Showing papers on "Tobit model published in 1996"
TL;DR: In this article, a large number of different pseudo-R2 measures for some common limited dependent variable models are surveyed, including those based solely on the maximized likelihoods with and without the restriction that slope coefficients are zero, those which require further calculations based on parameter estimates of the coefficients and variances, and those that are based on whether the qualitative predictions of the model are correct or not.
Abstract: A large number of different Pseudo-R2 measures for some common limited dependent variable models are surveyed. Measures include those based solely on the maximized likelihoods with and without the restriction that slope coefficients are zero, those which require further calculations based on parameter estimates of the coefficients and variances and those that are based solely on whether the qualitative predictions of the model are correct or not. The theme of the survey is that while there is no obvious criterion for choosing which Pseudo-R2 to use, if the estimation is in the context of an underlying latent dependent variable model, a case can be made for basing the choice on the strength of the numerical relationship to the OLS-R2 in the latent dependent variable. As such an OLS-R2 can be known in a Monte Carlo simulation, we summarize Monte Carlo results for some important latent dependent variable models (binary probit, ordinal probit and Tobit) and find that a Pseudo-R2 measure due to McKelvey and Zavoina scores consistently well under our criterion. We also very briefly discuss Pseudo-R2 measures for count data, for duration models and for prediction-realization tables.
456 citations
TL;DR: In this paper, the performance of a sample of French urban transit companies is evaluated using a broad selection of nonparametric reference technologies for two specifications of the production process, and the results corroborate results reported elsewhere: the relevance of ownership, the use of risk-sharing incentives in contracting, the harmful impact of subsidies, etc.
Abstract: The performance of a sample of French urban transit companies is evaluated using a broad selection of nonparametric reference technologies for two specifications of the production process. In particular, the variable returns to scale Data Envelopment Analysis (DEA) models with either strong or weak disposability in both inputs and outputs, and the Free Disposal Hull (FDH) are applied. An extensive comparison of the resulting radial output efficiency measures yields the following major methodological conclusions. First, the location of the efficiency distributions differs substantially depending on the methodology and especially on the output specification considered. The latter differences vanish if the impact of outliers is eliminated. Second, convexity has a stronger influence on the efficient-inefficient dichotomy than allowing for congestion by means of a weakly disposable DEA model. For policy purposes, these efficiency distributions are explained using a Tobit model. The findings corroborate results reported elsewhere: the relevance of ownership, the use of risk-sharing incentives in contracting, the harmful impact of subsidies, etc. Furthermore, the network structure seems to account for some differences in performance. Finally, a novelty in the urban transit context is the indirect monitoring effect of the French earmarked transportation tax.
188 citations
TL;DR: In this article, the Tobit model is used to explain the budget share that Dutch families spend on vacations. But, the authors take account of the substantial number of zero shares, and two types of models are used.
Abstract: We analyse several limited dependent variable models explaining the budget share that Dutch families spend on vacations. To take account of the substantial number of zero shares, two types of models are used. The first is the single-equation censored regression model. We estimate and test several parametric and semi-parametric extensions of the Tobit model. Second, we consider two-equation models, in which the participation decision and the decision on the amount to spend are treated separately. The first decision is modelled as a binary choice model; the second as a conditional regression. We estimate and test parametric and semi-parametric specifications.
139 citations
01 Jan 1996
TL;DR: In this article, a discrete-continuous choice model of time allocation to two types of discretionary activities is formulated based on the concept of random utility maximization, which takes into account the discreteness inherent in daily activity engagement and time allocation (i.e. one type of activity may not be engaged, therefore no time may be allocated to it at all, during a day).
Abstract: A discrete-continuous choice model of time allocation to two types of discretionary activities is formulated based on the concept of random utility maximization. The model development takes into account the discreteness inherent in daily activity engagement and time allocation (ie one type of activity may not be engaged, therefore no time may be allocated to it at all, during a day). Without requiring simplifying assumptions other than the one that at least one of the two types of activities will be engaged, the utility model is formulated as a doubly-censored Tobit model. This theoretical framework is applied to study daily time allocation to in-home and out-of-home discretionary activities while using, as explanatory variables, work schedules and commute characteristics as well as residential, household and personal attributes. A weekly time-use data set from the Netherlands is used in the statistical analysis. To account for heterogeneity in the data set comprising repeated measurements of daily time use for each respondent, an error component is introduced into the doubly-censored Tobit models. A non-parametric approach is adopted for model estimation using mass points. Behavioral implications of model estimation results are discussed. The validity of a procedure to correct estimated t-statistics proposed for repeated measurement data is also discussed. (A) For the covering abstract see IRRD 886400.
