Taxpayer perceptions of complexity and the effect of complexity on reporting positions

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content TAXPAYER PERCEPTIONS OF COMPLEXITY AND THE EFFECT OF COMPLEXITY ON REPORTING POSITIONS by Valerie Colleen Milliron 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 (Business Administration) May 1984 UNIVERSITY OF SOUTHERN CAUFORNIA Tim GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CAUFORNIA 90089 This dissertation, written by .. ':l •. l.f.r.!Jt ..... C..9..tl.~.~D. ... Mi.l.l.!r.P..rJ ....................... . under the direction of M.r........ Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY ouuuouuuu..K-..u. ...... r;;//.ff•••oooououuo U Dean Date .... ~:!!!~~:.! ... ~.~.! ... ~?.~~ ....... . ::~]!;2~1..~~~ ........ : .................. .. ~ .. /...~ ............................. := ·······e~~~-~············ (/J,,p Urwl 'gtl nt' f)" ,.~q~·.,/3 ~ DEDICATION In memory of ALICE E. KINNEY (1902-1971) To Nana, Whose love and spirit prevail in the best of my work ii ACKNOWLEDGMENTS I am deeply indebted to my doctoral committee for theii guidance, Paul Watkins for the methodology, Stewart Karlinsky for the tax expertise, and Ted Mock for the "partner" review. In particular, my gratitude is extended to Paul for his generous tutoring and technical support. Special thanks also go to Sheila Watkins for her friendship and understanding. In addition, I wish to express sincere appreciation to my parents, Jim and Joyce Kinney, for their constant love and encouragement. To the love of my life - Richard Scott Milliron, no words are adequate to acknowledge the gratitude I feel. Very simply, I could not have completed this project except for Rich's financial, emotional, and word processing support! iii ABSTRACT This study provides professionals involved in tax administration, compliance and policy matters with insight into the meaning and influence of tax complexity. Although extensive research has been conducted on the magnitude of the tax gap, little has been done to ascertain why this gap exists. This research attempts to define and measure one suspected tax gap influence--tax complexity. This is a two-part study. Phase 1 is devoted to empirically defining tax complexity, Phase 2 to testing its effects. For both Phase 1 and Phase 2 testing, the author selected subjects randomly from taxpayers awaiting jury duty at the Los Angeles County Courthouse. Thirty subjects, taking an average of 70 minutes each, participated in the Phase 1 testing. Eighty-six subjects, averaging 20 minutes each, participated in Phase 2. The demographic data collected indicates that these 126 subjects represeqt a cross section of Los Angeles area taxpayers. Phase 1 utilizes multi-dimensional scaling to derive ar empirically based definition of tax complexity. This definition helps explain the relationship between complexit' and other concepts and provides a scientifically defensible base for testing the effect of complexity. Phase 2 addresses the issue of whether complexity influences one's tax reporting position. Testing four distinct tax cases, iv ~he author shows that complexity has a significant effect in jevery case. A secondary issue in Phase 2 is whether taxpayers tend !toward noncompliant reporting positions as measured by what ~hey perceive or would advise as correct. In all four cases !the mean for "what you would report" reflected a more !aggressive tax stance than "what you would advise," which is still more aggressive than "what you think is correct." And in three of the four cases, "what you think is correct" is ~efinitely more aggressive than the IRS tax position. What is especially disturbing from a tax administration perspective is that 56 of the 86 subjects (67%) selected reporting positions that deviated in their favor from their own perception of the correct reporting position in one or more of the four cases. Moreover, all population subgroups revealed an equal propensity to evade. v TABLE OF CONTENTS DEDICATION • •••••••••••••••••••••••••••••••••••••••• ii ACKNOWLEDGEMENTS •••••••••••••••••••••••••••••••••• iii ABSTRACT • •••••••••••••••••••••••••••••••••••••••••• i v LIST OF TABLES AND FIGURES ••••••••••••••••••••••• viii Chapter 1. INTRODUCTION ••••••••••••••••••••••••••••• 1 Significance of the Study ••••••••••••• 5 Problem ..•.•. ~ ...•..•....•.........•.. 6 Objective ............................. 7 Limitations ••••••••••••••••••••••••••• 9 Definitions •••••••••••••••••••••••••• 10 Organization of Remaining Chapters ••• 14 2. LITERATURE REVIEW ••••••••••••••••••••••• 16 Summary of Previous Tax Studies •••••• 16 Synthesis of Findings of Previous Tax Studies •••••••••••••••••••••••••• 23 Review of Methodology Literature ••••• 29 3. RESEARCH METHODOLOGY, PHASE ONE ANALYSIS AND RESULTS •••••••••• 34 overview ............................. 34 Research Questions ••••••••••••••••••• 37 Phase One Methodology •••••••••••••••• 38 Research Task .....•.................. 40 Data Gathering ••••••••••••••••••••••• 43 Subjects ••••••••••••••••••••••••••••• 48 Model Selection •••••••••••••••••••••• 49 Introduction to Dimension Interpretation ••••••••••••••••••••••• 54 Interpretation of Dimensions One and Two • ••••••••••••••••••••••••• 59 Results: Other Phase One Questions ••• 71 summary of Phase One Research Results ••••••••••••••••••••• 78 vi Chapter 4. TABLE OF CONTENTS (Continued) PHASE TWO ANALYSIS AND RESULTS •••••••••• BO Phase Two Methodology ••••••••••• ~ •••• BO Scenario Construction •••••••••.•••••• 82 The Subjects ••••••••••••••••••••••••• 83 The Statistical Procedures ••••••••••• BS Complexity Effect Results •••••••••••• 86 Influence of Complexity Dimensions ••• 89 Direction of the Complexity Influence •••••••••••••••••••••••••••• 98 Task Effects •••••••••••••••••••••••• 103 Subjects' Aggression Propensities ••• 104 Background•Questions •••••••••••••••• 113 Summary of Phase Two Research Results •••••••••••••••••••• 118 5. SUMMARY, CONCLUSIONS, AND EXTENSIONS ••• 120 Scope and Limitations ••••••••••••••• 133 Extensions •••••••••••••••••••••••••• 136 ~~I>~()~~E) •••••••••••••••••••••••••••••••••••••••••• 138 REFERENCES • ••••••••••••••••••••••••••••••••••••••• 141 APPENDIX A: List of the Thirteen Phase One Stimuli ••••••••••••••••••••• 146 APPENDIX B: Sample Phase One Test Instrument •••••• lSO APPENDIX C: Phase One Data •••••••••••••••••••••••• 163 APPENDIX D: Phase Two Scenarios ••••••••••••••••••• 178 APPENDIX E: Sample Phase Two Test Instrument •••••• 190 APPENDIX F: Phase Two Data •••••••••••••••••••••••• 197 APPENDIX G: Multivariate and Univariate Analysis •• 204 APPENDIX H: Directional Effects of Complexity ••••• 213 vii TABLES 1 Summary of Tax Evasion Factor Studies •••••••• !? 2 Phase One Property Fitting Adjective Descriptors •••••••.•••••••••••••••• 45 3 General Background Questions ••••••••••••••••• 47 4 Evaluation of the Four Multiscale Models ••••• 53 5 Property Fitting Correlations •••••••••••••••• 56 6 Analysis of Dimension D2 Stimuli ••••••••••••• 64 7 Property Fitting Ratings •...•..•••••••••••••• 66 8 Subject's Weighting of the Complexity Dimensions •••.•••••••••••••••••••• 73 9 Part 1: Summary Canonical Correlation Statistics Part 2: Correlations Between the Biographical/Attitudinal Variables and the Complexity Dimensions ••••••••.••.•••. 76 10 Correlations Among the 10 Background V a r i ab 1 e s •••••••••••••••••••••••••••••••••••• 7 8 11 Phase Two Research Design •••••••••••••••••••• 81 12 MANOVA Analysis of the Effect of Tax Complexity on Reporting Position •••••• 86 13 MANOVA Analysis of the Effect of Complexity Dimesnion(s) on Reporting Positions •••.••••••••.•••••.••..••••••••••••• 90 14 Univariate Analysis of the Effect of Complexity Dimension(s) on Reporting Positions ••••••••••••.••.....••.••• 92 15 Summary Tax Position Statistics ••••••••.•••• lll 16 Subject's Background and Attitudinal Ratings ..................................... 114 FIGURES 1 Tax Positions •••••..•.•.•••••••••••••••..•••• 13 2 Factors Affecting Tax Evasion •••••••.•.•••••• 25 3 Relationship of Complexity to the Other Factors ........•••••..•........•... 28 4 Research Methodology Flowchart ••••.•...•••••. 36 5 Configuration Plots for Dimensions Dl & D2 ••. 61 6 Configuration Plots for Dimensions D3 & D4 ••. 68 vii' CHAPTER 1 INTRODUCTION Although it is not entirely clear what tax complexity is, or the exact nature of its impact, tax complexity has been identified as a factor that may threaten the viability of the u.s. tax system for individuals. Most of the criticism seems to center on the idea that the complexity o~ the current tax system mitigates against its functioning a~ a "good" tax system, and thereby contributes to the so called tax gap. Going back to the original standard for a good tax, 1 complexity can be viewed as compromising Adam Smith's canons of convenience, economy, certainty, and equity (Smith, 1776). It appears that complexity contributes to the income tax system being both less than convenient, and less than economical for the majority of Americans. Yearly tax preparation involves millions of hours and billions of dollars. The Treasury Department estimates that the American public spends over 600 million hours a year just filling out tax forms, which amounts to almost three hours for every man, woman and child in the country (Hall and Rabushka, 1982, p. 7). Furthermore, less than half of all tax filers try to do their own returns. The majority are sufficiently overwhelmed by the system to seek outside assistance (Consumer Reports, 1976, p. 130). The monetary 1 ~ost of this assistance is also awesome. Hall and Rabushka ~stimate that "taxpayers bear some $9 to $10 billion in real posts for preparing and verifying their taxes, above and beyond what they actually pay in taxes" (p. 8). In addition, they calculate the loss to the economy from all ~he effort to avoid or evade taxation to be $50 billion or ~ore per year (p. 9). Moreover, those who do tackle the ~ask of tax preparation themselves often toil for hours to complete the necessary forms. Complexity also appears to obstruct Smith's canon of pertainty. Even conscientious professional advisors can't precisely determine the correct tax liability in many ~ituations. "A reasonably certain conclusion cannot in some ~nstances be determined despite diligent and expert research" (Ginsburg, 1976, p. 317). Of course not every citizen is diligent or competent in tax matters, which explains the number of errors detected by the Internal Revenue Service each year. But the public is not alone in ~heir propensity to err. A Ralph Nader Tax Reform Research ~roup created a tax schedule for a fictional couple with one ~hild and sent copies to 22 IRS offices throughout the ~ountry. Each office calculated a different tax liability, ranging from a high refund of $811.91 to a tax underpayment pf $52.14 (Hall and Rabushka, 1982, p. 5). Thus, in the present environment, not even the Internal Revenue Service 2 ~an guarantee tax precision, or certainty. (These results suggest that tax complexity might also affect a revenue agent's ability to detect noncompliance.) This failure to meet the standard of certainty seems of special concern to tax accountants. Accountants operate in a competitive environment, but they must spend considerable time and resources to develop the expertise to perform tax services. Yet clients may not be able to differentiate between those who do and do not have the requisite skills. This phenomenon (also known as "Gresham's Law") may tend to put good tax practitioners at a competitive disadvantage. To quote the Committee of the New York Bar, "Simple economics must cause the 'bad' to drive out the 'good'; if the good practitioner cannot achieve a reasonably certain conclusion, why pay the fees that are required to arrive at that uncertainty?" (New York Bar Association, 1972, p. 331). After all, the "bad" practitioner may be able to provide a speedier, less expensive answer. Finally, the quest for equity may also be impeded by the complexity of the tax system. A 1972 Government Accounting Office report concluded that more than one million people of limited education did not file tax return~ because they could not cope with the complexity of the tax laws and filing requirements. Complexity appears to mitigate against equal application of the law in other ways 3 too. In every bracket of 1ncome earners there ex1sts considerable variation in the amount of taxes paid because of the complex structure of the tax law (Pechman and Okner, 1974, p. 10). This variation in the taxes paid among otherwise equal income earners may not, however, be due onl1 to the structural features of the law. Not all errors favor the taxpayer, but there is concern that a significant number of taxpayers may take advantage of the complexity to bend the law to their benefit (New York State Bar Association, 1972). The Committee of the New York State Bar Association states that "in numerous doubtful areas, the practice of ta> law tends to descend to a level, not of what the law provides, equally applicable to all like taxpayers, but of what will be discovered in a particular return. Of course there is a smaller group of taxpayers who fare even better: the marginal group with less conscience in separating the legitimate doubt from the minimal doubt, who rely upon the complexity of the law and audit deficiencies to adopt a position for which there is very little or no basis--for these taxpayers the advantages of playing the tax lottery may be even greater" (p. 330). Thus, whether it is to be attributed to the structural complexities of the tax law, or to the fact that some taxpayers use tax complexity to bend the law for their own gain, complexity appears to decrease the equity of the system. 4 Significance of the Study This study, therefore, addresses an area of vital social, political and economic concern. There appears to be a trend toward both increasing complexity of the tax law anc decreasing taxpayer compliance, and whether the phenomena are related is an important national issue. 2 Honest taxpayers suffer economically as the spread of tax evasion turns the tax system into a means of redistributing wealth from the honest to the dishonest. This redistribution coulc have further consequences, if honest taxpayers become demoralized and withdraw support for the tax law. Experience in other countries shows that a lack of support for the tax law can become pervasive (Rose, 1981, p.21). In the long-run, the erosion of support for the tax system would threaten the ability of the government to raise revenue. As former IRS commissioner Jerome Kurtz says, "A decline in compliance of one-half of one percent in a tax system that collects $700 billion a year could produce greater losses than would likely be gained in all the government's debt collection programs" ( Wall Street Journal , September 8, 1982; p. 31). If complexity does discourage taxpayer cooperation, then it is an element of one of the biggest economic problems facing the country today: the gap between the revenue due and revenue 5 collected. The revenue losses from noncompl1ance are already staggering. The IRS estimates that $72 billion was lost in 1981. Furthermore, they predict that the compliance gap will widen yearly, projecting a shortfall of over $102 billion by 1985 (Congressional Joint Committee on Taxation, 1982, p. 28). This study, then, provides professionals involved in tax administration, compliance and policy matters with additional insight into the meaning and influence of tax complexity. Although extensive research has been conducted on the magnitude of the tax gap, little has been done to ascertain why this gap exists. This research attempts to define and measure one suspected tax gap influence--tax complexity. The effort promises to benefit tax accountants and lawyers, governmental officials, and other professionals committed to the goal of achieving a more convenient, economical, certain, and equitable tax system. Moreover, the author hopes that ultimately all taxpayers will benefit as increased knowledge and understanding are translated intc improvements in the present system. The Problem The term "complexity" has been used to describe variou~ ills affecting the tax system. Unfortunately, most analyst~ use the term in an intuitive sense, precluding any 6 SCl!entifJ.:C teStl.ng Of .]. ts l.mpact ~ :~r Col'isequent!y li l. t r l.S • not yetr~ossible to assess'the nature or .scope:of tax : cotnplexi~y;.• .·NO:£ is it possible, given the lack of adequate definition·j. to ,examine the a. inks between ··camplexi ty and, the tax· reporting· system .. · 'fhis study .addresses.xtwo related aspects:of~this prbblem~·;bhe ·taxpayers' definit~~n of •co~plexity" iR reference~to different·tax:situations~ and the effect of:this complexity on reporting·position selections in· -different tax situations.· : -~ ' ''f I Objectives : ~· This~ study underta:tes··to··build··toward·.a- t.~eory of taxpayer·bebavior~ Acconding to Ker1inger~ "A-tkeory is a set of interrelated~-cons-tzructs. (concepts},. definitions, and propositions that present•a. systematic~iew of phenomena by specifying_rela~ions among·:variables~-with. the purpose of explaining·~nd·predicting·t:be phenomena" (Kerlinger, 1913, p. 9t: .ThEbimmed:iiate objective eft-this, study is to· examine tax complexity. f Compdexity. has·,beea•!alluded to: in some ·tax evasion!s~ud~es,~and represents a possible:compopent in the web ·of interrelated constructs, definitions~ :and , propositions.that may explain .voluntary ttax reporting-in de~ocratie ~ocieties. A th6rough study of one possible 7 concept affect1ng behavior may, therefore, contribute to a comprehensive theory that explains and predicts taxpayer reporting positions. The general objective of the first phase of this study is to define tax complexity. If scientifically defensible information is to be generated on the role of complexity in influencing taxpayer behavior, one must rigorously define complexity and the taxpayer behavior of interest. As detailed in the following section, behavior can be classified and described in sufficient depth to guide the design of an empirical study. Unfortunately, the other terrr vital to operationalizing the study--"complexity"--has never been rigorously defined in the literature. Consequently, the author devotes the first phase of the study to constructing an operational definition of complexity. To form this definition, psychometric procedures are employed to elicit taxpayers' impressions of the concept. As described in a subsequent section, these impressions are used to evaluate the underlying structure implicit in the empirical data. Such a structure then allows one to infer the definition of complexity one needs to test the effects of complexity. In other words, the ability to specify inherent properties in the concept, and to identify instances where the concept applies, is a prerequisite to ar experimental test of its effects. 8 The general objective of the second phase of this study is to test whether complexity has a significant effect on a tax reporting position. This phase of the research provides the first empirical test of the contention that taxpayer behavior is influenced by tax complexity. Related Phase 2 objectives (if complexity does influence behavior) are (1) to explore the direction and extent of this influence, and (2) to ascertain whether the influence is associated with any particular taxpayer background variables or attitudes. The author has used the previous studies on taxpayer behavior (the tax evasion studies reviewed in Chapter 2) to help identify variables and attitudes of interest. Limitations of the Study The scope of this study is limited to one aspect of ta> complexity--the tax law in the context of particular tax situations. This is not meant to infer that such other aspects of tax complexity as tax forms and instructions are less important. Rather, the scope has been deliberately narrowed to facilitate a more rigorous examination. The basic methodological approach is the controlled experiment. This approach enhances internal validity because it increases control over the experimental variables. In turn this increases confidence that an observed effect is attributable to the manipulated variable~ 9 rather than to an extraneous influence. The artificial environment allows more control, however, it also reduces realism. One challenge is to create instruments that the participants find plausible and meaningful, a related challenge is to motivate participants to devote sufficient time and attention to the task. Definitions An intregal part of theory building is concept definition, and there are three general ways to define a concept. Definitions can be expressed in an extensional, intensional, or ostensive manner (Bagozzi, 1980). An extensional definition lists all the things to which an empirical concept refers. An intensional definition lists the set of properties such that the empirical concept "applies to all things having that set of properties, and tc nothing else" (Bagozzi, 1980, p.7). Finally, an ostensive definition of an empirical concept provides a sample of the things implied by the concept. The review of the literature in Chapter 2 reveals that no rigorous definition of "tax complexity" exists. But the term is employed with reference to particular tax situations, providing a partial extensional definition of the concept. Thus this study begins with an extensive definition of tax complexity in order to imply an 10 intensional definition of the concept through the use of multi-dimensional scaling (MDS) methodology. The intent in using MDS is to help develop the intensional definition of tax complexity so that behavioral implications of this concept can be tested. By specifying real world referents (sample tax situations), one can identify characteristics (properties or dimensions): and once identified; these characteristics can then be used to build tax scenarios that test the effect of tax complexity. In a sense, the~e tax scenarios represent an empirically derived ostensive definition of tax complexity. The other term critical to the development of the stud} is "tax reporting position." Reporting position is used as a surrogate for taxpayer behavior in order to operationalize the study. Thus, behavior is defined by the tax position adopted. But how then is tax position defined? One way to approach a definition is legalistically. Congress in the "Tax Equity and Fiscal Responsibility Act of 1982", provideE some such guidance. The new tax law proposes a double standard to define the point where one passes beyond taking all legitimate tax saving steps to those which are subject to penalties if detected. For non-tax shelter items, the law states that penalties will not be assessed for understatements if there is or was "substantial authority" for the treatment of the 11 item in the return. Items tor wh~ch evas~on or avo~dance of income tax is the principal purpose are deemed tax shelter items. With respect to tax shelter items, the penalty may be avoided only if the taxpayer established that, in addition to having substantial authority for his position, the treatment claimed was "more likely than not" the proper treatment. The Senate Finance Committee Report, in an attempt to help clarify the meaning of "substantial authority", explains that this standard is intended to be more stringent than a "reasonable basis" standard but less stringent than the "more likely than not" standard. The Committee Report further indicates that the standard will be met if the weight of authorities supporting the position is substantial in comparison to authorities supporting other positions (Senate Finance Committee Report, 1982, p.13). Based upon the Congressional Committee report, the following definitions are inferred: Compliant positions arE those where the taxpayer selects a position that is either certain, or more likely than not, to reflect no tax underpayment. For tax shelters, noncompliant positions begin to the right of the double dotted line in Figure 1. Figure 1 attempts to graphically depict these relationships along the continuum from compliant to noncompliant positions. 12 FIGURE 1 TAX POSITIONS ACCEPTABLE REPORTING POSITIONS . Certain- More . Substain- . ty that Likely . tial Au- Position Than thorita- Reflects Not tive No Un- Support derpay- ment I COMPLIANT . . i ~~ .. .. .. .. :: .. H .. .. :: !! :· .. .. .. :: .. UNACCEPTABLE REPORTING POSITIONS (Subject to penalties if detected) Reason- Minimal Tax able or Weak Evasion Basis Authori- Due to tative Fraud Support or Neg- ligence NONCOMPLIANT The description of compliant and noncompliant position~ shown in Figure 1 appears adequate to form a working definition. Although there is some confusion over the precise application of the government's standards, from the guidelines given it is possible to develop scenarios and reasonably classify responses as compliant, or noncompliant. Although Figure 1 appears adequate to form a working legal definition, it omits relevant behavioral considerations. The behavioral view considers what the taxpayer believes is the correct (Internal Revenue Service) reporting position. In other words, there is a psychological as well as legal standard. In a legal sense, a subject is aggressive or evasive if they take a position unauthorized by the law. In a psychological sense, a taxpayer is aggressive or evasive only if they knowingly 13 take a position unauthorized by the law. Consequently, the legal standard can be objectively determined, but the psychological standard depends on each individual's attitude and understanding of the law. Since both measures are important from a policy perspective, the taxpayers evaluated in Phase 2 are asked to record both the tax reporting position they would choose and the position they believe to be correct. Consequently, it is possible to interpret the Phase 2 results in terms of psychological as well as legal noncompliance. Organization of Remaining Chapters The next chapter presents an overview of the tax evasion literature and also reviews studies using relevant research methodologies. The objective in Chapter 2 is to synthesize the literature and provide a foundation to guide the structure and design of the empirical portion of the study. Chapter 3 presents the methodology, design, and results of Phase 1 of the study. In Phase 1, tax complexit~ is empirically defined, and a basis is established for testing the effects of complexity in Phase 2. Chapter 4 presents the methodology, design, and results of Phase 2. In Phase 2, the effect of complexity on taxpayer reporting positions is empirically tested in four different tax situations. Chapter 5 contains the summary, conclusions, 14 limitations, and contributions of this research project. Chapter 5 also presents the author's recommendations for future research. 15 CHAPTER 2 LITERATURE REVIEW Summary of Previous Tax Studies The published work on income tax complexity provides little help for a study involving the effect of complexity on behavior. The majority of the studies are critical expositions on specific complexities within the tax law. The remaining literature, except for two empirical studies on measurement (Karlinsky, 1981, and Schroeder, 1975), discusses tax simplification and the political process of changing the tax law. No published academic work regarding taxpayers analyzes or relates complexity and compliant or aggressive tax behavior. In fact, no published academic work relates any factors to compliant or aggressive taxpayer behavior. All of the published articles deal with the extreme end of the spectrum shown in Figure 1--tax evasion. Tax evasion studies may be relevant to the study of tax compliance and aggression if behavior can be viewed as a series of possiblE acts prompted by identifiable stimuli. Table 1 presents a chronological ordering and summary of eleven key articles on tax evasion. The author discusse~ these articles in the pages following, emphasizing those with the greatest relevance to the present study. 16 All'tiiOal fftr.1CA!IO. 1. 1Cftlllle1 111111 Ceaviltucua of •• ...., CO-. Qllaaei.&at~YO aaa1reu La ,....u.c r~. lllitKitf A • .._. l,..at-UI. 2. ~e~~~~vu • owl .... 11H7l •aa t.e9al IUicuoaa•. !!Ia Uaivorai~ of ClLo ... taw b¥iw. Ht 274•2tt. J. AUi.OfltM 6 ...._ 111121 ·r- t'lla &wuuaa A !lleontual Aaalyau•. Jouul of ,.U.o leaaaal.•• Ia lZJ•UI. s. VOIJel 111741 ··t-t\00 .... -lwop~oa La ......_, Aa ta--.coucioa of locoat lui'YOy lieu", ... Laul t'lla Joaanel, 271 Pf 4ft-SU. '· Pri .. 1MIIl, -.1u1. eac lut .... rt, 11f711 "A ILaoalaUn ltlloly of faa II:Waaioa", J_..l of .. u..-.. !Oa107•1U. 7 ..... • taftnuta 11t111 "faa ltllioa &ad ! .... yu AUiClltHI A lui'YOy" ,.UoMa.UU.avnua ... , .. lla441•4S1 fur.& 1 SVBAU or tax lvura. racoa noora P1all _,.riMae i.:l COO .. raUOII ..,11:11 Cl'la tAl .. aaur1119 ~ollua .,...,. 011 cu racUBa. ~lylOO -:tacu. t.u...-c~c. aua:t~• faced wL:a _,.CbaCLoal cu ... aaioll ........... . l'IUJIClfLI PIIIDUIGI coerc1veaeaa of cu ... iaiavatl.on. penapci.oaa of tau-• of ..ao cu •re- an4 an.i.c- _ .. cu ou-.n all an-. cu c..,u_ w~cllLa a -Vf· ruac eo eoopllaaUo ia,o~ of uapayec nUClide. r...,..yor nona ua aa l.aportaac !actoc Wl4orly1119 cutaroc IM!Iavi.or an4 noCNUva appeala ace .. ra aftacc~va tbaft aaact~oa Cl'lruu a l.Aduo....., :oap11aaca. Ieuty ~~ u.a • ..,...,. ... &ad Ull4a :o nann S~l' 1 fial1119 tii&C aunula l.a aa Ulpo~~-~ factO~. -.11 cu evaa- u a lfOCLil o:aaa of rauoaal Ho-c onoico llftlor coatiUOIII llf IIIICIUC&iaCy, 1:118 c...-yer 11 Y~ .... II .-11119 tor -a ~ La u,ac ot ella UkoU- of 4a-tioa- peaaltiu. · auuau .. uu.~ t11e0ry •• :aa -lyl.af ~act.nl fOIIIIIIaUoa toe tllo at.uty. A&c-.ca co _..,. - uoqoayera ut1Ucy t.Mti.• fna a puraly -.uy oae co illllliiM -Lal •UIItia9· ......... tllot pe.....u- of ~Uy, nfw- - ... f-lf·· ... ...,.,,_ Vitll _ ......... _, .. ia,onUt 1ft-~ .... 11Ma. Al .. tile puui. .......... i.U.t.i.U of .._.... ..... "-- 1-1. _. 1:118 ~of ,_ lu1 ............... -· loUII w IM.coloted u ua nau-. ~ - .. - .... If ..... , ............ tu CO .... 1 1M ..,U1011. -~ HU.U llB ... ut ._..... ~ _. ret- tlfOIIf -ry toe -.ucral II&PfOCC· .._.1011 ot SUU.,.l' a _. .._,.. 6 Orlaaa•a atu1101 rotUI1119 ,..,...yu .. cu. ruta taat -~al :olaci.oaaa~p• 1111fur .. ~• u rete~ fC"P Meal play aa ut~ruat nola u atffttl.af cu ...,u-. laaaiael tAo effect~ llf 4itteroac ~~rolla.IIUi.Uoa of .. ,.acc:,e, lavale of fi.Ma aal U& lf&CU oa tAl .. UI.OB 4M11i.Oft. Pouat tlllc 1'11taor caa recea reduce tAo·fracti.oa of ~ r.,.rc... Apfeara ~• Do till fUn cu evaaioa .. ,..,. ual.af • a~laciooa perad.l.fO. Fouat tii&C u.,.yera 1ft 90ft0Ul IIAYO I ratlloa' loai.eac ucitvde -n• ua avaai.on, percei.v~ i.e co 1M a lua aori.oua offeftlo. tllla u-. -al-e, - l:lr~Mry eat oaly aU.fl'Ciy •co MltiDlll tbaft acaal1ft9 I l:li.cyolo. ltaei.n.l.cly ai.taifi.caac faccon lLakol co ..,aal.oa won 1 .. ~.a-. ftOftllllue. lw ,.uncal etUcaoy, 0111 1 .. uuac u people. ...auao cu coapiUi.cy waa l\OC l'li.fl'lly utot aa a 4ef1c1ancy 1n ..ao ~ law, ..ao autlloca conoluclad c.'>ac ic 1a of li.aor l.llpO~·~ 17 •. oeaa, &eenaa. ' •• nne,. tUIOI •t..,.yac•a At&Uwloa co 1110- taa &vaUOIU All lllficical Sl114y" IC1Uall tu _,_ u. Scott • aaaotck 111111 "De&on-• aiiC IIIC tu C~aenU.• teaUIIf xacacHUea lll'fO..._ ~ Vtt.lt.&art.aa taeocl. .... Jevuld .... Uell lel&aY1Ha11C1- 17olf5•401, -raau l I~ or tAX IVU1011 r~a ltoDlU t""ay • IIOIICIIMOD a-la ~~ 507 scocuaa c..,.yan. t.ala UfaC Ueal:. Sl:l&dy c:o11finacl co 57 =~u..,. nll4uu. Ate• r&llfacl fi'CIII 21•ZI yaua. lt&nef • ........, aalaccua ofZOO_f_aa Alltalll', - ta&'k I:Dl ....... tUDHO&'f• U -la cea,...aa van c_t.,... sunay • ..,. ... auple of Ut troe ':liD Oklall Ctcr uu. ltwiY aoutlll: ~~ 14ellc:Ofy aco:~cllacla l:llat :111911t 1llflua11Ce coepli.ance. Qllan1onaua quanacl faccou aucft u level of o:auuoa. wonll af jitiiDUc eapalldl.tll&'e, tal.:aaaa of tea, praiDUl.Ucy of ~auccioll, aaver1ty of pana1t1ea. opportllllitY for evaal.oo, .. cal paccepl:iolla, alld eco,._ic conu4arac1aaa af o:aapayau. One af cu 110at aul.killf UntU.a vaa l:llat ':liD aajarto:y d rupolldaftu talc c!wy W&'D payl.ll9 toO IIUCII 11\- tU Ul . &'alauoa to oeha&' paopla v•Ch aiailU auo•llf•· Til Ollly -uoe of eo~~p1..U.tr ill clW Ul:iCla waa a l:l&'ief DllftU&l.Oil l:ba& . C~~eplaaicy of o~>a tu ayai:U MY M ella cauaa of ':he reapoauaca eapa&'i.eiiC1119 1\o&'t&ooul ift041UUr. · SWijecta facacl wi':ll :.o rollllda of llypol:ftal:ical :ax evaal.oll daciaiolla iiiC:eaaacl ':!Ia .-.uac ava4acl wl\an thay paccaivacl ':lleUelvaa to M vtcuq of fiacal illO<IUUY aNI dee_.. tlla -- aYaclacl wi\DII they pacc1avell ~lYU u 1M -fUiutaa of f1aca1 iJielp&11:f. A .-n~Uo fUilD&'U. IIUJttHIIIIII aa4 ant....,.. iafo_.iOII wa -at.•~ alao. tile -lu iellt.n&ell cue .... ~ aNI .......... - .. 4t.4 - ap,.u ... ... •*1U-nr nla&e~~ • ua ..... -. .. . al....,.. - .... at.taifUaac tMloa ... - a .-.............. -1&'~1. 10 1toaUffU9 cJw tai!UU1Cy ot IMnU aNI -• ,_ ua &'aa,..._u .-eaur .autt.&ell a pau ~at.oa of ua· ua 1..,, unu. coc~lr oa ooly 10 • 40\ of U. !lypoclsatiCal c .. aa • - YD&'t.o\ID J:U . 11:- van aap1a1DH o- 7'1\ MU c~Wy -14 ~1M& co ~epo~~ - ar all of ua 111- 1-· coapaaoce vaa ai.flll.Ucuely n1a&ell eo - •tol' vu1aa1ea, cypa aa4 aue of a- 1uo. l.a1'9e caaa i.t.,.. var• CM •at lt.llaly o:o 1M l'eponacl. llllaU.I' cJw ~ UOD M4 -II l'aponacl ':0 l:fta liS 414 1101: ap,.u ':0 Ita a ••ta1f1CUC •c•vac~ fDCUI'. AU pa,_1a&LOII ..-.rooafa •- aa ..-al tn.U1ft911aea co avacle. A11:lla1191l · cJw auual'a tit aoc ""'- cJw uaua, i.l: u ia-tt.ot u ....,.lata aa u waal:ftD&' caaa i.- ua 1 ... CDIIflUJ aa4 fll&'l:ftD&'. -1' Chia U a fDCU&' 1ft clW fi'HCDI' «<OIPU.UCD &'Dta IIO&ell. ~a auc.'IOra pl'acli:t tax naai.oa IMftaYioc .. a tuacu•o of ~~~~~t•vauoa tpcojaccacl 1'-&'41 - 1Dil1ft1c- tpcatecucl coati. taay ,_ o:au ua aUec1:a of •cLva&t.oa oa ~ ua ClleaCillf "ua coelliU-1 oa ua ..... Of tnol'c-. · coo..waa1y, U. ~ of coau u a ~-c • ~ &UCIIoaCU. an coeiiUiHal oa - pee- O&' •- of ..uvact.olla eo -•· ftU I'HDUCI\ intt.M- - 4aaic81U.&y Of AD i.AI:al'actiOII -1 ~ uac189 ua DYaat.oa fae&Oca. taua •- -1891r ai14 facceca - pow&'hl ..- ca~ia cucwaauocaa. 