Turkish Voters and Losers’ Consent1

This paper tests four hypotheses of "losers' consent"—that is, the extent to which evaluations of democratic governance are shared by citizens who voted for the Justice and Development Party (AKP), the winner at the 2007 Turkish elections, and those who voted for the opposition parties, the Republican People's Party (CHP) and the Nationalist Action Party (MHP). It uses survey data about governance evaluations from the 2007 Turkish Election Study. The evidence rejects the idea that losers withhold consent, creating polarized pluralism. It gives some support to a winner's effect. It finds a substantial degree of similarity in views of governance across party lines. Differences between voters for the two losing parties are sometimes greater than differences with the AKP. Regardless of dissatisfaction with some features of governance, there is an overwhelming consensus rejecting a change of regime to military or Shari'a rule.

Electoral competition encourages opposition parties to describe the governing party's actions in extremely negative ways. Election results show that negative judgments of government are endorsed by some voters but rejected by others. The 2007 Turkish election illustrates the process of parties competing on terms that divide the electorate. The Republican People's Party (Cumhuriyet Halk Partisi, CHP) attacked the governing Justice and Development Party (Adalet ve Kalkınma Partisi, AKP) as threatening the secular foundations of Atatürk's secular republic. The National Action Party (Milliyetçi Hareket Partisi, MHP) attacked the governing party for its alleged failure to defend Turkey's national identity. Nonetheless, the AKP won reelection with a big majority of seats in the Grand National Assembly. However, it fell short of winning half the vote: more Turks voted for losing parties than for the party that won control of government.

Democratic elections involve a paradox. Differences of opinion between parties and partisans are the essence of democratic elections, but consensus about democratic institutions creating losers as well as winners is also necessary. As C. J. Anderson, A. Blais, S. Bowler, T. Donovan, and O. Listhaug emphasize: "The consent of losers is fundamental to the maintenance of any political system."[2] To achieve this goal:

Losers' support requires the recognition of the legitimacy of a procedure that has produced an outcome deemed to be undesirable. In the end the viability of electoral democracy depends on its ability to secure the support of a substantial proportion of individuals who are displeased with the outcome of an election.[3]

If this goal is not achieved, then there is a risk that losers may withhold consent from governance.

The dynamics of governance are in the hands of an election winner, but as William H. Riker has argued: "The dynamics of politics is in the hands of the losers. It is they who decide when and how and whether to fight on."[4] This point is particularly salient in Turkey, for the 2007 election was the second successive defeat suffered by the CHP, a party whose leaders had thought of themselves, rather than AKP, as a "natural party of government" and, even more importantly, as defending the secular state against a governing party committed to making major changes altering what the CHP sees as the fundamental constitution of the state.

One set of theories of electoral competition postulates that losers should be stimulated by defeat to adapt their policies and personalities in order to win the next contest. The opposition may alter its behavior somewhat and hope that the electoral pendulum will bring it victory due to the governing party becoming overconfident, suffering a cyclical downturn in the economy and unanticipated negative consequences of its policies. Insofar as opposition leaders take this view, they will support the system of governance on the assumption that they will be in charge after the next election. An alternative possibility is that competition can persist in the form of polarized pluralism, in which competing parties seek to mobilize votes by discrediting the other side as a threat to basic political values.[5] Insofar as this is the case, electoral competition can encourage negative attitudes and losers withholding consent from a system that their opponents control.

In long‐established democracies, differences between voters for winning and losing parties, whatever their numerical size, are unlikely to lead to political instability due to loss of support for the political system, since the system tends to be stable. However, many political systems, including Turkey's,[6] lack the stability of long‐established systems. In new democracies, voters are aware that the country has been governed by undemocratic means in the past and there is not a long enough period of time to create acceptance for the regime on positive grounds or because there is no alternative. Anderson et al. point to "inexperienced losers as a particularly weak link in the chain of stable democratic governance."[7]

The object of this essay is to analyze the extent to which evaluations of democratic governance are shared by citizens who voted for the winner, the AKP, in the 2007 elections and those who voted for the principal opposition parties, the CHP and the MHP. This will be done by utilizing evidence from the 2007 Turkish Election Study.[8]

The next section reviews alternative theories of electoral competition relevant to Turkey today and sets out four hypotheses about whether and to what extent partisans may differ in their views of democratic governance. The two sections that follow examine how voters for different Turkish parties evaluate the political system in terms of democracy and in terms of governance. The conclusion summarizes the evidence, which rejects polarized pluralism while giving some support to theories of a null effect and of a winner's effect. It also finds that, regardless of dissatisfaction with some features of governance, there is an overwhelming consensus rejecting a change of regime to military or Shari'a rule.

Whereas democracy is about what people want, governance is about the interaction between the governors and the governed. In Max Weber's succinct formulation: "Power is in the administration of everyday things."[9] As Joseph Schumpeter has emphasized, elections not only offer citizens an opportunity to give their verdict on who should govern but also on how they have been treated by their governors since the previous election.[10] Elections that do not affect what government does in the years between competitive elections are incompletely democratic.[11] The linkage of electoral behavior with governance thus contrasts with theories of political culture, which assume that government will act in accordance with the values of its citizenry, a highly contingent assumption in a world with more than 50 countries in which elections are held but that are far from complete democracies. It also differs from viewing governance solely in terms of the inputs of citizen participation[12] without regard to the outputs of government.[13]

The European Union has developed criteria for evaluating democratic governance as part of the process of considering whether applicant countries for membership meet its standards. At its Copenhagen Council in 1993 it identified five criteria for evaluating applicants.[14] The criteria are broad: a government is required to be a modern as well as a democratic state. Competitive elections are a necessary condition for democracy, but not sufficient for a system of democratic government. Adherence to the rule of law is equally fundamental. Whereas elections can produce an overnight change in who governs, adherence to the rule of law is critical to positive actions by the bureaucrats who are the main providers of services in an era of big government; it is also critical in preventing the abuse of human rights and minority rights by those in office.

Thinking of political systems in terms of both bureaucratic rule of law and competitive elections is consistent with a long tradition of European theorists going back to Max Weber[15] and before, who have integrated politics, economics, and the rule of law in the analysis of governance. Prime Minister Tayyip Erdoğan has endorsed the EU's Copenhagen criteria as "sensible and helpful" targets that could just as easily be described as "Ankara criteria."[16] However, agreement about the criteria for making an assessment is not the same as agreement about how the country is governed.

Four hypotheses offer different predictions about the extent to which voters for winning and losing parties agree in their assessment of a country's system of democratic governance and in the consequences of differences of opinion for the stability of the political system.

This hypothesis is most strongly supported when a large majority of winners and of losers holds the same view. Such circumstances not only reject theories about partisan differences but also identify a cross‐party consensus. Weaker support for this hypothesis occurs when the majorities endorsing a view are only slightly larger than half, thus rejecting a consensus interpretation, or when there are substantial interparty differences in the size of the majority.

The simplest theory accounting for differences in evaluations between voters for winning and losing parties is that people prefer winning to losing, and that disappointment with an election outcome will spill over into negative views about how the country is governed. The justification for these views may be based on principle, on specific shortcomings of performance, or on emotions.[17]

In a civic culture, a negative evaluation of the performance of the government of the day may be offset, at least in part, by positive pride and satisfaction with the performance of government, whichever party is in office. In such circumstances, differences between winners and losers will be differences of degree, not kind, with majorities in both groups positive about the system. In Easton's[18] terms, specific grievances that losing voters have about how winners govern will not spill over into diffuse disaffection with the political system. This assumes that voters are sufficiently sophisticated politically to distinguish between electoral institutions and election outcomes. In terms of education and longevity of experience with democracy, Turkish voters are hardly a match for the Anglo‐American citizens that constitute Easton's point of reference.

At some point, differences between groups of partisans can widen into a gap and deepen into a political cleavage. If these differences are about which party or leader is more competent, this will affect voting behavior but not assessments of democratic governance. However, in national contexts in which differences between parties are about strongly held values and interests, whether religious versus secular or propertied versus poor, this can create "polarized pluralism" in which partisan differences spill over to affect assessments of democratic governance. Turkish parties have reflected sharp differences on religion, economics, and constitutional issues.[19]

Regime change not only requires the rejection of democratic elections but also its replacement by one or another form of undemocratic rule, whether a civilian or military dictatorship. The practical choice facing election losers is whether to change in hopes of winning the next election or whether to organize to seize power without an election. The former alternative is easier and much less risky than trying to stage a revolt or a coup.[20] Moreover, as Winston Churchill argued, "Democracy is the worst form of government except all those other forms that have been tried from time to time."[21] Thus, a negatively assessed system of governance can still be preferred as a lesser evil to alternatives, a hypothesis consistent with Turkey's almost uninterrupted maintenance of competitive elections for almost half a century, notwithstanding major shortcomings of governance.

The circumstances of the 2007 Turkish election give special relevance to the fourth hypothesis. The winner, the AKP, had been in power since 2002; therefore, it was already identified with the system of governance and the opposition parties had had five years of experiencing how its actions differed from what they wanted. The election was called early because the polarization between the AKP and the chief opposition party, the CHP, intensified with the CHP preventing the Grand National Assembly from electing its choice as president in spring 2007. This led to millions protesting in the streets against the AKP government, challenges in the Constitutional Court, and veiled threats from military personages against the government. In such a context, the possibility of a change to an undemocratic regime, whether military or Islamic, could not be dismissed as groundless. The AKP denied allegations that it was seeking to act unconstitutionally and wanted to stress the country's booming economy since it took office in 2002. However, its opponents sought to discredit its intentions if reelected.

This essay uses data from the 2007 Turkish Election Survey (TES) to show how Turks evaluate governance according to the above standards of democratic governance. The pre‐election wave of this panel study included a battery of questions developed in the New Europe Barometer to evaluate the extent of democratic governance in Central and East European post‐Communist countries.[22] Given differences between the CHP and the MHP and both parties winning seats in the Grand National Assembly, the tables that follow report data for both opposition parties as well as for the victorious AKP.

The critical test of a democratic election is not who wins but whether it was fair. While elections are ubiquitous in the world today, the free and unfair elections of post‐Soviet states are a reminder that fairness cannot be taken for granted. Likewise, the outcome of Iraq's elections shows that if losers do not consent to defeat, then violence and civil war can follow. Among all respondents in the pre‐election TES wave, 72 percent said they regarded the election as fair. In the post‐election survey, 81 percent did so. The higher percentage that viewed the election as fair after the result was known is likely to reflect the fact that the Turkish proportional representation system's abnormally high barrier of 10 percent for winning seats can produce very disproportional results. In 2002, 45 percent of the vote was wasted, being cast for parties or candidates that did not win seats; in 2007, the proportion of wasted votes fell to 13 percent. The election also produced a clear‐cut winner, the AKP, which won 46 percent of the vote, and clear losers, the CHP with 21 percent and the MHP with 14 percent of the vote.

Although there was substantial agreement about the fairness of the election, the competition hypothesis is supported (see Figure 1). Even though a majority of voters for each of the parties thought the election was fair, there is a significant difference between winners and losers. Before the election, there was an eight percentage point gap between AKP and MHP voters in an expectation of a fair election, and a 25 percent gap between AKP and CHP voters. The post‐election survey found the gap had widened to 21 percentage points between AKP and MHP voters and 35 points between the winning party and CHP voters. AKP supporters were overwhelmingly of the opinion that the election was fair.

Graph: Figure 1 Partisan differences about fairness of election. Source: 2007 Turkish Election Survey. Pre‐election fieldwork: June 23–July 16, 2007. Post‐election fieldwork: August. Don't knows excluded: 2 percent pre‐election, 5 percent post‐election.

Since democratic governance is about much more than elections, Turks were also asked whether they thought the political system was more like a complete democracy or a complete dictatorship. Since respondents were offered an 11 point scale, this allowed people to assess the political system as falling exactly between the two polar extremes. Altogether, 28 percent of Turks placed the system at this halfway point between democracy and dictatorship. Among those who came down clearly on one side or the other, those seeing the system as more democratic were twice as numerous, 48 percent, as those seeing it more like a dictatorship, 24 percent. However, this less than enthusiastic assessment does not imply a rejection of democracy as an ideal. When asked how they would like to be governed, 89 percent endorsed democracy, 5 percent were for dictatorship, and 6 percent were neutral.

The competition hypothesis is again supported, for there is an 18 percentage point difference between AKP and MHP voters seeing Turkish government as democratic, and a 26 point difference between voters for the winning party and the CHP (see Figure 2). Moreover, a larger minority of voters for the losing parties regard the government as dictatorial rather than democratic. However, this does not mean that there is polarization between winners and losers, for the median group of voters for both losing parties see the political system as halfway between democracy and dictatorship, and more MHP voters see the system as democratic rather than a dictatorship. Insofar as there are signs of polarization, it is within the ranks of CHP voters, who are almost evenly divided as to whether the government is democratic or a dictatorship.

