Showing papers on "Tobit model published in 1997"

On the Determinants of Fiscal Centralization: Theory and Evidence

Abstract: This paper presents a simple positive model that unifies most of the results of the normative literature on fiscal federalism. The model describes an economy characterized by two levels of government, one public good, and a private good. The predictions of the model are tested by using a new set of measures of fiscal centralization. The main findings are that country size, income per capita, ethnic fractionalization, and level of democracy are negatively correlated with the degree of fiscal centralization. The model is tested using OLS, Tobit, and semi-parametric estimators. The paper also shows that the set of variables included in the regression are helpful to predict changes in the level of centralization.

Saving Accounts versus Stocks and Bonds in Household Portfolio Allocation

Abstract: We study the structure of household portfolios of financial wealth by analyzing both the determinants of total financial wealth and the choice between risky (stocks and bonds) and riskfree assets (saving accounts). The econometric specification is a generalized trivariate Tobit model, estimated on a cross section of 3,077 households in the Netherlands in 1988. We account for endogeneity of financial wealth and for selectivity due to nonreporting. Results show that the level of financial wealth and the marginal tax rate are major determinants of the allocation between riskfree and risky assets.

Privatization and economic growth: an empirical investigation from a sample of developing market economies

Abstract: This paper explores the economic determinants of privatization programmes and the macroeconomic impact of this institutional reform on the growth rate of the GDP. A sample of thirty five developing market economies is considered over the 1988-92 period. Probit and Tobit models are used to identify the determinants of successful privatization programmes. Then regressions are run where probabilities (Probit) and expected values (Tobit) are successively included as additional regressors beside the potential influence of other economic policy variables. A significant positive effect is highlighted, giving evidence of the importance of divestiture. The economic effect is found to be stronger when this institutional reform took place in industry or infrastructure.

Acquisition versus Greenfield Investment: The Location and Growth of Japanese Manufacturers in the United States

Abstract: O´ HUALLACHA´IN B. and REID N. (1997) Acquisition versus greenfield investment: the location and growth of Japanese manufacturers in the United States, Reg. Studies 31, 403‐416. Investigating foreign direct investment by principal modes of entry helps explain the location of Japanese manufacturers in the United States. Japanese greenfield start-ups and acquired plants had distinct interregional distributions in different decades. We use ordinary least squares and tobit regression analyses to account for interstate variation in the levels of Japanese acquired and greenfield establishment in 1979, 1989 and 1992 and their expansion in the 1970s, 1980s and early 1990s. The greenfield plants propelled the formation of automobile-based industrial complexes in the Midwest and Southeast regions. Greenfield investors also avoided states with strong unions and, over time, they became unconstrained by the general distribution of American manufacturing. Although supplies of procurable assets consistently constrained ...

Willingness to pay for rBST-free milk: a two-limit Tobit model analysis

Abstract: The labelling of dairy products containing milk from rBST-treated cows has rapidly emerged as an important issue in the US due to the controversy between state and federal legislation and between consumer and industrial groups. This paper evaluates consumer willingness to pay (WTP) for rBST-free milk using data collected from a consumer survey. Estimation results of a two-limit Tobit model suggest that a majority of consumers are willing to pay a premium for rBST-free milk and the WTP is affected by several sociodemographic factors such as income and attitudes toward the use of rBST.

Health care policy evaluation using longitudinal insurance claims data: An application of the Panel Tobit estimator

Abstract: The British Columbia Ministry of Health provides enhanced prescription drug insurance coverage to residents aged 65 and older. This exogenous change in the effective price of prescription drugs is used to investigate aspects of the drug use by seniors. Three sets of issues are of interest. First, what is the effect of enhanced insurance coverage on drug use and programme costs once drugs are provided free of charge? Second, is this effect permanent, or transitory? Third, are any increases in use observed concentrated among those with lower incomes? Longitudinal administrative claims payment data on 18 000 seniors over the period 1985–92 are used. All individuals in the sample turned 65 at some point and therefore became eligible for subsidized prescription drugs. Health status information is not collected; instead, health status is treated as an individual-specific fixed endowment, subject to a common rate of decay. Estimation is complicated by censoring of real drug expenditures for those under 65, rendering ‘first differencing’ methods invalid. A semi-parametric fixed effects Tobit estimator is used instead. For most individuals, the extension of insurance does not permanently increase drug use. Males with lower income were the exception. Little evidence of transitory effects to insurance coverage was found. Finally, the extension of insurance has made only a minor contribution to growth in seniors' drug use, relative to secular growth in drug use over time. © 1997 John Wiley & Sons, Ltd.