18 The first of the modern tax evasion studies is Strumpel's 1966 survey of European taxpayers'. Strumpel described tax evasion as a multi-factor problem and specifically identified taxpayer attitude as an important variable. The importance of taxpayer attitudes was reaffirmed in research done in this country a short time later. Schwartz and Orleans (1967) conducted a field experiment which, while methodologically very different, reached conclusions compatible with Strumpel's findings. P. few years later, Allingham and Sandmo (1972) used an analytical approach involving mathematical modeling. Predicated on utility maximization, they calculatd that tax evasion would decrease as the probability of detection, the magnitude of economic penalties, and the possible loss of social standing increased. Referring to these three studies, Spicer (1974) attempted to build a behavioral mode of tax evasion. He identified a few new background variables like age, income, and amount of income derived from wages as significantly related to tax resistance. But mainly his work tends to validate the previous three studief and demonstrates the applicability of social exchange and reference group theory to the study of tax evasion. Anothel source of validation for the importance of reference group norms is a Swedish survey by Vogel (1974). Friedland, Maital and Rutenberg's 1978 simulation 19 studied the same type of issues that Allingham and Sandmo studied six years earlier. The contribution of this later work is that it appears to be the first attempt to study ta~ evasion in a laboratory setting. Generally speaking, this methodology has greater internal validity than survey methodology and greater external validity than mathematical modeling. In addition, their application of this methodology appears to have been fairly rigorous. In fact, Spicer and Becker (1980) used this same experimental design two years later to study the effects of perceived inequity on evasion. Although neither of these laboratory experiments identifies new factors, they are more rigorous, thorough studies of previously identified factors. Song and Yarbrough's 1978 survey of taxpayer ethics anc attitudes is interesting in two respects. First, they founc certain background variables and attitude factors statistically significant. (This would tend to confirm Spicer's study except that Spicer found a different set of factors to be relevant.) Second, the Song and Yarbrough article is particularly relevant because they address the issue of tax complexity. Specifically, they asked respondents to rank these five commonly discussed shortcomings of the income tax: (1) the tax rate is too high, (2) there are too many loopholes, (3) the regulations are too complicated to understand, (4) the middle-class 20 bears the biggest burden, and (5) it penalizes the honest and law-abiding. Since item (3) was ranked the most serious problem by only 13 percent of the respondents and the least serious by 35 percent, they concluded that "tax complexity, in comparison with the other problems mentioned above, is only a minor concern for most of the taxpayers." (p. 451). But since the authors never defined "tax complexity," the relationships remain somewhat speculative. For instance, item 2 regarding loopholes may be related to the issue of complexity. (Also, stretching the matter further, item 5 rna} be a ramification of item 2.) Consequently, because Song and Yarbrough never really defined the term "complexity," it is difficult to assess the significance of their conclusions. Two other recent studies also touch upon the issue of complexity. In 1980, Dean, Keenam and Kenney published a broad and superficial study of tax evasion factors among Scottish taxpayers. The first mention of complexity in their article appears in the conclusion when they try to rationalize the cause of the horizontal inequities they noted. Hotaling and Arnold's 1981 survey is interesting from a complexity perspective because they presented respondents with three different income items that may possibly represent different degrees of complexity: (1) cash for moonlighting, (2) barter income in exchange for 21 serv1ces, and (3) noncash compensat1on which augments a regular salary. Cash appears to be the least complex of these items and, what is especially noteworthy, the authors report that cash is the item most likely to be reported. Another interesting conclusion Hotaling and Arnold draw is that none of the background variables were statistically significant: All population subgroups showed an equal willingness to evade. This finding, of course, conflicts with the conclusions Spicer, Song and Yarbrough reached. Studying tax evasion is only a secondary objective in the last article in Table 1. The primary objective of Scott and Grasmick's 1981 study is to demonstrate that the influence of many factors in human behavior is conditional. They found that the effects of motivation on tax evasion are conditional on the costs of deterrence, and conversely, that the effects of costs as a deterrent are conditional on the presence or absence of motivation. In essence, they assert that seemingly mild social factors may become extremely powerful when conditions are right. This methodology represents a departure from the preceding tax evasion studies which attempt to assess the impact of factors in isolation (that is, all the preceding studies involve a univariate rather than a multi-variate analysis of variance) • 22 A Synthesis of the F1nd1ngs of Prev1ous Tax Stud1es Although the articles listed in Table 1 all focus on tax evasion, they furnish some guidance for the present study, and this guidance is especially meaningful if tax aggression is viewed as a weaker form of tax evasion. Although most of the studies use relatively unsophisticated research methodologies, they identify several factors that may have an influence on tax behavior. Figure 2 represents an attempt to synthesize these factors and convey a sense of the intricacies involved in the studies from Table I. Unfortunately, in the majority of studies in Table I, the authors fail to define their terms. This problem is compounded by the fact that three of the studies involve European taxpayers, and thus the factors described may have different connotations related to differences in cultures. Because of these difficulties, several of the factors presented in Figure 2 almost certainly overlap. The author has arranged the factors examined in the studies according to whether they have a predominantly psychological, social or economic basis. The descriptions in the boxes are terms used in the articles. The arrow in each box represents the expected effect an increase in the factor described will have on tax evasion. The numbers above each box are keyed to the articles in Table I. For 23 example, the 2 above the "taxpayer norms" box on the upper left side of Figure 2 indicates that the Schwartz and Orleans (1967) article from Table 1 explicitly discusses taxpayer norms as a factor related to tax evasion. The dotted lines connecting the boxes represent an interpretation of the tax evasion literature regarding the likely relationships between factors. Finally, the plus and minus signs on the connecting lines indicate whether the studies reviewed generally report positive or negative relationships between the factors. Note that complexity is not included in Figure 2 because complexity is not explicitly examined as a factor in any of the studies. 24 FIGURE 2 FACTORS AFFECTING TAX EVASION PSYCHOLOGICAL SOCIAL ECONOMIC 2 7 8 9 10 11 4 7 3 8 10 TAXPAYER NORMS + NORMS OF - OPPORTUNITY + [',,------ ASSOCIATES """'---,.,,.... FOR TAX 1\ t EVASION • ,__ ______ \\ '------....---... - ...... \ \ ;.,. ,... ...'j \ \ I ,.... ,,/ 1- \ \"' I .,"' ,.. I \- \ _f./ ,,.. I \ \ .,"'I (,.. I ' \ .,.... I I : \ \ ,, I I 8 I 10 \ )(,.. 1 10j 11 1 2 3 6 1 11 \ ., ... \ I 1 4 10 11 PERCEPTION - , -t \ SOCIAL I ECONOMIC OF ,"'• \ ' CONSEQUENCES 1 SANCTIONS ~OERCIVENESS '---t-- OF EVASION J ./ OF IRS + \ t "' .,.....,.... ....... + \ ~ .... ... ---------\---------+ ---'1""'., __________ .... \ I .; ,.. 1 4 7 8 PERCEPTION OF FAIRNESS OF THE TAX I ,,.. \ ~,:,,... \ , .... ... \ _,,.. I , .... \ ;o: ;;..- I \ ~;;' ·- ~If' 11 ~~ \ I ~~ \ ATTITUDE ~~--.:. ___ j TOWARDS TAX 7 8 EQUITY OF 1------- _ TRADE WITH THE t SYSTEM ',~ EVADERS + ', t GOVERNMENT t ~------- ..... ' ' ' ..... ' ' ..... ..... ....... ........ ..... ..... ..... ........ ' ' ......... ; I I I I I 1- 618 I TAX RATES 25 Figure 3 uses the same symbols and is an evaluation of how complexity may relate to the factors presented in Figure 2. The purpose here is to present a model that may help one visualize some of the relationships explored in the methodology section that follows. First of all, there appears to be a link between complexity and the factor "perception of fairness of the tax system." Fairness is usually defined in terms of horizontal and vertical equity. Dean, Keenan and Kenney (1980) assert that complexity might be the cause of respondents experiencing horizontal inequity. But complexity is probably linked to vertical inequity more often. The common complaint (at least among middle-class taxpayers) is that the wealthy use tax loopholes to avoid paying their fair share of taxes. The second relationship depicted is between complexity and the factor "opportunity for tax evasion." Among the few authors to define their terms, Allingham and Sandmo (1972), described "opportunity" as a combination of the probability of detection and the costs of getting caught. Complexity may affect both aspects of opportunity. According to testimony before Congress, complexity lessens the probability of detection (Senate Finance Committee Report, 1982) • Apparently the IRS faces an arduous task in attempting to train its personnel in the complexities of thE tax law so that they can detect aggressive and evasive 26 positions. Complexity may also affect the costs of getting caught. Scott and Grasmick (1981) describe "cost" in terms of three factors shown in Figure 3: "taxpayer norms" (moral costs), "social consequences of evasion" (social costs), and "economic sanctions" (economic costs). Certain taxpayers may use complexity as a way to avoid some costs. For instance, they may use complexity as an excuse to placate their consciences or rationalize their evasive actions to others. In the case of economic costs, they may use complexity to argue for less severe sanctions when their noncompliance is detected. As Figure 3 indicates, a number of studies have relatec the perception of coerciveness to the opportunity for tax evasion. The consensus appears to be that an inverse relationship exists between these two factors (that is, the greater the perception of IRS coerciveness, the less the opportunity for tax evasion and, consequently, the less the amount of evasion). But Stumpel's early work indicated the reverse--that greater administrative coerciveness led to increased taxpayer resistance. Hotaling and Arnold's findings also conflict with the general sentiment. They found that reporting of income items to the IRS did not appear to be a significant motivating factor. 27 FIGURE 3 RELATIONSHIP OF COMPLEXITY TO THE OTHER FACTORS HORIZONTAL EQUITY PERCEPTION OF COERCIVENESS OF IRS PERCEPTION OF FAIRNESS OF THE TAX SYSTEM COMPLEXITY OPPORTUNITY FOR TAX EVASION PROBABILITY OF DETECTION MORAL COSTS (TAX PAYER NORMS) COSTS OF GETTING CAUGHT SOCIAL COSTS (NORMS OF ASSOCIATES & SOCIAL CONSEQUENCES OF EVASION) VERTICAL EQUITY ECONOMIC COSTS (ECONOMIC SANCTIONS) In conclus1on, the tax evas1on studies provide some indication of what might be expected from taxpayers. Although the results are mixed, a few of the articles indicate that taxpayer attitude and certain background variables (for example, sex, age, education and income) may be linked to the reporting positions selected. Furthermore, there is some indication that the concept of complexity may be linked to the notion of fairness and the opportunity for tax evasion. As the research methodology section suggests, these tax studies provide guidance for the background, attitude and "adjective descriptor" questions in the research instruments. A Review of the Methodology Literature This study is comprised of two phases. Phase 1 uses multi-dimensional scaling (MDS) to derive empirically a definition of "tax complexity". Phase 2 uses multi-variate statistics to test the impact of complexity on a reporting position. The statistical techniques used in Phase 2 (analysis of variance and canonical correlation analysis) are established and accepted analytical techniques (see, for example, Hair, Anderson, Tatham and Grablowsky, 1979). Consequently, there is no compelling need here to review the accounting studies that have utilized these techniques. 29 This is not true, however, of the MDS methodology used in Phase 1. Although this methodology has become increasingly popular in econometrics, finance, and marketing studies (Carroll and Arabie, 1980), it has not been widely used by accounting researchers. Consequently, the author will review an early accounting study using MDS and a recent one, both of which provide insights into the use of this methodology in accounting research. The technical discussion of MDS as a geometric model for representing data will be deferred until Chapter 3. The following discussion emphasizes the accounting issues addressed and the insights gained from utilizing MDS as a measurement tool in accounting. Multi-dimensional scaling has been used in accounting studies where the underlying attributes of interest were not well specified (for example, Libby, 1979; Frank, 1979; Belkaoui, 1980; Pratt, 1982; and Shockley and Holt, forthcoming). The primary reason for employing MDS in these studies has been "to help identify structure not obvious in the data that underlies attitudes and perceptions of accountants and users of accounting information" (Watkins, 1984). Libby's 1979 study represents one of the first accounting applications, and it illustrates the contributior the MDS methodology can make in providing new scientific insights. 30 Libby's objective was to test empirically bankers' and auditors' perceptions of the messages communicated by different audit reports. To do this testing, Libby used ten different types of audit reports as stimuli, and he asked his subjects to rate directly the similarity of the messages intended by the ten report types. Each subject rated all possible pairs of reports, a total of 45 combinations. Then they made numerical ratings of each of the ten reports on each of thirteen rating scales labeled with adjective phrases suggested as report descriptors in the author's literature search. Libby used multi-dimensional scaling to model the subjects' perceptions of the intended messages and. to measure any differences between them. Then he used property-fitting techniques (special regression procedures) to help label the messages with adjective phrases. Finally, he analyzed the relationships between individual differences in perceptions and a number of background variables. Following this methodology, Libby was able to demonstrate that his subjects differentiated the audit reports more or less along two dimensions, one based on the additional information provided and the other on the audit judgment rendered. This finding was especially interesting because, somewhat surprisingly, the subjects did not appear to differentiate according to the circumstances producing ar. uncertainty report (whether it was litigation or asset 31 realization). Also Libby's study provided a measure of the differences in the subjects' perceptions among unqualified, qualified, and disclaimer reports. The distance between the qualified and disclaimer reports was found to be more than twice as large as the distance between the unqualified and qualified reports. Shockley and Holt's forthcoming study of bankers' perceptions of Big Eight CPA firms also uses MDS. Their application is helpful because it reveals both the weaknesses and the strengths of this methodology. First of all, the Shockley and Holt study shows in an empirically defensible, unbiased manner, that bankers do differentiate among the Big Eight firms. Furthermore, it shows that they seem to differentiate along three distinct dimensions. What is interesting, however, is that the authors can interpret only one of the dimensions (conservatism) because of high correlations with adjective descriptors in their property fitting section. They interpret another dimension (industry specialization) by viewing the output and then performing some follow up work to verify the conclusions they reach from visual inspection. This was necessary because they never expected industry specialization to emerge as an attribute distinguishing the firms. Finally, Schockley and Holt could find no attributes with a sufficiently high correlation to warrant labeling the last dimension. These 32 results demonstrate (1) the usefulness of MDS in identifying structure in a data set, (2) the value of a spatial representation in helping to interpret the data, and (3) the fact that some aspects of the solution may remain indeterminant. As these two studies reveal, one of the benefits of MDS is the opportunity it gives to generate a solution not pre-specified or otherwise anticipated by the researcher. This advantage is particularly valuable in the present circumstances where the author did not know what elements ir a tax scenario subjects might cue upon in judging the complexity of a case, or which, if any, concepts from Figure 2 might relate to complexity, or how closely these concepts might relate. The following chapter details the MDS research methodology and describes the results of the analysis. 33 CHAPTER 3 RESEARCH METHODOLOGY, PHASE ONE ANALYSIS AND RESULTS Overview This chapter and the next detail the study's methodology. As the research methodology flowchart in IFigure 4 shows, the study is divided into two phases. Phas 11 is basically descriptive. Its purpose is to explore the !underlying structure of the taxpayers' complexity assessments in order to derive the intentions of the concept. One expected by-product of defining complexity is additional information about the relationship between complexity and other concepts . Another potential benefit of this phase is the opportunity to explore the relationships between complexity and the taxpayer backgroun I \variables. Phase 2, described in the following chapter, is the experimental phase. By analyzing in this phase the reporting positions selected in tax scenarios of varying complexity, the author seeks empirically to test the relationship between complexity and taxpayers' reporting positions. Figure 4 depicts the study's research methodology and shows that three distinct types of data have been gathered in Phase 1. By far the most important data involves the ta complexity judgments by each subject. These judgments 34 provlde the lnput necessary tor the MDS analysls. In addition, data has been collected on adjective descriptors to help the author interpret the MDS dimensions. A third set of data, involving demographic and attitudinal information on the subjects, has helped the author ascertain whether she can draw generalizations between the subjects' tax complexity judgments and their background data. In Phase 2, the author collected two types of data. The primary data is the subjects' reporting position selections based on tax scenarios of varying complexity. Phase 2 is predicated upon the successful completion of Phase 1, since one must have a definition of tax complexity in order to design tax scenarios of varying complexity. In addition to the reporting position data collected in Phase 2, the author also collected the same demographic and attitudinal information on the subjects as in Phase 1. In Phase 2, this data is collected in order to ascertain whether generalizations can be drawn between the subjects' reporting position selections and their background data. 35 FIGURE 4 RESEARCH METHODOLDBY FLOWCHART P~ase 1 Data Collection Phase 1 Analysis Objectives: T: str~cture the data gathered in order to define tax complexity and better understand the relationships between tax complexity and taxpayers~ Phase 2 Data C:e:llecticn Phase 2 Analysis Objectives: To assess the impact of complex- ity on reporting positions and better understand the relation- ~hips bet~eeG ta~ com plegi ty 5 ta.:~ repcrting pasi tions; and ta;,:paye~:R I Background J variables fDr prope:ty -. un ::.timuli _ (demogrt!phi:: fitting (13 ta~ & attitudinal (a ::.eries of situations) 1 quest i on s )J l.._ __ -r-.JJ l'---.,.--c---'1 ~ a;~t~~preting ~di[•ensions perceived complex ity dimensions (Pesearch Question 1 J MDS • Clustering of subjects into groups based on subject weights generated by MDS individual diff: scaling IRQ21 CLUSTER ANALYSIS and exa1ine the relationship af clusters {sub- ject weights) to background vari- abies CANONICAL CORRELATION the in- ati on ships among the bd.::k- / Rcnnrhn,o f , ...... r ~.~~ ~ £.1::! fBackground I ground towards ag gressive tax beha.vior, (F\Q6) A NOVA variables vari- position selections 1----ftDemographic a.bles scenerros ot \RQ4i ANOVA Effect of irdividual cwmpiet~i ty dimensions ~: & ~ttitudinal {RQ3) information)/ ~A~Q~~: F:elationship of ition select- ions tu back- ground variables {f~Q7) CANONICAL CQRF~ELATION ,,.., .. ,, .:.v-;:... CDF;t~EL- AT ION 36 Research Questlons The central research question in Phase 1 is What is tax complexity? More specifically, What are the intentions or properties inherent in taxpayers' perception of tax complexity? Focusing further on taxpayers, when one groups taxpayers according to their complexity assessments, can onE relate these groups to any background information on the subject (demographic data or attitude ratings)? In addition, do any significant interrelationships appear among the background variables? And if so, do these interrelationships provide any insight into taxpayer behavior? Phase 1, to repeat, is primarily the descriptive phase of the study. Its principal purpose is to simplify and order complex and diverse relationships so as to facilitate further research into a theory of the effect of complexity on taxpayer behavior. Accordingly, the specific Phase 1 research questions follow: 1. What are the perceived dimensions of tax complexity: 2. When taxpayers are clustered according to their weighting of the complexity dimensions, what is the relationship between these clusters and the background attitudinal ratings? 3. Are there any significant interrelationships among the background and attitudinal ratings? The results of Phase 1 will generate an empirical basis 37 upon which the author can build tax scenarios for Phase 2 testing. The primary focus in Phase 1 is to study the effects of complexity on tax reporting position selections. In addition to this primary focus, the study will also discuss secondary issues involving the complexity dimensions, tax scenarios, and background variables in order to increase understanding of the interrelationships involved. The Phase 2 questions follow: 4. Does complexity significantly influence the tax reporting position? 5. Do any of the individual complexity dimensions (or combinations of dimensions), or does the tax topic, exert a significant influence on the reporting positions? 6. Do subjects show a propensity towards aggressive tax behavior as measured by their perception of the correct tax position or by the legal standard? 7. What is the relationship between the tax reporting position selected and identified demographic or attitudinal variables? The Phase 2 research methodology, analysis, and results are detailed in the following chapter. This chapter focuses on the Phase 1 research. A general overview of the MDS research methodology appears first, followed by a description of the Phase 1 analysis and an interpretation of the results. 38 Phase 1 Methodology As the literature review indicated, MDS is most applicable to problems, like the proposed one, where the concept and its underlying dimensions are not well understood. As stated previously, one benefit of MDS is that the visual representation it produces can make the data easier to comprehend and may reveal otherwise hidden structures or dimensions in the data. Another benefit is that MDS tends to reduce researcher bias because, rather than requiring a priori knowledge of the attributes of the stimuli to be scaled, it provides a space in which dimensions relevant to the subjects can be scaled. An advantage of the MDS program used in this study is that weight and space measures are calculated for individual subjects. The subjects' weights are adjusted so that their average is one for each subject and for each dimension; that is, they are both row and column normalized. These weights indicate the relative importance of each dimension in the final solution to the individual subject. An MDS solution often shows a subject relying on only one or two dimensions, even though four or five dimensions may be relevant to the group. In a technical sense, MDS measures people's perceptions of the concept by identifying the inherent structure in a data set and depicting this structure as an n-dimensional 39 geometric representation. The axes (or dimensions) underlying this spatial representation are often interpretable as attributes that distinguish the objects of judgment (referred to as ''stimuli"). What one needs in order to use MDS procedures is a set of numbers that reflect proximity judgments about similarities in the stimuli presented to the subjects. The subjects are usually asked to indicate the similarity (some alternative words for "similarity'' are "relatedness", "dependence" or "association") of the two stimuli presented. Stimuli judged similar to one another are then represented as points close to each other in the spatial configuration. Stimuli judged dissimilar appear as points distant from one another on the configuration. Research Task The research task of Phase 1 required the author to gather three different types of data: (1) similarity judgments between pairs of stimuli, (2) ratings of stimuli on adjective descriptors, and (3) demographic and attitudinal information from the subjects. Similarity judgments are the heart of the MDS procedure. They are the primary means available for discovering the underlying structure in the relationships between the stimuli. In this study, the stimuli are 40 embedded in short tax scenarios. (Appendix A presents a complete list of these scenarios.) The author used thirteen scenarios covering the following topics: child care, charitable contributions, capital gains, entertainment costs, ACRS, gifts and awards, extra exemptions, health insurance, trusts, income averaging, interest deductions, non-recognition of gains and losses, and IRAs. These thirteen scenarios, covered a variety of revenue and expense topics and did not unduly constrain the potential dimensionality of the solution. 3 The author devoted considerable time and care to the scenarios' design. Since taxpayers are likely to relate to issues involving the tax law as it applies to individuals (rather than to corporations or partnerships), all the scenarios deal with individual tax matters. One benefit of covering a cross-section of topics is that one can ascertair from the MDS solution whether the topic influences a taxpayer's perception of tax complexity. In addition to topic selection, the author also took care to ensure that the scenarios were rich enough to reflect various aspects of tax complexity. Some of the manipulated aspects of interest include length, legal references, allusion to numbers, numerical manipulations, changes in the tax law, and population of benefit. Often, aspects of interest come to depend more on the subject's experience than on the 41 articular features embedded in the scenario. For example, epending on the taxpayer's experience, she or he might berceive a scenario as abstract or familiar, an area of I ~buse or benefit, vulnerable to cheating or an equity I provision. A taxpayer might use any one of these perceptions to define tax complexity. An advantage to using MDS, though, is that even if the researcher does not anticipate a feature relevant to the subjects' perceptions, it may still appear in an examination of the final MDS I configuration. (Recall the review of Shockley and Holt's study in Chapter 2.) I The research task also involved gathering the subjects' 1 ratings on seven adjective descriptors for each of the thirteen scenarios. The author asked the subjects to evaluate each scenario in terms of whether they felt the ta situation described was technical, fair, an area susceptibl to cheating, familiar, abusive, personally beneficial, or a changing area of the law. (Table 2 presents the question format.) The reason for this diverse group of descriptors is to help interpret the final MDS configuration. Ideally, one or more of the descriptors would correlate significant! with one or more of the dimensions subjects use to define tax complexity. The subjects were also asked to fill out a one-page information sheet indicating their sex, tax return 42 preparation experience, age, education, and earnings. They were also asked a series of questions designed to measure their attitude regarding the fairness of the tax laws, the percentage of taxes that can be evaded, the complexity of the tax system, the degree of tax compliance of their associates, and the percentage of revenue used wisely by the federal government. (This questionnaire appears as Table 3.) The author had identified each of these variables in the literature review (described in Chapter 2) to be of possible relevance in distinguishing tax evaders. Even though their reliability has yet to be established, it is interesting to test whether these variables have any relevance in distinguishing taxpayers according to their complexity judgments. Data Gathering The thirteen systematically varied scenarios were presented to the subjects in pairs. Thus, each subject was required to make 78 judgments to complete this section of the task. (Appendix B presents a sample of the test instrument.) The subjects were asked to rate the perceived similarity of pairs of scenarios by marking an "X" on a five inch undifferentiated line scale. 4 Since the line represented a continuum (with one pole marked "same" and the other marked "different"), the author used 43 o ass1gn scores to the marked line by takin a ruler and measuring the distance from the zero point (the "same" pole) to the "X." The five inch scale also helped in gathering adjective descriptor data. The objective in gathering adjective descriptor data is to facilitate the explanation, interpretation, and definition of the underlying attributes (complexity dimensions) in the MDS analysis. The adjective scales derived were related to the similarity judgments of complexity through a specialized regression procedure. As Table 2 shows, seven descriptors were used. (Ten descriptors were pilot tested and statistically 5 analyzed. From these ten, seven emerged as disinct, meaningful descriptors.) 44 TABLE 2 PHASE ONE PROPERTY FITTING ADJECTIVE DESCRIPTORS (Five-inch line not to scale for this example) PLEASE INDICATE HOW YOU FIND THIS TAX SITUATION BY MARKING AN X ALONG EACH LINE: Very Very Non technical 1----------------------t Technical Big Break for 1----------------------Very Undeserving Fair Very Easy to Impossibl Catch ---------------------ito Catch Cheaters Cheaters Very Very Unfamiliar 1----------------------t Familiar Area of Area of Very Little --------------------.... Widesprea Abuse Abuse Of No Personal Extremely Benefit 1--------------------- Beneficial To Me To Me Law Likely To 1--------------------- Law Like l Remain Same To Change 45 In addition to defining complexity by processing the first two types of data, the author gathered background and attitudinal information. (Table 3 contains a list of the questions asked.) A cluster analysis based on the background and attitudinal information grouped subjects intc profiles, and then these profiles were related by canonical correlation to individual subject weights of the underlying dimensionality of the tax scenarios. This procedure allowec the author to ascertain whether any significant relationships existed between the subjects' background and attitudinal variables and their complexity assessments. 46 TABLE 3 GENERAL BACKGROUND QUESTIONS 1. Please circle your sex: MALE FEMALE 2. Do you usually prepare your own tax return? Do you usually file a long (1040) form? y y N N Please mark an X on the line where it best expresses your feelings: 3. In terms of your earnings life cycle, how do you place yourself: 0. OF YRS. WORKED 0 5 10 15 20 25 30 35 40 45 50 OR MORE 4. Please indicate your education: LESS THAN GRADE 8 9 10 11 12 YEARS COMPLETED OF HIGH SCHOOL 1 2 3 4 5 6 OR MORE YEARS OF COLLEGE COMPLETED 5. Please indicate your (and your spouse, if applicable) average annual income: (in thousands of dollars) $0 10 20 30 40 50 60 70 80 90 100 OR MORE 6. How do you rate the general fairness of the income tax laws? (Note: "5 inch line" not to scale for this example) ery Unfair Very Fair a person wanted to evade income taxes, what percentage do you think they could evade and not get caught? one All 8. In terms of complexity, how do you rate the present income tax system? Very Simple ~------------------------------------------- Extremel Complex 9. Of the people you know, what percentage do you think pay all their legally owed income taxes? o One Everyone percentage of taxpayers' money used wisely by the federal government is: one All 47 The Subjects In order to test perceptions of tax complexity, the author needed a cross section of taxpayers. In addition, the pilot test results indicated that these subjects should have at least an hour of uninterrupted time to concentrate on the task. In view of these constraints, the author arranged to interview people awaiting jury duty at the Los Angeles County Courthouse. This group was suitable for the present study because the California jury selection system has been strengthened in recent years to ensure that it accurately reflects the general population (only working police officers and persons over 65 are exempted from service without going through a lengthy appeals process). Also, the prospective jurors were confined for approximatel1 one and a half hours between the morning roll call and the roll call for jury selection; consequently, most of them were willing to donate the time and attention necessary to complete the study. The author pulled names at random from a hopper containing the names of all those available that day for jury duty. The individuals selected were approached, told of the voluntary nature of the study, and asked to participate. Eighty-eight percent of the 34 peoplE approached agreed to participate. These 30 participants adjourned to quiet conference rooms and were given the task to complete. All of them finished. (The times for 48 completing the instrument ranged from 40 to 105 minutes with 70 minutes the average length of time.) The Model Selection The author selected the MULTISCALE MDS program developed by Ramsey (1978) for this study. MULTISCALE, uniquely among MDS programs, provides output that permits the user to apply some statistical tests (Schiffman, Reynolds, and Young, 1981). MULTISCALE allows one to fit four MDS models to the subjects' data, one of which yields only a group configuration while the other three yield both a group and an individual differences configuration. MULTISCALE is based on lognormal or normal distributional assumptions and provides maximum likelihood statistics that can be used to select the "best" model and the "best" number of dimensions, in a statistical sense. This feature of MULTISCALE is important since the number of meaningful dimensions cannot be specified a priori. Consequently, it requires a two stage analysis. The first stage is an exploratory MDS where one evaluates various MDS models for their ability to "fit" the data and tentatively to interpret the configurations. The second stage is a confirmatory MDS where the model selected in the exploratory phase is subjected to statistical tests of significance for the most meaningful number of dimensions to extract, given the goal 49 of interpretability. MULTISCALE is a metric model requiring interval or ratio data. Since the author had no a priori way of knowing which MDS model and number of dimensions was most appropriate for the analysis, she analyzed the similarity judgments collected in Phase 1 (Appendix C lists the 78 judgments recorded on each of the 30 subjects) using different MULTISCALE models and dimensions. All MDS models consist of an algebraic equation to summarize the similarity judgment data and a corresponding geometric model to represent the data in an Euclidean space. Four MULTISCALE models are available (Schiffman,et.al:,1981, pp. 211-17): M1: This model is a metric scaling of a single dissimilarity matrix (CMDS) or replicated matrices (RMDS). This model does not allow for differences in scale use by the subjects or for individual differences in dimensional saliences. However, this model forms the basis for the more elaborate models M2, M3, and M4. M2: Two response parameters are added to model M1 to yield model M2, which is a metric, matrix conditional RMDS model. The response parameters allow for differences between subjects in scale use whereby a subject's dissimilarities are assumed to be approximately a power function of the group distance. M3: This model is a metric, matrix conditional model that introduces weights for every subject on each dimension. The weights are both row and column normalized (they are adjusted so that their average is one for each subject and for each dimension) , and show the relative importance of the dimensions for each subject. M4: This model is an extension of M2. Both models are 50 metric, matrix conditional RMDS models. But, whilE M2 provides a common standard error measure, M4 provides variable standard errors for coordinate positions. Thus, this model provides a measure of cognitive uncertainty for each stimulus point. The author ran the similarity judgment data using all four models with two, three, four and five dimensional solutions. The MULTISCALE output on each model provided an unbiased standard error estimate and a log likelihood statistic. These key statistics and related chi square computations and levels of significance appear in Table 4. The standard error estimate indicates how well the data fits the model. In general, model M3 provides the best fit. But even with this model, the standard error is high and indicates that from .836 to .887 of the variation in the data may be noise. In attempting to assess the impact of the level of noise in the model, it should be noted that studies involving conceptual phenomena generally include a greater amount of noise than studies with physical phenomena. Physical phenomena are usually more perceptually distinct and hence have a lower standard error. However, Schiffman, et al. (1981), reporting on several studies involving physical phenomena (taste studies), indicated that even for the physically based, textbook examples they selected, "The size of the standard errors are at first disconcerting for the data examined. In the examples provided in this 51 chapter, the models were unable to fit some 50 to 70% of thE variability in the data. Nevertheless, with sufficient subjects, interpretable configurations were developed" (p. 2 3 5) • (The number of subjects in the taste studies ranged from 10 to 46 with a mean of 29.) 