Identifier

etd-06152010-201208

Degree

Master of Science (MS)

Department

Agricultural Economics

Document Type

Thesis

Abstract

This thesis studies the off-farm labor supply decision of farm operators and their spouses in the United States. The data used in this study is from the Agricultural Resource Management survey, 2006. The objective of this study is twofold. First, to identify those factors that affect off-farm labor supply of farm operators and their spouses. In particular, this study investigates the impact of human capital of farm operators and spouses, personal, family, farm and location characteristics on labor allocation for on- and off-farm work. Empirical results indicate that farm operators’ and their spouses’ human capital are positively correlated with off-farm labor supply. In addition, the number of children in a household is inversely related to a spouse’s off-farm employment. Similarly, a household’s net worth and farm size have a negative impact on off-farm labor allocation decisions by both farm operators and their spouses. Payments from government programs have a negative effect on labor allocation for non-farm work. The availability of health insurance to farm operators and their spouses from off-farm employment has a positive effect on labor supply for off-farm work. The second objective of this study is to compare results obtained from a parametric probit model and a semiparametric additive probit model of off-farm labor supply by farm operators and spouses. One of the most important aspects of semiparametric analysis is to identify smoothing or nonparametric variables in a regression model. The Blundell and Duncan (1998) approach shows that farm size is such a smoothing variable in the off-farm labor supply model. A semiparametric additive regression model identifies a few significant covariates as compared to a parametric probit model; however, the Hong and White (1995) specification test and likelihood ratio test favor a semiparametric model in this study. In particular, the graphical plots of fitted values from parametric and semiparametric models also show that a semiparametric model is preferred. The semiparametric model helps to formulate appropriate functional form of off-farm labor supply in the United States, which might be the subject of further study of this research.

Date

2010

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Mishra, Ashok K.

DOI

10.31390/gradschool_theses.1902

Download

Share

COinS