Showing papers on "Tobit model published in 2007"
TL;DR: This paper presents two alternative approaches to second stage DEA, the results of which are compared with the tobit approach through a case study for the Danish fishery, and it is shown that OLS may actually in many cases replace tobit as a sufficientsecond stage DEA model.
496 citations
TL;DR: In this paper, the authors examined the determinants of corporate dividend policy in Jordan using Tobit specifications and found that the proportion of stocks held by insiders and state ownership significantly affect the amount of dividends paid.
Abstract: This paper examines the determinants of corporate dividend policy in Jordan. The study uses a firm‐level panel data set of all publicly traded firms on the Amman Stock Exchange between 1989 and 2000. The study develops eight research hypotheses, which are used to represent the main theories of corporate dividends. A general‐to‐specific modeling approach is used to choose between the competing hypotheses. The study examines the determinants of the amount of dividends using Tobit specifications. The results suggest that the proportion of stocks held by insiders and state ownership significantly affect the amount of dividends paid. Size, age, and profitability of the firm seem to be determinant factors of corporate dividend policy in Jordan. The findings provide strong support for the agency costs hypothesis and are broadly consistent with the pecking order hypothesis. The results provide no support for the signaling hypothesis.
284 citations
TL;DR: This paper examined whether open market share repurchases deter takeovers and found that repurchase activity is inversely related to firm size, consistent with smaller firms having greater information asymmetry, and is related to temporary, but not permanent, cash flows.
Abstract: This paper examines whether open market share repurchases deter takeovers. We model pre-repurchase takeover probability as a latent variable and examine its impact on the firm's decision to repurchase shares. Given specification tests reject the Tobit model, we turn to the censored quantile regression method of Powell (1986, Journal of Econometrics 32, 143–155). We find a significantly positive relation between open market share repurchases and takeover probability, and we reconcile empirical findings in previous studies that contradict predictions. Repurchase activity is inversely related to firm size, consistent with smaller firms having greater information asymmetry, and is related to temporary, but not permanent, cash flows.
253 citations
TL;DR: In this article, a sample of 256 small firms who applied for bank loans was used to investigate the extent to which "innovativeness" is associated with a lower level of loan application success.
Abstract: Drawing upon a sample of 256 small firms who applied for bank loans, the current paper is concerned with the extent to which 'innovativeness' is associated with a lower level of loan application success. The paper records the proportion of loan successfully applied for and estimates a series of tobit models utilising a number of proxy measures for innovation (in terms of inputs, outputs, and commercial significance to the firm) and incorporating standard controls. In general, the models suggest (as anticipated) that the most innovative firms are less successful in loan markets than their less innovative peers - though there is some variation by proxy. Moreover, there is tentative evidence that 'a little innovation may be a good thing'.
236 citations
TL;DR: In this paper, a joint Tobit-ordered probit econometric model was used to address the possible endogeneity of household spending on private tutoring in the Vietnamese households.
202 citations
TL;DR: In this article, the authors used a contingent valuation survey to elicit Swedish households' willingness to pay (WTP) to avoid power outages and found that the WTP depends positively on the duration of the outages, and that WTP is significantly higher for unplanned outages.
Abstract: Using a contingent valuation survey, we elicit Swedish households' willingness to pay (WTP) to avoid power outages. In the study respondents are asked to state their WTP for avoiding nine different types of outages. We therefore apply a random parameter Tobit model since there is cross-section al heterogeneity and a proportion of zero responses. Based on the estimations, we find that the WTP depends positively on the duration of the outages, and that WTP is significantly higher for unplanned outages. The overall variation in the WTP due to observed heterogeneity in housing and socio-economic variables is small compared to the pure effects of power outages. Policy implications of those findings are discussed.
135 citations
TL;DR: In this article, the relationship between farm financial structure and technical efficiency in Central and Eastern European farming during the transition to a market economy is analyzed, which indicates substantial differences in the way in which the different management structures are treated by lenders.
