英语论文网

emales in lower-paying occupations, but a significant fraction remains unexplained, i.e. discriminatory fraction. It was estimated that from about one-quarter to one-half of the gender wage gap takes the form of wage differences between men and women with narrowly-defined occupation within establishments in the US (Bayard et al., 1999). However, analysing earnings by sex within narrowly defined occupations also has its limitations. Unequal pay for identical work only partially explains the pay disparity; the causes of differences in occupational employment patterns of men and women require independent explanation.

Blackaby, Booth, and Frank (2005) attempt to analyse the pay gap in the academic labour market in the UK. By employing the explanatory variables for productivity (such as teaching performance, publications, publications, experience, and career interruptions) they compare salaries of male and female economists. In contrast to the existing research findings, they reveal both a gender promotions gap and a within-rank gender pay gap. It is believed that the driving factor might be the role of outside offers as men receive a greater number of those than females of the same comparable characteristics. It, therefore, leads to a conclusion of discrimination in the UK academic labour market. However, the great limitation of direct productivity estimates is that datasets with direct measures of individual productivity are rarely available, limited in size, and not representative of the economy as a whole.

All the empirical procedures described above have their pros and cons. The method of narrowly defined occupations solves the issue of controlling for a wide range of pay determining factors, however, it ignores the effect of segregation across occupations. The audit studies, on the other hand, helps to detect discriminatory behaviour as the employer responses are measured given the same candidate characteristics but depending on his/her gender. Certain limitations yet come into the picture. The pay decisions as well as the hiring decisions are made by the marginal employer, which makes it challenging to draw any macroeconomic conclusions. Also, there is no information that we could obtain on the factors affecting the pay gap. Oaxaca-Blinder decomposition (or other regression decompositions) does, in fact, produce such information, but at the same time attaches the residual which is not explained by productivity-related characteristics to discrimination. A wide range of significant production factors gives a more solid ground to explore discrimination. However, some of those factors can occasionally reflect discrimination in the labour market themselves and cause ‘over-control’ in the regression. It is thus important to consider those elements with some moderation before including them in the model. What is most satisfactory, however, is the utilisation of multiple lines of evidence from the diverse empirical procedures. If a coincident pattern materialises and they all have a tendency to detect discrimination, then the results are more reliable.

In this section reviewed is the evidence ? from the empirical studies and official statistics on gender wage gap in the UK. As well, the following determinants of pay differentials are identified and analysed to create a clear diagram of causes in gender income inequality. Specific works are referred to in order to demonstrate specific causes.