Panel SVAR model of women's employment, fertility, and economic growth: A comparative study of East Asian and EU countries

Abstarct

This paper examines the relationship among women's employment, fertility, and economic growth using two regional panel data, those of 8 East Asian and 15 EU countries over the period 1980–2008. We use a panel structural vector autoregression model and estimate the variance decomposition. The endogeneity among women's employment, fertility, and growth is considerably reduced when we use the samples of East Asian and the EU countries rather than a specific country. In addition, we find that changes in women's employment and fertility rates affect the determination of growth rates in EU countries to the extent of about 15%, which is higher than the impact in East Asian countries—about 10%.

Acknowledgments

We have benefited from comments and suggestions on the editor of this journal and anonymous referees. Lee was supported by the Brain Korea 21 Project. Any remaining errors or ambiguities, of course, are ours.

Notes

1 We exclude some East Asian countries, such as Brunei, Cambodia, Laos, and Vietnam, due to the data availability. In addition, the excluded East Asian countries are less developed than the 8 East Asian countries considered in our sample.

2 Data from the World Bank, especially World Development Indicators is widely used in the existing literature, and is one of the most reliable data sources for cross-country comparisons.

3 Enders (2003) explains that a vector autoregression (VAR) model is helpful in examining the relationship among a set of economic variables. All variables in a VAR model are treated symmetrically: each variable has an equation explaining its evolution based on its own lags and the lags of all the other variables in the model. Moreover, the resulting estimates can be used for forecasting purposes. However, since the VAR approach has been criticized as being devoid of any economic content, we use the structural VAR (SVAR) model, which is theoretically identified from the reduced-form of VAR model.

4 Hondroyiannis and Papapetrou (2001), Maksymenko (2009), and Yip and Zhang (1997) analyze endogenous relationships among women's employment, fertility and growth based on similar comparative static analyses of Wang et al.’s (1994) argument.

5 The empirical results of panel cointegration tests are provided in Appendix A.

6 We try to estimate the structural variance decomposition of a few East Asian counties separately. The results are similar to those in Wang et al. (1994), namely the change in a variable that affects the variable itself is around 60–80%, although we used a smaller sample and a different measure of women's employment.

7 See Takayama and Werding (2011) for a detailed discussion of the issues.