6 After examining the standard error and selecting model M3, the next step was to select the appropriate dimensionality, a step that involved the log likelihood statistic. By comparing log likelihood statistics for two different solutions, one can perform a significance test. As indicated in Table 4, a chi-square value is given by twice the difference between the log likelihood values, X 2 =2 (HO-LD), where HD is the log likelihood for the model of higher complexity or higher dimensionality, and LD for the lower (Schiffman, et al., 1981, p. 219). Reviewing the computations in Table 4 for model M3, one sees that the chi-square values and the standard error show a significant improvement in moving from three to four dimensions, but not from two to three or four to five. Thus, the individual difference model M3 in four dimensions appears most appropriate for interpretation. 52 TABLE 4 EVALUATION OF THE FOUR MULTISCALE MODELS Log Unbiased Standard # of Deg. of Model Dim. Likelihood Error Estimate Freedom Ml 2 -1036.077 1. 000 2072 Ml 3 -972.028 .972 2062 M1 4 -957.623 .968 2053 M1 5 -953.692 .968 2045 M2 2 -757.045 .887 2019 M2 3 -743.952 .883 2009 M2 4 -735.356 .882 2000 M2 5 -729.638 .881 1992 M3 2 -732.898 .883 1992 M3 3 -717.951 .854 1954 M3 4 -598.953 .844 1916 M3 5 -557.780 .836 1878 M4 2 -751.841 .872 2007 M4 3 -735.754 .887 1997 M4 4 -727.884 .885 TEST FOR APPROPRIATE DIMENSIONALITY USING MODEL M3 Test Chi Square Degrees Model, LD Computation Result of vs HD 2(HD-LD)= X Freedom p M3,2 vs 3 2(-717.951 + 732.898)=14.947 38 > .10 M3,3 vs 4 2(-598.953 + 717.951)=118.998 38 > .01 M3,4 vs 5 2(-557.780 + 598.953)=41.173 38 > .10 LD = Model of lower dimensionality. HD = Model of higher dimensionality. 53 An Introduction to Dimension Interpretation The interpretation of the dimensions in a solution is, by its very nature, a subjective process. Schiffman, et al (1982) suggests that "the first step is to look at the properties of stimuli at each end of the dimension to determine if there is some attribute that changes in an obvious fashion" (p. 12). Shockley and Holt (1983) provide the following insights: "The interpretation of a solution's I \dimensions usually depends on a researcher's general knowledge of the research area or access to additional data which might aid in their identification" (p. 13). The additional data available in this study is the adjective descriptor (property) measures recorded for each of the 13 stimuli. Table 5 shows the results of the analysis of the descriptor measures. Part 1 focuses on the rho correlations between the subjects' measures on each property and the final placement of the 13 stimuli on the MDS configurations. The rho correlations reflect the importance of the respective properties to the MDS f . . 7 con lguratlon. Part 2 shows the positive or negative covariance of each of the seven properties with each of the four dimensions. These measures will be utilized to help interpret the MDS output shown in Figures 5 and 6. The first column of Table 5, Part 1, shows the propert1 numbers, reflecting the order in which the properties were 54 presentee to th-e sunJects. Tne second column shows the rho correlation between subjects' ratings on the property descriptors and the final MDS configuration. These measureE indicate the significance of the respective adjective descriptors in explaining the observed MDS judgments. The final column in Part 1 lists the property descriptors, whict were the adjectives used to label the anchors on the five-inch scales. 55 TABLE 5 PROPERTY FITTING CORRELATIONS PART 1: THE MAXIMUM CORRELATION BETWEEN THE PROPERTY AND THE MDS CONFIGURATION PART 2: PROP. NO. 1 2 3 4 5 6 7 PROPERTY # RHO 1 .89 2 .85 3 .78 4 .80 5 .75 6 .63 7 .63 PROPERTY DESCRIPTION Nontechnical - Technical Big Break for Undeserv ing - Very Fair Very Easy to Catch Cheaters - Impossible to Catch Cheaters Very Unfamiliar - Very Familiar Area of Very Little Abuse - Area of Wide spread Abuse Of No Personal Benefit - Extremely Beneficial Law Likely to Remain Same - Law Likely to Change DIMENSION AND PROPERTY CORRELATIONS BRIEF D1 D2 D3 D4 DESCRIPTION .69 -.24 -.16 .67 TECHNICAL -.39 -.49 .61 -.49 FAIR -.08 .39 -.75 .53 CATCH CHEAT. -.49 -.21 .13 -.84 FAMILIAR .14 .22 -.81 .53 ABUSE -.02 -.42 .68 -.60 BENEFICIAL .15 -.65 .19 .72 LAW CHANGING 56 In order of appearance, these descriptors attempt to capture a subject's perceptions of the stimuli regarding thE degree of technicality, fairness, likelihood of catching cheaters, degree of familiarity, existence of abuse, amount of personal benefit, and likelihood of the law changing. The rho correlation for each of these attributes follows: .89 for technical, .85 for fairness, .78 for catching cheaters, .80 for familiarity, .75 for abuse, .63 for personal benefit, and .63 for the law changing. Thus the first five properties, whose rho correlations range from .89 to .75, are particularly well correlated with the final MDS configuration (dimensions Dl-D4). The correlations between the seven properties and four dimensions are specified in Part 2 of Table 5. It is not whether a property is positively or negatively correlated that is important, but rather the magnitude of the correlation (that is the closer to 1.0 or -1.0, the more significant the correlation). A negative correlation indicates an inverse relationship. For example, the highest property and dimension correlation found in Part 2 is a -0.84 for familiarity on dimension D4. The adjective descriptor familiarity has two anchors "Very Unfamiliar" and "Very Familiar." Low scores are associated with the "unfamiliar" end of the spectrum, with increasing scores S7 reflecting increasing familiarity. Dimension D4 is associated with readability (a point to be elaborated upon in the following pages). More specifically, the low end of dimension D4 is associated with a low readability of the stimuli and the high end with easy reading. Thus, the negative correlation of -0.84 indicates that unfamiliarity is associated with stimuli that are difficult to read, and familiarity with stimuli that are easy to read. In essence, those correlations with a high absolute value are of interest, while those of low correlation convey little information and are ignored. Properties that correlate highly with a particular dimension may help interpret that dimension. As can be seer from the matrix in Part 2, property Pl (technical) has a 69 percent correlation with dimension Dl. All the remaining properties have less than a .50 correlation with dimension Dl and are, therefore, not particularly useful in interpreting this dimension. Dimension D2 is most highly associated with property P7 (law changing), with a -.65 correlation. Again, all the remaining properties have less than a 50 percent correlation with D2, and so they have little relevance in explaining this dimension. Dimension 03 is significantly, inversely correlated with properties P5 (abuse) with a .81 and P3 (catching cheaters) with a -.75. Dimension D3 is significantly and positively correlated with 58 proper~y Pb lpersonal oenerlt or law) Wlth a 68 percent, and property P2 (fairness) with a 61 percent. Accordingly, any explanation of dimension D3 will consider these four properties. As discussed earlier, dimension D4 is inversely associated with property P4 (familiar) with a -.84. In addition, dimension D4 is significantly associated with property P6 (changing law) with a .72, and property Pl (technical) with a .67. Consequently, these three properties will be considered in explaining dimension D4. Since the MDS solution is four dimensional, it is not possible to depict it on one graph. Consequently, in order to facilitate the analysis and discussion of the final MDS ~onfiguration, this discussion is divided into two sections. The first section relates to Figure 6 and focuses on the interpretation of dimensions Dl and D2. A discussion of ~imensions D3 and D4 shown in Figure 7 then follows. Viewed together, Figures 6 and 7 and the related analysis provide an empirical definition of tax complexity. An Interpretation of Dimensions One and Two The MDS output for dimensions Dl and D2 appears in Figure 5. Dimension Dl is reflected in the spacing of the ~3 stimulus points along the horizontal plane, while ~imension D2 is reflected in the spacing between points on Lhe vertical plane. The vertical spread between points-like 59 extra exemptions (stimulus S7) and child care expenses (stimulus Sl) - are irrelevant in interpreting dimension Dl, since only the horizontal spacing is of interest. Similarly, the spread between extra exemptions (S7) and ACRE (S5) is irrelevant in interpreting dimension D2, since only the vertical spacing is of interest. In the horizontal dimension in Figure 5 (dimension Dl), the subjects appear to be distinguishing the stimuli solely on the basis of topic. An examination of the stimuli reveals no other systematic differences. Topics dealing with personal matters (exemptions for age and blindness, child care, health insurance, entertainment expenses, charitable contributions, and gifts) are to the left of the axis, while topics dealing with financial matters (interest, individual retirement accounts, trusts, ACRS, income averaging, nonrecognition of gains and losses, and capital gains) are to the right of the axis. The adjective descriptor statistics support this conclusion. The propert] that correlates most highly with dimension Dl is property P1 (nontechnical-technical) with .68. This correlation is compatible with the interpretation that the subjects differentiate stimuli based on the topic, since the stimuli are perceived to be increasingly technical as the location of the stimuli proceeds from left to right (personal to financial). 60 0'1 I-' Dlft I PERSONAL TOPIC ILONI FIGURE 5 - CONFIGURATION PLOTS FOR DIMENSIONS Dl & D2 72.0 Extra Exuptions 7 49.0 21 .a 5. 7 -16.4 -38.5 -60.6 -82.7 Cbi ld Cue Expenses I -104.8 -120.9 -149.0 -250.3 -208.7 -167.1 Dlft 2 • IIUANTITATIVENESS ILDMI Health Insurance 8 Entertainaent Costs 4 Contributions 2 lifts l Aw•rd& 6 -125.5 -83.9 -42.2 -0.6 Interest II 41.1 DIN 2 • QUANTITATIVENESS IHI6HI Nonrecognition of Gains l Losses IRA 13 82.6 Trusts 12 9 ACRS 5 Capital Gains l Jnco1e Averaging 10 124.2 165.9 Dl" I FINANCIAL TOPIC IHI6HI The conf1gurat1on plot for d1mens1on D2 (the vertical plane in Figure 5) shows a distinct dispersion of stimuli with eight near the top, three spread in the middle, and two near the bottom of the figure. A careful reading of the 13 stimuli (Appendix A presents the text of all 13 stimuli) reveals that none of the stimuli above the horizontal axis contain any numerical manipulations while all the stimuli below the axis contain some calculations. The two highest and lowest stimuli points on the vertical plane illustrate this observation. The text for the two highest stimuli read as follows: 12 - NONRECOGNITION OF GAINS AND LOSSES Sec. 1031 is a very important tax law for real estate investors. This law allows some direct exchanges of business and investment property without taxation. Sec. 1031(a) states in part "No gain or loss shall be recognized if property • . is exchanged solely for property of a like kind. " The Treasury Regulations are rather liberal in interpreting what is "like kind''; a country farm can be exchanged for a city apartment without taxation. 8 - HEALTH INSURANCE Health insurance paid by an employer for an employee is not taxed to the employee. This can be very beneficial for the employee who wants insurance. For if an employee got an increased salary instead of insurance, he would have to pay income tax on the extra income, and then buy health insurance with what was left. 62 11n con~ras~, tne two lowest st1mu11 read as follows: 10 - INCOME AVERAGING What happens when a $30,000 a year baseball player gets a million dollar contract? Income averaging provisions were enacted in 1964 and liberalized in 1969 to help reduce the taxes of people with fluctuating incomes. The Code specifies the conditions and calculations for income averaging in Sees. 1301-1305. First the taxpayer's average taxable income for the prior four years is calculated. This average is then multiplied by 120 percent. The result is subtracted from the current year's taxable income. If the difference is $3,000 or more, then income averaging may be used. 2 - CONTRIBUTIONS Before 1982 if you didn't itemize deductions you didn't receive any tax breaks for charitable contributions, so Congress changed the tax law for the years 1982 through 1986. In 1982 and 1983, 25% of charitable contributions up to $100 may be deducted from gross income (a $25 deduction) . For 1984, the amount is 25% of $300 (a $75 deduction). In 1985, 50% of all contributions may be deducted (with no dollar limit). In 1986, the deduction is 100% of all contributions (again with no dollar limit). In 1987, however, this provision expires. Thereafter charitable contributions will once again be allowed only if deductions are itemized. Also, only one of the eight stimuli above the axis refers to a change in the tax law, while all except one of the five stimuli below the axis refer to a change in the tax law. Furthermore, the stimuli at the extremes show, respectively, ~he least and the greatest amount of change and computation ~s the Table 6 analysis reveals (proceeding from the highest pn the configuration to the lowest) . 63 TABLE 6 ANALYSIS OF DIMENSION D2 STIMULI TOPIC Nonrecognition of Gains and Losses Health Insurance Trusts Entertainment Expenses Exemptions Interest Expenses ACRS Gift Capital Gains IRA Child Care Income Averaging Contributions CALCULATIONS No No No No No No No No Moderate Minimum Moderate Heavy Heavy CHANGES No No No No No No Yes No No Yes Yes Yes Yes These conclusions are also supported by the property fitting results since property P7 (law likely to remain the same-lav, likely to change) is inversely correlated with dimension D2 (a .65 rating). For brevity, dimension D2 is labeled the quantitativeness dimension since the focus of the subject's discrimination is the degree of change and the amount of computation. The adjective descriptor data is also helpful in interpreting dimension D3. As was noted earlier, dimension 03 is negatively correlated with property P5 (abuse) with a -.81, and property P3 (cheating) with a -.75. In addition, dimension D3 is positively correlated with property P6 (personal benefit) with a 68 percent, and property P2 (fairness) with a 61 percent. Thus, these properties deserve special scrutiny. Table 7, showing the 30 subjects' 64 mean score on the adjective descriptors for each of the stimuli, is an aid in this regard. Since a five inch line is used to gather the property fitting data (Table 2 shows an example of the format) , the minimum score possible is zero and the maximum is 5.0. The properties are listed across (from property Pl to P7), and the stimuli are listed down (from stimuli Sl to 813) to form a matrix of the subjects' mean scores. Thus, the score appearing in the upper left of the first row indicates that the 30 subjects marked an X along the line an average of 2.05 inches from the left anchor. This score relates to property Pl, which is anchored by the adjective descriptor "Very Nontechnical" on the left (at the zero point), and "Very Technical" on the right (at the end of the five inch line). In essence, the mean scores in Table 7 detail data inherent in the dimensior and the property correlation statistics shown in Part 2 of Table 5. 65 TABLE 7 PROPERTY FITTING RATINGS Pl P2 P3 P4 P5 P6 P7 CATCH PERS STIMULI TECH FAIR CHEAT FAMIL ABUSE BEN CHANGE Sl Child Care 2.05 3.58 2.10 2.62 2.72 1. 03 1. 90 S2 Contributions 2.31 2.27 2.83 2.41 2.98 1. 95 2.17 S3 Capital Gains 2.75 1. 89 2.23 2.22 2.45 1. 21 1.45 S4 Entertainment 1. 4 7 1. 60 3.38 2.60 3.64 0.75 1. 44 S5 ACRS 2.42 2.35 2.11 1.92 2.33 1. 07 1.57 S6 Gift-Award 2.44 2.16 2.92 1. 85 2.75 1.56 1. 81 S7 Ex. Exemp. 0.88 3.81 1.86 2.95 1.19 1. 29 1.18 S8 Health Ins. 1.19 3.38 1.55 3.41 1.17 3.00 1. 96 S9 Trusts 2.57 1. 33 2.51 1. 65 2.91 1.13 1.51 SlO Income Avg. 2.23 3.02 1.77 2.48 2.03 1.82 1. 72 Sll Interest 1. 73 2.78 2.47 3.21 2.73 2.37 1. 4 7 Sl2 Nonrecog.G&L 2.42 2.09 2.41 2.03 3.21 1. 06 1. 98 S13 IRA 1. 51 3.52 1.42 2.99 1. 29 2.56 1. 75 In viewing dimension D3 on Figure 7, note that entertainment cost (stimuli S4) is isolated on the extreme left side of the plane, while IRA (stimuli Sl3) is the farthest to the right. Examining Table 6, observe that entertainment cost is rated the highest in abuse, highest in cheating, lowest in personal benefit, and second lowest in 66 fairness. On the opposite side of the plane, IRA is rated one of the lowest in abuse, the lowest in cheating, second highest in personal benefit, and one of the highest in fairness. Proceeding left to right, this pattern of decreasing abuse and cheating ratings, and increasing personal benefit and fairness ratings is confirmed across the dimension 03 stimuli. The composite of these four properties appears to be the concept of social justice; the concept distinguishes stimuli perceived to be subject to widespread abuse, and to be a haven for cheaters, from thosE thought to be personally beneficial and fair. Consequently, the author has labeled dimension 03 "the social justice dimension." 67 0"\ (X) FIGURE 6 CONFIGURATION PLOTS FOR DIMENSIONS 03 ~ 04 Dl" J 181.8 118.9 m.o 77.1 11.2 Enterhinaent Costs I Contributions 2 SOCIAL 5.3 Capital Gains !JUSTICE! -3-- ILOWI -30.6 -66.1 -102.3 -138.2 -174.1 -189.5 -159.4 -129.3 -99.1 ACRS 5 -69.0 Dl" 4 - READABILITY ILOil Nonrecognition of Gains l losses 12 I -38.9 lncoae Averaging 10 Interest llfducti ons II -8.8 21.3 Dl" 4 - READABILITY IHISHI Gi It L ANards 6 Child Care Expenses I Health Insurance 51. I Trusts 9 Extra Exeaptions 7 81.5 IRA 13 Ill. 7 mJ. SOCIAL IJUSTICEl !HIGH! Dimension D4 (the vertical plane of Figure 7) is interesting because 11 of the 13 stimulus are concentrated in the middle, with one stimuli at the extreme top of the configuration (stimuli S6) and one at the extreme bottom of the configuration (stimuli S11). The text of these two outlying stimuli follows: 6 - GIFTS AND AWARDS You are not taxed on a $5,000 gift from your rich uncle or on a $5,000 Pulitzer Prize. However, you are taxed on a $5,000 lottery ticket winning, or church raffle prize. Basically the law says that a person who receives a gift is not taxed, but a person who receives a prize or award is taxed unless they meet the following Sec. 74(b) exception: "Exception. - Gross income does not include amounts received as prizes and awards made primarily in recognition of religious, charitable, scientific, educational, artistic, literary, or civic achievement, but only if- (1) the recipient was selected without any action on his part to enter the contest or proceeding, and (2) the recipient is not required to render substantial future services as a condition to receiving the prize or award." 11 - INTEREST DEDUCTION Interest expense is the most important itemized deduction for individuals. In 1979 this deduction totaled $73.6 billion, or 40.7 percent of all itemized deductions. Virtually all interest paid is deductible. The major exceptions are interest on money borrowed to purchase tax free securities and a limit on the amount of interest deductible on funds borrowed to purchase investment property. Focusing on the two outlying stimuli, one quickly notices three differences. Stimuli S6 (gifts and awards) is the longest scenario, quotes the Internal Revenue Code, and 69 rated very low in familiarity. Stimuli Sll (interest deduction), at the opposite extreme, is a short scenario, does not quote or even refer to the code, and is rated very high in familiarity. In general, this pattern holds for the stimuli in the middle as well. Those stimuli above the horizontal axis tend to be longer, more closely associated with the code, and in less familiar areas of the law than the stimuli below the axis. The property descriptor ratings also appear relevant in interpretating this dimension. The most highly correlated property is a negative 84 percent with property P4 (familiarity). In addition, dimension 04 correlates positively with property Pl (technical) at 67 percent, and with property P7 (changing area of the law) at 72%. The stimuli above the horizontal axis on Figure 6 tend to be perceived as unfamiliar, technical, and liable to change, while stimuli below the axis are relatively familiar, nontechnical, and unlikely to change. This discrimination based on the length of the scenario and degree of association with the Internal Revenue Code is probably indicative of differences in the readability of the stimuli. Unfamiliar, changing, and technical scenarios alsc seem to be the most difficult to read 8 . 70 Results: Other Phase One Questions Two peripheral issues in the Phase 1 inquiry involve the subjects' weighting of the complexity dimensions and thE relationship of the weightings to their background information and attitudinal ratings. The subjects' weights provide a measure of the importance to them of each dimension. The weights are normalized so that the weights for each subject on the four dimensions total four. The weights for the first subject in Table 8 indicate that dimension D1 (topic), with a final weight estimate of .008, is virtually unutilized. Dimensions D2 (quantitativeness) and D3 (social justice), wit h weight estimates of 1.531 and 1.719, are relatively important cues. Dimension D4 (readability) is of less than average importance (1.0 represents an average weighting since the scores are normalized). Thus, for the first subject, the quantitativeness and social justice dimensions (D2 and D3) form over 80 percent of the basis for differentiating the tax complexity of the scenarios. Table 8 shows the final weight estimates for each of the 30 subjects on the dimensions. Note that two or three subjects in each dimension place the majority of the weight on that dimension. (Subjects 6, 16, and 24 in dimension D1; 11, 18, and 26 in dimension D2; subjects 3 and 28 in ~imension D3; and subjects 8 and 9 in dimension D4). Also, 71 if 1.5 is used as a cut-off, more subjects emphasize the readability dimension (04) than any other. Nine out of the 30 subjects indicated final weight estimates of 1.5 or above on dimension 04; while only 5 on dimension 01 and 02 (topic and quantitativeness), and 7 on dimension 03 (social justice) have final weight estimates of this magnitude. A review of the subjects' weighting of the complexity dimensions, however, shows no significant relationship between these weightings and the subjects' background information or attitudinal ratings, as measured by canonical analysis. 72 TABLE 8 SUBJECT'S WEIGHING OF THE COMPLEXITY DIMENSIONS FINAL WEIGHT ESTIMATES SUBJECT DIM 1 DIM 2 DIM 3 DIM 4 TOPIC QUANT SOC.JUS. READ 1 0.008 1.531 1.719 0.741 2 0.580 0.825 1.554 1.042 3 0.010 0.491 3.488 0.011 4 1. 420 1. 030 1. 540 0.011 5 0.800 0.854 1.294 1.052 6 2.663 0.125 0.459 0.753 7 0.420 0.938 0.785 1.858 8 0.250 0.950 0.686 2.114 9 1.758 0.074 0.012 2.155 10 0.648 0.951 0.435 1. 966 11 0.009 2.474 0.012 1.506 12 0.818 1. 4 76 1.099 0.607 13 0.778 1.028 1.331 0.863 14 1.455 0.514 0.817 1. 214 15 1.592 0.424 1. 579 0.404 16 2.597 0.011 0.507 0.885 17 0.838 0.698 0.868 1. 596 18 0.010 3.938 0.013 0.039 19 0.643 0.752 1.326 1. 278 20 1.152 0.990 1. 010 0.848 21 0.866 1.445 0.110 1.578 22 1.189 0.631 1. 426 0.754 23 1.948 1. 030 0.217 0.805 24 2.169 0.895 0.579 0.358 25 1. 482 0.368 0.264 1.885 26 0.147 2.260 0.012 1.582 27 1. 564 0.533 1.780 0.122 28 0.201 1. 636 2.153 0.010 29 1.141 0.149 1.765 0.945 30 0.863 0.967 1.141 1.029 73 Canonlcal correlatlon analysls provides information about the strength of the relationship between pairs of variates. The variates of interest in this analysis are the biographical-attitudinal scores and the complexity dimensions. When squared, the canonical correlation represents the amount of variance in one canonical variate that is accounted for by the other canonical variate. This can also be described as the amount of shared variance between the two canonical variates. Since canonical correlation analysis focuses on accounting for the maximum amount of the relationship between two sets of variables, the result is that the first function is derived so as to have the highest intercorrelation possible. The second pair of canonical variates (dimension D2, in this study) is ~erived to exhibit the maximum amount of relationship ~etween the two sets of variables that is not accounted for ~y the first pair of variates (dimension Dl). In short, successive pairs of canonical variates (related to ~imensions D3 and D4) are based on residual variance and their respective canonical correlations become smaller as ~ach additional function is extracted. As with all ~orrelation coefficients, canonical or otherwise, various statistics can be utilized to assess their level of significance. 74 The canonical correlation analysis of the biographical-attitudinal variables and the complexity dimensions revealed that the computed canonical correlation coefficients were not highly significant. Table 9 shows the summary canonical correlation statistics in Part 1. the squared canonical correlation (representing the shared variance) in Part 1 is .62 for dimension 1 (topic), .53 for dimension 2 (quantitativeness), .30 for dimension 3 (social justice), and .19 for dimension 4 (readability). Because of the canonical correlation procedures, the first pair of canonical variates (related to dimension D1) exhibit the highest intercorrelation, the next pair (dimension D2) involves the second largest correlation, and so forth through the fourth function, which is the smallest. The related F-statistics and levels of significance show that the "most significant" correlation was at the .17 level of significance. Consequently, it appears that none of the relationships between the background information and the complexity dimensions merit further investigation. 75 TABLE 9 PART 1: SUMMARY CANONICAL CORRELATION STATISTICS CANONICAL F TEST DEGREE OF LEVEL OF R-SQUARED FREEDOM SIGNIFICANCE DIM1(TOPIC) .62 1. 30 40 .17 DIM2(QUANT) .53 1. 07 27 .40 DIM3(SO.JUST) .30 .74 16 .74 DIM4 (READ.) .19 .64 7 .72 PART 2: CORRELATIONS BETWEEN THE BIOGRAPHICAL/ATTITUDINAL VARIABLES AND THE COMPLEXITY DIMENSIONS DIM 1 DIM 2 DIM 3 DIM 4 (TOPIC) (QUANT) (SO.JUST) (READ) A Sex -.16 0.24 -.05 -.05 B Ret Prep 0.18 0.23 -.33 -.09 c Age 0.12 0.01 -.27 0.16 D Education 0.17 0.00 -.27 0.12 E Income -.08 0.27 -.13 -.08 F Fair 0.33 -.29 -.12 0.13 G % Evade -.47 0.11 0.41 -.10 H Complex 0.10 0.17 0.02 -.36 I % Pay Tax -.07 0.20 0.03 -.21 J Used wise 0.01 0.16 0.05 -.28 Table 9, Part 2, shows the structure correlations between the biographical-attitudinal variables and the complexity dimensions. Generally, the larger the coefficient, the more important the variable in deriving th canonical correlation statistic in Part 1. Consequently, i appears that variable G, with a -.47, contributes the largest relative amount to dimension D1's canonical R-squared of .62. The negative sign before the .47 indicates an inverse relationship (omission of the sign indicates a direct relationship). Since all of the 76 structure correlations in Table 9 are relatively low, it appears that no one variable is of special importance in deriving the canonical correlations. The literature review in Chapter 2 indicates that certain biographical and attitudinal measures may relate to taxpayer behavior. This study focuses on the relationship between complexity and taxpayer behavior. Consequently, it is relevant to question whether any of the suggested background variables also relate to a subject's tax complexity assessments. In the gathered data no significant relationships appear. As noted in Chapter 2, previous studies conflict on the relevance of the biographical and attitudinal measures. The results of this analysis appear consistent with the results of two recent studies, Hotaling and Arnold (1981) and Spicer and Becker (1980), except that the latter found sex to be a significant variable. The matrix in Table 10 shows the correlations among the background variables. (See Table 3 for a list of the 10 background questions.) Of special interest is any relationship between biographical data and attitude measures. Any strong relationship between these variables may have implications for clustering taxpayers and predicting behavior. A review of the correlations in Table 10, however, reveals only one above .50 is recorded, a .60 between education and income. This leads to the 77 unremarkable conclusion that a moderate association exists between these two variables. This evident lack of correlation between biographical data and attitude measures is consistent with recent studies that show negative attitudes towards the tax system and government pervade all population subgroups. TABLE 10 CORRELATIONS AMONG THE 10 BACKGROUND VARIABLES Sex RetPrp Age Educ Inc Fair %Evad Cmplx %Pay %WiE Sex RetPrp Age Educ Income Fair %Evade Complx % Pay %Wise 1.00 -.36 .21 -.08 .18 .18 -.23 .02 .07 -.02 -.36 1.00 .04 .41 .24 .21 .05 .04 .14 .42 .21 .04 1. 00 .46 .44 .04 -.18 .38 .14 .06 -.08 .41 .46 1.00 .60 .12 -.02 .33 -.09 .13 .18 .24 .44 .60 1. 00 -.07 .17 .44 -.16 .16 .18 .21 .04 .12 -.07 1. 00 -.20 -.40 -.09 .04 Summary of Phase 1 Research Results -.23 .05 -.18 -.02 .17 -.20 1.00 -.04 -.27 .11 .02 .04 .38 • 3 3 .44 -.40 -.04 1.00 .15 .34 .07 .14 .14 -.09 -.16 -.09 -.27 .15 1.00 .10 This first research phase was conducted to address three questions: 1. What are the perceived dimensions of tax complexity? -.02 .42 .OE .1_ .lE -.0~ .1 .34 .10 1.00 2. When taxpayers are clustered according to their weighting of the complexity dimensions, what is the ·relationship between these clusters and the background or attitudinal variables? 3. Are there any significant interrelationships among the background and attitudinal ratings? The results indicate that there are four distinct 78 rcompleXlLY almenslons: personal 'Llnanclal tOplC or1entat1on, quantitativeness, social justice, and readability. And although subjects can be clustered into distinct groups based on their complexity assessments, these groupings appear to be related to no particular background or attitudinal measures. In addition, there appear to be no significant correlations between the biographical data and the attitudinal measures. The results of the MDS study provides a starting point for building scenarios of varying complexity in Phase 2 to test the effect of complexity on reporting position selection. 79 CHAPTER 4 PHASE TWO ANALYSIS AND RESULTS Upon completion and analysis of the first phase of the study, the author undertook the second phase. The four complexity dimensions from Phase 1 are manipulated in the Phase 2 tax cases. The second phase was experimental and required the subjects to estimate their preferred reporting position in four different tax situations of varying complexity. This second phase directly addressed the key question, the existence of a link between tax complexity anc reporting position. The Phase 2 Methodology Thirty-two distinct tax scenarios were constructed around four tax topics for Phase 2 testing. Each scenario was designed around the four complexity dimensions from the first phase and written to give a subject enough informatior to determine the correct tax reporting position (that is, the position likely to be judged compliant by the Internal Revenue Service) . The scenarios were also designed to introduce some ambiguity in order to give those subjects so inclined a basis for rationalizing noncompliant positions. Table 11 shows a full factorial design of 24 unique 80 ariations. This design provides for all high and low combinations on each of the four dimensions. The author selected a one-half replicated design (as indicated by the asterisk) from the full factorial. This one-half replicate is advantageous because it allows one to test for all main effects on each complexity level, as well as for the interaction between complexity levels, and still retain a manageable number of subjects (86 subjects were tested, with a minimum of 9 subjects per cell). TABLE 11 PHASE TWO RESEARCH DESIGN Design Topic Quant. So.Jus. Selected D1 D2 D3 L L L 1 L L L 2 L L H L L H 3 L H L L H L L H H 4 L H H 5 H L L H L L H L H 6 H L H H H L 7 H H L 8 H H H H H H D = Dimension L = Low Complexity Level H = High Complexity Level * = Combination selected for 1/2 replicated design Read D4 L H L H L H L H L H L H L H L H * * * * * * * * 81 Scenario Construction On the basis of the Phase 1 MDS results, the author constructed a quartet of four-dimensional cases, each dimension designed with a low and high complexity version modeled after the MDS analysis. Since dimension Dl is construed to be a personal versus financial topic orientation, dimension Dl in all four cases reflects this dichotomy. For instance, in Case A, this concept is operationalized in a charitable deduction context. The low complexity version describes a mechanic donating a van to his church, while the high complexity version describes gift giving through an annuity trust. The remaining three dimensions were structured in a manner similar to dimension Dl. Dimension D2 is interpreted to reflect taxpayers' perceptions of quantitativeness. Consequently, the low complexity version in all four cases was constructed so as to minimize numerical expressions and computations, while the high complexity versions were designed to emphasize these characteristics. Since dimension D3 appears to reflect social justice considerations, the low complexity versions attempt to depict taxpayers with untainted motives acting in socially acceptable ways. The high complexity versions, however, insinuate tax shelter motivations that may not necessarily be in the best interests of society. Lastly, the low and 82 ~igh complexity levels of dimension D4 are designed around the issue of readability. The goal in all four cases is to ~xpress the tax law in straightforward lay terms in the low ~omplexity version; and in long, intricate, legalistic passages in the high complexity version. The efficacy of the results in Phase 2 depends on how well one operationalizes the complexity dimensions developed in Phase 1. The objective is to design the high and low dimension descriptions with enough contrast and clarity to have an effect on reporting while still remaining true to the Phase 1 analysis. Appendix D contains the complete listing of the Phase 2 tax cases. The Subjects A total of 86 subjects were tested in Phase 2. The testing spanned the first week in June, 1983, and involved the same population and conditions described for the Phase 1 testing (Phase 1 testing preceded Phase 2 testing by five weeks and involved different subjects). The net participation rate in Phase 2 was 75 percent. The author selected 115 for testing, 83 percent of these agreed to participate, and 86 usable responses were received (an omission of any of the 12 reporting position selections invalidated the response) . 