GENDER DIFFERENCES IN VOCATIONAL SCHOOL TRAINING AND EARNINGS PREMIUMS IN TAIWAN

Abstract

International capital mobility and economic restructuring have brought training and skills acquisition to the forefront of policy dialogues. Taiwan has gone beyond most countries in promoting vocational education and setting strict quotas for schooling. Although the education plans do not have separate targets for men and women, they have gendered outcomes. Estimates of earnings premiums using ordinary least squares and quantile regression techniques indicate that only men have gained consistently higher premiums from vocational school compared to general schooling. Women who were denied access to the university system have forgone college premiums that exceed those of men. Also, the commerce track, in which women cluster, yields an earnings penalty compared to general schooling, while the technical track, in which men predominate, yields an earnings premium. Policy reforms based on relaxing education quotas and enforcing equal opportunity legislation would provide women with more rewarding education and career options.

Keywords:

• Education
• skills
• segregation
• wage gap
• Taiwan
• quantile regression
• JEL Codes: J24, O2, J31

Acknowledgments

The authors wish to thank Günseli Berik, Terri Boyer, Julie Hotchkiss, Bill Rodgers, and Patricia Roos for their useful comments. We also thank members of the 2004 – 5 seminar of the Rutgers Institute for Research on Women and panel members of the 2004 International Association for Feminist Economics conference and the 2004 Southern Economic Association conference. This research is supported by a Summer Research Grant from the College of William and Mary. This paper does not necessarily reflect the views of the Asian Development Bank or of the Urban Institute.

Notes

¹ Advocates and skeptics also dispute the social rates of return to vocational schooling in cost-benefit analyses. The conventional wisdom holds that social rates of return to vocational schooling are lower than returns to general schooling. For example, George Psacharopoulos (1994) finds that for a sample of twenty-five countries, the average rate of return to government investments in general secondary education (15.5 percent) is considerably higher than the average for vocational education (10.9 percent). Hence the labor market benefits of vocational-trained workers may not be large enough to justify the high costs to governments of delivering vocational schooling. In contrast, Paul Bennell (1996) concludes that the large majority of country studies find social rates of return for general secondary education that are not significantly higher than those for vocational secondary schooling.

² This section on Taiwan's education system, policies, and trends draws on Jennie Hay Woo (1991), Flora Tien (1996), Robert Wu (2000), and Ministry of Education, Republic of China (2002).

³ The data are obtained from the Executive Yuan's Directorate-General of Budget, Accounting, and Statistics (DGBAS Various years).

⁴ Individuals are asked to report weekly working hours for the previous week, which in most cases was the third week of May. Respondents whose incomes are relatively stable are asked to report their earnings for the month of April, while those with unstable or seasonally fluctuating incomes are asked to report their average monthly work income over a year's time. Non-monetary compensation is excluded. We impute monthly earnings for individuals with top-coded earnings as 1.2 times the top-code. This imputation affects only 4 observations per year on average.

⁵ See, for example, Lucie Cheng and Ping-Chun Hsiung (1994).

⁶ Yet further work in Linda Loury (1997) shows that increases in the market value of women's skills relative to those of men for a given major accounts for much of the closing of the gender gap. For other US evidence on college major and wage outcomes see Charles Brown and Mary Corcoran (1997) and Gill and Leigh (2000).

⁷ Closely related, critics also argue that the content of vocational school should be generalized, with an expansion of post-graduation academic options available to students in the vocational track. See, for example, Indermit Gill, Amit Dar, and Fred Fluitman (1999).

⁸ Within East Asia, relatively high premiums for vocational education compared to general education are found in Thailand (Thammarak Moenjak and Christopher Worswick 2003) and Singapore (Chris Sakellariou 2003). T. H. Gindling, Marsha Goldfarb, and Chun-Chig Chang (1995) report almost identical premiums for vocational school and high school in Taiwan for the 1978 – 91 period. This study does not further disaggregate vocational school by major areas of study.

⁹ Potential problems with selectivity bias in earnings premium estimates have received much attention in the literature (for a recent discussion, see David Card 2001). In particular, schooling is not exogenous and unobserved ability affects both wages and school attainment. The most common approach is to control for selection into school levels by using family background characteristics. Unfortunately, the data for Taiwan do not offer options for good instruments. We conducted some sensitivity tests using business owner within the household and other household income as instruments in a two-step selection correction procedure and results did not change in any substantial way.

¹⁰ Note that the coefficients for vocational school, junior college, and college are similar in magnitude, sign, and precision if we drop workers with only primary and junior high school educations completely out of the sample.

¹¹ For example, see Psacharopolous (1994), Philip Trostel, Ian Walker, and Paul Woolley (2002), Christopher Dougherty (2003), and George Psacharopolous and Harry Patrinos (2004).

¹² We define “recent” graduate in the workforce as any full-time employee in the worker sample who has up to six years of potential experience and up to six years of tenure with the same employer. Choosing a smaller experience category led to some education-sex cells with zero observations in some years.

¹³ We also compared starting salaries of recent graduates of academic high school with those of recent graduates of vocational school (with all majors grouped together). However, for both men and women, the differences across these two school levels were extremely small. Although vocational school students in Taiwan may be stigmatized, they are doing about the same in terms of starting salaries as academic high school students who fail to go on to college.