Tobit or OLS? An empirical evaluation under different diary window lengths

Researchers analysing time-use data often estimate limited dependent variable models because time spent must be nonnegative and cannot be more than the total amount of time in a given observation period. While the traditional empirical technique applied to such cases is maximum likelihood estimation of a Tobit (censored regression) model, recent debate has questioned whether linear models estimated via Ordinary Least Squares (OLS) are preferable. On the one hand, Tobit models are deemed necessary to address the significant censoring (i.e. large numbers of zeroes) typically found in time-use data, in the face of which OLS estimators would be biased and inconsistent. Yet, optimization occurs over a longer period than that covered by the typical time diary (often a day), and thus some argue that reported zeroes represent a measurement problem rather than true nonparticipation in the activity, in which case OLS would be preferred. We provide direct empirical evidence on this question using the Australian Time Use Surveys, which record time-use information for two consecutive diary days, by estimating censored and linear versions of a parental child care model for both 24-hour and 48-hour windows of observation in order to determine the empirical consequences of estimation technique and diary length.     

ESTIMATION WITH CENSORED REGRESSORS: BASIC ISSUES*

We study issues that arise for estimation of a linear model when a regressor is censored. We discuss the efficiency losses from dropping censored observations, and illustrate the losses for bound censoring. We show that the common practice of introducing a dummy variable to “correct for” censoring does not correct bias or improve estimation. We show how censored observations generally have zero semiparametric information, and we discuss implications for estimation. We derive the likelihood function for a parametric model of mixed bound‐independent censoring, and apply that model to the estimation of wealth effects on consumption.     

The Japanese Taylor rule estimated using censored quantile regressions

This paper conducts quantile regressions and obtains detailed estimates of monetary policy rules in Japan using a sample that includes recent periods of zero interest rates. Taking into account censoring and endogeneity, we compute censored quantile instrumental variable estimators and compare them with estimates from uncensored quantile regressions. The estimation results indicate that not accounting for censoring of interest rates tends to result in downwardly biased estimates. Moreover, our censored quantile regressions lead to relatively flat coefficients of inflation and insignificant coefficients of the output gap over the conditional interest rate distribution, suggesting that monetary policy in Japan may be well described by a linear rule.     

The complementarity between calls and messages in mobile telephony

This article estimates price elasticities of demand for mobile telephone calls and messages. We use panel data consisting of telephone bills to Portuguese mobile telephony consumers. In order to account for the unobserved individual heterogeneity and data censoring, we estimate a Tobit model for panel data with individual random effects. The demands for calls and messages are inelastic. Calls and messages are found to be complements.     

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 [U.S.] Bureau of Labor Statistics."     

食品安全标准促进还是抑制了我国谷物出口贸易?——基于重力模型修正贸易零值的实证研究

食品安全标准既可能促进也可能抑制国际贸易。文章在重力模型的基础上采用Tobit估计处理双边贸易零值,就食品安全标准对我国谷物出口贸易的影响进行了实证分析。结果发现:(1)进口国的食品安全标准对我国谷物出口具有抑制作用,这一结论不随进口国收入水平和食品安全标准度量指标的改变而改变。(2)我国的食品安全标准对我国谷物出口总体而言也具有抑制作用。因此,我国应当完善农产品质量安全保障体系,加强农产品进出口的标准化工作。     

Estimation and specification analysis of models of dividend behavior based on censored panel data

Dividends move in discrete jumps. Moreover, some companies pay dividends, others do not. Both these aspects necessitate the use of limited dependent variable models in the analysis of dividend behavior. Models of dividend behavior usually ignore these problems and treat dividends as a continuous variable. The present paper analyzes dividend behavior using panel data on 649 firms for 12 years (1976–1987). The model used is a censored regression model which allows for firm-specific and time effects. It is estimated using the maximum likelihood method under three different error covariance specifications. Based on specification tests, it is argued that it is important to allow for the zero observations, industry effects, and firm-specific and time effects in the estimation of models of dividend behavior.     