92 citations
TL;DR: According to the results, a high percentage of funding coming from central government matching grants and high taxable income per inhabitant are significant predictors of inefficiency.
Abstract: This paper examines the productive efficiency of Finnish health centres by applying data envelopment analysis (DEA) and econometric methods. The Tobit model was used in an attempt to find out how various economic, structural and demographic factors affect efficiency. The dependent variable of the model, the coefficient of inefficiency, was obtained by deducting the DEA efficiency score from one. According to the results, a high percentage of funding coming from central government matching grants and high taxable income per inhabitant are significant predictors of inefficiency. The results suggest that more generous resources tend to increase inefficiency since they may lessen incentives for tight cost and performance control. A high share of doctors and a low share of administrative, maintenance and support personnel promote efficiency. A low population share of the elderly and a long distance to the nearest hospital are positively associated with inefficiency.
79 citations
TL;DR: In this article, Tobit analysis is introduced into an adoption model to explain why larger and more educated firms adopt earlier, and dynamic economies of scale arise in learning by using that speed up adoption.
Abstract: Learning by using is introduced into an adoption model to explain why larger and more educated firms adopt earlier. Dynamic economies of scale arise in learning by using that speed up adoption. The empirical estimation of time of adoption using Tobit analysis integrates the concepts of adoption and diffusion, allowing the diffusion of the technology to be derived from the time of adoption analysis. In addition, by introducing heterogeneity among the adopters, Tobit analysis is shown to provide superior results to the traditional logit and probit analysis of the dichotomous adopt/not adopt variable.
76 citations
TL;DR: The authors examine the determinants of household alcohol expenditures by using a nonnormal and heteroskedastic double-hurdle model to accommodate zero observations in the sample.
Abstract: This paper examines the determinants of household alcohol expenditures by using a non normal an d heteroscedastic double-hurdle model to accommodate zero observations in the sample. The model is a generalization of the double-hurdle model estimated in previous studies of alcohol consumption. We also examine the effects of explanatory variables by calculating and decomposing the elasticities. Findings support the use of a more generalized error distribution. Income, region, education, and household demographics are among the significant determinants of alcohol expenditures. There is longstanding interest in the economic aspects of alcoholic beverages, although formal economic models of alcohol consumption are a more recent phenomenon (Hilton 1993). Findings on the price responses of alcohol demand have obvious policy implications--in promoting tax revenues and controlling consumption through taxes (Ornstein 1980). The identification of nonprice determinants of alcohol consumption is also important because such information often plays a crucial role in the prevention of alcohol abuse and alcohol-related problems (Hilton 1993). There have been numerous studies on the demand for alcoholic beverages. Ornstein (1980) and Ornstein and Levy (1983) provide surveys of earlier studies, and Leung and Phelps (1993) review studies that have appeared during the last decade. In addition, Selvanathan (1991) reviews the recent literature on alcohol demand systems. Most of the earlier studies have been based on aggregate time series. However, a notable trend among the more recent studies on alcohol consumption and expenditures is the growing use of survey data. Examples include Atkinson, Gomulka, and Stern (1990); Blaylock and Blisard (1993); Heien and Pompelli (1989); Pompelli and Heien (1991); and Yen (1994). Use of microdata allows examination of the effects of detailed household characteristics and provides the degrees of freedom to estimate a large number of parameters. However, microdata also present a unique problem: the sample often contains a significant proportion of observations with reported zero expenditures. Standard econometric techniques not accounting for this data feature lead to biased and inconsistent parameter estimates (Maddala 1983). Researchers have often used the Tobit model (Tobin 1958) to estimate demand relationships with limited dependent variables. However, parameterization of the Tobit model is known to be very restrictive, which makes it unpalatable for empirical analysis. Recent analysts have used the double-hurdle model in modeling alcohol demand (Blaylock and Blisard 1993; Yen 1994). The double-hurdle model features two separate stochastic processes for participation and consumption; it allows for examining the determinants of both participation and consumption decisions and provides more useful insights on consumer behavior than the Tobit model. Applications of the double-hurdle model in other areas of food demand include Haines, Guilkey, and Popkin (1988); Reynolds (1990); and Wang and Jensen (1994). The model specification and estimation procedure of the double-hurdle model require specification of the distribution for the error terms of the participation and consumption equations. At issue are a number of aspects related to the error distribution: (a) dependence between error terms, (b) heteroscedasticity, and (c) shape of the distribution. In previous studies of alcohol consumption, the error terms were found to be independent (Blaylock and Blisard 1993), while evidence of nonnormality and heteroscedasticity of errors has been reported (Yen 1994). The purpose of this paper is to evaluate determinants of alcohol expenditures and to explain important and recent empirical developments and considerations in modeling consumer expenditure behavior. We investigate household expenditures on alcohol, using a nonnormal and heteroscedastic double-hurdle model and accommodate skewness in the error terms by incorporating the inverse hyperbolic sine (IHS) transformation (Burbidge, Magee, and Robb 1988) in the dependent variable. …
67 citations
Posted Content•
TL;DR: In this article, a latent-class tobit model is proposed to estimate cardinal, segment-level preference functions based on consumers' preference ratings for product concepts considered worth adding to consumers' self-explicated consideration sets.
