Compensation policies and varieties of capitalism: The role of firms to generate inequality

Fátima Suleman, University Institute of Lisbon, DINÂMIA‘CET-IUL
Henrique Duarte, University Institute of Lisbon
Abdul K. Suleman, University Institute of Lisbon, DINÂMIA‘CET-IUL

It is well-known that firms respond differently to labour market regulations and develop an employment relationship accordingly. This study examines compensation policies of firms to illustrate how employers’ decisions regarding pay system generate inequality within and between firms. It compares firms within different models of capitalism and examines which compensation policies prevail in certain sectors, how they are associated with the use of flexible contracts, and workers characteristics.

The research is placed within the comparative institutional human resource management literature and attempts to give a picture of the segmentation of firms according to the pattern of their employment relationship. We therefore offer evidence on how these options affect inequality among workers between firms. More specifically, it presents a classification of firms according to their compensation policies and shows how those policies vary across varieties of capitalism (VoC) and sectors of activity, and interact with workers characteristics and contractual arrangements. The findings will help understanding whether firms are closer to a market-based model or an equity-focused system. This option will certainly affect workers’ income differently.

We draw on Structure of Earnings Survey (SES) data for 2014, which provide information on pay variables on eighteen European countries. Data are provided for Liberal, Continental European, Eastern Transitory Economies, Southern European Economies and Socio-Democratic countries.

The empirical strategy comprises two steps. The analysis starts with a fuzzy clustering to identify typical compensation policies of the sampled countries. The goal is to identify patterns of compensation policies according to four dimensions, notably pay structure, pay level, pay definition, and pay flexibility. These variables suit the theoretical models of compensation policies, namely the association between pay and job hierarchy suits the internal labour market model; and the ability to attract high-skilled candidates through pay has been highlighted by human capital and efficiency models.

In the next step, we estimate Tobit regression models to examine how the compensation policy interacts with the model of VoC, the sectors of activity and their possible relationships. The regression model includes also variables associated with the use of flexible contracts and demographic characteristics of workers.

The multivariate data analysis pointed to four fuzzy clusters based on the level of internal dispersion of earnings and relation to the market. Four fuzzy clusters of compensation policies can be ranged from low to high dispersion, while the other represents a policy oriented to the market. Internal Labour Market Undifferentiated (ILMU) policy show a negligible dispersion of earnings either within or between job levels; firms pay wages below the market rate and have lower returns to education and tenure. In the Internal Labour Market Qualified (ILMQ) policy, the key features are the highest return to education and tenure. The cluster labelled as Competitive evidences moderate dispersion and a wage rate higher than the general and industry market rates. Finally in the Incentive policy all the dispersion variables show the highest values.

These findings illustrate the variety of pay practices that seem to target a single outcome, the individualisation of earnings. While, some firms use within and between job dispersion to reward high-performers, others provide high return to education and qualifications. Some sampled firms prefer to be attractive in the market and pay higher wages, and offer at same time merit pay systems.

The Tobit regression model estimates show that highly dispersed policies prevail in Liberal labour market, while low dispersion policies are associated with Coordinated Markets Economy. The findings help dividing the Continental versus Southern and Transition European countries. In the former prevails Competitive and Incentive policies, while in the latter we found mixed patterns of ILMU and ILMQ. These results unveil the influence of labour market institutions, legal settings and probably cultural environments implicit in the capitalism model.

The results show that compensation policies also vary across sectors of activity, but the explanation power is weak. For example, pay structures in the public administration are less dispersed than in financial sector irrespective of the VoC model. The financial sector is indeed an example of the highest dispersion, showing that the banks and other financial institutions use pay system to be attractive and therefore differentiate workers. They consequently contribute to generate within and between firms inequality.

It has to be stressed that the models of VoC, which underlies the variation of institutions and economic characteristics of countries, are powerful predictors of patterns of compensation policies of firms. These policies are also associated with other elements of employment relationship, notably the types of contracts used by firms. Our findings show that firms that use Competitive model are less likely to use flexible contracts. In other words, they pay higher wages and gives job stability; their employment relationship model attempts to attract and retain workers. However, the firms in this cluster tend to differentiate more women and men. In contrast, firms implementing Incentive model are highly dispersed and make great use of temporary contracts; they associate pay flexibility and job flexibility. This model prevails, as reported, in Liberal market. Finally, flexible contracts prevail in Southern European and Transition countries, especially in firms having ILMU compensation policy. This is to say, low wages and low dispersion are associated with high-level of flexible contracts in those countries. Nevertheless, the gender wage gap is smaller herein. In sum, there is evidence of labour market segmentation; but firms combine different elements of pay and job flexibility.