Abstract: This paper provides the first analysis of the relationship between farm financial structure and technical efficiency in Central and Eastern European farming during the transition to a market economy. First, technical efficiency scores for a sample of Czech farms are computed with data envelopment analysis, accounting for sampling variability by using the Simar and Wilson bootstrapping procedure. The confidence intervals show that corporate livestock farms are the most homogenous in terms of technical efficiency. Second, a Tobit model is applied for the investigation of factors, including financial considerations, which influence the variation in technical efficiency. The potential endogeneity of the financial variables in the explanation of efficiency is examined, following Smith and Blundell. The Amemiya estimator is used for those cases which are revealed as endogenous. The analysis indicates substantial differences in the way in which the different management structures are treated by lenders. De novo, individual farms are subject to stringent evaluation of their loan applications, which is not the case for the successors of the former collective and state farms.
113 citations
Posted Content•
TL;DR: cmp as discussed by the authors is a mixed process model with hierarchical random effects, which is based on the Stata commands probit, ivprobit, treatreg, biprobit, tetrachoric, oprobit and opprobit.
Abstract: cmp estimates multi-equation, mixed process models, potentially with hierarchical random effects. "Mixed process" means that different equations can have different kinds of dependent variables. The choices are: continuous (like OLS), tobit (left-, right-, or bi-censored), probit, ordered probit or fractional probit. "Conditional" means that the model can vary by observation. An equation can be dropped for observations for which it is not relevant--if, say, a worker retraining program is not offered in a city then the determinants of uptake cannot be modeled there. Or the type of dependent variable can vary by observation. A dependent variable in one equation can appear on the right side of another equation. Such dependencies must have a recursive structure if the dependencies are on censored variables as observed, meaning that they split the equations into stages. If the dependencies are on (latent) linear dependent variables, they can be recursive or simultaneous in structure. So cmp can fit many SUR, simultaneous equation, and IV models. cmp's modeling framework therefore embraces those of the official Stata commands probit, ivprobit, treatreg, biprobit, tetrachoric, oprobit, mprobit, asmprobit, asroprobit, tobit, ivtobit, cnreg, intreg, truncreg, heckman, heckprob, xtreg, xtprobit, xttobit, xtintreg, in principle even regress, sureg, and reg3. It goes beyond them in offering far more flexibility in model construction. The routine runs under Stata 10 or later, faster under Stata 11.2 or later.
109 citations
TL;DR: The LCM and TPM with a log-transformed dependent variable are superior to other approaches in handling data with ceiling effects when predicting EQ-5D index scores from the MOS-HIV Health Survey.
Abstract: Health-related quality of life (HRQoL) is increasingly used to measure health status. Health profile and preference-based approaches are two types of methods frequently used to measure HRQoL. Health profiles tap multiple dimensions of health. Preference-based approaches quantify the extent to which an individual prefers specific health states using a single index from 0 (death) to 1 (full health). The MOS-HIV Health Survey (Wu et al. 1997) and EQ-5D (The EuroQol Group 1990) are typical health profile and preference-based instruments, respectively.
The availability of preference-based HRQoL data is limited because direct assessment is not always feasible. Therefore, researchers have attempted to predict preference-based scores from health profiles, generally using ordinary least squares (OLS) regression (Fryback et al. 1997; Franks et al. 2004).
A common phenomenon observed in both health profile and preference-based measures is that the score distribution tends to be skewed to the left. This is called the “ceiling effect.” When data have pronounced ceiling effects, the use of OLS regression violates the statistical requirement for linearity of conditional expectation, leading to inaccurate predictions of preference-based scores and inaccurate identification of predictor variables (Kennedy 1998). Several methods have been proposed to address ceiling effects, such as Tobit models, the censored least absolute deviation (CLAD) approach, two-part models (TPM), and latent class models (LCM) (Greene 2003).
The Tobit and CLAD models are preferable to OLS in the presence of a ceiling effect or censored dependent variable. The Tobit model assumes that there is a true unobservable dependent variable that depends linearly on independent variables. The observable dependent variable is defined to be equal to the unobserved variable whenever the true unobserved variable is greater than or equal to the ceiling value. The Tobit model assumes that the error terms are normally distributed with uniform variance (Greene 2003). Unfortunately, evidence suggests that the distribution of utility scores does not follow those assumptions (Austin 2002a). The CLAD approach is based on an assumption that the median will be more robust to ceiling effects than the mean. The coefficients are estimated so as to minimize the sum of the absolute deviations from the regression line (Arabmazar and Schmidt 1981; Johnston 1997).
