Title: The politics of social protection: social expenditure vs market regulation
Abstract: Abstract It has been argued that the notion of a European social model is misleading and that there are in fact different European social models with different features and different performances in terms of efficiency and equity. In this paper, we look at the welfare state from a political economy point of view and interpret the different regimes as possible outcomes of a political process through which heterogeneous preferences of voters are aggregated. In our model, agents differ in two respects: income and socio‐economic vulnerability. Policy‐makers have to decide on two policies: a proportional income tax to finance a social transfer, providing equal benefits to all citizens, and a market regulation policy which benefits only vulnerable workers, providing them with additional protection against unemployment risk. Market regulation is inefficient because it decreases aggregate resources. Individuals' heterogeneity generates a conflict over policies. We feature the political process as a two‐party electoral competition in a citizen–candidate model with probabilistic voting. We show that an inefficient equilibrium exists and that this outcome is more likely as income inequality and the proportion of vulnerable workers become greater. Intuitively, greater inequality raises the level of redistributive spending desired by the poor, making, at the same time, the rich more adverse to the welfare state. In this framework, both the rich and the poor, in order to win the election and realise the fiscal gain, have an incentive to support market restrictions, in the attempt to capture the votes of the vulnerable minority, who benefit from these policies. Keywords: welfare statesocial protectionmarket regulationspolitical processpolitical economyJEL classification code: D72H53L5J65 Acknowledgement Work for this article was partly funded by the EU FP6 project 'Inequality: Mechanisms, Effects, Policies' (INEQ, FP6‐029093). Notes 1. PMR measures the degree to which policies promote or inhibit competition in areas of the product market where competition is viable. The scale of the indicator is 0–6 from least to most restrictive of competition. For more information on the PMR indicator system, see Conway, Janod and Nicoletti (2005 Conway, P., Janod, V. and Nicoletti, G. 2005. "Product market regulation in OECD countries: 1998 to 2003". Paris: OECD. OECD Economic Department Working Paper No. 419 [Google Scholar]). 2. Esping‐Andersen (1990 Esping‐Andersen, G. 1990. The three worlds of welfare capitalism, Cambridge: Polity Press. [Google Scholar]) who was the first to emphasise cross‐national differences in European welfare state systems, had originally identified three regimes (social‐democratic in Nordic countries, liberal in Anglo‐Saxon countries and corporative‐conservative in continental Europe). Subsequent literature has taken up and extended his classification by adding a fourth model (Mediterranean). For a survey of the debate following Esping‐Andersen's classification, see Arts and Gelissen (2002 Arts, W. and Gelissen, J. 2002. Three worlds of welfare capitalism or more?. Journal of European Social Policy, 12(2): 137–58. [Crossref], [Web of Science ®] , [Google Scholar]). 3. There is consistent empirical evidence that strict EPL reduces unemployment turnover, with fewer dismissals but also fewer unemployed being hired. However, no consensus exists on the overall impact of EPL on unemployment. Recently, however, Djankov et al. (2004 Djankov, S., Botero, J., La Porta, R., Lopez‐de‐Silanes, F. and Shleifer, A. 2004. The regulation of labor. Quarterly Journal of Economics, 119: 1339–82. [Crossref] , [Google Scholar]) find that an increase in the employment laws index is associated with an increase in black market activity, a reduction in labour force participation and an increase in unemployment rates. 4. Since more restrictive product market regulations are associated with greater firms' rents, then increased product market regulation may increase workers' incentive to bargain for capturing part of these rents (Blanchard and Giavazzi 2003 Blanchard, O. and Giavazzi, F. 2003. Macroeconomic effects of regulations and deregulations in goods and labour markets. Quarterly Journal of Economics, 118(3): 879–907. [Crossref], [Web of Science ®] , [Google Scholar]). 5. A recent contribution from OECD (Arjona, Ladaique and Pearson. 2002 Arjona, R., Ladaique, M. and Pearson, M. 2002. "Social protection and growth". Paris: OECD. OECD Economic Studies No. 35 [Google Scholar]) finds that total social protection expenditures reduce growth. However, the empirical literature reviewed in the paper does not reach a consensus on a systematic effect of social spending on growth. 