83 Generally, the subjects took about 20 minutes to complete the task, which required each of them to read four different tax scenarios and select the position they would report on their own tax return, would advise another to report, and believed to be the correct tax position according to the IRS. (Appendix D contains a listing of the high and low complexity descriptions for the four dimensions and also details the three reporting position questions for each case.) Pilot testing indicated that some subjects tended to respond more conservatively when asked to report what they would advise rather than what they would do. To measure and minimize the confounding effect of this conservatism, the author asked the Phase 2 subjects to make separate tax position judgments on what they would report and what they would advise. As discussed in the first chapter, subjects were also asked for their assessment of the correct tax position in order to establish another benchmark (besides the legally correct position as determined by the Internal Revenue Service). Consequently, in addition to a legal standard, subjects were measured against this ••psychological standard" to evaluate their tax aggressiveness. Appendix E contains a sample of the test instrument used in Phase 2. In addition to presenting four tax cases, the test instrument included the same background 84 ~uestionnaire as the one used in Phase 1 (which appears in ~able 3). Although no evidence of significant relationships ~merged between the background data and subjects' complexity ~ssessments in Phase 1, significant relationships could still be evident when the background data was correlated with subjects' reporting position selections in Phase 2. (All of the data collected from the 86 Phase 2 subjects appears in Appendix F.) The Statistical Procedures The author used three statistical techniques for Phase 2 testing. First, a multi-variate analysis of variance (MANOVA) evaluated the influence of the complexity level on reporting position (research question four). MANOVA also ascertained the effect of the individual complexity dimensions and tax topic on reporting position (research question five). Second, a simple analysis of variance assessed the significance of the differences between the correct tax position and the subjects' preferred tax position (research question six). Third, a canonical correlation analysis, evaluated the relationship between reporting position selections and the biographical and attitudinal information collected on the subjects (research question seven) . 85 Complexity Effect Results The principal issue of Phase 2, the impact of complexity, is addressed in research question four. Secondary issues involving tax scenarios (task effects), aggression tendencies, and background data are addressed in research questions five through seven. The research results for each of these issues are examined in sequential order in the following pages. The central issue of the entire study is embodied in research question four: "Does complexity significantly influence the tax reporting position?" The results strongly suggest that it does. Using MANOVA, the author found that all three tax reporting position selections, in all four cases, showed significance to the .0000 level in the F-test. These results appear in Table 12. 86 TABLE 12 MANOVA ANALYSIS OF THE EFFECT OF TAX COMPLEXITY ON REPORTING POSITION =========================================================== DEGREE LEVEL VARIATE F TEST OF OF FREEDOM SIGNIFICANCE CASE A - CHARITABLE DEDUCTION OVERALL MEASURE 181.47 3,76 0.0000 MEASURE 1 (You) 521.64 1,78 0.0000 MEASURE 2 (Advise) 485.47 1,78 0.0000 MEASURE 3 (Correct) 363.56 1,78 0.0000 CASE B - INCOME OR GIFT OVERALL MEASURE 118.94 3,76 0.0000 MEASURE 1 (You) 222.47 1,78 0.0000 MEASURE 2 (Advise) 275.23 1,78 0.0000 MEASURE 3 (Correct) 358.24 1,78 0.0000 CASE c - TRAVEL EXPENSES OVERALL MEASURE 318.83 3,76 0.0000 MEASURE 1 (You) 865.35 1,78 0.0000 MEASURE 2 (Advise) 929.09 1,78 0.0000 MEASURE 3 (Correct) 595.48 1,78 0.0000 CASE D - REHABILITATION CREDIT OVERALL MEASURE 154.42 3,77 0.0000 MEASURE 1 (You) 415.69 1,79 0.0000 MEASURE 2 (Advise) 452.05 1,79 0.0000 MEASURE 3 (Correct) 379.43 1,79 0.0000 87 The extremely low level attained on the significance tests indicates that there is virtually no likelihood that the findings in Table 12 are a chance occurrence. The results conclusively demonstrate that complexity does have an influence on subjects' reporting position choice. These are striking results considering that there were four separate cases, and that these cases involved four distinct topics. (Subjects were asked to make judgments regarding two different expense deduction scenarios, one income scenario, and one tax credit scenario.) It is also surprising that the results are insensitive to the question asked. Not only did the complexity level influence the subjects' choice of the position they would select (M1), but it also influenced the position they would advise (M2), and their perception of the correct tax reporting position (M3). In essence, the findings in Table 12 provide strong evidence supporting the assertion of the Committee of the New York Bar Association (1972), and such authors as Dean, Keenan, and Kenney (1980), that complexity does have a pronounced effect on tax reporting. In addition, these results provide support for the meaningfulness of the efforts of such authors as Schroeder (1975), and Karlinsky (1981), who have attempted to measure the complexity of the tax law. 9 From a policy perspective, these results provide support for the credibility of such flat-tax author~ 88 such as Hall and Rabushka (1982), who argue that simplifying the tax system will affect compliance. Given the significance of these results, the next question (addressed in research question five) is whether we now have enough information to analyze the results further to determine which dimension or combination of dimensions produces the effect. Influence of Complexity Dimensions Research question five asks, "Do any of the individual complexity dimensions (or combinations of dimensions), or does the tax topic, exert a significant influence on the reporting positions?" Table 13 details the MANOVA analysis for the one- and two-dimensional effects that attained a minimum of a .10 level of significance. The F-test statistics and related levels of significance appear in Appendix G for all possible one, two, and three dimensional effects. Since there appears to be no systematic pattern of influence for the three dimensional effects, they are not discussed further. Some of the one- and two- dimensional effects, however, are interesting from the perspective of providing insight into the tax complexity concept. Out of 40 possible one and two dimensional effects, ten attained at least a .10 level of significance and are detailed in Table 13. 89 TABLE 13 MANOVA ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS =========================================================== DEGREES LEVEL VARIATE F TEST OF OF FREEDOM SIGNIFICANCE ----------------------------------------------------------- CASE A: DIM. DI & D3 (Charity) (Topic & So.Jus) 2.21 3,76 0.0941 CASE A: DIM. D2 & D4 (Charity) (Quant & Read) 2.21 3,76 0.0941 CASE B: DIM. D2 (Inc.-Gift) (Quant) 2.24 3,76 0.0902 CASE B: DIM. D4 (Inc.-Gift) (Read) 3.27 3,76 0.0256 CASE C: DIM. D1 (Travel) (Topic) 2.58 3,76 0.0594 CASE C: DIM. Dl & D3 (Travel) (Topic & So.Jus) 2.48 3,76 0.0674 CASE C: DIM. D2 & D4 (Travel) (Quant & Read) 2.48 3,76 0.0674 CASE D: DIM. D4 (Rehab) (Read) 2.17 3,77 0.0986 CASE D: DIM. D1 & D3 (Rehab) (Topic & So.Jus) 3.01 3,77 0.0353 CASE D: DIM. D2 & D4 (Rehab) (Quant & Read) 3.01 3,77 0.0353 90 A striking feature of Table 13 is that dimensions 01 and 03 (topic and social justice), and dimensions 02 and 04 (quantitativeness and readability) combine to produce a significant effect in three of the four cases. Notice from Table 13 that the dimension 01-03 and 02-04 combinations are the only ones to produce a significant two dimensional effect. As one can see from the uni-variate statistics in Table 14, the main cause of the multi-variate effect observed in Table 13 can be traced to measure M1. In all six situations where the dimension 01-03 and 02-04 combinations attained the .10 cut-off level in the multi-variate testing, the measure for what "you would report" (measure M1) is the most significant of the three uni-variate, repeated measure statistics. 91 TABLE 14 UNIVARIATE ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS VARIATE CASE A (CHARITY) DIM. D1~3- MEAS. 1 (You) (Topic & S.J) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D2 & D4 MEAS. 1 (You) (Quant&Read) MEAS. 2 (Advise) MEAS. 3 (Correct) CASE B (INC.-GIFT) DIM. D2 ---- -MEAS. 1 (You) (Quant.) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D4 MEAS. 1 (You) (Read) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D1 (Topic) CASE C (TRAVEL) -MEAS. 1 (You) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D1 & D3 MEAS. 1 (You) (Topic & S.J) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D2 & D4 MEAS. 1 (You) (Quant&Read) MEAS. 2 (Advise) MEAS. 3 (Correct) CASE D (REHAB. ) DIM. D4 MEAS. 1 (You) (READ.) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D1 & D3 MEAS. 1 (You) (Topic & S.J) MEAS. 2 (Advise) MEAS. 3 (Correct) DIM. D2 & D4 MEAS. 1 (You) (Quant&Read) MEAS. 2 (Advise) MEAS. 3 (Correct) F TEST 2.52 1. 26 0.35 2.52 1. 26 0.35 6.66 5.05 2.55 9.39 4.61 3.85 0.78 3.45 4.93 4.38 0.97 1. 66 5.96 0.97 1.66 6.42 4.37 2.85 5.96 1.38 2.54 5.96 1. 38 2.54 DF 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,78 1,79 1,79 1,79 1,79 1,79 1,79 1,79 1,79 1,79 p 0.1166 0.2655 0.5547 0.1166 0.2655 0.5547 0.0117 0.0274 0.1143 0.003C 0.035( 0.053...; 0. 40U 0.067( 0.013L 0.039 0.327q 0.201[:; 0.016q 0.327q 0.201[:; 0.013. 0.039~ 0.095 0.016C 0.244( 0.115 0.016< 0.244( 0.115 92 It appears that the factors in the D1-D3 and D2-D4 dimensions overlap and consequently have a reinforcing effect. The most statistically significant D1-D3 and D2-D4 dimension combinations are found in Case D. The high levels of these combinations illustrate the overlap of the concepts in the personal versus financial topic and the social justice descriptions (D1-D3), and the quantitativeness and readability descriptions (D2-D4). CASE D, D1 & D3 High (Rehab) (Topic & S.J) Carson, a real estate speculator, acquired an old but basically sound business building in August 1981 as a part of a tax-free exchange of property. Rehabilitation of older structures has become one of the latest tax shelter areas because of the amount of tax credit available. Aware of tax law changes between 1981 and 1982 that increased benefits even further, Carson withheld payments to contractors until January 1,1982. CASE D, D2 & D4 High (Rehab) (Quant & Read) During September, Carson solicited contractor's bids. Even though the structure was approximately 38 years old, estimates for the needed renovation were $1,500 plumbing, $2,750 electrical, $1,800 masonry, $1,250 glass work, and $2,700 general carpentry for a total of $10,000. Construction began in October 1981 and was completed within the six month period ending March 1982. A separate percentage for qualified rehabilitatior expenditures applies for costs incurred after 1981. The rehabilitation credit is 15% for structures that are at least 30 years old, 20% for structures atleast 40 years old, and 25% for certified historic 9~ structures. (Prior to 1982, only the 10% investment tax credit was available for qualified rehabilitation expenditures). Rehabilitation expenditures qualify only for real property if made in connection with a substantial rehabilitation of the building. According to Internal Revenue Code Section 48(g) (1) (C) (i), a rehabilitation is substantial if expenditures during the 24 months ending on the last day of the tax year exceed $5,000. Carson is a cash basis taxpayer, wrote all $10,000 worth of checks to contractors in 1982, and is not sure of the exact age of the building. Consequently, he wonders what amount of credit he should take for rehabilitating the structure. Notice from the preceding descriptions that the concept of a real estate speculator involved in a tax-free exchange of property (D1) appears to reinforce the notion of rehabilitating an older structure for the tax shelter benefits (D3). Likewise, the quantitativeness (D2) and readability (D4) dimensions obviously overlap and thus appear to have the same reinforcing effects. The four remaining variates attaining a .10 level of significance in Table 13 involve single dimensions. The first two significant single dimensions shown relate to CasE B. Both dimension D2 (quantitativeness) and dimension D4 (readability) produced a significant effect. To facilitate discussion, the high and low levels for these dimensions arE shown. CASE B, DIM. D2 Low (Inc.-Gift) (Quant.) Warren figures that he does half the amount of work that he would have to do anywhere else for the same amount of money. CASE B, DIM. D2 High (Inc.-Gift) (Quant.) warren figures he only does about $300 worth of services a month ($120 worth of errands, $85 cleaning, and $95 miscellaneous). CASE B, DIM. D4 Low (Inc.-Gift) (Read.) According to the tax law, some amounts received are taxable while others may be tax exempt. For students, two important areas of tax exempt income are gifts and scholarship grants. The logic governing taxability in both areas is the same. The general rule is that gratis funds are tax free, but funds received as compensation for services rendered are taxable to the recipient. CASE B, DIM. D4 High (Inc.-Gift) (Read.) According to the Internal Revenue Code Sec. 61 " ... gross income means all income from whatever source derived", and should be included for tax purposes unless it is specifically excluded under the law. For students, two important areas of tax exempt income are Section 102 on gifts and Section 117 on scholarship grants. Section 102 amounts are generally tax free if it is the donar's intent to give a gift. Section 117 amounts are generally excludable from gross income although any portion received that represents payment for services is taxable unless the services are of the type required of all students as a condition of receiving their degree. A careful review of the Phase 2 scenarios in Appendix I reveals that the Case B quantitativeness dimension (D2) is the only one of the four cases where the low level contains no numerical expressions. A close examination of all the high and low levels of the readability dimension (D4) reveals that the high level of Case B is also unique in that it references three different Internal Revenue Code 95 sections. None of the other high readability levels reference more than one code section. These special features in dimensions 02 and 04 of Case B may have created a greater contrast between the high and low dimensions in this case. A consequence of this increased contrast appears to be a stronger effect on reporting. Inspection of Table 14 shows that the usual pattern of measure Ml showing the strongest effect, followed by measure M2 and M3, is apparent for both dimensions 02 and 04 of Case B. In Case C, dimension 01 (topic), there is also a significant difference between the high and low complexity effect. In addition there seems to be a confounding betweer the subjects• perceptions of the topic and the applicable tax law. In fact, this confounding seems to have enough influence to account for the statistically significant effect noted. As shown in the following passage from Appendix 0, the low and high levels of dimension 01 are distinguished by a personal versus financial topic orientation. 96 CASE C, DIM. Dl Low (Travel) (Topic) Rogers lives in Minnesota but owns a home in Newport Beach, California which he rents out to life-long close friends of his family at fair market value so that they can enjoy a temperate climate during the winter. Around Christmas, Rogers arranges a trip to California to inspect and repair the property and also to visit his friends. CASE C, DIM. Dl High (Travel) (Topic) Rogers lives in Minnesota but owns a home in Newport Beach, California, which he rents out. Rogers bought the Newport Beach home on speculation that housing prices on the California coast would appreciatE significantly in the next few years. During the winter, Rogers arranges a trip to California to inspect and repair the property and also to visit friends and go sailing. A confounding results because the correct tax position (embedded in the readability dimension) depends on whether the majority of days on the trip are spent on personal or financial matters. Table 14 supports this interpretation, since, contrary to most of the results, measure M3 (the correct tax position) shows the greatest level of significance. In fact, measure Ml, which usually reflects the most pronounced difference between the high and low levels, indicates no significant difference in this instance. Both the high and low levels of Case D's dimension D4 (readability) appear the "most complex" of the readability dimensions in the four cases. A careful review of Appendix 97 D shows that the readability dimension in Case D is clearly the most conceptually difficult and the longest. Both the high and low level of this dimension deal with an area of the tax law that has recently changed and with several numerical expressions. These unique features of dimension 04 in Case D may have affected the level of influence of this dimension. Direction of the Complexity Influence The research questions in this study focus on whether tax complexity exerts an influence on reporting postion selections. An interesting and potentially valuable extension of this study, since complexity is shown to have a significant effect, is to explore the direction of the complexity influence. As Figure 3 in Chapter 2 indicates, there is basis for arguing that complexity mitigates against noncompliance; and there is also basis for arguing the opposite--that complexity creates a propensity toward noncompliance. Since the results of this study empirically establish that complexity does exert an influence, a logical extension is to try to ascertain the nature of this influence. Limiting the scope of this study to the issue of magnitude, the author did not collect data to help interpret the direction of the influence. However, some reasonable conjectures may be made. One can examine the direction of 98 the complexity influence for each reporting position in all four cases by reviewing Appendix H, which displays the mean response for the low and high level of each dimension. In order to structure the discussion of the complexity influence, the following paragraphs discuss the direction of the influence for all significant one dimensional effects listed in Table 13. A review of the mean responses in Appendix H demonstrates the mixed direction of the complexity influence. Although the effect of complexity appears to be weighted slightly towards tax aggression, the results are far from conclusive. (In at least one instance in each of the cases, and in each of the dimensions, low complexity is associated with greater tax aggression.) The four significant one-dimensional effects in Table 13 are indicative of the mixed directions generally observed in Appendix H. In Case C, dimension 01, and Case B, dimension D2, the mean scores on the low complexity level show a more aggressive response than the high complexity level. The opposite result is true in the two remaining one-dimensiona cases recorded in Table 13. In both dimension D4 situationE (Cases B and D) , the high complexity level responses are significantly more aggressive. Each of these four situations are now reviewed in detail. In Case C (travel), dimension D1 (topic), there may be 99 an overlap with the concept of tax equity or falrness. 1n constructing dimension Dl, the personal versus financial topic concept was operationalized as a home owner renting to friends versus a real estate speculator: CASE C, DIM Dl Low and High (Travel) (Topic) Rogers lives in Minnesota but owns a home in Newport Beach, California which he rents out to life-long close friends of his family at fair market value so that they can enjoy a temperate climate durinq the winter. Around Christmas, Rogers arranges a trip to California to inspect and repair the property and also to visit his friends. Rogers lives in Minnesota but owns a home in Newport Beach, California which he rents out. Rogers bought the Newport Beach home on speculation that housing prices on the California coast will appreciate significantly in the next few years. During the winter, Rogers arranges a trip to California to inspect and repair the property and also to visit friends and go sailing. The notion of an overlap between dimension Dl and the concepts of tax equity or fairness is supported by the fact that the dimension D3 (social justice) interaction with dimension Dl is significant. This interpretation is also supported by the fact that the complexity direction for dimension D3 is in the same as dimension Dl. For both dimensions, the mean response to the low level of complexit reflects a more aggressive tax reporting position than the high level of complexity. In dimension D3, the motive for the purchase in the low complexity level description is to 100 provide for a retirement house, while in the high level description there is an apparent tax shelter motive. In conclusion, the top of Figure 3 (Chapter 2) appears operative in both dimensions 01 and 03 of Case c. It seems that complexity is linked to some notion of equity, which in turn is inherent in the subjects' perceptions of the fairness in the tax system. When this connection occurs, it seems that higher complexity provokes less aggression because it is associated with a more inequitable and unfair position. In Case B (Income-Gift), dimension 02 (quantitative- ness), the lower complexity level is again associated with significantly more aggressive mean reporting position selections. The low and high levels of this dimension are reviewed below. CASE B DIM 02 Low and High (Inc.-Gift) (Quant.) Warren figures that he does half the amount of work that he would have to do anywhere else for the same amount of money. Warren figures he only does about $300 worth of services a month ($120 worth of errands, $85 cleaning, and $95 miscellaneous). Notice that the low complexity level description is far more ambiguous than the high complexity level description. In operationalizing the concept of quantitativeness, the numerical detail of the high level conveys more specificity 101 ~onsequently, the lower part of Figure 3 appears operative. Assuming that the ambiguity of the lower complexity level relates to a lower probability of detection, the result is an increased opportunity for tax evasion. Hence, the lower complexity level in this situation appears consistent with greater tax aggressiveness. In the other two significant one-dimensional cases, the high complexity level responses are significantly more aggressive. Both Case B (Income-Gift) and Case D (Rehabilitation Credit) show significant readability dimensions (D4). As discussed in the previous section, in both of these cases the high level appears substantially more complex than the low readability dimension. In Case B, the high level references three different code sections; anc in Case D, the high level is long and conceptually difficult. These situations appear to support the concern expressed by the Committee of the New York Bar Association (described in Chapter 1) that some taxpayers will use complexity to take advantage of the tax system. The inference is that the lower portion of Figure 3 is applicable. If tax law complexity does facilitate "playing the tax lottery," then it increases opportunities for tax evasion. The results from examining the direction of the complexity influence indicate that tax complexity is an 102 intricate, multi-faceted concept. Specifically, complexity seems closely related to two other tax compliance factors: fairness and opportunity for evasion. It appears that any definitive study of the direction of the complexity influence will benefit if these related concepts are explicitly incorporated into the research design. The Task Effects Research question five also asks whether the task affects the influence of complexity (or individual complexity dimensions). A tentatively negative answer appears warranted from the available evidence. As the discussion in research question four emphasized, there is definitely no task effect on the overall complexity measure. Complexity has a statistically significant influence in each of the four cases and is insensitive to the topic. But there is reason for more caution in concluding that no task effects influence the individual dimensions and the interaction of dimensions. At this point it seems reasonable to conclude that the items noted in Table 13 result from the construction and interaction of the complexity dimension descriptions, rather than from a task effect. This conclusion is reinforced by the fact that the significant effects are scattered among the dimensions and cases. Since the evidence on the influence of the 103 individual complexity dimensions is far from conclusive, however, further study involving a number of different tasks appears a fruitful area for future research. The Subjects' Aggression Propensities Research question six is of vital importance to tax authorities. Do taxpayers show a propensity toward aggressive tax behavior as measured either by their perception of the correct tax position or by the prevailing legal standard? This question is especially interesting since the author conducted the study in a structured, authoritarian environment (prospective jurors in a courthouse waiting room). In addition, the author provided enough information for the subjects to compute the legally correct amount (IRS measure). The information necessary to compute the correct tax position was woven into each scenario, with the legal rule embedded in dimension D4 in every case. To illustrate the guidance the subjects received to ascertain the correct tax position, the author has included the critical dimension descriptions. As a review of Appendix D confirms, the low and high dimension descriptions contain the same key facts and pertinent rules of law, differing only in the level of complexity of the presentation. (For example, in the readability dimension, dimension D4, the high complexity 104 ~ersion quotes or refers to the Internal Revenue Code, Whlle ~he low complexity version gives the same rule of law in lay terms.) Consequently, an objective reading of either the low or high version is designed to give a reader enough information to determine the correct reporting position. To illustrate the guidance subjects received for ascertaining the correct tax position in Case A (the charitable deduction scenario), the author has provided the low complexity level version of the scenario. (Note, in the test instrument the dimensions were not separately identified; rather, specific combinations--see Figure 5--of 01-04 were run together and presented in one paragraph.) CASE A CHARITABLE DEDUCTION 01 Low (Topic) Wilson, a mechanic, donated a van that he had thoroughly renovated to his church. This gift entitlec Wilson and his family to become life members of the church and thus eligible to attend all concerts, pageants, retreats, and camps free of charge. 02 low (Quant.) Appraisals on the donation averaged about $2,000. The Wilson family received a life membership because any gift with a value over $1,000 confers ttis special status ($1,000 is the present value of receiving free family admission to all special events). 105 D3 Low (Soc. Just.) Because of the clearly religious nature of its activities, Wilson's church is classified as a tax exempt organization and therefore contributions are tax deductible. D4 Low (Read.) After a review of his own tax situation, the nature of the gift, and the use of the gift by the church, Wilson determined that none of the possible charitable deduction limits applied to his gift. However, Wilson did find out that his deduction should be reduced by the value of benefits received as a result of the donation. Within the charitable contribution scenario description, the subjects received information regarding the value of the donation, the value of the benefits received, and the rule that the amount deducted should be the difference between the two. Thus, they had enough data to compute the correct tax position in this case ($2,000 donation minus $1,000 benefit received equals a permissible charitable contribution deduction of $1,000). Hence, for Case A, the proper deduction measure according to the Internal Revenue Service (abbreviated as M-IRS) would be $1,000. The low complexity version of Case B, the income-gift scenario, follows: 10E CASE B INCOME OR GIFT? 01 Low (Topic) Warren is a college student who spends several hours a week helping out an elderly neighbor by running errands, cleaning, and taking the gentleman to the doctor. Each month Warren receives a check from the man for $600. 02 Low (Quant.) Warren figures that he does half the amount of work that he would have to do anywhere else for the same amount of money. 03 Low (Soc.Just.) Warren really enjoys being able to help out and would continue the same activities regardless of whether he received a monthly check. 04 Low (Read.) According to the tax law, some amounts received are taxable while others may be tax exempt. For students, two important areas of tax exempt income are gifts and scholarship grants. The logic governing taxability in both areas is the same. The general rulE is that gratis funds are tax free, but funds received as compensation for services rendered are taxable to the recipient. Within the income-gift scenario, the subjects had information regarding the amount of money received and the rule that compensation for services is taxable. Consequently, the proper inclusion of M-IRS for Case B is $600. 107 The low complexity version of Case C, the travel expense deduction scenario, follows: CASE C TRAVEL EXPENSES D1 Low (Topic) Rogers lives in Minnesota but owns a home in Newport Beach, California, which he rents out to life-long close friends of his family at fair market value so that they can enjoy a temperate climate during the winter. Around Christmas, Rogers arranges a trip to California to inspect and repair the property and also to visit his friends. D2 Low (Quant.) Rogers trip to California cost a total of $1,400 for his Thursday through Monday stay. This total includes charges for air fare, car rental, meals and lodging for five days. Rogers figures that the primary purpose of his trip was investment related and that thE majority of his days were spent on business rather thar. personal matters. D3 Low (Soc.Just.) Rogers initially purchased the house in anticipation of providing a home for his own retirement in the next 5 to 10 years. Thus he is very interested in maintaining the appearance of the property and contributing to the betterment of the neighborhood and community. 10E D4 Low (Read.) The IRS allows a deduction for all travel costs as long as the primary purpose of the trip is investment related and the majority of the days during the trip are spent on investment-related, rather than personal matters. If the primary purpose is personal or if the majority of days are spent on personal matters, then the expenses are not deductible. Embedded in the travel expense scenario are the pertinent facts that the primary purpose of the trip is investment related and that the majority of the days are spent on business rather than personal matters. Also, contained in the last section of the description is the rule that, under the described circumstances, all of the travel costs are deductible. Hence, all $1400 of travel expenses may be deducted (that is, M-IRS equals $1400). The low complexity version of the last scenario, which involves the rehabilitation credit for older structures, reads as follows: CASE D REHABILITATION CREDIT D1 Low (Topic) Carson retired after years of working and decided to open an antique shop in town with his wife. After a thorough search, Carson discovered an old building in need of some internal repair, which he purchased in August, 1981. 109 D2 Low (Quant.) During September, Carson solicited contractors• bids. Even though the structure was estimated to be about 38 years old, none of the exterior walls needed to be replaced, so Carson found that the building could be restored for a total of $10,000. Construction began in early October, 1981, and was completed within six months. D3 Low (Soc.Just.) To encourage the preservation of American heritage, Congress has written certain tax incentives into the law. Actually, Carson would have purchased and restored the building regardless of the tax credits because of his appreciation and respect for older architecture and craftsmanship. D4 Low (Read.) Congress changed the tax law for years beginning in 1982 to allow a larger credit for rehabilitating older structures. For buildings at least 30 years old, a 15 percent credit is allowed, for 40 year old buildings, a 20 percent credit (pre 1982, the credit was limited to 10 percent in both cases). Carson is a cash basis taxpayer. He wrote all $10,000 worth of checks to contractors in 1982, and is not sure of the exact age of the building. Consequently, he wonders what amount of credit he should take for rehabilitatin~ the structure. 110 TABLE 15 SUMMARY TAX POSITION STATISTICS VARIATE MEAN STD ER STD DEV Z-TEST p Case A M1 You 1480 65 601 1vs3 2.0 .05 (Charity) M2 Adv. 1435 65 601 2vs1 .7 M3 Cor. 1347 69 641 3vsiRS 5.0 .01 M-IRS 1000 Case B M1 You 380 27 254 1vs3 2.8 .01 (Inc/Gift) M2 Adv. 411 25 235 2vs1 1.2 M3 Cor. 450 24 220 3vsiRS 6.3 .01 M-IRS 600 Case C M1 You 1213 41 380 1vs3 2.5 .05 (Travel) M2 Adv. 1186 38 357 2vs1 . 7 M3 Cor. 1102 46 427 3vsiRS 6.5 .01 M-IRS 1400 Case D M1 You 2003 103 962 1vs3 . 5 (Rehab) M2 Adv. 1968 92 862 2vs1 .4 M3 Cor. 1949 102 949 3vsiRS 4.4 .01 M-IRS 1500 It is striking that in all four cases the mean for "what you would report" (M1) reflects a more aggressive tax stance than "what you would advise" (M2), which is still more aggressive than "what you think is correct'' (M3). And in three of the four cases (all except Case C), "what you think is correct" (M3) is definitely more aggressive than the "IRS measure." In all four cases, the mean score for the amount "you would report" is more favorable for the taxpayer than the mean response for the amount "you would advise." In essence, the subjects were more conservative when advising. 111 Apparently, taxpayers are willing to bear additional risk ~hen they are the beneficiaries. So even though there is approximately a .90 correlation in the subjects' responses between the two positions, the direction of the difference raises an interesting question about whether tax advisors also tend to be more conservative in an advisory role. An interesting feature of Table 15 is that even when subjects have the correct tax rule, their perceptions of the correct tax law differ significantly. (The differences in all four cases are significant at the .01 level.) In three of the cases (A,B, and D), the subjects' perceptions of the correct tax law are closer to the correct tax law than either of their other responses. But in Case C, the opposite is true. Case C involves the travel expense deduction, and the rule is that all travel expenses are deductible if the majority of time is spent on business. Ir Case C, however, the subjects perceive the tax law as allowing almost $300 less in expenses than it actually allows. It may be that the subjects' intuitive understanding of the tax law, or their opinions of what the tax law ought to be, influenced their perception of the current rules. The divergence noted in the four cases regarding the correct tax law, regardless of the contrary direction in Case C, may indicate a serious deficiency in the general public's willingness or ability to comprehend 112 the existing tax law. The most potentially serious finding, from a public policy perspective, is that there appears to be a significant propensity towards aggressive tax behavior as measured by the difference between what subjects say they would report on their own tax returns (M1) and what they believe to be the correct tax position (M3). This tendency pervaded all the cases. Of the 86 subjects, 21 in Case A, 16 in Case B, 24 in Case C, and 26 in Case D indicated that they would skew the amount reported in their favor. And what is especially remarkable is that when all four cases are considered, 56 of the 86 subjects selected reporting positions that deviated in their favor from their own perception of the correct reporting position in one or more of the four cases. This finding implies that more than 67 percent of the subjects were willing to evade taxes under certain circumstances. Background Questions The last question in Phase 2 involves the background data collected from the subjects. This question attempted to ascertain whether the reporting position selection can bE related to any identified biographical or attitudinal variables. Like the results in Phase 1, the results in Phase 2 were negative. No significant correlations emerged 113 etween any of the background or attitudinal variables (see able 3 for the list of questions) and the subjects' tax reporting choices. (The greatest correlation was .27 and that was between the subjects' rating of the complexity of the income tax system and measure M1--the amount they would report on their tax return.) Even though the statistics collected produce no clues for grouping subjects or predicting reporting positions, they do present an interesting profile. The subjects were about evenly divided between male and female (44 male, 42 female), and 38 percent prepared their own tax returns. The remaining profile statistics appear in Table 16. TABLE 16 SUBJECT'S BACKGROUND AND ATTITUDINAL RATINGS VARIABLE DESCRIPTION MEAN STD. DEV. Years Worked 20 yrs. 13 Education 15 yrs. 2 Income $37,500 $22,000 Fairness of Tax Laws 25% 22% % Person Could Evade 43% 29% Tax Complexity Rating 75% 24% % Pay All Taxes 45% 32% % Taxes Used Wisely 36% 24% The background data shows the subjects to be a relatively mature, educated, and affluent group. As Table 16 shows, the respondents average 20 years of work experience, three years of college, and an annual family income of $37,500. Some authors, like Spicer and Becker 114 (1980} and Hotaling and Arnold (1981}, found background variables irrelevant when studying taxpayer compliance. But other studies reviewed in Chapter 2 found that older, more affluent people are the most likely to comply with the tax laws (for example, see Spicer, 1974; Song and Yarbrough, 1978; and the 1980 study of Dean, Keenan, and Keeney.} Thus, it appears that if any group might be expected to represent compliant taxpayers, the profile noted in Table 16 must be considered among the least likely to resist paying taxes. Considering this profile, the mean responses on the attitude questions are particularly interesting. The subjects were asked to mark a five-inch line, undifferentiated except for labels on each end (see Table 3 for a sample of the form}. The first attitude question was, "How do you rate the general fairness of the income tax laws?" The five inch line was marked "Very Unfair" on the left and "Very Fair" on the right. The mean response of thE 86 subjects was 1.25 inches, indicating a fairness rating or the tax laws of only 25 percent. The next question asked, "If a person wanted to evade income taxes, what percentage do you think they could evade and not get caught?" Surprisingly, considering income tax withholding payments or virtually all wages, the respondents' perceived that an average of 43 percent of all income could be evaded. 115 Perhaps not surprisingly, the subjects' mean score to the question "In terms of complexity, how do you rate the present income tax system?'' was toward the "Extremely Complex" anchor on the five-inch line. The response average was 3.75 inches, for a 75 percent complexity rating. The fourth attitudinal question attempted to ascertain the subjects' perceptions of their associates' activities. The question asked, "Of the people you know, what percentage do you think pay all their legally owed income taxes?" Several studies reviewed in Chapter 2, including Strumpel (1966), Schwartz and Orleans (1967), Spicer (1974), and Vogel (1974), found peer influence to be a strong factor affecting taxpayer behavior. The mean score on this question for the 86 subjects indicated that only 45 percent of the people they knew paid all their legally owed taxes. This perception of their peers' activities may help explain the propensity to choose aggressive tax positions on the four tax cases. The last question attempted to measure the respondents' perceptions of the percentage of taxpayer money used wisely by the federal government. Dean, Keenan, and Kenney (1980) and Scott and Grasmick (1981) hypothesized a link between ar attitude towards the use of tax revenues and a willingness to pay. Even though the hypothesized correlation did not emerge here, the mean score recorded on this last question 116 ~emonstrates a markedly negative view of government spending. The average response was that the government used only 36 percent of its tax dollars wisely. Even though no significant correlations appeared between biographical-attitudinal variables and the subjects' tax reporting choices, the biographical-attitudinal ratings provide a striking profile of a cross section of a randomly selected group of Los Angeles area taxpayers. The picture that emerges is discouraging for tax authorities. Even citizens gathered to perform a civic duty, who might otherwise be viewed as mainstays of American society, have a pronounced, negative view of the tax system. Specifically, they view the system as unfair, complex, and easily evaded. Furthermore, they expressed little faith that the tax money collected from them will be spent wisely by the government. 117 Summary of Phase Two Research Results The second research phase was conducted to address four questions: 1. Does complexity significantly influence the tax reporting position? 2. Do the individual dimensions(s) or does the task exert a significant influence on the reporting position? 3. Do subjects show a propensity towards aggressive tax behavior as measured by their perception of the correct tax position or by the legal standard? 4. What is the relationship between the tax reporting position selected and identified demographic or attitudinal variables? The results indicate that complexity does significantly influence tax reporting positions. This is true for all four tasks across all three reporting positions. Apparently, none of the individual dimensions are strong enough to have a pervasive effect when analyzed separately. Furthermore, although the results are not conclusive, there appear to be no significant task effects. A peripheral issue that produced interesting findings involves the propensity of taxpayers toward tax aggression. In all of the cases, the subjects' regarded their own returns more aggressively than they would advise others to do. In addition, a dramatic difference emerged between what they would report and both their perception of the correct tax 118 position and the IRS's view of the correct tax posltlon. These results pervaded the entire population tested, and an analysis of the background and attitudinal information collected revealed no basis for predicting reporting position responses. In the last chapter, the author will detail further the conclusions, limitations, and contributions of this research project. In addition, she will suggest what areas of future research this study delineates. 119 CHAPTER 5 SUMMARY, CONCLUSIONS, AND EXTENSIONS Complexity has been linked to the quality of an income ax system (Dean, Keenan, and Kenney, 1980). Moreover, the uality of a tax system may relate to the ability of that system to generate revenues (Committee of the New York State IBar Association, 1972). Consequently, the finding in this !study that tax complexity exerts a significant and I !pronounced effect on the tax reporting positions selected by ! a cross section of taxpayers sheds light on an issue of crucial importance. i I As described in Chapter 1, complexity represents a I !component in a web of interrelated constructs, factors, and I !propositions that affect tax reporting in a democratic I society. Tax aggression and evasion are pressing national issues. An estimated $100 billion in lost annual revenue i attributed to these propensities. Yet, as the literature !review in Chapter 2 reveals, a satisfactory explanation of this phenomenon is nonexistent. Since an fundamental part of theory building is concept definition, the author devote the first phase of this study to defining "tax complexity"; and through the use of multidimensional scaling, she developed an empirical definition of "tax complexity'' in Chapter 3. A definition of complexity has two benefits. 120 First, it provided empirical information about the ~elationship between complexity and other concepts. Second, it provided a scientifically defensible base for testing the ~ffect of complexity in the second phase. Phase 1 was primarily a descriptive study. The central !research question was "What is tax complexity?" or more specifically, "What are the intentions or properties inherent in taxpayers' perceptions of tax complexity?" Focusing further on taxpayers, the study asked whether taxpayers can be grouped according to their complexity assessments. And, if it is possible to form groups, whether these groups can be related to any demographic data or attitudinal ratings gathered on the subjects. In order to execute Phase 1, the author selected individuals waiting to serve jury duty at the Los Angeles County courthouse randomly and asked them to participate in the study. Thirty of the 34 people selected agreed. These people were isolated in quiet conference rooms and took an average of 70 minutes to complete the task. The research task required subjects to judge the similarity of tax scenarios in terms of tax complexity. Thirteen scenarios were systematically varied and presented to the subjects in pairs. Thus, each subject was required to make 78 judgment~ to complete this section of the task. In addition, the subjects were asked to assess each of the 13 scenarios in 121 ~erms of several adjective descriptors. This information ~as gathered to help interpret the results of the 78 tax ~omplexity judgments. The author also gathered background ~nd attitudinal information on the subjects in order to ascertain whether there was any significant relationship between such data and an individual's complexity assessments. Multidimensional scaling (MDS) was employed as the primary Phase 1 methodology. MDS appeared to be an especially appropriate methodology because the concept of tax complexity and its underlying dimensions were not well understood. An advantage of this methodology is that it produces a visual representation of the subjects' judgments that can help a researcher uncover and interpret hidden structure in the data. Another advantage of MDS is that it tends to reduce researcher bias because, rather than requiring a priori knowledge of the attributes of the stimuli to be scaled, it provides a space whereby dimensionE relevant to the subjects are scaled. The author chose the MULTISCALE MDS program developed by Ramsey (1978) because it provides statistics which could be used to select the most appropriate model and number of dimensions. In the first stage of a two-stage analysis, the author performed an exploratory MDS, evaluating various MDS models in terms of fit to the data, and the tentative interpretability 122 of the configurations. In the second stage, a confirmatory MDS, the author subjected the model selected in the exploratory phase to statistical tests of significance to ascertain the most meaningful number of dimensions to extract. The results of the first phase of the study identified four distinct complexity dimensions. The first dimension appears to reflect a personal versus financial topic orientation. Topics dealing with personal matters (like exemptions for age and blindness, child care, health insurance, entertainment expenses, charitable contributions, and gifts) appear on the left side of the configuration, while topics dealing with financial matters (like interest, individual retirement accounts, trusts, ACRS, income averaging, nonrecognition of gains and losses, and capital gains) appear to the right side. The adjective descriptor statistics supported this conclusion. The property that correlated most highly with dimension Dl was the "nontechnical-technical'' rating with a correlation of 68 percent. This correlation was compatible with the interpretation that the subjects differentiated stimuli according to the topic, since they indicated the stimuli to be increasingly technical as the location of a stimulus proceeded from left to right (personal to financial). The configuration plot for dimension D2 showed a 123 ~istinct dispersion of stimuli with eight near the top, ~hree spread in the middle, and two near the bottom of the figure. A careful reading of the 13 stimuli revealed that none of the stimuli above the horizontal axis contained numerical manipulations while all the stimuli below the axis contained some calculations. Also, only one of the eight stimuli above the axis referred to a change in the tax law, while all except one of the five stimuli below the axis referred to such a change. Furthermore, the stimuli at the extremes showed, respectively, the least and greatest amount of change and computation. These observations were supported by the property fitting results since the property ~escriptor "law likely to remain the same-law likely to change" was the most highly correlated with dimension D2, ~ith a 65 -percent correlation. For brevity, the author labeled dimension D2 the "quantitativeness dimension" since the focus of the subjects' discrimination was the degree of change and the amount of computation. Of all the dimensions, dimension D3 was associated most directly with the adjective descriptors. Dimension D3 was correlated with the property descriptor "area of very little abuse-area of widespread abuse" with an 81-percent correlation; "very easy to catch cheaters-impossible to catch cheaters" with 75 percent; "of no personal benefit to me-extremely beneficial to me"; with 68 percent and "big 124 ~reak for undeserving-very fair" with a 61-percent correlation. In the configuration plot for dimension 03, the stimuli represented by the entertainment cost scenario ~as isolated on the extreme left side of the plane while the IRA stimuli was farthest to the right. Correspondingly, entertainment cost rated the highest in abuse, highest in cheating, lowest in personal benefit, and second lowest in fairness. On the opposite side of the plane, IRA was rated one of the lowest in abuse, the lowest in cheating, second highest in personal benefit, and one of the highest in I !fairness. Proceeding from left to right on the !configuration plot, the pattern of decreasing abuse and !cheating ratings, and increasing personal benefit and l !fairness ratings was confirmed across the dimension D3 1stimuli. The composite of these four properties appears to I be the concept of social justice; here, stimuli perceived tc be subject to widespread abuse and difficult to enforce are distinguished from those thought to be personally beneficial and fair. Consequently, the author labeled dimension 03 thE "social justice dimension." Dimension 04 was interesting because 11 of the 13 stimuli clustered at the middle of the configuration, one stimuli was at the extreme top (the gift-award scenario), and one was at the extreme bottom (the interest deduction scenario). Focusing on the two outlying stimuli, one 125 hotices three differences. The gift-award scenario was the longest, quoted the Internal Revenue Code, and rated low in familiarity. The interest deduction scenario, at the opposite extreme, was a short scenario, did not quote or even refer to the code, and rated high in familiarity. In general, this pattern held for the stimuli in the middle as well. Those stimuli above the horizontal axis tended to be longer, to be associated more directly with the code, and to be with less familiar areas of the law than the stimuli below the axis. The property descriptor ratings also helped interpret dimension 04. The most highly correlated property was a negative 84 percent with the "familiarity" property descriptor. In addition, dimension 04 correlated positively with the "technical" property descriptor and the "changing area of the law" property descriptor. The stimuli above the horizontal axis tended to be perceived as unfamiliar, technical, and liable to change, while stimuli below the axis were relatively familiar, nontechnical, and unlikely tc change. It seems that discrimination based on length of the scenario and degree of association with the Internal Revenue Code is indicative of differences in the readability of the stimuli. Unfamiliar, changing, technical scenarios also seem to be the most difficult to read. Consequently, the author labeled dimension 04 the "readability dimension." 10 Two peripheral issues in the Phase 1 inquiry involve 126 the subjects' weighting of the complexity dimensions and the relationship of these weightings to the subjects' background information and attitudinal ratings. For each of the four dimensions, two or three subjects placed the majority of the emphasis on that dimension in making their tax complexity judgments. But a review of the subjects' weighting of the complexity dimensions showed no significant relationship between these weightings and the background information or attitudinal ratings. The results of Phase 1 indicated the existence of four distinct complexity dimensions: personal-financial topic orientation, quantitativeness, social justice, and readability. These four dimensions, each with a high and low complexity version, were embedded in the Phase 2 tax cases. The author constructed 32 distinct tax scenarios around four tax topics for Phase 2 testing. Each scenario was designed around the four complexity dimensions and written to give the subjects enough information to determinE the correct tax reporting position. A one-half replicated research design was chosen because it allowed for the testing of all main effects on each complexity level as wel as the interaction between complexity levels. A total of 86 subjects submitted to the Phase 2 test (a response rate of 75 percent) . These subjects were selected under the same conditions and tested by the same process as 127 ~he Phase 1 subjects. The research task required each subject to read four different tax scenarios and to choose the amount he or she would report, would advise another to report, and believed to be the correct tax position. The Phase 2 subjects were also asked the same background and attitude questions as the Phase 1 subjects since these questions were designed to investigate areas that the literature review indicated were relevant. The central question of the study was addressed in Phase 2: "Does complexity significantly influence the tax reporting position?" The results were resoundingly affirmative. Using MANOVA, the author found that all three tax reporting position selections, in all four cases, showed significance to the .0000 level in the F-tests. Given the significance of these results, the next research question is whether understanding of the complexity concept can be increased by analyzing the nature of the observed effect. Since there is a significant overall effect between thE high and low complexity dimensions, the next issue is whether any single dimension or combination of dimensions i~ strong enough to produce a statistically significant effect. In the MANOVA analysis of 40 one and two dimensional effects, 10 attained a .10 level of significance. Six of the 10 significant effects recorded involved two dimensiona combinations. Dimensions D1 and D3 (topic and social 128 justice), and also dimensions D2 and D4 (quantitativeness and readability) combined to produce a significant effect ir three of the four cases. The Dl-D3 and D2-D4 pairings were the only significant combinations recorded. Analysis of thE dimension descriptions revealed that the factors around which these pairs of dimenions were constructed tend to overlap. By overlapping, the dimensions appear to reinforcE one another. The four significant single dimension effects were scattered among three different dimensions and three different cases. Generally, it appears that the significance of these dimensions is attributable to their construction. The dimension descriptions in these instance~ seemed to reflect greater contrast between the high and low complexity levels, resulting in a more pronounced effect on reporting position selections. Although a tangential issue, it is especially interesting that the four significant single dimension effects are evenly divided on the direction of their influence. Dimension Dl (topic) in Case C, and dimension D. (quantitativeness) in Case B, provoked more conciliatory responses. Even though the empirical evidence is limited, analysis of the directional influences suggest that complexity has a conciliatory effect in cases where equity is a dominant issue. In these situations, it seems that higher complexity provokes less aggression because it is associated with a more inequitable and unfair position. This interpretation is consistent with the experimental wor~ of Spicer and Becker (1980) which indicates that subjects increase the amount evaded when they perceive fiscal inequity, and decrease the amount evaded when they perceive fiscal equity. The two other significant one dimensional influences involved dimension D4 (readability). In both Cases B and D, dimension D4 provoked more aggressive responses. Analysis of these two cases suggest that higher complexity may be linked to a lower probability of detection and hense an increased opportunity for evasion. These results, while not definitive, do lend credence to the concern that tax complexity may encourage aggression and thus facilitate "playing the tax lottery." The question of whether the topic significantly effects the influence of complexity was also addressed. Overall, complexity is apparently insensitive to the task, since it had a statistically significant influence in each of the four cases. On the one and two dimensional levels, the significant effects were scattered among the dimensions and cases. Consequently, it seems reasonable to conclude that the significant effects noted are a result of the construction of the dimension descriptions, rather than a task effect. 130 Phase 2 propounded two peripheral issues: First, "Do subjects show a propensity towards aggressive tax behavior ~s measured by their perception of the correct tax position or by the legal standard?" And second, "What is the relationship between the tax reporting position selected and identified demographic or attitudinal variables?" The results dramatically revealed the subjects' propensity toward tax aggression. In all four cases, the mean for "what you would report" reflected a more aggressive tax stance than "what you would advise," which was still more aggressive than "what you think is correct." And in three of the four cases, "what you think is correct" was definitely more aggressive than the IRS measure. It seems particularly remarkable that considering all four cases, 56 out of the 86 subjects selected reporting positions that deviated in their favor from their own perception of the correct reporting position in one or more of the four cases. This finding implies that more than 67 percent of the subjects appear willing to evade taxes under certain circumstances. Nevertheless, an analysis of the background and attitudinal information collected on each subject in Phase 2 revealed no basis for predicting which subjects might tend toward aggressive tax positions. As the literature reviewed in Chapter 2 indicates, there have been no previous behavioral studies involving ta) 131 complexity, notwithstanding political speeches laden with assertions on the effects of complexity. Consequently, this study expands our understanding of the behavioral ramifications of tax policy. As with most scientific investigations, however, some of the study's limitations may have impaired the results and thus need to be explicitly acknowledged. 132 he Scope and Limitations of this Study The potential scope and limitation issues that emerge ~rom this study involve both the subjects and the tax cases.! pne concern about the subjects is that the representative ~ess of the taxpayers may impose some limitations. Even if the taxpayer group is representative of Los Angeles area taxpayers, it is not clear that the results are generalizable to other geographic and cultural areas. Another difficulty is whether measuring the subjects' I jattitudes in hypothetical situations is fairly indicative of I !their retrospective or prospective behavior. But this is a i problem common to all behavioral research. Ideally, the topic of this study would be "the effect of complexity on taxpayer behavior," not "the effect of complexity on reporting position." But no one can control for complexity and observe taxpayer behavior directly. Consequently, the study proposes to measure reporting position, not behavior. Although taxpayer behavior is certainly of interest, any assumptions based on reporting position selections must be more reader inference than author implication. A third concern involving subjects relates to testing in a formal, authoritarian environment, and the concern here is whether an unwanted, unmeasured intervening variable may have been introduced. It may be reasonable to assume that subjects tend to act more lawfully in a situation where they have 133 been summoned by government officials to perform a civic duty. Unfortunately, it is not known whether, or to what extent, a courthouse environment induces a different response. Motivation is a related area of concern. Selected at random and asked to participate, the subjects received no external reward for their cooperation. Consequently, a question lingers regarding how motivated the subjects' were to read carefully and respond to the test instrument candidly. A review of the Phase 1 MDS data reveals that the last subject judged all of the 78 paired comparisons equally difficult. Equally dubious results appeared in a few of the Phase 2 responses. For instance, subject number seven's response to Case A indicates he would take a $1,800 charitable deduction although he believes the correct deduction is $3,000. Even though a few of the responses appear to lack credibility, the author refused to alter or discard this data for the following three reasons: First, the tests were administered in small groups and all of the subjects appeared to complete the instrument carefully. Second, the suspect data accounts for only a small minority of the data collected. Third, and most important, selectinc and editing the data might inadvertantly introduce researcher bias into the results. Thus, although the motivation of the subjects is a critical issue, they seemed 134 ~enerally interested in the task and conscientious about ~ompleting the test instrument. Therefore, the author feels ro uneasiness over the subjects• motivation. Another difficulty centers on the representativeness of the tax cases. The work of simplifying complex and diverse ~elationships in the MDS phase, proved to be particularly ~hallenging because there is no limited, natural group of stimuli available, as there is when one studies Big Eight ~PA firms (Shockley and Holt, forthcoming) or perceived differences in audit report opinions (Libby, 1979). Because one could describe hundreds of tax situations, the challenge ~as to design stimuli rich enough to reflect the important complexity dimensions, yet short enough to administer conveniently. In the second phase as well, countless tax scenarios could be designed. The challenge was to design scenarios that reflected the critical level of complexity while presenting sufficient realism to engage a reader and evoke a useful response. Thus, the efficacy of the study rests upon the adequacy of the stimuli in the first phase and the proper interpretation and incorporation of the resulting dimensions into tax scenarios in the second phase. Ideally, these limitations do not invalidate the results of the study; however, the viability and integrity of the findings can be accurately assessed only over time a~ the study is replicated and extended. 135 ~orne Possible Extensions The results of this study suggest several potentially fruitful areas of research. The ultimate objective of this study was to contribute to a theory of taxpayer behavior. As noted in Chapter 1, any comprehensive theory involves a whole set of interrelated constructs, definitions, and propositions. A logical extension of the present study, thus, would be to undertake parallel work on other concepts identified in Chapter 2, like opportunity for evasion, or fairness. A more direct extension of the present study would be to replicate the work on complexity with different tax scenarios in Phase 1 and different tax cases in Phase 2. The advantage of such an approach would be the opportunity it presents to examine the complexity dimensions more fully. Additional work is needed to ascertain the magnitude and direction of the influence of the individual complexity dimensions, as well as the circumstances under which they have an effect (that is, the task effects). Two peripheral issues in this study also merit further investigation. The first involves the advisory role. Half of the tax returns in the country are prepared by advisors, and it would be interesting and useful to know how this fac affects reporting. The other issue relates to taxpayer 136 ~ttitude. Ultimately, it seems that attitude plays a vital role in the dynamics between citizens and their government. Of special interest, then, are changes in this attitude that may portend shifts in behavior. 137 ENDNOTES 1. Adam Smith was the first political observer to publish criteria for evaluating a tax system. Subsequent authors have either added criteria or expanded upon Adam Smith's original ideas. A few later authors and their suggested additions follow: John Stuart Mill (1848), ability to pay and equality of sacrifice; Thomas Hobbes (1887), benefits received; E.R.Seligman (1894), horizontal equity, vertical equity; F.Y.Edgeworth (1925), minimum sacrifice; A.C.Pigou (1928), influence on savings and investment, and influence on work and leisure choice; John Maynard Keynes (1936), economic stability and the redistribution of wealth; Henry Simons (1938), the influence on economic choices of goods; John Kenneth Galbraith (1952), political stability and the funding of public goods; Joint Economic Committee (1959), promotion of economic growth and public interest; Louis Eisenstein (1961), minimization of class conflict and neutrality to all classes; Dan Troop Smith (1961), fairness and simplicity; Committee on Federal Tax Policy (1963), promote economic efficiency and administrative efficiency; Charles M. Allan (1971), taxpayer awareness of tax; and Joseph A. Pechman (1977), stabilization of consumption expenditures. Still, after over 200 years, Adam Smith's canons are still the most widely recognized standards for a good tax system (for example, see Sommerfeld, et al. 1983, Chapter.1, p. 16). 2. According to Frank Malanga, Director, Research Division of the Internal Revenue Service (speaking at the University of Michigan Tax Conference on November 18, 1983) the trend is toward a decreasing level of tax compliance. Malanga estimates that tax compliance has dropped at least five percent in the last decade. The current compliance rate is estimated to be about 80 percent. 3. The guidelines for selecting the number of stimuli and subjects are flexible. While too few stimuli may limit the number of dimensions that can be extracted, too many stimuli may actually reduce the power of the procedure by making the results less interpretable. Kruskal and Wish (1978) recommend between eight and eighteen stimuli. As noted in the literature review, Libby (1979) used ten stimuli and Shockley and Holt (forthcoming), eight. 4. The author used a five-inch undifferentiated line scale to collect data, an approach advocated by Schiffman, Reynolds and Young (1981). They maintain that "In our 138 experience, subjects seem to be comfortable with a 5 in. line--with a 4 in. line their judgments are compressed and with a 6 in. line they do not use the right-hand end of the scale often (p. 14) .•. We do believe it important, however, to provide an undifferentiated line scale rather than a series of boxes or numbers. This is for several reasons. First, people attach different meanings to verbal descriptions such as fairly and somewhat. Second, subjects tend to lose sight of the similarity task while debating the relative merits of fairly and somewhat. Third, many subjects feel uncomfortable with segmenting the line with words and especially with numbers" (p. 15). 5. The author eliminated following adjective descriptors as a result of the pilot testing: interesting-boring, clear-unclear, and easy to measure-difficult to measure. A consideration in designing the final test instrument was that the time for completion be kept as brief as possible in order to retain subject interest. This was a special concern since most of the 15 subjects in the pilot test took over an hour to complete the task. The pilot testing indicated that subjects spent an inordinate amount of time on the three adjective descriptors listed above. In addition statistical tests showed the first two descriptors listed to be highly correlated with other descriptors. For these reasons, and since the subjects appeared especially confused by the descriptor "easy to measure-difficult to measure," the three adjective descriptor pairs were eliminated. 6. Schiffman, Reynolds and Young (1981) maintain that MDS experiments "generate a large amount of information and generally yield stable spaces with only a few subjects" (p.4). Even though smaller samples may be defensible, in both the Libby (1979) and Shockley and Holt (forthcoming) studies, sample sizes of 30 were used. 7. When calculating rho correlations, ordinal or dichotomous data is assumed rather than interval data. 8. Karlinsky and Koch (1983) found that the Internal Revenue Code was perceived to be significantly more complex than a tax commentary on the same issue. 9. For a discussion of further readability issues, including the distinction between syntactical and conceptual complexity in the tax law, see Karlinsky and Koch (1983), p. 2. 139 lO.Both the low and high levels of the readability dimensior (D4) in Case D contain numerical expressions. Hense, it might be argued that a confounding results, negating the possibility of a low quantitativeness case (even when thE low level of dimension D2 is embedded). However, the relevant factor is not the absolute amount of quantitativeness in any one case, but rather the relativE difference in quantitativeness between the low and high level versions. Since the low and high readability levels of Case D contain approximately the same amount of quantitativeness, the absolute difference is preserved and confounding is not a problem. 140 REFERENCES Allan, Charles M. The Theory of Taxation. New York: Penguin Books, 1971. Allingham, Michael G., and Sandmo, Angar. "Income Tax Evasion: A Theoretical Analysis." Journal of Public Economics, 2 (1972), pp. 323-338. Bagozzi, Richard. "The Role of Measurement in Theory Construction and Hypothesis Testing: Toward a Holistic Model." Massachusetts Institute of Technology Working Paper, 1980. Belkaoui, A. "The Interprofessional Linguistic Communication of Accounting Concepts: An Experiment in Sociolinguistics." Journal of Accounting Research, Autumn, 1980, pp. 362-374. Carroll, J.D. and Arabie, P. "Multidimensional Scaling." Annual Review of Psychology, 1980, pp. 607-649. Committee on Federal Tax Policy. Financing America's Future: Taxes, Economic Stability and Growth. Government Printing Office, 1963, pp.1-18. Congressional Joint Committee on Taxation. Background on Federal Income Tax Compliance and Description of S. 2198. Hearing before the Subcommittee on Oversight of the Internal Revenue Service, March 22, 1982. Consumer Reports. pp. 130-137. "Where to Go for Tax Help." March, 1976, Dean, Peter~ Keenan, Tony and Kenney, Fiona. "Taxpayer's Attitudes to Income Tax Evasion: An Empirical Study." British Tax Review, 1980, pp. 28-44. Edgeworth, F.Y. Papers Relating to Political Economy. New York: MacMillan and Co., 1925. Eisenstein, Louis. The Ideologies of Taxation. New York: Ronald Press Co., 1961. Frank, Warner G. "An Empirical Analysis of International Accounting Principles." Journal of Accounting Research Autumn, 1979, pp. 593-605. Friedland, N.; Marital, S.; and Rutenberg, A. "A Simulatior Study of Tax Evasion." Journal of Public Economics, 10 (1978)' pp. 107-116. 1 41 Galbraith, John Kenneth American Capitalism: The Concept of Countervailing Power. Boston: Houghton Mifflin Co. , 19 52. Ginsburg, Martin D. "Tax Simplification - A Practitioner's View." National Tax Journal, 26 (1976), pp. 317-331. Government Accounting Office. Report to Congress by the Comptroller General of United States. Washington, D.C.: Government Printing Office, 1972. Hair, Joseph F; Anderson, Rolph E.; Tatham, Ronald L.; and Grablowsky, Bernie J. Multivariate Data Analysis. Tulsa: Petroleum Publishing Company, 1979. Hall, Robert E., and Rabushka, A. "Low Tax, Simple Tax, Flat Tax." Hoover Institute, Stanford University Working Paper, September 1, 1982. Hobbes, Thomas. Leviathan. 3rd ed. London: George Rutledge & Sons, 1887. Hotaling, Andrea w., and Arnold, Donald F. Economy." Massachusets CPA Review, 55 "The Undergroun (1981)' pp. 6-14 Internal Revenue Code of 1954. September 3, 1982 ed. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. Joint Economic Committee. "The Federal Revenue System: Facts and Problems 1959." Materials assembled by the committee staff for the Joint Economic Committee, 86th Congress, Senate Rep. No. 98, 1959. Karlinsky, Stewart S. Complexity in the Federal Income Tax Law: A Measurement Model. Unpublished Ph.D. Dissertation, New York University, 1981. Karlinsky, Stewart, and Koch, Bruce. "The Effect of Federa Income Tax Law Reading Complexity on Task Performance.' University of Southern California, Center for Accounting Research, Unpublished Working Paper, 1983 . . Kerlinger, Fred N. Foundations of Behavior Research. 2nd ed. New York: Holt, Rinehart and Winston Inc., 1973. Keynes, John Maynard. The General Theory of Employment, Interest and Money. London: Harcourt, Brace & Co., 1936. 142 Kruskal, Joseph B., and Wish, Myron. Multidimensional Scaling. Beverly Hills: Sage, 1978. Libby, Robert. "Bankers' and Auditors' Perceptions of the Message Communicated by the Audit Report." Journal of Accounting Research, Spring, 1979, pp.99-122. McDaniel, Paul. "Simplification Symposium Federal Income Tax Simplification: The Political Process." Tax Law Review, 34 (Fall, 1978), pp. 211-236. Mill, John Stuart. Principles of Political Economy. London: Longmans, Green & Co., 1923. Morrison, D.F. Multivariate Statistical Methods. McGraw-Hill (1976). New York Bar Association. "A Report on Complexity and the Income Tax." Tax Law Review, 27 (1972), pp. 329-376. Pechman, Joseph A. Federal Tax Policy. 3rd ed. Washington, D.C.: Brookings Institution, 1977. Pechman Joseph A., and Okner, Benjamin. Who Bears the Tax Burden? Washington, D.C.: Brookings Institution, 1974. Pigou, A.C. Study in Public Finance. New York: MacMillan and Co., 1928. Pratt, Jamie. "Post-Cognitive Structure: Its Determinants and Relationship to Perceived Information Use and Predictive Accuracy." Journal of Accounting Research, Spring, 1982, pp. 192-209. Ramsey, J.D. Multiscale. Chicago: National Educational Resources, 1978. Rose, Richard. "The Do-It-Yourself Tax Revolt." Across the Board, 18 (1981), pp.21-26 Schiffman, Susan S., Reynolds, M.L.; and Young, Forest W. Handbook for Multidimensional Scaling. New York: Academic Press, 1981. 143 Schroeder, Jack. Potential Simplification of the Federal Income Tax Law by Eliminating Special Treatment of Capital Gains and Losses. Unpublished Ph.D. Dissertation, Michigan State University, 1975. Schwartz, R., and Orleans, Sonya. "On Legal Sanctions." The University of Chicago Law Review, 34, (1967), pp. 274-299. Scott, Wilber J., and Grasmick, Harold G. "Deterrence and Income Tax Cheating: Testing Interaction Hypotheses in Utilitarian Theories." Journal of Applied Behavioral Science, 17, (1981), pp. 395-408. Seligman, E.R. Progressive Taxation in Theory and Practice Publication of the American Economic Association, Monographs, volume 9, numbers 1 and 2, 1894, pp.192-194. Senate Finance Committee Report. Tax Underpayment and Frauc I Penalty Revision. S 361:19, April, 1982. Shockley, Randolph A., and Holt, Robert. "A Behavioral Investigation of Supplier Differentiation in the Market for Audit Services." Journal of Accounting Research, forthcoming. Simons, Henry. Personal Income Taxation. Chicago: University of Chicago Press, 1938. Smith, Adam. The Wealth of Nations. New York: Dutton, 1910. ! Smith, Dan Throop. Federal Tax Reform, the Issues and a Program. New York: McGraw-Hill, 1961. Sommerfeld, Ray M.; Anderson, Hershal M; and Brock, Horace R. An Introduction to Taxation. 1983 ed. New York: Harcourt Brace Jovanovich, Inc., 1982. Song, Young-dahl, and Yarbrough, Tinsley E. "Tax Ethics anc Taxpayer Attitudes: A Survey." Public Administration Review, 38, (1978), pp. 442-457. 144 Spicer, Michael W. A Behavioral Model of Income Tax Evasion. Unpublished Ph.D. Dissertation, Ohio State, 1974. Spicer, Michael w., and Becker, Lee A. "Fiscal Inequity an Tax Evasion: An Experimental Approach." National Tax Journal, 33 (1980), pp. 171-175. Strumpel, B. "Contribution of Survey Research." Quantitative Analysis in Public Finance. Edited by A. Peacock. New York: Praeger, 1969. Tax Equity and Fiscal Responsibility Act of 1982. P.L. 97-2 8. September 3, 1982. Vogel, J. "Taxation and Public Opinion in Sweden: An Interpretation of Recent Survey Data." National Tax Journal, 27, (1974), pp. 499-513. Wall Street Journal. "Interview with Former IRS Commissioner Jerome Kurtz." September 8, 1982, p.31. Watkins, Paul R. "Multidimensional Scaling Measurement and Accounting Research." Journal of Accounting Research, forthcoming (Spring, 1984). Webster's. New Collegiate Dictionary. New York: Books Inc. , 19 7 3. 14 APPENDIX A LIST OF THE 13 PHASE 1 STIMULI 1. CHILD-CARE EXPENSES 2. CONTRIBUTIONS 3 . CAPITAL GAINS 4. ENTERTAINMENT COSTS 5. ACRS 6 . GIFTS/AWARD 7. ENTRA EXEMPTIONS 8. HEALTH INSURANCE 9. TRUSTS 10. INCOME AVERAGING 11. INTEREST DEDUCTION 12. NONRECOGNITION OF GAINS AND LOSSES 13. I.R.A. 146 APPENDIX A PHASE ONE STIMULI 1 - CHILD-CARE EXPENSES Is child care a personal expense or a business expense? If it is a personal expense then childless taxpayers shouldn't have to subsidize those with children. If it is a business expense, then it should be fully deductible. Congress has decided something in the middle. Prior to 1976, taxpayers were able to deduct limited child care expenses as "itemized deductions" if the costs were incurred to enable the taxpayer to be gainfully employed. The 1976 Tax Reform Act eliminated the deduction and substituted instead a credit for child-care expenses (Sec. 44A). The basic credit, subject to several limits, ranges from 20 percent to 30 percent of the amount expended for child care. 2 - CONTRIBUTIONS Before 1982 if you didn't itemize deductions you didn't receive any tax breaks for charitable contributions, so Congress changed the tax law for the years 1982 through 1986. In 1982 and 1983, 25% of charitable contributions up to $100 may be deducted from gross income (a $25 deduction). For 1984, the amount is 25% of $300 (a $75 deduction). In 1985, 50% of all contributions may be deducted (with no dollar limit). In 1986, the deduction is 100% of all contributions (again with no dollar limit). In 1987, however, this provision expires. Thereafter charitable contributions will once again be allowed only if deductions are itemized. 3 - CAPITAL GAINS The name of the tax game in the United States is capital gains. Section 1202 authorizes individuals to clairr a special deduction equal to 60 percent of the net capital gain realized in a year. This special deduction is the equivalent of a 60 percent tax deduction. If in 1983 an individual had only net capital gain income from stocks and bonds, then the real effective tax rate for the taxpayer would range from 4.4 to 20 percent rather than the normal rate range of 11 to 50 percent which applies to salaries, wages, interest, and dividend income. 4 - ENTERTAINMENT COSTS In 1978 the three martini lunch proved more powerful than the Carter Administration. Tax deductions for liquor, country clubs, and athletic events were continued despite proposed limits. Currently, to be tax deductible, entertainment costs must be "ordinary", "necessary", and 147 "reasonable". In addition, Internal Revenue Code Sec. 274 details certain record keeping requirements. 5 - ACRS In 1981, to stimulate the economy, Congress replaced ~he property depreciation laws with the Accelerated Cost ~ecovery System (ACRS). This system dramatically shortens ~stimated lives for most newly purchased properties and thus ~ramatically increases write-offs for most investors. Autos can now be written off in 3 years and apartment buildings and shopping centers in 15 years. 6 - GIFTS/AWARD You are not taxed on a $5,000 gift from your rich uncle or on a $5,000 Pulitzer Prize. However, you are taxed on a $5,000 lottery ticket winning, or church raffle prize. Basically the law says that a person who receives a gift is not taxed, but a person who receives a prize or award is taxed unless they meet the following Sec. 74(b) exception: "Exception. - Gross income does not include amounts received as prizes and awards made primarily in recognition of religious, charitable, scientific, educational, artistic, literary, or civic achievement, but only if- (1) the recipient was selected without any action on his part to enter the contest or proceeding, and (2) the recipient is not required to render substantia future services as a condition to receiving the prize or award." 7 - EXTRA EXEMPTIONS Should there be extra exemptions allowed for old age and blindness? These special exemptions have been in the tax law since the 1940s. Today these exemptions can total as much as $6,000 for a taxpayer and spouse if both are blind and both are over 64. 8 - HEALTH INSURANCE Health insurance paid by an employer for an employee i~ not taxed to the employee. This can be very beneficial for the employee who wants insurance. For if an employee got ar increased salary instead of insurance, he would have to pay income tax on the extra income, and then buy health insurance with what was left. 9 - TRUSTS How do trusts help the wealthy? Mainly, they allow thE shifting of income from high income to low income bracket taxpayers. Although wage income can't be diverted to a trust, income producing property like stocks and bonds can be put in a trust so its earnings (dividends/interest) isn' 148 taxed to the high income bracket taxpayer. 10 - INCOME AVERAGING What happens when a $30,000 a year baseball player gets a million dollar contract? Income averaging provisions were enacted in 1964 and liberalized in 1969 to help reduce the taxes of people with fluctuating incomes. The Code specifies the conditions and calculations for income averaging in Sees. 1301-1305. First the taxpayer's average taxable income for the prior four years is calculated. This average is then multiplied by 120 percent. The result is subtracted from the current years taxable income. If the difference is $3,000 or more, then income averaging may be used. 11 - INTEREST DEDUCTION Interest expense is the most important itemized deduction for individuals. In 1979 this deduction totaled $73.6 billion, or 40.7 percent of all itemized deductions. Virtually all interest paid is deductible. The major exceptions are interest on money borrowed to purchase tax free securities and a limit on the amount of interest deductible on funds borrowed to purchase investment property. 12 - NONRECOGNITION OF GAINS AND LOSSES Sec. 1031 is a very important tax law for real estate investors. This law allows some direct exchanges of business and investment property without taxation. Sec. 1031(a) states in part "No gain or loss shall be recognized if property ... is exchanged solely for property of a like kind ... ". The Treasury Regulations are rather liberal in interpreting what is ''like kind"; a country farm can be exchanged for a city apartment without taxation. 13 - I.R.A. As all the ads might lead you to suspect, Code Sec. 219, permitting individuals to establish their own individual retirement plan, is a far reaching change. Now all workers can deduct up to the lesser of $2000 or 100 percent of earned income for an IRA. Previously, only those without a qualified pension plan could take out an IRA, and then they were limited to $1,500 or 15 percent of earned income. 149 APPENDIX B SAMPLE PHASE 1 TEST INSTRUMENT There are 78 paired comparisons, sequenced in the following order: 1 2 3 4 5 6 7 8 9 10 11 12 13 c-c Cnt C-G E-C ACRS G/A E-E H-I Trst I-A I-D N-R IRA 1 c-c 2 Cnt 1 3 C-G 2 3 4 E-C 4 5 6 5 ACR 7 8 9 10 6 G/A 11 12 13 14 15 7 E-E 16 17 18 19 20 21 8 H-I 22 23 24 25 26 27 28 9 Tst 29 30 31 32 33 34 35 36 10 I-A 37 38 39 40 41 42 43 44 45 11 I-D 46 47 48 49 50 51 52 53 54 55 12 N-R 56 57 58 59 60 61 62 63 64 65 66 13 IRA 67 68 69 70 71 72 73 74 75 76 77 78 For the purposes of illustrating each of the stimuli with a paired comparison, the following subset of paired comparisons are included in this appendix: 1,3,6,10,15,21,28,36,45,55,66,78 (Note that this subset comprises the top tier of the triangle shown above.) 150 I-' U1 I-' CHILD-CARE EXPENSES Is ·Child care' a personal expense ~r a business expense? If it is a personal expense then childless taxpayers shouldn't have to subsidize those with children. If it is a business expense, then it should be fully deductible. Congress has decided something in the middle. Prior to 1976, taxpayers were able to deduct limited child care expenses as "itemized deductions" if the cosfs were incurred to enable the taxpayer to be gainfully employed. The 1976 Tax Reform Act eliminated the deduction and substituted instead a credit for child-care expenses (Sec. 44A). The basic credit, subject to several limits, ranges from 20 pe~cent to 30 percent of the amount expended for child care. · · SAME CONTRIBU'l'IONS Before 1982 if you didn't itemize deductions you didn't receive any tax breaks for charitable contributions, so Congress changed the tax law for the years 1982 through 1986. In 1982 and 1983, 25% of charitable contributions up to $100 may be deducted from gross income (a $25 deduction). For 1984, the amount is 25% of $300 (a $75 deduction) . In 1985, 50% of all contributions may be deducted (with no dollar limit). ·In 1986, the deduction is 100% of all contributions (again with no dollar limit). In 1987, however, this provision expires. Thereafter charitable contributions will once again be allowed only if deductions are itemized. DIFFERENT f-' U1 N CONTRIBUTIONS Before 1982 if you didn't itemize deductions you didn't receive any tax breaks for charitable contributions, so Congress changed the tax law for the years 1982 through 1986. ·In 1982 and 1983, 25% of charitable contributions up to $100 may be deducted from gross income (a $25 deduction). For 1984, the amount is 25% of $300 (a $75 deduction). In 1985, 50% of all contributions may be deducted (with no dollar limit). In 1986, the deduction is 100% of all contributions (again with no dollar limit). 'In 1987, however, this provision expires. Thereafter charitable contributions will once again be allowed only if deductions are 1temized. SAME CAPITAL GAINS The name of the tax game in the United States is capital gains. Section 1202 authorizes individuals to claim a special deduction equal to 60 percent of the net capital gain realized in a year. This special deduction is the equivalent of a 60 percent tax deduction. If in 1983 an individual had only net capital gain inGome from stocks and bonds, then the real effective tax rate for the taxpayer would range from 4.4 to 20 percent rather than the normal rate range of 11 to 50 percent which applies t0 salaries, wages, interest, and dividend income. DIFFERENT . 1-' Vl w CAPITAL GAINS The name of the tax game in the United States is capital gains. Section 1202 authorizes individuals to claim a special deduction equal to 60 percent of the net capital gain realized in a year. This special deduction is the equivalent of a 60 percent tax deduction. If in 1983 an individual had only net capital gain income from stocks and bonds, then the real effective tax rate for the taxpayer would range fro.m 4.4 to 20 percent rather than the normal rate range of 11 to 50 percent which applies to salaries, wages, interest, and dividend income. SAME ENTERTAINMENT COSTS In 1978 the three martini lunch proved more powerful than the Carter Administration. Tax deductions for liquor, country clubs, and athletic events were continued despite proposed limits. Currently, to be tax deductible, entertainment costs must be "ordinary", "necessary", and "reasonable". In addition, Internal Revenue Code Sec. 274 details certain record keeping requirements. DIFFERENT I-' lJl 0::. ENTERTAINMENT COSTS In 1978 the three martini lunch proved more powerful than the Carter Administration. Tax deductions for liquor, country clubs, and athletic events were continued despite-proposed limits. Currently, to be tax deductible, entertainment costs must be "ordinary", "necessary", and "reasonable".. In addition, Internal Revenue Code Sec. 274 details certain record keeping requirements. SAME . ACRS In 1981, to stimulate the economy, Congress replaced the property depreciation laws with th Accelerated Cost Recovery System (ACRS). This system dramatically · shortens estimated lives for most newly purchased properties and thu dramaticallly increases write-offs for most investors. Autos can now be written off in 3 years and apartment buildings and shopping centers in 15 years. --, DIFFERENT f-' Ul Ul ACRS In 1981, to stimulate the economy, Congress replaced the property depreciation laws with the Accelerated Cost Recovery System (ACRS). This system dramatically shortens estimated lives for most newly purchased properties and thus dramaticallly increases write-offs for most irivestors. Autos can now be written off in 3 years and apartment buildings and shopping centers in 15 years. SAME GIF'fS/ AWARD You are not taxed on a $5,000 gift from your rich .uncle or on a $5,000 Pulitzer Prize. However, you are taxed on a $5,000 lottery ticket winning, or church raffle prize. Basically the law says that a person who receives a gift is not taxed, but a person who receives a prize or award is taxed unless they meet the following Sec. 74(b) exception: "Exception. - Gross income does not include amounts received as prizes and awards made primarily in recognition of religious, charitable, scienti~ic, educational' artistic, literary, or civic achievement, but only if- (1) the recipient was selected without any action on his part to enter the contest or proceeding, and (2) the recipient is not required to render substantial future services as a condition to receiving the prize or award." DIFFERENT I-' U1 Cl" GIFTS/AWARD You are not taxed on a $5,000 gift from your rich uncle or on a $5,000 Pulitzer Prize. However, you are taxed on a $5,000 lottery ticket winning, or church raffle prize. Basically the law says that a person who receives a gift is not taxed, but a person who receives a prize or award is taxed unless they meet th~ following Sec. 74(b) exception: "Exception. - Gross income does not include amounts received as prizes and awards made primarily in recognition of religious, charitable, scientific, educational, artistic, literary, or civic achievement, but only if- (1) the recipient was selected without any action on his part to enter the contest or proceeding, and (2) the recipient is not required to render substantial future services as a condition to receiving the prize or award." SAME EXTRA EXEMPTIONS Should there be extra exemptions allowed for old age and blindness? These special exemptions have been in the tax law since the 1940s. Today these· exemptions can total as much as $6,000 for a taxpayer and spouse if both are blind and both are over 64. DIFFERENT t-' lJl -...J EX:rRA EXEMPTIONS Should there be extra exemptions allowed for old age and blindness? These special exemptions have been in the tax law since the 1940s. Today these exemptions can total as much as $6,000 for a taxpayer and spouse if both are blind and both are over 64. SAME HEALT~ INSURANCE Health insurance paid by an employer for an employee is not taxed to the employee. This can be very beneficial for the employee who wants insurance. For if an employee got an increased salary instead of insurance, he would have to pay incom.e tax on the extra income, and then buy health insurance with what was left. DIFFERENT t-' Vl 00 HEALTH INSURANCE Health insurance paid by an employer for an employee is not taxed to the employee. This can be very beneficial for the employee who wants insurance. For if an employee got an increased salary instead of insurance, he would have to pay income tax on ~he· extra income, and then buy health insurance with wh~t was left. SAME TRUSTS · How do trusts help the wealthy? Mainly, they allow the shifting of income from high income to low income bracket taxpayers. Although wage income can't be diverted to a trust, income producing property like stocks and bonds can be put in a trust so its earnings (dividends/interest) isn't taxed to the high income bracket taxpayer. DIFFERENT t-' U1 I..C TRUSTS How do trusts help the wealthy? Mainly, they allow the shifting of income from high income to low income bracket taxpayers. Although wage income can't be diverted to a trust, income producing property like stocks and bonds can be put in a trust so its earnings (dividends/interest) isn't taxed to the high income bracket taxpayer. SAME INCOME AVERAGING What happens when a $30,000 a year baseball player gets a million dollar contract? Income averaging provisions were enacted in 1964 and liberalized in 1969 to help reduce the taxes of people with - fluctuating incomes. The Code specifies the conditions and calculations for income averaging in Sees. 1301-1305. First the taxpayer's average taxable income for the prior four years is calculated. This average is then multiplied by 120 percent. The -result is subtracted from the current years t?xable income. If the difference is $3,000 or more, then income averaging may be used. DIFFERENT t-' 0\ 0 INCOME AVERAGING What happens when a $30,000 a year baseball player gets a million dollar contract? Income averaging provisions were enacted in 1964 and liberalized in 1969 to help reduce the taxes of people with fluctuating incomes. The Code specifies the conditions and calculations for income averaging in Sees. 1301-1305. First the taxpayer's average taxable income for the prior four years is calculated. This average is then multiplied by 120 percent. The result is subtracted from the current years taxable income. If the difference is $3,000 or more, then income averagin9 may be used. SAME INTEREST DEDUCTION Interest expense is the most important itemized deduction for individuals. In 1979 this deduction totaled $73.6 billion, or 40.7 percent of all itemized deductions. Virtually all interest paid is deductible. The major exceptions are interest on money borrowed to purchase 'tax free securities and a limit on the amount of interest deductible on funds borrowed to purchase investment property. --------------" DIFFERENT f-' 0'\ f-' INTEREST DEDUCTION Interest e~pense is the most important itemized deduction for individ~als. In 1979 this deduction totaled. $73.6 billion, or 40.7 percent of all itemized deductions. Virtually all interest paid is deductible. The major exceptions are interest on money borrowed to purchase tax free securities and a limit on the amount of interest deductible on funds borrowed to purchase investment property. SAME NONRECOGNITION OF GAINS AND LOSSES Sec. 1031 is a very important tax law for real estate investors. This law allows some direct exchanges of business and investment property without taxation. Sec. 1031(a) states in part "No gain or loss shall be recognized if property ••• is exchanged solely for property of a like kind ••. ". The Treasury Regulations are rather liberal in intepreting what is "like kind"~ a country farm can be exchanged for a city apartment without taxation. DIFFERENT 1-' 01 N NONRECOGNITION OF GAINS AND LOSSES Sec. 1031 is a very important tax law for real estate investors. This law allows some direct exchanges of business and investment property without taxation. Sec. 1031(a) states in part "No gain or loss shall be recognized if property ••• is exchanged solely for property of a like kind ••• ". The Treasury Regulations are rather liberal in i'ntepreting what is "like kind"; a country farm can be exchanged for a city apartment without taxation. SAME I.R.A. As all the ads might lead you to suspect, Code Sec. 219, permitting individuals to establish their own indiVidual retirement plan, is a far reaching change. Now all workers can deduct up to the lesser of $2000 or 100 percent of earned income for an IRA. Previously, only those without a · qualified pension plan could take out an IRA, and then they were limited to $1,500 or 15 percent of earned income. DIFFERENT APPENDIX C Phase 1 Data Part 1: MDS data. Each triangle represents one subject's 78 tax complexity judgments. Subjects are listed sequentially from 1 to 30. Data reflects measurements based on the subject's placement of an X on a 5 inch line. The measurements are recorded to the nearest tenth and the decimal point is omitted. Judgments are entered in the sequence shown in Appendix B forward. Part 2: Phase 1 adjective descriptor data. Part 3: Phase 1 background data. 163 PART 1: 78 MDS JUDGMENTS FOR EACH OF THE 30 SUBJECTS 0001 0 00 00020 4·=· ·-· 00 000:30 48 24 00 0004 0 4::: 2E· 4!3 00 00050 01 24 2.-. -~· 47 00 OOOE.O 4::: 4E: 4"" . I 2:;: 02 00 00070 47 4:=: 25 ~ .... e,._r 20 24 00 I) 0 o::: (I 02 45 .-.r: .::..-~ 27 47 46 02 I) (I 00090 47 25 02 4'7 . I 2E· 24 01 02 I) (I 001 00 47 04 24 24 2:3 46 46 22 46 (II) 0011 0 45 25· 2E· 4'=' 2::: 02 02 47 ·=-·-=· 25 (II) ._. ~--· 00120 04 2:;: 01 47 02 2:::: 45 25 02 46 46 00 oo1::::o 4·=· ·-· 02 o:::: 47 2:3 03 01 48 24 26 2E· 47 I) I) 00140 00 00150 1.-. .::.. 00 00160 4:::: 0'? (II) 00170 4'~ -· 2•3 49 00 00180 02 - .. -. -~··=· 14 1.-. ·=· 00 00190 45 •JC" ~--· 41 •'j·=· ·-·L- 2.::. 00 00200 4!3 46 47 16 .-.~ -:.·• 4.-. .::.. 00 0021 0 46 44 19 22 04 1·-:· ·-· 25 00 00220 2'? 44 0'? 46 24 41 4E. 08 I) (I 00230 44 0:3 ·=-·=- 44 45 .-.1:' o:::: 02 t·-· 00 ..... _. C.·-· .::.. 00240 ::::7 1 --=· ·-· 4E. 14 04 44 40 ·-·=- C.·-· 05 02 00 00251) 44 :;:::: 1 ~. 2~ 02 (r::: 46 47 ~ .... 0'? ~:·:;. 00 .::.. .,: .. _. 00260 12 0'? 44 16 24 4E. 46 ~:4 0'? 09 .-.c- C.·-· 4"":• -· I) (I 00270 00 00280 40 00 002';.(1 :36 04 00 oo::::oo 27 07 06 00 oo::::1 I) 1 0 4"":• ·-· 45 ·=oo~ ~ •. 00 oo::::2o 05 4·=· ·-· 45 24 04 (II) oo::::3o 01 45 4'=' ·:a If: o:::: 01 00 ·-· L...·-· o o::::4o o:::: 44 42 26 05 01 01 (I (I 00350 01 4'=' ·-· 42 2S' 02 01 01 01 00 oo::::E.o 01 44 40 1.-. .::.. o:::: 06 07 07 04 (II) oo::::?o 01 .-.=- .; .. _. ::::4 o:3 22 2E· .-.-, .:...· :;:o 1:::: 05 00 I) o::::::: 0 42 06 or:. 17 15 ::::1 31 .-.. -. -:.··=· 2:3 ::::I) 07 I) (I 003'?0 02 4"=• ...... 41 23 0'? 01 01 01 07 2€· 27 ::::o (II) 164 PHASE 1 MDS DATA {CONTINUED) 00400 00 00410 09 00 00420 25 47 00 00430 47 44 47 00 00440 47 45 46 06 00 00450 06 06 08 07 45 00 00460 08 28 04 06 07 06 00 00470 38 45 45 46 28 08 07 00 00480 47 47 17 47 05 45 45 40 00 00490 46 09 06 45 06 06 40 09 06 00 00500 11 08 13 07 08 09 46 07 36 28 00 00510 45 35 07 45 07 43 42 08 07 08 07 00 00520 41 09 27 43 41 08 08 38 10 24 07 40 00 00530 00 00540 33 00 00550 37 02 00 00560 49 18 17 00 00570 47 47 35 07 00 00580 48 48 48 48 48 00 00590 40 24 35 47 36 48 00 00600 48 48 39 33 48 04 47 00 00610 48 48 24 48 48 18 48 31 00 00620 48 48 49 49 25 35 45 44 48 00 00630 48 17 40 40 44 49 48 48 48 47 00 00640 46 46 48 32 47 47 47 48 25 28 48 00 00650 48 42 46 26 49 26 05 03 48 24 48 48 00 00660 00 00670 10 00 00680 36 39 00 00690 13 08 36 00 35 33 08 33 00 00700 '00710 38 00720 05 06 36 23 00 05 08 33 39 41 40 00 41 09 45 06 08 00 43 05 06 44 45 44 04 43 45 43 05 05 44 46 45 44 05 45 47 10 00 05 00 03 45 00 05 02 47 00 00730 08 08 00740 40 11 08 00750 44 46 05 00760 46 06 46 00770 46 47 03 00780 47 03 13 oo790 ·oo 49 10 12 44 25 06 08 33 05 00 00800 06 00 00810 26 25 00 00820 12 36 24 00 00830 47 16 01 15 00 00840 10 16 05 23 42 00 00850 42 38 20 11 11 38 00 00860 43 40 45 40 12 37 11 00 00870 39 38 14 36 08 40 09 10 00 00880 47 10 16 43 04 32 34 10 07 00 00890 42 42 42 43 14 40 12 41 11 47 00 00900 50 49 08 40 03 35 45 40 05 27 37 00 00910 40 44 40 41 :3 42 08 04 45 44 06 44 00 165 PHASE 1 MDS DATA (CONTINUED) 00920 00 00930 02 00 00940 47 05 00 00950 04 04 03 00 00960 04 44 40 41 00 00970 05 41 40 05 05 00 00980 42 42 07 43 44 44 00 00990 43 47 46 47 47 48 47 00 01000 45 26 27 25 40 45 47 43 00 01010 29 06 10 44 44 45 47 47 45 QO 01020 45 45 48 47 47 46 42 44 45 47 00 01030 43 46. 44 47 10 44 41 43 40 4? 4? 00 01040 48 46 07 43 46 48 47 46 48 26 46 48 00 01050 00 01060 11 00 01070 25 47 00 01080 04 04 44 00 01090 48 47 09 43 00 01100 44 06 45 41 20 00 01110 08 07 44 46 46 45 00 01120 46 05 47 03 46 46 05 00 01130 47 47 34 47 04 24 46 48 00 01140 47 48 24 46 08 40 48 47 36 00 01150 46 45 24 23 43 38 46 47 47 43 00. 01160 48 47 35 46 12 34 49 48 47 27 39 00 01170 37 46 ·02 39 18 26 47 47 38 05 41 37 00 01180 00 01190 23 00 01200 48 01 00 01210 01 24 24 00 01220 49 49 48 48 00 "01230 48 47 48 48 48 00 01240 48 47 48 48 48 48 00 01250 48 48 49 48 49 48 48 00 01260 47 48 26 48 48 48 49 49 00 01270 25 25 01 49 49 49 49 48 48 00 01280 48 49 48 49 49 49 49 49 49 49 00 01290 49 49 49 48 01 49 49 48 49 49 48 00 01300 49 25 02 48 48 48 49 49 49 25 48 48 00 01310 00 01320 23 00 01330 11 11 00 01340 08 24 11 00 01350 09 29 08 10 00 01360 17 08 21 23 22 00 01370 23 33 13 10 06 24 00 01380 15 30 15 04 12 28 06 00 01390 10 29 08 05 01 22 02 07 00 01400 03 11 12 18 26 05 25 12 19 00 01410 05 23 10 10 12 14 13 11 12 17 00 01420 06 18 07 01 12 23 05 04 08 22 09 00 01430 08 10 10 14 17 06 27 17 16 06 i7 10 00 166 PHASE 1 MDS DATA ~~qNTINUED} 01440 00 01450 02 00 01460 48 49 00 01470 49 49 49 00 01480 25 25 03 48 00 01490 02 02 02 48 25 00 01500 49 49 49 02 26 49 00 01510 02 49 49 02 26 49 02 00 01520 02 49 49 02 25 02 49 0~ 00 01530 02 02 48 49 02 02 48 48 02 00 01540 01 02 02 48 02 48 48 02 25 02 00 01550 02 48 02 02 02 47 48 02 02 48 22 00 01560 02 02 02 48 24 02 48 25 02 02 02 48 00 01570 00 01580 25 00 01590 48 48 00 01600 46 45 44 00 01610 46 45 47 47 00 01620 48 47 28 45 46 00 01630 46 47 45 47 46 47 00 01640 45 46 47 47 47 48 47 00 01650 47 47 26 47 47 47 48 47 00 01660 34 34 28 49 46 34 47 47 45 00 01670 12 14 15 46 30 48 14 45 46 48 00 01680 48 45 42 45 44 23 45 24 24 28 29 00 01690 46 28 26 47 47 47 22 48 46 47 49 48 00 01700 00 01710 33 00 01720 49 37 00 01730 06 05 49 00 01740 49 49 03 49 00 01750 25 03 49 49 49 00 01760 04 26 48 48 49 49 00 01770 03 37 49 48 48 11 24 00 01780 49 37 10 24 03 23 48 38 00 01790 09 23 12 49 24 35 48 49 40 00 01800 11 09 49 09 49 38 23 09 09 48 00 01810 49 36 25 1~ 09 10 49 49 10 39 37 00 01820 03 12 48 38 49 39 22 09 03 24 25 49 00 61830 00 01840 38 00 - 01850 22 08 00 01860 43 37 06 00 01870 40 06 08 09 00 01880 39 15 07 12 10 00 01890 05 39 39 38 41 40 00 01900 23 38 40 40 40 40 06 00 01910 41 09 08 08 08 15 43 39 00 01920 43 40 06 24 22 32 42 42 07 00 01930 19 09 09 06 06 11 39 41 09 19 00 01940 40 14 17 24 13 12 37 45 12 10 24 00 01950 10 36 37 38 39 35 15 13 35 24 23 12 00 167 PHASE 1 MDS DATA (CONTINUED) 01960 00 019?0 48 00 019:?.0 49 4'31 00 01990 23 03 48 00 02000 48 48 26 48 00 02010 48 48 27 48 48 00 02020 48 48 48 48 48 48 00 02030 48 48 48 48 4R 48 48 00 02040 48 48 04 48 04 26 48 48 00 02050 48 25 25 03 02 48 48 48 02 00 02060 48 48 02 02 4R 48 48 48 02 25 00 02070 48 48 02 48 02 48 48 48 02 02 28 00 02080 48 48 02 48 48 23 48 48 02 03 02 48 00 020'310 00 02100 48 00 02110 47 02 00 02120 36 03 30 00 02130 37 09 46 03 00 02140 24 45 10 24 05 00 02150 24 03 03 22 23 21 00 02160 46 25 05 45 04 45 24 00 02170 21 45 05 05 46 26 47 21 00 02180 25 48 23 04 44 46 26 47 46 00 02190 03 23 02 04 04 05 25 44 05 05 (II) 02200 45 23 44 2~ 05 46 25 27 46 45 25 00 02210 43 03 46 26 26 45 03 24 03 44 25 46 00 02220 00 02230 01 00 02241) 39 36 (1(1 02250 08 07 50 00. 