Beyond Democracy and Autocracy: System Responsiveness in Global Perspective. Michael S. Latner Cal Poly, San Luis Obispo

Transcription

1 Beyond Democracy and Autocracy: System Responsiveness in Global Perspective Michael S. Latner Cal Poly, San Luis Obispo Prepared for the 2014 World Congress of the International Political Science Association Montreal, Canada Draft Copy

2 Introduction From 2007 to 2011, a series of food price escalations and global recession created the greatest economic and political instability that the world has experienced since WWII. Commonly referred to as the Great Recession, real World GDP per capita declined in 2009, and the G-20 zone suffered a consecutive quarterly decline from Q through Q (IMF, 2009). A combination of drought, increased fuel prices, demand for resource-intense foods, financial speculation and falling stockpiles all contributed to global price spikes in basic goods such as rice, wheat and corn in Riots and outbreaks of violence spread across the globe, hitting developing countries especially hard. Although food prices plunged with the onset of recession, they peaked again by early As the world continues to recover, policy makers have an obligation to better understand how the design of political institutions shaped response capacity under such an exogenous shock. In the following research, I treat economic data from the global crisis as a quasi-natural experiment, and try to answer the question: Are some types of regime better able to absorb and respond to the threat of the global financial crisis? Background Political economists and comparative constitutionalists continue to battle over the relative importance and role that governing institutions play in the economic health and prosperity of political systems. While few would seriously discount the claim that institutions matter, as the last 50 years of development in places like North and South Korea, or East and West Germany demonstrate, there is genuinely little agreement over the question of how variation in regime type impacts economic and developmental trajectories. As Pippa Norris documents in her recent work on the subject (Norris, 2012), many scholars argue that democratic institutions are largely a product rather than a cause of economic development and prosperity (Inglehart, 2005), while others argue that the development of democratic (Diamond, 1999) or bureaucratic (Fukuyama, 2014; Rothstein, 2011) governance structures are a prerequisite to the creation of a responsive, healthy political system. Lack of agreement stems, in part, from conceptual ambiguity over the meaning and operationalization of terms like democracy and autocracy. In this research, I take a novel approach to the question of institutional design, integrating several concepts from the development and comparative democracy literature in order to ground the concept institutional design in a general theoretical framework. A major drawback of previous approaches to this question is the underspecified nature of institutional measures. Democracy measures often incorporate both behavioral and 2

3 institutional components that blur causal pathways (Lijphart, 1984; Powell, 2000) and are limited to analyzing only regimes that fit a pre-specified minimal definition of democracy. By contrast, broader measures of effective democracy or democracy/autocracy classifications typically truncate and cluster democracy on the high end of an index that limits analytic capacity ( Freedom in the World, 2014; Marshall, Jaggers, & Gurr, T., 2011). Conceptualizing governing institutions within the framework of a general policy design and implementation system, a system designed to regulate and coordinate human behavior (Easton, 1979)(Crozier, 2010)(Stoelhorst & Richerson, 2013), suggests several mechanisms required to sustain economic prosperity. If we start from the axiom that the benefits of life are disproportionately available to cooperating creatures (Axelrod & Hamilton, 1981), a system that reliably generates cooperative benefits is advantageous. This policy-making system should effectively access environmental and individual information and convert it into cooperative benefits. Immediately we face the basic trade off involved in facilitating decision inputs and converting them into policy outputs, a specification of the more general explore-exploit dilemma (Axelrod & Cohen, 2001). The dilemma arises from the inability to optimize both the exploration of information, where more information results in better decisions, and the exploitation, or conversion, of acquired information, necessary for generating benefits. Without the institutionalization of rules that solve both objectives, benefits cannot be generated. The exploration or policy-input aspects of a system can usefully be conceptualized as a question of permissiveness to illustrate the problem. A system that minimizes permissiveness allocates few resources to exploration in order to maximize action and decision-making autonomy. In political systems, pure dictatorship theoretically minimizes permissiveness, while optimizing decision-making authority in a single person. On the other end of the spectrum, a system that maximizes permissiveness would explore information from every member of a political community, as a pure democracy. A stable political system then requires some degree of institutionalized permissiveness in order to function reliably. Second, policy-making systems require the institutionalization of conversion or decisionmaking rules to exploit useful information. Just as individuals face the fundamental question of how much certainty, or internal consensus, is necessary to make decisions in our daily lives, in the aggregate the question is how much agreement, or concurrence, is necessary to enact policy. As with permissiveness, pure dictatorship reflects the design principle of minimizing the level of agreement required to act and implement policy, while the unanimity rule reflects the opposite. With permissiveness and concurrence, we have the elementary components of a policy-making system, but the institutionalization of explore-exploit rules is only the first set of challenges. Questions about system performance involve at least three distinct but interrelated challenges: Are the rules stable enough to function? Do they function efficiently? And are the policy outcomes fair, that is, what is the pay off for the cooperative effort? 3