6. See Cusack, Iversen and Rhem (2005 Cusack, T., Iversen, T. and Rehm, P. 2005. "Risk at work: the demand and supply sides of government redistribution". Berlin, , Germany: Social Science Research Center (WZB. [Google Scholar]). 7. On this point, see D'Orlando and Ferrante (2007 D'Orlando, F. and Ferrante, F. 2007. "The demand for job protection: some clues from behavioural economics". University of Cassino. Mimeo [Google Scholar]). 8. As is evident by our description of political competition (see below), we do not invoke lobbying to obtain our results. 9. In fact, median voter equilibrium is guaranteed in a one‐dimensional issue space under the assumption of singled peaked preferences (or single crossing). If preferences are not single peaked or if the issue space is multi‐dimensional Nash equilibrium of a majoritarian voting game may fail to exist. The literature has proposed various technical solutions to this problem reaching the conclusion that in case of multidimensional policies, in order to predict which groups will be the most powerful in the struggle for benefits, the details of the political process must be specified (see Persson and Tabellini 2000 Persson, T. and Tabellini, G. 2000. Political economics, explaining economic policy, Cambridge, MA: MIT Press. [Google Scholar]). 10. Roemer (1998 Roemer, J. 1998. Why the poor do not expropriate the rich: an old argument in a new garb. Journal of Public Economics, 70: 399–424. [Crossref], [Web of Science ®] , [Google Scholar]) was the first to examine how the existence of a second issue, other than general redistribution, affects policy outcomes in a model with political parties. He shows that the existence of another salient issue (e.g. religion) can work against the pure economic interests of the poor if this non‐economic issue is sufficiently important. 11. For a survey of the multidimensional political economy approach to redistributive public policies, see Profeta (2007 Profeta, P. 2007. "Income inequality and redistributive public policies: a survey of the multidimensional political economy approach". Research Report, Project VI European Research Framework Programme 'Inequality: Mechanisms, Effects, Polices [Google Scholar]). 12. This result is in contrast with the one in Boeri, Conde‐Ruiz and Galasso (2007 Boeri, T., Conde‐Ruiz, I. and Galasso, V. 2007. "The political economy of flexicurity". Mimeo [Google Scholar]); the reason is that while in our framework a more unequal distribution of income implies that redistribution is the salient issue – thus leaving room for the minority to obtain their preferred policy on the other issue (market regulation) – in Boeri et al. a large wage differential induces low‐skilled workers, if they are pivotal, to support generous UB and therefore low EPL. 13. Besides simplicity, the assumption of rigid labour supply allows us to not deal with the disincentive argument, according to which social protection distorts economic decisions. 14. We focus on unemployment and assume, for simplicity, that the average productivity of employed workers does not change over time. 15. By the term 'deep‐rooted' we mean sectors well equipped to cope with international competition. 16. If productivity and vulnerability are independent then the proportion of vulnerable (non‐vulnerable) among high income individuals is the same as the one among low income individuals. Moreover, if productivity and labour market status are independent then the proportion of high (low) productivity individuals among employed workers is the same as the one among unemployed workers. Formally, the proportion of individuals of type i−k is: for i = NV, V, U and k = H, L. Thus, for k = H, L, the proportion of the different types in the population are: and . 17. Therefore, at t = 1 the probability of being dismissed for vulnerable workers is equal to ϕ. 18. For simplicity, we are assuming that all unemployed workers have the same probability of being hired. 19. Social public transfers, can be cash or in‐kind and include, public services such as health care and education provision, unemployment benefits and active labour policies. Social transfers can be classified by eligibility and benefit criteria (see discussion in Bertola et al. 2001 Bertola, G., Jimeno, J.F., Marimon, R. and Pissarides, C. 2001. "EU welfare systems and labour markets: diverse in the past, integrated in the future?". In Welfare and employment in a United Europe, Edited by: Bertola, G., Boeri, T. and Nicoletti, G. Cambridge, MA: MIT Press. [Crossref] , [Google Scholar]). Here we focus on (basic) security transfers to which all citizens are entitled and which are established on a flat‐rate basis. 