A TPM approach is a more flexible way to address the fact that the HRQoL scores usually have mixed distributions that contain a disproportionate mass of observation at the ceiling and a left-skewed distribution of observations not at the ceiling. The “two parts” model (1) the probability that the dependent variable (e.g., EQ-5D index) can attain the full score and (2) the distribution if the full score is not attained. Both parts are combined in a joint model to predict the dependent variable using the subject's predictors.
The LCM is another approach to address data with ceiling effects. This approach is considered by some to be more flexible than the TPM because LCM can handle situations where there are two (or more) classes, with one class being very healthy (i.e., higher utility index) and another group being less healthy (i.e., lower index), as opposed to a TPM which models perfectly healthy (index=1) versus other than perfectly healthy (index <1) (Deb and Trivedi 1997, 2002).
Although several studies analyzing HRQoL have used the Tobit model to address ceiling effects (Austin, Escobar, and Kopec 2000; Austin 2002b; Clarke, Gray, and Holman 2002; Sullivan, Lawrence, and Ghushchyan 2005), only three have applied the CLAD (Austin 2002b; Clarke, Gray, and Holman 2002; Sullivan, Lawrence, and Ghushchyan 2005) and none have used a TPM or LCM.
The purpose of this study was to compare these statistical methods to predict preference-based HRQoL scores from health profiles for persons with HIV, taking into account ceiling effects. We compared performance (predictive accuracy and error) of these methods to a standard regression model. As an example, we estimated the EQ-5D index from the MOS-HIV. We used a cross-validation method to evaluate the performance of each statistic method.
106 citations
TL;DR: An alternative statistical approach, considered more appropriate for many microbial datasets is suggested, is suggested and its application in a study on the effectiveness of a decontamination process for chicken carcases is demonstrated.
77 citations
01 Jan 2007
TL;DR: In this article, gender differences in the allocation of time between market work, domestic work and leisure over the life-cycle were analyzed using the Time Use National Surveys (TUNS).
Abstract: This article analyses the extent to which changes in household composition over the life course affect the gender division of labour. It identifies and analyses cross-country disparities between France, Italy, Sweden and United States, using most recent data available from the Time Use National Surveys. We focus on gender differences in the allocation of time between market work, domestic work and leisure over the life-cycle. In order to map the lifecycle, we distinguish between nine key cross-country comparable life stages according to age and family structure such as exiting parental home, union formation, parenthood, and retiring from work. By using appropriate regression techniques (Tobit with selection, Tobit and OLS), we show large discrepancies in the gender division of labour at the different life stages. This gender gap exists in all countries at any stage of the life course, but is usually smaller at the two ends of the age distribution, and larger with parenthood. Beyond social norms, the impact of parenthood on time allocation varies across countries, being smaller in those where work-family balance policies are more effective and traditionally well established.
Journal Article•
TL;DR: This paper proposes a method to implement maximum likelihood estimation of the dynamic panel data type 2 and 3 tobit models using a two-dimensional indefinite integral evaluated using "two-step" Gauss-Hermite quadrature.
Abstract: This paper proposes a method to implement maximum likelihood estimation of the dynamic panel data type 2 and 3 tobit models. The likelihood function involves a two-dimensional indefinite integral evaluated using "two-step" Gauss-Hermite quadrature. A Monte Carlo study shows that the quadrature works well in finite sample for a number of evaluation points as small as two. Incorrectly ignoring the individual effects, or the dependence between the initial conditions and the individual effects results in an overestimation of the coefficients of the lagged dependent variables. An application to incremental and radical product innovations by Dutch business firms illustrates the method.