02260 27 34 50 02 00 02270 26 26 49 02 02 (1(1 02281) 36 49 49 02 02 01 00 022'310 o2::::oo 38 28 49 02 02 02 02 00 39 28 49 02 02 14 02 02 (II) 02;310 10 02 27 49 34 26 49 49 49 00 02::::20 49 023::::o 40 02340 17 02::::50 00 28 39 02 02 26 0~ 02 02 42"0(1 26 14 02 02 26 02 02 02 39 02 00 28 49 16 29 l2 13 09 11 02 12 12 00 02::::6 (I 09 0 (I 023?Cr 19 ::::6 oo 02380 46 05 03 00 02390 22 22 21 19 00 02400 02 03 13 18 27 00 02410 47 16 46 26 33 28 00 02420 45 22 33 42 11 23 32 00 0243 0 0'? 12 2';. 0244 0 "(i2 1 '? 02 43 43 05 44 2:?. 24 o:::: 42 02 2:;: 24 04 o:::: cr:=:450 11 42 05 1)(1 02 04 00 03 25 42 00 D2460 46 40 25 24 32 22 14 41 03 47 34 00 02470 46 41 10 25 13 24 13 45 46 22 00 168 PHASE 1 MDS DATA (CONTINUED) 02480 1)1) 02490 :.::2 00 02500 33 42 00 02510 15 37 35 00 02520 43 45 07 41 00 02530 41 44 18 30 34 00 02540 42 13 47 40 48 4? 00 02550 36 30 45 46 46 46 23 00 02560 45 45 28 44 20 25 47 39 00 02570 40 41 41 35 43 45 17 44 29 00 02580 29 35 19 2? 32 23 40 47 34 35 00 02590 37 45 28 34 30 28 44 45 28 35 31 00 02600 37 42 30 37 35 35 46 43 31 31 31 29 00 02610 00 02E.20 32 00 02E.~: 0 41 2:.::: 0 0 02640 05 25 31 00 02650 03 07 36 13 (II) 02660 38 32 30 08 12 00 02670 03 33 0? 06 16 36 00 02680 44 41 41 13 39 42 05 00 02690 45 42 42 08 0? 40 06 17 00 02700 27 06 05 33 42 09 47 40 45 00 02710 40 38 16 14 09 28 33 04 05 41 00 02720 41 15 05 18 41 08 46 42 07 06 10 00 02730 44 44 41 10 06 21 41 08 06 25 43 39 00 . 02740 00 02750 25 00 02760 :3(1 11 00 02770 40 35 16 00 02?80 43 38 44 07 00 02790 11 03 42 23 45 00 02800 34 05 41 40 44 25 00 02810 30 28 38 39 47 0? 11 00 02820 48 48 44 42 42 38 27 38 00 02830 47 42 26 32 38 48 45 32 14 00 02840 32 41 38 29 26 38 44 30 11 10 00 02850 46 48 30 26 43 44 46 42 47 34 26 00 02860 46 29 1? 26 26 34 35 36 46 45 45 46 00 02870 00 02880 47 00 02890 47 4? 00 02900 47 47 47 00 02910 4? 47 46 46 00 02920 46 46 46 46 46 00 02930 46 46 46 46 46 46 00 02940 47 47 47 4? 4? 4? 45 00 02950 46 47 04 47 03 4? 47 46 00 02960 47 48 48 4R 47 47 48 47 48 00 02970 47 04 48 47 47 4? 48 48 03 48 00 02980 48 48 03 48 48 48 47 48 03 48 4? 00 02990 47 47 47 48 4R 48 48 48 03 03 03 48 00 169 PHASE 1 MDS DATA (CONTINUED) o:;: (lrJI) 0 0 o::::: I) 1 (I :::::4 (I 0 o::::: 02 (I ::;::=: 35 (II) 0:3030 15 :37 38 00 03040 37 11 12 :38 00 o::::: 05 I) ·::::.:. 12 11 17 :36 (II) 03060 14 :34 23. :38 36 10 00 o3o7o 39 12 1::::: 12 12 12 n oo 03080 38 1:3 10 :35 12 10 46 46 00 03090 42 18 05 25 09 18 42 :34 18 00 03100 46 43 02 43 08 44 44 44 02 45 00 03110 45 43 1)4 45 06 44 45 43 06 42 34 00 03120 42 :39 17 37 12 06 38 25 02 41 01 39 00 o::::: 13 0 I) 0 1):314 0 :::::9 0 0 03150 43 4::::: ·00 03160 13 33 24 00 03170 39 46 35 15 00 03180 22 35 47 25 38 00 03190 07 23 10 03 04 16 00 03200 07 03 03 06 05 01 06 00_ 03210 08 10 03 02 12 10 43 37 00 03220 40 48 24 33 18 42 47 45 11 00 03230 40 07 23 44 44 46 42 24 47 44 00 . 03240 41 49 05 12 04 38 48 39 06 16 38 00 0:3250 42 :37 15 24 22 45 47 10 23 05 34 44 00 o:::::260 oo 1):327 0 :::::2 0 0 1:.:::::2::: o o·;. 32 o o 0:3290 28 45 28 00 0:3:300 32 43 31 09 00 '03310 23 38 26 33 16 00 1)::;:::;:2 0 :::::9 4E. 4E. o::::::::::::::o 16 41 22 1):3::;:4 0 44 ::::E. 06 06 03 15 00 20 07 19 07 (II) :32 04 2E· 34 o::::::::::50 09 29 07 11 11 24 34 13 07 00 o:::::::::.:. o 26 36 :::::4 12 12 21 12 11 :=:4 24 1)1) 1):::::370 (r:3:::::::: (I :39 45 39 21 :34 1:3 :36 42 20 34 28 00 05 17 23 35 20 11 19 19 17 07 17 22 00 o3:::::·;.o oo 1):34 0 0 4::: 0 I) 0341 0 24 2::::: 0 0 03420 47 05 24 00 03430 45 25 04 25 00 03440 04 45 46 03 46 00 03450 03 46 47 46 46 25 00 03460 25 44 45 43 44 25 26 00 03470 45 46 26 27 44 46 46 47 00 03480 47 26 27 27 27 43 47 46 27 00 03490 03 04 04 24 25 44 46 47 05 04 00 03500 03 24 03 45 03 03 47 48 03 25 26 00 03510 25 24 26 46 24 04 48 02 47 46 27 25 00 ·170 PHASE 1 MDS DATA (CONTINUED} 03520 00 03530 4R 00 03540 ~Q ~~ 12 00 03550 48 04 03 00 03560 40 y 03 1? ~ 03 00 03570 40 u 1~ 0 03 48 4~ 0 00 03580 04 37 03 09 47 05 00 03590 09 ~~ ~~ 05 48 ~~ ~0 04 06 00 03600 05 29 05 04 09 04 03 45 00 0361 0 ~~ 1? 09 44 30 ~~ 16 47 44 00 cc ~ ~( 03620 04 07 15 24 06 08 38 07 07 47 00 03630 4? ~ 41 19 30 17 05 04 47 03 47 18 00 03640 46 04 08 47 17 19 47 19 31 15 1 1 ?7 ~· 00 03650 00 03660 42 00 03670 09 49 00 03680 08 39 38 00 03690 44 1~ ~ 08 1 0 00 03700 40 y 07 05 44 37 00 0371 0 04 37 44 1 1 05 46 00 03720 06 44 1 1 ~~ ~( 12 05 06 00 03730 4? ~ 05 07 39 07 08 44 09 00 03740 ~~ 45 42 4? 08 40 42 39 14 00 ~~ ~ 03750 42 09 07 42 08 ~~ ~( 1 0 41 06 4~ ~ 00 03760 45 17 1 1 41 40 41 14 ~1 07 06 40 00 03770 47 47 12 42 44 46 46 44 1? ~ 44 42 13 00 03780 00 03790 02 00 03800 02 02 00 0381 0 02 02 02 00 03820 02 02 02 02 00 03830 02 02 02 02 02 00 03840 02 02 02 02 02 02 00 03850 02 02 02 02 02 02 02 00 03860 02 02 02 02 02 02 02 02 00 03870 02 02 02 02 02 02 02 02 02 00 03880 02 02 02 02 02 02 02 02 02 02 00 03890 02 02 02 02 02 02 02 02 02 02 02 00 03900 02 02 02 02 02 02 02 0~ 02 02 02 02 00 END OF DATA 171 PART 2 : PHASE 1 PROPERTY FITTING 7 PROPERTY FITTING SCORES EOR EACH OF THE 13 TAX SCENARIOS (A TOTAL OF 210 JUDGMENTS) FOR EACH OF THE 30 PHASE 1 SUBJECTS 0001 0 4::: ·=··":• 01 25 24 21 ,.,J:' 01 ·=-·-=- 01 .-.C" 47 22 &.;;..·-· ~--' &.-·-· C.·-· 00020 25 2:~: -·= 01 24 21 01 24 2:3 01 .-.c:- .-.-. 02 C.·-· .::.-· .::. .,:. ooo::::o 01 -=--·-· 01 4'3 01 21 24 24 2:3 2:::: .-.~ 47 03 ~--=· C.·-· 00040 -·= ·~·-· 01 25· -·= 48 24 4E: 01 01 4·=· 2:::: 24 C·-• .:: . .;. C.·-' ·-· 00050 47 01 25 01 4::: 4•::0 ·-· 01 .-.r:: .:.._• 2:::: -..I:" C.·-' -=-= t...·-· 02 47 OOOE.O 02 -=-= L...·-' .-.r::: c.-· 48 21 4::: 01 48 49 01 4':;. 2~: 26 00070 2E. 4·=· 25 25 22 4·=· 4'? 01 4'? 01 .-.c::" 47' 49 ·-· ·-· C.·-· ooo:::o 40 41 2E. .-,.., 15 1 "? 40 22 2';.t 3€. ·-=··-=· 24 .-.c::- .;, ... I --··-· ·=··-· (r(I0'30 44 25 42 07 15 .-.~: :;:E. :;::3 :;:? 21 24 .-.c::" .-,.-. C.·-· C.·-· -:.··=· 001 00 05 :;::!: o·;. 45 15 25 04 1 I) .-.c- C.·-· 27 ·:··:· ._ .. _. 17' o::: 001 1 0 ,.,.,. o~: 20 22 1 0 1 0 14 :;:~: .-,.., 20 -.c:- -=·-=- 40 C,._l ·=··· .,; .. _. '""-· 00120 .-.. -. t::.·=· o:::: 1 .-. c. 34 2'~ 15 06 17 15 :3(1 ~:E. ·=-I:' L-·-· 1 ':• ·-· oo1::::o 07 25 2::: o::: o::: .-.r:: 05 :~::3 30 15 1 1 1·-:· 40 C.·-· ·-· 00140 17 16 .-,-, t::: 21 1::: 05 09 1·=· 14 1 1 .-,,-, 40 ·=··· ·-· .::.o 00150 15 25 24 1 1 07 1·=· ·-· 01 01 16 14 14 o::: 05 00160 42 16 o:::: OE. ,.,.,. c.._; :;:~~: 49 4::: -·= C:·-· 24 2:::: 02 4'? 00170 42 4'? 47 45 40 4·:· ·-· o::: 24 41 24 3:=: 41 1 "":• ·-· oo1:::o 17 o:::: 01 ::::1 02 o:::: 43 17 04 o·;. 02 01 17 001'~0 25 2E. .-..-. -.:.··=· 4.-. -=· ·=-·=- "-'-' ::::4 o::: 02 34 t:3 23 2E. OE· 00200 01 08 01 01 02 06 01 48 (18 06 01 01 1::: 00210 01 01 1)1 01 01 01 01 07 08 1)1 (11 16 02 00220 29 28 40 42 4E. 44 09 :~:7 45 05 15 45 12 002::::0 45 1)5 05 04 I):::: 24 42 2E. 04 (13 46 05 17 00240 ~:1 35 .-.c:"' .,; .. _. ~:2 31 .-.. -. ·.:··=· 05 2E· 27 2€. 1 I) ::::o o·;. 00250 45 :;:5 04 04 04 o::: 05 ,.,.,. C.·-· 04 47' 45 06 45 002E.O ~:E. 4:::: 44 45 44 ~-·C" . ,.: .. _. 05 ,.,.,. .:;:._ . 47 46 25 :;:·;. 26 00270 .-.. -. .:.·=· 45 1 .-, c. 04 04 05 04 o:::: OE. 4€. 45 02 45 002:::0 12 16 07 06 04 07 05 42 fl"? -' 04 1)4 25 26 002'?0 2~: 17 1 0 -:-.-:- .-.. -. 01 01 01 22 1.-. 1)2 27' 01 '"'---' c. c. 0::• 01):300 49 21 •jC:: 24 49 4•=. 4'3 50 ..-.. ~ 49 4:3 - .. -. 4'? 1,.;.·-· -· c;.;. c.·=· 0031 I) 24 23 43 4:::: 02 1 ·;. 01 01 -=-~· 02 .-.. -. 04 01 ._._. e,.;, • oo::::2o 4'':• -=---=· 45 :=:E. 4'3 24 'l'::! 4::: 02 49 48 02 .-.c- ·-· L;;;;.·-· C.·-· oo::::::::o 4':0 -":t7 45 4::: 01 .-.. -. 01 01 45 01 01 4'3 02 ·-· ·-·· c.c 0Ct:340 4'=· ·-· 14 07 OE: 4'3 01 01 49 (12 49 4'? 01 02 00:350 01 02 02 02 02 01 01 01 02 01 01 02 _03 oo::::E. o o:::: 03 .:::1 1 0 1 1 06 05 1)2 39 09 05 07 02 l)(r370 2E· 45 ::::6 1·=· 46 4'=· 4"':• 34 .-.r:: 4':• ~::::: 2::::: 48 - ·-· ·-· ·-· C.--· --· 0 o::::::: 0 4':0 47' 1·:· 40 1 ·:· 40 29 09 .-.c::::.- 06 14 :;:::: 1'0 ·-· ·-· ·-· C.·-· oo::::9o 2::=: 4::: .-.C" 4·=· 49 4·=· 40 48 :31) 49 4::: -.~ 4'? .,: .. _. . ·-· . ·-· .,; ... 00400 .-.c:- 40::• 1 1 . -.. -. 1 I) :~:E. 31 15 :3.: . 04 21) 49 01 . .: .. _. ·-· -.:.··=· 0041 0 02 4';:. ·-· 12 07' 1 0 1 1 02 41 1 ':• ·-· 47 4'3 07 40 00420 19 1-· c. OE. ::::7 21 24 02 17 14 07 09 40 OJ 00430 2:~: 2:.:: ·-:-·:· 0'? 1 ·=· 31 1::: 02 20 '3:=: 02 .-.~ 01 ·-'"- . ·-· C.f 00440 00 00 00 00 00 00 00 41 00 1)(1 2~::: 2~::: 4.-. c. 00450 o::: 22 09 .-,r:: ·=··-· 00 12 22 12 24 07 25 .-.~ ·=··-· ·=··-=· L-·-· 00460 4'3 ::::6 24 47 :::;:4 4(1 13 41 09 24 42 o::: 44 00470 2'? 36 25 44 flO 24 00 2:3 00 00 24 ::::o 22 172 PHASE 1 PROPERTY FITTING (CONTINUED) 1) (14:::: I) 46 25 05 2.5 41 4.5 1)~. 46 01 03 ·;:c.· 01 .... -. ~t::. 1) 049 0 ·=··:· '-·-· 0? 03 ~~::=: :~:(I 2:~: 05 :~:6 01 25 4.-. . .::. 01 41 00501) 12 45 0'3 12 0? 45 1 (I 01 02 07 o::: 1 0 09 0051 0 4? 0'3 0'3 t·-· c 4? o:=: 45 50 06 46 1 I) 46 00 (I ~)52 0 09 44 4~. 4'=' ·-· 46 44 09 01 44 1 0 41 42 o·~ I) 1)5:~: 0 45 1 1 47 46 1 1 09 1 1 46 (1:3 46 42 45 00 00540 46 45 47 46 44 45 1 1' 01 4":0 ·-· 0'? 43 46 OS 00550 09 1 1 o::: 1 0 1 0 1 0 1 1 00 0:3 09 07 1 0 OS 00560 09 0'3 09 1 0 10 1 1 1 1 47 4:3 44 4·-=· ·-· 43 0'3 0057(1 04 02 4C• ·-· 02 47 25 04 02 49 2? 09 46 :~:4 (I (15::: 0 48 4·=· .-.=: 27 2:3 47 49 4? 26 4·=· 4:~: :;: 1 .-,...., ·-· .:,._. ·-· .; ... 0 05'3 0 .-.c: .::,._, 01 :~:.:. 2? 22 ::::5 o:~: 1 '3 27 14 o:::: 1 1 01 00600 02 4::: ~:.:. 1·-=· ·-· 01 1E. 48 1"':• ·-· 04 04 44 1 ·:· ·-· 4? 0061 (I 2E. 2:::: 25 :;:::: 24 2E. 01 04 27 2E. 09 24 01 00620 02 48 :3~. 02 o:;: 2E· 4E. 05 0:3 03 4.5 04 48 006:30 02 2E. 24 2~! 2:~: 2:3 02 46 04 04 1 0 25 02 0 OE.4 0 01 49 01 01 01 49 01 01 4'3 01 01 01 01 oo.:.so 4:=: 01 01 01 01 01 4'~ -· 49 01 4'3 01 4'3 4? 0 OE.6 0 01 49 4'3 49 49 49 02 01 4'3 01 49 49 01 OOE.70 48 01 4·=· ·-· 01 01 01 49 4'~ -· 4'=· ·-· 4q -· 49 4:=: 49 I) o.:.::: 0 4·=· ·-· 48 4•:0 ·-· 49 4'3 01 02 01 4.8 01 49 4'7.1 01 OOE.'30 01 01 01 01 01 4'3 4!3 48 01 49 01 4::: 4'3 00700 4C• ·-· 01 01 01 01 4::: 49 4'3 01 4·=· ·-· 01 4q -· 4'3 0071 0 1 ·=· ~ .... 20 05 1 .-, 2'=.t 02 04 10 45 22 ~:(I 25 ·-· ·-=··-' .::. 00720 :32 22 2€. 20 19 .-.. -. :;:6 -:-·:· . ..., ... ·-:.·;:. 46 ·:··-=· 40 ·-=·c. ....... .:. ._, ·-·~ ._ .. _. oo7:;:o 1 1 0? 1 0 ::::::: o::: :~:4 15 01 1 1 12 05 .-.. -. c...:.• 02 00740 46 :31 ~:4 :~:6 12 - .. -. . ,; .. ;. 40 4:3 46 35 49 ::::? 4::: 00750 15 04 1 0 ....... .;, .. _. 1.-, .:. ::::E. 25 o:::: .-.-. .::.-.:· 1 I) 04 ·=··J .._._. .(15 00760 34 0~: 05 02 02 0'3 o::: 4::: 01 -=---=· .__. 4'=· ·-· 01 22 00?70 ·~·-· :~:4 15 13 --:.·:· 1·:· o::: 02 01 o:=: 01 01 .-,,-. .:.c. ._.._, ·-· -:.··=· oo7:::o 24 3€. 44 02 40 o::: 02 22 1 E· 42 22 o:=: 24 00?90 28 1:~: 05 4'3 4E. 42 4::: -··=. .:. -· (r;:: 27 :~:7 45 45 oo:::oo 18 :=:4 24 4E: 14 40 48 24 36 .-.c::" c.-• 21 24 ~:o OOS1 (I 40 1 C• ,_, 1)2 48 14 OE. 46 12 06 1)7 47 45 2€· oo::::2o .-,.-, 24 24 4€. ·=-·:. 24 02 ~ .... o.: . -=-·-=· 05 ·-·-=- 05 .::. . .;. ... -· ..:··-· L...·-· C.·-· oo:::::::o (1:~: 07 05 02 44 01 Of' 1)~: 06 02 46. 4C• ,_, o:~: 00840 -=-·=- 24 2:~: 44 4"':• 24 02 44 ~--. ~--. 24 46 25 ... ~ ·-· c. c. .:. . .;. 00850 1 0 o::: 07 04 22 :;a:. 24 05 05 46 -=··=- .-,r::_ 0"5 ... -· C.·-· oo:::6 o 00 4.-, 00 4·:· 22 26 2:1 44 ~--. 2.-, 2:~: 00 46 .::. ·-· c. c. ·.::0 o o:::7 o 00 41 00 07 4·-=· ·-· 42 07 26 22 2:::: 2:=: (II) 2:3 •Jo:::::::o 00 00 00 00 06 05 07 05 01 03 04 03 OJ oo:::9o 00 07 •:•o 00 2:3 2.:. o::: t)? 21 ·J"":• I,;.·-· 24 (II) .-,.-, C.·:.· 00'300 49 2:::: 00 00 ·06 05 o::: o=== 02 03 05 o:::: 04 0091 0 47" 24 00 00 .-,C" 25 0'3 0'3 21 (1:~: .-.. -. 00 2:~: c._. c. c. 00'320 02 02 22 02 4? 1 1 02 o:::: 25 12 1)5 .-.t: .::,._. 02 00930 4'3 ~3? 22 03 0'? 1 1 47 25 03 .-.c:- . .; .. _. 2:~: 1 0 4'3 0 0'34 (I ::::::: 4'3 1 0 4'3 15 :~:·:.- 4? I) (I ~ .... . .; .. _. 24 47 24 ~:::: 00'?50 4•::. -· 4'3 4::: ·-:·~ . _,, 34 .3'? 4? 46 35 46 47 I . ::::5 4'::.. 00'36 0 24 4'3 1 :::: 4'3 22 ·;::=: ~:·;.. 1):~: ~ .... .:, .. _. 24 47 2:::: ·=··:· .._._. I) 0'37 I) 01 01 02 02 0:3 1.-, .::. 0:3 46 0:3 24 :~:6 04 47 173 PHASE 1 PROPERTY FITTING (CONTINUED) 00980 1 1 ~~ ~· 09 03 36 •23 03 37 05 ~.~ ~~ 46 ~~ ~~ 23 00990 38 33 ~~ 07 31 21 1 1 02 09 23 1~ 21 05 ~~ ( 01 000 ~~ ~~ 12 08 09 09 22 36 41 1~ I • 04 26 22 22 01 01 0 31 36 09 30 33 22 ~Q ~~ 29 ~0 ~~ ~~ ~~ 1~ ~ 10 ~ 1 1 01 020 22 23 01 02 02 02 22 48 03 02 42 09 46 01 030 ~~ ~-~ 31 1 ·~ ~ 41 20 24 23 ~ 33 ~1 -.-. ~~ 23 12 01 040 01 31 01 02 01 03 04 48 02 02 39 24 46 01 050 ~~ ~~ 14 24 12 09 15 07 09 15 ~~ 24 07 ~~ c~ ~~ 01 060 03 01 49 01 4'? 4'? 02 50 50 50 50 50 01 01 070 4'? 50 50 50 01 01 49 01 01 50 01 01 50 01 080 02 01 01 01 01 01 01 01 01 01 50 01 01 01 090 49 50 50 50 01 01 50 01 01 01 01 50 50 01 1 00 01 01 01 01 01 01 02 01 50 50 50 50 01 01 1 1 0 01 02 ~2 01 01 01 49 01 01 01 01 01 50 01 120 02 02 02 01 01 01 01 01 01 01 01 01 01 01 130 03 22 44 01 47 46 03 ~~ ~~ 25 46 47 05 47 01 140 40 ~ 02 45 03 46 05 04 1 0 06 20 23 ~~ ~~ 31 01 150 22 24 34 4A 22 24 27 46 21 22 09 37 16 01 160 22 24 1~ ~ 46 47 24 47 ~~ ~~ ~? ~I ~~ c~ 19 26 06 01 170 4~ . I 46 46 46 47 16 25 23 46 03 ~~ c~ 47 26 01 180 ~~ c~ 23 46 46 03 25 26 07 09 05 25 20 18 01 190 ~~ 33 03 03 04 24 ~~ 22 40 24 05 ~~ 36 cc ~~ ~~ 01200 ~~ ~· 39 50 01 26 09 01 01 01 50 12 01 24 0121 0 4Q ~~ 1~ 01 1 0 01 50 49 01 49 ~~ 12 41 ~ ~~ ~ c~ 01220 49 ~Q ~~ 4'? 26 50 01 01 12 26 39 27 01 ~~ co 01230 ~~ 26 01 12 01 01 ~~ 50 02 01 26 01 50 c~ ~~ 01240 ~~ 40 ~~ 50 28 ~~ 01 01 13 27 39 27 1~ ~~ co c~ ~ 01250 01 ·27 01 02 01 01 01 50 01 01 01 01 50 01260 40 50 41 27 28 15 01 26 01 01 ~~ co 40 01 01270 50 20 22 24 49 03 50 47 ~:4 03 1~ ~ 10 u 42 01280 49 21 22 14 49 49 0~ -~~ 05 25 22 21 17 41 01290 24 22 22 18 26 27 21 21 24 ~0 ~~ 35 18 36 01300 49 50 21 10 ~ 35 4'? 46 47 01 22 ~~ ~~ ~Q ~~ ~~ ~~ 0131 0 40 ~ 25 21 18 26 37 48 48 23 22 34 40 u 36 01320 24 02 02 02 02 02 03 01 ~~ c~ 02 01 01 44 01330 26 24 22 19 02 03 02 01 23 21 01 01 04 01340 30 21 ~~ ~~ 31 40 30 10 18 34 32 15 ~~ ~~ 39 01350 31 0~ ~~ 13 04 17 26 43 40 08 41 ~~ co 19 32 01360 26 25 36 36 24 18 06 13 34 25 24 10 ~ 24 01370 15 30 05 24 1 0 10 ~ 1n 0 40 04 40 20 1 0 32 01380 26 30 36 4~ ~ ~0 ~~ 24 05 14 42 ~= ~~ 24 31 ~~ c~ 01390 93 36 03 04 07 07 24 46 03 36 22 03 ~~ ~I 01400 29 26 30 17 1- ( OQ ~~ 2:3 26 19 28 24 0~ ~~ -.~ c~ 0141 0 00 00 00 00 00 00 00 00 00 00 00 00 00 01420 45 26 00 00 24 00 47 00 00 24 00 ~~ c~ 46 01430 00 00 00 00 00 00 00 00 00 00 00 00 00 01440 00 00 2? ~~ cr 00 24 00 00 26 00 00 00 00 01450 00 00 00 00 00 00 00 00 00 00 00 00 00 Ot460 00 00 00 00 00 00 00 44 00 00 25 00 00 01470 ~ 00 00 00 00 00 00 00 00 00 00 00 00 174 PHASE 1 PROPERTY FITTING (CONTINUED) 014:::0 1 ·=· ·-· 25 39 24 ·-·= t:_._l 1 0 1 ·:· ·-· 18 :2:6 .-.~ .,:. ( ~~:2 21 24 01491) 21 o::: (I (I or:. 0-;" 24 45 44 00 24 (I (I I)!) 00 01500 :~:2 24 00 30 00 00 ·-=··=- ·-''-- 44 (I (I (II) 24 1.-, Co 00 0151 0 2~;: 24 04 1 .-. ·=· 04 .-.~ C.f 05 :~:7 o:::. (14 24 22 1""=' ·-· 01520 2::: 4.-, Co 00 .,..,~ . .;.,. 00 24 19 06 (II) 00 00 00 00 015:2:0 o:::: 02 03 04 05 1):2: o::: 46 04 1):2: 00 1 .-. ·=· 2:3 01540 ·-:··:· ·-··-· 24 25 .-.. - C. I:• 00 •J"':• ~--· 24 14 I) (I I) (I 00 00 00 01550 49 01 01 01 02 49 02 02 49 02 02 49 02 015E.O 4'3 49 4'31 49 4'31 02 49 4'? 02 4'? 4'31 02· 4'31 01570 01 01 01 01 02 49 49 02 02 02 02 02 49 015:::0 4C! -· 4'3 4'3 49 49 02 49 49 02 02 49 02 02 01590 01 4'3 01 4'? oE· 4'3 02 02 4'? 49 4C! -· 4':;:. 02 01E.OO 01 01 01 01 02 02 02 4'31 02 02 02 02 02 01E·1 0 01 02 01 01 02 49 02 49 4'? 49 02 02 02 01 E.20 09 42 40 ·-· 45 :34 2:::: 0:2: 41 4'::0 ·-· 4'3 :~::::: 44 47 016:2:0 ::::9 2::=:-. 1.-, Co o:::: 1.-, c :~;:~: 4:3 4E. 09 01 22 2:.:: 2:;: 01640 34 45 34 :=:o 4.-. -.:· 2:~: 47 17 35 41) .-.. -. .:..:.· :32 30 01650 2:~: 15 1 0 07 25 .-.c:" C.·-· 4E. 25 07 04 2:~: 1 (I 1 I) 01660 ~:4 4E. :;:::: :~:2 44 .-.-. c....:· 05 25 :37 -.c:- .,;..._. 3? 27 24 01E.70 02 .-.~ c_._. 04 02 0:2: 30 26 :37 01 01 24 01 1 0 016:::0 26 41 or:. 03 04 14 o::: 24 02 01 1.-, Co - .... C.·-· 24 01690 1.-, ·-=· 09 20 05 22 1 ·:o ·-· 01 09 14 19 05 20 15 01700 27 02 09 12 :3::: 1 0 01 02 14 ~:2 4·:· ·-· 1 ':• ._. 02 01710 20 06 40 .-.~ ·:··-=· 21 07 15 09 -.-. 24 ~:~: 02 C·-• ·-··-· c:..:.· 01720 07 05 44 44 40 06 02 ::=:7 14 36 48 40 o·31 01730 13 09 :2: I) 1 .-. •:0 ·-=··-=· ·-·-· 1-. Co 05 20 22 22 2E. 24 o·3 01740 02 41 44 02 03 02 01 01 4!3 46 2:9 02 02 01750 1 0 4::: :31 1 1 1 E. 1·:· 02 01 42 ~=1 04 06 - .... ·-· . .; . .._. 01 ?E. 0 1·:· ·-· 46 27 0? 1.-, c 42 o::: 1.-. Co :;: 1 l3 1 ·;. :;:o 1 0 01770 1? 10 2!:: 07 12 .-.. -. 44 42 23 4"""' .-. ... 24 42 c.•:• Co ~··-· 017:::0 :;::3 ~35 1'3 44 1 E. 27 2:;: 25 ,..,.-. ~·C. o:=: :;:o 1? 0? 01790 15 .-,.-, 27 o::: :~::;: 20 24 .-.r::: 32. -':•·=- ~:4 17 ·-:-·::. .::,..:,. .:, .. _ . ._ .. _. ·-··-· (11:300 15 ·?-==' 1'3 42 12 :31 o·3 07 ·:··-=- 1)9 :32 ~:2- o·31 ·-·· ~-· 0"1 :31 0 01 04 3:;: 02 :;:::: ~ .- .::. '=· ~:6 4? 2'3 40 1 I) 22 o·3 01820 44 41 -·-=· :2:4 2C! 28 42 3:3 3::: 40 38 ":•·=- 41 .:_.._, -· ·-•&:;;;. 01 ::::=:(I 1 1 24 .-.co 27 2? .-..-. OE.: 01 47 1 ~: -~--. 46 05 .::;._. c.·.? .::; . .:. 01:=:40 40 .-.~ C.·-' 2E. 17 2:=: .-,c: ~·-· 4? 4? 25 28 24 2E· 47 01850 25 2E. 27 27 2'3 2E· 25 -.c:- C·-' 27 27 24 2E. 25 1)1 :::.; (I o4 06 o::: 17 1-· .:,. 1 I) 2E. 47 27 4E. 05 06 27 01:::7 0 2E. 2€· 27 4? 2::: :3::: 26 --·=- 27 27 4·=· 41 07 .::;._. ·-· 01 ::::::0 05 05 27 2E. 2'3 .-.r:: C.·-· OE. 45 2:3 -=-'? ..... 25 27 o·3 o1:::·=..o 27 25 27 26 :30 .-,c- C·-' 2::: 26 -,.-. .-=.·=· 2::: 25 .. -.c- C·-• 0'3 01900 4::: :3·=.- 47 4? :31 4:=: 04 03 ·-·C" .,: .. _. 04 07 05 12 01'31 0 2E. 3::: 25 04 :2:1 05 40 49 07 4::: 45 4:3 4·=· ·-· 01920 2E· 07 ~==== 4'.:0 ·-· 22 4? 26 25 48 40 27 07 2E· 019:30 0:3 0? 02 04 1):2: 05 o:=: 4':;. OE. 48 :.::•S 1)7 1 ·:o ·-· 01 '340 4? 2E. 4..? 48 :~:·;. 4'=· ·-· 20 06 49 25 46 4':;:. 25 01'?50 0:3 47 0:3 0:3 1'? :3::: (1:2: 4·=· ·-· 0:3 49 4? 1):2: 25 01 '?E. 0 03 06 24 04 4:• 20 2:~: 0:3 04 24 .-.co 2S• 02 ·-· c_._. 01'?70 14 t·-=· ·-· 41 :31 1 1 02 02 1)1 07 30 (1:.::: 12 01 175 PHASE 1 PROPERTY FITTING (CONTINUED) 01 '?:::: 0 40 :~:::: 41 .-.c:- 42 4:3 4'~ ·-·=- :~:::: :32 4E. 1·=· 46 . ,: .. _. .; .. _ . ·-· 019'~0 14 45 1.-. .::. 4'3' 24 ·~::3 16 01 14 1 1 02 41 02 02000 06 :3•3 44 4E. ~;:·:;. 4'3' ·-:·7 ._., 40 50 40 50 47 50 02010 12 44 1 0 4'3' o·;. 1 .-. . .::. 27 1)1 I)? ·=-C" ~·-· 01 4:=: 01 02020 04 ::::o 42 o:=: 01 4'3' 01 01 50 01 24 25 47 02o::::o 1E. 4'3' 06 1 .-, .:,. 1 0 01 01 01 01 34 t·-=· ·-· 14 o:;: 02040 27 20 .-.~ C.f o:::: 01 24 01 25 01 1)1 4'3' 50 01 02050 :3? 22 01 50 01 01 50 .-.. -. . :,e, 01 40 40 01 2E . 02060 3E. 50 ·":•7 2'=-' 34 27 .-,~::: 22: 50 24 4'~ 50 .-,~ ._., .::_._. -· ~--· 02070 r:r::: 24 15 14. 1·=· 01 50 15 50 4q 01 01 50 ·-· -· o2o:::o ~::~: 4';. 50 4'3' 14 50 01 01 50 1'~ 4'~ 50 15 020'3'0 02 24 o:::: 01 01 01 50 02 01 50 01 01 02 02100 02 02 02 02 01 01 01 01 01 01 01 01 01 END OF DATA 176 PART 3: PHASE 1 BACKGROUND QUESTIONS 00010 2 1 41 26 1 1 03 20 1 1 10 07 1 2 31 41 1 2 32 49 38 01 31 33 23 25 00020 00030 00040 00050 00060 00070 00080 05 08 19 01 23 10 43 13 50 32 38 11 33 13 43 42 30 22 28 11 1 2 26 43 29 28 1 1 35 41 32 12 25 38 43 24 14 40 34 13 33 41 14 14 2 1 25 25 25 25 29 04 25 03 2 48 32 25 25 25 25 07 23 00090 1 00100 2 00110 1 2 1 24 27 24 25 24 02 22 01 15 31 2 2 40 34 33 14 13 32 32 17 32 29 15 31 38 15 00120 00130 2 1 00140 1 2 00150 2 1 25 34 27 21 18 11 22 07 09 33 28 12 38 34 10 22 35 37 26 14 10 46 08 11 00160 1 1 22 21 00170 2 1 50 50 12 22 01 01 2~ 50 12 12 33 35 35 01 26 50 42 26 22 19 40 01 00180 2 2 00190 1 1 08 27 18 00200 2 1 24 22 15 00210 2 1 25 24 25 00220 2 1 50 31 21 00230 1 2 45 38 25 06 12 32 07 17 14 11 37 46 08 05 11 45 43 14 22 09 38 46 11 20 04 30 48 25 00240 00250 00260 00270 00280 00290 00300 2 1 24 24 31 03 04 46 34 08 1 1 49 42 31 01 20 47 40 05 2 1 41 29 25 17 32 45 17 14 2 1 25 26 27 26 24 48 26 48 1 1 25 37 35 09 40 44 13 04 2 1 32 38 35 30 14 41 40 14 1 1 50 23 24 01 ENn OF DRTR 12 50 50 13 177 APPENDIX D Phase 2 Scenarios Part 1: High and low dimensions for each scenario. Part 2: Reporting position questions for each scenario. 178 APPENDIX D Part 1: Listed below are the elements of the phase two tax scenarios. Each subject was presented one topic A, B, C, D scenario. The scenarios were constructed according to the 1/2 factorial design in Figure 5 (see chapter 3). Consequently, every scenario had either a high or low complexity level on each dimension (but never both). Also, every scenario consisted of all four dimensions and only the four dimensions. TOPIC A - CHARITABLE DEDUCTION DIMENSION D1 PERSONAL/FINANCIAL TOPIC D1 Low Wilson, a mechanic, donated a van that he had thoroughly renovated to his church. This gift entitled Wilson and his family to become life members of the church and thus eligible to attend all concerts, pageants, retreats, and camps free of charge. D1 High Wilson has set up a charitable remainder annuity trust for the benefit of his church. This gift entitled Wilson and his family to become life members of the church and thus eligible to attend all religious concerts, pageants, retreats and camps free of charge. DIMENSION D2 QUANTITATIVENESS D2 Low Appraisals on the donation averaged about $2,000. The Wilson family received a life membership because any gift with a value over $1,000 confers this special status. ($1,000 is the present value of receiving free family admission to all special events). D2 High Three appraisers valued the donation from $1,600 to $2,400, with the mean and median value computed to be $2,000. Because the value was over $1,000, the Wilson family received a life membership in the church ($1,000 is the present value of receiving free family admission to all special events). 179 DIMENSION D3 SOCIAL JUSTICE TOPIC A CONTINUED D3 Low Because of the clearly religious nature of its activities, ~ilson's church is classified as a tax exempt organization and therefore contributions are tax deductible. D3 High Even though Wilson's church does not have regular congregation meetings, or a minister, its tax exempt status has not been challenged by the IRS, therefore contributions are deductible. DIMENSION D4 READABILITY D4 Low After a review of his own tax situation, the nature of the gift, and the use of the gift by the church, Wilson determined that none of the possible charitable deduction limits applied to his gift. However, Wilson did find out that his deduction should be reduced by the value of benefits received as a result of the donation. D4 High Section 170(e) of the Internal Revenue Code, which was addec by the Tax Reform Act of 1969, imposes limitations on the value that can be deducted for contributions of specific items of appreciated property falling within certain categories. The classes of property and affected donees arE these: 1. "Ordinary income" property regardless of the kind of charitable donee. 2. Tangible personal property if the donees use is unrelatec to its charitable function. 3. "Capital gain" property given to certain private foundations. After a review of his own tax situation, the nature of the gift, and the use of the gift by the church, Wilson determined that none of the possible charitable deduction limits applied to his gift. However, Wilson did find out that his deduction should be reduced by the value of benefits received as a result of the donation. 180 TOPIC B INCOME OR GIFT? DIMENSION D1 PERSONAL/FINANCIAL TOPIC D1 Low Warren is a college student that spends several hours a week helping out an elderly neighbor by running errands, cleaning, and taking the gentleman to the doctor. Each month Warren receives a check from the man for $600. D1 High ~arren is a college student that spends several hours a week helping out in the school research lab by running errands, cleaning, and giving directions to freshmen students. Every month Warren receives a check out of a lab miscellaneous fund for $600. DIMENSION D2 QUANTITATIVENESS D2 Low Warren figures that he does half the amount of work that he would have to do anywhere else for the same amount of money. D2 High Warren figures he only does about $300 worth of services a month ($120 worth of errands, $85 cleaning, and $95 miscellaneous). DIMENSION D3 SOCIAL JUSTICE D3 Low Warren really enjoys being able to help out and would continue the same activities regardless of whether he received a monthly check. D3 High Warren figures that his situation is better than most students who have part time jobs. For one thing, Warren notes that taxes are not taken out of his check. 181 DIMENSION D4 READABILITY TOPIC B CONTINUED D4 Low According to the tax law, some amounts received are taxable while others may be tax exempt. For students, two important areas of tax exempt income are gifts and scholarship grants. The logic governing taxability in both areas is the same. The general rule is that grates funds are tax free, but funds received as compensation for services rendered are taxable to the recipient. D4 High According to the Internal Revenue Code Sec. 61 " ••. gross income means all income from whatever source derived", and should be included for tax purposes unless it is specifically excluded under the law. For students, two important areas of tax exempt income are Section 102 on gifts and Section 117 on scholarship grants. Section 102 amounts are generally tax free if it is the donar's intent to give a gift. Section 117 amounts are generally excludable from gross income although any portion received that represents payment for services is taxable unless the services are of the type required of all students as a condition of receiving their degree. 182 TOPIC C TRAVEL EXPENSES DIMENSION D1 PERSONAL/FINANCIAL TOPIC D1 Low Rogers lives in Minnesota but owns a home in Newport Beach, California which he rents out to life-long close friends of his family at fair market value so that they can enjoy a temperate climate during the winter. Around Christmas, Rogers arranges a trip to California to inspect and repair the property and also to visit his friends. D1 High Rogers lives in Minnesota but owns a home in Newport Beach, California which he rents out. Rogers bought the Newport Beach home on speculation that housing prices on the California coast will appreciate significantly in the next few years. During the winter, Rogers arranges a trip to California to inspect and repair the property and also to visit friends and go sailing. DIMENSION D2 QUANTITATIVENESS D2 Low Rogers trip to California cost a total of $1,400 for his Thursday through Monday stay. This total includes charges for air fare, car rental, meals and lodging for 5 days. Rogers figures that the primary purpose of his trip was !investment related and that the majority of his days were 1 spent on business rather than personal matters. D2 High Rogers trip to California cost a total of $1,400 ($800 air fare, $180 car rental, $200 meals and $220 lodging). He stayed a total of 5 days and figures the 3 weekdays of his I visit (Thursday, Friday and Monday) were primarily devoted t to investment related matters. Rogers is in the 50% marginal tax bracket, so a $1 deduction results in a 50 cen tax savings. I I 183 DIMENSION D3 SOCIAL JUSTICE TOPIC C CONTINUED D3 Low Rogers initially purchased the house in anticipation of providing a home for his own retirement in the next 5 to 10 years. Thus he is very interested in maintaining the appearance of the property and contributing to the betterment of the neighborhood and community. D3 High Rogers initially purchased the house because he was able to acquire the property with a minimal cash outlay. In addition, the depreciation and interest expense on the house considerably reduce the federal income taxes that Rogers would otherwise owe on his salary income each year. DIMENSION D4 READABILITY D4 Low The IRS allows a deduction for all travel costs as long as the primary purpose of the trip is investment related and the majority of the days during the trip are spent on investment-related, rather than personal matters. If the primary purpose is personal or if the majority of days are spent on personal matters, then the expenses are not deductible. D4 High Section 212 of the Internal Revenue Code allows as a deduction ordinary and necessary expenses paid or incurred: 1. For the production or collection of income, and 2. For the management, conservation, or maintenance of property held for the production of income. Regulation 1.212-1(d) stresses that " ... expenses must be reasonable in amount and must bear a reasonable and proximate relation to the production or collection of taxable income or to the management, conservation, or maintenance of property ... ". In applying this law, the IRS generally allows the deduction of all travel costs as long as the primary purpose of the trip is investment related anc the majority of the days·during the trip are spent on investment-related, rather than personal, matters. 184 TOPIC D REHABILITATION CREDIT DIMENSION D1 PERSONAL/FINANCIAL TOPIC D1 Low Carson retired after years of working and decided to open an antique shop in town with his wife. After a thorough search, Carson discovered an old building in need of some internal repair, which he purchased in August 1981. D1 High Carson, a real estate speculator, acquired an old but basically sound business building in August 1981 as a part of a tax-free exchange of property. DIMENSION D2 quantitativness D2 Low During September, Carson solicited contractor's bids. Even though the structure was estimated to be about 38 years old, none of the exterior walls needed to be replaced, so Carson found that the building could be restored for a total of $10,000. Construction began in early October 1981 and was completed within six months. D2 High During September, Carson solicited contractor's bids. Even though the structure was approximately 38 years old, estimates for the needed renovation were $1,500 plumbing, $2,750 electrical, $1,800 masonry, $1,250 glass work, and $2,700 general carpentry for a total of $10,000. Construction began in October 1981 and was completed within the six month period ending March 1982. 185 DIMENSION D3 TOPIC D CONTINUED 1 SOCIAL JUSTICE D3 Low I To encourage the preservation of American heritage, Congress has written certain tax incentives into the law. Actually, Carson would have purchased and restored the building regardless of the tax credits because of his appreciation and respect for older architecture and craftsmanship. D3 High Rehabilitation of older structures has become one of the latest tax shelter areas because of the amount of tax credi available. Aware of tax law changes between 1981 and 1982 that increased benefits even further, Carson withheld payments to contractors until January 1,1982. DIMENSION D4 READABILITY D4 Low Congress changed the tax law for years beginning in 1982 to allow a larger credit for rehabilatating older structures. For buildings at least 30 years old, a 15% credit is allowed, for 40 year old buildings, a 20% credit (pre 1982, the credit was limited to 10% in both cases). Carson is a cash basis taxpayer. He wrote all $10,000 worth of checks t contractors in 1982, and is not sure of the exact age of th building. Consequently, he wonders what amount of credit h should take for rehabilitating the structure. D4 High A separate percentage for qualified rehabilitation expenditures applies for costs incurred after 1981. The rehabilitation credit is 15% for structures that are at least 30 years old, 20% for structures at least 40 years old, and 25% for certified historic structures. (Prior to 1982, only the 10% investment tax credit was available for qualified rehabilitation expenditures). Rehabilitation expenditures qualify only for real property if made in connection with a substantial rehabilitation of the building. According to Internal Revenue Code Section 48 (g) (1) (C) (i), a rehabilitation is substantial if expenditures during the 24 months ending on the last day of the tax year exceed $5,000. Carson is a cash basis taxpayer, wrote all $10,000 worth of checks to contractors 186 in 1982, and is not sure of the exact age of the building. Consequently, he wonders what amount of credit he should take for rehabilitating the structure. 187 Part 2: Listed below are the directions given the subjects for each of the four (A - D) tax scenarios. FOR CASE A - CHARITABLE DEDUCTION PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Wilson, what amount would you take as a charitable deduction on your Federal income tax return? .... --+---+---+----+-----+-----+-----+-----+-----·-----·---~ t----T-- $0 250 500 750 1000 1250 1500 1750 2000 2250 2500 OVER $2500 If Wilson sought you advice, what amount would you suggest he take as a charitable deduction? .... --+---+---·----+-----T-----+-----+-----T-----+-----+---4 ~--+-- $0 250 500 750 1000 1250 1500 1750 2000 2250 2500 OVER $250C If you were to be completely objective, what amount do you think a person should deduct under the current tax law? .... --·---+---+----+-----+-----+-----+-----1-----+-----+---~ r--+-- $0 250 500 750 1000 1250 1500 1750 2000 2250 2500 OVER $250C FOR CASE B - INCOME/GIFT PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Warren, what amount of income would you include in your Federal tax return? .... -------+---------+---------+---------+---------+--------~- $0 100 200 300 400 500 $60( If warren sought you advice, what amount of income would yol suggest he include in his Federal tax return? .... --------+---------+---------+---------+---------+-------+ $0 100 200 300 400 500 If you were to be completely objective, what amount of income do you think a person should include under the current tax law? $ 60( .... --------+---------+---------+---------+---------+-------+ $0 100 200 300 400 500 $60( 188 FOR CASE C - TRAVEL EXPENSE PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Rogers, what amount of travel expense would you deduct on your Federal tax return? .... -- .... ---+---+--- .... ----t-----1-----1-----1-----1-----1-----; t--1--- $0 200 400 600 800 1000 1200 1400 1600 1800 2000 OVER $2000 If Rogers sought you advice, what amount of travel expense would you suggest he deduct on his Federal tax return? t---+---+---+---+----l-----i-----l-----l-----l-----l---1 ~-~-- $0 200 400 600 800 1000 1200 1400 1600 1800 2000 OVER $2000 If you were to be completely objective, what amount of travel expense do you think a person should deduct under thE current tax law? .... --+---+---+---+----1-----I-----I-----1-----1-----1--- -1 t--+-- $0 200 400 600 800 1000 1200 1400 1600 1800 2000 OVER $2000 FOR CASE D - REHABILITATION CREDIT PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Carson, how much "rehabilitation of an older structure" credit would you take on your Federal income tax return? t---+---+---+---~-----l-----1-----i-----i-----t-----t---~ f--+-- $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 OVER $350C If Carson sought you advice, how much credit would you suggest that he take for the 1982 tax year? t---+---+---+---~-----1-----l-----i-----i-----i-----l---~ ~--+- $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 OVER $350C If you were to be completely objective, what amount of credit do you think a person should take for the year under the current tax law? .... - -+----+--- ... ----1-----1-----t-----1-----1-----1-----1--- _., }----i- $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 OVER $350( 189 APPENDIX E ~ample Phase 2 Test Instrument and Background Questionaire ~his sample consists of the following configuration: Case A: D1 Low D2 Low D3 Low D4 High Case B: D1 Low D2 High D3 High D4 High Case C: D1 High D2 Low D3 High D4 High Case D: D1 High D2 High D3 High D4 Low This sample configuration depicts the following research designs from Figure 5: 1,4,6 and 8 respectively. For the actual testing of the 86 subjects the ordering of the cases (A through D) was randomized. In addition each subject was presented a different research design in each of the four cases (i.e., a subject was never presented cases A and B which both had research design #1). 