Academic journal article The American Journal of Economics and Sociology

Stereotypes, Asian Americans, and Wages: An Empirical Strategy Applied to Computer Use at Work

Academic journal article The American Journal of Economics and Sociology

Stereotypes, Asian Americans, and Wages: An Empirical Strategy Applied to Computer Use at Work

Article excerpt

I

Introduction

ASIAN AMERICANS have long been labeled as a "model minority" who use high investment in education (Xie and Goyette 2003; Kao 1995; Barringer et al. 1990; Chiswick 1988), family solidarity (Kao 1995), and group connectedness (Oyserman and Sakamoto 1997) to succeed. Asian Americans have also been recognized for high motivation (Mar 2000), stereotyped as having superior quantitative and technical skills compared to other ethnic groups (Xie and Goyette 2003; Shih et al. 1999; Steele 1997), and considered more likely to succeed as engineers, computer scientists, or mathematicians (Tan 1994; Leong and Hayes 1990).

In this article, we study the impact of positive stereotypes of Asian Americans' mathematical and technical ability on wages and the role this stereotype may play in explaining racial wage differences for both men and women. We do so by examining racial and gender differences in the wage premium associated with computer use. (1)

Stereotypes are likely to influence wages through two mechanisms. The first mechanism is statistical discrimination. An employer will be more willing to hire a member of a positive stereotype group and will pay that person a higher wage than someone else with the same observed characteristics. The second is the effect of stereotypes on the behavior and performance of individuals both in schooling and at work. A positive stereotype improves academic performance and affects one's field of study (Steele 1997; Steele and Aronson 1995), and possibly also one's work performance and earnings. If there is a positive stereotype of the technical and computer skills of Asian Americans, we'd expect that Asian Americans would have a higher wage premium for computer use than whites.

Stereotypes may themselves reflect aspects of racial or ethnic identity. For example, some work activities or tasks may be associated with "being Asian" while other tasks are not. Performing a work task that is associated with "being Asian" yields benefits beyond the earnings associated with the job if it contributes positively to group identity. If computer use on the job contributes positively to the Asian social identity but not to white identity, we'd expect a larger computer wage premium for Asians than for whites.

Because stereotypes are likely to be tied to race and gender: (1) positive stereotypes of Asians regarding quantitative skills could increase work performance and, in turn, lead to a higher wage premium for computer use for Asian men and women alike; (2) negative stereotypes of women regarding quantitative skills (Quinn and Spencer 2001; Spencer et al. 1999; Hedges and Nowell 1995) could reduce their work performance and, in turn, lower their computer use wage premium. The wage effect of stereotypes for Asian-American women thus may depend on which of these two is dominant.

Studies of wage differences between Asian Americans and whites have yielded mixed results. Some studies show evidence of discrimination (Sharpe and Abdel-Ghany 2006; Waters and Eschback 1995; Duleep and Sanders 1992) with Asians earning less than whites with similar education and experience, while others suggest that there are wage differences unexplained by traditional human capital theory (Sharpe and Abdel-Ghany 2006; Chiswick 1988). Others find no earning differences among those (whether white, Asian American, or Asian) who completed their education in the United States (Zeng and Xie 2004).

In general, wage regressions have difficulty distinguishing discrimination from other factors that might contribute to "unexplained" wage differences by race and gender. There is the possibility of both Type I and Type II errors. A hypothesis of discrimination may be accepted as true when it actually is false (a Type I error) if the wage regression did not include a measure of school quality and the average quality of schools attended varies by race. A hypothesis of discrimination may be rejected as false when it actually is true (a Type II error) if the wage regression did not include a measure of technical or mathematical skills and those are higher on average for one racial group than another. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.