DOI•
01 Sep 2007
Abstract: Travel is the price we pay to be with others, at least it is a very substantial part of the generalised cost of meeting them. If we accept this proposition, and consider that the vast majority of travel serves activities with others, then one has to wonder, why the social content of activities and the constraints arising from coordination with others has received so little attention so far in trans port research. This paper supplements the previous extensive qualitative research by the analysis of a large quantitative survey of egocentric social networks in Zurich. The exact geocoding of the ego’s and alteri’s home location builds the basis for this paper’s analysis of the size and structure of social network geographies and the distances involved. The size and structure of so cial network geographies is measured by the confidence ellipse of the spatial distribution of the members in egocentric social networks. Young, well educated people with a low and middle in come and an eventful work and educationbiography tend to maintain bigger social network geographies. These parameters as well as a low ratio between the minor and the main axis and a strong eastwest direction are the main determinants for big social network geographies
TL;DR: In this paper, the authors examined the cost efficiency of 16 Greek cooperative banks over the period 2000-2004 and found that the dominant source of cost inefficiency is allocative rather than technical.
Abstract: We follow a two-stage procedure to examine for the first time the cost efficiency of Greek cooperative banks. Our sample consists of 16 banks over the period 2000-2004. We first use data envelopment analysis (DEA) to estimate the technical, allocative and cost efficiency for each bank in sample. Then, we use Tobit regression to determine the impact of internal and external factors on banks’ efficiency. The results of DEA indicate that Greek cooperative banks could improve their cost efficiency by 17.7% on average as well as that the dominant source of cost inefficiency is allocative rather than technical. The results of Tobit regression indicate that size has a positive impact on all measures of efficiency. However, the impact of capitalization, branches and ATMs depends on the efficiency measure and whether we control for market conditions or not. GDP per capita has a negative and significant impact on all measures of efficiency, while unemployment rate has also a negative and significant impact on technical and cost efficiency although not on allocative efficiency. Finally, banks operating in regions with higher disposal income of households in relation to the total disposal income of households in Greece are more efficient in terms of allocative and cost efficiency.
Posted Content•
TL;DR: In this paper, gender differences in the allocation of time between market work, domestic work and leisure over the life-cycle were identified and analyzed between France, Italy, Sweden and United States, using most recent data available from the Time Use National Surveys.
Abstract: This article analyses the extent to which changes in household composition over the life course affect the gender division of labour. It identifies and analyses cross-country disparities between France, Italy, Sweden and United States, using most recent data available from the Time Use National Surveys. We focus on gender differences in the allocation of time between market work, domestic work and leisure over the life-cycle. In order to map the life-cycle, we distinguish between nine key cross-country comparable life stages according to age and family structure such as exiting parental home, union formation, parenthood, and retiring from work. By using appropriate regression techniques (Tobit with selection, Tobit and OLS), we show large discrepancies in the gender division of labour at the different life stages. This gender gap exists in all countries at any stage of the life course, but is usually smaller at the two ends of the age distribution, and larger with parenthood. Beyond social norms, the impact of parenthood on time allocation varies across countries, being smaller in those where work-family balance policies are more effective and traditionally well-established.
TL;DR: The authors empirically extend the gravity model, which has been widely used to analyze volumes of trade between pairs of countries, by allowing for the inclusion of country-specific effects into the analysis and also explore the relationship between trade volumes and a given covariate via a nonparametric approach.
Abstract: This paper seeks to empirically extend the gravity model, which has been widely used to analyze volumes of trade between pairs of countries. We generalize the basic threshold tobit model by allowing for the inclusion of country-specific effects into the analysis and also show how one can explore the relationship between trade volumes and a given covariate via a non-parametric approach. We use our derived methodology to investigate the impact of a particular aspect of institutions—the enforcement of contracts—on bilateral trade. We find that contract enforcement matters in predicting trade volumes for all types of goods, that it matters most for the trade of differentiated goods, and that the relationship between contract enforcement and trade in our threshold tobit exhibits some nonlinearities. Copyright © 2007 John Wiley & Sons, Ltd.
TL;DR: In this article, the authors investigated the effects of household, farm productive characteristics and the policy variables on the demand and supply of credit in the rural credit markets using household data from India.
Abstract: The demand and supply of credit in the rural credit markets is investigated in this article using household data from India. The aim is to study the effects of household, farm productive characteristics and the policy variables on the demand and supply of credit. A type 3 Tobit model is estimated which corrects for sample selection and endogeniety bias. In addition, a generalized Double Hurdle model is estimated where the information on the household's access to credit is included to estimate the demand and supply of credit. The results suggest that the size of the operational holdings, net-wealth, dependency ratio, educational level of the household and the wages and output prices are important determinants of the demand and supply of credit for farm households. The Double Hurdle model confirms that the ‘size of land owned’ plays a crucial role in whether the household has access to a loan or not.