190 In the following cases you may want to have more information before making a judgement. However, the objective of this study is to ascertain the effect of different amounts of information on tax position choice. S please try to make the best judgement you can, based on the information available in each case. DIRECTIONS: For each question, place an X on the line to indicate your choice. Feel free to mark anywhere on the line. Two examples are given below. In the first example, the ! taxpayer feels that $1,700 should be deducted. In the second example, the taxpayer feels slightly more familiar than unfamiliar with the tax situation. 1----+-----1-----1------1------1-----1-----1-----1-----1-----1--i 1$0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 + ~-j- $5000 I Very Unfamiliar------------------1 Very Familiar (Five inch line not to scale for this appendix example) All questionnaires are anonymous. Your responses will not be traced to you or associated with you in any manner, so please feel free to respond to the questions as you consider appropriate. 191 Wilson, a mechanic, donated a van that he had throughly rennovated to his church. This gift entitled Wilson and his family to become life members of the church and thus eligible to attend all concerts, pagents, retreats, and camps free of charge. Appraisals on the donation averaged about $2,000. The Wilson family received a life membership because any gift with a value over $1,000 confers this special status. ($1,000 is the present value of receiving free family admission to all special events). Because of the clearly religious nature of its activities, Wilson's church is classified as a tax exempt organization and therefore contributions are tax deductible. Section 170(e) of the Internal Revenue Code, which was addec by the Tax Reform Act of 1969, imposes limitations on the value that can be deducted for contributions of specific items of appreciated property falling within certain categories. The classes of property and affected donees arE these: 1. "Ordinary income" property regardless of the kind of charitable donee. 2. Tangible personal property if the donees use is unrelatec to its charitable function. 3. "Capital gain" property given to certain private foundations. After a review of his own tax situation, the nature of the gift, and the use of the gift by the church, Wilson determined that none of the possible charitable deduction limits applied to his gift. However, Wilson did find out that his deduction should be reduced by the value of benefits received as a result of the donation. PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Wilson, what amount would you take as a charitable deduction on your Federal income tax return? ..... --+---+---+----1----_,_-----t-----1-----1-----1-----t-- -i ~--1-- $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 + $3500 If Wilson sought your advice, what amount would you suggest he take as a charitable deduction? ..... --+---+---+----1-----1-----1-----1-----1-----r-----1---; r--1--- $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 + $3500 If you were to be completely objective, what amount do you think a person should deduct under the current tax law? 1---+---+---~-----t-----l-----t-----l-----l------t-----t--~ t----t-- $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 + $3500 192 arren is a college student that spends severa ours a wee elping out an elderly neighbor by running errands, leaning, and taking the gentleman to the doctor. Each onth Warren receives a check from the man for $600. arren figures he only does about $300 worth of services a month ($120 worth of errands, $85 cleaning, and $95 miscellaneous). arren figures that his situation is better than most students who have part time jobs. For one thing, Warren notes that taxes are not taken out of his check. According to the Internal Revenue Code Sec. 61 " ... gross income means all income from whatever source derived", and should be included for tax purposes unless it is specifically excluded under the law. For students, two important areas of tax exempt income are Section 102 on gifts and Section 117 on scholarship grants. Section 102 amounts are generally tax free if it is the donar•s intent to give a gift. Section 117 amounts are generally excludable from gross income although any portion received that represents payment for services is taxable unless the !services are of the type required of all students as a condition of receiving their degree. PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Warren, what amount of income would you include in your Federal tax return? ~--------+-------+-------+-------+----------+---------~ $0 100 200 300 400 500 $600 If Warren sought your advice, what amount of income would you suggest he include in his Federal tax return? ~-------+--------+--------+--------+--------+---------; $0 100 200 300 400 500 $600 If you were to be completely objective, what amount of income do you think a person should include under the current tax law? ~-------+--------+--------+--------+--------+---------~ $0 100 200 300 400 500 $600 193 Rogers lives in Minnesota but owns a horne in Newport Beach, California which he rents out. Rogers bought the Newport Beach horne on speculation that housing prices on the California coast will appreciate significantly in the next few years. During the winter, Rogers arranges a trip to California to inspect and repair the property and also to visit friends and go sailing. Rogers trip to California cost a total of $1,400 for his Thursday through Monday stay. This total includes charges for air fare, car rental, meals and lodging for 5 days. Rogers figures that the primary purpose of his trip was investment related and that the majority of his days were spent on business rather than personal matters. Rogers initially purchased the house because he was able to acquire the property with a minimal cash outlay. In addition, the depreciation and interest expense on the house considerably reduce the federal income taxes that Rogers would otherwise owe on his salary income each year. Section 212 of the Internal Revenue Code allows as a deduction ordinary and necessary expenses paid or incurred: 1. For the production or collection of income, and 2. For the management, conservation, or maintenance of property held for the production of income. Regulation 1.212-1(d) stresses that " ... expenses must be reasonable in amount and must bear a reasonable and proximate relation to the production or collection of taxable income or to the management, conservation, or maintenance of property ... ". In applying this law, the IRS generally allows the deduction of all travel costs as long as the primary purpose of the trip is investment related anc the majority of the days during the trip are spent on investment-related, rather than personal, matters. PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Rogers, what amount of travel expense would you deduct on your Federal tax return? ~ --+ ---+---+--- + ----1---- _, __ ---·---- ...,._-----1-----1-----; ,_--1-- $0 200 400 600 800 1000 1200 1400 1600 1800 2000 OVER $2000 If Rogers sought your advice, what amount of travel expense would you suggest he deduct on his Federal tax return? ~--+---+---+---+----~----t-----~---~-----1-----~---~ ~--r- $0 200 400 600 800 1000 1200 1400 1600 1800 2000 OVER $2000 If you were to be completely objective, what amount of travel expense do you think a person should deduct under thE current tax law? ~---l---~---~---+----l-----1----~-----~----~----l----~ ~--~- $0 200 400 600 800 1000 1200 1400 1600 1800 2000 OVER $2000 194 Carson, a real estate speculator, acquired an old but basically sound business building in August 1981 as a part of a tax-free exchange of property. During September, Carson solicited contractor's bids. Even though the structure was approximately 38 years old, estimates for the needed renovation were $1,500 plumbing, $2,750 electrical, $1,800 masonry, $1,250 glass work, and $2,700 general carpentry for a total of $10,000. Construction began in October 1981 and was completed within the six month period ending March 1982. Rehabilitation of older structures has become one of the latest tax shelter areas because of the amount of tax credit available. Aware of tax law changes between 1981 and 1982 that increased benefits even further, Carson withheld payments to contractors until January 1,1982. Congress changed the tax law for years beginning in 1982 to allow a larger credit for rehabilatating older structures. For buildings atleast 30 years old, a 15% credit is allowed, for 40 year old buildings, a 20% credit (pre 1982, the credit was limited to 10% in both cases). Carson is a cash basis taxpayer. He wrote all $10,000 worth of checks to contractors in 1982, and is not sure of the exact age of thE building. Consequently, he wonders what amount of credit hE should take for rehabilitating the structure. PLEASE PLACE AN X ON EACH LINE TO INDICATE YOUR CHOICE If you were Carson, how much "rehabilitation of an older structure" credit would you take on your Federal income tax return? I---+----+---+----·-----·-----!-----1-----1-----1-----1-- _, $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 If Carson sought your advice, how much credit would suggest that he take for the 1982 tax year? ,_ --+----+---+----·-----t-----1-----1-----1:-----1-----\--~ $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 + f---;- + $3500 you f---t- $3500 If you were to be completely objective, what amount of credit do you think a person should take for the year under the current tax law? J---+----t----+----1-----1-----1-----1-----f-----1-----1:--i $0 250 500 750 1000 1250 2000 2250 2500 3000 3500 f--1- + $3500 195 GENERAL BACKGROUND QUESTIONS 1. Please circle your sex: MALE FEMALE 2. Do you usually prepare your own tax return? Do you usually file a long (1040) form? y y N N Please mark an X on the line where it best expresses your feelings: 3. In terms of your earnings life cycle, how do you place yourself: NO. OF YRS. WORKED 0 5 10 15 20 25 30 35 40 45 50 OR MORE 4. Please indicate your education: LESS THAN GRADE 8 9 10 11 12 YEARS COMPLETED OF HIGH SCHOOL 1 2 3 4 5 6 OR MORE YEARS OF COLLEGE COMPLETED 5. Please indicate your (and your spouse, if applicable) average annual income: (in thousands of dollars) $0 10 20 30 40 50 60 70 80 90 100 OR MORE 6 • How do you rate the general fairness of the income tax laws? (Note: "5 inch line" not to scale for this example) Very Unfair Very Fai 7. If a person wanted to evade income taxes, what percentag do you think they could evade and not get caught? None Al 8. In termes of complexity, how do you rate the present income tax system? Very Simple ~------------------------------------------~Extremel Complex 9. Of the people you know, what percentage do you think pay all their legally owed income taxes? No One ~------------------------------------------~ Everyon 10. The percentage of taxpayers' money used wisely by the federal government is: None Al 196 APPENDIX F Phase 2 Data Part 1: Phase 2 tax reporting position selections for cases A - D. Data is presented by case, and in the following order: subject number, subject's responsE for the amount "you" would report, for the amount you would ''advise", and the amount you think is "correct''. The last sequence of numbers for each subject indicate whether the dimensions were high (coded with a 2) or low (coded with a 1) and the corresponding design number (see Figure 5 in the text for details of the design number) . Part 2: Data corresponding to the 10 background questions for each of the 86 subjects. 197 PHASE 2 JUDGMENTS ON TAX CASE A (CHARITABLE DEDUCTION SCENARIO) SUBJ YOU ADV CORR DESIGN 00010 1000 1000 1 2212 7 00020 2000 2000 2000 1222 4 00030 1000 1000 250 1121 2 00040 1000 1000 1 1121 2 00050 2000 2400 2000 2221 8. 00060 1000 1000 1000 2122 6 00070 1800.2100 3000 2122 6 00080 2000 2000 1000 2122 6 00090 1000 1000 1000 1121 2 00100 2000 2000 2000 1121 2 00110 2125 2125 2375 2212 7 00120 1000 1000 1000 1222 4 00130 2000 2000 1000 1211 3 00140 1600 600 1600 1211 :;: 00150 2000 2000 2000 2221 8 00160 1 1 1700 1211 3 00170 2000 2000 1000 2221 8 00180 2000 2000 2000 2111 5 00190 2000 2000 2000 2122 6 00200 2000 2000 2000 2111 5 00210 1400 1400 1000 2221 8 00220 1900 900 700 2111 5 00230 2000 2000 1400 2212 7 00240 1100 1100 1000 2221 8 00250 1000 1000 1000 1121 2 00260 2000 2000 2000 1222 4 00270 2000 2000 2000 1121 2 00280 ~~0 990 990 1222 4 00290 1000 1000 1000 1112 1 00300 2300 2000 2300 1222 4 00310 1000 1000 2000 1112 1 00320 2000 2000 2000 1112 1 00330 2000 2000 2000 1211 3 00340 1000 1000 1000 2212 7 00350 2000 2000 2000 1211 3 003E.O 1240 1250 1250 1112 1 00370 2000 2000.2000 2221 8 00380 1000 1000 1000 2111 5 .o o:;:·:;e o 1 o o o 10 o o 1 o o o 1112 1 00400 2000 2000 2000 1222 4 00410 1750 1750 1?50 2111 5 00420 2000 2000 2000 1121 2 00430 50 50 50 1112 1 00440 2000 1850 2000 2111 5 00450 2000 2000 2000 2122 6 00460 1300 1300 1000 2212 ? (I 04 7 0 1 ::;: I) 0 . 1 :;: 0 (I 1 ~::: 0 0 1211 :~: SUBJ YOU ADV CORR DESIGN 00480 2000 2000 2000 2221 8 00490 2000 1000 1000 2212 7 00500 2000 2000 2000 1112 1 00510 2000 2000 1500 1222 4 00520 1000 1000 1000 2122 6 00530 2000 2000 1000 1121 2 00540 1000 1000 1000 1211 3 00550 2400 2000 !000 2221 8 00560 2000 2000 1000 1121 2 00570 1000 1000 1000 2212 7 00580 600 850 850 2212 7 00590 2000 2000 2000 2111 5 00600 2000 2000 2000 2221 8 00610 75 90 80 1112 1 00620 2000 2000 2000 1121 2 00630 1950 1950 1950 2212 7 00640. 650 700 650 1121 2 00650 1100 1100 1100 2221 8 00660 1000 1000 1000 2221 8 00670 125 125 125 1211 3 00680 1000 1000 1000 1112 1 00690 1000 1000 1000 2111 5 00700 1800 1800 1900 1211 3 00710 1000 1000 1 2122 6 00720 1000 1000 1000 2122 6 00730 1000 1000 1000 1222 4 00740 1900 1900 1900 1211 3 00750 1500 1500 1500 1211 3 00760 1050 1050 1300 1112 1 00770 2000 2000 1000 1222 4 00780 1000 1000 1000 11~1 2 00790 550 550 550 2111 5 00800 1100 1100 1100 2221 8 00810 1550 1550 600 2212 7 00820 2000 2000 2000 1112 1 00830 2000 1000 1000 2111 5 00840 2000 2000 1000 2111 5 00850 50 50 2100 2122 6 00860 2000 2000 2000 2212 7 END OF DATA . 198 CCI ~ U) o::e uo H xo:: o::Co::C E-IZ ILl zu OUl UlE-1 E-1 (Lo ZH 1Lll9 ?.;-...... ~ILl O::E: :::>0 t-:lU z NH ILl U) o::C :r:: ~ Z~~~M~-~ru~~--~ru~M~~~NMM~~~--OOMOOMru~ooru~~ru- l9 Hru----ruru-rurururu--ru--ruru---rururururu-----ru--ruru-ru Ulru-----ruru-ru---ru-----ru--rururu--ru-ru-rururururururu wru--ru----ruru----ruru-ruru-rururururu--rurururu--ru---- o-ruru-ru-ru-ru---ru-ru-rururu--------ru-ru--ruru-ruru-- O::oo-oo-ooooooooooo-ooooooooooo-o~6ooooo-~· O::oo ~~ oooo~OMOO~~ ~-oooooo~o~ o~~~~ooo ~ 0~~ ~~ ~~~~M~ ~~~M ~-~M~M~M~~~ ~~ ~~~~~ 0 u ~ oo-oo-oooo~oooooo-ooooooooooo-o~oooo-o-o >oo ~~ oooo~o0oo~~ ~-oooooo~o~ o~~~oo o 00~ ~~ ~M~~M~ ~~~M ~-~M~MMM~~~ ~~ ~~~ ~ A < ~ ~ :::> 0 >t w -o-oo-o-ooo-ooooo-ooooooooooo-o~ooo--o o 00~ o oo~ ~oo~~ ~-oooooo~o~ o~~~o o ~ ~~ ~ W~M ru~~~M ~-~M~MMM ~~ ~~ ~· ~ 000000000000000000000000000000000000000 t-:~oo~o-ruM~~~~oo~o-ruM~~~~oo~o-ruM~~~~oo~o-ruM~~~ ca~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~oooooooooooooo ::>000000000000000000000000000000000000000 U)OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ~oooo-~~~ru~M~~~-~~ooru---ruMru-oo~M~oo~~oo~~oo~M~Mruru~M·~~~~ H--rururu--ru--rururururu--rururu---ru-ru-ru-ru--ru--ru-ru---ru-ru-ruru Ulruru-ru--ruru--ru--ru-ruru---ru-ru-ruru--ruru-ruru-ruru---rururu-ru-ru Wruru--ru--ruru--ru--ruru-----ru--ru-rururu--ruru-rururururu--ruru-·-ruru Oruru-rururu---rururu-rururu--------ruru-rururururu-ruru--ru-----ruru-ru O::ooooooooooooo-ooooooooooo-ooo-ooooooooooooooooo O::oooo~oruoooo~~o ooo~ooo~ooo o~o ooooo~ooooo~~oooo O~~ru~~~ M~~~~M M~~ ~~~ruM~~ ~~~ W~M~~ ~~M~~M~~~~0 u oooooooooooo--oooooooo-oo-ooo-oooooooooooooo-oo >oooo~oruoooo~~ o~~ru~~~ 0~~~~ o::C 000~000~ M~~ ~M~ru ·=· ·=· • . .(1 •.£t ·=· ,,., ·=· (•) l.f:O •.[.1 00000~000000~0 ~~M~~ ~~~~~M~~ ·=· ·=· •.(1 f•) oooooooooooo--oooooooo-oo-ooo-oo-oo~oooooooo-oo ::>oo~o~oruoooo~~ ooo~oooo oo o~o oo ooru~.ooooo~o oo 0 •-L• •.£.• .-. •.[., Ll":t •.£a (•) 1,-:r •.£t V IJ":• f•:• •.£t •.£• •.£t (•) •.£t ·~·) •.J) •.£t •.£• II~•. •.J) •.f..t •.£1 •.£• •.£t •.f.t •.£• •.£.• •.ft •.f.• f•) tf:- •.J) •.J) f•) >t ooooooooooooooooooooooooooooooooooooooooooooooo t-:~-ruM~~~~oo~o-ruM~~~~oo~o-ruM~~~~oo~o-ruM~~~~oo~o-ruM~~~~ caooooooooo----------rurururururururururuMMMMMMMMMM~~~~~~~~ ::>OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO U)OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 0"1 0"1 ..-I PHASE 2 JUDGMENTS ON TAX CASE C (TRAVEL EXPENSE SCENARIO) SUBJ YOU ADV CORR DESIGN 00010 1400 1400 1400 2111 5 00020 1150 1150 1150 2212 j 00030 400 750 1 2212 7 00040 900 900 900 2221 8 00050 700 700 700 2111 5 00060 1100 1100 1100 1211 3 00070 2500 1500 1100 2221 8 00080 1400 1400 1400 1112 1 00090 700 700 75 2122 6 00100 1400 1400 1400 2212 7 00110 1400 1400 1600 1121 2 00120 600 600 1 1121 2 00130 1400 1000 100~ 2122 6 00140 1400 1400 1400 2111 5 00150 1400 1400 1400 2122 6 00160 1400 1400 1400 1112 1 00170 1400 1400 1400 2122 6 00180 1400 1400 1400 1211 3 00190 1000 1000 1000 2111 5 00200 1400 1400 1400 1211 3 00210 1450 1450 1 2212 7 00220 900 700 400 2212 7 00230 1400 1400 1000 2122 6 00240 1400 1400 1400 1112 1 00250 1400 1400 1400 2111 5 00260 1400 1400 1400 2111 5 00270 1400 1400 1400 1112 1 00280 1375 1375 1175 1211 3 00290 7BO 700 700 1211 3 00300 1000 1000 1000 2111 5 00310 1400 1400 1400 1112 1 00320 650 700 700 2111 5 00330 1400 1400 1400 2221 8 00340 1400 1400 1400 1121 2 00350 1400 1400 1400 2111 5 00360 100 100 1050 1121 2 00370 ~nn 800 400 2212 7 00380 1400"1200 1000 1112 1 00390 600 600 600 2111 5 00400 1400 1400 1400 1211 3 00410 1400 1400 1400 2212 7 00420 1200 1200 1200 2122 6 00430 1475 1450 1450 2111 5 00440 1400 1400 1400 2212 7 00450 1400 1400 1 1121 2 00460 1400 1400 1400 1211 3 00470 1400 1400 1400 1121 2 SUBJ YOU ADV CORR DESIGN 00480 900 900 900 2221 8 00490 1400 1400 1400 1121 2 00500 1400 750 750 2122 6 00510 1500 2000 1400 1121 2 00520 1500 1600 2000 1121 2 00530 900 900 1200 1222 4 00540 1400 1400 1400 1211 3 00550 1400 1400 1200 1211 3 00560 800 1400 700 1222 4 00570 1400 1400 1300 2221 8 00580 500 500 500 1121 2 00590 1400 1400 800 2122 6 00600 1400 1400 1400 1211 3 00610 1 1 200 2221 8 00620 1200 1200 1200 2111 5 00630 1350 1350 1350 1222 4 00640 1700 900 900 2221 8 00650 900 900 900 1222 4 00660 875 800 800 1222 4 00670 1300 1300 1300 1112 1 00680 1400 1400 1200 1222 4 00690 800 800 800 2212 7 00700 1400 1400 1400 1112 1 00710 1800 1800 1800 2221 8 00720 1400 1400 1400 2221 8 00730 1000 1000 1000 2122 6 00740 1300 1300 1300 1121 2 00750 1500 1400 1000 1112 1 00760 650 650 700 222i 8 00770 1400 1400 1400 2122 6 00780 1400 1400 1600 1112 1 00790 450 450 450 2212 ~· 00800 1500 1500 !500 1222 4 00810 1400 900 900 1222 4 00820 1400 1400 1400 1222 4 00830 1400 1400 1400 1222 4 00840 1400 1¢00 1400 1211 3 00850 1000 1000 1000 1112 1 00860 1400 1400 1400 2212 7 END OF DATA 200 PHASE 2 JUDGMENTS ON TAX CASE D (REHABILITATION CREDIT SCENARIO) SUBJ YOU ADV CORR DESIGN 00010 1500 1500 1500 1112 1 00020 1500 151)0 1500 1121 2 00030 2200 1200 750 2221 8 00040 3500 3500 3500 2212 7 00050 2100 4000 2000 1112 1 00060 1700 1700 1700 1112 1 00070 4000 2300 2700 2212 7 0 (11)::: 0 15 0 0 1 50 0 1 5!) 0 1211 3 00090 1500 1500 1500 2111 5 00100 1500 1500 1500 1112 1 00110 3000 3000 4000 2111 5 00120 1000 1000 1000 2122 6 00130 1500 1700 1500 2111 5 00140 1060 1000 1000 1121 2 00150 1500 1500 1500 2111 5 00160 2000 2000 1500 2122 6 00170 4000 2000 2000 2212 7 00180 4000 4000 4000 1222 4 00190 1250 1250 1250 2212 7 00200 ~000 4000 4000 1222 4 0021 (I 1500 1500 1-500 2111 5 00220 4000 1900 1900 1222 4 00230 2000 2000 1500 1222 4 00240 2000 2000 2000 1211 3 00250 1 1 1 1112 1 00260 2000 2000 2000 2212 7 00270 4000 4000 4000 1222 4• 00280 L260 1260 1260 2122 6 00290 2000 2000 2000 1121 2 00300 4000 4000 4000 1211 3 00310 2250 2250 2250 1211 3 00320 1500 1500 1500 1121 2 oo:~:·~:o 1500 15oo 15oo 1121 2 00340 2000.2000 125Q 2122 6 00350 1500 1500 1500 1112 1 00360 1250 1150 3015 2212 7 00:~:71) ·1 000 2000 2000 2122 6 00380 750 750 750 1211 3 00390 1000 1000 1000 2221 8 00400 1500 1500 1500 1112 1 00410 2000 2000 2000 2122 6 00420 1000 1000 1000 2221 8 00430 1500 1500 1500 2221 ::: 00440 1500 1500 1500 1222 4 00450 2000 2000 2000 2212 7 00460 2000 2000 1500 1222 4 00470 2000 2000 2000 1121 2 SUBJ YOU ADV CORR DESIGN 00480 3300 3300 3300 2212 7 00490 2000 2000 2000 2122 6 00500 3500 2500 2000 2212 ( 00510 4000 4000 4000 2122 6 00520 2500 2500 3500 2221 8 00530 2000 2000 2000 2221 8 00540 1400 1400 1350 2212 7 00550 1500 1500 1500 1222 4 00560 1500 1500 1500 2221 8 00570 2000 2000 2000 1112 1 00580 2750 2750 3000 2111 5 00590 2000 2000 2000 1121 2 00600 4000 4000 4000 2122 6 00610 220 280 130 2111 5 00620 2250 2250 2250 1112 1 00630 1500 1500 1500 1112 1 00640 0 2100 2100 1211 3 OOE.50 1500 1500 1501l. 1211 :~: oo.: . .:.o 15(1(1 1500 1500 1211 ·~: 00670 180(1 1800 1800 1222 4 00680 1500 1500 1100 2111 5 00690 2000 2000 1500 1121 2 00700 1700 1500 1500 2111 5 00710 2000 2000 1500 1121 2 00720 2000 2000 1500 1121 2 00730 3000 3000 3000 1112 1 00740 1500 1500 1500 2221 8 00750 2000 2000 1000 1121 2 00760 400 400 400 2122 6 00770 2000 2000 2000 1211 3 Q0780 2000 2000 2000 2221 ~ 00790 1250 2000 2500 2221 8 I) (1::: 0 (I 1 5 I) 0 1 5 0 (I 1 5 0 0 1 211 3 00810 2100 2100 2100 1121 2 00820 3250 2750 4000 2111 5 00830 4000 4000 4000 2212 7 00840 2000 1500 1250 1211 3 00850 2000 2000 1500 1112 1 00860 2100 2100 3500 1222 4 00870 1500 1500 1500 1222 4 END OF DATA 201 PHASE 2 BACKGROUND DATA SUBJ. QUESTIONS 1-10 IN SEQ. (SEE TABLE 3) 00010 1 2 39 14 40000 07 41 41 37 16 00020 1 2 20 16 50000 14 39 35 15 16 00030 2 1 5 11 0 48 46 48 14 09 00040 2 2 15 18 30000 36 26 00 10 33 00050 1 2 40 15 86000 30 37 48 45 18 00060 2 1 23 14 29~00 11 07 42 34 12 00070 1 1 17 14 32000 01 01 49 01 01 00080 1 1 17 17 21000 13 14 39 05 24 00090 2 2 50 16 45000 08 25 45 45 30 00100 1 1 35 16 60000 34 30 47 41 13 00110 2 1 7 11 15000 20 21 21 24 34 00120 2 1 17 12 20000 31 08 18 ~9 16 00130 1 1 1 17 2000 17 33 29 12 19 00140 2 2-12 14 25000 01 32 49 09 01 00150 2 1 14 14 15000 12 13 37 35 24 00160 1 2 7 18 38000 04 21 37 49 07 00170 2 1 20 12 20000 01 34 47 11 01 00180 2 1 1 16 99000 01 25 49 18 25 00190 2 1 5 13 25000 22 26 35 28 16 00200 1 1 15 19 55000 07 13 49 06 06 00210 1 2 4 19 40000 20 28 39 10 22 00220 1 2 50 19 60000 01 09 50 45 34 00230 1 2 10 17 40000 17 14 29 07 32 00240 2 1 25 15 60000 29 29 40 25 20 00250 1 1 23 19 50000 02 39 48 05 08 00260 1 2 20 14 25000 01 50 50 01 03 00270 1 1 15 15 25000 02 11 40 25 25 00280 2 2 32 13 22000 12 01 25 49 09 00290 1 1 15 16 30000 03 13 38 19 18 00300 2 1 10 13 40000 11 03 36 16 16 00310 1 2 25 15 50000 20 40 2~ 33 17 00320 1 2 35 14 35000 12 11 38 32 32 00330 2 2 17 12 55000 01 11 50 30 10 ocr::4o 2 1 ·7 14 1sooo 05 12 .;:s 17 12 00350 2 1 10 16 31000 09 33 25 25 18 00360 1 1 5 15 30000 01 01 49 48 02 00370 1 2 35 13 70000 01 01 50 01 01 00380 1 1 35 14 20000 22 22 23 13 04 00390 2 2 35 19 30000 15 45 45 36 37 00400 2 1 17 14 70000 09 46 41 07 11 oo41 o 1 1 :;:o 12 4oooo 01 01 5o ot 01 00420 1 1 5 12 7000 18 38 31 30 12 00430 1 2 40 16"44000 02 05 49 24 30 00440 2 1 15 16 85000 40 21 46 04 25 00450 2 1 10 14 65000 06 18 32 03 06 00460 1 2 10 17 40000 31 32 36 29 34 00470 1 2 15 16 45000 02 39 07 43 44 "202 PHASE 2 BACKGROUND DATA (CONTINUED) SUBJ. QUESTIONS 1-10 IN SEQ. (SEE TABLE 3) 00480 2 1 12 14 18000 01 25 49 0? 04 00490 2 2 15 14 25000 28 10 31 31 08 00500 2 1 22 14 25000 05 07 49 39 04 00510 2 1 20 17 65000 02 22 43 13 17 00520 2 1 20 12 0 01 01 50 01 01 00530 2 1 10 16 20000 01 43 50 47 15 00540 1 2 16 19 50000 08 49 30 03 11 00550 1 2 5 16 25000 02 38 42 15 30 00560 1 2 20 19 40000 19 18 29 38 20 00570 1 2 25 17 45000 18 12 41 37 26 00580 1 1 5 14 26000 26 32 32 31 18 00590 1 1 5 16 20000 24 37 42 25 11 00600 1 1 30 16 0 02 27 36 14 13 00610 2 1 10 12 40000 12 11 16 50 36 00620 2 1 22 16 55000 01 01 50 01 25 00630 2 2 14 14 48000 20 12 30 40 43 00640 1 1 50 14 30000 03 01 49 49 47 00650 1 2 12 16 37000 19 16 28 24 18 00660 1 1 45 13 30000 02 49 48 01 37 00670 1 1 35 12 30000 15 28 28 16 14 00680 2 1 20 13 47000 13 11 39 01 11. 00690 2 1 4 16 24000 08 07 21 31 06 00700 2 1 2 16 14000 28 15 39 19 24 00710 1 1 40 19 99000 13 05 41 41 33 00720 1 2 20 1'::, 60000 1:3 :39 40 08 11 00730 2 1 30 14 20000 23 08 49 38 38 00740 1 1 35 17 30000 34 38 23 09 39 00750 1 1 12 14 35000 10 07 34 38 14 00760 2 1 45 12 4?000 07 05 34 05 07 00770 1 2 25 14 47000 26 25 50 01 25 00780 1 1 40 17 30000 24 10 31 47 23 00790 2 1 0 9 0 02 02 02 49 48 00800 2 1 15 12 0 12 24 0 05 18 00810 2 1 40 12 60000 25 25 25 25 10 00820 2 1 15 14 30000 01 03 41 20 22 00830 2 2 15 18 50000 02 35 45 24 09 00840 1 2 18 16 35000 16 42 45 04 03 00850 1 1 20 14 40000 10 ~0 40 18 15 00860 2 1 20 16 99000 02 0 50 0 01 E~m OF DATA 203 APPENDIX G MULTIVARIATE AND UNIVARIATE ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS 204 CASE A MULTIVARIATE AND UNIVARIATE ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 1-All .40 3,76 0.7557 Meas. 1 .81 1,78 0.3712 Meas. 2 .46 1,78 0.4997 Meas. 3 .01 1,78 0.9171 DIM 2-All .59 3,76 0.6242 Meas. 1 1.77 1,78 0.1869 Meas. 2 1.57 1,78 0.2140 Meas. 3 .49 1,78 0.4851 DIM 3-All 1.76 3,76 0.1611 Meas. 1 .90 1,78 0.3454 Meas. 2 2.65 1,78 0.1077 Meas. 3 .31 1,78 0.5779 DIM 4-All .76 3,76 0.5214 Meas. 1 1.08 1,78 0.3021 Meas. 2 .38 1,78 0.5387 Meas. 3 .00 1,78 0.9760 DIM 1 & 2-All 1.44 3,76 0.2379 Meas. 1 .29 1,78 0.5936 Meas. 2 .01 1,78 0.9194 Meas. 3 1.86 1,78 0.1767 DIM 1 & 3-All 2.21 3,76 0.0941 Meas. 1 2.53 1,78 0.1166 Meas. 2 1. 26 1,78 0.2655 Meas. 3 .35 1,78 0.5547 DIM 1 & 4-All .33 3,76 0.8038 Meas. 1 .87 1,78 0.3539 Meas. 2 .53 1,78 0.4702 Meas. 3 .29 1,78 0.5943 DIM 2 & 3-All .33 3,76 0.8038 Meas. 1 .87 1,78 0.3539 Meas. 2 .53 1,78 0.4702 Meas. 3 .29 1,78 0.5943 205 CASE A (Continued) VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 2 & 4-All 2.21 3,76 0.0941 Meas. 1 2.52 1,78 0.1166 Meas. 2 1. 26 1,78 0.2655 Meas. 3 .35 1,78 0.5547 DIM 3 & 4-All 1.44 3,76 0.2379 Meas. 1 .29 1,78 0.5936 Meas. 2 .01 1,78 0.9194 Meas. 3 1.86 1,78 0.1767 DIM 1,2&3-All .76 3,76 0.5214 Meas. 1 1.08 1,78 0.3021 Meas. 2 .38 1,78 0.5387 Meas. 3 .00 1,78 0.9760 DIM 1,2&4-All 1. 76 3,76 0.1611 Meas. 1 .90 1,78 0.3454 Meas. 2 2.65 1,78 0.1077 Meas. 3 .31 1,78 0.5779 DIM 1,3&4-All .59 3,76 0.6242 Meas. 1 1.77 1,78 0.1869 Meas. 2 1.57 1,78 0.2140 Meas. 3 .49 1,78 0.4851 DIM 2,3&4-All .40 3,76 0.7557 Meas. 1 .81 1,78 0.3712 Meas. 2 .46 1,78 0.4997 Meas. 3 .01 1,78 0.9171 206 CASE B MULTIVARIATE AND UNIVARIATE ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 1-All .20 3,76 0.8933 Meas. 1 .46 1,78 0.4978 Meas. 2 .12 1,78 0.7340 Meas. 3 .08 1,78 0.7780 DIM 2-All 2.24 3,76 0.0902 Meas. 1 6.66 1,78 0.0117 Meas. 2 5.05 1,78 0.0274 Meas. 3 2.55 1,78 0.1143 DIM 3-All .85 3,76 0.4696 Meas. 1 1. 25 1,78 0.2663 Meas. 2 .17 1,78 0.6830 Meas. 3 .01 1,78 0.9376 DIM 4-All 3.27 3,76 0.0256 Meas. 1 9.39 1,78 0.0030 Meas. 2 4.61 1,78 0.0350 Meas. 3 3.85 1,78 0.0533 DIM 1 & 2-All .45 3,76 0.7146 Meas. 1 .08 1,78 0.7830 Meas. 2 .66 1,78 0.4181 Meas. 3 .19 1,78 0.6620 DIM 1 & 3-All .50 3,76 0.6862 Meas. 1 .70 1,78 0.4045 Meas. 2 .07 1,78 0.7850 Meas. 3 .23 1,78 0.6357 DIM 1 & 4-All .80 3,76 0.4967 Meas. 1 1.75 1,78 0.1898 Meas. 2 .46 1,78 0.5011 Meas. 3 .11 1,78 0.7457 DIM 2 & 3-All .80 3,76 0.4967 Meas. 1 1.75 1,78 0.1898 Meas. 2 .46 1,78 0.5011 Meas. 3 .11 1,78 0.7457 207 CASE B (Continued) VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 2 & 4-All .50 3,76 0.6862 Meas. 1 .70 1,78 0.4045 Meas. 2 .07 1,78 0.7850 Meas. 3 .23 1,78 0.6357 DIM 3 & 4-All .45 3,76 0.7146 Meas. 1 .08 1,78 0.7830 Meas. 2 .66 1,78 0.4181 Meas. 3 .19 1,78 0.6620 DIM 1,2&3-All 3.27 3,76 0.0256 Meas. 1 9.39 1,78 0.0030 Meas. 2 4.61 1,78 0.0350 Meas. 3 3.85 1,78 0.0533 DIM 1,2&4-All .85 3,76 0.4696 Meas. 1 1.25 1,78 0.2663 Meas. 2 .17 1,78 0.6830 Meas. 3 .01 1,78 0.9376 DIM 1,3&4-All 2.24 3,76 0.0902 Meas. 1 6.66 1,78 0.0117 Meas. 2 5.05 1,78 0.0274 Meas. 3 2.55 1,78 0.1143 DIM 2,3&4-All .20 3,76 0.8933 ~1eas. 1 .46 1,78 0.4978 Meas. 2 .12 1,78 0.7340 Meas. 3 .08 1,78 0.7780 208 CASE C MULTIVARIATE AND UNIVARIATE ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 1-All 2.58 3,76 0.0594 Meas. 1 .71 1,78 0.4014 Meas. 2 3.45 1,78 0.0670 Meas. 3 4.93 1,78 0.0132 DIM 2-All .27 3,76 0.8501 Meas. 1 .08 1,78 0.7837 Meas. 2 .33 1,78 0.5661 Meas. 3 .56 1,78 0.4563 DIM 3-All .66 3,76 0.5766 Meas. 1 .00 1,78 0.9805 Meas. 2 .45 1,78 0.5050 Meas. 3 .36 1,78 0.5528 IM 4-All .62 3,76 0.6074 Meas. 1 .01 1,78 0.9220 Meas. 2 .02 1,78 0.8988 Meas. 3 .82 1,78 0.3690 IM 1 & 2-All .24 3,76 0.8681 Meas. 1 .07 1,78 0.7869 Meas. 2 .17 1,78 0.6811 Meas. 3 .64 1,78 0.4244 DIM 1 & 3-All 2.48 3,76 0.0674 Meas. 1 4.38 1,78 0.0397 Meas. 2 .97 1,78 0.3279 Meas. 3 1. 66 1,78 0.2015 DIM 1 & 4-All 1.19 3,76 0.3206 Meas. 1 .34 1,78 0.5599 Meas. 2 .01 1,78 0.9295 Meas. 3 1.65 1,78 0.2023 DIM 2 & 3-All 1.19 3,76 0.3206 Meas. 1 .34 1,78 0.5599 Meas. 2 .01 1,78 0.9295 Meas. 3 1.65 1,78 0.2023 209 CASE c (Continued) VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 2 & 4-All 2.48 3,76 0.0674 Meas. 1 5.96 1,78 0.0169 Meas. 2 .97 1,78 0.3279 Meas. 3 1.66 1,78 0.2015 DIM 3 & 4-All :; ;~ I~ '' 76 0.8681 Meas. 1 .07 1,78 0.7869 Meas. 2 .17 1,78 0.6811 Meas. 3 .64 1,78 0.4244 DIM 1,2&3-All .62 3,76 0.6074 Meas. 1 .01 1,78 0.9220 Meas. 2 .02 1,78 0.8988 Meas. 3 .82 1,78 0.3690 DIM 1,2&4-All .66 3,76 0.5766 Meas. 1 .00 1,78 0.9805 Meas. 2 .45 1,78 0.5050 Meas. 3 .36 1,78 0.5528 DIM 1,3&4-All .27 3,76 0.8501 Meas. 1 .08 1,78 0.7837 Meas. 2 .33 1,78 0.5661 Meas. 3 .56 1,78 0.4563 DIM 2,3&4-All 2.58 3,76 0.0594 Meas. 1 .71 1,78 0.4014 Meas. 2 3.45 1,78 0.0670 Meas. 3 4.93 1,78 0.0293 210 VARIATE CASE D MULTIVARIATE AND UNIVARIATE ANALYSIS OF THE EFFECT OF COMPLEXITY DIMENSION(S) ON REPORTING POSITIONS F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 1-All 1.43 3,77 0.2408 Meas. 1 .23 1,79 0.6329 Meas. 2 .04 1,79 0.8488 Meas. 3 .51 1,79 0.4778 DIM 2-All 2.15 3,77 0.1010 Meas. 1 3.08 1,79 0.0831 Meas. 2 .79 1,79 0.3766 Meas. 3 2.62 1,79 0.1096 DIM 3-All .19 3,77 0.9028 Meas. 1 .00 1,79 0.9651 Meas. 2 .00 1,79 0.9719 Meas. 3 .11 1,79 0.7416 DIM 4-All 2.17 3,77 0.0986 Meas. 1 6.42 1,79 0.0132 Meas. 2 4.37 1,79 0.0398 Meas. 3 2.85 1,79 0.0951 DIM 1 & 2-All .45 3,77 0.7198 Meas. 1 .07 1,79 0.7860 Meas. 2 .51 1,79 0.4778 Meas. 3 .87 1,79 0.3549 DIM 1 & 3-All 3.01 3,77 0.0353 Meas. 1 5.96 1,79 0.0169 Meas. 2 1.38 1,79 0.2440 Meas. 3 2.54 1,79 0.1152 DIM 1 & 4-All .32 3,77 0.8091 Meas. 1 .34 1,79 0.5622 Meas. 2 .25 1,79 0.6202 Meas. 3 .01 1,79 0.9408 211 CASE D (Continued) VARIATE F TEST DEGREES LEVEL OF OF FREEDOM SIGNIFICANCE DIM 2 & 3-All .32 3,77 0.8091 Meas. 1 .34 1,79 0.5622 Meas. 2 .25 1,79 0.6202 Meas. 3 .01 1,79 0.9408 DIM 2 & 4-All 3.01 3,77 0.0353 Meas. 1 5.96 1,79 0.0169 Meas. 2 1.38 1,79 0.2440 Meas. 3 2.54 1,79 0.1152 DIM 3 & 4-All .45 3,77 0.7198 Meas. 1 .07 1,79 0.7860 Meas. 2 .51 1,79 0.4778 Meas. 3 .87 1,79 0.3549 DIM 1,2&3-All 2.17 3,77 0.0986 Meas. 1 6.42 1,79 0.0132 Meas. 2 4.37 1,79 0.0398 Meas. 3 2.85 1,79 0.0951 DIM 1,2&4-All .19 3,77 0.9028 Meas. 1 .00 1,79 0.9651 Meas. 2 .00 1,79 0.9719 Meas. 3 .11 1,79 0.7416 DIM 1,3&4-All 2.15 3,77 0.1010 Meas. 1 3.08 1,79 0.0831 Meas. 2 .79 1,79 0.3766 Meas. 3 2.62 1,79 0.1096 DIM 2,3&4-All 1.43 3,77 0.2408 Meas. 1 .23 1,79 0.6329 Meas. 2 .04 1,79 0.8488 Meas. 3 .51 1,79 0.4778 212 APPENDIX H DIRECTIONAL EFFECTS OF COMPLEXITY ON REPORTING POSITION 213 DIRECTIONAL EFFECTS OF COMPLEXITY ON REPORTING POSITION SELECTIONS CASE A (CHARITY) FACTOR LEVEL VARIATE COUNT MEAN STD ST[ ERROR DE\ DIM 1 LOW MEAS. 1 (You) 43 1409 98 6 4; (Topic) MEAS. 2 (Advise) 43 1380 98 64 MEAS. 3 (Correct) 43 1335 97 64 HIGH MEAS. 1 43 1551 84 55 t MEAS. 2 43 1490 84 55( MEAS. 3 43 1360 98 6 4 f DIM 2 LOW MEAS. 1 43 1398 94 62 (Quant) MEAS. 2 43 1357 94 61 MEAS. 3 43 1295 107 70t HIGH MEAS. 1 43 1561 87 5 7t MEAS. 2 43 1514 88 57~ MEAS. 3 43 1400 87 57 DIM 3 LOW MEAS. 1 44 1417 95 63t (Soc.Just) MEAS. 2 44 1329 94 6 2 t MEAS. 3 44 1315 95 63( HIGH MEAS. 1 42 1545 86 56 MEAS. 2 42 1546 86 55~ MEAS. 3 42 1381 101 6 5 ~ DIM 4 LOW MEAS. 1 46 1547 85 57 (Read) MEAS. 2 46 1480 87 59 MEAS. 3 46 1351 86 5 8E HIGH MEAS. 1 40 1402 99 6 2 E MEAS. 2 40 1384 96 61 MEAS. 3 40 1344 111 7 0 E 214 DIRECTIONAL EFFECTS OF COMPLEXITY ON REPORTING POSITION SELECTIONS CASE B (INCOME/GIFT) FACTOR LEVEL VARIATE COUNT MEAN STD ST ERROR DE DIM 1 LOW MEAS. 1 (You) 46 365 37 25 (Pers/Fin) MEAS. 2 (Advise) 46 404 32 21 MEAS. 3 (Correct) 46 444 30 20 HIGH MEAS. 1 40 397 41 25 MEAS. 2 40 419 40 25 MEAS. 3 40 457 37 23 DIM 2 LOW MEAS. 1 44 313 41 27 (Quant) MEAS. 2 44 356 40 26 MEAS. 3 44 412 37 24 HIGH MEAS. 1 42 450 33 21 MEAS. 2 42 469 27 18 MEAS. 3 42 489 27 17 DIM 3 LOW MEAS. 1 44 405 36 23 (Soc.Just) MEAS. 2 44 420 34 23 MEAS. 3 44 446 32 21 HIGH MEAS. 1 42 354 41 27 MEAS. 2 42 402 37 24 MEAS. 3 42 453 35 22 DIM 4 LOW MEAS. 1 44 460 34 22 (Read) MEAS. 2 44 465 32 21 MEAS. 3 44 497 29 19 HIGH MEAS. 1 42 297 39 25 MEAS. 2 42 355 37 24 MEAS. 3 42 401 36 23 215 I DIRECTIONAL EFFECTS OF COMPLEXITY I I ON REPORTING POSITION SELECTIONS I I I CASE C (TRAVEL EXPENSES) I I I I I I FACTOR LEVEL VARIATE COUNT MEAN STD ST~ ERROR DE I I i I DIM 1 LOW MEAS. 1 (You) 43 1251 47 3131 1 (Pers/Fin) MEAS. 2 (Advise) 43 1259 50 33~ I MEAS. 3 (Correct) 43 1202 58 384j I I HIGH MEAS. 1 43 1175 66 436 I I 369 I MEAS. 2 43 1113 56 I I MEAS. 3 43 1002 68 44/1 DIM 2 LOW MEAS. 1 44 1223 49 33~ (Quant) MEAS. 2 44 1207 53 353 MEAS. 3 44 1137 64 42g HIGH MEAS. 1 42 1202 66 42~ MEAS. 2 42 1164 55 362 MEAS. 3 42 1065 66 429 DIM 3 LOW MEAS. 1 45 1212 45 30~ (Soc.Just) MEAS. 2 44 1209 43 29] MEAS. 3 45 1127 59 40~ HIGH MEAS. 1 41 1214 70 449 MEAS. 2 41 1160 65 4H~ MEAS. 3 41 1074 71 456 DIM 4 LOW MEAS. 1 44 1208 67 448 (Read) MEAS. 2 44 1181 62 41~ MEAS. 3 44 1143 64 429 HIGH MEAS. 1 42 1218 45 295 MEAS. 2 42 1190 44 286 MEAS. 3 42 1059 65 424 216 ~--~~--~~----- DIRECTIONAL EFFECTS OF COMPLEXITY ON REPORTING POSITION SELECTIONS CASE D (REHABILITATION CREDIT) I FACTOR LEVEL VARIATE COUNT MEAN STD. STD I ERROR DEV DIM 1 LOW MEAS. 1 (You) 46 1947 130 887 (Pers/Fin) MEAS. 2 (Advise) 46 1977 124 841 MEAS. 3 (Correct) 46 1872 124 845 HIGH MEAS. 1 41 2067 163 1047 MEAS. 2 41 1957 139 894 MEAS. 3 41 2037 165 1057 DIM 2 LOW MEAS. 1 44 1823 1·20 8001 (Quant) MEAS. 2 44 1879 126 835 MEAS. 3 44 1777 136 908 HIGH MEAS. 1 43 2187 165 1082 MEAS. 2 43 2058 135 887 MEAS. 3 43 2126 147 967 DIM 3 LOW MEAS. 1 44 2004 153 1015 (Soc.Jus·t) MEAS. 2 44 1964 133 8831 MEAS. 3 44 1981 143 9501 HIGH MEAS. 1 43 2003 139 916 MEAS. 2 43 1971 129 8491 MEAS. 3 43 1918 146 957 DIM 4 LOW MEAS. 1 43 1755 107 707 1 (Read) MEAS. 2 43 1776 95 627 MEAS. 3 43 1777 128 844 HIGH MEAS. 1 44 2246 167 1114 MEAS. 2 44 2155 152 1014 MEAS. 3 44 2118 154 1022 217

Asset Metadata

Creator Milliron, Valerie Colleen (author)

Core Title Taxpayer perceptions of complexity and the effect of complexity on reporting positions

Contributor Digitized by Interlibrary Loan Department (provenance)

Degree Doctor of Philosophy

Degree Program Business Administration

Defense Date 05/01/1984

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

Tag OAI-PMH Harvest

Format application/pdf (imt)

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-499315

Unique identifier UC11297741

Identifier etd-Milliron-577096.pdf (filename),usctheses-c3-499315 (legacy record id)

Legacy Identifier etd-Milliron-577096.pdf

Dmrecord 499315

Document Type Dissertation

Format application/pdf (imt)

Rights Milliron, Valerie Colleen

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

Linked assets

University of Southern California Dissertations and Theses