to a different set of research questions presented in the Introduction. The first two research questions concerned the extent and the effectiveness of state NH regulatory activity. The answers to these questions are provided in this study through the collection of state NH deficiency citation data. These results offer insight into the consistency with which national NH policy is administered by the states. The volume and severity of NH deficiency citations in each state provide the basis for interstate comparisons of the extent of the enforcement actions taken by state surveying agencies against NHs. Analyses of the scheduling of state annual NH inspections and the timeliness of NH deficiency resolution by state agencies reveal the quality of each state’s implementation of federal NH mandates. The second section focuses on the factors expected to predict the extent and effectiveness of state NH regulatory activity. The answers to the third and fourth research questions concern which external and internal determinants impact state enforcement in a particular year and which determinants predict change in state enforcement from year to year. This section also provides answers to the fifth research question: which theoretical model best explains the extent and effectiveness of state NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 4 2 The Extent and Effectiveness of State Nursing Home Regulation Extent of State Regulation In this dissertation, the volume and severity of NH enforcement actions by the states were recorded. Volume of state regulatory activity was computed as the number of NH deficiency citations issued by the state NH surveying agency per 100 NH beds in the state. Severity was assessed as the percentage of a state’s deficiency citations in the two highest severity categories; those causing actual harm or the potential of death or serious injury to NH residents (see Table 1). The following section describes the volume and severity of state enforcement during each year of observation and discusses change in the extent of regulatory activity over this three- year period. Volume of State Nursing Home Deficiency Citations Volume in Each Year o f Observation California had the highest volume of deficiency citations in the first year of observation, with 11.4 citations per 100 NH beds. Idaho (9.9) and Alaska (9.9) had the second and third highest citation volume in Year 1. In Year 2, Wyoming (12.1), California (11.3) and Alaska (10.8) had the most citations per 100 NH beds. In Year 3, California (11.0), Hawaii (10.4), and Maine (10.2) had the highest citation volume (see Table 31). New York (2.4) had the fewest citations per 100 NH beds in Year 1, slightly fewer than Wisconsin (2.4) and Rhode Island (2.5). The following year, Maryland R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 143 (2.8), Rhode Island (2.9), and Wisconsin (3.0) were the three states with the fewest citations per 100 NH beds. Finally, during the third annual inspection period, Virginia (2.5) had the lowest volume, closely followed by Wisconsin (2.6) and New York (2.9) (see Table 31). Change in Volume over the Three-Year Period The overall volume of state NH survey agency activity was Stable over the three-year period of this study (1999-2002). The 50 state NH surveying agencies averaged 6.2 (SD = 2.3) deficiency citations per 100 NH beds in Year 2, slightly higher than the volumes in Years 1 and 3 respectively. 21 states issued their most deficiency citations per NH bed in Year 2. In most cases, these peaks were modest; however, in Arizona and Wyoming, citation volume spiked in Year 2 as compared to Years 1 and 3 (see Table 31). In 13 states, deficiency citations per NH bed increased throughout the three- year period; in Louisiana and Maine, citation volume nearly doubled from Year 1 to Year 3. In nine states, citation volume decreased over the three-year period, with Delaware and South Carolina declining by more than 40 percent from Year 1 to Year 3. Finally, seven states issued their fewest deficiency citations per NH bed in Year 2; however, in each case, citation volume was not notably lower in Year 2 than in Years 1 and 3 (see Table 31). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 144 Table 31: Number of State NH Deficiency Citations per 100 NH Beds Year 1 Year 2 Year 3 1 . California (11.4) 1. Wyoming (12.1) 1 . California (11.1) 2. Idaho (9.9) 2. California (11.3) 2. Hawaii (10.4) 3. Alaska (9.9) 3. Alaska (10.8) 3. Maine (10.2) 4. Washington (9.2) 4. West Virginia (9.9) 4. Idaho (9.5) 5. Kansas (9.1) 5. Kentucky (9.5) 5. Kansas (9.5) 6. Kentucky (9.1) 6. Kansas (9.1) 6. Wyoming (9.4) 7. Nevada (9.0) 7. Idaho (9.0) 7. West Virginia (8.8) 8. West Virginia (8.4) 8. Washington (8.8) 8. Washington (8.7) 9. South Carolina (8.1) 9. Hawaii (8.5) 9. Kentucky (8.4) 10. Hawaii (7.9) 10. Arizona (8.0) 10. Nevada (8.1) 11. Wyoming (7.8) 11. South Dakota (7.7) 11. Michigan (7.7) 12. Michigan (7.7) 12. Arkansas (7.6) 12. Oklahoma (7.5) 13. Mississippi (7.7) 13. Nevada (7.4) 13. Alaska (7.3) 14. Montana (7.6) 14. Michigan (7.2) 14. South Dakota (7.2) 15. Arkansas (7.3) 15. Mississippi (7.1) 15. Louisiana (7.1) 16. Alabama (6.7) 16. Oklahoma (7.0) 16. Tennessee (6.8) 17. Oregon (6.2) 17. Oregon (6.7) 17. Arkansas (6.7) 18. Indiana (6.1) 18. Maine (6.6) 18. Oklahoma (6.5) 19. Delaware (6.1) 19. South Carolina (6.4) 19. Georgia (6.4) 20. Arizona (5.7) 20. Alabama (6.4) 20. Florida (6.1) 21. North Carolina (5.7) 21. Florida (6.2) 21. Montana (6.1) 22. New Mexico (5.6) 22. Indiana (6.0) 22. Texas (6.0) 23. Maine (5.6) 23. North Carolina (6.0) 23. Arizona (6.0) 24. South Dakota (5.5) 24. Tennessee (5.8) 24. Colorado (6.0) 25. Texas (5.4) 25. Texas (5.8) 25. Nebraska (5.9) 26. Missouri (5.4) 26. Missouri (5.8) 26. New Mexico (5.7) 27. Florida (5.3) 27. Colorado (5.8) 27. Missouri (5.7) 28. Ohio (5.1) 28. Montana (5.5) 28. Mississippi (5.6) 29. Nebraska (5.0) 29. Delaware (5.5) 29. Minnesota (5.6) 30. Oklahoma (4.9) 30. Georgia (5.5) 30. North Dakota (5.4) 31. North Dakota (4.8) 31. Nebraska (5.5) . 31. Utah (5.1) 32. Minnesota (4.7) 32. Louisiana (5.3) 32. Ohio (5.1) 33. Tennessee (4.6) 33. Ohio (5.1) 33. Alabama (5.0) 34. Utah (4.5) 34. New Mexico (5.0) 34. Indiana (4.9) 35. Iowa (4.3) 35. Minnesota (4.7) 35. Maryland (4.7) 36. Illinois (4.3) 36. Iowa (4.7) 36. South Carolina (4.6) 37. New Hampshire (4.1) 37. North Dakota (4.6) 37. Iowa (4.6) 38. Georgia (4.1) 38. Connecticut (4.6) 38. North Carolina (4.6) 39. Colorado (4.1) 39. New Hampshire (4.5) 39. Connecticut (4.3) 40. Massachusetts (3.8) 40. Massachusetts (4.3) 40. Massachusetts (4.2) 41. Louisiana (3.7) 41. Utah (4.1) 41. Rhode Island (3.9) 42. Vermont (3.6) 42. Illinois (4.1) 42. New Hampshire (3.8) 43. Pennsylvania (3.5) 43. Vermont (3.9) 43. Delaware (3.5) 44. Connecticut (3.5) 44. New Jersey (3.6) 44. Illinois (3.4) 45. Virginia (3.1) 45. Pennsylvania (3.4) 45. New Jersey (3.2) 46. New Jersey (2.8) 46. Virginia (3.3) 46. Vermont (3.2) 47. Maryland (2.6) 47. New York (3.0) 47. Pennsylvania (2.9) 48. Rhode Island (2.5) 48. Wisconsin (3.0) 48. New York (2.9) 49. Wisconsin (2.4) 49. Rhode Island (2.9) 49. Wisconsin (2.6) 50. New York (2.4) 50. Maryland (2.8) 50. Virginia (2.5) Mean (5.7) Mean (6.2) Mean (6.0) SD (2.3) SD (2.3) SD (2.2) Source: CMS (2003c) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 145 Severity of State Nursing Home Deficiency Citations Severity in Each Year of Observation Among the 50 states, Oregon (22.5) had the highest proportion of citations at the G level and above in Year 1, followed by Wyoming (19.8) and New York (17.8). The following year, New Jersey (14.8), New York (14.7), and Oregon (14.3) led the nation in percentage of severe citations. In Year 3, New Hampshire (15.8), Massachusetts (15.1), and Utah (14.3) had the highest proportions (see Table 32). Maine (3.0), West Virginia (3.1), and Rhode Island (3.5) had the lowest percentages of G level and above deficiencies in Year 1. In Year 2, Nevada (1.2) had the lowest proportion, followed by California (2.3) and Arizona (3.4). California (0.7) had the lowest percentage of severe deficiencies of any state in any year of this study in Year 3, with Arizona (0.8) and Maine (1.2) also issuing very few severe deficiency citations that year (see Table 32). Change in Severity over the Three-Year Period The proportion of NH deficiency citations in the highest two levels of severity defined by CMS (see Table 1) steadily declined over the three-year period of observation. In Year 1, the average proportion of actual harm deficiency citations by the 50 state surveying agencies was 10.5 percent (SD = 4.5). This proportion dropped to 7.9 percent (SD — 3.2) in Year 2. By Year 3, an average of only 6.8 percent (SD - 3.7) of NH citations issued by the states were for deficiencies at the G level and above (see Table 32). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. In 27 states, the proportion of actual harm citations decreased throughout the three years of observation. This trend was particularly evident in California, which saw an eightfold decrease in its percentage of deficiency citations at the G level and above from Year 1 to Year 3 (see Table 32). In 13 states, citation severity was lowest in Year 2; in seven states, citation severity was highest in Year 2. However, only three of these 20 states (Colorado, South Dakota, and Utah) had a higher proportion of actual harm deficiencies in Year 3 than in Year 1; the other 17 states mirrored the overall national downward trend in citation severity. Finally, three states (Alaska, Oklahoma, and South Carolina) showed a steady increase in the proportion of deficiencies at the G level and above over the three-year period (see Table 32). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 147 Table 32: Percentage of State NH Deficiency Citations G and Above Year 1 Year 2 Year 3 1. Oregon (22.5) 1. New Jersey (14.8) 1 . New Hampshire (15.9) 2. Wyoming (19.8) 2. New York (14.7) 2. Massachusetts (15.1) 3. New York (17.8) 3. Oregon (14.3) 3. Utah (14.4) 4. Massachusetts (17.4) 4. Connecticut (14.3) 4. Connecticut (13.1) 5. Idaho (16.7) 5. Massachusetts (13.3) 5. South Dakota (11.8) 6. New Mexico (16.0) 6. Arkansas (12.4) 6. Minnesota (11.3) 7. New Jersey (16.0) 7. Montana (12.0) 7. Alaska (10.9) 8. North Carolina (14.8) 8. New Hampshire (11.0) 8. South Carolina (10.8) 9. Maryland (14.7) 9. New Mexico (11.0) 9. Washington (10.3) 10. Kansas (14.3) 10. North Dakota (10.3) 10. New York (10.1) 11. Indiana (14.2) 11. Alaska (10.1) 11. Oregon (9.9) 12. Washington (13.8) 12. Delaware (10.0) 12. North Carolina (9.6) 13. Arkansas (13.7) 13. North Carolina (9.6) 13. Idaho (9.2) 14. Delaware (13.6) 14. Tennessee (9.3) 14. Kentucky (8.9) 15. Connecticut (13.5) 15. Wyoming (9.3) 15. Wyoming (8.6) 16. N ew Hampshire (13.5) 16. Washington (9.2) 16. Mississippi (8.1) 17. Minnesota (13.4) 17. Indiana (9.0) 17. Tennessee (8.0) 18. Nebraska (13.4) 18. Texas (9.0) 18. New Mexico (7.9) 19. Kentucky (12.4) 19. Colorado (8.8) 19. Vermont (7.8) 20. Tennessee (12.0) 20. South Dakota (8.8) 20. Indiana (7.8) 21. Pennsylvania (11.7) 21. Minnesota (8.6) 21. Kansas (7.7) 22. Vermont (11.1) 22. Kentucky (8.5) 22. Colorado (7.5) 23. Montana (11.0) 23. Kansas (8.4) 23. New Jersey (7.1) 24. Virginia (10.0) 24. Vermont (8.2) 24. Texas (7.0) 25. Alaska (9.9) 25. Idaho (7.6) 25. Ohio (6.6) 26. South Dakota (9.7) 26. Virginia (7.5) 26. Oklahoma (6.4) 27. Texas (9.6) 27. Georgia (7.3) 27. Montana (6.3) 28. Mississippi (9.0) 28. Nebraska (7.2) 28. Georgia (6.2) 29. Arizona (8.6) 29. Louisiana (7.2) 29. Arkansas (6.0) 30. Missouri (8.5) 30. Maryland (6.9) 30. Maryland (5.5) 31. Utah (8.3) 31. South Carolina (6.8) 31. Virginia (5.3) 32. Ohio (8.0) 32. Mississippi (6.7) 32. Pennsylvania (5.2) 33. Illinois (7.9) 33. Ohio (6.6) 33. Michigan (5.1) 34. Georgia (7.7) 34. Illinois (6.1) 34. Wisconsin (5.1) 35. Alabama (7.6) 35. Pennsylvania (6.0) 35. Nebraska (5.1) 36. North Dakota (7.4) 36. Oklahoma (5.9) 36. Illinois (5.0) 37. Colorado (7.3) 37. Alabama (5.9) 37. North Dakota (4.7) 38. Louisiana (7.2) 38. Rhode Island (5.6) 38. Delaware (4.2) 39. Michigan (7.1) 39. Utah (5.5) 39. Louisiana (4.0) 40. Wisconsin (6.9) 40. Hawaii (5.3) 40. Missouri (3.7) 41. Nevada (6.9) 41. Maine (5.2) 41. Hawaii (3.3) 42. Hawaii (6.6) 42. Wisconsin (5.1) 42. Alabama (2.6) 43. Florida (6.0) 43. Florida (4.6) 43. West Virginia (2.4) 44. California (5.6) 44. West Virginia (4.4) 44. Florida (2.3) 45. South Carolina (5.5) 45. Michigan (4.1) 45. Iowa (2.1) 46. Iowa (5.2) 46. Iowa (4.0) 46. Nevada (1.7) 47. Oklahoma (3.9) 47. Missouri (3.6) 47. Rhode Island (1.5) 48. Rhode Island (3.5) 48. Arizona (3.4) 48. Maine (1.2) 49. West Virginia (3.1) 49. California (2.3) 49. Arizona (0.9) 50. Maine (3.0) 50. Nevada (1.2) 50. California (0.7) Mean (10.6) Mean (7.9) Mean (6.8) SB (4.5) SD (3.2) SB (3.8) Source: CMS (2003c) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 148 Summary of the Extent of State Nursing Home Regulation Correlation analyses revealed strong intravariable relationships among the two measures of extent (volume and severity). Citation volume in Year 1 strongly correlated with citation volume in Years 2 and 3. The proportion of citations at the G level and above was also highly correlated across the three years of the study. However, citation volume and citation severity were unrelated to each other. Among the 50 states, there were some parallels between volume and severity of NH citations, as observed in the states that consistently ranked near the top or the bottom of both lists. For example, Rhode Island ranked consistently among states with the fewest citations per NH bed and the lowest proportion of severe deficiencies. At the opposite end of the scale, enforcement in Wyoming was both frequent and severe, particularly in Year 2. At the same time, there were also states in which volume and severity of deficiency citations did not reflect one another. New York, for instance, was a state with low numbers of citations per NH bed, but a high percentage of these citations were for severe deficiencies. In contrast, California and Maine had high volumes of deficiency citations, but low proportions of citations at the G level and above. Table 33: Correlations among Dependent Variables, Extent Variable 1 lb lc 2a 2b 2c la. Volume State Citations (Year 1) — .888" .764" -.029 -.229 -.080 lb. Volume State Citations (Year 2) — .846** -.015 -.202 -.084 lc. Volume State Citations (Year 3) — -.140 -.296* -.201 2a. Severity State Citations (Year 1) 2b. Severity State Citations (Year 2) 2c. Severity State Citations (Year 3) .874** .564** .638** N = 50; p <05; pc.O l R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 149 Effectiveness of State Regulation In this dissertation, the implementation by state surveying agencies of two CMS initiatives designed to improve state NH regulation was assessed over the three-year period of observation. First, repetitive scheduling of annual state NH inspections was determined by calculating the percentage of NHs surveyed by a state during the same month in Years 1 and 2, and in Years 2 and 3. Second, the timeliness of NH deficiency resolution by states during the three years of the study was defined as the percentage of C to G level deficiencies resolved within 60 days and the percentage of deficiencies at the H level and above resolved within 30 days, as required under current CMS regulations (USGAO 2000, 2003). Repetitive Scheduling of Annual State Nursing Home Inspections Inspection Scheduling in Each Two-Year Interval Oklahoma was the most successful state in avoiding repetitive annual NH inspections in Year 2. None of that state’s NHs were inspected during the same month in Years 1 and 2. Georgia (0.6) and Utah (1.2) also had low percentages of annual NH inspections occurring in the same month in Years 1 and 2. Mississippi (0.5) had the lowest percentage of NHs inspected during the same month in Years 2 and 3. Minnesota (1.9 percent) and Georgia (2.0 percent) also avoided predictable annual NH inspections in Year 3 (see Table 34). In Year 2, the most predictable annual inspections were in Virginia, where 35.4 percent ofNHs were inspected during the same month in Years 1 and 2. North R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 5 0 Dakota (31.6 percent) and Iowa (30.1 percent) were also predictable in scheduling annual surveys during Year 2. Iowa had the most predictable surveys in Year 3. Over half (51.9 percent) of Iowa NHs were inspected in the same month in Years 2 and 3. More than a third of annual surveys were predictable in North Dakota (35.1 percent) and Maryland (34.2 percent) in Year 3 (see Table 34). Change in Inspection Scheduling over the Three-Year Period During the three-year period observed in this dissertation, the 50 state surveying agencies inspected an average of 15.0 percent (SD = 8.7) of NHs in the same month during consecutive years (see Table 33), despite the CMS mandate to avoid this practice (GAO 2000, 2003). In fact, this problem of survey predictability worsened during the three-year period of observation in this study. Across the 50 states, an average of 13.7 percent (SD = 8.9) of NHs were inspected during the same month in Years 1 and 2, while an average of 16.1 percent (SD = 10.7) of annual NH surveys in Years 2 and 3 occurred during the same month (see Table 34). During the three years (1999-2002) observed in this study, the three states that were most successful in avoiding predictable annual inspections were Georgia, Oklahoma, and Minnesota. The three states least successful in implementing this federal initiative were Iowa, North Dakota, and Virginia (see Table 34). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 151 Table 34: Repetitive Scheduling of Annual State NH Inspections Years 1 and 2 Years 2 and 3 1. Virginia (35.4) 1 . Iowa (51.9) 2. North Dakota (31.6) 2. North Dakota (35.1) 3. Kansas (30.1) 3. Maryland (34.2) 4. Arkansas (29.9) 4. Nevada (33.3) 5. Pennsylvania (27.4) 5. Delaware (31.0) 6. Washington (24.5) 6. Alabama (29.2) 7. Delaware (24.3) 7. New York (27.9) 8. Nevada (23.1) 8. Arizona (27.9) 9. South Dakota (22.5) 9. Michigan (26.3) 10. Louisiana (27.4) 10. Wisconsin (24.7) 11. Wisconsin (21.5) 11. New Jersey (23.7) 12. Maryland (19.8) 12. Colorado (23.7) 13. Kansas (19.5) 13. Virginia (22.9) 14. Oregon (18.2) 14. Washington (22.4) 15. New Jersey (17.6) 15. New Mexico (21.3) 16. Massachusetts (17.2) 16. South Dakota (21.1) 17. New York (17.0) 17. California (19.6) 18. Texas (17.0) 18. Massachusetts (19.1) 19. Connecticut (16.6) 19. West Virginia (18.6) 20. West Virginia (15.6) 20. South Carolina (18.3) 21. Vermont (15.4) 21. North Carolina (18.0) 22. Arizona (14.3) 22. Indiana (17,2) 23. Missouri (14.2) 23. Texas (16.6) 24. Rhode Island (13.6) 24. Rhode Island (16.1) 25. North Carolina (13.6) 25. Oregon (15.3) 26. Arizona (13.3) 26. New Hampshire (14.9) 27. Kentucky (13.3) 27. Vermont (14.6) 28. Indiana (13.0) 28. Kansas (13.7) 29. Hawaii (12.2) 29. Missouri (13.5) 30. New Mexico (12.2) 30. Louisiana (12.2) 31. Michigan (10.0) 31. Kentucky (12.1) 32. New Hampshire (9.9) 32. Wyoming (10.5) 33. Colorado (9.2) 33. Florida (10.2) 34. South Carolina (8.8) 34. Alaska (10.0) 35. Wyoming (8.8) 35. Connecticut (9.5) 36. Illinois (8.8) 36. Hawaii (9.1) 37. Tennessee (7.8) 37. Illinois (8.6) 38. Florida (7.1) 38. Tennessee (8.2) 39. California (6.7) 39. Idaho (7.6) 40. Montana (6.2) 40. Pennsylvania (6.1) 41. Alabama (3.7) 41. Maine (5.5) 42. Mississippi (3.6) 42. Arkansas (4.7) 43. Nebraska (3.2) 43. Nebraska (4.5) 44. Idaho (2.5) 44. Montana (4.1) 45. Ohio (2.5) 45. Ohio (3.8) 46. Maine (1.9) 46. Utah (3.6) 47. Minnesota (1.4) 47. Oklahoma (3.1) 48. Utah (1.2) 48. Georgia (2.0) 49. Georgia (0.6) 49. Minnesota (1.2) 50. Oklahoma (0) 50. Mississippi (0.5) Mean = 13.79 Mean = 16.19 SD = 8.88 SD = 10.67 Source: CMS (2003 c) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 152 Resolution of C to G Level Deficiencies C to G Level Deficiency Resolution in Each Year of Observation The states resolved non-actual harm deficiencies (C to G level), which account for the majority of NH deficiencies (see Table 35) quickly over the course of the study. In Year 1, North Carolina took the shortest amount of time (23.7 days) to correct C to G level deficiencies, followed by Alaska (25.2 days) and Maine (26.2 days). In contrast, Connecticut (81.5 days), New York (75.9 days), and Alabama (75.7 days) had the slowest average resolution of C to G level deficiencies. During this annual inspection period, 42 states resolved C to G level deficiencies in an average of less than 60 days (see Table 36). During the second annual inspection period, Alaska had the fastest resolution of C to G level deficiencies (1.3 days), followed by Maryland (12.0 days) and Massachusetts (15.8 days). Oklahoma (76.4 days), New York (76.3 days), and Montana (75.5 days) took the most time to ensure correction of these problems. As in the previous year, 42 states operated within CMS guidelines for timely resolution of C to G level deficiencies (see Table 36). Alaska resolved C to G level deficiencies in an average of less than one day (0.20) in Year 3. Maryland (11.5 days) and Kentucky (15.3 days) were the second and third fastest states, respectively in resolution of these deficiencies. New Hampshire (69.7 days), Montana (68.4 days), and New York (64.2 days) took the most time to correct these problems. In Year 3, 46 of the 50 states resolved C to G level deficiencies by the end of the 60-day grace period (see Table 36). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 153 Table 35: Number/Severity of State Nursing Home Deficiency Citations State Total C to G H to L Total C to G H to L Total C to G H to L Y1 (#) Y1 (%) Y1 (%) Y2 (#) Y2 (%) Y2 (%) Y3 (#) Y3 (%) Y3 (%) AL 1,691 93.97 1.71 1,603 93.76 1.37 1,289 94.88 .62 AK 81 93.83 0 89 97.75 0 64 85.94 3.13 AZ 1,026 85.96 1.56 1,390 84.46 1.08 909 72.50 0 AR 1,851 81.63 6.97 1,921 79.70 8.07 1,658 86.19 3.80 CA 15,159 83.31 .96 14,865 81.10 .40 14,156 78.32 .08 CO 826 87.65 1.94 1,165 90.73 1.46 1,133 91.88 .44 CT 1,119 98.30 .54 1,477 98.65 .61 1,343 98.66 .22 DE 308 85.71 .97 265 82.26 2.26 136 72.06 .74 FL 4,443 96.74 .99 5,206 96.81 .83 5,135 96.73 .25 GA 1,642 80.09 .73 2,185 83.71 1.74 2,500 86.68 1.32 HI 303 88.78 0 339 85.25 0 420 89.52 0 ID 624 94.23 .80 555 89.01 .18 606 89.93 .83 IL 4,636 87.60 0.52 4,331 88.36 .42 3,513 88.96 .63 IN 3,731 93.73 2.20 3,364 90.93 1.55 2,737 86.41 2.05 IA 1,621 88.71 .31 1,730 93.35 .52 1,682 97.03 .24 KS 2,541 97.21 2.13 2,463 98.13 1.22 2,521 98.61 .52 KY 2,308 93.28 3.73 2,415 94.70 2.28 2,131 95.96 1.13 LA 1,463 93.16 1.30 2,070 94.69 1.50 2,542 96.38 1.26 ME 466 95.06 .86 541 91.87 1.85 701 89.73 .14 MD 777 85.84 2.45 877 89.62 .68 1,371 87.75 .29 MA 2,167 94.60 1.43 2,382 95.55 .46 2,264 95.76 .66 MI 3,910 90.10 .33 3,611 87.04 .25 3,808 87.21 .39 MN 2,050 94.83 1.17 1,950 93.85 1.23 2,248 93.59 2.18 MS 1,312 79.12 3.43 1,209 83.46 1.99 886 89.16 2.60 MO 2,970 94.38 1.78 3,175 93.86 .72 3,035 95.91 .53 MT 581 91.57 .86 425 92.47 .94 458 83.62 1.09 NE 906 93.38 1.32 983 94.00 .20 1,009 94.85 .20 NV 466 87.34 .64 412 84.47 .24 384 86.20 1.04 NH 325 94.15 .62 354 95.48 .85 302 92.05 5.96 NJ 1,420 86.62 2.89 1,866 89.01 3.54 1,661 89.52 2.89 NM 411 88.08 7.54 364 92.58 4.40 416 94.47 3.61 NY 2,787 84.97 3.73 3,610 87.12 2.85 3,455 91.46 1.33 NC 2,302 90.49 1.00 2,471 89.88 .97 1,913 89.75 1.15 ND 338 84.62 .89 318 83.96 .63 363 85.40 0 OH 5,246 93.12 1.58 5,293 92.90 .89 4,876 93.05 .59 OK 1,693 92.79 .53 2,365 91.29 2.33 2,258 92.12 2.39 OR 852 93.08 4.11 901 93.67 2.33 834 93.88 .48 PA 3,370 93.03 1.31 3,227 93.28 .22 2,650 91.74 .19 RI 254 95.28 0 298 93.96 .34 378 96.56 0 SC 1,443 93.56 1.18 1,160 90.52 2.07 836 86.72 5.62 SD 435 90.11 0 606 88.78 1.32 544 87.87 1.65 TN 1,802 95.62 3.22 2,228 95.92 2.96 2,655 95.56 3.35 TX 6,788 87.09 2.42 7,152 86.45 2.15 7,187 87.27 1.43 UT 339 90.86 3.54 311 96.14 1.29 392 79.59 8.93 VT 134 93.28 0 143 82.52 0 115 82.61 0 VA 921 87.95 1.41 977 87.92 1.33 350 88.86 1.71 WA 2,418 88.88 1.49 2,275 88.00 1.32 1,940 86.91 1.19 WV 945 92.80 .32 1,047 92.65 .19 841 92.87 .24 WI 1,139 94.29 1.32 1,367 93.49 .80 1,062 96.33 .28 WY 247 90.28 2.83 376 93.88 1.06 292 94.52 2.74 Mean 1,932 90.62 1.67 2,034 90.58 1.36 1,919 89.99 1.44 SD 2,404 4.48 1.58 2,386 4.80 1.36 2,292 5.99 1.76 Source: CMS (2003 c) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 154 Change in C to G Level Deficiency Resolution over the Three-Year Period The states also improved in the timeliness of their resolution of non-actual harm deficiencies (C to G level) over the course of the study. State NH surveying agencies needed an average of 47.5 days (SD = 14.2) from the initial citation of a deficiency to its correction in Year 1. The average interval for C to G level deficiency resolution across the states declined to 43.4 days (SD =15.9) in Year 2 and to 40.5 days (SD = 13.7) in Year 3. In 40 states, resolution of non-actual harm deficiencies was faster in Year 3 than in Year 1 (see Table 36). In 10 states, resolution of C to G level deficiencies was slower in Year 3 than in Year 2. In two (Louisiana and Maine), the average time needed to correct these minor problems became progressively longer over the three-year period of observation, though still within the 60-day period mandated by CMS (see Table 36). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 155 Table 36: Average Number of Days for Deficiency Resolution by the States C to G Res. Y1 CtoGRes. Y2 C to G Res. Y3 H to L Res. Y1 H to L Res. Y2 H to L Res. Y3 1.CT (81.51) 2.NY (75.90) 3.AL (75.65) 4.PA (73.99) 5.NV (70.99) 6.NJ (68.47) 7.MT (67.97) 8.OK (67.06) 9.WY (65.69) 10.MO (58.64) 1.OK (76.38) 2.NY (76.30) 3.MT (75.47) 4.CT (68.91) 5.NV (66.95) 6.PA (66.72) 7.NJ (61.77) 8.WY (61.76) 9.NV (55.30) IQ.MO (54.47) 1.NH (69.73) 2.MT (68.44) 3.NY (64.21) 4.CT (60.06) 5.NJ (55.96) 6.WY (54.60) 7. AL (53.56) 8.LA (52.65) 9.VT (52.57) 10.HI (52.00) 1.NH (134.00) 2.PA (124.64) 3.DE (111.33) 4.NE (92.00) 5.NV (92.00) 6.MT (89.60) 7.NY (83.72) 8.VA (79.85) 9.NJ (76.93) 10. WY (75.00) l.ID (94.00) 2.NH (89.33) 3.NV (87.00) 4.WY (85.75) 5.PA (74.71) 6.OK (72.02) 7.NY (69.60) 8.VA (67.62) 9.CT (65.22) 10.CA (64.17) 1.AK (147.00) 2.NH (79.39) 3.RI (73.11) 4.NE (69.00) 5.MI (65.00) 6.CT (63.33) 7.MO (59.06) 8.CA (58.75) 9.NY (56.43) 10.NM (56.27) 11.HI (55.52) 12.MN (54.82) 13.NE (54.69) 14.WV (52.57) 15.MS (52.03) 16.SD (51.82) 17.UT (50.81) 18.DE (49.49) 19.NH (48.67) 20.CA (48.33) 11.MI (50.40) 12.LA (49.42) 13.SD (49.29) 14.IL (48.59) 15.MS (48.08) 16.AR (48.00) 17.NM (47.97) 18.HI (47.91) 19.NE (47.86) 20.GA (47.80) 1 l.SD (50.94) 12.MS (49.16) 13.MO (47.66) 14.0K (47.36) 15.MI (47.32) 16.NM (47.02) 17.0R (45.81) 18.NV (45.72) 19.PA (45.25) 20.IL (44.66) '11.OK (71.33) 12.IA (68.80) 13.CA (68.62) 14.AL (66.45) 15.IL (65.13) 16.WV (64.33) 17.ID(63.40) 18.MO (61.92) 19.CT (59.67) 20.WI (56.53) 11.UT (62.25) 12.NE (61.50) 13.MT (60.00) 14.MA (59.09) 15.MO (57.48) 16.DE (56.50) 17.RI (55.00) 18.IA (54.44) 19.NJ (53.94) 20.MI (52.89) ll.IL (55.00) 12.MT (54.40) 13.LA (54.09) 14.MS (50.83) 15.WA (50.35) 16.NV (48.00) 17.ID (46.60) 18.WI (46.33) 19.MA (45.67) 20. VA (45.17) 21.LA (47.02) 22.VA (46.84) 23.IL (46.46) 24.MI (46.13) 25.ID (46.08) 26.0R (45.27) 27.0H (44.52) 28.WI (43.90) 29.GA (43.49) 30. WA (43.48) 21.AZ (46.45) 22 ND (45.04) 23.OH (44.08) 24.1A (44.04) 25.MN (43.37) 26.0R (42.59) 27.UT (42.45) 28.WI (41.60) 29.FL (41.11) 30.CQ (41.11) 21.NE (44.54) 22.DE (44.07) 23.GA (43.29) 24.ND (42.59) 25.0H (41.69) 26.MN (40.51) 27.AZ (40.44) 28.VA (40.34) 29.UT (40.28) 30.WA (38.65) 21.MA (56.35) 22.MS (52.62) 23.LA (52.16) 24. AR (51.54) 25.UT (50.25) 26.MN (49.58) 27.MD (49.32) 28.MI (48.08) 29.0R 45.77 30.KY (44.36) 21.LA (52.29) 22.0R (51.76) 23.WV (51.00) 24.SD (50.00) 25.AZ (49.93) 26.AR (49.39) 27.MN (48.96) 28.ND (48.00) 29.AL (47.27) 30.IL (46.56) 21.WV (44.00) 22.0K (42.91) 23.MD (42.00) 24.UT (41.69) 25.MN (40.82) 26.OH (40.76) 27.1A (40.75) 28.PA (37.60) 29.OR (36.75) 30. CO (35.40) 31 .IA (42.42) 32.ND (42.12) 33.FL (41.52) 34.AZ (41.03) 35.AR (40.05) 36.MD (39.70) 37.TN (39.35) 38.KS (39.11) 39.TX (38.27) 31.WA (40.93) 32. VA (40.71) 33.ID (40.05) 34. WV (39.80) 35.DE (39.41) 36.TX (38.06) 37.CA (37.30) 38.TN (36.50) 39.NH (35.63) 31.FL (37.21) 32.WV (36.80) 33.ID (36.79) 34.TX (36.30) 35.RI (34.78) 36.WI (34.45) 37.CA (34.02) 38.IA (32.92) 39.ME (32.48) 31.KS (43.46) 32.WA (43.06) 33.NM (41.26) 34.OH (40.92) 35.NC (37.52) 36.IN (35.51) 37.GA (31.83) 38.ND (31.33) 39.FL (27.34) 31.CO (44.71) 32.WI (38.09) 33.MD (37.17) 34. WA (36.30) 35.MS (35.83) 36.OH (33.68) 37.KS (33.67) 38.GA (31.87) 39.FL (29.28) 31 .NJ (34.00) 32.GA (33.51) 33.IN (33.46) 34. AR (27.14) 35.KS (26.85) 36.FL (26.54) 37.TX (26.07) 38.SC (25.00) 39.DE (23.00) 41.NM (37.09) 41.KS (33.83) 41.SC (28.21) 41.TX(26.20) 41.NM (26.31) 41.AL (21.37) 42.KY (36.42) 42.VT (28.73) 42.MA (27.95) 42.SC (23.18) 42.TX (25.84) 42.NC (19.68) 43. VT (31.97) 43 .ME (27.51) 43. AR (27.67) 43.CO (19.25) 43.SC (24.67) 43.TN (17.49) 44.CO (30.19) 44.NC (25.13) 44.IN (27.20) 44.TN (17.93) 44.NC (18.04) 44 .ME (11.00) 45.SC (29.73) 45.IN (23.39) . 45.CO (27.09) 45 .ME (7.50) 45.ME (17.50) 45.KY (10.04) 46.IN (29.54) 46.KY (22.37) 46.KS (26.57) AKN/A 46.TN (16.71) AZ N/A 47.RI (27.52) 47.RI (17.29) 47.NC (17.93) HIN/A 47.KY (13.95) HI N/A 48.ME (26.15) 48.MA (15.76) 48.KY (15.31) RIN/A AKN/A . ND N/A 49.AK (25.22) 49.MD (11.95) 49.MD (11.54) SD N/A HIN/A SDN/A 50.NC (23.71) 50.AK (1.28) 50.AK (0.20) VTN/A VTN/A VTN/A Mean (47.53) Mean (43.44) Mean (40.50) Mean (56.85) Mean (49.57) Mean (44.08) SD (15.23) SD (15.92) SD (13.69) SD (27.22) SD (19.64) SD (22.57) Source: CMS (2003c) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 156 Resolution of H to L Level Deficiencies HtoL Level Deficiency Resolution in Each Year of Observation In Year 1, Maine (7.5 days) was the fastest in resolving H to L level deficiencies, followed by Tennessee (17.9 days) and Colorado (19.3 days). New Hampshire (134 days), Pennsylvania (124 days), and Delaware (111 days) were the slowest states. While New Hampshire and Delaware had very few H to L level deficiencies, the same is not true of Pennsylvania, which had 44 citations at the H level and above. Overall, only seven states with deficiencies at H level and above resolved them within the 30-day grace period allowed by CMS in Year 1 (see Table 36). In Year 2, Kentucky (14.0 days), Tennessee (16.7 days), and Maine (17.5 days) were the fastest states in correcting actual harm deficiencies. The slowest states in resolving H to L level deficiencies were Idaho (94.0 days), New Hampshire (89.3 days), and Nevada (87.0 days). However, there were only five actual harm deficiencies combined in three slowest states. On average, nine states resolved H to L level deficiencies within the 30-day grace period in Year 2, as opposed to seven in Year 1 (see Table 36). Kentucky (10.0 days), Maine (11.0 days), and Tennessee (17.5 days) were the fastest states in resolving actual harm deficiencies in Year 3. Alaska was the slowest, with an average of 147 days (both at the same NH), followed by New Hampshire (79.4 days) and South Dakota (73.1 days). As in the previous year, nine states corrected these problems, on average, within the time frame specified by CMS. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 157 In 36 states, the average interval for resolution of H to L level deficiencies exceeded the 30-day grace period during Year 3 (see Table 36). Change in H to L Level Deficiency Resolution over the Three-Year Period As with non-actual harm deficiencies, the states became faster in resolving actual harm (H to L level) deficiencies over the three-year period of observation. The average interval among states for correcting deficiencies at the H level and above decreased from 56.8 days (SD - 27.2) in Year 1 to 49.6 days (SD = 19.6) in Year 2 and 44.1 days (SD = 22.6) in Year 3 (see Table 36). However, this was still considerably longer than the 30 days allowed under CMS regulations (USGAO 2000, 2003). In fact, it took the states longer to resolve H to L level deficiencies than to correct C to G level problems, despite the shorter grace period allowed for the former. Overall, 34 states resolved H to L deficiencies more quickly as the study progressed, while 12 states became slower. In two states (Alaska and Rhode Island), deficiencies at the H to L level were cited in only one year, while in two others (Hawaii and Vermont), no citations at the H to L level were issued during the entire three-year period (see Table 36). Summary o f Resolution o f Nursing Home Deficiencies by States Correlation analyses revealed strong intravariable relationships among resolution of C to G level deficiencies and of H to L level deficiencies across the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 5 8 three years of observation in this study. In Years 1 and 2, intervariable correlations were strong as well. The time it took states to correct minor problems corresponded with the length of time required to resolve major issues. However, in Year 3, the resolution of these two categories of deficiencies was unrelated (see Table 37). Table 37: Correlations among Dependent Variables, Deficiency Resolution Variable la lb lc 2a 2b 2c la. C-G Resolution (Year 1) .848" .703" .645" .634" .072 lb. C-G Resolution (Year 2) — .775" .415" .468" -.089 lc. C-G Resolution (Year 3) — .553" .569" .062 2a. H-L Resolution (Year 1) — .750” .476” 2b. H-L Resolution (Year 2) — .524" 2c. H-L Resolution (Year 3) — JV=50; ><.01 During these three years, patterns emerged among the states in their compliance with the CMS mandate for swift deficiency resolution. Two states, Kentucky and Maine, were particularly successful in quickly correcting both minor and major problems in accordance with CMS guidelines. Alaska, although the fastest in dealing with issues that pose the potential for harm to NH residents, was slow in resolving its few cases of actual harm. New Hampshire was not timely in handling either set of problems during this three-year period. Summary of the Effectiveness of State Nursing Home Regulation Correlation analyses revealed that compliance of the states with the CMS mandate to avoid repetition in scheduling annual inspections was unrelated to state compliance with the CMS mandate to resolve NH deficiencies by the end of the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 159 prescribed grace period (see Table 38). Two different groups of states led the nation in successful implementation of these two initiatives. Similarly, states that had the most repetitive annual NH inspections were not necessarily the states with the slowest resolution of NH deficiencies. Based on these disparate findings, it is difficult to determine which states are the “most” and “least” effective in terms of implementing federal NH initiatives. Table 38: Correlations among Dependent Variables, Effectiveness Variable la lb lc 2a 2b 2c 3a 3b 3c la. Repetitive State Inspections (Yl-2) — .587“ .866“ -.010 -.060 -.090 .212 .107 .029 lb. Repetitive State Inspections (Y2-3) .913“ -.039 -.-077 -.084 .071 .056 -.023 1 c. Repetitive State Inspections (Yl-3) -- -.029 -.080 -.101 .153 .089 .003 2a. C-G Resolution (Yl) 2b. C-G Resolution (Y2) 2c. C-G Resolution (Y3) 3a. H-L Resolution (Yl) 3b. H-L Resolution (Y2) 3c. H-L Resolution (Y3) —------:— n-------------- .848“ .703“ .775“ .645“ .415“ .553“ .634" .468“ .569“ .750“ .072 -.089 .062 .476" .524“ N = 50; < .01 Summary of the Extent and Effectiveness of State Nursing Home Regulation Correlation analyses among the dependent variables revealed no significant relationships between the extent and the effectiveness of the regulatory activities of the 50 state NH surveying agencies (see Table 39). During the three years of this study, the volume and severity of state deficiency citations were unrelated to state implementation of federal initiatives to avoid repetitive scheduling and to quickly resolve deficiencies. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 160 _____________ Table 39: Correlations among Dependent Variables_________ Variable la lb lc 2a 2b 2c 3a 3b 3c “ I I ~ .888** .764** -7029 7229 7080 Tl70 -.111 -.154 Volume (Yi) lb. — .846 -.015 -.202 -.084 -.186 -.154 -.188 Volume (Y2) lc. — -.140 -.296 -.201 -.273 -.219 -.273 Volume (Y3)____________________________________________ _________________________ 2a. ::: .874** .564** 7l45 7041 .048 Severity 2b. — .638 .175 -.092 .031 Severity (Y2) 2c. — -.113 -.267 -.222 Severity 0*3)_____________________________________________________________________ 3 T = 587" .'866^ Repetitive Inspections (Y1-2) 3b. — .913 Repetitive Inspections 0 *2-3) 3c. Repetitive Inspections 0*1-3)_____________________________________________________________________________ 4a. C-G Resolution 0 *1 ) 4b. C-G Resolution (Y2) 4c. C-G Resolution 0*3)_________________________________________________; __________________ __________ 5a. H-L Resolution (Yl) 5b. H-L Resolution 0 *2) 5c. H-L Resolution 0*3) _ __________________________________________ N = 50; p < .05; p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 161 Table 39: Correlations among Dependent Variables (Continued) Variable 4a 4b 4c 5a 5b 5c la. -.078 -.115 -.232 -.077 .007 .101 Volume (Yl) lb. -.067 -.091 -.229 -.157 .037 .094 Volume (Y2) lc. -.073 -.059 -.168 -.298* -.025 -.036 Volume (Y3) 2a. .124 .008 .078 .209 .168 -.027 Severity (Yl) 2b. .079 .071 .162 .078 .005 .060 Severity (Y2) 2c. -.016 -.117 .035 .060 .099 .247 Severity (Y3) ... 3 a. -.010 -.060 -.090 .212 .107 .029 Repetitive Inspections (Yl-2) 3b. -.039 -.-077 -.084 .071 .056 -.023 Repetitive Inspections (Y2-3) 3c. -.029 -.080 -.101 .153 .089 .003 Repetitive Inspections (Yl-3) 4a. C-G .848" .703** .645’* .634" .072 Resolution (Yl) 4b. C-G .775" .415" .468" i © o o Resolution (Y2) 4c. C-G .553" .569" .062 Resolution (Y3) 5a. H-L -- .750" .476 Resolution (Yl) 5b. H-L .524 Resolution (Y2) 5c. H-L Resolution (Y3) _ _★ _ _ * * __ iV= 50; *p < ,05; * * p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 162 Factors Predicting State Nursing Home Regulation For each outcome measure of state enforcement, sets of determinants related to the states’: 1) external characteristics, 2) NH surveying agencies, 3) legislatures, 4) interest groups (NH industry and aging advocacy groups) and 5) governors were tested. In this section, combined and individual effects of determinants on: 1) outcome measures (in each year) and 2) change in the outcome measures (over the three-year period of the study) are described. Factors found to have significant effects on a particular outcome measure were included in theoretical models. External Determinants The outcome measures of the extent and the effectiveness of state NH regulation were regressed onto four external determinants (urbanization, education, proportion of oldest-old and political culture) for each of the three annual inspection periods analyzed in this study. Correlation analyses revealed three significant relationships between external determinants and outcome measures. First, there was a strong negative correlation between education and volume of deficiency citations in Year 1 (see Table 40a) and Year 2 (see Table 40b) and a moderate negative correlation in Year 3 (see Table 40c). Second, education was positively correlated with the severity of deficiency citations in all three years of the study, particularly Year 1 (see Table 40a) and Year 3 (see Table 40c). Third, the proportion of oldest- old negatively corresponded with citation volume in Year 1 (see Table 40a). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 163 Table 40a: Correlations among External Determinants and Outcome Measures, Year 1 Variable 1 2 3 4 5 6 7 8 1 . Urbanization — .494** -.217 -.049 -.174 .125 .182 .082 2. Education -.058 -.436** -.409** .313* -.012 .112 3. Proportion of Oldest-Old — -.265 -.313* -.035 .165 .083 4. Political Culture — .058 -.121 .006 -.105 5. Volume Deficiency Citations — -.029 -.078 -.077 6. Severity Deficiency Citations . . . .124 .209 7. C-G Deficiency Resolution . . . .645 8. H-L Deficiency Resolution . . . V=50;><.05; *><.01 Table 40b: Correlations among External Determinants and Outcome Measures, Year 2 Variable 1 2 3 4 5 6 7 8 9 1. Urbanization — .494** -.217 -.049 -.186 .125 .058 .055 .201 2. Education — -.058 i 45* 0\ -.394“ .379“ .063 -.142 ,188 3. Proportion Oldest-Old — -.265 -.272* .086 .205 .234 .005 4. Political Culture — .103 -.096 .013 .055 -.277 5. Volume Citations — -.202 -.186 -.091 .037 6. Severity Citations 7. Repetitive Scheduling -- .175 .071 -.060 .005 .107 Annual Inspections 8. C-G Deficiency — .468** Resolution 9. H-L Deficiency Resolution ________________________________________________________________________ 7V = 50; > < .05;’><.01 Table 40c: Correlations among External Determinants and Outcome Measures, Year 3 Variable 1 2 3 4 5 6 7 8 9 1. Urbanization — .494“ -.217 -.049 -.116 -.044 .161 .016 .089 2. Education — -.058 -.436** -.320* .353* .112 .082 .207 3. Proportion Oldest-Old — -.265 -.154 -.144 .062 .267 -.240 4. Political Culture — .042 -.192 -.124 -.127 -.211 5. Volume Citations — -.201 -.219 -.168 -.036 6. Severity Citations — -.267 .035 .247 7. Repetitive Scheduling — -.084 -.023 Annual Inspections 8. C-G Deficiency — .062 Resolution 9. H-L Deficiency Resolution _______ __________________ _ _____ _ N = 50; > <.05; *><.01 ~ ~ ~~~~~~~ R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 164 Extent of State Nursing Home Regulation . Extent in Each Year o f Observation Multiple regression revealed that the combined effect of the four external determinants on the volume of state NH deficiency citations was highly significant in Year 1 (F = 5.69, p < .005) and Year 2(F = 4.28, p < .005) and moderately significant in Year 3 (F= 2.99, p < .05). Several individual effects were also significant during the three years of observation (see Table 41a). Education had a strong negative effect on the volume of state deficiency citations in all three years of the study. The proportion of oldest-old had a negative effect on state citation volume in all three years of observation, particularly in Year 1. Political culture had moderate negative effect in Years 1 and 3; a moralistic subculture predicted a higher volume of citations. The combined effect of the four external determinants on the severity of state NH deficiency citations was non-significant in Year 1, with only education having a modest individual effect. However, the combined effect of external measures was significant in Year 2 (F = 3.20,p < .05) and in Year 3 (F=3.22,p < .05). During Year 2, one variable (education) had a strong positive effect on the proportion of state deficiency citations at the G level and above. Education also had a moderate positive effect on the severity of citations in Year 3 and urbanization had a slight negative effect that year (see Table 42a). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 165 Change in Extent over the Three-Year Period External determinants did not have a significant combined effect on change in the volume (see Table 41b) or severity (see Table 42b) of state NH deficiency citations. One variable (proportion of oldest-old) had a modest negative effect on change in state citation volume. A higher proportion of state residents age 85 and older led to a decrease in deficiency citations per NH bed from Year 1 to Year 3 (see Table 41b). Table 41a: Multiple Regression of Volume of State Deficiency Citations on External Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 18.2 17.0 15.3 Urbanization .000 -.005 .010 Education -.297**" -.257**** -.261**** Proportion of Oldest- -2.68**" -2.28** -1.57* Old Political Culture -.251** -.173 -.267** Adjusted R2 .289 .211 .139 F ___ * 4 * * 5.69 4.28**** 2.99** N 50 50 50 * . * * _ _ * * * SKI** ^ ^ _ p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 41b: Multiple Regression of Change in Volume of State Citations on External Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -1.21 -1.72 -2.93 Urbanization -.004 .014 .010 Education .039 -.004 .036 Proportion of Oldest- .402 .705 1.11* Old Political Culture .077 -.094 -.017 Adjusted R2 -.048 .043 .009 F .434 1.55 1.11 N 50 50 50 * „ * * ^ * * * p<.\\ p<. 05; p < .01; ***><.005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 166 Table 42a: Multiple Regression of Severity of State Deficiency Citations on External Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 3.13 -1.19 6.30 Urbanization -.015 -.059* -.082*’ Education .364* .452 .421*** Proportion of Oldest- -.244 .996 -2.094 Old Political Culture .037 .216 -.071 Adjusted R2 .020 .152 .154 F 1.25 3.20** 3.23” N 50 50 50 * * * ^ _ # * * * * * * _ _ _ p < . 1; p < .05; p < .01; p< .005 Unstandardized Beta coefficients Table 42b: Multiple Regression of Change in Severity of State Citations on External Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -4.32 -7.49 3.17 Urbanization -.044 -.023 -.068 Education .088 -.031 ..057 Proportion of Oldest- 1.24 -3.09** -1.85 Old Political Culture .179 -.287 -.108 Adjusted R2 -.016 .063 -.017 F .807 1.82 .800 N 50 50 50 * * * „ _ # * * _ „ * * * * ^ ^ _ p<.l\ p < .05; /?<.01; p < .005 Unstandardized Beta coefficients Effectiveness of State Nursing Home Regulation Effectiveness in Each Year o f Observation In this study, effectiveness of state NH regulation was assessed through the repetitive scheduling of annual state inspections and the timeliness in which states resolve non-actual harm (C to G level) and actual harm (H to L level) deficiencies. As shown in Table 43 a, multiple regression revealed that neither the combined nor individual effects of external determinants on repetitive scheduling were significant, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 167 with one exception. A higher proportion of oldest-old led to more repetitive scheduling in Year 2. Multiple regression also revealed that the combined effect of the four external determinants on state resolution of C to G level deficiencies was non-significant in all three years of the study. Urbanization (Year 1) and proportion of oldest-old (Years 2 and 3) had modest positive effects on the time needed to resolve C to G deficiencies (see Table 44a). This indicated that greater urbanization and a higher proportion of a state’s residents age 85 and over led to slower resolution of minor NH problems. For the most part, external determinants had no significant effects, combined or individually, on the resolution of H to L level deficiencies by the states in Years 1, 2 or 3. The exception was the proportion of oldest-old, which had a modest positive effect on the correction of major NH problems in Year 3 (see Table 45a). Change in Effectiveness over the Three-Year Period External determinants did not affect whether the scheduling of a state’s annual NH inspections was more or less repetitive in Year 3 than it was in Year 2 (see Table 43b). The combined effects of external determinants on change in the time needed to resolve non-actual harm (see Table 44b) and actual harm (see Table 45b) deficiencies from year to year were also non-significant. However, two individual determinants had significant effects on change in the resolution of C to G level deficiencies. Greater urbanization led to faster state R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 168 correction of non-actual harm problems in Year 3, as compared to Years 2 and 1. In contrast, a higher level of education led to slower resolution of these deficiencies in Year 2 than in the previous year (see Table 45b). Table 43a: Multiple Regression of Repetitive State Inspection Scheduling on External Determinants Independent Variable Year 1 Year 2 Year 3 Intercept -6.12 6.66 Urbanization .043 .129 Education .202 -.045 Proportion of Oldest- 6.46* 2.17 Old Political Culture .420 -.409 Adjusted R2 -.018 -.041 F .788 .522 N 50 50 * * ** _ * * * _ . * * * * p<. 1; p < .05; p<- 01; p < .005 Unstandardized Beta coefficients Table 43b: Multiple Regression of Change in Repetitive Inspection Scheduling on External Determinants Independent Variable Year 1 to Year 22 Year 2 to Year 3 Year 1 to Year 33 Intercept 12.8 Urbanization .086 Education -.248 Proportion of Oldest- -4.29 Old Political Culture -.829 Adjusted R2 -.011 F .871 N 50 * „ * * _ _ * * * _ . * * * * * A /,. p<.\\ p<. 05; p <.01; p < .005 Unstandardized Beta coefficients 1 Predictability of annual state inspections was not assessed in the first year of the study (1999-2000). 2 Comparison between the first and second years of the study (1999-2000 and 2000-2001) was unavailable. 3 Comparison between the first and third years of the study (1999-2000 and 2001-2002) was unavailable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 169 Table 44a: Multiple Regression of State Resolution of C-G Deficiencies on External Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 23.4 24.6 17.1 Urbanization .287* .241 .038 Education -.459 -.806 .214 Proportion of Oldest- 9.04 12.6* 10.5* Old Political Culture .111 .273 -.121 Adjusted R2 .013 .033 .001 F 1.16 1.42 1.01 N 50 50 50 V « ! | XXB XXXX p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 44b: Multiple Regression of Change in C-G Deficiency Resolution on External Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept 1.23 -7.50 -6.27 Urbanization -.046 -.203* -.248** Education -.348 1.02** .673 Proportion of Oldest- 3.53 -2.11 1.42 Old Political Culture .162 -.394 -.232 Adjusted R2 .008 .118 .045 F 1.09 2.63 1.58 N 50 50 50 p<. 1; p<. 05; p< .01 ;***><.005 Unstandardized Beta coefficients Table 45a: Multiple Regression of State Resolution of H-L Deficiencies on External Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 29.0 48.2 76.7 Urbanization .131 .241 -.189 Education .353 -.154 .814 Proportion of Oldest- 7.95* -1.71 -19.1 Old Political Culture -.441 -1.98 -1.88 Adjusted R2 -.073 .020 .055 F .250 1.23 1.64 N 50 50 50 p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 7 0 Table 45b: Multiple Regression of Change in H-L Deficiency Resolution on External Determinants Independent Variable Intercept Urbanization Education Proportion of Oldest- Old Political Culture Year 1 to Year 2 17.1 .143 -.531 -9.61 -1.50 Year 2 to Year 3 -29.0 -.104 .532 7.67 .856 Year 1 to Year 3 -6.69 .030 .152 -8.94 -.508 Adjusted R2 -.030 .069 -.085 F .682 .304 .176 N 50 50 50 p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Agency Determinants The outcome measure of the extent and the effectiveness of state NH regulatory activity was regressed onto four agency determinants (state proportion of total state agency funding, total state agency funding per NH bed and average number of years of experience of state surveyors) for each year of the study. Four correlations were revealed (see Tables 46a-c). First, total state survey agency funding was strongly correlated with citation volume in Years 1, 2 and 3. Second, average number of years of surveyor experience was negatively correlated with citation volume in all three years. Third, surveyor experience corresponded with slower state resolution of C to G level deficiencies in Years 1 and 2. Fourth, total state surveying agency funding per NH bed corresponded with slower resolution of H to L level deficiencies in Year 3. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 171 Table 46a: Correlations among Agency Determinants and Outcome Measures, Year 1 Variable 1 2 3 4 5 6 7 1 . State % of Total Funding — -.063 .429" -.110 .085 .200 .050 2. Total Funding/NH Bed — -.035 .496** .054 -.214 -.146 3. Experience of Surveyors — -.362* -.145 .365* .292 4. Volume Deficiency Citations — -.029 -.078 -.077 5. Severity Deficiency Citations . . . .124 .209 6. C-G Deficiency Resolution . . . .645** 7. H-L Deficiency Resolution — N= 50; *p <05; p < .01 Table 46b: Correlations among Agency Determinants and Outcome Measures, Year 2 Variable 1 2 3 4 5 6 7 8 1 . State % of Total Funding -.072 .344* -.093 -.213 -.093 .142 .034 2. Total Funding/NH Bed -.085 .438" .087 -.089 -.275 .067 3. Experience of Surveyors — -.340* .075 -.176 .325* .135 4. Volume Deficiency Citations — -.202 -.186 -.091 .037 5. Severity Deficiency Citations — .175 .071 .005 6. Repetitive Scheduling of State ... -.060 .107 Inspections 7. C-G Deficiency Resolution ... .468" 8. H-L Deficiency Resolution .... . " " ..* { £ . . . ..... .......... ............ ... IV=50; p <.05; pc.O l Table 46c: Correlations among Agency Determinants and Outcome Measures, Year 3 Variable 1 2 3 4 5 6 7 8 1 . State % of Total Funding -.112 .301 -.123 -.037 -.069 -.034 .053 2. Total Funding/NH Bed -.072 .374" .093 -.020 -.244 .423** 3. Experience of Surveyors . . . -.346* .128 -.262 .253 .283 4. Volume Deficiency Citations 5. Severity Deficiency Citations 6. Repetitive Scheduling of State Inspections 7. C-G Deficiency Resolution 8. H-L Deficiency Resolution -.201 -.219 -.267 -.168 .035 -.084 -.036 .247 -.023 .062 T V =50; p <.05; p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 7 2 Extent of State Nursing Home Regulation Extent in Each Year o f Observation Multiple regression revealed that the combined effect of the three state surveying agency determinants on the volume of state NH deficiency citations was significant in Year 1 (F= 7.04, p < .005), Year 2 (F = 4.90,p < .01) and Year 3 (F = 3.78,/? < .05). As Table 47a demonstrates, the positive effect of total state agency funding per NH bed (Years 1, 2 and 3) and the negative effect of average number of years of experience of state surveyors (Years 1, 2 and 3) on citation volume were significant. The combined effect of the three agency determinants on the severity of state NH deficiency citations was non-significant in all three years of observation, as were the individual effects of these three variables, with the exception of state proportion of total agency funding, which had a negative effect on citation severity in Year 2 (see Table 48a). Change in Extent over the Three-Year Period Agency determinants did not have significant combined or individual effects on change in citation volume over the three years of observation (see Table 47b). However, as Table 48b demonstrates, the combined effect of agency determinants on change in state citation severity from Year 1 to Year 3 was significant (F - 2.85, p < .1). An increase in the state share of total agency funding led to increased severity in R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. citations from Year 1 to Year 3. More experienced state surveyors also predicted increased severity (see Table 48b). Table 47a: Multiple Regression of Volume of State Deficiency Citations on Agency Determinants Independent Variable Intercept State % of Total Funding Total Funding/NH Bed Average Number of Years of Experience of State Surveyors Year 1 4.90 .002 .012” * * -.222** Year 2 5.64 -.016 .010*** -.191* Year 3 5.72 -.011 .008** -.209** Adjusted R2 .312 .226 .172 F 7.04 4.90*** 3.78** N 50 50 50 p< - 1; p<- 05; p<. 01; p < .005 Unstandardized Beta coefficients Table 47b: Multiple Regression of Change in Volume of Citations on Agency Determinants Independent Variable Intercept Change in State % of Total Funding Change in Total Funding/NH Bed Average Number of Years of Experience of State Surveyors Year 1 to Year 2 .162 -.055 .008 .020 Year 2 to Year 3 -.187 .001 -.001 -.008 Year 1 to Year 3 .025 -.063 .004 -.002 Adjusted R2 .001 -.080 -.047 F 1.02 .011 .402 N 50 50 50 p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 174 Table 48a: Multiple Regression of Severity of State Deficiency Citations on Agency Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 8.86 9.90 6.70 State % of Total .195 -.207* -.119 Funding Total Funding/NH Bed .003 .002 .003 Average Number of -.287 .171 .180 Years of Experience of State Surveyors Adjusted R2 -.032 .009 -.040 F .583 1.12 .487 N 50 50 50 * . * * _ _ * * * _ - * * * * ^ _ p<.l; p<. 05; jo < .01; p < .005 Unstandardized Beta coefficients Table 48b: Multiple Regression of Change in Severity of State Citations on Agency Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -3.86 -1.83 -5.52 Change in State % of .087 .152 .262* Total Funding Change in Total -.018 .001 -.016 Funding/NH Bed Average Number of .244* .081 .350* Years of Experience of State Surveyors Adjusted R2 .059 -.056 .122 F 1.84 .298 2.85* N 50 50 50 * * * _ _ * * * _ ^ p<A\ p<. 05; p<.01 ; ***> < .005 Unstandardized Beta coefficients Effectiveness of State Nursing Home Regulation Effectiveness in Each Year o f Observation The combined effect of agency determinants on whether state inspections were scheduled during the same month two years in a row was non-significant, (as were the effects of the three individual agency measures), in Years 2 and 3, the two R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 175 years in which it was possible to observe repetitive state inspection scheduling in this study (see Table 49a). In contrast, as Table 50 demonstrates, the combined effect of agency determinants on state resolution of C to G level deficiencies was significant in Year 1 (F = 2 M ,p < .05), Year 2 (F -3.24,p < .05) and Year 3 (F=2.28,p < .1). Greater total state agency funding resulted in faster resolution of non-actual harm deficiencies in Years 2 and 3, while more experienced state surveyors resulted in slower resolution in Years 1 and 2 (see Table 50a). Finally, the combined effect of agency determinants on state resolution of H to L level deficiencies was significant in Year 3 (F= 5.34,p < .005). Greater total state agency funding and more experienced state surveyors both led to slower resolution of actual harm deficiencies in Year 3 (see Table 51a). Change in Effectiveness over the Three-Year Period Agency determinants did not lead to a change from Year 2 to Year 3 in repetitive scheduling of state inspections (see Table 49b) or in the timeliness of resolution of actual harm deficiencies (see Table 5 lb), as neither the combined nor the individual effects of agency determinants on these outcome measures were significant. In contrast, agency determinants did have a significant combined effect (F = 2.61,p < .1) on change in the resolution of C to G level deficiencies from Year 2 to R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 76 Year 3. In particular, an increase in total agency funding from Year 2 to Year 3 led to faster resolution of non-actual harm deficiencies (see Table 50b). Table 49a: Multiple Regression of Repetitive State Inspection Scheduling _______________ on Agency Determinants________________ Independent Variable Year l 4 Year 2 Year 3 Intercept 23.0 23.9 State % of Total -.164 .017 Funding Total Funding/NH Bed -.017 -.009 Average Number of -.403 -.803 Years of Experience of State Surveyors Adjusted R2 -.010 .001 F .874 1.01 N 50 50 * ** ^ _ * * * „ _ -ft*** _____ p<. 1; p<. 05; p <. 01; p < .005 Unstandardized Beta coefficients Table 49b: Multiple Regression of Change in Repetitive Inspection Scheduling ________________________ on Agency Determinants__________________ ___ Independent Variable Year 1 to Year 2s Intercept Change in State % of Total Funding Change in Total Funding/NH Bed Average Number of Years of Experience of State Surveyors Year 2 to Year 3 5.10 -.091 -.015 -.346 Year 1 to Year 3^ Adjusted R5 -.061 F .234 N * # « > _ # * * _ „ * * * * __ p < .1; p < .05; p<. 01; p < .005 50 Unstandardized Beta coefficients 4 Predictability of annual state inspections was not assessed in the first year of the study (1999-2000). 5 Comparison between first and second years of the study (1999-2000 and 2000-2001) was unavailable. 6 Comparision between first and third years of the study (1999-2000 and 2001-2002) was unavailable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 177 Table 50a: Multiple Regression of Resolution of C-G Deficiencies on Agency Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 42.4 44.8 48.3 State % of Total .178 .104 -.369 Funding Total Funding/NH Bed -.038 -.053” -.040* Average Number of 1.42” 1.25* 1.01 Years of Experience of State Surveyors Adjusted R2 .122 .144 .088 F 2.86** 3.24** 2.28* N 50 50 50 * * * _ _ * * * _ „ * * * * _ _ _ p<. 1; p < .05; p < . 01; p><.005 Unstandardized Beta coefficients Table 50b: Multiple Regression of Change in C-G Deficiency Resolution on Agency Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -4.41 1.37 -2.89 Change in State % of .105 .933 .506 Total Funding Change in Total .044 -.127* -.018 Funding/NH Bed Average Number of -.055 -.306 -.512 Years of Experience of State Surveyors Adjusted R2 -.034 .107 -.006 F .556 2.61* .914 N 50 50 50 > < .1 ; p < .05; p < .01; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 178 Table 51a: Multiple Regression of Resolution of H-L Deficiencies on Agency Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 58.9 40.8 -3.64 State % of Total -.455 -.065 .340 Funding Total Funding/NH Bed -.052 .023 .117**** Average Number of 2.40 .918 2.08** Years of Experience of State Surveyors Adjusted R2 .025 -.062 .266 F 1.32 .279 5.34**** N 50 50 50 *p < .1; *'p < .05; * * p < .01; * * * * /? < .005 Unstandardized Beta coefficients Table 51b: Multiple Regression of Change in H-L Deficiency Resolution on Agency Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept 4.88 -12.2 -17.4 Change in State % of -.187 1.93 .815 Total Funding Change in Total -.078 -.102 .049 Funding/NH Bed Average Number of -1.63* .816 -.462 Years of Experience of State Surveyors Adjusted R2 .032 -.004 -.047 F 1.41 .953 .472 N 50 50 50 * * * - _ # * * * * * * s \ m p < . \ ; p < .05; jP<.01; p < .005 Unstandardized Beta coefficients Legislative Determinants Measures for extent and effectiveness of state NH regulation were regressed onto three legislative determinants (political control of the state legislature, legislative professionalism and presence of an aging or LTC committee) for each year of observation. There were no significant correlations between the five outcome measures and the three state legislature variables (see Tables 52a-c). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 179 Table 52a: Correlations among Legislative Determinants and Outcome ________________________Measures, Year 1________________________ Variable 1 2 3 4 5 6 7 1 . Political Control of Legislature — -.025 .170 -.107 -.158 -.009 -.053 (Democratic) 2. Legislative Professionalism — .538** -.045 -.067 .161 .095 3. Aging/LTC Committee — -.131 .049 .114 .140 4. Volume Deficiency Citations . . . -.029 -.078 -.077 5. Severity Deficiency Citations . . . .124 .209 6. C-G Deficiency Resolution . . . .645“ 7. H-L Deficiency Resolution — N= 50; ><.05; *><.01 Table 52b: Correlations among Legislative Determinants and Outcome _______ Measures, Year 2_______________________ Variable 1 2 3 4 5 6 7 8 1 . Political Control of .019 .086 -.019 -.063 -.139 -.061 -.261 Legislature (Democratic) 2. Legislative Professionalism .538“ -.069 -.116 .027 .103 .093 3. Aging/LTC Committee -.153 .043 .033 .130 .052 4. Volume Deficiency Citations — -.202 -.186 -.091 .037 5. Severity Deficiency Citations . . . .175 .071 .005 6. Repetitive Scheduling of State . . . -.060 .107 Inspections 7. C-G Deficiency Resolution . . . .468“ 8. H-L Deficiency Resolution — N = 50; 'p <.05; "p < .01 Table 52c: Correlations among Legislative Determinants and Outcome Measures, Year 3 ___________________ Variable 1 2 3 4 5 6 7 8 1 . Political Control of .032 .086 .134 -.202 -.088 -.095 -.204 Legislature (Democratic) 2. Legislative Professionalism - .538“ -.068 -.156 .162 -.005 .217 3. Aging/LTC Committee — -.199 .122 -.113 .107 .123 4. Volume Deficiency Citations — -.201 -.219 -.168 -.036 5. Severity Deficiency Citations — -.267 .035 .247 6. Repetitive Scheduling of State . . . -.084 -.023 Inspections 7. C-G Deficiency Resolution . . . .062 8. H-L Deficiency Resolution -----------------— -----S----------- — ----ijrg------------- — ------------------------------------------------- . . . N= 50; 'p <.05; p<.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 180 Extent of State Nursing Home Regulation Extent in Each Year o f Observation The combined effect of the three legislative determinants on the volume of state NH deficiency citations was non-significant in all three years of observation, as were the individual effects of the three legislative measures (see Table 53a). Legislative measures also did not have a combined effect on citation severity in Years 1, 2 or 3. However, two individual effects were significant in Year 3. Greater legislative professionalism led to lower citation severity, while the presence of an aging or LTC committee led to higher proportion of severe citations (see Table 54a). Change in Extent over the Three-Year Period The combined effect of legislative determinants on change in citation volume from year to year in this study was non-significant. A change in political control of the legislature (from Republican to Democratic) led to an increase in the number of citations per 100 NH beds from Year 1 to Year 2 and from Year 2 to Year 3 (see Table 53b). Legislative determinants did not affect change in citation severity over the three-year period, either collectively or individually (see Table 54b). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 181 Table 53a: Multiple Regression of Volume of State Deficiency Citations on Legislative Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 5.90 6.36 6.18 Political Control of State -.219 -.012 .398 Legislature (Democratic) Legislative .486 .413 1.07 Professionalism Aging/LTC Committee -.659 -.815 -1.15 Adjusted R2 -.038 -.040 .004 F .408 .390 1.06 N 50 50 50 * * * _* * * ~ „ * * * * ___ p<.l; p<. 05; p<.01; p < .005 Unstandardized Beta coefficients Table 53b: Multiple Regression of Change in Volume of State Citations on Legislative Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept .589 -.236 .378 Change in Political .740* 1.25* .760 Control of State Legislature (to Democratic) Legislative -.982 .440 -.499 Professionalism Aging/LTC Committee .154 -.241 -.088 Adjusted R2 .010 .004 -.014 F 1.15 1.06 .772 N 50 50 50 tf _ Jit* * * „ „ _ p<- 1; p < .05; p<.01; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 182 Table 54a: Multiple Regression of Severity of State Deficiency Citations on Legislative Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 11.1 8.55 7.87 Political Control of State -1.02 -.280 -.965 Legislature (Democratic) Legislative -6.24 -5.23 -9.65* Professionalism Aging/LTC Committee 1.75 1.06 2.43* Adjusted RJ -.011 -.031 .071 F .831 .523 2.23 N 50 50 50 * *# — * * * ~ _ _ p<. 1; p<. 05; /?<.01; p < .005 Unstandardized Beta coefficients Table 54b: Multiple Regression of Change in Severity of State Citations on Legislative Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -2.47 -.580 -3.30 Change in Political -.228 -2.50 -1.66 Control of State Legislature (to Democratic) Legislative .116 -4.05 -2.76 Professionalism Aging/LTC Committee -.334 1.22 .499 Adjusted R2 -.064 .010 -.017 F .041 1.17 .739 N 50 50 50 £ > < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Effectiveness of State Nursing Home Regulation Effectiveness in Each Year o f Observation In this study, legislative determinants did not affect state implementation of two CMS mandates designed to improve the effectiveness of NH regulation. Multiple regression revealed that the combined effect of the three legislative R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 183 determinants on repetitive scheduling of state inspections was non-significant, as were individual effects, with two exceptions. In Year 3, greater legislative professionalism increased repetitive scheduling, while the presence of an aging or LTC committee decreased repetition (see Table 55a). The combined effect of legislative determinants on actual harm (see Table 56a) and non-actual harm deficiency resolution (see Table 57a) were non-significant, as were the effects of individual measures, with one exception. Democratic control of the state legislature led to faster resolution of actual harm deficiencies in Year 2 (see Table 57a). Change in Effectiveness over the Three-Year Period The combined effect of legislative determinants on whether repetitive scheduling of state inspections increased or decreased from Year 2 to Year 3 was non-significant. Two individual determinants affected this outcome measure, albeit in opposite directions. Greater legislative professionalism was associated with an increase in repetitive scheduling of state inspections from Year 2 to Year 3. The presence of an aging or LTC committee in the state legislature led to a decrease in repetition from Year 2 to Year 3 (see Table 55b). Legislative measures did not predict change in the timeliness of C to G deficiency resolution (see Table 56b). However, the combined effect of these determinants on change in H to L deficiency resolution from Year 1 to Year 2 was modestly significant (F= 2.38,p < .1). A change in legislative control (to R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 184 Democratic) increased the time needed to correct actual harm problems from Year 1 to Year 2 and from Year 1 to Year 3 (see Table 57b). Table 55a: Multiple Regression of Repetitive State Inspection Scheduling ____________________ on Legislative Determinants____________________ Independent Variable Year 1 Intercept Political Control of State Legislature (Democratic) Legislative Professionalism Aging/LTC Committee Year 2 13.6 -1.49 1.27 .361 Year 3 13.5 -.908 28.1* -6.74* Adjusted R2 -.045 .038 F .314 1.64 N 50 50 p<. 1; p><.05; p<. 01; p<.005 Unstandardized Beta coefficients Table 55b: Multiple Regression of Change in Repetitive Inspection Scheduling on Legislative Determinants Independent Variable Year 1 to Year 28 Year 2 to Year 3 Year I to Year 39 Intercept -.140 Change in Political -.456 Control of State Legislature (to Democratic) Legislative 26.6** Professionalism Aging/LTC Committee -6.98** Adjusted R2 .062 F 2.05 N 50 * * * „ _ _ _ # * * * „ „ _ p < .l; p < .05; jd<.01; p<.005 Unstandardized Beta coefficients 7 Predictability of annual state inspections was not assessed in the first year of the study (1999-2000). 8 Comparison between first and second years of the study (1999-2000 and 2000-2001) was unavailable. 9 Comparison between first and third years of the study (1999-2000 and 2001-2002) was unavailable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 185 Table 56a: Multiple Regression of Resolution of C-G Deficiencies on Legislative Determinants Independent Variable Intercept Political Control of State Year 1 44.0 -.234 Year 2 41.0 -1.38 Year 3 40.7 -1.73 Legislature (Democratic) Legislative Professionalism 15.5 5.36 -10.4 Aging/LTC Committee 1.57 4.02 4.99 Adjusted R2 F -.037 .434 -.040 .378 -.035 .458 N 50 50 50 *p < .1; *> < .05; **> < .01; ***> < .005 Unstandardized Beta coefficients Table 56b: Multiple Regression of Change in C-G Deficiency Resolution on Legislative Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -2.34 -.432 -3.47 Change in Political 3.89 2.39 -1.89 Control of State Legislature (to Democratic) Legislative -13.6 -16.2 -24.1 Professionalism Aging/LTC Committee 3.40 .968 2.66 Adjusted R2 -.016 -.032 .005 F .749 .507 1.08 N 50 50 50 p<. 1; p<. 05; p<. 01; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 186 Table 57a: Multiple Regression of Resolution of H-L Deficiencies ________________ on Legislative Determinants________________ Independent Variable Year 1 Year 2 Year 3 Intercept 52.5 46.3 36.2 Political Control of State -2.92 -6.27* -5.55 Legislature (Democratic) Legislative -1.19 11.4 33.3 Professionalism Aging/LTC Committee 10.5 1.91 3.08 Adjusted R2 -.037 .014 .027 F .494 1.21 1.40 N 50 50 50 > < .1; **p < .05; **> < .01; * ’*> < .005 Unstandardized Beta coefficients Table 57b: Multiple Regression of Change in H-L Deficiency Resolution ____________________on Legislative Determinants____________________ Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -4.19 -11.6 -19.7 Change in Political 16.3* 4.71 16.8* Control of State Legislature (to Democratic) Legislative -2.72 .618 2.18 Professionalism Aging/LTC Committee -2.96 7.95 7.04 Adjusted R2 .088 -.023 .016 F 2.38* .684 1.22 N 50 50 50 p<A\ p < .05; p < .01; ;> < .005 Unstandardized Beta coefficients Interest Group Determinants The effects of interest group determinants were evaluated by regressing the state NH enforcement measures onto four factors (influence and access of NH industry and aging advocacy groups) for each annual inspection period. For the most part, the five outcome measures and the four interest group variables were unrelated R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 187 during the three years of observation (see Tables 58a-c). Correlation analyses revealed one relationship between an outcome measure and an interest group variable. In Year 3, agency advocacy group influence in a state was positively correlated with the period of time to resolve deficiencies at the H level and above (see Table 58c). Table 58a: Correlations among Interest Group Determinants and Outcome Measures, Year 1 Variable 1 2 3 4 5 6 7 8 1 . NH IG Influence .129 -.066 .086 -.056 .029 -.105 -.132 2. Aging IG Influence — -.151 .067 -.132 -.101 -.061 .055 3. NHIG Access — .465** .000 -.105 -.227 -.226 4. Aging IG Access — -.156 -.145 -.091 -.149 5. Volume Deficiency Citations — -.029 -.078 -.077 6. Severity Deficiency Citations — .124 .209 7. C-G Deficiency Resolution . . . .645’* 8. H-L Deficiency Resolution N - 50; 'p <.05; p < .01 Table 58b: Correlations among Interest Group Determinants and Outcome Measures, Year 2 Variable 1 2 3 4 5 6 7 8 9 1. NH IG Influence — .129 -.066 .086 -.143 .127 .074 .016 -.223 2. Aging IG Influence — -.151 .067 -.151 -.046 .117 .093 -.031 3. NHIG Access — .465“ .061 -.199 -.159 -.085 -.103 4. Aging IG Access — -.136 -.070 -.116 .099 -.130 5. Volume Citations — -.202 -.186 -.091 .037 6. Severity Citations — .175 .071 .005 7. Repetitive Scheduling State — -.060 .107 Inspections 8. C-G Deficiency Resolution — .468** 9. H-L Deficiency Resolution__________________________________________________________— jV= 50; p <.05; ^<.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 8 8 Table 58c: Correlations among Interest Group Determinants and Outcome Measures, Year 3 Variable 1 2 3 4 5 6 7 8 9 1 . NHIG Influence .129 -.066 .086 -.127 -.030 .012 -.194 -.242 2. Aging IG Influence -.151 .067 -.115 -.226 .371*' .026 .005 3. NHIG Access — a y e - * * .465 .001 -.134 -.100 -.134 -.089 4. Aging IG Access — -.178 -.094 -.166 .040 -.166 5. Volume Citations — -.201 -.219 -.168 -.036 6. Severity Citations — -.267 .035 .247 7. Repetitive Scheduling of State Inspections 8. C-G Deficiency Resolution 9. H-L Deficiency Resolution -.084 -.023 .062 N= 50; "p <.05; p < .01 Extent of State Nursing Home Regulation Extent in Each Year o f Observation Interest groups did not affect the extent of state NH regulation in this study. As demonstrated by multiple regression, the combined and individual effects of the four interest group variables (influence of NH industry and aging advocacy groups and access of these groups to decision makers) on the volume (see Table 59a) and severity (see Table 60a) of state NH deficiency citations were non-significant, with one exception. Greater influence of aging advocacy groups (AARP and the Alzheimer’s Association) led to lower citation severity in Year 3 (see Table 60a). Change in Extent over the Three-Year Period Interest groups did not affect change in the extent of state NH enforcement over the three-year period of observation. The combined and individual effects of factors related to NH industry and aging advocacy groups on change in citation volume and severity were all non-significant (see Tables 59b and 60b). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 189 Table 59a: Multiple Regression of Volume of State Deficiency Citations on Interest Group Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 6.39 6.84 6.83 NH Industry Group -.062 -.292 -.252 Influence Aging Advocacy Group -.462 -.460 -.325 Influence NH Industry Group .184 .344 .215 Access Aging Advocacy Group -.507 -.503 -.559 Access Adjusted R2 -.042 -.021 -.027 F .505 .752 .676 N 50 50 50 * „ * * * * _ _ _ p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 59b: Multiple Regression of Change in Volume of State Citations on Interest Group Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept .459 -.012 .446 NH Industry Group -.231 .041 -.190 Influence Aging Advocacy Group .001 .136 .137 Influence NH Industry Group .160 -.129 .030 Access Aging Advocacy Group .004 -.055 -.051 Access Adjusted R2 -.037 -.071 -.076 F .566 .187 .139 N 50 50 50 * - ~ _ * * * „ _ * # * * _ _ _ p<. 1; p < .05; y?<.01; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 190 Table 60a: Multiple Regression of Severity of State Deficiency Citations on Interest Group Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 12.1 8.43 8.23 NH Industry Group -.062 .484 -.042 Influence Aging Advocacy Group -.887 -.596 -1.79* Influence NH Industry Group -.429 -.907 -.843 Access Aging Advocacy Group -.593 .112 .022 Access Adjusted R2 -.052 -.022 -.001 F .390 .740 .982 N 50 50 50 p < A ; p<. 05; p<. 01; p < .005 Unstandardized Beta coefficients Table 60b: Multiple Regression of Change in Severity of State Citations _______ on Interest Group Determinants__________________ Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -3.63 -.202 -3.83 NH Industry Group .546 -.526 .020 Influence Aging Advocacy Group .291 -1.19 -.902 Influence NH Industry Group -.478 .064 -.414 Access Aging Advocacy Group .704 -.090 .615 Access Adjusted R2 -.021 -.007 -.065 F .751 .914 .256 N 50 50 50 * 4 * * - _ * * * _ * * * * _ . _ p<A\ p<. 05; p<. 01; p < .005 Unstandardized Beta coefficients Effectiveness of State Nursing Home Regulation Effectiveness in Each Year o f Observation In this study, interest groups had very little impact on the effectiveness of state NH regulation as well as its extent. The combined effects of the interest group R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 191 measures on repetitive scheduling of state inspections (see Table 61a), resolution of non-actual harm deficiencies (see Table 62a) and resolution of actual harm deficiencies (see Table 63a) were all non-significant during the three years of observation. The individual effects of the four interest group variables on the extent and effectiveness of NH regulation were also non-significant, with one exception. The greater the influence of aging advocacy groups, the more repetitive were state annual inspections during Year 3 (see Table 61a). Change in Effectiveness over the Three-Year Period The combined effect of interest group variables on whether state inspections had become more or less repetitive by Year 3 was non-significant. Aging advocacy group influence led to increased repetition in the scheduling of annual inspections from Year 2 to Year 3 (see Table 61b), despite the CMS mandate. As Table 62b demonstrates, interest group determinants had a significant combined effect on change in C-G deficiency resolution from Year 1 to Year 2 (F - 3.21,p < .05). Aging advocacy influence led to an increase from Year 1 to Year 2 in the average time needed to correct non-actual harm problems. In contrast, NH industry group influence resulted in a shorter time period needed to resolve C-G level deficiencies in Year 3 as compared to Year 2 (see Table 62b). Interest group variables did not affect change in the resolution of H to L level deficiencies, as neither the combined nor the individual effects of these determinants were significant (see Table 63b). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 192 Table 61a: Multiple Regression of Repetitive State Inspection Scheduling on Interest Group Determinants__________________ Independent Variable Year l 1 0 Intercept NH Industry Group Influence Year 2 15.1 .658 Year 3 17.2 -.217 Aging Advocacy Group Influence 1.67 8.03** NH Industry Group Access -1.17 .839 Aging Advocacy Group Access -.910 -2.98 Adjusted R2 F -.043 .491 .104 2.42 N 50 50 * * * __ * * * - . * * * * p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 61b: Multiple Regression of Change in Repetitive Inspection Scheduling on Interest Group Determinants Independent Variable Year 1 to Year 21 1 Year 2 to Year 3 Year 1 to Year 31 2 Intercept 2.13 NH Industry Group -.875 Influence Aging Advocacy Group 6.36** Influence NH Industry Group 2.01 Access Aging Advocacy Group -2.08 Access Adjusted R2 .070 F 1.92 N 50 *p < .1; *p < .05; ’*> < .01; ***> < .005 Unstandardized Beta coefficients 1 0 Predictability of annual state inspections was not assessed in the first year of the study (1999-2000). 1 1 Comparison between first and second years of the study (1999-2000 and 2000-2001) was unavailable. 12 Comparison between first and third years of the study (1999-2000 and 2001-2002) was unavailable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 193 Table 62a: Multiple Regression of Resolution of C-G Deficiencies on Interest Group Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 54.4 42.4 44.9 NH Industry Group -2.00 -.332 -3.75 Influence Aging Advocacy Group -2.46 1.84 .282 Influence NH Industry Group -4.98 -3.17 -3.91 Access Aging Advocacy Group .929 3.41 2.82 Access Adjusted RJ -.007 -.051 -.002 F .910 .409 .979 N 50 50 50 * _ _ * * * _ , * # * * p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 62b: Multiple Regression of Change in C-G Deficiency Resolution on Interest Group Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -12.0 2.51 -9.54 NH Industry Group 1.67 -3.42* -1.75 Influence Aging Advocacy Group 4.30* -1.56 2.74 Influence NH Industry Group 1.81 -.742 1.07 Access Aging Advocacy Group 2.48 -.590 1.89 Access Adjusted R2 .153 .015 -.020 F 3.21** 1.18 .764 N 50 50 50 * _* * * „ „ * * * * „ „ _ p<. 1; p < .05; p<.01; p<.005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 194 Table 63a: Multiple Regression of Resolution of H-L Deficiencies on Interest Group Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 72.0 59.6 55.6 NH Industry Group -5.29 -5.31 -6.27 Influence Aging Advocacy Group 1.62 -.713 1.35 Influence NH Industry Group -7.68 -2.43 -1.22 Access Aging Advocacy Group -.984 -1.25 -3.05 Access Adjusted R2 -.014 -.022 -.015 F .849 .749 .833 N 50 50 50 * ** _ _ * * * * „ * * * * . „ _ p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 63b: Multiple Regression of Change in H-L Deficiency Resolution on Interest Group Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -11.7 -10.6 -27.5 NH Industry Group -.176 -1.77 .278 Influence Aging Advocacy Group -2.50 3.02 -.264 Influence NH Industry Group 5.12 -1.39 3.95 Access Aging Advocacy Group -.447 4.31 4.50 Access Adjusted R2 -.033 -.061 -.043 F .648 .383 .564 N 50 50 50 sk # * # * & * * * * p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Gubernatorial Determinants Gubernatorial determinants were evaluated by regressing state NH enforcement measures onto three variables. Three correlations existed between outcome measures and gubernatorial variables (political party, institutional power R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. and reference to NH policy in the state-of-the-state (SOS) address). First, a Democratic governor corresponded with faster resolution of C to G level deficiencies in Year 2 (see Table 64b). Second, the mention of NH policy in the governor’s state- of-the-state address was related to faster resolution of C to G level deficiencies during Year 2 (see Table 64b). Third, the greater the governor’s institutional power in Year 3, the slower was the state’s resolution of H to L level deficiencies (see Table 64c). Table 64a: Correlations among Gubernatorial Determinants and Outcome _________________________ Measures, Year 1_________________________ Variable 1 2 3 4 5 6 7 1. Political Party of Governor — -.332' .012 .267 .047 -.251 .003 (Democratic) 2. Institutional Power of Governor — .068 -.111 .051 .164 .043 3. NH Policy in Governor’s SOS — .060 -.069 -.036 -.037 4. Volume Deficiency Citations . . . -.029 -.078 -.077 5. Severity Deficiency Citations — .124 .209 6. C-G Deficiency Resolution . . . .645** 7. H-L Deficiency Resolution . . . T V = 50; *p <.05; **p < .01 Table 64b: Correlations among Gubernatorial Determinants and Outcome Measures, Year 2_________________________ Variable 1 2 3 4 5 6 7 8 1 . Political Party of Governor -.268 -.110 .244 -.111 -.054 -.371** -.256 (Democratic) 2. Institutional Power of -.107 -.109 .174 -.063 .068 .035 Governor 3. NH Policy in Governor’s — .040 -.096 .104 .331* .088 SOS 4. Volume Citations — -.202 -.186 -.091 .037 5. Severity Citations — .175 .071 .005 6. Repetitive Scheduling State Inspections 7. C-G Deficiency Resolution 8. H-L Deficiency Resolution -.060 .107 .468** N= 50; p <.05; p < .01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 196 Table 64c: Correlations among Gubernatorial Determinants and Outcome _________________________ Measures, Year 3__________________________________ Variable 1 2 3 4 5 6 7 8 1. Political Party of Governor (Democratic) 2. Institutional Power of -.214 .234 .283* -.079 -.003 -.030 .188 .149 -.061 -.159 -.012 -.014 .304* Governor 3. NH Policy in Governor’s — -.152 -.079 .262 .180 .018 SOS 4. Volume Citations — -.201 -.219 -.168 -.036 5. Severity Citations 6. Repetitive Scheduling State Inspections 7. C-G Deficiency Resolution 8. H-L Deficiency Resolution -.267 .035 -.084 .247 -.023 .062 N = 50; "p <.05 Extent of State Nursing Home Regulation Extent in Each Year o f Observation In this study, factors related to the state governors did not affect the extent of state NH regulation. The combined and individual effects of political party, institutional power, and reference to NH policy in the SOS address on the volume of state NH regulation (see Table 65a) and the severity of state NH regulation (see Table 66a) was non-significant, with one exception. A Democratic governor led to one more deficiency citation for every 100 NH beds in Year 1 (see Table 65a). Change in Extent over the Three-Year Period Gubernatorial variables did not affect change in either the volume (see Table 65b) or the severity (see Table 66b) of state NH deficiency citations from year to year in this study, as the combined and the individual effects of these measures were all non-significant. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 9 7 Table 65a: Multiple Regression of Volume of State Citations on Gubernatorial Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 5.50 6.19 6.19 Political Party of 1.23* 1.13 -.207 Governor (Democratic) Institutional Power of -.083 -.157 -.018 Governor NH Policy in .382 .302 -.927 Governor’s SOS Adjusted R2 .014 .000 -.043 F ' 1.22 1.01 .350 N 50 50 50 * * * * * * ^„ *#** _ _ v _ p < A ; p < .05; p<. 01; p<- 005 Unstandaxdized Beta coefficients Table 65b: Multiple Regression of Change in Volume of State Citations on Gubernatorial Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept .520 -2.21 -1.46 Change in Political 1.10 --314 -.393 Party of Governor (to Democratic) Institutional Power of -.043 ^554 .468 Governor Change in NH Policy in .239 -.469 -.146 Governor’s SOS Adjusted R2 -.026 .005 -.048 F .605 1.08 .285 N 50 50 50 p<A; p<- 05; p < .0 1 ; p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 9 8 Table 66a: Multiple Regression of Severity of State Deficiency Citations on Gubernatorial Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 6.86 3.82 -1.60 Political Party of .738 -.526 .506 Governor (Democratic) Institutional Power of 1.11 1.33 2.46 Governor NH Policy in -1.21 -.734 -2.17 Governor’s SOS Adjusted R2 -.042 -.015 .001 F .372 .771 1.01 N 50 50 50 * * * ^ _ * * * ~ . * * * * . _ _ p < .1; p < . 05; p < . 01; p < . 005 Unstandardized Beta coefficients Table 66b: Multiple Regression of Change in Severity of State Citations on Gubernatorial Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept -4.94 -2.42 -7.66 Change in Political 3.76 -2.66 -1.82 Party of Governor (to Democratic) Institutional Power of .637 .423 1.19 Governor Change in NH Policy in .809 -.089 .468 Governor’s SOS Adjusted RJ .006 -.016 -.035 F 1.10 .747 .464 N 50 50 50 p < . 1; p < .0 5 ; p < .0 1 ; p < .005 Unstandardized Beta coefficients Effectiveness of State Nursing Home Regulation Effectiveness in Each Year o f Observation State governors had very little impact on repetition in the scheduling of annual state inspections. The combined and individual effects of the gubernatorial measures on repetitive scheduling were non-significant (see Table 67a). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 199 However, as Table 68a shows, the combined effect of gubernatorial determinants on resolution of C to G level deficiencies was significant (F = 4.13,p < .05) in Year 2, as were the individual effects of political party and the presence of NH policy in the SOS address that year. In Year 2, Democratic governors were associated with faster resolution of non-actual harm deficiencies; while the governor’s reference to NH policy led to slower correction of these problems. Reference to NH policy in the governor’s SOS address also led to slower resolution of C to G level deficiencies in Year 3 (see Table 68a). The combined effect of gubernatorial measures on resolution of actual hatm deficiencies was non-significant in Years 1, 2 and 3. However, greater institutional power led to slower resolution of H to L deficiencies in Year 3 (see Table 69a). Change in Effectiveness over the Three-Year Period Gubernatorial determinants had a significant combined effect (F= 3.50,p < .05) on change in the repetitive scheduling of state inspections from Year 2 to Year 3 (see Table 67b). A change in the political party of the governor (from Republican to Democratic) decreased repetitive scheduling. The presence of NH policy in the governor’s SOS address led to an increase in repetitive inspections from Year 2 to Year 3 (see Table 67b). The combined effect of gubernatorial determinants on change in C to G level deficiency resolution from Year 1 to Year 3 was significant (F = 3.19,p < .05). Change in the political party of the governor (to Democratic) increased the speed R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 200 with which non-actual harm problems were corrected (see Table 68b). Gubernatorial variables had no effect on change in H to L deficiency resolution (see Table 69b). Table 67a: Multiple Regression of Repetitive State Inspection Scheduling on Gubernatorial Determinants Independent Variable Year I 1 3 Year 2 Year 3 Intercept 14.8 21.3 Political Party of -.848 1.57 Governor (Democratic) Institutional Power of -.150 -1.81 Governor NH Policy in 1.32 8.82 Governor’s SOS Adjusted R2 -.060 .014 F .112 1.22 N 50 50 ><.1;*V< .05; **> < .01; "*> < .005 Unstandardized Beta coefficients Table 67b: Multiple Regression of Change in Repetitive Inspection Scheduling ______________________ on Gubernatorial Determinants___________________ Independent Variable Year 1 to Year 21 4 Year 2 to Year 3 Year 1 to Year 31 5 Intercept 7.82 Change in Political -13.2*’ Party of Governor (to Democratic) Institutional Power of -1.07 Governor Change in NH Policy in 5.29** Governor’s SOS____________________________________________________________________ Adjusted R2 .138 F 3.50** N ^ ^ ^ 50 p < .1; p < .05; p < .01; * *p < . 0 0 5 Unstandardized Beta coefficients 1 3 Predictability of annual state inspections was not assessed in the first year of the study (1999-2000). 1 4 Comparison between first and second years of the study (1999-2000 and 2000-2001) was unavailable. 1 5 Comparison between first and third years of the study (1999-2000 and 2001-2002) was unavailable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 1 Table 68a: Multiple Regression of Resolution of C-G Deficiencies on Gubernatorial Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 36.0 41.9 59.8 Political Party of -6.07 -10.8** -7.06 Governor (Democratic) Institutional Power of 4.27 .971 -4.98 Governor NH Policy in -2.43 10.4” 12.6* Governor’s SOS Adjusted Rz .018 .166 .028 F 1.29 4.13” 1.46 N 50 50 50 * * * „ _ * * * * * * * ~ ~ _ p < .1; p < .05; p < .01; p < .005 Unstandardized Beta coefficients Table 68b: Multiple Regression of Change in C-G Deficiency Resolution on Gubernatorial Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept 1.56 -29.7 -58.4 Change in Political -3.94 -21.6* -35.6**’ Party of Governor (to Democratic) Institutional Power of -2.74 9.66 12.3 Governor Change in NH Policy in -.207 .662 -8.50 Governor’s SOS Adjusted R2 -.071 .047 .141 F .067 1.68 3.19” N 50 50 50 * p < .1; *> < .05; ” > < .01; ***> < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 2 Table 69a: Multiple Regression of Resolution of H-L Deficiencies on Gubernatorial Determinants Independent Variable Year 1 Year 2 Year 3 Intercept 27.1 54.7 -50.9 Political Party of 3.09 -10.2 6.72 Governor (Democratic) Institutional Power of 9.03 -.285 27.1*’* Governor NH Policy in -4.25 .762 -12.3 Governor’s SOS Adjusted R2 -.059 -.002 .095 F .219 .965 2.46 N 50 50 50 # # * * * * * p< . 1; p<. 05; p < .0 1 ; p < .005 Unstandardized Beta coefficients Table 69b: Multiple Regression of Change in Resolution of H-L Deficiencies on Gubernatorial Determinants Independent Variable Year 1 to Year 2 Year 2 to Year 3 Year 1 to Year 3 Intercept 4.02 11.9 12.1 Change in Political -7.61 -6.30 -8.09 Party of Governor (to Democratic) Institutional Power of -2.30 -4.13 -5.42 Governor Change in NH Policy in 1.26 2.59 -.811 Governor’s SOS Adjusted R2 -.022 -.009 .042 F .667 .863 1.69 N 50 50 50 ♦ * * . _ * * * . . * * * * _ „ _ p <.1; p < .05; p < .0 1 ; p < .005 Unstandardized Beta coefficients Theoretical Models and HLM Analysis Results from the previous multiple regression analyses were used to create theoretical models for each outcome measure in this study. The five models were: 1) volume of state NH deficiency citations, 2) severity of state NH deficiency R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 3 citations, 3) repetitive scheduling of annual state NH inspections, 4) resolution of C to G level deficiencies, and 5) resolution of H to L level deficiencies (see Table 70). A pooled analysis of each theoretical model was then completed using hierarchical linear modeling (HLM). For each outcome measure, the fit of the theoretical (conditional with time) model is compared with that of the intercept (unconditional) model and intercept plus slope (unconditional with time), using the deviance statistic. The closer the deviance value is to 0, the better a model fits. Table 70: Summary of Multiple Regression Analyses: Outcome Measures and their Determinants Outcome Measure Determinants and Effects on Outcome Measure Volume of State Deficiency Citations Proportion of Oldest-Old (-) Education (-) Political Culture (+) Total State Agency Funding/NH Bed (+) Years of Experience of State Surveyors (-) Political Party of Governor (+) Severity of State Deficiency Citations Urbanization (-) Education (+) State % of Total State Agency Funding (-) Legislative Professionalism (-) Aging/LTC Cmte. in State Legislature (+) Aging Advocacy Group Influence (-) Predictability of State Inspections Proportion of Oldest-Old (+) Legislative Professionalism (+) Aging/LTC Cmte. in State Legislature (-) Aging Advocacy Group Influence (+) Resolution of C-G Deficiencies Urbanization (+) Proportion of Oldest-Old (+) Total State Agency Funding/NH Bed (-) Years of Experience of State Surveyors (+) Political Party of Governor (-) NH Policy in Governor’s SOS Address (+) Resolution of H-L Deficiencies Proportion of Oldest-Old (+) Total State Agency Funding/NH Bed (+) Years of Experience of State Surveyors (+) Political Control of State Legislature (-) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 4 Volume of Deficiency Citations The theoretical model for volume of deficiency citations included six covariates (education, proportion of oldest-old, political culture, total state surveying agency funding per NH bed, average number of years of experience of state surveyors and political party of the governor). The theoretical model for volume of deficiency citations fit the data significantly better did than the intercept model {deviance difference = 50, df= 15, p < .001) and the intercept plus slope model {deviance difference = 45, df= 12,p < .001). Three individual covariates had significant effects on citation volume. A lower proportion of college-educated residents, a moralistic subculture and a Democratic governor predicted a higher volume of citations in a particular year. All three trends were previously observed in multiple regression analyses. However, political party of the governor had the opposite effect on change in citation volume; a Democratic governor predicted a decrease in citation volume from one year to the next (see Table 71). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 5 Table 71: HLM Estimates of Volume of State Deficiency Citations Volume Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE 0 SE Intercept 6.01 .292 5.86 .318 10.7 2.73 Education — — — — -.225**** .054 Proportion of Oldest- — — - - -.616 .722 Old Political Culture — — — — .011**** .101 Total Funding/NH Bed — — - - -.130 .003 Average Number of - - - - 1.06“ .060 Years of Experience of State Surveyors Political Party of — — — — 1.06** .448 Governor (Democratic) Change in Volume Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE 0 SE Slope — — .150 .105 -.409 .024 Education — — — — .024 .023 Proportion of Oldest- — — — — .337 .289 Old Political Culture — — — — -.002 .029 Total Funding/NH Bed - — — — -.001 .039 Average Number of - - - - -.015 .030 Years of Experience of State Surveyors Political Party of — ~ — — -.525“ * .448 Governor (Democratic) Deviance 547 542 497 Number of Estimated 3 6 18 Parameters * * * ~ ^ * # * * „ ^ _ p < .05; p < .01; p < .005 Unstandardized Beta coefficients Severity of Deficiency Citations Six covariates were represented in the theoretical model for citation severity: urbanization, education, state proportion of total state surveying agency funding, legislative professionalism, presence of an aging or LTC committee in the state legislature, and aging advocacy group influence. The theoretical model provided a R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 206 better fit than either the intercept model ( < deviance difference — 77, df= 15 ,p < .001) or the intercept plus slope model (deviance difference = 45, df= 22, p < .050). As seen in the multiple regression analyses, higher education led to a higher proportion of severe citations in the HLM estimates. However, this was the only determinant in the theoretical model with a significant effect on citation severity and none o f the individual covariates predicted change in citation severity (see Table 72). Table 72: HLM Estimates of Severity of State Deficiency Citations Severity Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE a SE Intercept 8.44 .469 10.3 .597 3.75 4.59 Urbanization — — — — .011 .047 Education — — — — .322" .130 State % of Total — — — — -.048 .165 Funding Legislative - - - - -7.30 6.53 Professionalism Aging/LTC Committee - - „ - .905 1.26 Aging Advocacy Group - - - — -.446 1.18 Influence Change in Severity Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE 0 SE Slope — — -1.85 .274 -1.89 2.08 Urbanization — — — — -.032 .021 Education — — — — .053 .054 State % of Total — — — — .078 .071 Funding Legislative - — - — -1.04 2.69 Professionalism Aging/LTC Committee — — — — .456 .624 Aging Advocacy Group - - - -.387 .394 Influence Deviance 821 766 744 Number of Estimated 3 6 1 8 Parameters *p < .1; **p < .05; *"p < .0 1 ; '"'p < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 7 Repetitive Scheduling of Annual State Inspections The theoretical model for repetitive scheduling of state inspections comprised four covariates: proportion of oldest-old, legislative professionalism, presence of an aging or LTC committee in the state legislature and aging advocacy group influence. The theoretical model provided a better fit than both the intercept model {deviance d iffe re n c e -2 A ,d f-9 ,p < .005) and the intercept plus slope model {deviance difference = 21, df= 8,p < .010). Legislative professionalism, the presence of an aging or LTC committee and aging advocacy group influence predicted change in repetitive scheduling of annual state inspections from Year 2 to Year 3. As expected, the presence of an aging or LTC committee led to state improvement in this measure of effectiveness (scheduling became less repetitive from Year 2 to Year 3). However, as observed in multiple regression, greater legislative professionalism and greater aging advocacy group influence predicted more repetitive inspections from one year to the next (see Table 73), contrary to initial hypotheses. Finally, it is interesting to note that while these three variables predicted change in repetitive scheduling from Year 2 to Year 3, they did not predict repetitive scheduling itself, in either Year 2 or Year 3 (see Table 73). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 8 Table 73: HLM Estimates of Repetitive Scheduling of Annual State Inspections Repetitive Scheduling Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE 0 SE Intercept 15.0 1.22 13.8 1.37 5.55 6.13 Proportion of Oldest- - — — - 4.85 3.72 Old Legislative — - - - 2.59 9.90 Professionalism Aging/LTC Committee — — — - -.333 3.25 ' Aging Advocacy Group — — ~ - 1.51 2.25 Influence Change in Repetitive Scheduling Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE 0 SE Slope — — 2.41 1.94 2.88 5.24 Proportion of Oldest- - — — — -3.01 2.86 Old Legislative - — — - 26.2** 11.3 Professionalism Aging/LTC Committee — - ~ - -6.98** 3.19 Aging Advocacy Group — - ~ - 6.11**** 1.68 Influence Deviance 721 718 697 Number of Estimated 3 4 12 Parameters * ** „ _ *** ^ „ **** _ p < A ; p < . 05; /><.01; p < .005 Unstandardized Beta coefficients Resolution of C to G Level Deficiencies Six covariates were analyzed in the theoretical model for resolution of C to G level deficiencies: urbanization, proportion of oldest-old, total state surveying agency funding per NH bed, average number of years of experience of state surveyors, and political party of the governor. The theoretical model fit the data significantly better than did the intercept model (deviance difference = 43, df= 15,p < .001), but not the intercept plus slope model (deviance difference = 17, df= 12, p > .1). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 0 9 Higher state agency funding let to faster correction of non-actual harm problems, while more experienced surveyors led to slower resolution. None of the covariates predicted change in resolution of C to G level deficiencies (see Table 74). Table 74: HLM Estimates of Resolution of C-G Deficiencies C-G Resolution Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE & SE Intercept 43.8 1.89 47.3 2.10 38.1 18.2 Urbanization — — ~ — .113 .150 Proportion of Oldest- - - — - 1.43 6.02 Old Total Funding/NH Bed — - — - -.035* .019 Average Number of - — — - 1.36** .573 Years of Experience of State Surveyors Political Party of - - - ~ -5.96 3.71 Governor (Democratic) NH Policy in - - - ~ -2.51 3.47 Governor’s SOS Change in C-G Resolution Variable Unconditional Unconditional + Conditional + Time Time 0 SE 0 SE 0 SE Slope — — -3.50 .754 .095 6.49 Urbanization — — — — -.062 .058 Proportion of Oldest- - — — -- 1.86 1.97 Old Total Funding/NH Bed - — — — -.000 .010 Average Number of - - - - -.279 .233 Years of Experience of State Surveyors Political Party of — — — — .263 1.54 Governor (Democratic) NH Policy in — - - - -.767 1.57 Governor’s SOS Deviance 1150 1124 1107 Number of Estimated 3 6 18 Parameters > < .1; *> < .05; **> < .01; ***> < .005 Unstandardized Beta coefficients R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 0 Resolution of H to L Level Deficiencies Four covariates (proportion of oldest-old, total state surveying agency funding per NH bed, average number of years of experience of state surveyors and political control of the legislature) were included in the theoretical model for H to L level deficiency resolution. The theoretical model provided a better fit than did the intercept {deviance difference = 56, df= 7,p < .001) and the intercept plus slope models {deviance difference = 36, df= 4 ,p < .001). In the HLM estimates, greater state survey agency funding led to slower resolution of H to L level deficiencies, a result previously observed in the multiple regression analyses. This was the only individual determinant in the theoretical model that predicted correction of actual harm problems. HLM was not able to measure the slope of change from year to year, possibly due to the relative infrequency of severe deficiency citations (see Table 75). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 1 Table 75: HLM Estimates of Resolution of H-L Deficiencies H-L Resolution Variable Unconditional Unconditional + Conditional + Time Time J 8 SE 0 SE 0 SE Intercept 51.5 3.22 58.8 4.15 58.8 4.15 Proportion of Oldest- - — - .728 3.40 Old Total Funding/NH Bed - — - .054**** .014 Average Number of - — -- .436 .333 Years of Experience of State Surveyors Political Control of — — . . 1.36 2.25 Legislature (Democratic) Change in H-L Resolution Variable Unconditional Unconditional + Conditional + Time Time1 6 0 SE 0 SE SE Slope — — -7.34 1.71 Proportion of Oldest- — — - Old Total Funding/NH Bed — — — Average Number of - — - Years of Experience of State Surveyors Political Control of — — . . Legislature (Democratic) Deviance 1288 1248 1232 Number of Estimated 3 6 10 Parameters *p < .1; "p < .05; **> < .01; ***> < .005 Unstandardized Beta coefficients Summary of Factors Predicting State Nursing Home Regulation HLM revealed that the theoretical models improved the fit of unconditional models for four of the five outcome measures (volume of state citations, severity of state citations, repetitive scheduling of annual state inspections and resolution of H to 1 6 Iterations stopped due to small change in likelihood function. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 2 L level deficiencies). For the other measure (resolution of C to G level deficiencies), model fit was not improved by the addition of covariates (see Table 76). Several individual determinants predicted state nursing home regulation in both multiple regression and HLM analyses. A Democratic governor led to higher volume of citations. A higher level of education predicted more severe citations. An aging or LTC committee led to improved state performance in avoiding repetitive inspection scheduling. Greater surveying agency funding led to faster resolution of non-actual harm deficiencies (see Table 76). These findings were all anticipated. However, other results were surprising. Professional legislatures and aging advocacy group influence predicted poorer performance by the states in avoiding repetitive inspections by the end of the three-year period, while greater state agency funding led to slower resolution of actual harm deficiencies (see Table 76). These three findings were in direct contradiction to hypotheses proposed in Chapter 1. Table 76: Summary of HLM Analyses: Outcome Measures and their Determinants Theoretical Model/Outcome Measure Model Fit Improved? Determinants and Effects on Outcome Measure or Change in Outcome Measure Volume of State Deficiency Citations YES Education (-) Political Culture (+) Political Party of Governor (+) CHANGE Political Party of Governor (-) Severity of State Deficiency Citations YES Education (+) Predictability of State Inspections YES CHANGE Legislative Professionalism (+) Aging/LTC Cmte. in State Legislature (-) Aging Advocacy Group Influence (+) Resolution of C-G Deficiencies NO Total State Agency Funding/NH Bed (-) Resolution of H-L Deficiencies YES Total State Agency Funding/NH Bed (+) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 213 Chapter Summary In this study, bivariate statistics revealed that interstate differences exist in the extent and effectiveness of NH regulation. California had nearly five times the volume of citations as did Wisconsin. However, citation volume was not linked with citation severity, as evidenced in California, which had the smallest proportion of severe citations among the states in Year 3. The states varied in the implementation of CMS initiatives. However, states that successfully avoided repetitive annual inspections were not necessarily the same states that quickly resolved deficiencies. The results of HLM analysis indicated that the theoretical models proposed in this dissertation provided a better fit than did the unconditional models for the various outcome measures of extent and effectiveness. The inclusion of the covariates selected following multiple regression improved the model fit for four of the five outcome measures (citation volume, citation severity, repetitive scheduling of state inspections, and H-L deficiency resolution). For the fifth measure (C-G deficiency resolution), the theoretical model did not provide a better fit. General findings of this research were: 1) external and gubernatorial determinants predict interstate variations in the volume and severity of NH deficiency citations, 2) agency and legislative determinants correspond with differences in the resolution of deficiencies by the states and 3) theoretical models integrating determinants both external and internal to the state NH policy ACF were largely successful in explaining interstate variation in state NH regulation. These findings and their implications are discussed in the final chapter. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 4 Chapter 5 Discussion This dissertation addressed the administration of national NH policy by the states. This discussion chapter first reviews the extent and the effectiveness of NH regulation in the 50. states. Extent was represented by the volume and severity of state NH deficiency citations. Effectiveness was measured as the success of state implementation of two recent federal mandates requiring states to: 1) avoid repetitive scheduling of annual state inspections and 2) resolve NH deficiencies in a timely manner. The present study is the first in which implementation of these two mandates is compared across the 50 states. Second, this chapter evaluates the theoretical models tested through hierarchical linear modeling (HLM) for each outcome measure of extent and effectiveness. In this study, theoretical models were tested for five outcome measures: 1) volume of state deficiency citations, 2) severity of state deficiency citations, 3) repetitive scheduling of annual state inspections, 4) resolution of-C to G level deficiencies and 5) resolution of H to L level deficiencies. Previous studies have examined aspects of state agencies that are believed to impact state NH regulation. However, this study is the first in which the integrated effects of these external and internal determinants on state NH regulation have been considered. In summary, this discussion reviews the efforts of this study to measure the extent and effectiveness of state NH regulation and to explain interstate variation in the implementation of national NH policy. This chapter then considers contributions R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 5 made by this research to the aging policy and comparative state policy literatures, as well as to NH residents and their families. Finally, the discussion identifies the limitations of this research and suggests future topics for investigation. This dissertation concludes by affirming the completion of an important study both of aging policy and of state regulatory policy. The Extent and the Effectiveness of State Nursing Home Regulation This section reviews the collection of data and the construction of outcome measures for this study and discusses the results that were obtained. In this dissertation, the extent of state NH regulation was assessed as the volume and severity of state nursing home citations. Effectiveness of state NH regulation was measured as the success with which the states implemented federal mandates concerning the repetitive scheduling of annual inspections and the timely resolution of deficiencies. The Collection of State Nursing Home Regulation This dissertation used the NH deficiency citations issued by state NH surveying agencies to gauge the extent and effectiveness of state NH regulation. Information on state citations was obtained in January 2003 from CMS’s Nursing Home Compare website. For the purposes of the present study, Nursing Home Compare facilitated a fast, efficient and reliable collection of state NH regulatory R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 6 activity. Altogether, more than 600,000 deficiency citations issued by state surveying agencies over a three-year period were analyzed in this study. Extent of State Nursing Home Regulation The volume of state deficiency citations was fairly consistent over the three- year period of observation. Nationally, the level of citations per NH bed rose slightly from Year 1 to Year 2, and then declined from Year 2 to Year 3, though not to the level of Year 1. Citation volume in individual states was also consistent; for example, California issued a relatively high volume of citations in each year of the study. The stability in state citation volume observed in this study follows a period of fluctuation (1991 to 2001). The six years from 1991 to 1997 saw falling numbers of citations (Harrington & Carrillo, 1999); this was shortly followed by a three year period (1998-2001) of increased citations (USDHHS, 2003). The present study was also unique among recent studies of state NH policy in comparing not only the volume of deficiency citations, but also the percentage of severe deficiency citations across the 50 states. While previous studies (Harrington & Carrillo, 1999; USDHHS, 2003) reported only national percentages of actual harm deficiencies, this study provided a state-by-state breakdown of the proportion of citations at the G level and above for each year of observation. This yields valuable insight into the prevalence of serious NH deficiencies in different states, providing a clearer picture of the scope and severity of the problems each state NH surveying agency must confront. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 7 While citation volume remained level during the three years of this study, citation severity steadily declined from Year 1 to Year 3. Nationally, more than 10 percent of state NH citations were for actual harm deficiencies (G level and above) in Year 1; this decreased to less than seven percent by Year 3. In some states, the change in citation severity was dramatic, particularly in California, where severe deficiencies had all but disappeared by Year 3. This finding underlines the importance of context to reports of rising or falling numbers of NH deficiencies in previous studies (Harrington & Carrillo, 1999; USDHHS, 2003). A high frequency of citations alone may not indicate substandard NH care in a state. Effectiveness of State Nursing Home Regulation The use of NH deficiency citations as the sole means of evaluating the performance of the state surveying agency that issues them is problematic. For example, it is difficult to decide whether high deficiency frequency should reflect negatively on NHs, or positively on the state agency that recognizes their infractions (Mullan & Harrington, 2001; Walshe, 2001). The present study addressed this conundrum by also investigating the effectiveness of state implementation of federal requirements to: 1) avoid predictable annual inspections and 2) resolve non-actual harm deficiencies within 60 days and actual harm deficiencies within 30 days. The US GAO (2000) studied implementation of these measures in a small sample of states; this dissertation examined implementation in all 50 states. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 8 This dissertation concluded that repetitive scheduling of annual state inspections remains a problem. In this study, 15 percent of the NHs in a state were inspected during the same month two years in a row. This means nearly one in five NHs in a typical state could anticipate the state’s “surprise” inspection based on the timing of the previous year’s survey, despite the federal directive to avoid this situation. Further, this problem worsened over the three-year period. In one state (Iowa), more than half of the annual inspections in Year 3 were predictable. This is a disturbing finding. The integrity of an annual inspection is compromised when a NH has time to prepare. This can enable a NH to conceal chronic quality-of-care problems from state surveyors (Vladeck, 1980). In the implementation of the other CMS mandate of interest, timely resolution of NH deficiencies, this study found mixed results. In all three years of the study, the average length of time needed by the states to correct non-actual harm problems was well within the 60-day grace period allowed for resolution of C to G level deficiencies under federal regulations, and by Year 3, only four states (Connecticut, Montana, New Hampshire, and New York) exceeded this limit. In 2000, the US GAO identified slow deficiency resolution as a problem among state surveying agencies. The present study’s findings indicate that the states improved from year to year in quickly resolving the non-actual harm problems that constitute a majority of NH deficiency citations. However, in most states, resolution of actual harm (H to L) deficiencies was slow and showed little improvement over the course of the study. In each year of R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 9 observation, the average length of time for correction of these problems by the states was generally longer than the required 30-day limit. Only nine of the 45 states with H to L level deficiencies in Year 3 removed these problems within 30 days. One state (Alaska) took nearly five months to verify that two life-threatening deficiencies had been removed. Surprisingly, while the states were generally successful in swiftly resolving the minor deficiencies in a typical NH, they were unable to quickly correct the relatively small number of major problems that place the lives of NH residents in jeopardy. This is another troubling discovery. Finally, the present study found no consistent pattern among the 50 states in implementation of the two federal mandates. The states that successfully avoided repetitive scheduling of annual NH inspections were not necessarily the same states that quickly corrected NH problems. Because this is the first study to analyze implementation of these two CMS mandates in all 50 states, the finding that states seem to enforce one federal mandate and not the other is important. This suggests that there are no clear-cut “good” or “bad” states in terms of implementation of the CMS mandates. Perhaps the question to be asked is not “How well do the states implement federal NH policy?” (Thompson and Scicchitano, 1987), but rather, “ ‘Which federal policies do the states choose to implement?” R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 0 Factors Predicting State Nursing Home Regulation » m m ill nnw In this dissertation, the extent and effectiveness of state NH regulation in each year of observation were regressed onto five sets of variables: 1) external determinants, 2) agency determinants, 3) legislative determinants, 4) interest group determinants and 5) gubernatorial determinants. A theoretical model integrating external and internal determinants with significant effects was then proposed for each outcome measure and these five models were tested using HLM. External Determinants Extent o f State Nursing Home Regulation External determinants had a significant combined effect on state citation volume in all three years. Among individual measures, a moralistic subculture led to greater citation volume, which supported Hypothesis Id. However, two results contradicted earlier hypotheses. First, higher levels of education did not lead to higher citation volume, as Hypothesis lb predicted. Second, a higher proportion of oldest-old predicted lower citation volume, undermining the notion of Hypothesis lc that a large LTC target population would favor more NH citations. Finally, the effects of urbanization (negative in Years 1 and 2, positive in Year 3) were not statistically significant, suggesting that urbanization has little bearing on citation volume, contrary to Hypothesis la, possibly due to the small sample size (50 states). External determinants had a significant combined effect on state citation severity in Years 2 and 3, but not in Year 1. This may reflect the dramatic decline in R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 1 severe citations in highly developed states such as California during the latter two years of observation. The more educated a state’s population, the higher was the state’s proportion of severe citations. On the other hand, the more urbanized a state, the lower was its proportion of severe citations. While the first finding supports Hypothesis lb, the second contradicts Hypothesis la and suggests that severe NH deficiencies may be more prevalent in rural areas. Finally, a higher oldest-old proportion and a moralistic political culture did not lead to more frequent issuance of severe citations, as was predicted by Hypotheses lc and Id, respectively. While each external determinant in this study affected either volume or severity of NH citations in at least one year, these variables had minimal impact on change in volume or severity over the three-year period. This was not surprising, as the values for the external determinants measures themselves did not change from year to year. Data from the 2000 U.S. Census was used for three measures (urbanization, education and proportion of oldest-old), while Elazar’s 1984 description of political culture was utilized for that variable. Effectiveness o f State Nursing Home Regulation External determinants generally did not explain state variance in repetitive scheduling of state inspections. A higher proportion of oldest-old in a state led to a greater number of state NH inspections during the same month in Years 1 and 2. This finding was surprising, for it was expected that states with a higher percentage of potentially frail elders would be vigilant in employing new federal NH inspection R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 2 guidelines. In fact, several states with older populations (North Dakota and Iowa, for example) were among the most predictable in scheduling annual NH inspections. Further investigation of the link between a state’s aged proportion and repetitive scheduling may be necessary. Urbanization, education and political culture did not impact repetitive scheduling of annual NH inspections. Three external determinants (urbanization, education and political culture) had a minimal impact on deficiency resolution. One measure, proportion of oldest- old, did impact state deficiency resolution; however, the effect of this measure contradicted Hypothesis lc. A higher oldest-old proportion led to slower resolution of C to G level deficiencies in Years 2 and 3 and to slower resolution of H to L level deficiencies in Year 1. One explanation is that deficiency resolution may take more time in states with a large aged population, despite the urgent need for swift deficiency resolution in these states. Although negative effects (faster correction) of oldest-old proportion on H to L level deficiency resolution were found in Years 2 or 3, these did not reach statistical significance. External determinants did not affect change in the effectiveness of state NH regulation. This is likely due to the fact that the values for these determinants were based on 2000 U.S. Census data and did not change from year to year in this study. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 3 Agency Determinants Extent o f State Nursing Home Regulation Agency determinants largely explained state variance in NH deficiency citation volume. Greater funding for state agencies led to more citations per NH bed. This supports Hypothesis 2b, which predicted that greater financial resources would lead to more frequent state NH regulatory activity. What was not expected, however, was that less experienced surveyors predicted greater citation issuance. One explanation is the confounding effect of NH quality. In states with poorly performing NHs, deficiencies may be blatant enough to be detected by even a rookie surveyor. Finally, a state’s financial stake in its NH surveying agency, as indicated by the state’s proportion of agency funding, did not affect citation volume. In contrast to the above findings concerning citation volume, agency determinants had little bearing on the percentage of severe deficiency citations in the states. The one exception was that greater agency funding per NH bed led to lower citation severity in Year 2. While this finding may suggest that well-funded state agencies have eradicated actual harm deficiencies, it is more likely due to the paucity of severe deficiencies in certain states that have few NH beds (e.g., Hawaii and Vermont). Total agency funding and surveyor experience had no impact on citation severity, another indication that poor quality-of-care in NHs is the primary trigger for serious enforcement actions. Finally, agency determinants had very little effect on change in the volume and severity of state deficiency citations from year to year. This may be an R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 4 indication that increases to funding of state agencies during the three years of observation were too minimal to impact the extent of state enforcement actions. Effectiveness o f State Nursing Home Regulation Agency factors did not predict repetitive scheduling of state inspections. Greater agency funding and surveyor experience were expected to enhance the state’s ability to avoid repetitive scheduling; however, financial and human resources of state agencies did not affect implementation of this mandate. This indicates that the states may require more time to comply with national NH policy. A longer period of observation in future studies may be necessary to detect this. The combined effect of agency determinants on resolution of C to G level deficiencies was significant in all three years; however, only in Year 3 was state resolution of H to L level deficiencies affected by agency variables. That state agency resources may have a greater effect on less serious NH deficiencies is potentially important. This suggests that states may place too high a priority on minor violations that are easy to resolve (often related to facility maintenance issues) and too low a priority on the major quality of care problems that are most harmful to NH residents. Greater total agency funding per NH bed led to faster resolution of non-actual harm deficiencies (in Years 2 and 3), as was predicted in Hypothesis 2b. The greater the number of years of experience of state surveyors, the longer was the time needed to resolve C to G level deficiencies (Years 1 and 2) and H to L level deficiencies R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 5 (Year 3). This result was unexpected. Whether experienced surveyors were more thorough in ensuring NH problems were corrected, or simply slower, could not be determined from the information available. Surprisingly, greater agency funding led to slower resolution of H to L level deficiencies in Year 3. That greater agency funding would lead to faster resolution of minor problems but slower resolution of major ones suggests the disproportionate influence of Alaska, the state with the most abundant funding. Alaska fixed minor problems quickly (usually on the same day they were detected), while major problems remained for months, possibly due to the distance surveyors have to travel to verify correction. Agency determinants did not affect change in repetitive scheduling or in the resolution of H to L deficiencies, but they did lead to increased speed in the resolution of C to G level deficiencies from Year 2 to 3. This suggests that while the states have been slow in implementing federal initiatives, the mandate for surveying agencies to resolve non-actual harm deficiencies quickly has gained momentum. An important issue for future studies is whether state agencies begin to treat actual harm deficiencies with similar dispatch. Legislative Determinants Extent o f State Nursing Home Regulation The impact of legislative determinants on the volume of state deficiency citations was negligible in each of the three years of observation, despite the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 6 expectations that Democratic control, more professional legislators and the presence of an aging or LTC committee in the state legislature would all lead to more citations per 100 NH beds. The legislature did not sway state agencies in overall citation issuance in any one year. In terms of severity, the presence of an aging or LTC committee in the state legislature led to higher proportion of actual harm citations in at least one year of the study, as was predicted in Hypothesis 3c. However, a more professional legislature resulted in a lower proportion of severe citations, in contrast to Hypothesis 3b, which predicted that greater professionalism would correspond with greater severity in regulatory activity. Perhaps one consequence of having a more engaged state legislature is that fewer high-profile problems in the state’s regulated industries (like widespread instances of severe abuse and neglect in NHs) occur in the first place, Democratic control of the legislature did not lead to increased citation severity, as predicted in Hypothesis 3a. In terms of change in citation volume, a change in party control of a legislative chamber from Republican to Democratic led to an increase in citation volume from year to year, in contrast to the nonsignificant effect of partisan control of the legislature during each year. This finding supports Hypothesis 3 a. However, the effects of other legislative variables on change in citation volume and severity were non-significant. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 7 Effectiveness o f State Nursing Home Regulation Legislative determinants did not explain state variance in repetitive scheduling of annual inspections. Among individual measures, the presence of an aging or LTC committee predicted fewer repetitive state inspections (in Year 3). This finding was not surprising, as states with a regular forum for aging issues were expected to be the ones most likely to comply with a federal mandate designed to improve NH inspections. However, greater legislative professionalism led to more repetitive inspections; this result was not anticipated. Perhaps the ability of professional legislatures to influence the activities of state regulatory agencies was overestimated at the beginning of this research. There was no effect of legislative determinants on deficiency resolution (non actual harm and actual harm) in the three years of observation. In Year 2, Democratic control led to faster resolution of H to L level deficiencies, as was predicted. On the whole, however, legislative influence on the speed of deficiency resolutions was minimal in each of the three years of observation. This may suggest that legislators are reluctant to become involved in the day-to-day operations of the state surveying agencies. Change in two outcome measures was predicted by legislative determinants. First, an aging or LTC committee decreased repetitive state inspections from Year 2 to Year 3, while legislative professionalism increased it. This mirrors the previously described findings for repetitive state inspections in individual years and posits an aging or LTC committee as a catalyst for stronger regulation on behalf of NH R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 8 residents. Second, a change in party control to Democratic led to slower resolution of H to L level deficiencies over the three-year period. This result was the opposite of the effect that Democratic control had in Year 2. It is not clear whether a Democratic-controlled legislature leads to more effective regulation. Interest Group Determinants Extent o f State Nursing Home Regulation The finding that interest group determinants did not predict either the volume or the severity of state NH deficiency citations was surprising, especially considering the influence of the NH industry, the emotional appeal of groups that advocate on behalf of NH residents and their families, and the access that both interest group coalitions have to the state agencies due to their presence in the state capitals. In fact, only one interest group variable had a significant effect on the extent of NH regulation, and not in the direction that was predicted. Greater aging group influence led to a lower proportion of severe citations, which belied Hypothesis 4b. This was surprising because states in which seniors have greater clout were expected to have more frequent NH regulatory activity. This finding suggests instead that aging advocacy group influence decreases the likelihood of serious problems occurring in the first place and that it is the NHs, not the state surveying agencies, which are impacted the most by aging interest groups. Interest group determinants did not predict change in the volume or severity of state NH deficiency citations. This was not surprising. As was true of the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 2 9 external determinants, the measures operationalized for the interest group determinants (the Hrebenar-Thomas scale for influence, presence of offices in the state capital city for access) were unchanged over the three-year period. Effectiveness o f State Nursing Home Regulation Collectively, interest group determinants did not explain the effectiveness of state NH regulation in any of the three years. The one variable that impacted an outcome belied earlier expectations. Greater aging advocacy group influence led to more repetitive state inspections in Year 3. This finding is another example of the often counterintuitive nature of these results. Iowa, where senior citizens are described as an influential interest group (Thomas & Hrebenar, 2003), had the most repetitive scheduling of state inspections. Due to the unexpected results in states such as Iowa, change in the implementation of this federal mandate was opposite to what was expected, with influential senior citizens leading to an increase in repetitive state inspections from Year 2 to Year 3. No other interest group variable affected change in the repetitive scheduling of state inspections. This may be an indication that the deleterious effect of predictable inspections on the quality of NH regulation is not yet an issue of concern to aging advocacy groups. Another surprising finding was that aging advocacy group influence led to an increase in the time needed for deficiency resolution from year to year, while NH industry influence led to a decrease. This was exactly the opposite of what was R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 0 hypothesized. One conclusion that can be drawn from is that the NH industry also has a vested interest in fast removal of deficiencies, due to its financial stake. The industry may apply pressure to state agencies to quickly resolve citations and restore a clean bill of health to a facility. Gubernatorial Determinants Extent o f State Nursing Home Regulation The effect of the governors on the extent of regulatory activity by state NH surveying agencies was practically non-existent. Democratic governors did lead to a greater volume of state deficiency citations in Year 1; this was expected, as Democrats were expected to be more supportive of regulatory activity. However, the otherwise non-significant effects of gubernatorial variables on extent during each year of the study and change in these measures over the three- year period indicate that the governors are not important players in state NH policy. The explanations for interstate variance in the extent of NH regulation must lie outside the gubernatorial realm. Effectiveness o f State Nursing Home Regulation Gubernatorial measures had no effect on repetitive annual inspections in either year in which this was measured (Year 2 or Year 3). However, in terms of change from year to year in this outcome, a change from a Republican to a Democratic governor between Years 2 and 3 decreased repetitive scheduling. This R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 1 finding supported Hypothesis 5a, which predicted that Democratic governors would more strongly support state implementation of federal initiatives designed to strengthen NH regulation. In Year 2, the effect of the governors on state resolution of non-actual harm deficiencies was significant. Two individual variables were significant; however, these two measures produced conflicting results where deficiency resolution was concerned. A Democratic governor resulted in faster resolution of C to G level deficiencies, and a change to a Democratic governor reduced the time needed to correct these problems from one year to the next. However, the presence of NH policy in the governor’s SOS address unexpectedly led to slower resolution of C to G level deficiencies. Perhaps NH policy is higher on the governor’s agenda in the states with poor performing state surveying agencies. A gubernatorial mention of NH policy in the SOS address also resulted in an increase in repetitive annual inspections from Year 2 to Year 3 and in slower resolution of non-actual harm deficiencies in Year 2. Both findings contradicted the hypothesis presented earlier in the study that NH policy on the governor’s agenda would lead to more effective enforcement. It is difficult to determine why a governor who prioritizes NH policy would hinder, rather than help, a state surveying agency in fulfilling its federal requirements. Finally, in Year 3, the greater the institutional power of the governor, the slower was the state’s resolution of H to L level deficiencies. The unfortunate R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 232 implication of this finding is that powerful governors do not wield this clout on behalf of NH residents impacted by the most serious quality-of-care problems. Theoretical Models Volume of State Nursing Home Deficiency Citations HLM estimates revealed that the theoretical models proposed in this study were generally successful in explaining the volume of state NH deficiency citations. Controlling for factors found in multiple regression to impact citation volume, three individual measures continued to stand out as predictors. A higher proportion of college-educated residents in a state led to a lower number of citations per NH bed. Conversely, greater total state agency funding per NH bed led to higher citation volume. Again, while the latter result was expected, the former was not. The hypothesis that greater state agency funding would lead to more frequent regulatory activity has been supported, while the prediction that higher education would lead to greater citation volume has not been borne out by this study. Inexperienced state surveyors continued to lead to greater citation volume, contrary to expectations. This is again attributed to the confounding effect of NH quality in evaluating NH enforcement noted in previous research: does higher citation volume reflect poorly performing NHs or particularly vigorous state regulation (Walshe & Harrington, 2002; Wiener, 2003)? Finally, three variables found in previous models to impact citation volume during each year of observation did not have significant effects when controlling for R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 3 other key external and internal determinants in the integrated model. These were proportion of the oldest-old, political culture, and political party of the governor, although a larger proportion of oldest-old did lead to an increase in citation volume from Year 2 to Year 3. For the most part, however, citation volume was driven by a state’s economic development and by the resources of the state surveying agencies. Severity of State Nursing Home Deficiency Citations The theoretical model also provided a good fit for data related to citation severity. This finding, and that of citation volume, suggests that the integrated models proposed in this dissertation are useful in explaining the extent of state NH regulation. In the theoretical model for citation severity, three individual measures had significant effects on citation severity. As observed previously, higher levels of education led to greater citation severity, a result that was predicted at the beginning of this study. The presence of an aging or LTC committee in the state legislature continued to have the anticipated effect, greater citation severity, in Year 3. However, as observed previously, greater legislative professionalism led to a smaller proportion of severe citations in Year 3, contrary to expectations. One can conclude from the findings of citation severity that: 1) political factors do not affect the extent of state NH regulation, and 2) that education is the most important predictor among socioeconomic factors. The effects of three measures (urbanization, state proportion of total state agency funding and aging advocacy group influence) on citation severity were diminished in the integrated R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 4 model. None of these factors affected citation severity during any of the three years, and only one (aging advocacy influence) affected change in citation severity from year to year. As previously described, the finding that aging advocacy group influence led to decreased citation severity suggests that senior clout leads not to stricter enforcement, as was hypothesized, but perhaps to better NH care. Repetitive Scheduling of Annual State Nursing Home Inspections HLM revealed that the theoretical model fit the data on repetitive scheduling of state inspections better than the unconditional model. Three measures (legislative professionalism, presence of an aging or LTC committee and aging advocacy group influence) predicted change in state implementation of this federal mandate from Year 2 to Year 3. However, only one of these effects was in the direction that was anticipated. While an aging or LTC committee improved state implementation, greater legislative professionalism and aging advocacy group influence led to a decline in state effectiveness on this measure from Year 2 to Year 3, defying the initial hypothesis. Proportion of oldest-old did not affect the predictability of state inspections in the theoretical model. These unexpected findings are due in part to the surprisingly poor performance by states with large aged populations, highly professional legislatures and/or strong aging advocacy groups on this outcome measure. For example, state inspections were most repetitive in Iowa and North Dakota, two states with large and influential senior populations. These results also suggest that an aging or LTC R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 5 committee can be crucial in ensuring that state agencies comply with federal directives designed to improve NH regulation. Perhaps the creation of such committees in places where they are currently absent (such as Iowa and North Dakota) would help encourage the surveying agencies in these states to avoid repetitive scheduling of annual NH inspections. Resolution of C to G Level Deficiencies The theoretical model significantly predicted state resolution of C to G level deficiencies as well as change in the time needed by the states to resolve these deficiencies from year to year. State agency variables were the most significant determinants in this integrated model. As expected, greater total funding for state surveying agencies led to faster correction of non-actual harm problems in the three years of observation. More experienced state surveyors led to slower resolution of non-actual harm deficiencies, contrary to earlier hypotheses. As previously described, it is unclear whether this reflects diligent work, or simply slowness, on the part of experienced surveyors. Finally, the effects of urbanization, proportion of the oldest-old, a Democratic governor and NH policy on the governor’s agenda were non-significant in the theoretical model. These findings suggest that state agency variables, more than any other determinants, predict the speed with which the states correct minor NH problems. In particular, state agencies with greater financial resources are better R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 236 equipped to quickly dispatch the non-actual harm deficiencies that represent the majority of state NH citations. Resolution of H to L Level Deficiencies The theoretical model successfully predicted state resolution of H to L deficiencies, but HLM was not able to determine whether this model explained change in correction of these problems. This may be attributed to the small number of H to L level deficiencies in many states across the years of observation. In the integrated model, only one variable predicted the swiftness with which states correct serious problems. Total state agency funding per NH bed led to slower resolution of actual harm problems. This can be attributed to the outsized influence of Alaska. The other agency measure analyzed in the theoretical model (surveyor experience) did not predict faster resolution of H to L level deficiencies. Similarly, a Democratic legislature did not lead to faster correction of actual harm problems as expected. Finally, the connection between aged proportion and faster deficiency resolution was mitigated in the integrated model. These nonsignificant findings suggest that variance among states in resolution of actual harm deficiencies is largely explained by factors not considered in the present study. Refinement of this theoretical model appears necessary. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 237 Section Summary HLM estimates suggested that four theoretical models proposed in this study (volume of state NH deficiency citations, severity of state NH deficiency citations, repetitive scheduling of annual state inspections and resolution of C to G level deficiencies) provided a good fit for both the outcome measure and change in the outcome measure. One model (resolution of H to L level deficiencies) did not fit the data analyzed in this study, particularly in regard to change in this outcome measure. The results of HLM suggest that while the models proposed in this study are useful in predicting the extent of state nursing home regulation, little of the interstate variation in effectiveness of implementing CMS mandates can be explained at this stage. It is possible that as implementation of these two federal initiatives is observed over additional years, the factors that predict effectiveness will begin to emerge. Revisions to these theoretical models are recommended. Dissertation Summary This dissertation assessed the quality of NH regulation across the 50 states through the extent and effectiveness of state enforcement. Extent of regulatory activity, long considered the “gold standard” for measuring state agency performance, fails to fully separate the regulated from the regulator, and may falsely identify states with frequent minor citations as “poor” performers. Effectiveness of state implementation of CMS mandates is also a necessary measure because it enables the agencies themselves to be evaluated. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 8 The results of this dissertation support findings in previous studies that the state NH surveying agencies are inconsistent in the extent of the deficiency citations they issue. The present study also reveals uneven state performance in effectively implementing federal mandates that are designed to improve NH regulation. Moreover, contradictory findings across measures of extent and effectiveness are often found within the same state. Consequently, it is difficult to define a “good” or “bad” state based on one measure of extent or effectiveness. At the beginning of this research, state NH regulation was believed to be shaped by factors both external and internal to the NH policy subsystem. Multiple regression was used to identify external and internal determinants of state NH regulation. For each outcome measure, a theoretical model was constructed comprising factors found in the previous five models to predict the extent or effectiveness of NH regulation and tested using HLM. This provided a parsimonious way to consider a broad range of determinants shape state NH regulation. This evaluation led to three general conclusions. First, factors related to the external state environment (particularly education) and the governor are the most important predictors of the extent of state NH regulation. Second, factors related to the state surveying agencies and state legislatures are the most important predictors of the effectiveness of state NH regulation. Third, theoretical models integrating variables external and internal to the state NH policy ACF are useful in explaining interstate variation in NH regulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 3 9 Limitations Although this dissertation constructed and evaluated novel determinants of state NH regulatory activity, the value of many of these variables has not yet been established. Specifically, do legislative, interest group and gubernatorial variables truly affect state NH policy, or does the inclusion of these factors further muddy the waters? Further testing of the theoretical models proposed in this study may be necessary in order to identify the state-level factors that impact the actions of state NH inspectors. In addition, results of this study indicate that these enforcement actions may tell less about the regulating agency and more about the regulated industry than is true of other regulatory areas. For this reason, the factors that distinguish the NH industry from state to state (for-profit vs. non-profit, chain vs. free-standing, etc.) perhaps needed to be taken into account here more than has been the norm in other areas of regulatory policy research, in which variables related to the regulated industry have been marginalized. Finally, the duration and timing of the period of observation limit the conclusions that can be drawn from this study. The three years of state surveying activity sampled (1999-2002) may represent an interval too short and too soon after the issuance of the CMS NH initiatives of interest in 1997 to assess state compliance with these federal mandates. Further analysis of these outcome measures over a longer period of time is needed to more fully understand how well the states implement national NH policy. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 0 REFERENCES AARP. (2004). AARP in Your State. Available on the World Wide Web at http://www.aarp.org/states. Aeschleman, Heather K. (2000). White world of nursing homes: The myriad barriers to access facing today’s elderly minorities. Elder Law Journal, 8 (2): 367- 391. Allison, P.D. (1999). Multiple Regression: A Primer. Thousand Oaks, CA: Pine Forge Press. Alzheimer’s Association. (2004). Find Your Chapter. Available on the World Wide Web at http ://www. alz.org/findchapter.asp. American Association of Homes and Services for the Aging (AAHSA). (2004). AAHSA State Association Partners. Available on the World Wide Web at http://www2.aahsa.org/Document/Display.asp?MsgID==8-@- MC7152003123735. eml&SC=Taxonomy/Categories/About%20AAHSA/Contacts% 20and%20Directories/Contacts%20and%20Directories. American Health Care Association (AHCA). (2004). State Affiliates o f the American Health Care Association. Available on the World Wide Web at http://www.ahca.org/about/pubstate.htm. Angelelli, J., Mor, V., Intrator, O., Feng, Z., & Zinn, J. (2003). Oversight of nursing homes: Pruning the tree or just spotting bad apples? The Gerontologist, 43 (Special Issue II): 67-75. Barrilleaux, C. (1999). Governors, bureaus, and state policymaking. State and Local Government Review, 31 (1): 53-59. Beer, S.H. (1977). Political overload and federalism. Polity, 10 (1): 5-17. Begun, J.W., Crowe, E.W., & Feldman, R. (1981). Occupational regulation in the states: A causal model. Journal o f Health Politics, Policy and Law, 6 (1): 229-254. Berry, F., & Berry, W. (1992). Tax innovation in the states: Capitalizing on political opportunity. American Journal o f Political Science, 36: 715-742. Beyle, T. (2002). Governors: Elections, powers and priorities. In The Council of State Governments (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 1 Beyle, T. (2003). The governors. Chapter 7 in V. Gray & R.L. Hanson (Eds.), Politics in the American States: A Comparative Analysis. (8th Edition). Washington, DC: CQ Press. Blomquist, W. (1991). Exploring state differences in groundwater policy adoptions, 1980-1989. Publius, 2P. 101-115. Blomquist, W. (1999). The policy process and large-N comparative studies. Chapter 8 in P. Sabatier, Ed., Theories o f the Policy Process. Boulder, CO: Westview Press. Boeckelman, K. (1991). Political culture and state development policy. Publius, 2P . 49-62. Bohmstedt, G.W., & Knoke, D. (1994). Statistics for Social Data Analysis (3rd ed.). Itasca, IL: F.E. Peacock Publishers, Inc. Breyer, S. (1982). Regulation and its Reform. Cambridge, MA: Harvard University Press. Brown, R.S. (2002). Trends in state government progress in environmental protection. In The Council of State Governments (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. Browne, W.P. (1987). Some social & political conditions of issue credibility: Legislative agendas in the American states. Polity, 20 (2): 296-315. Bmdney, J.L., & Hebert, F.T. (1987). State agencies and their environments: Examining the influence of important internal actors. Journal o f Politics, 49\ 186- 206. Bryk, A.S., & Raudenbush, S.W. (1992). Hierarchical Linear Models: Applications and Data Analysis Methods. Newbury Park, CA: Sage Publications, Inc. Burwell, B. (2001). Medicaid long-term care expenditures in Fiscal Year 2000. The Gerontologist, 40 (5): 687-691. Centers for Disease Control and Prevention (CDC). (2003). Health, United States, 2003. Hyattsville, MD: National Center for Health Statistics. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 2 Centers for Medicare and Medicaid Services (CMS). (2003a). Nursing Home Compare. Available on the World Wide Web at http://www.medicare.gov.NHCompare/home.asp. Centers for Medicare and Medicaid Services (CMS). (2003b). Annual Performance Plan/Annual Performance Report. Available on the World Wide Web at http://www.cms.hhs.gov/about/performanceplan/APP2004.pdf. Centers for Medicare and Medicaid Services (CMS). (2003c). State Survey Agency Budget/Expenditure Reports. Obtained from the Centers for Medicare and Medicaid Services. Coble, R., & Watts. S. (2002). Governance and Coordination o f Public Higher Education in All 50 States. Raleigh, NC: North Carolina Center for Public Policy Research. Coggbum, J.D., & Schneider, S.K. (2003). The quality of management and government performance: An empirical analysis of the American states. Public Administration Review, 63 (2): 206-213. Cohen, M., Miller, J., & Weinrobe, M. (2001). Patterns of informal and formal caregiving among elders with private long-term care insurance. The Gerontologist, 41 (2): 180-187. Coleman, B. (1991). Nursing Home Reform Act o f1987: Provisions, Policy, Prospects. Boston, MA: Gerontology Institute, University of Massachusetts at Boston. Corazzini, K. (2003). How state-funded home care programs respond to changes in Medicare home health care: Resource allocation decisions on the front line- Public Policy Impact. Health Services Research, October 2003. Electronic journal retrieved May 3, 2004 from http ://www. findarticles. com/ cf_0/m4149/5_3 8/110461721/print.jhtml. Council of State Governments, The (CSG) (Ed.). (2000). The Book o f the States. Lexington, KY: The Council of State Governments. Day, C.L. (1990). What Older Americans Think. Princeton, NJ: Princeton University Press. DiLeo, D. (1997). Dynamic representation in the United States: Effects of the public mood on governors’ agendas. State and Local Government Review, 29 (2): 98-109. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 3 Dye, T. R. (1966). Politics, Economics, and the Public: Policy Outcomes in the American States. Chicago, IL: Rand McNally & Company. Elazar, D.J. (1972). American Federalism: A View from the States. New York: Thomas Crowell. Elazar, D.J. (1984). American Federalism: A View from the States (3rd Edition). New York: Harper & Row. Elling, R.C. (2003). Administering state programs: Performance and politics. Chapter 9 in V. Gray &, R.L. Hanson (Eds.), Politics in the American States: A Comparative Analysis. Washington, DC: Congressional Quarterly Books. Erikson, R.S., Mclver, J.P., & Wright, G.C., Jr. (1987). State political culture and public opinion. American Political Science Review, 81 (3): 797-813. Eshbaugh-Soha, M., & Meier, K.J. (2003). Economic and social regulation. Chapter 13 in V. Gray & R.L. Hanson (Eds.), Politics in the American States (8th Ed.), Washington, DC: CQ Press. Gerber, B.J., and Teske, P. (2000). Regulatory policymaking in the American states: A review of theories and evidence. Political Research Quarterly, 53 (4): 849- 886 . Gilligan, T., & Krehbiel, K. (1987). Collective decision-making and standing committees: An informal rationale for restrictive amendment procedures. Journal o f Law, Economics, and Organization, 3: 287-335. Gormley, W.T., Jr. (1983). Regulatory issue networks in a federal system. Paper presented at the Annual Meeting of the American Political Science Association, Chicago, IL, September 4, 1983. Gormley, W.T., Jr. (1986). The representation revolution: Reforming state regulation through public representation. Administration & Society, 18 (2): 179-196. Gormley, W.T., Jr. (1987). Intergovernmental conflict on environmental policy: The attitudinal connection. Western Political Quarterly, 40 (2): 285-304. Gormley, W.T., Jr. (1996). Accountability battles in state administration. Chapter 8 in C.E. Van Horn (Ed.), The State o f the States (3rd Edition). Washington, DC: Congressional Quarterly, Inc. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 244 Grabowski, D.C. (2001). Does an increase in the Medicaid reimbursement rate improve nursing home quality? Journal o f Gerontology: Social Sciences, 56B (2): S84-S93. Graddy, E., & Nichol, M.B. (1990). Structural reforms and licensing board performance. American Politics Quarterly, 18 (3): 376-400. Gray, V. (2003). The socioeconomic and political context of states. Chapter 1 in V. Gray & R.L. Hanson (Eds.), Politics in the American States (8th Edition). Washington DC: CQ Press. Gray, V., & Hanson, R.L. (Eds.). (2003). Politics in the American States (8th Edition). Washington, DC: CQ Press. Gray, V., & Lowery, D. (1999). Interest representation in the states. Chapter 11 in R.E. Weber & P. Brace (Eds.), American State and Local Politics: Directions for the 21st Century. New York: Seven Bridges Press, LLC. Greenberg, M.R., & Amer, S. (1989). Self-interest and direct legislation: Public support of a hazardous waste bond issue in New Jersey. Political Geography Quarterly, 8 (1): 67-78. Hamm, K.E., & Moncrief, G.F. (2003). Legislative politics in the states. Chapter 6 in V. Gray & R.L. Hanson, Politics in the American States (8th Edition). Washington DC: CQ Press. Hanson, R.L. (1991). Political cultural variations in state economic development policy. Publius, 21\ 63-81. Hanson, R.L. (2003). Intergovernmental relations. Chapter 2 in V. Gray & R.L. Hanson (Eds.), Politics in the American States: A Comparative Analysis. (8th Edition). Washington, DC: CQ Press. Harrington, C., & Carrillo, H. (1999). The regulation and enforcement of federal nursing home standards, 1991-1997. Medical Care Research and Review, 56 (4): 471-494. Harrington, C., Mullan, J.T., & Carrillo, H. (2004). State nursing home enforcement systems. Journal o f Health Politics, Policy and Law, 29 (1): 43-74. Harrington, C., Swan, J.H., Nyman, J.A., & Carrillo, H. (1997). The effect of certificate of need and moratoria policy on change in nursing home beds in the United States. Medical Care, 35 (6): 574-588. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 5 Harrington Meyer, M. (2001). Medicaid reimbursement rates and access to nursing homes. Research on Aging, 23 (5): 532-551. Heclo, H. (1978). Issue networks and the executive establishment. In A. King (Ed.), The New American Political System. Washington, DC: American Enterprise Institute. Hembree, T. (2002). Future challenges of state emergency management. In The Council of State Governments (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. Holzer, M. (2002). Performance measurement and improvement in state agencies. In The Council of State Governments (Ed.), The Book o f the States. (V olume 34). Lexington, KY: The Council of State Governments. Hrebenar, R.J., & Thomas, C.S. (1993). Interest Group Politics in the Northeastern States. University Park, PA: The Pennsylvania State University Press. Hwang, S-D, & Gray, V. (1991). External limits and internal determinants of state public policy. Western Political Quarterly, 44 (2): 277-298. Institute of Medicine (IOM). (1986). Improving the Quality o f Care in Nursing Homes. Washington, DC: National Academy Press. Institute of Medicine (IOM). (2001). Improving the Quality o f Long-Term Care, Washington, DC: National Academy Press. Jaccard, J., Turrisi, R., & Wan, C.K. (1990). Interaction Effects in Multiple Regression. Newbury Park, CA: Sage Publications, Inc. Jacoby, W.G., & Schneider, S.K. (2001). Variability in state policy priorities: An empirical analysis. Journal o f Politics, 63 (2): 544-568. Jennings, E.T. Jr. (1979). Competition, constituencies, and welfare policies in American states. American Political Science Review, 73 (2): 414-429. Kane, R.L. (1995). Long-term care in the United States: Problems and promise. Chapter 11 in T.R. Marmor, T.M. Smeeding, and V.L. Greene (Eds.), Economic Security and Intergenerational Justice: A Look at North America. Washington, DC: The Urban Institute Press. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 246 Kane, R.L. (1998). Managed care as a vehicle for delivering more effective chronic care for older persons. Journal o f the American Geriatrics Society, 46 (8): 1034- 1039. Kaskie, B. (1998). State Legislation Concerning Individuals with Dementia: An Evaluation o f Three Theoretical Models o f Policy Formation. Ph.D. Dissertation. University of Southern California, Graduate School. Kelly, C.M., & Liebig, P.S. (2003). Nursing Homes. In Lois A. Vitt (Ed.), Encyclopedia o f Retirement and Finance. Westport, CT: Greenwood Press. Key, V.O. Jr. (1949). Southern Politics in State and Nation. Knoxville, TN: The University of Tennessee Press. Kim, J-O., & Mueller, C.W. (1978). Introduction to Factor Analysis: What It Is and How To Do It. Newbury Park, CA: Sage Publications. Kincaid, J. (2002). State-Federal Relations: Continuing Regulatory Federalism. In The Council of State Governments (Ed.), The Book o f the States. Volume 34. Lexington, KY: The Council of State Governments. Kingdon, J.W. (1995). Agendas, Alternatives, and Public Policies. (2n d Edition). New York: Addison Wesley Longman, Inc. Kirst, M.W., Meister, G., & Rowley, S.R. (1984). Policy issue networks: Their influence on state policymaking. Policy Studies Journal, 13 (2): 247-264. Kish, L. (1965). Survey Sampling. New York: John Wiley & Sons, Inc. Lake, L.M. (1983). The environmental mandate: Activists and the electorate. Political Science Quarterly, 98 (2): 215-233. Lammers, W.W. (1989). Prospects for innovation in state policies and the elderly. Journal o f Aging and Social Policy, 1 (3-4): 37-66. Lammers, W.W., & Klingman, D. (1984). State Policies and the Aging: Sources, Trends, and Options. Lexington, MA: Lexington Books. Lester, J.P., & Bowman, A. O’M. (1989). Implementing environmental policy in a federal system: A test of the Sabatier-Mazmanian model. Polity, 21 (4): 731-753. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 7 Licari, M. J. (1998). The policy process. Chapter 2 in Meier, K., et al. (Eds.), Regulation and Consumer Protection: Politics, Bureaucracy and Economics. Houston: Dame. Liebig, P.S. (1992). Federalism and aging policy in the 1980s: Implications for changing interest group roles in the 1990s. Journal o f Aging & Social Policy, 4 (1/2): 17-33. Liebig, P.S. (1998). Gerontology 540: Social Policy and Aging (Fall 1998 lecture notes). Los Angeles: University of Southern California, Leonard Davis School of Gerontology. Lovrich, N.P., Daynes, B.W., & Ginger, L. (1980). Publius, 10 (2): 111-126. Lynk, W.J. (1981). Regulatory control of the membership of corporate boards of directors: The Blue Shield case. Journal o f Law & Economics, 24 (1): 159-173. Mazmanian, D., & Sabatier, P. (1981). Effective Policy Implementation. Lexington, MA: Lexington Books. Mazmanian, D., & Sabatier, P. (1989). Implementation and Public Policy. New York: University Press of America. Meier, K.J. (2000). Politics and the Bureaucracy: Policymaking in the Fourth Branch o f Government. Fort Worth, TX: Harcourt College Publishers. Miller, D.Y. (1991). The impact of political culture on patterns of state and local government expenditures. Publius, 21: 83-100. Milward, H.B., & Francisco, R.A. (1983). Subsystem politics and corporatism in the United States. Policy and Politics, 11 (3): 273-293. Morehouse, S.M. (1981). State Politics, Parties and Policy. New York: CBS College Publishing. Morgan, D.R., & Watson, S.S. (1991). Political culture, political system characteristics, and public policies among the American states. Publius, 21: 31-48. Mullan, J.T., & Harrington, C. (2001). Nursing home deficiencies in the United States: A confirmatory factor analysis. Research on Aging, 23 (5): 503-531. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 8 Nathan, R.P. (1990). Federalism: The great “compromise”. Chapter 8 in A. King (Ed.), The New American Political System (2n d Version). Washington, DC: AEI Press. National Conference of State Legislatures (NCSL). (2003). Directory o f State Legislatures. Available on the World Wide Web at http://www.ncsl.org/programs/pubs/eronline.htm. National Governors Association (NGA). (2003). State o f the State Addresses. Available on the World Wide Web at http://www.nga.org/nga/1,1169,C_SEARCH,00.html. Netting, F.E., Huber, R., Paton, R.N., & Kautz, James R. III. (1995). Elder rights and the long-term care ombudsman program. Social Work, 40 (3): 351-357. Nice, D.C. (1987). Federalism: The Politics o f Intergovernmental Relations. New York: St. Martin’s Press. Omnibus Budget Reconciliation Act o f1987. (1987). Public law 100-203. Subtitle C: Nursing home reform. Washington DC, December 22. Osborne, J. W. (2000). Advantages of hierarchical linear modeling. Practical Assessment, Research & Evaluation, 7 (1). Electronic journal retrieved April 30, 2004 from http ://PAREonline.net/getvn. asp?v=7&n= 1. O’Toole, L.J., Jr. (2000). American intergovernmental relations: An overview. In L.J. O’Toole, Jr. (Ed.), American Intergovernmental Relations (3rd Edition). Washington, DC: CQ Press. Paddock, J. (1997). Political culture and the partisan style of state party activists. Publius, 27 (3): 127-132. Parsons, T. (1947). Introduction to M. Weber, The Theory o f Social and Economic Regulation (Translated by A.M. Henderson & T. Parsons). New York: The Free Press. Polivka, L., Salmon, J.R., Hyer, K., Johnson, C., & Hedgecock, D. (2003). The nursing home problem in Florida. The Gerontologist, 43 (Special Issue II): 7-18. Radcliff B., & Saiz, M. (1998). Labor organization and public policy in the American states. Journal o f Politics, 60 (1): 113-125. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 4 9 Redfoot, D.E. (2003). Changing consumer: the social context of culture change in long-term care. Journal o f Social Work in Long-Term Care, 2 (1-2): 95-110. Ringquist, E.J. (1993). Does regulation matter? Evaluating the effects of state air pollution control programs. Journal o f Politics, 55(4): 1022-1045. Ritf L.G. (1974). Political culture and political reform: A research note. Publius, 4 (1): 127-134. Rosenthal, A., & Jones, R. (2002). Trends in state legislatures. In The Council of State Governments (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. Rowland, D. (2002). Medicaid: A future challenge for the states. In The Council of State Governments (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. Sabatier, P.A., & Jenkins-Smith, Hank C. (1999). The advocacy coalition framework: An assessment. Chapter 6 in P. A. Sabatier, (Ed.), Theories o f the Policy Process. Boulder, CO: Westview Press. Sabatier, P,A., & Mazmanian, D. (1983). Can Implementation Work? The Implementation o f the 1972 California Coastal Initiative. New York: Plenum Press. Sabato, L. (1983). Goodbye to Good-time Charlie: The American Governorship Transformed (2n d Edition). Washington, DC: CQ Press. Sato, H. (1999). The advocacy coalition framework and the policy process analysis: The case of smoking control in Japan. Policy Studies Journal, 27 (1): 28-44. Savage, R.L. (1981). Looking for political subcultures: A critique of the rummage- sale approach. Western Political Quarterly, 34 (2): 331-336. Schlager, E. (1995). Policy making and collective action: Defining coalitions within the advocacy coalition framework. Policy Sciences, 28: 243-270. Schneider, S.K. (1988). Intergovernmental influences on Medicaid program expenditures. Public Administration Review, 48 (4): 756-763. Schneider, S.K. (1993). Examining the relationship between public policies: AFDC and Medicaid. Public Administration Review, 53 (4): 368-380. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 5 0 Scholz, J.T., Twombly, J., & Headrick, B. (1991). Street-level political controls over federal bureaucracy. American Political Science Review, 85 (3): 829-850. Schroeder, L.D., Sjoquist, D.L., & Stephan, P.E. (1986). Understanding Regression Analysis: An Introductory Guide. Newbury Park, CA: Sage Publications. Shaffitz, J. M. (1974). Political culture: The determinant of merit system viability. Public Personnel Management, 3 (1): 39-43. Sharkansky, I. (1969). The utility of Elazar’s political culture: A research note. Polity, 11 (1): 36-63. Shelley, F.M. (1988). Structure, stability and section in American politics. Political Geography Quarterly, 7 (2): 153-160. Sigelman, L. (1976). The quality of administration: An exploration in the American states. Administration & Society, 8 (1): 107-144. Sigelman, L., & Smith, R.E. (1980). Consumer legislation in the American states: An attempt at explanation. Social Science Quarterly, 61 (1): 58-70. Squire, P. (1992). The theory of legislative institutionalization and the California Assembly. Journal o f Politics, 54 (4): 1026-1054. Squire, P. (2000). Uncontested seats in state legislative elections. Legislative Studies Quarterly, 25 (I): 131-146. Steuerle, C.E., & Bakija, J.M. (1994). Retooling Social Security fo r the 21st Century: Right and Wrong Approaches to Reform. Washington, DC: The Urban Institute Press. Stone, R. (2001). Providing long-term care benefits in cash: Moving to a disability model. Health Affairs, 20 (6): 96-108. Stonecash, J., & Hayes, S.W. (1981). The sources of public policy: Welfare policy in the American states. Policy Studies Journal, 9 (5): 681-698. Swan, J., Bhagavatula, V., Algotar, A., Seirawan, M., Clemena, W., & Harrington, C. (2001). State Medicaid nursing home reimbursement rates: Adjusting for ancillaries. The Gerontologist, 41 (5): 597-604. Teske, P.E. (1990). After Divestiture: The Political Economy o f State Telecommunications Regulation. Albany, NY: State University of New York Press. . R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 5 1 Thomas, C.S., & Hrebenar, R.J. (2003). Interest groups in the states. Chapter 4 in V. Gray & R.L. Hanson (Eds.), Politics in the American States (8th Ed.). Washington, DC: CQ Press. Thompson, F.J., & Scicchitano, M.J. (1987). State implementation and federal enforcement priorities: Safety versus health in OSHA and the states. Administration & Society, 19 (1): 95-124. U. S. Census Bureau (2003). United States Census 2000. Available on the World Wide Web at http://www.census.gov. U. S. Department of Health and Human Services, Center for Disease Control, Vital and Health Statistics (USDHHS). (2002). The National Nursing Home Survey: 1999 Summary. Hyattsville, MD: U. S. Department of Health and Human Services. U. S. Department of Health and Human Services, Office of Inspector General (USDHHS). (2003). Nursing Home Deficiency Trends and Survey and Certification Process Consistency. Washington, DC: U.S. Department of Health and Human Services. U. S. General Accounting Office (USGAO). (1998). California Nursing Homes: Care Problems Persist Despite Federal and State Oversight. Report to the Special Committee on Aging, U. S. Senate. GAO/HEHS-98-202. Washington, DC: U. S. General Accounting Office. U. S. General Accounting Office (USGAO). (1999a). Nursing Home Care: Enhanced HCFA Oversight o f State Programs Would Better Ensure Quality. Report to the Special Committee on Aging, U. S. Senate. GAO/HEHS-OO-6. Washington, DC: U. S. General Accounting Office. U. S. General Accounting Office (USGAO). (1999b). Nursing Homes: Additional Steps Needed to Strengthen Enforcement o f Federal Quality Standards. Report to the Special Committee on Aging, U. S. Senate. GAO/HEHS- Washington, DC: U. S. General Accounting Office. U. S. General Accounting Office (USGAO). (1999c). Nursing Homes: Proposal to Enhance Oversight o f Poorly Performing Homes Has Merit. Report to the Special Committee on Aging, U. S. Senate. GAO/HEHS-99-157. Washington, DC: U. S. General Accounting Office. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 5 2 U. S. General Accounting Office (USGAO). (2000). Nursing Homes: Sustained Efforts Are Essential to Realize Potential o f the Quality Initiatives. Report to Congressional Requesters. GAO/HEHS-OO-197. Washington, DC: U. S. General Accounting Office. U. S. General Accounting Office (USGAO). (2003). Nursing Home Quality: Prevalence o f Serious Problems, While Declining, Reinforces Importance o f Enhanced Oversight. Report to Congressional Requesters. GAO-03-561. Washington, DC: U. S. General Accounting Office. Van Lare, B. (2002). Emerging issues in welfare reform. In The Council of State Governments (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. Van Meter, D.S., & Van Horn C.E. (1975). The policy implementation process: A conceptual framework. Administration & Society, 6 (4): 445-488. Vladeck, B. (1980). Unloving Care: The Nursing Home Tragedy. New York: Basic Books. Walker, J. (1969). The diffusion of innovations across the American states. American Political Science Review, 63 (3): 880-899. Walshe, K. (2001). Regulating nursing homes: Are we learning from experience? Health Affairs, 20 (6): 128-144. Walshe, K., & Harrington, C. (2002). Regulation of nursing facilities in the United States: An analysis of resources and performance of state survey agencies. The Gerontologist, 42 (4): 475-486. Weber, M. (1947). The Theory o f Social and Economic Organization. (Translated by A. M. Henderson & T. Parsons). New York: The Free Press. Wiener, J.M. (2003). An assessment of strategies for improving quality of care in nursing homes. The Gerontologist, 43 (Special Issue II): 19-27. Willms, J.D. (1999). Basic concepts in hierarchical linear modeling with applications for policy analysis. Chapter 19 in G J. Cizek (Ed.), Handbook o f Educational Policy. San Diego: Academic Press. Wood, B.D. (1991). Federalism and policy responsiveness: The clean air case. Journal o f Politics, 53 (3): 851-859. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 5 3 Wood, E.F. (2002). Termination and Closure o f Poor Quality Nursing Homes: What are the Options? Washington, DC: AARP Public Policy Institute. Wright, G.C. Jr., Erikson, R.S., & Mclver, J.P. (1987). Public opinion and policy liberalism in the American states. American Journal o f Political Science, 31 (4): 980-1001. Wulf, H.S. (2002). Trends in state government finances. In The Council of State Governments, (Ed.), The Book o f the States. (Volume 34). Lexington, KY: The Council of State Governments. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Out -of -pocket health expenditures by older adults in relation to age, race, and insurance

