Against the political expectations and theoretical models: how to implement austerity and not to lose political power

Abstract

During the global financial crisis in 2008–10 Estonia symbolized a pathway of fiscal consolidation and austerity. Moreover, despite opting for austerity and all the social and political consequences of achieving fiscal consolidation, the governing coalition in Estonia succeeded also in avoiding a negative political reaction from the voters and remained in power. The current article analyses the variables that made the austerity reforms in Estonia in 2008–11 electorally successful. The economic success of Estonia has otherwise been attributed to a combination of political, institutional, and economic factors: timing; a fiscal policy that was not pro-cyclical; the availability of reserves in Estonia; and the ownership structure of the banks. The present study asks more specifically about the variables that saw the economic success in the implementation of the austerity measures accompanied by a positive electoral outcome. As it will be argued, these factors include the communication strategy chosen intentionally or otherwise by the government, the design of the austerity measures, the peculiarity of the electoral cycle, lack of political alternatives, and the performance of neighbouring countries in implementing austerity measures.

Keywords:

• austerity
• electoral behaviour
• euro area
• fiscal consolidation

JEL classification codes:

• E42
• E58
• G18
• H12

1. Introduction

During the global financial crisis in 2008–10, austerity was often described as the best possible solution for achieving fiscal consolidation in the euro area and its member states. However, even when austerity was accepted on the ideological level as an economically rational measure, the politicians of most euro area member states have introduced it only to a limited extent or have chosen less stressful and politically less risky strategies. Austerity was rejected by most politicians not because it was argued to be economically irrational or unachievable, but primarily because it was considered as a certain way to lose political power.

The explanation was that austerity would cause additional economic and social pressure as it tends to lower the quality and quantity of the services and benefits provided by the government, to create additional tax pressure on the public, and bring a negative political reaction from the electorate to replace the ruling government that had initiated the reforms (see e.g. Pierson, 2011; Guajardo, Leigh, & Pescatori, 2011).

Estonian politicians, who had an earlier experience of shock therapy from the early years of re-independence, were among the minority of the European governments who decided to adopt austerity measures with all the economic and social consequences. Surprisingly, in contradiction to the theoretical models and general expectations, the governing parties in Estonia stayed in power after the elections of the national parliament in 2011.

Raudla and Kattel (2013) state that the economic success of Estonia in recent years can be attributed to a combination of political, institutional, and economic factors (timing, a fiscal policy that was not pro-cyclical, the availability of reserves in Estonia, the ownership structure of banks, the electoral cycle, and the significantly smaller revenue shock in Estonia) and conclude that the Estonian experience cannot be reproduced elsewhere in the EU. They also argue that other countries cannot use entry to the euro area as a crisis exit strategy as Estonia did. While the current study agrees with the main conclusions of their study, it emphasizes additionally the fact that the positive electoral outcome of the austerity policy in Estonia was also produced by the country-specific factors including the government's communication strategy, the specific design of the austerity measures, the peculiarity of the electoral cycle, the lack of political alternatives, and the performance of neighbouring countries during the crisis.

As there might be a need for a fiscally successful and politically sustainable austerity model for a number of euro area member states, Estonia can provide valuable evidence as to why austerity triggered no negative public political reaction that was generally predicted in the respective theoretical models and feared by politicians.

The analysis concentrates on the results of the parliamentary elections in 2011, but the results of the local elections from 2009 and in general terms the European Parliament (EP) election results from 2009 will also be analysed. It will also be discussed in what aspects Estonia meets the logic of theoretical models and evaluates the central variables for successful austerity and public support.

2. The theoretical debate: economic actions and political reaction of voters` during the austerity

The following section is aimed at analysing the main theoretical concepts and options about how voters are predicted to react to radical reforms and policy changes initiated by a national government. In this context, austerity measures like increasing taxes or cutting spending to reduce public expenditure can be classified as a specific type of short-term radical reform willingly initiated by a national government, which could be held responsible by voters for the socio-economic consequences (Gourevitch, 1986; Pierson, 2011). This research is based on a following logic: provided that austerity as a fiscal tool first creates additional social pressure, this should lead to an electoral reaction aiming at replacing the ruling coalition that initiated the reforms (Pierson, 2011, pp. 540–541).

The theoretical debates related to austerity cover a wide range of studies including political culture (Almond & Verba, 1989; Hirschman, 1994; Inglehart, 1997), electoral studies (Lipset & Rokkan, 1967, Reif & Schmitt, 1980), public policy communication models (Katzenstein, 1985; Lustick, 2000), and models of economic voting (Alvarez, Nagler, & Willette, 2000; Lewis-Beck & Stegmaier, 2007).

In terms of political and civic culture Almond and Verba (1989) analyse the past experience of political reforms of democratic countries and point out that long-term and successful experience of democracy contributes to the development of a participant political culture, while citizens of countries with only a limited experience of democracy tend to be more subordinate. Accordingly, in countries with a limited experience of democracy (such as Estonia) the governing party is less likely to face the negative consequences of painful social reforms. This idea is also supported by Moore (1966) who states that democracies that are not historically consolidated have a stronger tendency towards non-democratic and non-participatory development when faced with pressure and instability. A similar logic could also be applied to a wider social and philosophical debate on the sustainability of the welfare state, the role of the state in general, and society's moral judgements on the acceptable level of certain economic indicators like the budget deficit, gross debt, or the money supply. Some societies apply more inclusive, but thereby also socially sensitive, models with restrictions on the government for reducing salaries, social security benefits, or healthcare costs. For those countries austerity is almost impossible to implement and internal devaluation of the national currency should be considered as the main key to crisis solution (Pierson, 2011). In contrast, countries that traditionally prefer what could be called socially ignorant models (like Estonia) more easily accept government actions in cutting pensions and salaries and even reducing healthcare accessibility to combat the fiscal imbalances.