4 The advantages of large-scale cooperation are easily undermined by within-group competition, because policies that are optimal from a systemic perspective will not always align with what is optimal for specific individuals. A multitude of scholars argue that rule of law, bureaucratic capacity, or government effectiveness is a key component of successful performance (Fukuyama, 2014)(Evans & Rauch, 1999). While the specific mechanisms linking structure to performance vary among scholars, the underlying claim is that performance requires the maintenance of order, security, and the bureaucratic capacity to implement effective policies. Effective government is thought to enhance performance by providing a stable regulatory environment, and an efficient, rational division of labor and resources. Further, bureaucratic capacity can facilitate fairness through impartiality and professionalism. The advantages of large-scale cooperation are also easily undermined by the exploitation of leaders. Indeed, it has been argued that the emergence of representative democracy is an adaptive response to the parasitism associated with authoritarian rule (Stearns, 2012; Fukuyama, 2014). Along these lines, the facilitation of public communication and freedom of the press may also contribute to system stability and efficiency through the generation of public information and a reduction in coordination costs associated with collective action (). Finally, numerous scholars have argued that while it may be as much a product of, as a cause of development, strong civil society liberal democratic culture are strongly associated with regime performance over the long run (). If that is so, democratic culture may also facilitate short-term performance. Research Design In order to assess how variations in these regime traits shaped responsiveness to the global economic crisis, I modeled political and economic performance of political systems from as a function of institutional design prior to 2007, along with pre-recession measures of economic performance and a set of other controls. I rely primarily on data from the Quality of Government dataset provided by the Quality of Government Institute at University of Gothenberg (Teorell et al., 2014). With the exception of the GDELT data summarized below, all datasets are included as part of the Quality of Government compilation. Economic performance indicators were constructed from two primary sets of data, so as to capture a variety of performance traits. Economic efficiency and productivity were modeled from World Bank estimates of Gross Domestic Product in current USD, GDP per capita based on purchasing power parity (PPP), and GDP per capita growth based on annual local currency ( World Development Indicators Data, 2013). Measures of economic fairness and inequality were based on gini coefficients of disposable household income, gross income, and inequality reduction through redistribution compiled by Frederick Solt for the World Income Inequality Database (Solt, 2009). 4

5 For each of these six measures, pre- and post-2007 estimates of performance were calculated in three forms: the average absolute value over three years ( , and ) was used as a starting point for comparative analysis and for pre-recession economic control variables; next, the slopes over each three-year period were calculated in order to capture the direction (positive or negative) and degree of change in annual economic performance; finally, the means were subtracted from means to estimate average change in each of these measures over the pre- and post-shock periods. Each of the primary political measures was constructed using a variety of sources. Democratic political culture: The Economist s Economic Intelligence Unit provides a cultural measure of the extent to which there is a societal consensus supporting democratic principles as part of their widely used Democracy Index (Economist Intelligence Unit, 2014). Based on expert surveys and World Values Survey data where available, the indicator reflects assessment of a society s support for democracy, belief in the performance of democracy compared to alternatives, and cultural susceptibility to authoritarian or military rule. Estimates from the 2007 index from 165 countries are used. Values range from 0-10 with 10 equal to the highest level of democratic culture. Government effectiveness: Part of the World Bank s Worldwide Governance Indicators, effectiveness is measured from combined expert responses regarding the quality of public service provision, the quality of the bureaucracy, the competence of civil servants, the independence of civil service from political pressures, and the credibility of the government s commitment to policies. Average pre-recession values range from -2.3 to 2.3 for 191 countries. Press freedom: Freedom House s annual measures of freedom in the world include this score of freedom of the press, which combines laws and regulations, political pressures and controls, and economic influences. Data scores were reversed with higher values from indicating greater press freedom ( Freedom in the World, 2014) Permissiveness: Institutional permissiveness is measured as a function of the capacity for political organization and association, the size (capacity) of national representative assembly, the magnitude of electoral districts (and capacity for electoral expression), and institutional incentives to cultivate personal votes. Beginning with organizational and expressive capacity, the 2006 Freedom House measure surely smuggles in some normative/cultural components of organization, but in the absence of data on suffrage and formal associational restrictions, this indicator serves as a sufficient base estimate for freedom of assembly, demonstrations and open public discussion. Values of assembly permissiveness (A) are transformed to range from 1-78 as part of a weighting procedure. Phillip Keefer s updated Database of Political Institutions provides two additional variables for this measure (De Melo et al., 2001). Assembly permissiveness is multiplied by the size (number of seats) in the national assembly (S) for each country (lower house if bicameral) and the mean district magnitude (M), or average number of seats contested 5