PDF

Nurse executive leadership and the relationship to organizational commitment among nurses

PDF

The effects of case mix, hospital competition, and managed care penetration on hospital costs and revenues

PDF

Using network perspective to examine the organization of community -based elder care systems across four communities

PDF

The effects of Medi -Cal managed care on access to mammography screening for women residing in Los Angeles County

PDF

Service evaluation and quality improvement of a geriatric evaluation and management program

PDF

Physician profiling and clinical pathways: Combining the tools to change physician resource utilization

PDF

Marriage and health: The effect of marital status on health and the mechanisms by which marital status affects health among elderly people

PDF

Performance management in health care in Iceland

PDF

Toward a theory of networks and network effectiveness: An analytical framework for the study of multilevel networks in health care service delivery

PDF

Dual eligible beneficiaries and managed care: Ensuring consumer protection through state Medicaid contracts

PDF

Variations in physician practice patterns for eye care under the National Health Insurance of Taiwan

PDF

Health care finance reform and market failure at King/Drew Medical Center, Los Angeles: An analysis of pediatric social outcomes in an urban public safety -net hospital

PDF

Recruitment of family physicians into rural California: Predictors and possibilities

PDF

A quantitative and qualitative evaluation of the least burdensome provisions of the Food and Drug Administration Modernization Act of 1997

PDF

Evidence of age-based rationing of health care to the elderly: The effect of age and death on the use of health care services by elderly enrollees in an HMO

PDF

Evaluation of the Care Advocate Program: Bridging managed care and home community -based services

PDF

The role of educational legislation and policy in the California Gubernatorial Recall Election of 2003

PDF

Gambling with public health: How government officials brought Los Angeles County to the brink of disaster

PDF

The effects of East Asian values, coping styles, social support on physical health among Korean American caregivers

Asset Metadata

Creator Kelly, Christopher Michael (author)

Core Title The extent and effectiveness of nursing home regulation in the 50 states

Degree Doctor of Philosophy

Degree Program Gerontology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Gerontology,health sciences, health care management,OAI-PMH Harvest,Political Science, public administration

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-405346

Unique identifier UC11340793

Identifier 3145215.pdf (filename),usctheses-c16-405346 (legacy record id)

Legacy Identifier 3145215.pdf

Dmrecord 405346

Document Type Dissertation

Rights Kelly, Christopher Michael

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA