Introduction
Ever since Pierson (Reference Pierson1996) claimed that all-out welfare state retrenchment was risky given the electoral punishment associated with it, researchers have been at pains to explain cases in which governments introduced cut-backs in the welfare state. Particularly puzzling are those cases in which social democrats retrenched. Since the electorate punishes social democrats for welfare state retrenchment (Schumacher et al., forthcoming), and a less generous welfare state does not equate well with traditional social democratic goals, it is difficult to see what social democrats gain with welfare state cut-backs. The rich literature on the consequences of social democratic government participation for welfare state generosity has produced inconclusive results. Although some studies find a general positive effect of social democratic incumbency (Allan and Scruggs, Reference Allan and Scruggs2004), there has been a tendency for some social democrats to support neoliberal reforms, such as reductions of unemployment benefits and more flexible dismissal protection (Green-Pedersen et al., Reference Green-Pedersen, Van Kersbergen and Hemerijck2001).
Our question is why social democrats engage in this kind of reform. Environmental incentives – such as party competition (Kitschelt, Reference Kitschelt2001), economic crises (Vis, Reference Vis2009) or public opinion shifts (Stimson et al., Reference Stimson, Mackuen and Erikson1995) – help to explain some cases. Nevertheless, social democratic parties have had different policy responses to, for example, the economic crises of the late 1970s, ranging from traditional social democratic approaches to neoliberal reforms. We explain this variation by studying intra-party conflict between office-motivated leaders and policy-motivated activists. Environmental conditions, such as the threat of economic voting against the government, may motivate party leaders to push for neoliberal reforms to serve their office goals. Activists, on the other hand, push for traditional social democratic goals, regardless of the state of the economy. In the end, the balance of power between activists and leaders decides the policy response.
We explain the link between environmental incentives and the intra-party balance of power by analyzing the reform trajectories of social democratic parties in three European countries: the Social Democratic Party in Germany (Sozialdemokratische Partei Deutschlands, SPD); the Labour Party in the Netherlands (Partij van de Arbeid, PvdA); and the Socialist Party in Spain (Partido Socialista Obrero Español, PSOE). As will be argued, the literature is theoretically and empirically ambiguous about the agency of social democrats in labor market reforms and for that reason we focus on these parties in this paper (Schmidt, Reference Schmidt1996). We concentrate on a key policy area for social democrats – labor market policy. We broadly define our dependent variable as the difference between what is commonly understood as neoliberal reforms (unemployment benefit retrenchment and employment deregulation) and traditional social democratic approaches toward the labor market (maintaining benefit generosity and employment regulation). We focus on turning points in which parties switch from social democratic to neoliberal approaches or vice versa. Both in partisan orientation and voter orientation, labor market policies are more divisive than other areas of the welfare state. While the elderly and those suffering from bad health are usually considered as the ‘deserving poor’, the unemployed are more often seen as ‘undeserving’ (e.g. Van Oorschot, Reference Van Oorschot2006). Hence, there is more consensus about insuring against income deprivation in cases of old age and sickness than in the case of unemployment. Moreover, the mass unemployment of past decades pressured policy makers to reform the institutional framework of the labor market. For these reasons labor market policies are a fertile ground to test hypotheses about office-seeking and policy-seeking behavior. Our independent variables are environmental incentives, which include public opinion shifts, economic conditions, and party competition. Their importance is mediated by the intra-party balance of power, with the expectation of leadership dominance being associated with a higher responsiveness to these incentives. Our cases represent different types of intra-party balances of power, which allows us to analyze the implications of our proposition (Kitschelt, Reference Kitschelt1994).
We show that office-seeking social democrats retrench once an underperforming economy endangers the party's electoral position. Policy-seeking social democrats do not retrench, but do experience party organizational change if prolonged exclusion from office enfranchises more pragmatic elites at the expense of policy-seeking activists. This feedback effect transforms a policy-seeking party into an office-seeking party. We contribute to the retrenchment literature by indicating that different intra-party balances of power elicit different policy responses. This leads to a better understanding of the seemingly inconsistent behavior of social democratic parties – both cross-sectionally and across time.
Social democrats and neoliberal reforms
Social democratic parties are traditionally seen as agents of the working class, thus pursuing expansion of social benefits and employment regulation (Korpi, Reference Korpi1983; Huber and Stephens, Reference Huber and Stephens2001). In this perspective, varying levels of welfare state generosity are primarily explained by the relative strength of social democrats and other left-wing parties vis-à-vis conservative or liberal parties, which represent business interests. Hence, partisanship is seen as a key explanatory factor of labor market policy change.
Arguing that fiscal pressures since the late 1970s have changed the logic of welfare state politics, the new politics literature challenges the explanatory power of partisanship (Pierson, Reference Pierson1996). Without the possibilities for large-scale budgetary expansion, the left can no longer expand the welfare state, while the right refrains from welfare state retrenchment because even their constituents have become supporters of the welfare state. As a result, partisan differences recede. While there is evidence of a declining role of partisan differences (e.g. Huber and Stephens, Reference Huber and Stephens2001; Kittel and Obinger, Reference Kittel and Obinger2003), evidence in other literature indicates that social democrats are more likely than liberals and conservatives to cut-back on the welfare state and deregulate the labor market. Because of their positive role in welfare state expansion, social democrats have greater credibility to adapt existing policies and institutions responsibly. Hence, voters trust reformist social democrats to find a balanced mode of modernization – a level of trust liberal and conservative parties do not enjoy. This provides social democrats with greater leeway for reforms than openly liberal governments (Levy, Reference Levy1999; Ross, Reference Ross2000). Social democrats exploit this leeway in order to maintain welfare state support as their crucial power resource (Klitgaard, Reference Klitgaard2007). On the other hand it has also been argued that social democrats in post-industrial societies focus on their traditional core constituency (male workers in manufacturing). To sustain acquired privileges for this group, social democrats are willing to enforce insider–outsider divisions, for instance by deregulating temporary forms of employment while maintaining protection for regular workers (Rueda, Reference Rueda2007).
However, claims of a neutral (or even positive) impact of social democratic government participation on retrenchment are contested. Various empirical studies show persistent partisan differences, with left-wing incumbency still positively correlated to welfare state generosity (Korpi, Reference Korpi2003; Allan and Scruggs, Reference Allan and Scruggs2004; Amable et al., Reference Amable, Gatti and Schumacher2006).
Hence, existing research is theoretically and empirically ambiguous about the agency of social democrats with regards to neoliberal labor market reforms. In other words, what is the motivation for social democrats to introduce neoliberal reforms? Social democrats may respond to right-wing public opinion shifts or anticipate electoral punishment for poor economic performance. This might explain why socio-economic pressures (e.g. unemployment) translate into retrenchment (Vis, Reference Vis2009; Eichhorst and Marx, Reference Eichhorst and Marx2011). Kitschelt (Reference Kitschelt2001) shows that the spatial configuration of party systems shapes the possibilities for social democrats to reform without getting punished. However, these are office-seeking reasons for why social democrats retrench. They do not take into account that party leaders may fail to convince policy-motivated activists of the necessity of shifting to a neoliberal approach (Müller and Strøm, Reference Müller and Strøm1999).Footnote 1 Hence, parties should not be analyzed as unitary actors, as is often assumed in political-economy approaches to retrenchment (Merkel and Petring, Reference Merkel and Petring2008). In the following section we develop an explanation of why social democrats embark on neoliberal reforms, taking into account the intra-party balance of power.
Social democrats and the intra-party balance of power
Parties care about both policies and office (Müller and Strøm, Reference Müller and Strøm1999). Therefore, the (dis)utility of retrenchment for a party lies in the effect the measure has on the sum of the party's office and policy pay-offs.Footnote 2 Hence, parties enact a retrenchment measure if they believe that the measure will have a positive effect on the party's pay-off. Not every party values office and policy pay-offs to the same extent (Müller and Strøm, Reference Müller and Strøm1999). Achieving policy goals often presupposes access to office, but is a party in office to achieve policy goals or office goals (Budge and Laver, Reference Budge and Laver1986)? Parties sometimes have to make hard choices between policies that serve office goals and policies that serve ideological goals. Which approach a party adopts is influenced by the intra-party balance of power (Harmel et al., Reference Harmel, Heo, Tan and Janda1995). In one extreme a party is completely dominated by exclusively policy-motivated activistsFootnote 3 and therefore only accepts the party's policy preference as government policy. Standard social democratic ideology dictates high levels of generosity and strong employment regulation. If party activists dictate party policy, we expect that these activists choose policies along the lines of standard social democratic ideology or seek to retain the policy choices already made in that direction (Hypothesis 1).
Hypothesis 1: If party activists are dominant, social democrats retain social democratic policies.
In the other extreme, a party only cares about office pay-offs and has no policy preference. To obtain or maintain office pay-offs, a party needs votes and, if necessary, coalition partners. To keep (or first win) votes, the party should support policies that are supported by a large number of voters. But public opinion can shift. Bad economic performance (i.e. rising inflation or unemployment) is known to influence the short-term decision to vote for or against the incumbent, that is, retrospective economic voting (Duch and Stevenson, Reference Duch and Stevenson2008). More generally, voters dissatisfied with economic performance can be expected to shift to the right and become favorable to liberal or conservative policies. In such cases social democratic parties, whether in office or in opposition, must consider neoliberal policies to prevent losing office pay-offs, since the (rightwards shifting) center of a policy distribution has the highest density of voters. Capturing the center of a policy distribution should be a sufficient condition for winning elections in a majoritarian electoral system. However, in proportional representation systems, adopting centrist policies ensures that one is proximate to other parties, which helps the formation of a multiparty coalition. Hence, in both electoral systems social democrats have an office-seeking incentive to shift to the center.
However, if opposition parties or unions mobilize effectively against a social democratic government intending to retrench, it severely damages the prospects of social democrats remaining in office (either by losing votes or coalition partners – or both). If there is a party as pro-welfare as social democrats (left-libertarian parties, Christian democrats), voters have an alternative to vote for once they become dissatisfied with the neoliberal policies of the social democrats. Labor unions can also become instrumental in mobilizing public opinion against social democrats. However, if unions and opposition parties fail to mobilize public opinion or are not credible, social democrats do not need to fear electoral punishment and subsequent departure from office. Hence, in economically troubled times party leaders face a choice between losing votes due to economic mismanagement or due to punishment for implementing unpopular policies. If leaders calculate that the impact of the economy on their future office pay-offs is larger than the impact of credible opposition mobilization, then social democratic party leaders shift to neoliberal policies (Hypothesis 2). If there is no economic crisis, or credible opposition mobilization, leaders refrain from shifting towards neoliberal policies.
Hypothesis 2: If party leaders are dominant, poor economic performance and the absence of left-wing mobilization motivate party leaders to adopt neoliberal policies.
The choice for a particular labor market strategy may result in election defeat, exclusion from office or even party splits. Such negative feedback effects can persuade either side (activists or leaders) that a different strategy is necessary – or they enfranchise new, more pragmatic or more ideological groups at the expense of older groups. Hence, there is a dynamic interaction between policy choices and their success. These strategic choices change the distribution of power and enfranchise groups that shift the party's strategy from office-seeking to policy-seeking or vice versa (Hypothesis 3).
Hypothesis 3: Negative feedback effects from earlier labor market choices prompt parties to shift from office-seeking to policy-seeking strategies or vice versa.