6 per electoral district. These two components measure overall electoral permissiveness and the capacity for political demands to be accurately reflected in political competition and representation. These data were also cross-checked with estimates from Johnson and Wallack (see below). Finally, the potential for hyper-permissiveness through personalistic vote cultivation was measured using the Electoral Systems and the Personal Vote variable for incentives (I) to cultivate personal votes in the populous tier of an electoral system (Johnson & Wallack, 2012). This comes from an update and expansion of Shugart and Carey s original classification of incentives based on single-member districts, electoral formula, candidate list control, and related factors to create a rank ordering from 1-13 (Carey and Shugart, 1995). Thus, the empirical estimate of Permissiveness is calculated with the equation logp = log(asmi). Base values for non-electoral regimes and related missing data were imputed, where possible, as described in the supplementary materials appendix (see below). The curvilinear fitted.67log(p+1) was used for much of the multivariate analysis. Concurrence and proximity to Majority Rule: Concurrence, like Permissiveness, is constructed out of a number of distinct components. Starting with the Polity IV data on executive constraints (Marshall et al., 2011), regimes receive a score ranging from 1-7, where a score of one reflects unlimited authority by the executive, while seven reflects decision rules where legislatures or accountability groups share effective authority equal or greater than the executive in most areas. Transitional or failed states are coded with a value of 24, representing maximum anarchy, or the extreme opposite of unlimited executive authority. Additional coding rules were applied to states differently depending on their initial categorization. Data from the Institutions and Elections Project (Regan & Clark, 2014) were used to construct measures of separations of power and supermajoritarianism, with an additional point added to political systems scoring at least a five on the Polity IV measure for each of the following: strong bicameralism, separately elected executive, and executive veto power. Similarly, the Political Constraints Index (Henisz, 2004) is composed of the number of independent branches of government with veto power over policy change, and dichotomous measures of bicameralism, federalism, and judicial independence are incorporated to reflect deviation from simple majority rule. These scores were subtracted from failed states (executive constraints = 24) to indicate greater institutional structure and political constraint, and added to established regimes to reflect their supermajoritarian structure. Finally, data from the Comparative Constitutions Project (Elkins, Ginsburg, & Melton, 2014) were used to identify levels of judicial independence and strength of review, a final additional constraint on simple majority rule. For much of the multivariate analysis, the scores are folded to maximize the value of simple majority rule. Size of government: An additional institutional control and proximate constraint on policy flexibility, this variable reflects high values (up to 10) for large government 6

7 consumption, large transfer sector and many government enterprises in addition to a high marginal tax rate (Gwartney, Lawson, & Easterly, 2006). Economic globalization: A final control to account for the extent to which a regime is embedded in, and susceptible to, fluctuations in the global economy (Dreher, 2006). A set of proximate political stability variables was estimated using data from the Database of Political Institutions (DPI) (De Melo et al., 2001) and the Global Database of Events, Language and Tone ( Protest: The GDELT archives were queried for frequency of citizen protests from , and counts were used to create a slope to measure the increase/decrease in protests across time for all events, as well as the relative (per capita) level of protest in Fractionalization: DPI measures of electoral (effective number of parties) and government (herfindahl index) fractionalization for the and periods were used to calculate expansion or contraction of party systems after The DPI also provides time-series estimates of governing coalitions (right, center, left) during the time periods under consideration, which allows me to assess the impact of institutional design on changes in party system control. A preliminary analysis examines the strength of association between institutional design features and overall levels of economic performance prior to the global financial crisis. After identifying the static linkage between political institutions and economic stability and productivity, I turn to political stability, assessing changes in party systems and government in the post-2007 period. Economic performance and ability to withstand the shock of global crisis are tested in the next set of models, followed by an analysis of how regime variation shaped economic inequality and distributive fairness in the midst of widespread economic instability. Results Figure 1 displays bivariate correlations between the various cultural and institutional traits used in the analysis. Both of the new institutional design measures are modestly correlated with one another. As expected, the more permissive a system is, the less likely it is to be governed with excessively low levels of concurrence. This is true for both the raw and the fitted or transformed scores used in multivariate analysis. System permissiveness is positively, but modestly, correlated with other regime traits commonly associated with democracy. As expected, the strongest correlation is between freedom of the press and the fitted values of permissiveness (R=.528), but the two are clearly distinct regime traits. Examples of permissive systems that exhibit restrained press freedom include pre-2007 Russia, Ukraine, and Thailand. 7

8 (Figure 1 here) Concurrence is only loosely linked to democratic political culture (R=.185) or government effectiveness (R=.187) because of the concentration of authoritarian regimes at the low end, and failed regimes at the high end. The link between proximity to simple majority rule (C fitted) and other regime traits associated with democracy is stronger, especially in the case of freedom of the press (R=.569). The strongest bivariate correlations are found between press freedom, democratic political culture and government effectiveness. While this analysis does not examine the extent to which press freedom is a requirement for, or result of, government effectiveness, or the possible feedback loops that link government effectiveness to democratic political culture, clearly these components are mutually reinforcing. Together, these five traits constitute the primary architecture of political systems, with permissiveness and concurrence operating as distinct institutional features. Stability In Table 2, pre-recession economic performance is modeled as a function of these system traits, in addition to the size of government and economic globalization. Without taking into account any other natural or historic features (other than population size), these variables effectively model current GDP and per capita GDP. Government effectiveness is the strongest political indicator of economic productivity in the first two models, both before (above) and after (below) Greater integration into global markets was also a reliable indicator of a regime s GDP during both time periods. (Table 2) Neither associational permissiveness nor simple majoritarian decision-making institutions contributed to economic performance on their own, however. Of course, some of the least permissive regimes in the world derive considerable wealth from natural resources (Humphreys, Sachs, & Stiglitz, 2007), and strategic authoritarian regimes have considerable capacity to make autocracy work (Besley & Kudamatsu, 2007). Indeed, press freedom is weakly, but negatively, associated with higher GDP in both time periods. GDP growth was not modeled well before or after 2007, but institutional linkages to growth were more pronounced before In the period just prior to the beginning of the global crisis, GDP growth was higher in regimes that exhibited less democratic political culture and press freedom. These findings contrast more comprehensive accounts of the impact democratic traits on long-term growth (Norris, 2012), but support the hypothesis that authoritarian regimes might be more effective at manipulating shortterm growth, especially under the pressures of globalization, also positively associated with pre-recession growth. The only statistically significant indicator of growth during the period is government effectiveness, and it is negative. Globalization also drops out as indicator of 8