20. See Sen (1997 Sen, A. 1997. Inequality, unemployment and contemporary Europe. International Labour Review, 136: 155–72. [Web of Science ®] , [Google Scholar]) for a detailed description of the non‐pecuniary costs associated to job deprivation. 21. We have in mind measures that limit competition in the product market, such as legal and administrative barriers to entrepreneurship and to international trade and investment, and, in the labour market, such as EPL and immigration laws. 22. Note that ϕ(π) denotes the probability to be dismissed for vulnerable workers, which coincides with the shock probability ϕ when π = 0. 23. As discussed in the introduction, MR is also likely to affect productivity. Accordingly, we could have assumed that unemployed workers hold more up‐to‐date competence than vulnerable workers. Note, however, that because our concern is with aggregate resources, the two assumptions point in the same direction, i.e. the inefficiency of MR. Moreover, because of non‐pecuniary costs associated with job deprivation, increased unemployment has additional social costs. 24. Recall that (∂u 2/∂π) > 0, (∂ϕ(π)/∂π) < 0 and (∂ϑ/∂π) < 0. Also note that when τ = 0 non‐vulnerable workers are indifferent to market regulation: if there is nothing to be re‐distributed then MR cannot reduce social public expenditure, which is already nil. 25. Note that this assumption implies that the majority of employed workers is of type NV, that is γNV > γV . 26. One of the most widely quoted facts is that under the systems of plurality rule there are two main parties (this is the so‐called Duverger's Law, after Duverger (1954 Duverger, M. 1954. Political parties, London: Methuen. [Google Scholar])). One possible explanation of this stylised fact is that if there are three or more parties, at least one can withdraw giving its vote to the 'closest' party and triggering the victory of the latter (see Osborne (1995 Osborne, M. 1995. Spatial models of political competition under plurality rule: a survey of some explanations of the number of candidates and the positions they take. Canadian Journal of Economics, 27: 261–301. [Crossref] , [Google Scholar]), for an account of ideas that explain Duverger's Law). 27. This is consistent with Assumption 4. 28. Alternatively, we could have assumed that party members select candidates via some type of bargaining process. In this case the candidate chosen would maximise the expected payoff of a pivotal party member. Under majority voting the pivotal party member is the median one. For a discussion of inter and intra party competition in general elections, see Roemer (2004 Roemer, J. 2004. "Modelling party competition in general elections". In Oxford Handbook of Political Economy, Edited by: Weingast, B. and Wittman, D. Oxford: Oxford University Press. [Google Scholar]) and Levy (2004 Levy, G. 2004. A model of political parties. Journal of Economic Theory, 115(2): 250–77. [Crossref] , [Google Scholar]). 29. Party A's median member is of type NV if (1 − u 1)γNV γL > (1 − u 1)γV γL + u 1 that is, γNV > γV + [u 1/(1 − u 1)γL ] which is a stronger condition than Assumption 4. As for party B, having assumed γNV > γV the median member must be of type NV. 30. The description of the noise vote is based on Besley and Coate (2000 Besley, T. and Coate, S. 2000. "Issue unbundling via citizens' initiatives". NBER Working Paper Series 8036 Cambridge, MA: National Bureau of Economic Research. [Crossref] , [Google Scholar], 2003 Besley, T. and Coate, S. 2003. Elected versus appointed regulators: theory and evidence. Journal of European Economic Association, 1: 1176–206. [Crossref], [Web of Science ®] , [Google Scholar]). 31. If redistribution were the only issue at stake, unemployed and low productivity workers would vote for party A while high productivity workers would vote for party B; the difference in the share of the vote for parties is therefore (1 − u 1)(γL − γH ) + u 1. 32. Note that parties are not Downsian: namely, they do not wish to maximise the probability of winning the election per se, but rather to maximise the expected payoff of their median member. 33. If redistributive policy were salient for the whole population, the equilibrium of the policy game would be straightforward: rational voters would vote for the candidate who shares their redistributive policy preferences and the equilibrium outcome would depend on the median voter preferences. 34. Recall that .
Publication Year: 2009
Publication Date: 2009-05-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 5
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot