What and How

The project is built around two different “dependent variables” (the “explanandum”) and three sets of independent variables (the “explanans”). 

The explanandum 

The dependent variables are at the level of each policy field (e.g. pensions, etc.) (Mares, I., 2003. The politics of social risk: Business and welfare state development): 

1. employers’ preferences for social policy. “No social policy” (employers are not interested in any form of social coverage in respect to a given need); “private social policy” (employers prefer “occupational welfare” as the answer to given social needs, which means leaving to companies and, eventually, social partners, the choice to discretionally develop social programmes by virtue of the employment status); “contributory social policy” (employers accept or are willing to develop social policies financed by compulsory social contributions); “universalistic social policy” (employers accept or are willing to develop social policies financed by general taxation); the four types of preferences have clearly different costs for companies but also different redistributive and social dialogue effects; 
2. employers’ associations preferences and behaviours in relation to social policies. We adopt the same model as in the previous point, taking into consideration also the fact that preferences of employers’ associations among the four ideal-types are shaped by processes of negotiation within the business community (where interests and preferences might be divergent), with trade unions and governments (for example, “political exchanges” between different policy goals might affect preferences and actions). 

The explanans 

In general, the larger the company size, the more pressing the need of (high general or specific) skilled workers and the higher the risk of labour (skilled) shortages, the more probable is the likelihood that companies will move from a “no social policy preference” to one of the other three. To determine which one of the other preferences is influenced by the interplay among the three company level variables. For example, firms with: a) a high incidence of risk, a small size and low general skills should prefer “universalistic” social policies; b) a low incidence of risk, a small size and low general skills should opt for no policy at all; c) a high incidence of risk, a big size and general high or specific skills should prefer “contributory” social policies; d) a low incidence of risk, a big size and general high or specific skills should prefer “private” social policies. 

Meso-variables: the shapes of employers’ policy preferences taken by their collective action through employer associations. It is possible to differentiate among three types of labour market coordination by firms: 

• the “macrocorporatist” model, where employers are organized into hierarchically ordered groups, and the peak association negotiates broad political agreements with labour and the state through collective bargaining and tripartite policy-making committees; 
• the “sectoral coordination” model, where employers wield power largely at the industry level – employers’ associations within specific industries engage in significant coordination with corresponding labour unions but the encompassing multisector peak associations are much weaker and the state is largely absent from negotiations; 
• the “pluralist” model, where employers are represented by a panoply of conflicting groups, with many purporting to aggregate business interests and none having much policy-making authority. 

These three models not only represent employers’ interests to different degrees and with different outcomes, but they are also able to partially shape single employers’ preferences. In sum, employers’ associations can play a pivotal role to foster and to shape social policy preferences by their members, reaching its maximum strength in macrocorporatist models and its weakest one in the pluralist model. The strength of each model depends also on the share of companies joining employers’ associations. 

Macro-variables (the country political economy model): in typical comparative political economy approaches, actors’ preferences and behaviours (in our case, employers and their associations) are also influenced by the environment they belong to and in particular by: 

• the prevalent model of industrial relations in the country; 
• the state of the economy (around 10-12 years afterwards the 2007-08 economic crisis), with particular reference to impact of economic crises and their effects on socio-economic institutions and welfare policies. In the context of Covid-19 we propose to focus on the peculiar traits of the pandemic. The latter is a typical systemic crisis (with potential massive effects on the future economic growth potential of European countries). It is important to address the issue of whether and how employers have changed their own priorities on welfare reforms in the post-Covid-19 context; 
• the structure of the economy (more based on internal demand vs more export-led, the role of the four different macro-sectors, the average size of companies); 
• the welfare state model (systems based more on cash transfers than services, using the traditional classifications adopted by Esping-Andersen and Ferrera); 
• the functioning of the labour market (with risks, on one side of unemployment, on the other of – skilled and/or low skilled – labour shortages).