TL;DR: In this paper, the authors studied the change in total productivity in travel agencies in two stages, in the first stage, they break down total productivity into technically efficient change and technological change by means of data envelopment analysis.
Abstract: This article estimates changes in total productivity in travel agencies in two stages. In the first stage, it breaks down total productivity into technically efficient change and technological change by means of data envelopment analysis (DEA) applied to a representative sample of travel agencies operating in the Portuguese market. The aim of this procedure is to seek out those best practices that will lead to improved performance in the market. We rank the companies according to their change in total productivity for the period 2000-2004, concluding that some companies experienced productivity growth while others experienced a decrease in productivity. In the second stage, the inefficient scores are regressed into contextual variables of inefficiency with a bootstrapped Tobit model. The implications arising from the study are considered in terms of managerial policy.
Journal Article•
TL;DR: In this paper, the authors investigated the factors that influenced the soybean farmers' decisions about adoption and use intensity of these improved soybean varieties and found that farmers' adoption decisions were significantly influenced by soybean yield, expenditure on labour, memberships of associations and meeting groups, exposure to extension services and ecology.
Abstract: The deficient agronomic qualities of the local soybean germplasm and the need to develop varieties with desirable attributes, such as promiscuous nodulation, low pod shattering, sustained seed viability and resistance to bacterial diseases and viruses necessitated the introduction of new varieties to farmers in northern Nigeria. This study investigates the factors that influenced the soybean farmers’ decisions about adoption and use intensity of these improved soybean varieties. A multistage random sampling design was used to select a sample of 307 respondents from a list of soybean growers obtained through households listing conducted by trained enumerators. Data were collected from respondents using a structured questionnaire. The logit and Tobit regression models were used in the analysis for comparison. Empirical results reveal that the farmers’ adoption decisions were significantly (P<0.01) influenced by soybean yield, expenditure on labour, memberships of associations and meeting groups, exposure to extension services and ecology. The two models confirm that increases in each of the identified adoption characteristics will lead to increases in the probability of adoption and use intensity. However, the effect of a one-percent increase in each variable is higher for the Tobit than for the logit adoption elasticities. Moreover, the decomposed elasticity values of the Tobit model reveal that a one percentage increase in the variables would result in a higher change in the elasticity of use intensity than in the elasticity of adoption of the improved soybeans. The results have significant implications for research scientists, extension agents and policy-makers. Apart from providing justification for strengthening the extension capacities of the national research systems, the findings underscore the relevance of networks that promote farmer-to-farmer interactions in the circulation of new technologies.
Posted Content•
TL;DR: The authors examined whether open market share repurchases deter takeovers and found that repurchase activity is inversely related to firm size, consistent with smaller firms having greater information asymmetry and is related to temporary, but not permanent, cash flows.
Abstract: This paper examines whether open market share repurchases deter takeovers. We model pre-repurchase takeover probability as a latent variable and examine its impact on the firm's decision to repurchase shares. Given specification tests reject the Tobit model, we turn to the censored quantile regression method of Powell (1986). We find a significantly positive relation between open market share repurchases and takeover probability, and we reconcile empirical findings in previous studies that contradict predictions. Repurchase activity is inversely related to firm size, consistent with smaller firms having greater information asymmetry, and is related to temporary, but not permanent, cash flows.
TL;DR: In this article, a three-year window analysis together with the Data Envelopment Analysis (DEA) approach is employed to investigate the effects of mergers and acquisitions on the Singapore banking groups' efficiency.
Abstract: A three-year window analysis together with the Data Envelopment Analysis (DEA) approach is employed to investigate the effects of mergers and acquisitions on the Singapore banking groups’ efficiency The results suggest that the merger has resulted in a higher Singapore banking groups’ mean overall efficiency We do not find evidence of more efficient acquirers compared to the targets and that the acquiring banks’ mean overall efficiency tends to improve from the merger with a more efficient bank The Tobit regression results suggest that bank profitability has positive impact on bank efficiency, whereas poor loan quality has negative influence on bank performance (JEL: G21, D24)
Posted Content•
TL;DR: In this article, a method to implement maximum likelihood estimation of the dynamic panel data type 2 and 3 to-bit models is proposed, which involves a two-dimensional indefinite integral evaluated using "two-step" Gauss-Hermite quadrature.
Abstract: This paper proposes a method to implement maximum likelihood estimation of the dynamic panel data type 2 and 3 tobit models. The likelihood function involves a two-dimensional indefinite integral evaluated using "two-step" Gauss-Hermite quadrature. A Monte Carlo study shows that the quadrature works well in finite sample for a number of evaluation points as small as two. Incorrectly ignoring the individual effects, or the dependence between the initial conditions and the individual effects results in an overestimation of the coefficients of the lagged dependent variables. An application to incremental and radical product innovations by Dutch business firms illustrates the method.
TL;DR: This paper investigated why some economics departments in Australian universities are more research productive than others, and found that research productivity depends upon the human capital of department members and the department-specific conditions under which they work.
Abstract: This study investigates why some economics departments in Australian universities are more research productive than others. The hypothesis is simple: research productivity depends upon the human capital of department members and the department-specific conditions under which they work. A Tobit model is used to estimate the magnitude of the two effects. Both are found to be important. Our results help explain why a small number of departments consistently outperform the others in studies that rank Australian economics departments according to research output.
Posted Content•
TL;DR: The estimator is implemented with the lf method and supports all the usual maximum likelihood features and Computation time is an issue with many equations.
Abstract: mvtobit estimates M-equation tobit models (including bivariate tobit models). To evaluate the multidimensional integrals, mvtobit relies on Cappellari & Jenkins' mdraws and egen mvnp commands to implement the Geweke-Hajivassiliou-Keane (GHK) simulator. The estimator is implemented with the lf method and supports all the usual maximum likelihood features. Computation time is an issue with many equations. Users of Stata 9 and higher should have a look at the user command cmp.
Posted Content•
TL;DR: In this article, the authors present empirical evidence to show how socioeconomic factors affect the adoption of and investment in agrochemicals in the cocoa sector of Cote d'Ivoire.
Abstract: This paper presents empirical evidence to show how socioeconomic factors affect the adoption of and investment in agrochemicals in the cocoa sector of Cote d’Ivoire. The analysis uses primary farm-level data collected in 2002 from a nationally representative sample of more than one thousand cocoa farmers. The study describes the status of the adoption of various chemical inputs and uses a multiplicative heteroscedastic Tobit model to identify and quantify the impact of the socioeconomic environment on the incentive to invest. The results generally show that farmer, household and village characteristics are all important in explaining the farmers’ decisions. The paper concludes by outlining a number of implications for strategic targeting of farmers and locations. These should serve as entry points for a successful diffusion of efficient pest, disease and soil management programs.
TL;DR: In this article, the Tobit model was used to identify the variables affecting the decision to make contributions to personal pension plans and the amount of such contributions and found that individuals decide to invest in pension plans on complex grounds combining the wish to benefit from tax savings and to ensure they will receive supplementary income upon retirement.
Abstract: The aim of this article is to identify the variables affecting the decision to make contributions to personal pension plans and the amount of such contributions. For this purpose, we specify and estimate a Tobit model for a sample based on the 1995 Personal Income Taxpayers Panel prepared by the Institute of Fiscal Studies (Spanish Ministry of Economy and Finance) formed by 3041 taxpayers, of whom 358 made contributions to pension plans. Our results suggest that individuals decide to invest in pension plans on complex grounds combining the wish to benefit from tax savings and to ensure they will receive supplementary income upon retirement.
Posted Content•
TL;DR: In this article, a simple QML estimator based on the sum of univariate Tobit models is introduced and a Monte Carlo simulation comparing the three estimators is performed on three different sample sizes.
