March divided, fight united? Trade union cohesion and government appeal for concertation

Abstract

Why does the government appeal for concertation? Starting from the principal‒agent framework and delegation theory, the article argues that the government is more willing to share decision-making power with trade unions when the policy preferences endorsed by the unions are closer to those of the cabinet. Furthermore, it maintains that government propensity to negotiate with trade unions increases as the heterogeneity of union policy preferences grows because the cabinet can exploit its agenda-setting power to divide the union front. The article tests these two hypotheses through a longitudinal analysis of the Italian case (1946–2014). In detail, it takes advantage of two original datasets built through content analysis that provide unique in-depth information on the policy preferences of parties and cabinets and measures the policy positions of the main Italian trade unions, thus allowing assessment of their reciprocal heterogeneity. The results confirm the expectations.

Keywords:

• Government
• trade unions
• tripartite institutions
• collective bargaining
• transaction costs
• content analysis

Acknowledgements

An early version of this paper was discussed by Andreas Dür and Caelesta Braun at the ECPR General Conference in Prague. We are grateful for their detailed and helpful suggestions. We also thank the other participants at that conference for their useful comments.

Notes

1. The assumption that the status quo lies far away from the ideal point of trade unions seems credible for other reasons. Empirically, social pacts have been formulated either to increase welfare state expenditures (starting from lower levels of public spending, that were far away from trade union viewpoints) or manage times of crisis (to reduce unemployment, contain inflation, or avoid a national default by cutting public debt and public expenditure). In the latter context, despite the presence of high levels of wages and welfare expenditures, we argue that the status quo can be pushed to the right (away from trade union preferences) with the eruption of a crisis: e.g. the status quo can drift to the right if the level of welfare expenditure is so high as to cause national bankruptcy, or if the excessive wage level causes high levels of inflation and unemployment.

2. The original 56 CMP categories have been extended to 68 to account for some Italian peculiarities (e.g. references to the Catholic Church or the Soviet Union).

3. The literature paid less attention to the role of employer associations. Even so, we can assume that their policy preferences are close to those of the cabinet (Regini and Regalia 1997), or at least not skewed to the extreme, compared to those of trade unions.

4. We focused on the two most important unions, leaving aside the smaller Unione Italiana dei Lavoratori (Italian Union of Workers – UIL), founded in 1949 by republican and social democratic union members, which only represents approximately 10% of the unionised workers, and lies (ideologically) between CGIL and CISL.

5. According to the ICTWSS database, which reports information on social pacts in 51 countries between 1960 and 2014 (Visser 2015), we observe overall a social pact every 10 years (9.8% of the time), although this rate is double in Italy (11 pacts in 55 years).

6. This definition is equal to that employed in the CMP. As an example, we classified in this category – amongst others – the following quasi-sentences (own translation): ‘The cabinet aims to open a dialogue with the trade unions’ (Moro cabinet, 1974); ‘The cabinet hopes to meet, as a spokesman, a trade union that is strong, autonomous and representative of the various viewpoints existing within it’ (Craxi cabinet, 1983); ‘The dialogue with the social partners will be open and fair’ (Dini cabinet, 1995); ‘If the cabinet passes the confidence vote, it will first start a dialogue with the trade unions’ (D’Alema cabinet, 1998).

7. We do not distinguish whether the appeal is sincere or instrumental (and this can be a limitation of this study), although even instrumental appeals are costly and it is not easy to turn back from there.

8. The frequency of this category in PM speeches is on average higher than the other 29 categories. For instance, the salience of concertation is higher if compared to PM quasi-sentences devoted to the judiciary and slightly lower than quasi-sentences related to multiculturalism, traditional morality, or the environment.

9. To compute positions on the economic dimension, we employed the formula reported in the online appendix (Table A2).

10. The dataset will be made available at http://andreaceron.com. Results are robust even when controlling for other variables (see online appendix), such as the lagged dependent variable, trade union polarisation on concertation, the actual negotiation and signature of social pacts in a given year (Table A3); trade union density, number of strikes/strikers (Table A4); GDP, GDP growth, public debt and unemployment (Table A5) or alternative measures of government partisanship (Table A6). Notice that in M12 (Table A4) the number of observations is particularly low due to several missing values in union density (which has a positive though not statistically significant effect on the outcome).

11. The ‘indirect’ solution of proving their willingness to reach a compromise by moderating their own positions therefore seems more effective than making an open, explicit appeal to the cabinet.

12. This is no longer true if we consider the run-up to joining the Eurozone, when Italian governments had to bargain with trade unions to meet the Maastricht criteria (Ahlquist 2010; Hancké and Rhodes 2005).