Discrete Endogenous Variables Research Articles

Discrete endogenous variables in weakly separable models

Summary This paper contains an extension of the identification method proposed in Jun et al. (2011), hereafter JPX, which is based on a generated collection of sets, that is a ‘Dynkin system’. We demonstrate the usefulness of this extension in the context of the model proposed by Vytlacil and Yildiz (2007), hereafter VY. VY formulate a fully non-parametric model featuring a nested weakly separable structure in which an endogenous regressor is binary-valued. The extension of the JPX approach considered here allows for non-binary-valued discrete endogenous regressors and requires weaker support conditions than VY in the binary case, which substantially broadens the range of potential applications of the VY model. In this paper we focus on the binary case for which we provide several alternative simpler sufficient conditions and outline an estimation strategy.

Disability and Labour Supply during Economic Transition: Evidence from Bulgaria

. This paper focuses on the interrelationship between disability and labour force participation for the adult population in Bulgaria. We estimate simultaneous equation models with discrete endogenous variables to account for the different forms of interdependence between health and labour supply. Using the Bulgarian Integrated Household Surveys conducted in 1995, 1997 and 2001, our econometric results indicate that disability exerts a negative effect on labour participation, but labour supply has little effect on disability. Evidence from panel data also suggests that disability reduces labour force participation.

Tenure Security, Investment and Productivity in Gambian Agriculture: A Generalized Probit Analysis

AbstractThe determinants of investment, input use, and productivity are investigated under customary tenure in peri‐urban areas of the Gambia. A structural model is specified to investigate the role of tenure security on farm investments and input use and thereby on yield. Testing of the structural form hypotheses requires simultaneous equation estimation. Containing both continuous and discrete endogenous variables, the model is estimated as a feasible generalized least squares Amemiya's generalized probit. Some of the positive relationships hypothesized between tenure security, investment, and yields are corroborated. In particular, tenure security is found to enhance long‐term investments, which in turn enhance yields.

Planting strategies of maize farmers in Kenya: a simultaneous equations analysis in the presence of discrete dependent variables

A fairly comprehensive range of planting choices made by maize farmers in Kenya (including discrete endogenous variables creating self-selectivity) is modelled and estimated as one system of interrelated decisions. Two-stage and three-stage probit procedures are used to handle the simultaneity and self-selectivity problems. Results showed that population pressure and agroclimatic diversity are important determinants of crop intensification and planting regimes among maize farmers and further supported the importance of focusing maize research in terms of agroclimate and socio-economic domains. Shorter maturity and efficient double and multiple cropping methods are needed to increase land productivity and intensity of labour use in areas of high population pressure and bimodal rainfall, i.e. mid-altitude zones. On the other hand, technologies that would lead to increased productivity of capital and higher response to external inputs are desired for the highlands of Kenya. Access to extension and machine services, distance to the maize plot, and time of onset of the rains were also found to significantly influence the planting strategies of maize farmers.

Planting strategies of maize farmers in Kenya: a simultaneous equations analysis in the presence of discrete dependent variables

AbstractA fairly comprehensive range of planting choices made by maize farmers in Kenya (including discrete endogenous variables creating self‐selectivity) is modelled and estimated as one system of interrelated decisions. Two‐stage and three‐stage probit procedures are used to handle the simultaneity and self‐selectivity problems. Results showed that population pressure and agroclimatic diversity are important determinants of crop intensification and planting regimes among maize farmers and further supported the importance of focusing maize research in terms of agroclimate and socio‐economic domains. Shorter maturity and efficient double and multiple cropping methods are needed to increase land productivity and intensity of labour use in areas of high population pressure and bimodal rainfall, i.e. mid‐altitude zones. On the other hand, technologies that would lead to increased productivity of capital and higher response to external inputs are desired for the highlands of Kenya. Access to extension and machine services, distance to the maize plot, and time of onset of the rains were also found to significantly influence the planting strategies of maize farmers.

Aid to families with dependent children (AFDC) and non-marital births in the USA: an examination of causality

Utilizing a sample of 2964 unmarried women over the period 1979-88 from the US National Longitudinal Survey of Youth, a test was performed to determine the causal direction of relationship between receipt of aid to families with dependent children (AFDC) and the decision to have a non-marital birth. The existence of ‘causality’ is defined as the lack of a simultaneous relationship or joint dependency between these variables. One of the most interesting findings was a failure to reject the hypothesis that these choices are jointly determined. Both were found to depend upon variables that are demographic, economic, personal, and family-related. The profiles derived from the specification and estimation of a simultaneous equation system with discrete endogenous variables depict a woman whose fertility and economic decisions were predominantly based on economic deprivation, demographic situation and family instability.