The model of Inglehart (1997) focuses on the attitudes and orientation of the public towards the political system as a playground and their role and existing choices as the players in that playground. First, the orientation of citizens – which should be the basis for their further actions – could be cognitive, affective, or evaluative, and thus, trust, participation, and attitudes towards central government policy could either be ‘participant’, ‘subjective/subordinate’, or ‘parochial’. Theoretically, these categories should cause different reactions to the political decisions of the government, e.g. the ‘participants’ tend to be most reactive and the ‘parochial’ are most ignorant or silent in accepting the government's decisions. If there is active pressure, the measures taken by the government and its values are ignored as far as possible by the public in the parochial model because this type of citizen do not perceive themselves to be components or members of the political system. In the subjective/subordinate political culture, voters act according to the expected rules, express their opinion in elections, and tend to support government parties. In a participant political culture, voters are also aware of their political power in between the elections but may act decisively when needed by voting against the ruling coalition (Inglehart, 1997). In that connection the parochial model could be related to the statement by Putnam (1993) that voters’ ignorance and lack of political reaction should be understood as signs of low social capital, lack of trust, or unconventional political participation, as citizens do not believe they have the ability to change the government in elections and therefore might prefer to withdraw themselves from political participation.

In the context of political culture, Hirschman (1994, pp. 272–273) analyses an individual's choices in reaction to socio-economic stress and comes to the conclusion that:

there are two types of response to unsatisfactory situations in one’s firm, organization or country. The first is ‘exit’ or leaving without trying to fix things. The second is ‘voice’, that is, speaking up and trying to remedy the defects.

Hirschman also points out that the response is based on the individual's loyalty and preferences, and also on what the individual is able to do, since not all inhabitants are citizens with the right to vote and not all groups have the rational option of leaving if they are dissatisfied. He claims that in the choice of ‘exit’ or ‘voice’, the level of ‘loyalty’ depends on national history and earlier experience. The tragedy is that when a nation is pushed to choose between exit and voice, the most competitive groups tend to leave and fighting is left to those who are unable to leave (Hirschman, 1994). Whether the majority of the disappointed voters tend to ignore the elections, a vote for the government or the opposition or a decision to emigrate also depends on the ideological line-up of the existing parties and the existence of ideologically acceptable alternatives to the government (Lipset & Rokkan, 1967).

At the same time, it also needs to be stressed that mobilization is not equal among voters, as negatively minded citizens are more likely to participate in elections than satisfied groups, which leads to a negative overreaction to austerity. The voters most opposed to the government in this case tend to be those groups which have had personal experience of austerity, and groups which were traditional supporters of the government but suffered from austerity or were disappointed in the government's policy (Kittel & Obinger, 2003, pp. 22–24).

Attention also needs to be paid to the aspect that voters’ behaviour tends to be different in ‘first order’ and ‘second order’ elections. The ‘first order’ elections are the national parliamentary elections in parliamentary systems and the national presidential elections in presidential systems, while municipal elections and various regional elections are considered ‘second order’ elections. The distinction is based on the logic that there is less at stake in secondary elections and so fewer voters consider them sufficiently important to bother to cast their votes (Reif & Schmitt, 1980). Since the ‘first order’ elections are considered a priority for voters as well as for politicians, the ‘punishment’ reaction should rather follow ‘first order’ elections, i.e. national parliamentary elections or presidential elections. The time factor is also of particular importance – reactive punishment will follow the social stress caused by the government in the first possible occasion.

The government's conduct and the image it projects of itself and its ability to influence social processes also play significant roles in the electoral feedback to austerity measures (see e.g. Katzenstein, 1985; Lustick, 2000; Skocpol, Evans, & Rueschemeyer, 1985). In this connection, Katzenstein (1985) analyses the role of corporatist governance in maintaining the political stability in a small European state under conditions of high economic openness and economic fluctuations. Structural issues mean that the government is placed in the role of ‘taker’ instead of ‘maker’ and therefore the government needs to follow a different path to legitimize its decisions in the eyes of voters (Katzenstein, 1985). This statement is supplemented by the model of partial actor and partial force majeure (see Lustick, 2000), which offers another explanation of how the negative reaction of voters could be avoided in elections. More precisely, should the austerity be understood by the citizens as a voluntary choice of the government or the one and only rational solution forced by economic reality, the reaction of voters is perceived to be different. Therefore, should the government succeed in convincing the electorate that they were not active decision-makers or that austerity was not their free political choice among other alternatives, but the one and only rational option left because of the financial or institutional force majeure, a more positive electoral outcome could be expected from austerity reforms, as in this case the government maintains its image as a sovereign member of the political system, but escapes the responsibility in the voters’ eyes as ‘it did its best’. As Katzenstein (1985) says, this strategy requires that countries perceived by voters to be in a similar situation behave in the same way. If neighbouring countries in a comparable situation have decided to take different and less stressful measures, then the government needs to rely on the partial actor approach where austerity is shown as the one and only reasonable option forced by the supranational institutions or induced by market forces. In this case the government presents itself as merely a policy-taker, not a policy-maker.

Unexpected support for austerity measures could also be related to voters’ current experience of austerity and their future expectations, i.e. whether voters have personal negative experience of austerity or whether public expectations for the crisis solution are even more negative than the reality. The model by Lipset and Rokkan (1967) shows that if a government wants to avoid negative consequences in elections, two points are crucial – as negative as possible an image of the crisis needs to be created in order to lower public expectations, while the number of people personally affected by the austerity measures should be reduced as low as possible. In the Estonian case of austerity, also a wider social philosophical debate and media image over comparative sustainability of the welfare state and Thatcherism has been in the centre of attention. Here the main variable is the coalition's ability to create future images of a successful state and convince the electorate passively to support it by actively monopolizing policy feedback in the media (Pierson, 1996, pp. 176–177).

Finally, the voters’ reaction to the austerity reforms also depends on the economic outcome of the measures taken. According to the models of economic voting (Lewis-Beck & Stegmaier, 2007), voters punish the government parties for worsening economic conditions. This is amplified due to the fact that the economy is a stronger factor in bad times than in good (Alvarez, Nagler, & Willette, 2000).

3. The Estonian government's actions to overcome the financial crisis

There are two central questions to be addressed in the following empirical overview. First, it should establish if the Estonian government can be considered as a policy-taker or a policy-maker during the financial crisis. And second, which were the central variables of governmental policy and strategic communication?

The recessions in Estonia can generally be divided into two distinct phases: the first phase lasted from the second half of 2007 to the intensification of the global financial crisis in the autumn of 2008, and the second phase from the autumn of 2008 until the end of the recession by the end of 2009.

The starting economic circumstances in Estonia in 2007 were quite specific when compared to its regional neighbours or euro area members. First, there were several positive variables easing the entrance to the crisis: there existed a tradition of balanced central budgets, small central government debt (4.4% from the gross domestic product (GDP)), and a special reserve fund for a crisis situation. There were also no government bonds issued in the whole history of re-independence.

As Estonia was using the currency board system, government was a passive actor in terms of money supply and interest rates. Additionally, the Estonian budgetary law did not allow to submit budgets even with minor deficit, even when in the EU level the Maastricht rules allowed a 3% budget deficit for the states applying for euro area membership. There were no limitations to sovereign debt, but at the beginning of the crisis Estonia did not have a tradition of national bonds. The government's ability to finance a budget deficit was also influenced by the actual interest rate asked by the commercial banks.

The first phase of the recession was relatively mild and saw a domestic demand-led adjustment that was primarily related to a deceleration in credit growth. The second phase of the recession, which started with the deepening of the global financial crisis in September 2008, saw a sharp reversal of foreign capital flows and a steep fall in exports. The austerity measures taken in Estonia in 2008–11 were triggered by the fall of foreign capital inflows and lending in the first quarter of 2008, which resulted in FDI dropping from its previous level of 2 billion euros in 2007 to 1.2 billion euros in 2008, causing an immediate drop in GDP in the second quarter of 2008 and a fall in budget revenues (Statistics Estonia, 2013a). As a reaction to the fall in budget revenues, a budget reduction was passed by the parliament of Estonia in June 2008, reducing spending for the rest of 2008 by 384 million euros, which was almost 7% of the budget.

To reduce growing expenditures, public sector wages were first reduced by around 20% in the last quarter of 2008 and investments and numerous state procurements were cancelled (see Table 1). Despite these austerity measures, even after the correction the total revenue side of the budget remained in deficit in 2008. The government's austerity policy caused a chain reaction where the reduction of public wages and cancellation of investments and state procurements reduced the government's tax revenues and payments to commercial banks too; as a result, the purchasing power and the ability of the private sector to service the existing debt declined. The only group which fully escaped the cuts was pensioners, as the average monthly pension even grew by about 20% in 2008, which has been described as the government's ‘insurance policy’ to maintain political support among voters in an ageing society.

Table 1. Main economic indicators for Estonia, 2007–12.

Indicator	2007	2008	2009	2010	2011	2012
GDP (chain-linked volume growth, %)	7.5	−4.2	−14.1	2.6	9.6	3.9
Total general government revenue (EUR b)	5.843	5.961	5.978	5.841	6.269	6.831
Change in total general government revenue (%, in comparison to previous year)		2.02	0.29	–2.29	7.33	8.96
Total general government expenditure (EUR b)	5.46	6.44	6.251	5.813	6.088	6.872
Change in total general government expenditure (%, in comparison to previous year)		17.95	–2.93	–7.01	4.73	12.88
Net lending (+)/net borrowing (−)	0.383	–0.479	–0.273	0.028	0.181	–0.041
Average monthly gross salaries (EUR)	724.5	825.2	783.8	792.3	839	887
Change in average monthly gross wages and salaries (%, in comparison to previous year)		13.90	–5.02	1.08	5.89	5.72
Civil servants wages from average wages (%)	108	106	96	98	99	101
Monthly average old-age pension (EUR)	226.3	278.4	301.3	304.5	305.1	312.9
Change in monthly average old-age pension (%, in comparison to previous year)		23.02	8.23	1.06	0.20	2.56
Unemployment rate (%)	4.6	5.5	13.5	16.7	12.3	10
Consumer price index (%)	6.6	10.4	–0.1	3	5	3.9
Harmonized competitiveness indicator (%)	4.2	6.1	1.6	–2.8	1.3	–0.3

Source: Bank of Estonia (2013); author's calculations in italics.

In 2009, Estonian budget planners had the even more complicated task of fiscal consolidation. As a subsequent response to the economic chain reaction that had erupted at the end of 2008, the Estonian government decided to make drastic expenditure cuts across the board in December 2008. The latter were preferred to the targeted cuts which were not used as consensus was not reached on the priority of sectors. The majority of the expenditure cuts were seen as temporary and planned to last for two years (Government of Estonia, 2008). The cuts resulted in a central government budget deficit of 2.9% which narrowly met the Eurozone accession criteria on government budget deficit (maximum 3% from GDP), but was extraordinarily large under the current budget law.

Thus for the first time since Estonia had regained independence, a budget with deficit was planned in 2009. As a result of the drastic budget cuts Estonia experienced a fall in GDP reaching 14.1% and a decline in industrial output of 24%, while average monthly gross wages and salaries dropped by 5% and unemployment increased by 8 percentage points. At the same time, as in 2008, pensions were not cut in 2009 but grew again by 8% from an average of 278 euros per month in 2008 to 301 euros in 2009.

Although the Estonian economy showed some encouraging signs of growth in 2010, the budget was planned in a conservative way. The revenue side of the budget remained in general on the same level as in 2009 at around 5.8 billion euros while spending was cut by about 7% from 2009. Spending and revenues were balanced mainly due to the government's drastic decision to halt payments into the second pillar of the Estonian pension system. Although payments were restored after two years and the state even increased its contributions to compensate for the losses, the popular belief that the pension system is firm and unshakable has not been restored. In 2010, Estonia also fulfilled the Maastricht criteria and decided to adopt the euro and join the euro area in 2011.

The government budget for 2011 was planned rather optimistically, based on the view that both the Estonian economy and government spending would be growing. As the Estonian economy saw growth both in terms of GDP and industrial output, a budget deficit was avoided, and revenues exceeded the expenditures by 1.2% (relative of GDP). Salaries grew together with pensions, and while the growth in the average pension was largely nominal at 1 euro per year, average salaries grew by around 5% (Statistics Estonia, 2013a). While fiscal factors began to normalize, social tensions and pressure continued, caused by the high unemployment level (12.3%), reduced healthcare accessibility, high inflation (5%) and limited credit options by private banks. As a result, labour migration especially towards Nordic countries grew.

In 2012, the GDP growth in Estonia stabilized at 3.9% and aggregate gross domestic product reached the highest level in the country's history (Bank of Estonia, 2013). Although the budget proposal for 2012 was initially planned to be in deficit with expenditures exceeding revenues by 0.68 billion euros, at 6.80 billion euros to 6.12 billion euros, in actual terms of fulfilment the budget was more balanced at 6.4 billion euros (Ministry of Finance, 2012). Both wages and pensions continued to grow as average salaries increased by 5.7% and the average monthly pension by 2.5% (Bank of Estonia, 2013).

Which were the social effects of austerity measures taken by the government in 2008–11? Despite the government's efforts, costs for social policy were also growing during the crisis, pushed by a quickly growing consumer price index and political promises to raise pensions every year. According to Organization for Economic Co-operation and Development (OECD) data, Estonian social costs grew from 12% of GDP in 2007 to 18% of GDP in 2011 (OECD, 2011). To the Estonian labour market, the influence of crisis and budget cuts was evident but arriving late. If before the crisis in 2007 the unemployment rate was 4.7%, then in 2008 it rose only to the 5.6% level since employers tended to hope that the crisis would be short and preferred temporary salary cuts instead of reducing the labour force (Statistics Estonia, 2013a). In 2009, however, unemployment more than doubled to the level of 13.8%, and in 2010 it reached the top level of 16.9%, and lowered to the 12.5% level only in 2011 (Masso & Krillo, 2011).

Average monthly wages reached 724 euros in 2007, followed by more than 10% growth to 825 euros in 2008, a decline to 783 euros in 2009, a small growth in 2010 to 792 euros and a 5% growth in 2011 to 839 euros. What differed in the Estonian wage correction process and dynamics from the other Baltic States was that at the beginning of the crisis Estonia's public sector salaries were higher than private sector salaries (Masso & Espenberg, 2013).

During the years of crisis, social pressure started to influence migration (starting from the year 2010), when emigration grew by 25% on a yearly basis. In 2011 this growth continued with the speed of 20% and reached a remarkable 80% growth in 2012 (Statistics Estonia, 2013b).

How did the Estonian government use the toolbox to achieve fiscal consolidation prescribed by the laws, which were the main actions indicating the government's role in initiating the austerity policy and which were the main principles of Estonian austerity policy?

