A vast body of literature emphasizes that successful political development relies on stable and united political parties that underpin clear and responsive policymaking. By this perspective, divided government and party volatility represent pathologies that imperil political accountability and development. Indeed, studies contend that divided government – where different parties control the executive and legislative branches of government – lead to policy deadlock while party volatility – characterized by frequent party splits, solo switches, mergers, dissolutions or the establishment of new parties – confounds representation and accountability. Clearly, each on its own is seen as a considerable threat to political performance and democratic development; the confluence of the two points virtually to an inevitable political doom.

These conceptions of divided government and party volatility sound the alarm over two current political phenomena: First, the institutional arrangements of presidential and semi-presidential regimes that give rise to divided governments is the predominant constitutional choice of emergent democracies (see Cheibub and Chernykh of this issue). Second, party-system volatility can be observed not only in emergent democracies but also established ones (see Kato and Kannon; Mershon of this issue). Do the divided government and party volatility spell a spate of regime instability and collapses on the horizon or, with apologies to Mark Twain, is their depiction as pathologies to political and democratic development greatly exaggerated?

The collection of six papers here follow from a workshop on ‘Governability Across Regime Types’, convened by the School of Social and Political Studies, University of Edinburgh, on 24 October 2007 to evaluate the effects of divided government and party volatility. Our research is methodologically broad as we seek to arrive at conclusions with generalizable and comparative implications. Thus, our papers use large-N data vis-à-vis issues of democratic stability in divided government and semi-presidential systems. We also adopt small-N comparative studies with in-depth qualitative methodology that analyze the stability of minority or divided government and how party volatility influences party performance and government stability.

Importantly, our diverse methodological approaches and cases underscore several consistent findings of interest to the literature: First, we find no systematic evidence that divided governments impede legislative effectiveness or democratic development. Second, we find that party volatility also does not pose the threats to political or democratic stability as previously supposed. Third, even where the two conditions dovetail, political and democratic effectiveness are not jeopardized. Fourth, we find that, contrary to expectations, legislators and parties are not strident on policies or positions but, instead, change positions as needed to win elections. These results suggest that party volatility and divided governments are not mistakes from political or democratic development that threaten stability. Rather, they represent conscious and strategic responses to manifested problems in political and democratic development and, hence, pose as vehicles of progress towards representation and political accountability.

The series begins with two papers examining the effects of divided government on democratic and political performance. Scholars on regime types typically ask if presidential or semi-presidential systems are unstable (Linz, Reference Linz1990; Stepan and Skach, Reference Stepan and Skach1993). Often, the literature casts the internal conflicts between the legislature and the executive as destined to impasse or deadlocks in the legislature unless averted by lameduck presidents in their late tenure. The papers by Elgie and McMenamin and also Cheibub and Chernykh show otherwise: through logit and survival analyses, the two papers independently report systematic findings that semi-presidential or mixed systems are not more likely to be politically or democratically weak. Importantly, the two studies fill the need for rigorous study on a topic that has been mostly supported by anecdotal evidence. The expansiveness of the studies, in terms of countries and time periods covered, underscores that their findings of insignificant statistical relationships between divided governments and political and democratic performance are meaningful and represent substantive contributions to the literature.

Elgie and McMeamin point out that arguments cited against semi-presidential systems rest primarily on anecdotal evidence. They identify three that are commonly cited: first, a strong president can ‘personalize’ the political process to the detriment of democratic performance; second, the problem of dual legitimacy, especially if the president and prime minister are from different parties, may compromise political performance; and, third, the absence of a majority in parliament can lead one of the two executives to ignore the rule of law and undermine democratic performance. The authors proceed to test for these effects of semi-presidential systems through cross-national logit analyses of 38 countries with semi-presidential systems that are neither democratically consolidated nor autocratic. Specifically, Elgie and McMenanim evaluate the effects of three measures – power of the president, cohabitation, and divided minority government – on democratic performance. Of the three arguments used against semi-presidential systems, only one is supported statistically: Elgie and McMenamin find that semi-presidential systems with strong presidents pose challenges to democratic development. There were no substantive statistical support for the remaining two arguments; that is, divided executives are not statistically related to democratic weakness, and divided minority governments are not statistically associated with poor democratic performance. The significance of these findings, as Elgie and McMenaim point out, lies not in the rebuttal of two perceived wisdoms or the corroboration of one academic convention; rather, their findings underscore the need to pay attention to differences in the types of semi-presidentialism. Indeed, the authors suggest that in some forms, the dual nature of the executive in semi-presidentialism may enhance rather than erode democratic performance because it gives more political actors a stake in the system.

Cheibub and Chernykh similarly consider semi-presidential systems with a directly elected president and a cabinet who is responsible to the national assembly, which the authors denote as mixed systems or constitutions. The authors address four perceived weaknesses in mixed systems: first, that they are more unstable than parliamentary ones; second, they are less accountable; third, they are lacking in legislative effectiveness; and, fourth, they are less likely to survive as democracies. Using data of all democracies between 1919 and 2006 where the governments are based on assembly confidence, Cheibub and Chernykh systematically compare the performance of mixed systems against parliamentary ones. Thus, the authors evaluate two measures of government stability – tenure of the prime minister (PM) and tenure of the party of the PM – and show that the higher instability we observe in mixed systems is not due to the type of constitution but to the type of electoral system they adopt. Mixed or semi-presidential democracies overwhelmingly adopt proportional representation rules, which lead to more frequent ‘minority situations’ and more frequent coalition governments; prime ministers in the latter, in turn, tend to have shorter lives. Thus, it is the way they elect legislators, and not their constitutional framework, that make prime ministers in mixed democracies have shorter lives in office. Cheibub and Chernykh also found no evidence that mixed systems contributed to lower government accountability, which they operationalize as the probability that the head of the government will remain in office even as economic outcomes deteriorate. Instead, in both parliamentary and mixed systems, the probability that the PM remains in office is not affected significantly by economic outcomes. The authors also found no statistical difference in legislative effectiveness (defined as the proportion of executive proposals approved by the legislature in a given year) across types of government (single-party majority, single-party minority, coalition majority, coalition minority) in either mixed systems and its pure parliamentary counterparts. Finally, Cheibub and Chernykh found that the mixed systems are no more likely than parliamentary ones to transition from democracy to non-democracy. As a further test of the robustness of their results, Cheibub and Chernykh evaluated the impact of presidential constitutional powers on the performance of mixed systems. They find that constitutional powers of the president have no statistically significant effect on the turnover of the PM or the party of the PM.

The two papers, then, challenge the conventional wisdom that divided governments imperil political accountability and democratic development. Two papers on party volatility, Mershon focusing primarily on Italy and Kato and Kannon on Japan, similarly disprove its eviscerating effect on representation and accountability. Coalition research has traditionally focused on the size and formation of coalitions; however, recent trends are looking at intra-party politics and the dynamics of conflict and negotiation within the legislature. In this regard, Japan and Italy may represent cases of ‘instability out of stability’ in parliamentary systems (Laver and Kato Reference Laver and Kato2001; Laver and Giannetti Reference Laver and Giannetti2001; Mershon Reference Mershon2002; Heller and Mershon Reference Heller and Mershon2009). Importantly, the two papers by Mershon and Kato and Kannon underscore that party volatility follows from strategic decisions and lends to, rather than undermines, political performance and democratic accountability. Specifically, Mershon finds that legislators switch parties strategically to jockey for rewards – electoral advantage, offices, or policy influence – most prominent, during different stages of the legislative term. Kato and Kannon find that party volatility results from parties reconstituting or redefining their policy space in order to be more successful with forming and maintaining coalitional partnerships that are necessary for successful governance. Thus, their findings fundamentally challenge the conventional perception that party volatility is inherently destabilizing to politics and democracy.

