Employability and Wage Compensation in an Asian Economy: Evidence for Female College Graduates in Taiwan

Abstract

The purpose of this study is to analyze the influence of employability skills on wage compensation for female college graduates in Taiwan. We find that employability skills can explain some variation in wage compensation whether we include conventional human capital variables or not. For example, the career management skills category exhibits a consistent and significant influence on wage compensation and could raise the earnings level by 5–6 percent. In addition, employability skills have more diverse effects on wage compensation across various occupations, while conventional human capital variables are shown to have more consistent effects on wage compensation across occupations.

KEY WORDS:

• employability
• female college graduates
• wage compensation

Notes

1. The Big Five factors are openness, conscientiousness, extraversion, agreeableness, and neuroticism.

2. Cognitive abilities refer to arithmetic reasoning, word knowledge, paragraph comprehension, mathematical knowledge, and coding speed.

3. Communication skills, teamwork skills, problem-solving skills, initiative and enterprise skills, planning and organizing skills contribute to long-term and short-term strategic planning, self-management skills, learning skills, and technology skills.

4. The key questions regarding employability contain many variables that may be highly correlated. Thus, there may be a multicollinearity problem in the estimation. In the regression analysis, there are two commonly-used methods that are employed to detect the severity of multicollinearity. The first relates to the correlation coefficients and the other concerns the magnitude of the variance inflation factor (VIF). We have used these two approaches to examine the multicollinearity problem in this study. According to the results of the correlation coefficient matrix, the twelve key questions indeed exhibit some correlation among each other, with the correlation coefficient ranging between 0.2 and 0.5. The category with the highest correlation relates to the career management skills (CAREER), where the correlation coefficient reaches as high as 0.5. However, the value of the VIF in our study is less than 10. In general, if the value of the VIF is less than 10, the multicollinearity problem will be regarded as acceptable and the empirical results will therefore be considered reliable.

5. The currency unit in this study is the NT$. The exchange rate is about US$1 = NT$30.

6. There are five groups of occupations classified by the original survey. The first group, Salesmen, contains businessmen, chefs, police, salesmen, babysitters, and so on. The second group, Associated professionals, contains occupations such as legislators, government administrators, business executives, managers, lawyers, doctors, accountants, engineers, dispensers, reporters, nurses, musicians, directors, etc. The third group, Technicians and professional assistants, includes technicians, realtors, tax collectors and personnel working in social welfare institutions. The fourth group, Clerks and other staff, contains stenographers, typists, tellers and counter clerks, counterpersons, secretaries, and telephone switchboard operators. All other occupations are classified into the fifth group. There may be some reservation about the classification of occupations. However, since the original survey does not allow us to identify specific occupations for each respondent, we cannot exclude some strange occupations from a specific group, such as the policemen and babysitters in the group of Salesmen.

7. The control variables for family background contain the parents’ educational levels, their working sectors (private or public sector), and the father’s occupation. The college/university-related variables are college/university types, majors, grade point average, certifications of finance, language, computer, or a license issued by the government, and whether the college or university is located in a metropolitan area or not. The work-related variables include working locations, firm size, whether one is a manager, and industries.

8. A possible explanation of the independence between labor force participation and occupation choice is that the participation decision is normally made prior to the choice of occupation. Once a female college graduate decides to participate in the labor market, she will start searching for a job. She will face the choice of occupation when she receives job offers. As a result, the participation decision usually comes before the choice of an occupation. This may justify the independence assumption between these two selections. One may be curious about how the estimation results will differ if the independence assumption does not hold. We have conducted an experiment on this matter by estimating the wage equation with the selection of occupation only. We have found that the implications of the employability on the wage of each occupation are similar to our main findings from the estimation results having considered both the selection of occupation and labor force participation.

9. In Taiwan, there is usually a three-month probation period for new recruits. During this period, supervisors or employers closely evaluate the skills, abilities, attitudes, and other aspects of new employees. After the probation period, the employer offers a formal contract to the employee.