Estimation of econometric models of some discrete games

AbstractThis paper discusses, estimates and compares some microeconometric models for simultaneous discrete endogenous variables. The models are based on the assumption that observed endogenous variables represent the outcome of a static discrete game. I discuss models based on non‐cooperative equilibrium concepts (Nash, Stackelberg), as well as models which presume Pareto optimality of observed outcomes. The models are estimated using data on the joint labor force participation decisions of husbands and wives in a sample of Dutch households.

Simultaneous‐Equation Systems Involving Binary Choice Variables

Simultaneous‐equation systems of continuous variables are well known in the social sciences. Unfortunately, these methods cannot be used when dealing with discrete endogenous variables. Although for discrete types of data a broad range of methods and techniques have been developed, their applicability is restricted to single‐equation systems (for example, logit or probit formulations) or to association‐type models (log‐linear models), mainly because of the lack of suitable estimation methods and computer programs. The basic problem is the multivariate error structure of the endogenous variables. This paper presents a simultaneous‐equation system of binary endogenous variables. The model is applied empirically to longitudinal data on mode choice and is compared with the well‐known log‐linear model.

Econometric Evaluation of New Technology with an Application to Integrated Pest Management

In this paper we provide a method for the microeconomic evaluation of a new technology using a panel data approach. With panel data, the choice of whether to adopt is endogenous, and firms (observations) for which it is more profitable to adopt may have self-selected the new technology as opposed to the case of experimental design data where the choice of technology is randomly assigned by the experimenter. Heckman showed that in the presence of self-selection common error structures lead to self-selection bias of ordinary least squares estimators. Duncan, however, noted some economic conditions, summarized by his Quantal Response Regression (QREG) condition, under which self-selection bias would not occur for ordinary least squares estimators. Both models are of the mixed continuous/discrete type, a system of equations with continuous and discrete endogenous variables. In addition to self-selection bias, a second type of bias may result from the use of panel data. Because a random sample could easily not contain enough observations in both groups of adopters and nonadopters to enable the researcher to draw statistical inference of the differences between the two groups, one might use stratified random sampling (Cochran); but stratification upon an endogenous choice-called choice-based sampling-can result in choice-based sampling bias (Cosslett, Hausman and Wise, Manski and Lerman, and Manski and McFadden). Our data are subject to both types of problems, so our econometric evaluation of a new technology using panel data requires a model which can account simultaneously for the possibility of both selfselection and choice-based sampling. We apply our model to the evaluation of integrated pest management (IPM) which is of interest to the general public, farm workers, growers, and policy makers. In Silent Spring, Rachael Carson popularized the growing literature on risks to the environment and human health from pesticides. From penguins in the Antarctic to mothers' milk, pesticides are ubiquitous. Residues of persistent pesticides on food, in water, and in the air resulting in continual long-term exposure at low doses pose to the general public largely unknown chronic disease risks, such as carcinogenicity, mutagenicity, and teratogenicity (Ames 1979). In response to these risks, entomologists embarked upon research to develop programs which integrate biological controls, such as predators and parasites, with pesticides (Stern et al.). IPM programs are not, however, always designed to reduce risks to human health and the environment, nor do IPM programs necessarily achieve that objective even when designed to do so. For example, Ames (1983) notes that crops which are naturally resistant to pests have natural pesticides which may be just as hazardous or even more so than artificially applied pesticides, and crop resisThe authors are, respectively, an assistant professor, University of California, Riverside, and an associate professor, Washington State University. Giannini Foundation Paper No. 714. The authors gratefully acknowledge financial support from the University of California Agricultural Experiment Station and the Consortium for Integrated Pest Management, funded by the U.S. Environmental Protection Agency and the U.S. Department of Agriculture through Texas A&M University.

Dummy Endogenous Variables in a Simultaneous Equation System

This paper considers the formulation and estimation of simultaneous equation models with both discrete and continuous endogenous variables. The statistical model proposed here is sufficiently rich to encompass the classical simultaneous equation model for continuous endogenous variables and more recent models for purely discrete endogenous variables as special cases of a more general model.