Mershon argues that party switches represent strategic decisions by individual members of parliament (MPs) throughout the legislative term. Legislators time any switches they make in response to distinct stages of the parliamentary cycle. They also understand the essence of the constitutional design of a parliamentary system: the executive's dependence on the legislature. Thus, party switches are unlikely out of governing parties and are likely to affect the redistribution of executive offices during the legislative term. To test this argument, Mershon focuses primarily on the Italian lower house between 1996 and 2001, and supplements that analysis with data from Australia, Britain, Canada, France, 1963–1972 Italy, and Spain. For the in-depth case of 1996–2001 Italy, Mershon shows that party switching corresponds to the legislators’ strategic response to obtain different benefits over the course of the legislative term. Specifically, she classifies switching inducements as they appear in four active stages of the legislative term: initial announcement of party affiliation, assignment of legislative and executive offices, control of policymaking on the legislative agenda, and pre-electoral positioning for the next elections. She also identifies an inactive dormant stage, where switching should not appear. Mershon finds significant support in the 1996–2001 Italian lower house for the proposition that party switching is relatively frequent in each of the distinct active stages of the parliamentary term. Further, in analysis of over 2,000 monthly observations of switching behavior across the full set of six countries, she finds that most switches occur between governing parties or between opposition parties. That is, switching out of governing parties tends to be rare, which Mershon ascribes to institutional responsiveness between the legislature and executive in a parliamentary system. Moreover, Mershon finds that legislative party switching stimulates a redistribution of offices in the executive only in those parliamentary system with relatively high switching rates − that is, not semi-presidential France, not Australia and Britain, but rather Canada, 1963–1972 Italy, Spain, and above all 1996–2001 Italy.

Kato and Kannon's analysis of the organization and policy positions of Japan's political parties since 1993 also show that legislators’ switching is strategic and related to institutional performance. Also significant is their finding that party volatility is the result of parties redefining policy positions in order to enhance legislative competitive advantage and improve their odds of forming the government since the 1994 electoral reforms. That is, Kato and Kannon find that legislators are not the only ones with an incentive for party switching as a result of the electoral reforms. Instead, parties also realign or reorganize in pursuit of a policy space that leads to successful coalition formation. This is demonstrated in their analysis of expert survey data regarding the ideological positions of political parties in Japan. The authors find that the policy positions that predict the parties’ ideological positions change from year to year. Thus, for instance, Kato and Kannon find that the parties staked out less strident positions on finance and economics than previously held in order to demonstrate fiscal responsibility and financial accountability. But the authors consider their most persuasive finding to be on foreign policy and national identity issues, which typically predict a left–right ideological stance: they find that parties now hold different policy orientations on defense and national identity issues throughout the period from 1996 to 2005. Their results also show that the major opposition party, the Democratic Party of Japan (DPJ), rather than the governing Liberal Democratic Party (LDP), benefits most from the increasing volatility. In fact, the DPJ uses its ideological diversity among legislative members as leverage to attract more switchers and increase its size. These results provide a lens for understanding the electoral outcome of the recent elections to the House of Councilors and presage the political fortune of the governing Liberal Democratic Party (LDP) in the upcoming elections to the House of Representatives. By these papers, the party volatility witnessed in Italy and Japan represent a critical component of party changes in stable democracies that does not jeopardize political or democratic performance.

The papers in this series thus far have looked at political and democratic performance of divided governments broadly in large-N studies, or considered such performances in the context of party volatility in industrialized democracies. An interesting and important consideration is to examine if the confluence of party volatility and divided government poses significantly more risks to political or democratic performance, especially in emergent democracies. Curiously, this consideration is largely neglected in extant studies: scholars have typically focused on minority governments under parliamentary systems (Bogdanor, Reference Bogdanor1983; De Winter et al., Reference De Winter2002; Strøm, Reference Strøm1990; Strøm and Müller, Reference Strøm and Müller2001) or divided government under presidential systems, with a particular focus on Latin America (Colomer and Negretto, Reference Colomer and Negretto2005; Elgie, Reference Elgie2001; Mainwaring and Shugart, Reference Mainwaring and Shugart1997). Two final papers in this series fill that gap: the papers by Yap and Kim examine political or democratic performance under coalition politics and divided governments in two non-Western European systems, Taiwan and South Korea. Although the two papers differ methodologically – Yap's paper is based on statistical analyses of Taiwan, while Kim derives her findings through qualitative research and elite interviews during the Kim Dae-jung administration in South Korea – they arrive at the same conclusion: the confluence of party volatility and divided government do not jeopardize political or democratic performance.