Methods and case selection
To answer the question ‘Why do social democrats retrench?’, we analyze the trajectory of the PSOE, SPD, and PvdA social democrats in the period 1975–2010. We select this time frame because social democrats in this period had to adapt from the politics of ever-expanding government budgets to the more restricted politics of austerity (Pierson, Reference Pierson1996). We choose the case study method because the theoretical concepts of interest are context-dependent and should therefore be analyzed from a qualitative perspective. Although a quantitative or formal analysis would surely yield interesting results, the explicit goal here is to conduct an in-depth investigation into intra-party decision-making. The diachronic perspective of each case study allows us to incorporate the feedback effects of earlier decisions.
The unit of analysis is the turning point signifying a party's departure from a previous policy trajectory. We identify turning points by comparing actual policy (or the presentation of policy choices in party manifestos) with earlier policy choices. Coherent reversals in language and policy decisions signify a turning point. In each case study we indicate what explains turning points and the absence of a turning point, that is, the continuation of the status quo. As stated in the previous section, our explanatory variables are economic context, public opinion, and credible opposition mobilization, with the intra-party balance of power considered an intervening variable.
We selected the PSOE, SPD, and PvdA because they cover meaningful variation in terms of the intra-party balance of power. Kitschelt (Reference Kitschelt1994) combines two expert survey questions about the balance of power between party leaders and party activists to create a ratio measure of the intra-party balance of power between activists and leaders.Footnote 4 Using this measurement, the PSOE is identified as a strong office-seeking party and the PvdA as a policy-seeking party. Therefore we expect the PSOE to follow the predictions of H2 and the PvdA to follow the predictions of H1. The SPD has a value in-between these two extremes – therefore, we expect it to zigzag between office-seeking (H2) and policy-seeking (H1) motivations. Furthermore, Kitschelt (Reference Kitschelt1994) compares this expert survey to more qualitative measures of party organization. He finds that policy-seeking parties such as the PvdA are characterized by decentralization, union linkage and the presence of left-wing factions. Conversely, office-seeking parties such as the PSOE are centralized, have no linkages with civil society organizations and no significant intra-party factions. Choosing parties from three different organizational types allows us to make some tentative generalizations about party behavior in general.
Netherlands: initial radicalization, eventual moderation
We start our analysis with the PvdA, a party which has been characterized as a policy-seeking party (Kitschelt, Reference Kitschelt1994; Hillebrand and Irwin, Reference Hillebrand and Irwin1999). In the mid-1960s the PvdA was moderately left-wing. There had been increasing frustration with the lack of electoral success and the compromises the party had to make when in office. The PvdA advocated welfare state expansion, but was duly aware of the inflationary consequences of excessive government spending, as their efforts to curb spending in the 1965–66 cabinet under Cals demonstrate (van Praag, Reference Van Praag1990). At the same time, an influential paper from the party's research bureau argued that the welfare state did not alleviate social inequality and poverty still persisted (Bleich, Reference Bleich1986; van Praag, Reference Van Praag1990). In the late 1960s a group of young party activists emerged from the party and called themselves the New Left (Nieuw Links). Frustrated with the politics of accommodation, they complained that the PvdA swapped office pay-offs for policy purity too anxiously. They demanded a more democratic party and a stronger emphasis on social and economic equality (Wolinetz, Reference Wolinetz1977). Because the party elites partly agreed with the New Left and because they feared party split-offs, the party elite gave in to some of the demands (Bleich, Reference Bleich1986; van Praag, Reference Van Praag1990; van Praag, Reference Van Praag1994). The activism and media connections of the New Left also gave them a powerful position in relation to the party elite and the more aloof, older party members (Wolinetz, Reference Wolinetz1977). This informal change in the balance of power between new party activists and the old party elite elicited formal changes in the intra-party balance of power. In 1969 the candidate selection procedure was decentralized to powerful regional councils, which operated independently of the party executive (Koole and Leijenaar, Reference Koole and Leijenaar1988; van Praag, Reference Van Praag1994).Footnote 5 It was also decided that only a limited number of members of parliament (MPs) were to be allowed on the party executive board (van Praag, Reference Van Praag1990). Both changes limited the power of the party leadership in favor of the party activists. Consequently, the New Left penetrated all layers of the party organization and demanded extensive programmatic commitments (Wolinetz, Reference Wolinetz1977; van Praag, Reference Van Praag1990).
Triggered by a negative feedback effect of the office-seeking strategy of accommodation (Hypothesis 3), these shifts in the intra-party balance of power led to the adoption of a new policy-seeking strategy of polarization (van Praag, Reference Van Praag1990, Reference Van Praag1994; Green-Pedersen and van Kersbergen, Reference Green-Pedersen and Van Kersbergen2002). It was believed by some that by emphasizing left-right differences support for the centrist confessional parties would plummet and the PvdA would gain, and that in combination with other parties (Democraten 66, Politieke Partij Radikalen, and Pacifistisch Socialistische Partij) a progressive majority could be a distinct possibility (van Praag, Reference Van Praag1990). This new strategy to obtain policy and office pay-offs marks an important turning point. There was a strong preference for welfare state expansion, and the power of the activists vis-à-vis the party leadership ensured policy purity. Hence, office was less important than ideology. In addition, supporting the welfare state brought the PvdA a gain in votes (from 37 seats in 1967 to 52 seats in 1986). Consequently, the PvdA proposed rather dramatic demands in coalition negotiations till 1989 (van Praag, Reference Van Praag1990) and radicalized its demands for social welfare (Schumacher, forthcoming), even when the economic crisis of the 1970s meant a rising budget deficit and ever increasing government debt.