A note on the use of the Tobit approach for tests scores with floor or ceiling effects

Abstract: Muthen (1989) developed the Tobit approach to deal with censored variables in structural equation modelling. We performed a simulation study to evaluate the effectiveness of this approach for the type of censoring that occurs for sums of Likert items. Results indicated that estimated correlations and factor loadings were substantially biased downwards, and that goodness-of-fit tests were inaccurate. Furthermore, the magnitude of distortion depended on characteristics of the scales such as the amount of censoring, scale length, the number of response categories of the items, and the size of the correlation. When we adjusted the Tobit approach by assuming that all scores at the lower end of the scale were affected by censoring and not just the lowest score, results improved considerably. In addition, this adjusted Tobit approach was superior to alternative procedures that deal with censoring and was easy to implement using the existing software.

Efficiency differences of Finnish senior secondary schools: an application of DEA Tobit-analysis

Abstract: Summary form only as given. The authors study efficiency differences among Finnish senior secondary schools by data envelopment analysis (DEA). The results also show that the rankings of schools by matriculation examination scores (an output variable) differed markedly from rankings by efficiency. As a second stage after DEA analysis, they explained the degree of inefficiency by a statistical Tobit model. Somewhat surprisingly, private schools were inefficient relative to public schools, and grant ratios, i.e. ratios of state grants to accepted or actual educational expenditures, had in some models a positive and significant effect on efficiency.

Monte carlo evidence on the robustness of conditional moment tests in tobit and probit models

Abstract: This paper numerically examines the size robustness of various conditional moment tests in misspecified tobit and probit models. The misspecifications considered include the incorrect exclusion of regressors, ignored heteroskedasticity and false distributional assumptions. An important feature of the experimental design is that it is based on an existing empirical study and is more realistic than many simulation studies. The tests are seen to have mixed performance depending on both the original null hypothesis being tested and type of misspecification encountered.

Time on market and sales price of residential housing: A note

Abstract: This note extends the study of Kalra and Chan (1994). A simultaneous TOBIT equations model is established to address the simultaneity nature of time on the market (TOM) and sales price (SP) in the presence of censored sample bias. We find that both TOM and SP are positively related to each other.

Rent-seeking and political tenure: First estimates

Abstract: This paper computes a measure of rent-seeking based on costs of tenure. Within a political tenure type model, a new incumbency variable is introduced and a variant to existing datasets is presented. Consistent ML rent-seeking estimates are computed for our sample of OECD democracies. The statistical advances in limited dependent variable models in particular, the decomposition of the Tobit estimates enabled us to compute rent-seeking estimates as costs of tenure the Tobit partial derivatives and to compute the impact of explanatory variables on the probability of being above the limit which we interpret as the probability of tenure.

A double-hurdle expenditure analysis of nonconsumptive wildlife related recreation

Abstract: This study utilized cross-sectional data obtained from the 1991 National Survey o f Fishing, Hunting, and Wildlife-Associated Recreation to analyze the expenditures and consumption o f nonconsumptive wildlife related recreation in the Lower Mississippi Valley Ecosystem area In the process of selecting an appropriate model most consistent with individuals’ consumption behavior associated with nonconsumptive wildlife related recreation, the tobit model and the double-hurdle model for both the primary nonresidential and residential expenditure models were evaluated. Based on the Lagrange multiplier test and the likelihood ratio test results, the double-hurdle model fit the data much better than the tobit model. The hypothesis that there are heteroscedastic problems associated with the error terms was rejected based on the likelihood ratio test result. In the primary nonresidential expenditure model, income, education, ethnicity, public lands, and forest lands had a significant effect on nonconsumptive wildlife related recreation expenditures. The total consumption was predicted to increase $0.000337 with income growth, increase $16.92 with increases in educational status, increase $16.29 when participants are Caucasian, increase $27.44 with use o f public lands, and increase $30.04 with use o f forest lands. In the primary residential expenditure model, gender, employment status, ethnicity, wildlife including birds, mammals, insects, and fish, maintaining natural areas for fish or wildlife, and visiting public parks or natural areas had a significant effect on

Tobit analysis with a natural non-response rate

Abstract: A Tobit model is considered with a natural non-response rate. The model features simultaneous determination of the standard Tobit model and a natural non-response rate. An example on the determinants of extramarital affairs is used to illustrate the method. The natural non-response rate is shown to be statistically significant in this example.

An Empirical Analysis of Determinants of Geographic Differentials in the Savings and Loan Failure Rate, 1989-1991, Using the Heteroskedastic-Tobit Model

Abstract: This study empirically examines determinants of geographic (interstate) differentials in the savings and loan failure rate in the United States from 1989 to 1991. The study uses the heteroskedastic-Tobit estimation technique because 12 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 S&L failure rate differential is affected by the average annual growth rate in gross state product, the percentage of gross state product deriving from oil and natural gas extraction, the S&L cost of deposits, the volatility of the S&L cost of deposits, and the remaining average balance on outstanding fixed-rate mortgage loans at S&Ls.

Effects of Sample Selection on Estimates of Economic Impacts of Outdoor Recreation

Abstract: Estimates of the economic impacts of recreation often come from spending data provided by a self-selected subset of a random sample of site visitors. The subset is frequently less than half the onsite sample. Biased vectors of per trip spending and impact estimates can result if self-selection is related to spending patterns, and proper corrective procedures are not employed. This paper shows a method for accounting for both sample selection and the censored nature of reported expenditures, via a Tobit model with sample selection. Results from a sample of visitors to Cumberland Island National Seashore indicate a naive (uncorrected) approach overestimates per trip visitor spending by 15 percent and economic impacts to industrial output by 10 percent.

A tobit model with garch errors

Abstract: In the context of time series regression, we extend the standard Tobitmodel to allow for the possibility of conditional heteroskedastic error processes of the GARCH type.We discuss the likelihood function of the Tobit model in the presence of conditionally heteroskedastic errors.Expressing the exact likelihood function turns out to be infeasible, and we propose anapproximation by treating the model as being conditionally Gaussian. The performance of theestimator is investigated by means of Monte Carlo simulations. We find that, when the errorterms follow a GARCH process, the proposed estimator considerably outperforms the standardTobit quasi maximum likelihood estimator. The efficency loss due to the approximationof the likelihood is finally evaluated.

Valuing Risk Tradeoffs and Voluntary Insecticide Reduction

Abstract: Fixed search costs associated with locating and purchasing pecans invalidate the Tobit model. Factors such as perceived quality, ease of purchase, and familiarity with marketing outlets influence the fixed costs of pecan purchases. These factors have differing impacts on the probability of purchasing and the amount purchased based on the Heckman model. Failure to apply self- selectivity corrections produces misleading assessments of key variables influencing pecan purchases.

A model with mixed binary responses and censored observations

Abstract: In this paper, we study the maximum likelihood estimation of a model with mixed binary responses and censored observations. The model is very general and includes the Tobit model and the binary choice model as special cases. We show that, by using additional binary choice observations, our method is more efficient than the traditional Tobit model. Two iterative procedures are proposed to compute the maximum likelihood estimator (MLE) for the model based on the EM algorithm (Dempster et al, 1977) and the Newton-Raphson method. The uniqueness of the MLE is proved. The simulation results show that the inconsistency and inefficiency can be significant when the Tobit method is applied to the present mixed model. The experiment results also suggest that the EM algorithm is much faster than the Newton-Raphson method for the present mixed model. The method also allows one to combine two data sets, the smaller data set with more detailed observations and the larger data set with less detailed binary choice observa...