Abstract: Recently a number of authors have suggested to estimate censored demand systems as a system of Tobit multivariate equations employing a Quasi Maximum Likelihood (QML) estimator based on bivariate Tobit models. In this paper I study the efficiency of this QML estimator relative to the asymptotically more efficient Simulated ML (SML) estimator in the context of a censored Almost Ideal demand system. Further, a simpler QML estimator based on the sum of univariate Tobit models is introduced. A Monte Carlo simulation comparing the three estimators is performed on three different sample sizes. The QML estimators perform well in the presence of moderate sized error correlation coefficients often found in empirical studies. With absolute larger correlation coefficients, the SML estimator is found to be superior. The paper lends support to the general use of the QML estimators and points towards the use of simple etimators for more general censored systems of equations.
TL;DR: Morrison Paul et al. as mentioned in this paper used a nonparametric, output-based data envelopment analysis (DEA) to estimate the technical, allocative, and scale efficiencies of farm households in southern Minnesota using a time series, panel dataset from 1993-2005.
Abstract: Inter-firm differences in economic efficiency are major factors explaining differences in firm survival and growth and changes in industry structure. Thus, factors explaining differences in efficiency are of major interest to many involved in or affected by the industry. This study was undertaken to improve our understanding of the inter-farm differences in farm household efficiency in utilizing their land, labor, and capital resources to achieve household objectives. It estimated the technical, allocative, and scale efficiencies of farm households in southern Minnesota using a nonparametric, output-based data envelopment analysis (DEA) of a time series, panel dataset from 1993-2005. A bootstrap method was used to establish statistical properties of the technical efficiency estimates. Tobit analysis was conducted to identify significant factors explaining differences in efficiency scores. This study also expands the use of efficiency estimation and Tobit analysis of factors to identify educational needs for improving efficiencies. This study extends current research in several ways. It uses a true panel dataset versus the pseudo panel used by Morrison Paul et al (2004). To our knowledge, this study is the first estimating U.S. agricultural production efficiencies to use bootstrapping procedures to correct the bias generated by the deterministic DEA approach. It is the first to use a weighted Tobit procedure to correct for that same bias. The study is also the first to extend the results of estimating efficiencies and the Tobit identification of explanatory factors to identifying both potential policies and educational opportunities for improving efficiencies. While most previous studies did not consider nonfarm income and labor in their study, the fact that nonfarm activity now accounts for a large percentage of household income and resources means that they should be incorporated in the calculating of production frontier. As in Morrison Paul et al. (2004) and Chavas et al. (2005), this study incorporated nonfarm income as an output and nonfarm labor as an input in the production technology. Following Chavas et al. (2005), Morrison Paul et al. (2004), and others, we first used nonparametric (DEA) methods to estimate output-based technical, allocative, and scale efficiencies. Based on the smoothed bootstrap procedure for DEA estimators proposed by Simar and Wilson (1998, 2000), the study estimated the bias and the confidence interval of the DEA estimators for TE. Initial technical efficiency was estimated to be 0.90; scale efficiency to be 0.88, and allocative efficiency to be 0.77. These efficiencies improved over the period. Using the bootstrapping results, the average bias-corrected technical efficiency was 0.77 or 86% of the initial estimate. The bias-corrected technical efficiencies were similar for small farms and large farms even though the difference between the lower and upper bounds was larger for large farms. Tobit analysis showed that more specialized farms had higher levels of efficiency by all three measures. A higher proportion of rented land was associated with higher allocative and scale efficiency but lower technical efficiency. A lower debt-asset ratio and a higher percentage of current debt were associated with higher levels of all three measures of farm efficiency. A higher proportion of nonfarm household income and higher hired labor, capital-to-labor and land-to-labor ratios had positive effects on efficiency. Several conclusions and suggestions for improving farm efficiencies can be drawn from these results. First, since a lower debt-asset ratio and higher percentage of current debt were associated with higher efficiency levels, management skills that improve financial condition will likely improve efficiency. So improvement of management skills in general, through education of current and future farmers, appears to be needed (as we have always striven to do). Increasing the amount of rented land relative to owned land has a positive impact on allocative and scale efficiency so improved land markets and the ability to obtain and hold additional land is critical. Also, improvement in land market negotiation skills and intra-personal skills dealing with