In terms of office pay-offs the strategy worked only once during the progressive Den Uyl cabinet, from 1972 to 1977, which needed parliamentary support from the confessional parties. The PvdA was expelled from government between 1977 and 1989 with the exception of a brief interlude in 1981. In that year the party leader, Den Uyl, was probably convinced that the utility of retrenchment was greater than the utility of the status quo because he negotiated a cut on sickness benefits.Footnote 6 However, the party and the unions resisted reform and eventually the party chairman, Max van den Berg, forced the PvdA ministers to resign from the cabinet (Bleich, Reference Bleich1986). Following Hypothesis 1, policy-motivated activists ensured that policy purity took precedence over office pay-offs. By stressing the anti-social character of the welfare state cut-backs by the first and second Lubbers governments, the PvdA was even able to play up its pro-welfare position during this period (Schumacher, forthcoming). This was well received by the voters, but left the party bereft of coalition partners. Eventually, this heralded the second turning point in the PvdA's welfare reform trajectory.
Deprived of policy influence and the spoils of office, the PvdA implemented drastic changes in the period 1989–92. Radical party activists had become more moderate and the PvdA electorate shifted to the right (Wolinetz, Reference Wolinetz1993; Schumacher, forthcoming). This negative feedback effect (Hypothesis 3) changed the balance of power within the party. A number of internal reports [Schuivende Panelen (PvdA, 1987); Bewogen Beweging (PvdA, 1988)] recognized the failures of polarization. However, recommendations for party organizational change were initially rejected by the party executive (van Praag, Reference Van Praag1994). The sudden breakdown of the coalition of the Christian Democratic Appeal (Christen-Democratisch Appèl, CDA) and the People's Party for Freedom and Democracy (Volkspartij voor Vrijheid en Democratie, VVD) in 1989 forced the PvdA to fight an election unreformed but with a new leader, Wim Kok. The PvdA mildly toned down its social welfare program and now proposed that in some areas the status quo in entitlements was acceptable instead of expansion (Schumacher, forthcoming). Since the CDA had also moved towards the position of the PvdA, they were able to form a coalition government (Lucardie and Voerman, Reference Lucardie and Voerman1989). Intended as a cabinet with a social image that would not engage in welfare state retrenchment, the third government led by Lubbers was quickly confronted with an unexpectedly high budget deficit. In 1991 almost a million people lived off disability benefits and a further million received unemployment benefits. In the summer of 1991 the cabinet put forward a drastic measure: for all people below 50 the worker disability benefit would be gradually decreased to 70 percent of the minimum wage – previously this had been 70 percent of last-earned income (Green-Pedersen et al., Reference Green-Pedersen, Van Kersbergen and Hemerijck2001). Party activists and unions protested, yet the PvdA congress eventually agreed to this drastic measure when Wim Kok forced the congress to choose between him or the retrenchment (van Praag, Reference Van Praag1994).
Winning 80 percent of the vote, Wim Kok broke the internal opposition against retrenchment. Following Hypothesis 2, the office-seeking party leadership believed retrenchment was preferable to the status quo because its purpose was to signal to the electorate and the other parties that the PvdA had financially sound policies in order to safeguard future office pay-offs. Hence, not responding to the budgetary pressures and overwhelming use of welfare benefits would probably have sent the PvdA back to the opposition benches – once again confirming their financial irresponsibility. Dire electoral and office prospects motivated the leadership to retrench. Why did party activists agree? As argued, the moderation of activists and the right-wing shifts of the PvdA's members had probably swung back the informal intra-party balance of power towards favoring office pay-offs instead of policy pay-offs. In addition, there was no longer a clear policy alternative or a strongly organized left-wing group within the party. The unions organized exceptionally large demonstrations against government plans, but failed to obtain significant support from parliament. With the two major pro-welfare parties in cabinet (the PvdA and CDA), the only alternative for voters was the smallish Green-Left party.
Shortly after the summer of 1991 17 percent of PvdA members, including a number of prominent party activists and MPs, decided to leave the party, further reinforcing the balance of power tilted towards office-seeking behavior (van Praag, Reference Van Praag1994). This also expressed itself in formal party organizational changes. In two reforms in March 1992 the powerful party council was abolished, and in December 1992 the party executive was once again made responsible for candidate selection. The PvdA had once more become a centralized party (van Praag, Reference Van Praag1994). This enabled new, more liberal-oriented ‘Third Way’ politicians to move up to the party elite level (Green-Pedersen and van Kersbergen, Reference Green-Pedersen and Van Kersbergen2002). In the 1994 election the PvdA drastically moved towards the center and even formed a coalition with the liberal party (VVD). The content of its party manifesto shifted completely from emphasizing decommodification to stressing efficiency and more liberal issues such as freedom of choice (Schumacher, forthcoming). By shifting to the center the PvdA lost votes but won office, as it had enhanced its strategic position within the party system. With a centralized leadership and a more centrist electorate, the PvdA leadership was now in a position to implement retrenchment. The first government of Wim Kok, still responding to the economic difficulties of the early 1990s, introduced a relaxation in statutory dismissal protection for regular workers and tightened eligibility criteria for unemployment benefits (Green-Pedersen et al., Reference Green-Pedersen, Van Kersbergen and Hemerijck2001). His second cabinet implemented only a few reforms, aimed at the flexibilization of the labor market (Green-Pedersen and van Kersbergen, Reference Green-Pedersen and Van Kersbergen2002), because the economy had picked up considerably. After the second government under Wim Kok, the PvdA continued a moderate stance vis-à-vis the labor market and the welfare state, supporting occasional retrenchment when deemed necessary.Footnote 7
Spain: immediate moderation
Now we analyze the trajectory of an office-seeking party, the Spanish PSOE. Emerging out of the shadows of the Francoist era the PSOE initially adopted a Marxist stance. Its slogan in 1976 was ‘mass, Marxist and democratic’, and it even operated, arguably, to the left of the Communist Party (Partido Comunista de España, PCE; Share, Reference Share1988). However, the pragmatic leadership around Felipe González and Alfonse Guerra battled with party activists to strike Marxism off the party manifesto. They succeeded in this because they disempowered party activists by changing rules concerning which party members were eligible to vote at party conferences (Share, Reference Share1988). In addition, the closed party list electoral system in Spain benefits parties’ leadership because it gives them a more prominent role in candidate selection (Field, Reference Field2006). Taking into account the comparatively low member/vote ratio, the centralization of decision-making, the large patronage networks that were effectively built up by the party leadership, and the weak links with the General Union of Workers (Unión General de Trabajadores, UGT), we conclude that the party leadership was absolutely dominant in decision-making processes within the PSOE. As a consequence, the PSOE leadership met with few obstacles to their office-seeking ambitions.
