| Summary: | 博士 === 國立臺灣大學 === 農業經濟學研究所 === 104 === This dissertation includes two essays on computer use and gender differences.
Essay 1 (Chapter 2 in this dissertation):
Exploring Gender Digital Divide in Taiwan from the Perspectives of Internet Adoption, Use and Usage Patterns
This paper explores gender digital divide in Taiwan from the perspectives of internet adoption, use and usage patterns. Using 2012 Survey of Digital Divide in Taiwan, it is found that gender is not a significant factor determining internet adoption. However, after correcting for Heckman’s sample selection bias, the result reveals that females in Taiwan spend less time on internet than do males. The Blinder-Oaxaca decomposition results further suggest that the gender gap in usage time is the mixed product of both socio-economic difference, accounting for 34-47% of the gap, and gender specific factors, accounting for 48-66% of the gap. Finally, all else equal, the simulation shows that females are inclined to use internet for health information and e-government as the gender digital divide in internet use is eliminated, suggesting a boost to the elimination of the gender gap in internet use is important for Taiwanese government.
Essay 2 (Chapter 3 in this dissertation):
Computer Use and Gender Wage Differential: Evidence from Taiwan
Although there is a considerable body of literature examining the impact of on-the-job computer use on personal wage, little of them are known about the association between computer use and gender wage differential. In an effort to fill this knowledge void, the present study investigates the differential impacts of computer use on gender wage structures. Using the data from Taiwan, the findings first suggest that working experience, education, firm size and occupation are important determinants of on-the-job computer use. In addition, after correcting for the endogeneity of computer use, it is found that, on average, the rate of returns to computer use is 23% for males while it is not evident for females. This finding suggests the OLS estimates of the returns to computer use for females are biased upward as a result of positive self-selection. Third, by applying unconditional quantile regression, it is found that, no matter for males or females, the returns to computer use are more pronounced for employees at the lower quantile of wage distribution, suggesting the impact of computer use are heterogeneous across different quantiles of wage distribution. Moreover, the result suggests that the females at the lower quantile of wage distribution have higher returns to computer use than their male counterparts. Finally, the result of Blinder-Oaxaca decomposition reveals that most of the gender wage differential can be explained by the discrimination, particularly for those females at the lower quantile of wage distribution. However, the contributions of the coefficient and endowment effects of computer use to the gender wage gap are quite small.
|
|---|
