Page 162 |
Save page Remove page | Previous | 162 of 229 | Next |
|
small (250x250 max)
medium (500x500 max)
Large (1000x1000 max)
Extra Large
large ( > 500x500)
Full Resolution
All (PDF)
|
This page
All
|
153
categorical variables reflecting the degree of fractionalization of each state. Religion
and Language are represented by three groupings. Ethnic is represented by two
groupings.
Cox Regression Results: Model 3 – Time-Dependent Covariates
Table 40 summarizes results of three specifications of a Cox regression with
time-dependent covariates. The separate specifications address the high correlation
between the GDP per capita growth rate and the GDP growth rate. Specification 1
includes both growth rates. Specifications 2 and 3 include only one of the growth
rates. The dependent variable for all three specifications is Days 3.
In contrast to the results derived from categorical variables representing
income, use of time-dependent covariates for income fails to produce non-zero
coefficient estimates that are statistically significant. The coefficient estimates for
the GDP and CGDP covariates are not statistically significant in any of the
specifications. Similarly, the freedom time-dependent covariate is not statistically
significant. This finding is surprising given the importance attributed to income
groupings in the Kaplan-Meier and the Cox regression analyses. Similarly, the
coefficient estimates for the growth rate time-dependent covariates are not
statistically significant. Inclusion of the time-dependent covariates, however, does
seem to influence the individual region estimates as evidenced by the complete lack
of any statistically significant result and the aforementioned sign change. The results
Object Description
| Title | Riding the wave: an interdisciplinary approach to understanding the popularity of RTA notifications to the GATT/WTO |
| Author | McClough, David Andrew |
| Author email | mcclough@usc.edu; dmcclou@bgsu.edu |
| Degree | Doctor of Philosophy |
| Document type | Dissertation |
| Degree program | Political Economy & Public Policy |
| School | College of Letters, Arts and Sciences |
| Date defended/completed | 2008-08-07 |
| Date submitted | 2008 |
| Restricted until | Unrestricted |
| Date published | 2008-10-18 |
| Advisor (committee chair) | Katada, Saori N. |
| Advisor (committee member) |
Nugent, Jeffrey B. Cartier, Carolyn |
| Abstract | The proliferation of Regional Trade Agreements (RTAs) notified to the GATT/WTO since the early 1980s deviates from the long-term trend and reflects participation of nearly every member of the United Nations. This dissertation seeks to explain the current wave of RTA notifications by supplementing the economic model of supply and demand with diffusion theory. Application of the supply and demand model is useful in distinguishing between changes in demand and changes insupply of RTAs. This distinction is seldom emphasized in the current literature examining RTAs. Recent applications of diffusion theory in the discipline of international relations offer a unique opportunity to include a dynamic force in the static analysis of the supply and demand model. Empirical analysis assesses the fit of the RTA diffusion pattern by comparing the RTA diffusion pattern to a cumulative standard normal distribution. The analysis indicates that the diffusion pattern of RTAs resembles the diffusion of an innovation through a social system.; The implication of this finding is that the adoption of an RTA as trade policy is not made independently of the decision by other states. Indeed, the analysis suggests interdependency between states. Further empirical analysis explores economic and political variables that may explain the decision to adopt the RTA as trade policy. The empirical analysis is unique in that survival analysis is utilized to assess the variation in duration to adopt an initial RTA since the early 1980s. A central discovery is that regional designation explains the variation in duration to adopt an initial RTA. Multiple regression analysis confirms the results generated using survival analysis and support the assertion that the proliferation of RTAs likely reflects changes in both the demand for RTAs and the supply of RTAs. This dissertation concludes by considering implications for the WTO resulting from the increase in RTA notifications. |
| Keyword | trade agreements |
| Coverage date | after 1980 |
| Language | English |
| Part of collection | University of Southern California dissertations and theses |
| Publisher (of the original version) | University of Southern California |
| Place of publication (of the original version) | Los Angeles, California |
| Publisher (of the digital version) | University of Southern California. Libraries |
| Provenance | Electronically uploaded by the author |
| Type | texts |
| Legacy record ID | usctheses-m1675 |
| Contributing entity | University of Southern California |
| Rights | McClough, David Andrew |
| Repository name | Libraries, University of Southern California |
| Repository address | Los Angeles, California |
| Repository email | cisadmin@lib.usc.edu |
| Filename | etd-McClough-2338 |
| Archival file | uscthesesreloadpub_Volume32/etd-McClough-2338.pdf |
Description
| Title | Page 162 |
| Contributing entity | University of Southern California |
| Repository email | cisadmin@lib.usc.edu |
| Full text | 153 categorical variables reflecting the degree of fractionalization of each state. Religion and Language are represented by three groupings. Ethnic is represented by two groupings. Cox Regression Results: Model 3 – Time-Dependent Covariates Table 40 summarizes results of three specifications of a Cox regression with time-dependent covariates. The separate specifications address the high correlation between the GDP per capita growth rate and the GDP growth rate. Specification 1 includes both growth rates. Specifications 2 and 3 include only one of the growth rates. The dependent variable for all three specifications is Days 3. In contrast to the results derived from categorical variables representing income, use of time-dependent covariates for income fails to produce non-zero coefficient estimates that are statistically significant. The coefficient estimates for the GDP and CGDP covariates are not statistically significant in any of the specifications. Similarly, the freedom time-dependent covariate is not statistically significant. This finding is surprising given the importance attributed to income groupings in the Kaplan-Meier and the Cox regression analyses. Similarly, the coefficient estimates for the growth rate time-dependent covariates are not statistically significant. Inclusion of the time-dependent covariates, however, does seem to influence the individual region estimates as evidenced by the complete lack of any statistically significant result and the aforementioned sign change. The results |
