The Measurement of Sex and Marital Status Discrimination at the Workplace

Abstract

This paper attempts to measure market discrimination against women, defined as receipt of unequal pay for equal productivity, at the level of the establishment.1 Such measurements are of relevance to current policy since, for example, the Sex Discrimination Act (1975) empowers the Equal Opportunities Commission (EOC) to undertake investigations of establishments where there is reason to believe discriminatory practices are taking place. The first such investigation, at the Luton site of Electrolux Limited, began in February 1977. The EOC has also recommended that all managements undertake regular audits of the relative position of men and women in their organizations (see EOC, 1978a, b). In measuring discrimination at establishment level we follow other studies (see for instance Malkiel and Malkiel, 1973; and Chiplin and Sloane, 1976b), but we differ in that separate regressions are run not only according to sex but also according to marital status. Five cases as opposed to a single establishment are considered in order to cover a variety of employment situations. Data on individual employees obtained from personnel records are used to estimate earnings functions relating weekly or annual earnings to variables reflecting the education and job experience of the worker. The estimated equations are subsequently used to determine how much of the differences in mean earnings by sex and marital status can be determined in terms of differing characteristics. The question of whether factors such as education and service with the firm should influence pay is not taken up here. Our choice of standardizing variables is essentially dictated by availability of personnel record data. The legislation however has a wider conception of discrimination. For example, the Sex Discrimination Act defines indirect discrimination as a requirement or condition with which a considerably smaller proportion of one sex than the other (or of married as opposed to unmarried persons) can comply, which is to the detriment of that group, and which cannot be justified by the requirements of the job (our emphasis). This is a different problem to that taken up in this paper. Nevertheless, the decisions of the industrial tribunals and the Employment Appeals Tribunal so far hold that differences in factors such as experience, age, length of service and qualifications can be justifiable requirements for enhanced payment. There are advantages and disadvantages in attempting to measure discrimination at the establishment rather than the national level. The advantage is that managers themselves have to combat discrimination at the establishment level. The sort of tests we employ below could be readily implemented on a wide basis using information available to management. Indeed, the Equal Pay and Sex Discrimination Acts are oriented towards the establishment: equality may be claimed only with comparable workers in the same firm or establishment, so that group differences in wage rates for similar occupations or in occupational distributions within an industry or over the economy as a whole do not constitute