The asymptotic efficiency of semiparametric estimators for censored linear regression models

This paper presents numerical comparisons of the asymptotic mean square estimation errors of semiparametric generalized least squares (SGLS), quantite, symmetrically censored least squares (SCLS), and tobit maximum likelihood estimators of the slope parameters of censored linear regression models with one explanatory variable. The results indicate that the SCLS estimator is less efficient than the other two semiparametric estimators. The SGLS estimator is more efficient than quantile estimators when the tails of the distribution of the random component of the model are not too thick and the probability of censoring is not too large. The most efficient semiparametric estimators usually have smaller mean square estimation errors than does the tobit estimator when the random component of the model is not normally distributed and the sample size is 500–1,000 or more.     

地方政府卫生支出效率核算及影响因素实证研究——以中国31个省份面板数据为依据的DEA-Tobit分析

文章利用1997—2007年中国31个省份的面板数据,在DEA-Tobit两阶段分析框架下研究了中国地方政府的卫生支出效率。首先,通过数据包络分析方法(DEA)核算了各省份政府卫生支出的综合技术效率、纯技术效率和规模技术效率。在此基础上,利用受限因变量Tobit模型对效率得分与其影响因素之间的关系进行了实证研究。研究结果表明:中国地方政府的卫生支出效率存在显著的地区差异,而人口密度、居民受教育水平、人均GDP、财政分权和医疗体制改革等社会、经济和政策变量则是造成这种效率差异的重要原因。     

IPO发行价溢价异象与投资者情绪研究

以2003—2011年沪深两市所有A股作为样本进行研究可以发现,IPO发行价与IPO溢价之间存在显著的负相关关系,IPO发行价越低,IPO溢价现象越明显,称之为“发行价溢价异象”。采用受限因变量Tobit模型对投资者结构进行分析又可发现,IPO发行价越低,个人投资者持股比例越高,机构投资者持股比例越低．由此表明,正是由于低发行价股票投资以个人投资者为主,个人投资者易于受到投资者情绪影响,高估股票价值,增加非理性投资需求,寸导致了“发行价溢价异象”产生。     

The degree of monopsony power in agricultural labour markets,and the impact of the agricultural minimum wage: an application to craft workers in England and Wales

This paper uses cross-section data to examine the determinants of wages for agricultural craft workers over the period 1991–1994. Using limited dependent variable models to take account of the censoring in the data resulting from the minimum wage, we have investigated the degree to which agricultural and broader labour markets are integrated, the impact of tied housing on wages, and the extent to which the minimum wage truncates the wage distribution.     

Generalized spatial autocorrelation in a panel-probit model with an application to exporting in China

This paper proposes a generalized spatial panel-data probit model with spatial autocorrelation of the dependent variable, the time-invariant individual shocks, and the remainder disturbances. It proposes its estimation with a Bayesian Markov chain Monte Carlo procedure. Simulation results show that the proposed estimation method performs well in small- to medium-sized samples. This method is then applied to the analysis of export-market participation of 1451 Chinese firms between 2002 and 2006 in the prefecture-level city of Wenzhou in the province of Zhejiang. Empirical results show that two of the three forms of the hypothesized spatial autocorrelation are significant, namely the spatial lag for the dependent variable and the time-invariant firm-specific shocks, but not the time-variant shocks. Ignoring any of these significant spatial effects would lead to misspecification.     

The impact of software piracy on inclusive human development: evidence from Africa

The study examines the effect of software piracy on inclusive human development in 11 African countries for which software piracy data is available for the period 2000–2010. The empirical evidence is based on instrumental variable panel Fixed Effects (FE) and Tobit models in order to control for the unobserved heterogeneity and limited range in the dependent variable. The modeling exercise is based on the inequality adjusted human development (IHDI) and its constituents. The following main findings are established. First, from the FE regressions, software piracy consistently improves the IHDI and its constituents. Within this framework, the positive relationship between inclusive human development and software piracy is driven by all its constituents. Second, for Tobit regressions, the positive relationship between software piracy and inclusive human development is confirmed exclusively in IHDI and literacy specifications. Within the latter framework, the positive relationship between software piracy and inclusive human is driven fundamentally by the literacy rate. Policy implications are discussed.     

The lower bound and the causes of monetary policy inertia: evidence from Sweden

In 2009, in the midst of a global recession, Sweden’s Riksbank approached a lower bound on nominal interest rates. This encounter with the lower bound provides a natural experiment for investigating the causes of monetary policy inertia. To exploit this experiment, we estimate Taylor rules with Tobit specifications that permit both interest rate smoothing and persistent shocks (serial correlation) as explanations for inertia. The interest rate smoothing hypothesis leads to a specification in which lagged actual values of the dependent variable appear on the right-hand side of the Taylor rule, while the persistent shocks hypothesis leads to a specification in which lagged values of an unobserved latent dependent variable appear on the right-hand side of the Taylor rule. The divergence of actual and latent dependent variables that occurs at the lower bound provides leverage in distinguishing the two hypotheses. For a conventional Taylor rule, we find evidence of both sources of inertia. For a modified Taylor rule that includes a measure of financial stress, our evidence suggests that interest rate smoothing is the principal source of monetary policy inertia.     

地方政府公共文化支出效率及影响因素——基于广东21个地级市的实证分析

在DEA-Tobit两阶段分析框架下,使用2000～2009年广东21个地级市的面板数据,研究了广东地方政府公共文化支出效率。通过数据包络分析方法(DEA)评价各地方政府公共文化支出的综合效率、纯技术效率和规模技术效率,利用受限因变量Tobit模型对效率评价结果与其影响因素之间的关系进行了实证研究。研究结果表明,广东地方政府的公共文化支出效率存在显著的差异,而造成这种效率差异的重要原因主要包括财政分权和文化政策等政策变量、人均GDP、人口密度和居民受教育水平等经济及社会因素。     

Expenditure elasticities of the demand for leisure services

Although some research has already focused on the analysis of expenditure elasticities of leisure demand, some shortcomings with regard to the content and the underlying theoretical model as well as the applied methods exist. This article aims at avoiding these problems to provide consistent derivatives of leisure service expenditure elasticities. Therefore, a regular demand system is derived from microeconomic duality theory. To implement leisure specific demand factors (i.e. demand- and supply-based sports and recreational opportunities as well as sports and recreational preferences) while still being consistent with neoclassical demand theory, the basic model is extended by applying the demographic translation framework. Data of the continuous household budget survey (n?=?7724) from Germany is used for the estimation of the derived demand system. It is shown how sensitive the results are depending on the applied (censored) regression model: 16 out of 18 analysed services are indicated as luxury goods based on the findings of the Tobit model type I but as necessities based on the findings of the Tobit model type II. Possible implications are presented and discussed.     

Efficiency of two-step estimators for censored systems of equations: Shonkwiler and Yen reconsidered

This study analyses a parametric estimator for a system of equations with limited dependent variables that was recently proposed. Its performance is compared with those of alternative estimation procedures using Monte Carlo methods. The comparison shows that this new estimator is less efficient for a wide range of parameter regions than multivariate generalizations of the classical Heckman model. This result can be explained by its variance depending on the squared conditional mean of the dependent variables. Additionally, it turns out that within the class of generalized Heckman estimators, rather simple ones display the best performance.     

Stock return predictability: the role of inflation and threshold dynamics

This paper argues that the nature of stock return predictability varies with the level of inflation. We contend that the nature of relations between economic variables and returns differs according to the level of inflation, due to different economic risk implications. An increase in low level inflation may signal improving economic conditions and lower expected returns, while the opposite is true with an equal rise in high level inflation. Linear estimation provides contradictory coefficient values, which we argue arises from mixing coefficient values across regimes. We test for and estimate threshold models with inflation and the term structure as the threshold variable. These models reveal a change in either the sign or magnitude of the parameter values across the regimes such that the relation between stock returns and economic variables is not constant. Measures of in-sample fit and a forecast exercise support the threshold models. They produce a higher adjusted R², lower MAE and RMSE and higher trading related measures. These results help explain the lack of consistent empirical evidence in favour of stock return predictability and should be of interest to those engaged in stock market modelling as well as trading and portfolio management.     

The Tobit cointegrated vector autoregressive model: An application to the currency market

The Tobit cointegrated vector autoregressive model proposed in this study extends the existing methodology by allowing the censored variable to be nonstationary. The approach requires deriving the distribution of the cointegration rank test and simulating new critical values. The empirical application refers to the currency market. It has confirmed that the exchange rate is driven by four main forces: inflation, terms of trade, the perception of the country-specific risk, and the state of the currency market. Temporary disequilibria in the currency market arise not only from the “fundamental” factors, but also from the contagion effect.