Abstract: The authors model product consideration as preceding choice in a segment-level conjoint model. They propose a latent-class tobit model to estimate cardinal, segment-level preference functions based on consumers' preference ratings for product concepts considered worth adding to consumers' self-explicated consideration sets. The probability with which the utility of a product profile exceeds an unobserved threshold corresponds to its consideration probability, which is assumed to be independent across product profiles and common to consumers in a segment. A market-share simulation compares the predictions of the proposed model with those obtained from an individual-level tobit model and from traditional ratings-based conjoint analysis. The authors also report simulations that assess the robustness of the proposed estimation procedure, which uses an E-M algorithm to obtain maximum likelihood parameter estimates.
49 citations
TL;DR: In this article, a latent-class tobit model is proposed to estimate cardinal, segment-level preference functions based on consumers' preference ratings for product concepts considered worth adding to consumers' self-explicated consideration sets.
Abstract: The authors model product consideration as preceding choice in a segment-level conjoint model. They propose a latent-class tobit model to estimate cardinal, segment-level preference functions based on consumers’ preference ratings for product concepts considered worth adding to consumers’ self-explicated consideration sets. The probability with which the utility of a product profile exceeds an unobserved threshold corresponds to its consideration probability, which is assumed to be independent across product profiles and common to consumers in a segment. A market-share simulation compares the predictions of the proposed model with those obtained from an individual-level tobit model and from traditional ratings-based conjoint analysis. The authors also report simulations that assess the robustness of the proposed estimation procedure, which uses an E-M algorithm to obtain maximum likelihood parameter estimates.
43 citations
TL;DR: In this paper, the authors empirically studied determinants of interstate differentials in the S&L failure rate, and the heteroskedastic-TOBIT model was used in estimation.
Abstract: This article has empirically studied determinants of interstate differentials in the S&L failure rate. The heteroskedastic-TOBIT model used in estimation turns out to perform much better than either OLS or the homoskedastic-TOBIT. Significant efficiency gain is obtained in formulating the multiplicative heteroskedasticity. Four types of variables are important: regional economic conditions, asset/liability management, regulatory structure, and politics.
30 citations
TL;DR: The authors analyzes underreporting through the two extended tobit models, p-tobit and double-hurdle models, and shows that misreporting plays an important role in underreporting.
TL;DR: In this paper, the Tobit model is itself rejected in favour of separate Probit and Truncated Regression investigations of incidence and intensity, respectively, and the results obtained indicate that the standard list of explanatory variables, which one thinks of in the context of the theoretical literature as dealing primarily with intensity, perform well in explaining incidence but very poorly in accounting for the non-limit observations.
Abstract: Variables suggested by theoretical studies of indexation are considered with a view to examining their utility in the context of the decision to index (incidence) and the desired degree of indexation when escalator clauses are put into effect (intensity). The Tobit model, which can address both of these issues, suggests that most of the effects of explanatory variables on the regressand occur by modifying the incidence of indexation, not its intensity. However, the Tobit model is itself rejected in favour of separate Probit and Truncated Regression investigations of incidence and intensity, respectively. The results obtained indicate that the standard list of explanatory variables, which one thinks of in the context of the theoretical literature as dealing primarily with intensity, perform well in explaining incidence but very poorly in accounting for the non-limit observations. The latter are influenced by bargaining power proxies such as the unemployment rate and union density; a very clear trade-off be...
01 Jan 1996
TL;DR: In this paper, a multi-disciplinary approach was developed to evaluate participation decisions in IBM's such as the Wetland Reserve Program (WRP), which extended the traditional utility maximization approach to choice behavior by including alternative measures of environmental attitudes based on the theory of reasoned action or the New Environmental Paradigm.
Abstract: Environmental policy in the United States is increasingly relying on incentive-based mechanisms (IBM). Incentive-based mechanisms, are voluntary environmental management instruments based on free market environmentalism, a framework that attempts to harness market forces. A multi-disciplinary approach was developed to evaluate participation decisions in IBM’s such as the Wetland Reserve Program. This approach extended the traditional utility maximization approach to choice behavior by including alternative measures of environmental attitudes based on the theory of reasoned action or the New Environmental Paradigm. The addition of these psychological constructs conceptually improved the utility maximization by allowing the consideration of well established determinants o f behavior. Probit and Tobit models derived from the conceptual framework were empirically tested using primary data collected via a mail survey of Louisiana wetland owners. Results presented suggest that the acreage of wetlands owned, the level of information about the WRP, respondents’ involvement in environmental organizations, education level, income, the number of people living in the household, and attitudes were significant in explaining Louisiana wetland owners’ decision to offer to participate as well as the level of participation in the WRP. The significance of attitude measures as explanatory factors suggests that a successful implementation of IBM programs depends, in addition to getting the economic incentive “right”, on properly addressing attitudinal concerns. Comparison between the
TL;DR: Four statistical methods are examined-ordinary least square multiple regression, logit and tobit models, and a multivariate procedure for comparing expected and observed outcomes-to evaluate the extent to which each method is able to both remove spurious and detect valid estimates of racial disparity when relevant control variables are added.
TL;DR: The authors examined the response of fiscally incentivized savings to income, wealth and a series of individual characteristics using a multiple-step estimation strategy, using a Tobit technique to predict the desired savings, year by year.
Abstract: The objective of this paper is to examine the response of fiscally incentivated savings to income, wealth and a series of individual characteristics. We use a multiple step estimation strategy. Our dependent variable is double censored. Therefore, we first apply a Tobit technique to predict the desired savings, year by year. Then we stack these predictions and build a system of simultaneous equations that is estimated taking into consideration individual effects and cross-equation correlations induced by panel temporal structure and measurement errors in the dependent variables.
TL;DR: In this paper, the authors considered the semiparametric efficiency bound for the Type 3 Tobit model under a symmetry restriction and investigated the relationship between this bound and those of Type 1 and 2 Tobit models and its implications.
01 Jan 1996
TL;DR: The authors used pooled, cross-sectional data in a simultaneous Probit-Tobit-Generalized Least Squares framework and extended the traditional analyses by examining alternative timing specifications, election year and nonelection year samples, and Senate committee members and nonmembers.
Abstract: Based on the economic theory of regulation, our model portrays U.S. senators as utility-maximizing agents and the American Medical Association (AMA) as a rational contributor. We use contributions from the AMA’s political arm. the American Political Action Committee (AMPAC), and roll call votes in the U.S. Senate from 1979-1992 to test for AMPAC contributions effects, to test the hypothesis that the AMA “rewards” senators for their voting behavior, and to determine the role of issue specificity. We use pooled, cross-sectional data in a simultaneous Probit-TobitGeneralized Least Squares framework and extend the traditional analyses by examining alternative timing specifications, election year and nonelection year samples, and Senate committee members and nonmembers. We obtain estimates of the standard errors by using bootstrap procedures which does not rely on asymptotic assumptions. We find no strong evidence that the AMA is successful in capturing legislation nor any evidence that the AMA rewards (or punishes) voting behavior. There is weak evidence that senators respond to future AMA contributions when voting on bills especially important to the AMA. We also find weak evidence that (future) key votes explain AMA contributions. Tests show that the underlying process generating AMA contributions is different for election years and nonelection years, and for Senate committee members and nonmembers.
01 Jan 1996
TL;DR: In this article, the use of synthetic estimation as a means of establishing need levels and expenditure requirements of small areas, by means of worked illustrations zoith a simplified example, is introduced.
Abstract: This note introduces the use of synthetic estimation as a means of establishing need levels and expenditure requirements of small areas, by means of worked illustrations zoith a simplified example. It explains a number of practical problems with the use of such indicators, in particular the reason for using linear rather than the theoretically preferable logistic and tobit regression. It compares the method of estimating need and unit cost separately, compared with a direct expenditure need approach.
TL;DR: The authors empirically examined determinants of geographic (interstate) differentials in the bank failure rate in the United States for the period 1989-1994, using the Heteroskedastic-TOBIT estimation technique.
Abstract: This study empirically examines determinants of geographic (interstate) differentials in the bank failure rate in the United States for the period 1989-1994. The study uses the Heteroskedastic-TOBIT estimation technique because 20 percent of the observations on the dependent variable are zeros and because of the need to correct for heteroskedasticity. The findings indicate that the interstate bank closing rate differential is affected by the cost of deposits, the savings and loan failure rate, the percentage of gross state product deriving from oil and natural gas extraction, the tangible capitaUasset ratio, and state regulations that restrict bank branching.