The purpose of this study was to identify the major components of affirmative action (AA) fairness and to develop a valid and reliable measuring instrument, which could be used to measure the perceptions of employees on AA fairness. A literature study and an empirical study were conducted. The literature study focused on fairness principles, outlined AA practices and identified work behaviours related to commitment. On completion of the literature study, a questionnaire was developed to collect information on respondents’ biographical details and their perceptions of the fairness of AA, the treatment of AA employees and employees’ commitment. Using SPSS, principal axis factor analysis was performed on the data, with a Varimax rotation, in order to uncover the different factors related to the three behavioural domains. Four factors for each of the behavioural domains with latent roots greater than unity (Kaiser’s criteria) were extracted from the factor matrix of each domain. The factors postulating the different behavioural domains are as follows: AA fairness: interactional justice, procedural justice (input), procedural justice (criteria) and distributive Justice Treatment of AA employees: task autonomy, respect, responsibility and realistic expectations on the part of supervisors Employees’ commitment: obedience, participation, satisfaction and loyalty The influence and effect of the biographical variables on fairness perceptions were determined by xiv comparing the responses of various employee subsets with one another by means of univariate and multivariate analysis of variance. The results of the t-tests revealed that staff category, marital status and ethnicity have a statistically significant effect on employees’ perceptions of the distributive justice of AA. Decisions such as granting AA employees token positions, paying unrealistically high salaries to AA managers and appointing less-qualified employees, play a key role when managers, married employees and whites form perceptions of the fairness of AA. Women consider procedural fairness, adhering to rules and regulations and loyalty to be of paramount importance. The MANOVAs and associated ANOVAs indicated that ethnicity, staff category and age must be considered for their significant effect on perceptions of distributive justice, autonomy, respect and loyalty. Blacks, contrary to whites, believe that AA employees are not treated with respect and are not accorded autonomy. Older employees and clerical staff tend to be more loyal than their counterparts. Multiple regression statistics were used to predict how the treatment of AA employees relates to perceptions of the fairness of AA. The results indicated that the more task autonomy and respect accorded to employees, the more likely they will perceive decisions about AA to be fair. This study represents a vital step towards a better understanding of the dimensionality of AA fairness and should ultimately contribute to more effective management of AA in the workplace.
Thesis (PhD (Human Resources Management))--University of Pretoria, 2006.