Wage policy patterns at the firm level and vulnerable employees

Segmentation in the labour market is a long known phenomenon and large and substantial evidence supports the hypothesis that employees who get locked into disadvantaged segments suffer significant penalties in employment opportunities, career development and earnings. Recent changes in the labour markets of industrialized countries have revived this phenomenon by creating new forms of segmentation. Skill biased technological change increases the gap between the highest segment and the lowest segment of the labour market and penalises those employees endowed with wrong or insufficient skills. The increasing speed of technological and organisational change marginalises workers who cannot keep pace with new skill requirements due to insufficient education or inadequate training opportunities. The diffusion of temporary contracts reduces the incentives to invest in firm-specific or occupation-specific professional skills and jeopardises the development of professional paths for affected employees. The literature suggests that the risk of getting caught into an unfavourable segment of the labour market is larger for more vulnerable categories of employees, e.g. younger workers, women, or older workers. However, despite consistent and growing evidence on the negative labour market outcomes for individuals restrained to lower labour market segments, there is less evidence about the impact of firm characteristics on segregation into less favourable segments. This paper argues that the wage policy at the firm level is a significant driver of segregation into disadvantaged segments of the labour market. Recruitment patterns, retention patterns and further career steps both in the internal and in the external labour market significantly depend on the wage policy pursued by firms. In support of the proposed thesis, the paper develops an empirical analysis based on WHIP, a longitudinal dataset including a representative sample of employment relationships in Italy. WHIP provides a dynamic panel including about 862,000 employment relationships held by about 350,000 individuals between 1985 and 2004. For each employment relationship, identified by a unique code, WHIP provides information about start date, end date, gross reward per year, equivalent worked days per year, occupation, collective labour agreement in force, job level and administrative events that each year may affect an employment relationship, such as maternity leave or illness leave. The database provides also additional information on employers and employees. Wage policy patterns are identified by means of a two-step clustering algorithm based on variables that capture wage levels, wage structure and wage dynamics of a sample of medium and large Italian firms. The comparison of the career paths developed by workers initially employed at firms adopting different models of wage policy allows identifying the impact of wage policy on the segregation into unfavourable segments of the labour market for different categories of vulnerable workers, including younger workers, older workers and women. The preliminary results suggest that the sampled employers actually adopt diversified wage policy patterns. In addition, the identified wage policy patterns are significantly associated with the earnings dynamics and the inward and outward labour flows of vulnerable categories of employees.