In the 1980s and early 1990s the PSOE did not face a highly competitive electoral system, despite the fact that Spain has a majoritarian electoral system. Also, after the rapid demise of the centrist Union of the Democratic Centre (Unión de Centro Democrático, UCD) in the early 1980s no real competitor for power existed. The Peoples Party (Partido Popular, PP) – and its predecessors – were generally considered to be too far to the right and the communists (and later, to a lesser extent, the Izquierda Unida) as too radical and too divided (Share, Reference Share1988). Hence, looking both at the intra-party balance of power and the potential for credible mobilization against the government, the PSOE leadership was in very comfortable position to capture the center.
Table 1 Operationalization of independent variables
The outgoing UCD government had introduced employment legislation (Ley Básica de Empleo) that strongly decreased levels of unemployment benefit coverage and unemployment income replacement rate (Gutiérrez and Guillén, Reference Gutiérrez and Guillén2000; Mato, Reference Mato2011). In its 1982 election platform the PSOE promised job creation, redistribution of income, vastly increased social expenditure, a 40-hour working week with 30 days of vacation (Share, Reference Share1988; Royo, Reference Royo2000). However, when the PSOE came to power in 1982 it had to deal with very high levels of unemployment and a very rigid labor market. Instead, the first government under Felipe González opted for large-scale austerity measures, shedding large numbers of workers in the industrial sector. Following Hypothesis 2 the office-seeking PSOE – rather than satisfying the core social democratic voters – adopted centrist policies in response to economic decay, and tried to balance the budget and bring down inflation. Both union and left-wing opposition were very weak at the time and were unable to mobilize significant opposition. In 1984, the government addressed the issue of social security for the unemployed. Although they extended the coverage and generosity of the benefits system, it was targeted solely at the elderly and agricultural workers (Mato, Reference Mato2011). But the government also lowered the maximum unemployment benefit and introduced labor market deregulation by introducing temporary contracts (Gutiérrez and Guillén, Reference Gutiérrez and Guillén2000). In conclusion, the 1984 reforms were generally not expansionary, and solely benefited important electoral groups (mostly agricultural workers in Andalucia and Extremadura; Recio and Roca, Reference Recio and Roca1998). In terms of employment regulation it heralded the first signs of a neoliberal approach.
A more generous expansion of the welfare state came in 1989. The government extended the duration of unemployment benefits and assistance (Gutiérrez and Guillén, Reference Gutiérrez and Guillén2000; Mato, Reference Mato2011). But what triggered this turning point? The unions, who were initially weak and divided, had managed to organize large-scale strikes and demonstrations against the government in 1988. Since the economy was starting to pick up, the unions felt it was time for real social democratic policies (Share, Reference Share1988). The group of supporters around Alfonse Guerra (the Guerristas) had disapproved of the alienation between the PSOE and the UGT and fears of electoral reprisals eventually turned the tide within the party (Recio and Roca, Reference Recio and Roca1998).
The 1989 reforms were more of an aberration of PSOE centrist policies regarding the labor market than a real turning point, and in 1992 and 1993 the party continued on its course of labor market deregulation. This time they also combined this with retrenching the unemployment benefit and unemployment assistance systems. In 1992, the third cabinet under Felipe González decreased unemployment benefits from 70 percent to 60 percent of the last-earned salary, reduced the duration of the benefit, reduced the level of social assistance given after unemployment benefits had ended, and also introduced strict labor market activation measures. In 1993 they further reduced the scope of unemployment benefit and social assistance programs. What could have triggered the third González cabinet to embark on this unpopular course of action? Firstly, there was little to fear from other parties and intra-party dissent was also weak. The unions tried to repeat their successes of 1988 but failed. Apparently the sudden weakening of the Spanish economy in 1992, which had pushed up spending on passive welfare state measures (Gutiérrez and Guillén, Reference Gutiérrez and Guillén2000), combined with stricter European Union rules with regard to budget deficits, were incentives to implement deep cuts when the economy stalled. All of which verifies Hypothesis 2.
To provide an intellectual counterweight against the PSOE's policies, the Guerristas formulated a manifesto for a social democratic program for the future (Programa 2000). Despite large numbers of people that contributed to drafting the manifesto, it remained a factional attempt to set the agenda, which ultimately failed (Gillespie, Reference Gillespie1993).
González led the PSOE from 1974 to 1997. His policies are generally regarded as neoliberal (Share, Reference Share1988; Recio and Roca, Reference Recio and Roca1998). The policies generally benefited the elderly and agricultural workers at the expense of younger workers, and keeping spending down appealed to centrist and right-wing voters. With limited options on both the left and the right, the PSOE managed to gain a strong electoral alliance. By leading the PSOE from Marx to the market, it managed to stay in office from 1982 to 1996, which made it the most successful European social democratic party of that period.
The post-González era began with a dramatic electoral defeat in 2000 and renewed intra-party dissent – exemplified by the fact that the party conference twice elected a new party leader who was not chosen by the party leadership.Footnote 8 When José Zapatero was elected party leader in 2000, a gradual rejuvenation of the party began. The negative feedback effect (H3) from electoral defeat and damaging corruption practices heralded a change of generation in the PSOE leadership (Encarnación, Reference Encarnación2006) as well as the introduction of more democratic measures to elect candidates (Méndez Lago, Reference Méndez Lago2006). These changes led to an increased emphasis on post-material and identity issues. Despite the negative feedback effect, the minority Zapatero government that came to power in 2004 continued the economic policy of Gonzaléz (Royo, Reference Royo2006). José Zapatero even appointed the same minister for economic affairs (Pedro Solbes) as Felipe Gonzalez had done in his last two cabinets. In terms of unemployment benefits, the Zapatero government implemented two minor reforms, one easing access to unemployment benefits, the other slightly decreasing unemployment benefits (Mato, Reference Mato2011). Zapatero did not change Gonzalez’ economic policies because these policies had been effective in capturing the center of the electorate. This was necessary because the Partido Popular had slowly moved to the center, which culminated in their winning the 1996 elections. Fighting for the median voter, Zapatero could not afford to shift to the left on economic policies, nor was this necessary since intra-party discussion focused on post-material issues and the left-wing Izquierda Unida became largely irrelevant in electoral terms after 2000.Footnote 9
Given the economic boom in Spain, José Zapatero was in a comfortable position in the run-up to the 2008 elections. However, with the financial crisis in the United States, the economy suddenly became an important issue during the 2008 elections, which were narrowly won by the PSOE. His initial policy response to the crisis was the introduction of a number of temporary measures that would alleviate the income fall of the newly unemployed. Such measures were very much in line with how other countries responded to the crisis. The initial understanding of the problem was that consumer spending should be kept at a certain level to avoid recession. However, facing great pressures from the financial markets and other European countries, Zapatero implemented some radical austerity measures such as freezing pensions and civil servants’ pay and raising the retirement age. By cutting popular welfare programs and thereby working towards a more balanced budget, Zapatero tried to evade electoral punishment of voters dissatisfied with economic performance. This response is very similar to what his predecessor Gonzalez did and follows the predictions of Hypothesis 2. However, for the PSOE Zapatero's drive for austerity was probably too little and too late, because the party was defeated in the November 2011 elections by the Partido Popular, which proposed even more radical reforms.
In conclusion, the PSOE followed a centrist trajectory, allowing neoliberal reforms, but choosing a traditional social democratic approach once unions mobilized effectively and threatened the maintenance of office pay-offs.
Germany: between traditionalists and modernists
The SPD is characterized by a rather fragmented party structure (regionally as well as ideologically) and strong organizational ties to trade unions. Both aspects constitute constraints for office-seeking behavior and make programmatic reorientation contentious, particularly in areas such as labor market policy (Merkel and Petring, Reference Merkel and Petring2008).
In the 1970s the SPD's blend of economic pragmatism and traditional social democratic values formed the foundation for electoral success. Its first government participation in 1966 was followed by 16 consecutive years in office, most of them as a senior partner in a coalition with the liberal Free Democratic Party (Freie Demokratische Partei, FDP) (1969–82). Union interests (most importantly extended co-determination) ranked high on the SPD's agenda, particularly under the chancellorship of Brandt (1969–74). Although Schmidt's ascension to office meant a strengthened role for the pragmatist's intra-party faction and more contentious relations with trade unions (von Beyme, Reference Von Beyme1990), the policy response to the oil crisis remained very much in line with traditional social democratic ideology. It relied mainly on costly labor-supply reduction by means of passive benefits. Despite deteriorating labor market performance and growing deficits, the party was reluctant to adopt far-reaching retrenchment or deregulatory measures. This immobility eventually led the FDP to walk out of the government and form a new coalition with the Christian Democratic Union (Christlich Demokratische Union, CDU), which seemed to be more open to labor market modernization at the time.
The following 16 years of opposition clearly revealed the SPD's difficulties to compete with a strongly welfare-oriented Christian democratic rival. After some half-hearted deregulation of temporary employment in 1985, the CDU–FDP coalition followed an expansionary social policy course. After reunification especially the CDU boosted its popularity by implementing generous passive labor market policies cushioning the transition process (Zohlnhöfer, Reference Zohlnhöfer2003). In this situation the SPD could hardly win votes by moving to the right. At the same time, however, the decline of its traditional working-class constituency forced the SPD to open up to more liberal middle class voters (Egle and Henkes, Reference Egle and Henkes2004b).
In this context, the SPD struggled to develop a successful electoral strategy at the onset of the 1990s. Being torn between ‘traditionalists’ and ‘modernizers’ posed an additional challenge. The fragile intra-party power balance prevented a programmatic reorientation, which would have allowed the party to follow a coherent office-seeking strategy. This was expressed, for instance, in the high turnover of party leaders from the late 1980s (Meyer, Reference Meyer1999).Footnote 10 However, in the post-reunification period the opportunity structure changed for the better and the SPD managed to temporarily overcome internal divisions. This was mainly due to the apparent failure of the government's labor market policy. As a consequence of persistently high unemployment and related concerns about the sustainability of the costly German employment model, the electorate shifted to the right on economic issues. This, in turn, forced the CDU–FDP government into unpopular benefits cuts and employment regulation (Picot, Reference Picot2009).
The government's rightward programmatic shift opened up space for the SPD in the middle of the political spectrum. Yet the orientation of the party remained unclear, especially because competition for the party leadership continued. While party chairman Oskar Lafontaine represented the traditional wing of the SPD, Gerhard Schröder (who eventually became the candidate for chancellor) clearly belonged to the modernizing camp (Debus, Reference Debus2008). This dual leadership led to a rather vague program. However, it turned out to be beneficial during the 1998 campaign, as contradictory political demands could be catered to (Egle and Henkes, Reference Egle and Henkes2004b; Picot, Reference Picot2009). In terms of labor market policy, the SPD developed a promising electoral strategy in line with the demands of its traditional wing: it announced it would repeal the unpopular reforms introduced by the government, which eventually contributed to its victory together with the Greens.
The overall tendency of the first term under Schröder (1998–2002) was characterized by traditional social democratic labor market policy. The new government kept its pledge to repeal its predecessor's reforms. Instead of decreasing benefit generosity, revenues from the social insurance system were strengthened by obliging additional groups to pay contributions (Egle and Henkes, Reference Egle and Henkes2004a). With regard to temporary work, the SPD hesitantly tried to counteract the pattern of ‘dual’ labor law reforms under the CDU–FDP government, which had led to a constant increase of fixed-term contracts and agency work (Eichhorst and Marx, Reference Eichhorst and Marx2011). Although the new government did not fundamentally change the dual structure of employment protection, fixed-term contracts without a valid reason were made somewhat stricter. By and large, the initial labor market policy approach reflected the demands of the traditional unionist wing of the SPD. During the pre-2001 boom there was no reason to abandon this popular (and election-winning) strategy. Hence, the interests of activists and leaders temporarily converged.
However, it soon became clear that the balance between the intra-party factions remained fragile. The favorable economic situation had biased the policy approach towards popular measures in line with traditionalists’ interests. Yet towards the end of the legislative term, unemployment began to rise again, which made reconciling interests more difficult and provided incentives for the office-seeking leadership to shift to the right.
Schröder II (2002–05) brought about an important turning point in labor market policy, one which illustrates how the intra-party balance of power and external reform incentives interact. Against the background of a labor market crisis, falling approval ratings and disastrous state-level elections, Schröder announced ‘Agenda 2010’ (including the ‘Hartz reforms’) in 2003, by German standards a major package of labor market deregulation and welfare state retrenchment.Footnote 11 The reform proposals, launched by the leadership in the face of open opposition within the party, can be seen as a pragmatic response to the changing incentive structure, in particular centrist voters’ discontent with labor market performance and budget deficits. In line with Hypothesis 2, the office-seeking leadership pursued neoliberal reforms with the aim of reducing unemployment and social expenditure, thereby safeguarding future office pay-offs. Besides public opinion, the behavior of other parties supported this strategy. During the 1990s, all major parties had shifted rightwards in their social and economic policy positions, indicating a consensus on the need to reform German labor market institutions (Saalfeld, Reference Saalfeld2006). The SPD was therefore able to successfully negotiate the reforms with the CDU, and because there was no significant left-wing competitor, the SPD leadership believed they would avoid negative vote pay-offs (Picot, Reference Picot2009).
Traditionalists, who still formed a majority within the party as well as in the parliamentary group, were clearly opposed to what they perceived as neoliberal reforms (Debus, Reference Debus2008; Conradt, Reference Conradt2010). To overcome this veto power, Schröder opted for a confrontational top-down approach. By abandoning intra-party dialogue and skillfully creating pressure via the media instead, he managed to push through his agenda (Korte, Reference Korte2007). One important element was to frame the changes as an inherent necessity, given Germany's structural problems. Moreover, Schröder relied on extra-parliamentary commissions to prepare reform proposals rather than on experts within the SPD (Saalfeld, Reference Saalfeld2006). At the 2004 party convention, the party base reluctantly accepted the reforms, which were predominantly seen to contradict social democratic principles (Meyer, Reference Meyer2007).
But the strategy backfired. A large part of the SPD's core constituency, traditionalist party activists, and the unions felt betrayed. As a direct response, a new far-left party split off and soon merged with the post-communist PDS (Partei des Demokratische Sozialismus), which previously had its stronghold in the former East Germany (the new party was called ‘Linkspartei’, later ‘Die Linke’). Die Linke quickly gained support from disappointed social democratic party members and competed aggressively with the SPD for left-leaning voters on welfare issues (Vail, Reference Vail2010). Die Linke's explicit aim was to revoke the liberal changes of the renegade social democrats. In the subsequent federal elections in 2005 Die Linke did surprisingly well, which contributed to the downfall of the Schröder government (Conradt, Reference Conradt2010).
The emergence of a powerful left-wing competitor was an immediate feedback effect of the Hartz reforms (Hypothesis 3), which created substantial constraints for the SPD's course in the subsequent Grand Coalition (2005–09). Moreover, internal conflict increased in the post-Schröder SPD. While the traumatized left majority pushed for corrections to the previous reforms, the party leadership was still formed by modernizers who, in principle, sought to protect the political achievements of the ‘Hartz reforms’ (Raschke, Reference Raschke2010). Thus, the SPD had both electoral and intra-party reasons to shift leftwards. This tendency was supported by the socio-economic context. Arguably, because of the institutional changes introduced under Schröder, Germany experienced job-intensive growth (until entering the economic crisis in 2008). This shifted public attention from unemployment to growing labor market inequalities and increased the government's fiscal leeway for expansionary social policy (Schmidt, Reference Schmidt2010; Eichhorst and Marx, Reference Eichhorst and Marx2011).
New left-wing competition, internal discontent, and the socio-economic background explain why the SPD rapidly turned from retrenchment to reversing some of the most unpopular elements of the Hartz legislation.Footnote 12 It also put minimum wages on the agenda and agreed with the reluctant CDU on the extension of minimum wage agreements in specific sectors. This leftward trajectory continued after a disastrous performance in the 2009 elections, which put an end to the SPD's government participation. After obtaining veto power in the upper house of the German parliament, it pushed the CDU–FDP coalition to implement more generous modifications of social assistance and better working conditions for agency workers in 2011.