Forecasting Nonpayment Behavior of Customers for a Mail Order Company

Abstract: Tobit regression works well for forecasting the nonpayment behavior of customers .. longer the delinquency, the greater the chances of nonpayment ... demonstrates the results by using the real data of a mail order company. Predicting nonpayment behavior is of considerable interest to industries such as banking and retail. Corporations are interested in many issues surrounding nonpayment, ranging from various levels of delinquency all the way to write-offs. Typical forecasting solutions involve the construction of credit scoring models using either standard linear or logistic regression which attempt to classify each individual in one of two categories. If the payment behavior was acceptable (or good ) over a certain span of time, then the dependent variable in the model might take on a value of one. If behavior was unacceptable (or bad), it would be assigned a zero value. The dependent variable is then measured against information from the credit bureau and/or the corporation to form a statistical profile to predict future delinquencies. These procedures, however, often stop at the classification step, that is, acceptable or unacceptable behavior. Sometimes, the two-way classification process (zero or one) is not enough. For example, as in the case of a mail order company, a customer may default by 100%, 75% or 25% on his/her account, or have a completely satisfactory payment record. In that case, the forecast variable, write-offs, is no longer a dichotomous variable taking on values of zero or one, but a continuous variable. It is also bounded between zero and one hundred, often with many observations at one extreme. Because of the clustering problem around the limits, ordinary least square estimates are biased, sometimes producing even negative predicted values. This bias can distort the true relationship between the nonpayment rate and the profile characteristics. The purpose of this paper is to compare the predictability of Tobit regression which can handle a continuos dependent variable with other models which can not. In the econometric literature, the models that can handle these types of continuous dependent variables are called models with limited dependent variables. In statistics, they are often called truncated or censored models. The following discussion is centered around a credit forecasting application which attempts to predict nonpayment rates which are bounded between 0 and 100 with a large cluster of observations near the lower bound. Benchmark comparisons are made with two different neural network algorithms as well as ordinary least squares. In 1958 James Tobin wrote a paper making the best of both worlds - least squares regression and models using dichotomous dependent variables. He referred to many applications where this type of model could be applied. In particular, his paper centered around Engel curves curves which show the relationship between household expenditures and various levels of income. For example, in any given year a large number of households would have made zero purchases on automobiles. Those who did purchase would have a wide variety of expenditure levels. Tobit regression is an extension of Probit regression which is based on the cumulative normal distribution. The cumulative normal distribution is viewed as a desirable transformation since it relates a variable (number of standard deviations from the mean) which has a range from minus infinity to plus infinity to another variable (a probability) which has a range from zero to one. In this manner an unconstrained variable can be transformed into a new variable that is bounded. The procedure is unique in the sense that it yields insight into two questions: ( 1 ) What are the chances that an observation will be over a predefined limit? (in this case zero) and (2) if it is over the limit, what will be the predicted value? For the procedure used in Tobit regression see Appendix 1. The regression algorithm, based on the 1958 Tobin article, is available in a number of computer programs such as SHAZAM and LIMDEP. …

A microsimulation model of Dutch tourist behaviour an exploration of Dutch panel data

Abstract: In this paper the holiday behaviour of the Dutch population is described, using Dutch panel data. Several econometric models have been applied and the results of these models have been compared. A Tobit model with heteroscedasticity has been selected as the best specification, and the results of this specific model have been used to construct a microsimulation model (MSM). The paper shows that microsimulation is a rich instrument to describe the (changes in) participation probability and holiday expenditure of Dutch households.

Demand and Quality Uncertainty in Pecan Purchasing Decisions

Abstract: Fixed search costs associated with locating and purchasing pecans invalidate the Tobit model. Factors such as perceived quality, ease of purchase, and familiarity with marketing outlets influence the fixed costs of pecan purchases. These factors have differing impacts on the probability of purchasing and the amount purchased based on the Heckman model. Failure to apply self- selectivity corrections produces misleading assessments of key variables influencing pecan purchases.

Cigarette Consumption by U.S. Men

Abstract: Heckman’s sample selection model is derived from maximization of a utility function with stigma and health concern effects. The model outperforms the Tobit model and suggests that social stigma and health concerns play important roles in cigarette consumption.