Graph: Figure 2 Partisan differences about whether Turkey is democratic. Source: 2007 Turkish Election Survey. Fieldwork: June 23–July 16, 2007. Respondents giving a neutral answer not reported in the above figure: 24 percent of AKP voters, 27 percent MHP voters, and 28 percent CHP voters.

New regimes are much more vulnerable to regime change than are established regimes. Since the founding of the state, the Turkish Republic has had three constitutions; the current Constitution was prepared in 1982 after military intervention. The Atatürk revolution made the state rather than a particular constitutional order the inviolate protector of governance, and the army has not hesitated to force an elected government from office in efforts to promote what it deems good governance, as distinct from popular government.[23] The present Constitution was only adopted after the army deposed a government presiding over intense civil disorder in 1980.

The army receives by far the highest rating when Turks are asked about their trust in eight different political institutions. On a 0–10 scale, an absolute majority, 51 percent, give the army the highest possible rating: a total of 85 percent rate it positively, 8 percent are neutral, and only 7 percent negative. However, an equally great majority does not want the army to govern Turkey today. Moreover, there is a consensus on this among voters for all three parties (Figure 3). Thus, all three hypotheses about partisan differences are rejected. The consensus against military rule, notwithstanding substantial signs of dissatisfaction with the contemporary political system, supports the Churchill hypothesis: voters for all parties prefer the political system as it is, if only as a lesser evil.

Graph: Figure 3 No disagreement in rejecting undemocratic alternatives. Source: 2007 Turkish Election Survey. Fieldwork: June 23–July 16, 2007. Don't knows excluded: 3 percent army rule, 6 percent Shari'a.

The Turkish Constitution explicitly endorses secular authority. However, the majority of Turks are practicing Muslims.[24] In the TES survey, 71 percent said they attend mosque services at least weekly, and only 7 percent said they go to a mosque less than once a year or not at all. Thus, in addition to asking about army rule as an alternative to the current political system, a question was also asked about whether people would prefer to be governed by Shari'a, a strict Islamic code of behavior. Over all, a very large majority reject this type of regime: only 12 percent favor it, compared to 83 percent who reject it, and a small portion expressing no opinion.[25] Moreover, the difference among parties is only in the size of the majority rejecting Shari'a rule. Rejection is virtually unanimous among CHP and MHP voters, and even though the AKP appeals to committed Muslims, four‐fifths of AKP voters are also against application of Shari'a rule.

The Churchill hypothesis is doubly supported. Asked about secular or religious alternatives to the current regime, three‐quarters of Turks reject both alternatives. The small minorities that endorse undemocratic alternatives are divided almost evenly between the 13 percent endorsing army rule but not Shari'a and the 10 percent endorsing the opposite view, with 2 percent endorsing both alternatives. A substantial proportion of supporters of all three parties—84 percent of CHP voters, 77 percent of MHP voters, and 71 percent of AKP voters—reject both military and clerical domination of the government.

The rule of law is doubly central to the governance of a modern, democratic state. On the one hand, it introduces predictability into governance, imposing obligations on public officials to behave consistently rather than in a clientelistic or sultanistic manner. Adherence to the rule of law is also critical for implementing European Union obligations, such as Single Europe Market and the thousands of pages of regulations and directives in its acquis communitaire. The rule of law can be maintained without free elections, as in nineteenth century Imperial Germany or contemporary Singapore, and competitive elections can be held without the government being bound by the rule of law, as in many successor states to the Soviet Union.[26]

Corruption is the commonest way in which public officials break laws, benefiting financially. In a clientelistic political system they may provide favors in order to build a base of political support. When asked what proportion of public officials are corrupt, respondents divide into two almost equal groups (Table 1). However, perceptions of corruptions are often based on hearsay information rather than experience.[27] In the 2007 Turkish Election Survey, only 5 percent said they had paid a bribe to a public official in the past year.

Table 1. Popular Evaluation of the Rule of Law and Bureaucracy (%)

Since citizens spend far more time receiving public services such as education, refuse collection, and healthcare than voting, fairness in delivering public services to citizens is central to governance. The fair delivery of services does not mean that everyone gets what they want but that the refusal to provide benefits is due to the impersonal application of rules regulating entitlements rather than to corruption or clientelistic favoritism.[28] Turks are equally divided about the behavior of public officials: half think people like themselves are usually treated fairly, and half do not (see Table 1).

There is limited support for the competition hypothesis: those who voted for the AKP are more likely to expect fair treatment and less likely to see most public officials as corrupt than are supporters of opposition parties. However, the differences are of degree, not kind. AKP voters are 3 percent less likely to see officials as corrupt than MHP voters and 13 percent less than CHP voters. Moreover, any partisan bias in the bureaucracy is limited: the difference between AKP and MHP voters in the expectation of fair treatment is eight percentage points and between AKP and CHP supporters 17 percent. Insofar as there is polarization in the application of the rule of law, it is due to divisions of opinion within all three groups of partisans, resulting in the percent of those with positive and with negative views of governance being almost equal in number.

The rule of law is also important in institutionalizing the rights of citizens. Only if the rule of law effectively constrains governors can citizens feel secure in their liberty and in political rights. In the Republic of Turkey, citizens have not been subject to totalitarian abuses of freedom and rights, as happened in Central and Eastern Europe under Communist rule. However, the tradition of the Republic is different from that of Anglo‐American democracies, where individual rights and freedom from the state have priority.[29] For example, Article 1 of the Constitution states that "no protection shall be afforded to thoughts or opinions contrary to Turkish national interests" and Article 301 declares it an offense to make statements "insulting Turkishness, the Republic, the Parliament and state institutions."

When asked whether everybody can say what they think, 39 percent say that this is the case, while a majority of all three parties feel that there are difficulties in doing so. There is a limited degree of support for the electoral competition hypothesis. While there is no significant difference between AKP and MHP voters, there is a 14 percentage point difference between AKP and CHP voters, with four in seven of the supporters of the government feeling less than fully free in speaking out, as compared to five in seven of CHP voters (see Table 2).

Table 2. Popular Evaluation of Political Rights (%)

In the absence of constraints from the law, a government need not show respect for human rights, even if they are embodied in the national Constitution. Almost two‐thirds of Turks believe that the government has some or a lot of respect for individual rights. Here again there is some support for the competition hypothesis. Three‐quarters of AKP voters see respect for human rights, compared to three‐fifths of MHP voters and one‐half of CHP voters. Nonetheless, there is no polarization, for a majority of supporters of all three parties believe the government respects individual rights.

The issue of rights can also be related to the collective status of ethnic minorities. In Turkey the largest such group is Kurdish, albeit the confounding of universalistic rights and the Turkish nation in the Constitution means that Kurds do not have a legal status as an ethnic minority. Nor does the Kurdistan Workers Party (Partiya Karkerēn Kurdistan, PKK) recognize this status, since for more than two decades it has carried out a terrorist campaign against the Turkish state in pursuit of its goal of an independent Kurdistan uniting territories in Turkey, northern Iraq, and Iran. Since the vote for explicitly Kurdish parties has tended to be half or less than the estimated proportion of Kurds in the electorate, partisan attitudes toward minorities cannot be analyzed in the terms of views of voters for a Kurdish party.

Voters for all three parties were divided in their opinion of the treatment of minorities. The effects of competition show to a limited degree but differently than hypothesized. The nationalist MHP is most likely to see minorities treated fairly and CHP voters second most likely, while a majority of AKP voters (which include more Kurds than the opposition parties) think that there is not fair treatment of minorities.

An alternative approach to evaluating the treatment of minorities is to compare the way in which ethnic Kurds assess their situation, whatever party they vote for, with the evaluation of ethnic Turks. In the Turkish Election Survey, ethnic Kurds are identified as individuals who spoke Kurdish to their parents in childhood. Giving priority to the primordial significance of family and language classifies 9 percent of respondents as Kurds. If Kurdish respondents agree among themselves and disagree with the majority of Turks, this would demonstrate ethnic divisions about democratic governance.

A comparison of the replies of Turks and ethnic Kurds to questions in the figures above, reveals no statistically significant differences between the two groups for a majority of indicators. Big majorities in both groups think that elections are fair, and differences of opinion are similar about the democratic character of the state. The same is true in the assessment of corruption in government. On two key measures of special salience to interethnic relations—receiving fair treatment by public officials and respect for human rights—there is no significant difference between Turks and Kurds.

There are limited differences of degree but not kind on three measures of governance. The majority of Kurds who feel there are restrictions on freedom of speech is 12 percentage points larger than the Turkish majority. Whereas Turks divide almost evenly about whether national minorities are a threat to the country's security, more than three‐quarters of ethnic Kurds do not think this is the case. The Kurdish minority (28 percent) endorsing Shari'a law is larger than that of Turks (11 percent). Nonetheless, almost three‐quarters of Kurds agree with almost nine‐tenths of Turks in rejecting Islamic rules governing society.

Altogether, the evaluation of governance by AKP, MHP, and CHP supporters is most consistent with the consensus or null hypothesis. Whatever their party, on the critical issues of endorsing army or Shari'a rule, overwhelming majorities of voters are of the same mind: they reject these forms of undemocratic governance. There is also a negative majority expressing dissatisfaction with freedom to say what one thinks. Since many Turks are negative about some aspects of the current political system, the rejection of undemocratic alternatives also supports the Churchill hypothesis. Dissatisfaction with the status quo is accepted as a lesser evil than alternative forms of governance, such as army rule or theocratic rule. Even when majorities in all three parties make the same assessment of governance, they may not do so to the same degree. When this occurs, the competition hypothesis—voters for winning parties are more positive about governance than are losers—is also supported in response to questions about the fairness of elections and respect for human rights.

The multiparty system of Turkey shows that juxtaposing governing and opposition parties can be misleading. Even though opposition parties have a common goal, turning out the government of the day, their voters do not think alike on many assessments of how the country is governed. Interparty differences tend to show a consistent pattern: AKP and CHP partisans are furthest apart in their views, with MHP voters in between. About half the time the profile of MHP voters is closer to that of AKP supporters than to supporters of the other opposition party, the CHP.

The Turkish Election Survey shows that the polarization between government and opposition parties in the Grand National Assembly is not reflected in the mass electorate. Instead of polarization between AKP voters who think Turkey is well governed and opposition supporters who think it is being governed badly or in defiance of Republican principles, divisions are more likely to occur within parties than between them. For example, not only are Turks as a whole divided about whether government is corrupt or unfair in dealing with citizens, but the same is also true of AKP, MHP, and CHP supporters (see Table 1). Broad endorsement of the political system is sustained by the absence of a coalition of losers against the winners.[30]

As a crosscheck on the limited influence of partisanship on attitudes to democratic governance, a variety of social, economic, and political influences—plus dummy variables for AKP or CHP support—were regressed on each of the indicators of governance reported above. While there were substantial differences between indicators in the amount of variance that could be explained, the dummy variables for party support were either not significant—for example, in accounting for the extent to which the political system was seen as democratic—or of only secondary importance—for example, in assessing the fairness of the election. Consistent with the second hypothesis, when party preferences did have some influence, it was favoring the AKP, and this boosted endorsement of the political system.

When questions were asked about expectations for the future, replies were optimistic and showed increasing consensus. Whereas there was disagreement between a majority of AKP and CHP voters about whether the country today is more a democracy or dictatorship (Figure 2), in five years' time a clear majority of supporters for all three parties expect the political system to be more democratic. This includes 59 percent of CHP votes, 67 percent of MHP voters, and 77 percent of AKP partisans. After discounting those who gave a neutral answer, almost three times as many AKP voters are optimistic as pessimistic about the future of democracy in Turkey, and four times as many MHP voters are.[31]

In new democracies, a change in regime cannot be dismissed as entirely wishful thinking. In Turkey, the Grand National Assembly has been shut down by military action within the memory of the median Turkish voter, and the pro‐Islamist Welfare Party (Refah Partisi, RP) was forced to disband in 1998. Hence, people were asked whether they thought the Assembly could be closed down and parties abolished in the next five years. Overwhelming majorities of voters think this will not happen (see Figure 4). Moreover, there is no statistically significant difference between the 91 percent of AKP voters, 90 percent of CHP voters, and 89 percent of MHP voters holding this view. Even though only 29 percent of Turks trust members of the Assembly, there is a consensus in favor of the Churchill hypothesis. With all its faults, the country's system of democratic governance ought to remain in place, for only 9 percent say they would approve if the Grand National Assembly was shut down and parties abolished, and there is no statistically significant difference between voters for different parties. Thus, in Turkey as in every new member state of the European Union, the current political system, whatever its faults, is now "the only game in town."[32]

Graph: Figure 4 No expectation of an end to party government. Source: 2007 Turkish Election Survey. Fieldwork: June 23–July 16, 2007. Don't knows excluded: 4 percent.