Steven B. Caudill

Frontier Estimation and Firm-Specific Inefficiency Measures in the Presence of Heteroscedasticity

The purpose of this article is to illustrate a straightforward and useful method for addressing the problem of heteroscedasticity in the estimation of frontiers. A heteroscedastic cost-frontier model is developed and estimated using bank cost data similar to that used by Ferrier and Lovell. Our results show dramatic changes in the estimated cost frontier and in the inefficiency measures when accounting for heteroscedasticity in the estimation process. We find that the rankings of firms by their inefficiency measures is affected markedly by the correction for heteroscedasticity but not by alternative distributional assumptions about the one-sided error term.

Crime in the Classroom: An Economic Analysis of Undergraduate Student Cheating Behavior

A logit model is used to study the cheating behavior of students in two large principles of microeconomics classes: cheating is inversely related to GPA and directly related to the perception of the number of students who routinely cheat.

Biases in frontier estimation due to heteroscedasticity

Abstract This paper investigates the effects of heteroscedasticity on the parameters in frontier regression models. A Monte Carlo experiment is employed to determine that heteroscedasticity leads to overestimation of the intercept and underestimation of the slope coefficients.

The Impact of Participation in Intercollegiate Athletics on Income and Graduation

Males who participated in intercollegiate athletics are estimated to receive 4 percent higher annual incomes than similar nonathletes. No such income premium associated with college athletics is revealed among females. Both male and female athletes who attended colleges and universities in the early 1970s had higher graduation rates than other students. Since the models used to estimate income and graduation differentials included many measurable determinants of labor market and academic outcomes, these findings suggest that athletic participation may enhance the development of discipline, confidence, motivation, a competitive spirit, or other subjective traits that encourage success. Copyright 1991 by MIT Press.

Financial development and economic growth: the case of Taiwan

This paper examines the relationship between financial development and economic growth in Taiwan from 1962 to 1998. Using a four-variable VAR model, the competing hypotheses of demand-following versus supply-leading are empirically tested. The results from Granger causality tests based on vector error-correction models (VECM) suggest unidirectional causality running from financial development (measured as the ratio of M2 to GDP) to economic growth. This result supports the supply-leading hypothesis for Taiwan. This finding highlights the importance of financial development in Taiwans recent growth.

Tax-and-spend, spend-and-tax, or fiscal synchronization: new evidence for ten countries

Cointegration and vector autoregression are used to test the ‘Tax-and-Spend’, ‘Spend-and-Tax’, and ‘Fiscal Synchronization’ for ten countries using annual time-series data over the period 1951 to 1996. Three of them are part of the newly industrialized countries of Asia (South Korea, Taiwan, and Thailand) and seven are industrialized countries (Australia, Canada, Japan, New Zealand, South Africa, UK, and the USA). This paper includes GDP as a control variable into the model like Baghestani and Mcnown (1994), Ross and Payne (1998), and Koren and Stiassny (1998). The Johansen (1988) and Johansen and Juselius (1990) cointegration test results indicate that these three variables are cointegrated with two cointegrating vectors for South Korea, one vector for Australia, Canada, South Africa, Taiwan, UK, and the USA, and no vector for Japan, New Zealand, and Thailand. The results from Granger causality tests suggest unidirectional causality running from revenues to spending, supporting the ‘Tax-and-Spend’ hypothesis, for Japan, South Korea, Taiwan, UK, and the USA. The opposite relationship, supporting the ‘Spend-andTax’ hypothesis, holds only for Australia and South Africa. In the case of Canada, this study finds a feedback existing between revenues and spending, supporting the ‘Fiscal Synchronization’ hypothesis. For New Zealand and Thailand, these results support none of the hypotheses.

A re-examination of Wagner's law for ten countries based on cointegration and error-correction modelling techniques

Following Manns (National Tax Journal, 33, 189–201, 1980) study, five different versions of Wagners law are empirically examined using annual time-series data on ten countries over the period 1951 to 1996. Included are three of the emerging industrialized countries of Asia: South Korea, Taiwan, and Thailand, and seven industrialized countries: Australia, Canada, Japan, New Zealand, USA, the United Kingdom, and South Africa. The analysis is an advance over previous work in two respects. First, the stationarity properties of the data, the order of integration using the Augmented Dickey–Fuller (Journal of American Statistical Association, 74, 427–31, 1979, Econometrica, 49(4), 1057–72, 1981) test and the Kwiatkowski et al. (Journal of Econometrics, 1, 159–78, 1992) test are empirically investigated. Second, the hypothesis of a long-run relationship between income and government spending is tested using bivariate cointegrated systems and by employing the methodology of cointegration analysis as suggested by Johansen and Juselius (Oxford Bulletin of Economics and Statistics, 52, 169–210, 1990) and Johansen (Journal of Policy Modelling, 14, 313–34, 1992). Unidirectional Granger causality is found running from income to government spending for the newly industrialized countries of South Korea and Taiwan, and the industrialized countries of Japan, the United Kingdom, and the United States, supporting Wagners hypothesis for those countries. For the five remaining countries in this study: Australia, Canada, New Zealand, South Africa, and Thailand, no causal relationship between income and government spending is found.

Finite Mixture Estimation of Multiproduct Cost Functions

This paper presents a technique of cost-function estimation, based on the theory of finite mixture distributions, which allows for the simultaneous existence of multiple technologies of production when the researcher does not know which observations correspond to which technologies. The finite mixture technique provides estimates of the proportions of firms using the various technologies, facilitates comparisons between technologies, and preserves the traditional interpretations of cost estimation. After describing the mixture procedure, the technique is illustrated on a large sample of savings and loan associations, and it is concluded that this industry exhibits multiple technologies of production. Copyright 1991 by MIT Press.

A discrete-time hazard model of lottery adoption

This paper uses conventional logit probablitities to estimate a discrete-time hazard model of lottery adoption. The data set consists of a time-series of cross-sections on states in the US. Our findings suggests that politicians are more likely to support the decision. A lottery is more likely to be adopted of ther is a favourable climate for gambling in the state and if a bordering state has adopted a lottery.

Modeling household fertility decisions: estimation and testing of censored regression models for count data.

This paper adds to the recent body of research on fertility by estimating and testing censored poisson regression models and censored negative binomial regression models of household fertility decisions. A novel feature of this study is that in each case the censoring threshold varies from individual to individual. Also, a Lagrange multiplier or score test is used to investigate overdispersion. In these regression models the dependent variable is the number of children. In this situation, censored poisson regression models and censored negative binomial regression models have statistical advantages over OLS, uncensored poisson regression models, and uncensored negative binomial regression models. The censored models employed in this study are estimated using panel data collected from the Consumer Expenditure Survey compiled by the Bureau of Labor Statistics. The findings of this study support the fertility hypothesis of Becker and Lewis (1965-70).