Yap challenges the assumption of parties as unitary and stable purveyors of ideological policy-framers and -makers and offers an alternative explanation – ambition – that contravenes dooms-day predictions of the effects of divided government and volatile parties on performance. She points out that the ambition that motivates office-seekers in new democracies also contributes to divided government and party volatility as office-seekers split, merge, or create new vehicles to attain office. However, because ambitious office-seekers cannot afford to slacken on political or democratic performance, it does not lead to political and democratic instability, To test her argument, Yap examines budgetary spending in Taiwan since the 1960s to 2004, to find out if ambition leads the executive to emphasize an executive agenda at the expense of a legislative spending, and if opposition fragmentation in the legislature increases as a result of such executive domination in party development. Her analyses show persistent spending patterns since the 1960s: spending in favor of an executive agenda that also sidelines or penalizes a legislative welfare-spending agenda. Such spending patterns may make sense under the one-party system prior to democratization, where the Kuomintang (KMT) oversaw spending in favor of the executive office and maintained party discipline and loyalty by dispensing offices and related favors. However, the same patterns foster the development of an executive party at the expense of party development for legislative members, even as transparency and electoral controls since democratization eliminated the party favors necessary to instill loyalty and discipline. As a result, Yap finds more parties composing the legislature as legislators split, switch, or create alternative routes to office in reaction to executive-dominated spending in Taiwan. Anecdotal evidence in Yap's paper complements the statistical analyses and completes her findings. First, the informal pan-green and pan-blue coalitions in Taiwan, comprising the coalescence of parties that split or fracture from the major parties, emphasize the effort by candidates and politicians to ensure governability and effectiveness. Second, the recent effort and success in restructuring electoral rules point to sustained efforts to pursue political and democratic performance.

Kim zeroes in on the internal dynamics of Korean political parties since democratization to challenge the prevailing argument that divided government and party volality are the sources of political deadlocks in the country. Instead, Kim emphasizes the salience of party organization as a source of political deadlock. Through an in-depth analysis of the Kim Dae-jung administration, which is characterized by a high number of party fissions, fusions, mergers, and breakups, Kim contends that ungovernability under the Kim Dae-jung administration does not derive from government size because the minority status is not a sufficient condition for the lack of governability; further, instability persisted even with the attainment of majority status. Kim's field research and semi-structured interviews of the Korean political elite and study of the minority coalition government in South Korea under the Kim Dae-Jung administration show that deadlock persisted under two conditions. First, when the party leadership fails to provide mechanisms for compromise with the opposition party and a linkage with ordinary citizens, it contributes to deadlock in the legislature. Second, failure to consolidate party organization within the ruling coalition parties and among factions within the party increases conflict within the party and between ruling and opposition parties. Thus, Kim contends that party organization and internal mechanism need more attention as determinants of governability and she shows how intra-party politics impacts governability to establish that it is not size that matters: indeed, the size of the ruling coalition does not secure legislative effectiveness. By discussing party organization in terms of leadership, factionalism, funding, and party–citizen linkages, Kim's paper provides the qualitative equivalent of statistical controls of the Kim Dae-jung's administration to rule out party volatility and divided government as the sources of political and democratic weakness. Instead, she shows that understanding how the party system and parties are run and function sheds significant light on the instability of political and democratic performance.

In sum, the papers here provide statistical and qualitative evidence to show that party volatility and divided governments do not represent problems in political or democratic development that threaten stability. Rather, by considering the internal structures and institutional mechanisms that affect governability and democratic stability, the research points to these developments as conscious and strategic responses to existing political and democratic conflicts. That is, contrary to expectations, legislators and parties are not strident on policies or positions but, instead, change positions as needed to win elections. Importantly, because these are strategic responses for achieving representative offices, political accountability and government performance are not threatened in the process. Indeed, our findings suggest that divided government and party volatility may provide the vehicles of progress towards increased representation, political accountability, and institutional performance. We look forward to further tests and applications of the findings.