Testing dependence using copulas: the case of dual employment

Abstract

Copulas are functions that parameterize the dependence between univariate marginal distribution functions to form a joint distribution function. Copulas provide a consistent procedure for testing dependence and guiding the choice of empirical model. In this study we use copulas to measure the dependence in labour supply by married farm couples. Using individual data our research provides an easy method that researchers can use to address the issues of dependence. The method outlined here reduces computational time and provides a more efficient method of testing dependence early in the investigative process. Unlike previous studies findings from this study show with robustness that labour supply decision of married farm couples is best modelled jointly.

Keywords:

• copula
• dependence
• dual employment
• married farm couples
• off-farm labour supply

JEL Classification:

• J12
• J22
• J38
• Q12

Notes

¹ For example, although Lass et al. (1989), Lass and Gempesaw (1992), Mishra and Goodwin (1997), El-Osta et al. (2004) and El-Osta et al. (2008) use a theoretical model that jointly determines labour allocation decision, none of the empirical testing and evaluation of the models showed an evidence of joint labour decision-making in US farm households.

² Goodwin et al. (2003) pointed out that the jackknife procedure may suffer from some limitations and they proposed bootstrapping procedure as an alternative.