The development of the SPD's labor market policy approach after 1998 illustrates the complex interactions between the intra-party balance of power and reform incentives as well as the importance of feedback effects. First, in the ideologically fragmented SPD the socio-economic background and the perceived pressure for reform had a strong impact on the balance between traditionalists and modernizers. The party leadership was only able to overcome policy-seeking veto factions by pointing to the factual constraints created by the labor market crisis. In times of economic upswings a policy-seeking strategy usually prevailed, because the interests of leaders and activists converged. Second, the results of internal conflict created feedback effects. This is illustrated by the emergence of Die Linke, but also by the internal reorientation within the SPD, which took place after the political failure of retrenchment.
Comparing cases
Based on the qualitative information gathered in the case studies, we can now specify the conditions conducive to neoliberal shifts of Social Democratic parties. Table 2 summarizes the explanatory variables at times of turning points in labor market policy approaches. They indicate that when the party leadership dominates party activists, party leaders responded to adverse economic situations (and related public opinion shifts) and had the reasonable expectation they would not be punished for the reforms (i.e. there was no credible opposition mobilization). This applies to the SPD (2002), the PvdA (1991) and PSOE (1984 and 1991). This confirms Hypothesis 2, as is most apparent in the case of the PSOE. The constant leadership dominance allowed the PSOE to maneuver the party without significant internal constraints. The high responsiveness to environmental incentives explains the rapid shifts between social democratic and neoliberal approaches, which all were in line with an office-seeking strategy. Hence, the PSOE adopted a neoliberal approach when the economy was going badly and credible opposition mobilization was absent (1984, 1994), but stressed its social democratic values when the economy improved and union mobilization threatened its access to office (1989).
Table 2 Comparing SPD, PvdA, and PSOE 1975–2010
SPD = Sozialdemokratische Partei Deutschlands; PvdA = Partij van de Arbeid; PSOE = Partido Socialista Obrero Español.
The PvdA's organizational changes confirm the importance of the intra-party balance of power. Following Hypothesis 1, activists’ dominance ensured that the party remained on a social democratic track, despite exacerbating economic conditions and exclusion from office. Under similar economic conditions, the Dutch social democrats rejected the retrenchment policies their Spanish sister-party implemented. The PvdA switched to a neoliberal response in 1991, because organizational reforms, motivated by negative feedback effects, had empowered party leaders over party activists (Hypothesis 3). Admittedly, the potential for opposition mobilization in the Netherlands decreased somewhat over this period, providing an alternative (or complementary) explanation for changed party behavior.
Finally, the SPD represents an in-between case, for which the balance between activists and leaders is rather unstable. Under favorable economic conditions, the interests of activists and leaders converged and both pushed for a traditional social democratic agenda. However, under unfavorable economic conditions clashes between traditionalists and modernists emerged, and only in 2002 did the modernists win this battle and push for neoliberal reforms. Hence, like the PvdA the response of the SPD to economic troubles was much delayed, in contrast to the swift response of the PSOE. Also, the unanticipated consequences of the Hartz IV reforms created a grave crisis for party leaders and re-established activist dominance (Hypothesis 3). This partly explains the SPD's return to social democratic policies after 2005.
Conclusion
Although we examine only three case studies, we contend that additional cases can be explained by our framework. Comparing reform strategies of other European social democratic parties along the party organizational dimension, Kitschelt (Reference Kitschelt1994) reveals broad similarities to our cases. First, the UK Labour Party is much like the PvdA in terms of party organization. In the early 1980s the radical left obtained a powerful position within the Labour Party, which radicalized its party program. After rather dramatic party organizational changes in 1995, Labour shifted to the right, adopting neoliberal policies. Second, the Spanish and French socialists are both dominated by their party leaderships. Similarly to the Spanish socialists, the French PS departed from its ideological position and implemented retrenchment measures in the 1980s when the economic context called for it. In contrast to the PSOE, the PS’ more social democratic approach in the 1990s is explained by a pattern of party competition with left-wing rivals whose support was necessary to obtain a majority in parliament (e.g. the ‘Gauche plurielle’ under Jospin). Third, the Swedish social democratic party organization bears resemblance to the SPD's party organization. Both parties were reluctant to retrench and fluctuated from traditionalism to neoliberalism and back. Hence, despite very similar economic conditions, these social democratic parties differ in their responses to these conditions.
While in-depth research is necessary to strengthen these claims, this brief overview shows that our explanatory framework provides plausible hypotheses for studies into the behavior of social democratic parties. The main argument of the paper is that the intra-party balance of power explains the different behavior of social democratic parties in a comparative perspective. Our findings show that if the party leadership is dominant, social democrats turn from a social democratic to a neoliberal labor market approach if the economy is going badly and credible opposition fails to materialize (Hypothesis 2). The leadership anticipates that losses due to economic voting or decreased trust among coalition partners in the party's ability to weather economic storms are larger than potential losses due to credible opposition mobilization. Hence, the office-seeking social democratic leadership pursues a neoliberal course of action to safeguard future office pay-offs. However, social democratic parties under similar economic conditions choose more social democratic approaches to the labor market if party activists are strong enough to veto a shift away from the party's traditional policies (Hypothesis 1). The choice for social democracy or neoliberalism has feedback effects and electoral defeat or exclusion from government motivate party change and the choice for another strategy (Hypothesis 3).
In conclusion, the larger lesson for the welfare state literature is that political ideology and economic conditions do not have a generalizable impact on the welfare state irrespective of the organizational structure of parties. This structure shapes the strategies parties choose and thereby determines which environmental incentive is relevant for explaining welfare reform.
Acknowledgements
An earlier version of this paper was presented at the Political Parties and Welfare State Reform workshop at the University of Aarhus (19/1). Many thanks to Christoffer Green-Pedersen, Carsten Jensen, Patrick Emmenegger, Christian Elmelund-Praestekaer, Christoph Arndt, Georg Picot, Barbara Vis, and Kees van Kersbergen.