To implement austerity the Estonian government chose to achieve fiscal consolidation by taking the path of across-the-board measures (Pollitt, 2010, pp. 17–18) rather than targeted cuts. Although it is easier to achieve consensus for across-the-board measures, the targeted cuts approach would in an ideal case offer higher economic returns or less social stress. Analysing the duration of the measures (i.e. whether the cuts and the immediate burdens are being introduced as a temporary or a long-term measure), we can see that the government acted exactly according to the theory described by Rubin (1980) and started with the minimum possible cuts in expenditures across the board to compensate for the lack of revenues while hoping that the crisis would prove temporary. Later the government continued ‘slicing’ the costs to reach a balance, which it followed with unsuccessful attempts to increase revenues as the economic situation deteriorated and drastic cuts in spending were needed to avoid a budget deficit in 2009 and 2010. There were not many alternatives to austerity for the Estonian government. Since Estonia had a currency board at the time, the government was unable to influence money supply and interest rates. Although there were no limitations on sovereign debt, Estonia had no tradition of issuing national bonds at the beginning of the crisis, and although the government had a legal right to change the exchange rate, in practical terms it rejected this option as it would have resulted in failure to meet the euro area accession criteria. And so austerity focused only on budget corrections.

With regard to long-term crisis management, the fundamental issue facing the Estonian government during the economic crisis was the question of whether fiscal consolidation should be achieved through cutting costs, increasing revenues or going for both options simultaneously. Since meeting the euro area accession criteria (the Maastricht criteria) was a parallel goal for the government alongside crisis stabilization, the Estonian political elite decided to choose the first option. In this way the Estonian experience also offers a valuable contribution to the theoretical framework of fiscal consolidation which would generally tend to support the opposite model of immediate gains through revenue-based measures, with spending cuts following only in case revenue raising proves insufficient (see Maher & Deller, 2007; Wolman & Davis, 1980).

In terms of strategic political communication the Estonian government followed the logic discussed by Katzenstein (1985) and Pierson (1996) for a small state corporatist government where the government presents itself as a policy ‘taker’ in a partial force majeure situation.

The government purposely took the role of partial actor (policy-taker) who merely tries to save the situation caused by a global crisis. And the government did succeed in convincing the voters that the fiscal situation was desperate and its choices were limited but that full commitment to austerity would save the situation and even if the government did not succeed, it was morally and economically the most sustainable solution.

Was austerity a necessity or the only option for the Estonian government to achieve fiscal consolidation? In terms of macroeconomic conditions, the decline of the GDP by 14.1% in 2009 was one of the highest in the EU, but it was lower compared to the Latvian economy which dropped by −17.1% and quite equal to the Lithuanian economy decline of −14.8%. At least in the Baltic context, the Estonian GDP suffered less than the Latvian or Lithuanian GDPs (Statistics Estonia, 2014). In the Baltic context Estonia was also the most stable country in terms of budget balance. The highest deficit in Estonia occurred in 2008 reaching 2.9% only to be followed by 2% in 2009. In 2007 and 2011 the Estonian budget saw surpluses. Estonia was also the only country among the Baltic States having a special stabilization reserve for the case of economic shocks. Next to witnessing a balanced budget, also Estonian tax revenues were growing during the years of crisis – starting from 31% of GDP in 2007 and reaching 34% in 2010. While this level is significantly lower than the EU-27 average (39% in 2010) it is still higher than the tax levels in Latvia (27.5% in 2010) or Lithuania (27.4% in 2010). Estonia started the debt crisis with the lowest debt level (4.6% of GDP in 2008) in the EU and euro area; Latvian debt reached 19.5% in 2008 and Lithuania met the crisis with a debt of 15.6% in 2008. In 2009 governmental debt was growing in Estonia, in all selected reference countries and in the EU on average. While Estonia almost doubled its debt level to a mere 7.2%, Latvia reached 36% and Lithuania 29%. The average governmental debt of the European Union's member states in the same time increased from  61.6% in 2008 to 73.6% in 2009. One further indicator which can explain Estonia choosing austerity during the financial crisis is inflation. Estonian inflation increased by 6.6% in 2007 and reached 10.4% already in 2008. While there was no direct economic need to counter inflation with austerity, austerity nevertheless offered Estonia an excellent possibility to bring inflation below 3% which was the main obstacle in its way to fulfilling Maastricht criteria and entering the Eurozone.

Based on the above described indicators, it can be admitted that austerity was not the one and only fiscal choice for the Estonian government. In terms of GDP dynamics the Estonian situation was better or equal to its Baltic neighbours and in respect of the fiscal deficit and governmental debt Estonia scored the best results in the euro area and the whole European Union. Hence, the main reason for choosing austerity seemed to be the willingness of the Estonian government to reduce the inflation level in order to be able to access the euro area.

One major argument used to legitimize austerity as a partial force majeure was the fact that in 2007–13 Estonia was allocated more than 3.4 billion euros from the EU Structural Funds and so Estonia simply had to follow European guidelines. The need to find active support among the member states of the euro area for Estonia's entry was used as another argument.

At least the Estonian government itself considered its ability to manage the economic crisis and simultaneously maintain voters’ support to be an example to others – the Estonian prime minister and the government coalition have expressed the opinion that Estonia was one of the most successful EU members in combating the financial crisis and that other member states of the EU should learn from the experience of Estonia (Ansip, 2012).

4. Voters’ reaction to the austerity measures: the case of Estonia

The following part will research which citizens’ political reactions to austerity in elections were detected during its implementation and shortly after. During this task, the main research focus is on the results of the parliamentary (Riigikogu) elections in 2011, but also the European Parliament election results from 2010 and elections of the local municipalities in 2009 will be analysed. The focus of the current study will be only on election results as no major protest or strike against salary cuts or social pressure took place in Estonia in the years of austerity.

As according to the theoretical models, reactive punishment by voters should follow the social stress caused by the government at the first suitable option during or after spending cuts (which started in Estonia in December 2008 and continued until January 2011), it could be expected that the voters’ dissatisfaction would be first visible in the elections of the local authorities in October 2009 and the ‘punishing’ behaviour should occur particularly in national parliament (Riigikogu) elections in 2011, but also in the EP elections in 2010.

The Estonian elections’ timing seems almost ideal for the testing, as voters’ reaction can be evaluated in all the main stages of austerity, while the main case – the parliamentary elections followed the end of austerity almost immediately. It is important to add, that even when the Estonian economy was in general growing in 2011 (particularly the private sector), the governmental budget was to the end of 2010 and for first the quarter of 2011 still in track of austerity and therefore also the quality of social services and public sector salaries suffered until to the parliamentary elections in beginning of March 2011.

In analysing the political reaction of the voters to the austerity measures, a first level conclusion could be drawn in terms of participation, indicating whether voters chose the ‘exit’ strategy and ignored the elections or opted for the ‘voice’ strategy and perceived themselves as members of the political system (see Hirschman, 1994). A comparison of the levels of participation in the elections to the local municipalities in 2009 and in the previous elections of the same type and in the elections to the local municipalities in 2013 (see Figure 1, panel (a)) reveals that the number of people participating in the elections in 2009 was higher and that more than 60% of those who had the right to vote, did so (Electoral Committee of Estonia, 2013).

Figure 1. The dynamics of the electorate and participation in the Estonian elections in 1992–2013. Source: Electoral Committee of Estonia (2013).

The level of participation in 2011 parliamentary elections was higher than in the previous equivalent four years before at 63.5%, but was still not comparable to the high participation in the first years of independence, when 67.8% turned out in 1992 and 68.9% in 1995. Electronic voting had a positive effect, especially in the case of the European Parliament elections (participation almost doubled), but the effect was less visible in the parliamentary elections in 2011.

The voters did not use their option of negative voting in ‘first order elections’ as suggested by Reif and Schmitt (1980), as the parliamentary elections in Estonia in 2011 showed growing support for the governing coalition of the Estonian Reform Party and the Pro Patria and Res Publica Union (it needs to be added that the Pro Patria and Res Publica Union joined coalition just in the middle of austerity in May 2009 and austerity reforms for 2009 budget were conducted by Social Democrats controlling the Ministry of Finance until May 2009).

The Reform Party received 28.6% of the votes and its coalition partner, the Pro Patria and Res Publica Union, increased its support and received 20.5% of the votes. The Estonian Social Democratic Party, which was also in the coalition until May 2009, received 17% of the votes, a surprising increase in support of 6.5 percentage points from the results of the parliamentary elections in 2007 (Table 2).

Table 2. Estonia's election results for 2007 and 2011 Riigikogu elections.

Party	Ideology	Year	Votes	% of votes and change	Seats in parliament
Estonian Reform Party	Classical liberalism	2007	153,044	27.8% (+10.1%)	31 (+12)
		2011	164,275	28.6% (+0.8%)	33 (+2)
Estonian Centre Party	Social liberalism	2007	143,518	26.1% (+0.7%)	29 (+1)
		2011	134,090	23.3% (−2.8%)	26 (−3)
Union of Pro Patria and Res Publica	Conservatism	2007	98,347	17.9% (−14%)	19 (−16)
		2011	118,023	20.5% (+2.6%)	23 (+4)
Social Democratic Party	Social democracy	2007	58,363	10.6% (+3.6%)	10 (+4)
		2011	98,302	17.1% (+6.5%)	19 (+9)
Estonian Green Party	Green politics	2007	39,279	7.1% (+7.1%)	6 (+6)
		2011	21,828	3.8% (−3.3%)	0 (−6)
People's Union	Agrarianism	2007	39,215	7.1% (−5.9%)	6 (−7)
		2011	12,192	2.1% (−5%)	0 (−6)
Russian Party in Estonia	Russian minority	2007	1084	0.2% (0)	0 (0)
		2011	5027	0.9 (+0.7)	0

Source: Electoral Committee of Estonia (2013).

The main losers in the parliamentary elections were the Estonian Green Party and the People's Union, both strong opponents of austerity, who both lost their previous representation of six seats in parliament and were excluded from Riigikogu (Electoral Committee of Estonia, 2013). The leader of the opposition and the main opponent of austerity, the Estonian Centre Party, received 23.3% of votes, which was 2.8 percentage points less than in 2007.

After the parliamentary elections in 2011 the number of seats of the coalition in the Estonian parliament increased by two for the Estonian Reform Party (ERP) and four for the Pro Patria and Res Publica Union (PPRPU) compared to the previous elections. At the same time, the main opposition party, the Estonian Centre Party (ECP) lost three seats and the Social Democrats (ESDP) as the other opposition party gained nine seats.

This indicates that all three parties conducting the austerity reforms (ERP, ESDP, and PPRPU) were the winners of the elections. This was in stark contrast to that in more than 10 other EU member states where governments have fallen or been voted out of office, and where a direct link could be drawn between the fall of the government and its austerity measures in most cases.

However, any comparison of the seats in parliament before and after the parliamentary elections in 2011 needs to account for some changes in the distribution of seats among all the four main parties in Riigikogu between 2008 and January 2011. For example, in February 2011, one month before the parliamentary elections, the ESDP not only had its own nine seats in the parliament but had also become a home to three members of the People's Union Party, who had joined the ESDP in 2010 and were by then official members of the party.

The success of the Social Democrats in the parliamentary elections in 2011, compares starkly with the results of the European Parliament elections in March 2009 when the Party witnessed quite a sizeable failure in the elections to the European Parliament by receiving only one seat instead of the three seats received five years earlier. However, the failure of the ESDP in the 2009 EP elections was more related to their loss of their main vote-magnet T. H. Ilves and the general growth in the popularity of independent candidates.

The elections of the local municipalities with their broader electorate two years earlier were more ambiguous in terms of the voters’ support for the ruling coalition. Although an immediate negative reaction of voters should be considered more likely in 2009 following the social stress caused by the austerity measures passed since 2008 and a loss for government parties should be expected because of the cyclical pattern of the electoral behaviour shown in the theoretical pattern of ‘second order’ elections, in fact the elections revealed no negative response from voters to the austerity measures in general, but rather showed the polarization between the citizens of the Republic of Estonia and citizens of a non-EU member state or stateless people residing lawfully in Estonia. To illustrate this situation, in the Estonian capital Tallinn, where only 55% of the population are Estonian citizens while most of the rest hold a permanent living permit also allowing them to vote, the clear winner of the elections was the main opposition of austerity, the ECP with 53.5% of the votes, 12.4 percentage points more than in 2005. The leading party in the coalition, the ERP, received 16.6% of the votes, and lost 4.1 percentage points of voter support from the previous elections to the local municipalities in 2005. The PPRPU, a member of the governing coalition, received 15.4% of the votes, and the ESDP received 9.8% of the votes, losing 11.1 percentage points of support from 2005. The results were completely different in Tartu, the second biggest city of Estonia, where 80% of the population are Estonian passport holders. The ERP scored here a clear victory with 30.8% of the votes and the PPRPU received a further 23.5% of the votes. The Estonian Centre Party received only 17.4% and the ESDP got 15.9% of the votes (Electoral Committee of Estonia, 2013). The results of the elections in 2009 and 2011 offer a clear example of how the support for the governing coalition differs in general between Estonian citizens and the Russian-speaking population with permanent living permits.

High migration in 2010–11 may also have reduced the negative impact of the electorate, especially among the Estonian citizens. Estonian migration trends are similar to those in Latvia and in Lithuania (despite the significant difference in quantities), where emigration also intensified in the years of crisis. At the same time, in the Estonian case it should be noted that the high migration may have reduced the negative reaction among the citizens, but not among the non-Estonian citizens, as most of those who have chosen to leave Estonia had Estonian citizenship and thus had the right to vote and ‘punish’ the government if they wanted. Accordingly, these citizens who sensed the social injustice and lack of welfare voted with their legs and chose labour migration; those ones who were able to keep their pensions, salaries, and jobs in Estonia continued to support the government's economic policy.

5. Discussion and conclusions: variables influencing and explaining voters’ behaviour

The following analytical part will discuss the main variables causing the theoretically unexpected positive reaction from voters to austerity measures implemented in Estonia, by evaluating the role of local political culture and the government's communication with the aim to conclude if the Estonian austerity model could be in case of need used in other euro area member states.

The theoretical models suggest that a positive electoral outcome from austerity reforms could be expected if: (1) the voters act according to the subjective/subordinate political model and follow the expected rules, expressing their opinion in elections and tending to support government parties, and if the country has only a limited experience of democracy; (2) the government succeeds in convincing the electorate that they were not active decision-makers or that austerity was not their free political choice from among other alternatives, but the one and only rational option left because of a financial or institutional force majeure; (3) the most active groups of voters in elections are not particularly affected by the austerity measures; (4) the negative image of the crisis is amplified bringing along a strongly negative public expectation about the crisis solutions; (5) neighbouring countries in a comparable situation have decided to take even more stressful measures; (6) there are no ideologically acceptable alternatives to the government; and (7) individuals are more focused on the public welfare rather than their own welfare, or the degree of economic voting is low.