absentee landowners can lead to efficiency improvements. However, since a higher tenancy ratio was associated with lower technical efficiency, improvements in managing larger operations and rented properties appears to be needed. The positive impact of nonfarm income shows the need for farm households to take advantage of nonfarm opportunities as well as the need for rural communities to expand and develop those opportunities. Better access to both debt and nonfarm equity capital can improve efficiencies. This includes the identification and use of nonfarm capital (such as partnerships and investments by nonfarmers) and the identification and use of lower cost-debt capital for expansion and improvements as well as the increased management ability to manage higher debt loads. The positive impact of higher capital-to-labor and land-to-labor ratios indicates the need for more intensive use of available labor through increased mechanization and expansion of the land base. These steps can be seen as needing to accompany the ability to access more debt and equity capital. The positive hired labor ratio illustrates the impact of hiring labor and thus, presumably, freeing the farm household to spend more time on management following the highest and best use argument for the owner's time allocation. The need to increase the relative amount of hired labor points to the need to increase personnel management ability in farmers and thus personnel management educational opportunities for current and future farmers. The positive impact of the Herfindahl index shows the need to increase management skills, and risk management skills especially, to handle more specialized operations that will rely on off-farm tools for protection from uncertainty versus relying on on-farm diversification as a risk decreasing tool. For the AAEA meetings, the presentation will include an explanation of the estimation methods used including the smoothed bootstrapping methods (Simar and Wilson). Quantitative results will be presented including the initial TE estimates compared to the bias corrected TE estimates. The significant explanatory factors identified through in the Tobit analysis will be presented. We will expand our discussion of the educational needs identified for improving efficiencies. We anticipate discussion will occur on the use of the bootstrapping method and our identification of education needs for improving household efficiency. Chavas, J-P, Petrie, R., and Roth, M. (2005). i?½Farm Household Production Efficiency: Evidence from The Gambia.i?½ American Journal of Agricultural Economics 87 (1): 160-179. Morrison Paul, C., Nehring, R., Banker D. and Somwaru A. (2004). i?½Scale Economics and Efficiency in U.S. Agriculture: Are Traditional Farms History?i?½ Journal of Productivity Analysis, 22 (2004):185-205. Simar, L., and P., Wilson (1998), i?½Sensitivity Analysis of Efficiency Scores: How to Bootstrap in Nonparametric Frontier Models.i?½ Management Science 44(1): 49-61. Simar, L., and P., Wilson (2000), A general methodology for bootstrapping in nonparametric frontier models, Journal of Applied Statistics, 27(6):779-802.
TL;DR: In this article, a joint time-allocation model that explicitly considers a child in addition to parents was proposed, based on results of an activity-based time allocation model that analyzes activity duration as well as activity type.
Abstract: This paper describes research on intrahousehold interactions, based on results of an activity-based time allocation model. The paper describes a joint time-allocation model that explicitly considers a child in addition to parents. It analyzes activity duration as well as activity type. This model was formulated as a nonlinear Tobit model for econometric analysis. The model parameters were estimated with data from a household timeallocation survey completed during 2003 in Toyama, Japan. Two models were estimated with the sample data sets: one for weekdays and one for weekend days. The results indicate that social and cultural backgrounds relating to the husband's preference of child gender, household childcare role sharing between husband and wife, and the husband's immersion in work influence the household interactions significantly. The paper presents a summary of model results, conclusions, and recommendations for further research.
01 Jan 2007
TL;DR: In this article, a censored linear approximate Almost Ideal Demand System (LA-AIDS) model was employed in the study, and major economic parameters were estimated for different meat items, and a Tobit model for away-from-home consumption expenditure was estimated.
Abstract: This paper focuses on both at-home and away-from-home meat consumption in China, aiming to provide a deeper insight into the changing meat consumption patterns and their determinants in China. Data were collected in 2005 from two separate consumer surveys – one urban and one rural. A censored linear approximate Almost Ideal Demand System (LA–AIDS) model was employed in the study, and major economic parameters were estimated for different meat items. A Tobit model for away-from-home meat expenditure, and a system of multivariate Tobit equations were estimated. This enabled an examination of the links between at-home and away-from-home meat consumption. Moreover, it may be more important than it seems, and under-recording of away-from-home consumption could partly explain the widening gap in official statistics between China’s livestock production and meat consumption.