In practice most of these conditions were met in Estonia in the parliamentary elections in 2011. First, the level of participation was high and did not indicate any ignorance on the part of the electorate. To analyse the elections according to the categories described by Hirschman (1994), the utmost priority for the Estonian governing coalition when opting for austerity was not to mobilize groups which may choose the ‘voice’ strategy in an active negative form. ‘Exit’ – especially a temporary one – was considered to be an acceptable choice by both citizens and non-citizens and ‘loyalty’ was seen as a crucial variable in maintaining the support of the civil servants, young people, and young families. Pensioners were approached with material motivation. At the same time, the strategy of keeping pensions growing to win votes did not have any visible effect. Post-election research conducted after the parliamentary elections by age group showed that aggregated support for the coalition parties was highest in the 31–35 age group at 48.6% and the 18–25 group at 39.5%, but was smallest in the 51–55 age group at 23.7% and the 61–65 group at 28.6% (ENES, 2011). Second, the austerity measures implemented in Estonia were presented as the most rational as well as the only choice for the government and the political elite in the crisis situation. Third, the ‘punishment’ effect in parliamentary elections in 2011 was also minimized by the wide ideological cleavage in the centre of the ideological axis in terms of budgetary choices, and by the successful demonizing of the opposition in the media. As a result, even if numerous citizens wanted to vote for alternatives to the governing coalition and its radical austerity measures, the only real alternative was to support the Estonian Centre Party. However, its Keynesian approach and partially non-democratic image was too extreme to be accepted as a rational option.

The readiness of the voters to vote for the alternative candidate was proved by the elections to the European Parliament in 2009, when independent candidate Indrek Tarand received 25.8% of the total votes. Fourth, the negative reaction of the electorate was also tempered by the government's successful media communications focusing mainly on the crisis in Latvia and also those in Hungary, Greece, and Portugal, while avoiding any debate on the crisis solutions implemented in Finland, Sweden, Poland, or Lithuania – countries where fiscal consolidation was achieved without austerity. Fifth, the increase in the support for the governing parties in Estonia could be related to the accession to the euro area in January 2011, which in this particular case could be described as a clear sign of the voters’ cyclical behaviour where in certain periods individuals focus on the public welfare (as membership of the euro area should enhance economic stability), but in other periods they put their own interests to the fore (personal negative experience of austerity measures, high unemployment rate, etc.). As fulfilment of the euro area accession criteria was presented as the goal of the government, a phenomenon could be observed where individuals were more focused on the public welfare rather than on their own welfare.

Loyalty of citizens as a key variable inside political culture (debated by Almond & Verba, 1989) could be argued to have played an important role in the choice between not voting, voting for the government, or voting for the opposition. The results of the elections show that the majority of voters acted as subjective members of the society by channelling their reaction into the elections but remaining supportive, while only a minority of voters distanced themselves from political participation. Citizens were more active participants in the elections and tended to support the coalition parties more (parliamentary elections; local elections in Tartu), while non-Estonians voted more for the parties opposing austerity. Young people were the main supporters of the government's policy, while pre-pensioners and pensioners were the main opponents of the government policy, including its austerity measures, even though pensioners were spared most during the budget cuts. These results are also in correlation with the most popular self-image of Estonians, who see themselves as a protestant but obedient nation (Estonian Cooperation Assembly, 2012).

In general, no single pattern of voting behaviour can be said to have been formed during the different elections analysed here. While the European Parliament elections in 2009 carried a protest message against the governing coalition and its political choices, the parliamentary elections in 2011 indicated strong support for the Prime Minister's party. The voters’ behaviour in the elections to the local municipalities was divided internally, with strong opposition to the governing parties among voters with non-Estonian citizenship, but simultaneous support from voters with Estonian citizenship. It needs to be stressed additionally, that next to cleavages caused by economic factors in Estonia, the different pattern of voters’ reactions was also caused by national identity and citizenship status.

Which government actions played a central role in minimizing the negative reaction and mobilizing supportive voters? The Estonian government, by keeping pensions constantly growing, aimed to minimize the number of individuals who were personally affected by the austerity measures. In practice pensioners continued to vote for opposition parties, which promised even higher pensions and no austerity in the future. At the same time young civil servants, who were the main victims of salary cuts caused during austerity continued to support the ruling coalition, which indicates that models of economic voting were not fulfilled in the Estonian case.

The philosophical debate on pro-Thatcherism and the opposite classical welfare state model was also used to support the government's argument that austerity is the one and only strategically acceptable solution to the crisis.

As a result the governmental coalition was successful in preventing punishment by voters, as non-citizens who preferred voting against the coalition were not able to use their power in the parliamentary elections, and Estonian citizens were either loyal to the government or silently emigrating when not satisfied.

The Estonian government's success in the political management of austerity will of course be hard to repeat. But in the case of the tradition of a balanced budget, very low governmental debt, stabilization reserves, Thatcherism among majority voters, high loyalty to the government among citizens and most important – a targeted and focused policy of national government, it has been proven to be possible to implement austerity and receive even more votes in the next elections.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Viljar Veebel is a political scientist, who defended his PhD in 2012 at the University of Tartu, focusing on the European Union pre-accession conditionality. Since 2013 he has been working at the Estonian National Defence Academy as an Associate Professor of Social Sciences. Viljar Veebel is also actively participating as a consultant in European Union-related state building projects in Moldova, Georgia, Ukraine, and the Balkan area in cooperation projects in cooperation with the Estonian Ministry of Foreign Affairs, the Estonian School of Diplomacy, the OSCE, and Swedish International Development Agency. He is also participating in active media debate on European Union enlargement, fiscal stability, and EU–Russia relations.

Liina Kulu is an economist; she completed her masters studies at the University of Tartu, where she is also continuing her doctoral studies and has been conducting both teaching and research. At the moment Liina Kulu is working as senior specialist at the Bank of Estonia. Her main research interests consist of the CEE models of transition, fiscal stability, and Eurozone financial stability.