9 growth, suggesting that growth patterns during this tumultuous period more sensitive to local contingencies. So far, of the institutional traits examined, government effectiveness appears to play a stronger supporting role than permissiveness or majoritarian decision rules in stabilizing economic performance. While short-term growth prior to the financial crisis was linked to greater permissiveness, post-2007 growth was actually higher under governments with less effective bureaucracies, in contrast to expectations. Next, I examine the impact of institutional design on political stability, defined as the level of protest exhibited after 2007, regime collapse, and changes in coalitional politics in regimes with enough permissiveness to allow them. Figure 2 provides an overview of the landscape of political stability using permissiveness (P) and concurrence (C) as guideposts. (Figure 2) A cluster of regimes with low permissiveness and concurrence collapsed in the midst of the Arab Spring of , resulting in either new constitutional design (Tunisia, Yemen, Lybia), continued transitional struggles (Egypt) or civil war (Syria). These regimes are all marked in black, but it is notable that the two regimes with some permissiveness already built into their systems, Egypt and Syria, are actually having the most difficulty reestablishing stability. Moreover, when compared to other high protest regimes in the same space, it appears that in these and other countries with low permissiveness (Jordan, Oman, Morrocco), political reforms were more difficult not because of permissiveness, but where moderate permissiveness was combined with serious retardation of press freedoms, as in Syria and Egypt. That combination of traits appears to have been especially unstable. Elsewhere, as in Iceland, high levels of protest resulted in considerable political reform, but the process was less destabilizing as a result of the participatory architecture employed. Indeed, as Figure 1 and Table 3 show, high levels of protest were still taking place across all levels of permissiveness by 2011, though extreme levels were more common in low concurrence regimes. There was also greater variance in the protest score, standardized by population and ranging from 0-1, within these low concurrence regimes. Several, including Saudi Arabia and Qatar, experienced minor protests and were able to alleviate demands through minor reforms (and subsidies). (Table 3) In addition to protest levels, Table 3 displays the frequency of change in party system types (Left or Right) among regimes from the pre- and post-periods. Overall, stability in party system ideology was widespread, that is, the dominant legislative party, whether right, center, or left, stayed in power throughout both time periods. At the same time, institutional constraints clearly matter. Significantly higher levels of change in both directions were observed in systems with greater permissiveness, and in systems with 9

10 higher levels of concurrent decision-making. More governments transitioned to the left than to the right overall, but that pattern was most pronounced in regime with at least majoritarian concurrence rules, where about one in five (22%) regimes transitioned from right or centrist to more leftist governments after In Table 4, the slopes of protest from were regressed against the political indicators and economic controls (including pre-recession lagged variables for GDP and GDP per capits, not shown). Surprisingly, proximity to simple majority rule is not associated with lower protest, as might be expected from Figure 1 or Table 3. Instead, it is democratic culture, or the lack thereof, that turns out to be the strongest indicator of where protests increased the most. While protest rose in the low concurrence regimes of the Arab Spring, it was also rising in countries like Spain and the U.S. However, whether in tightly controlled authoritarian regimes or failed states, the highest levels of protest occurred in countries with little democratic cultural inheritance. Moreover, the sparks of protest were also more pronounced in countries more deeply embedded in the global economy. (Table 4) Table 4 also displays the results of two tests employed on a subset of regimes, those where permissiveness allows for significant (though not necessarily free or fair) electoral competition. Among these regimes, there was a global contraction in average party system fractionalization, or the number of parties competing effectively for office. Similarly, there was a slight contraction in the fractionalization of the average governing coalition in these same regimes. In other words, rather than emerging and expanding competition for political power, there is evidence that, on the whole, electoral regimes consolidated power during the global crisis. Modeling fractionalization as a function of the performance indicators reveals a number of interesting patterns. First, government effectiveness played no role in shaping increased electoral or government fractionalization. Whereas effectiveness plays a greater role in policy outputs, press freedoms, permissiveness and proximity to majority rule had a greater impact on these expressive inputs. Both permissiveness and majority rule were associated with reduced party fractionalization, in part because more permissive systems start out with greater fractionalization, but also because some of the most permissive systems (Italy, Ukraine, Slovakia) exhibited substantial consolidation after Of course, the Italian and other systems have since reversed course, which points to an important result not revealed in regression estimates. Both permissiveness and majority rule exhibit considerable heteroskedasticity; at the high ends, some party systems expanded considerably, while others consolidated. In particular, and as hypothesized, permissive, majoritarian systems are noisier with respect to coalition stability, whereas more constrained systems exhibited greater consistency in the consolidation of political power. In addition, while regimes with higher levels of democratic culture fragmented electoral power, that same trait contributed to consolidation within governments. This is possible 10

11 wherever minor parties achieve greater electoral success but fail to win seats or become members of a governing coalition. Globalized economies, including India and Mexico, were also more likely to experience party fractionalization along with government consolidation. Efficiency and Productivity Next, changes in economic performance, or the capacity of systems to respond to global crisis, are modeled as a function of the performance measures. Table 5 displays the results, and reveals that short-term economic fluctuations, even as a result of a global financial crisis, are not particularly sensitive to differences in institutional design. For each of the three economic performance variables (GDP, GDP per capita and GDP growth) change from pre- to post-period is modeled in two ways. First, the slope over the period is regressed against the indicator variables, controlling for prior GDP and the slope of the prior years. Next, the average value from the prior three years is subtracted from the average of the later three years to create a measure of change in the average. (Table 5) As Table 5 shows, strength of democratic political culture did not directly contribute to a regime s ability to increase wealth or rate of growth over this period. Similarly, freedom of the press was not a significant predictor in any but one of the models (average change in GDP per capita), and the linkage was weak and negative. The institutional indicators that are the primary focus of analysis also lack any clear impact on economic performance during the crisis. Permissiveness does not show any statistically significant association with these economic variables, and majoritarian systems appear to have done slightly less well at sustaining overall GDP compared to highly supermajoritarian or authoritarian regimes. Indeed, several regimes with moderate permissiveness, including the U.S. and non-electoral regimes like China, sustained impressive economic performance relative to more permissive democracies in Europe and South America. Overall, there was greater variation in performance among more permissive regimes, a point to which we will return. Even government effectiveness only reached conventional levels of statistical significance as an indicator of average change in GDP, though the sign of the coefficient was in the hypothesized direction in all but one model. Similarly, size of government was a successful indicator for GDP per capita, but it contributed to an increase, contrary to expectations. Finally, economic globalization was positively associated with increased GDP and GDP per capita, but not with increasing growth or slopes of change. Overall, the models were better fitted to explain average change than differences in slopes. Fairness 11

12 In the last set of models, the political indicators are used to account for changes in levels of economic inequality and redistribution that resulted after the financial crisis struck. The first set of columns in Table 6 displays the impact of political institutions and culture on changes in economic inequality based on gross income. None of the primary variables of interest show any influence on shaping changes in this gini coefficient, whether it is measured in terms of average change across the two time periods, or the slope of annual change in gross income inequality from However, in the next set of models, regime Permissiveness emerges as a modest contributor to reduced levels of disposable household income, in line with expectations. Conversely, economic globalization appears to have contributed to increased disposable income inequality since 2007, but not increases in the average level of inequality from to (Table 6) In the final set of models, government effectiveness reemerges as an indicator of enhanced reduction of inequality through redistribution (the difference between gross and disposable income inequality). While there is no impact on the direction of inequality reduction since 2007, the importance of government capacity for commitment is indicated by a positive association with government effectiveness. By contrast, the size of government in 2006 is associated with decreasing reduction in inequality after 2007, both in terms of average levels of redistribution and in the direction of policy since What is not clear is whether this results from overstretched bureaucracies being incapable of committing to enhanced redistribution, or whether the crisis itself reduced differences between gross and disposable incomes. Discussion Having explored the impact of several distinct components of governance on regime stability, efficiency and productivity, and distributive fairness, it is now possible to pull together the findings and reassess the question of whether some types of regimes have outperformed others during the global financial crisis. Clearly, the design of national institutions for policy-making is a crucial component of long-term regime performance. But in the years immediately following a financial crisis, it would appear that suffering is widespread across various regime types. Moreover, short-term fluctuations are difficult to account for without taking the analysis deeper to examine the budgetary policies actually adopted and implemented by regime authorities. Nevertheless, this analysis does point to a number of important conclusions for policy makers concerned with institutional design and system responsiveness. First, these findings support the claim that permissive representation and the institutionalization of political equality through majority rule do not, in themselves, stabilize economic prosperity better than alternative configurations (Norris, 2012). 12

13 Rather, grounding impartiality through effective government appears to more directly impact the capacity for economic prosperity, creating conditions for productive coordination and regulation of economic behavior. On the other hand, government effectiveness did not secure absolute levels of short-term growth prior to or after the Great Recession. In fact, more permissive systems enjoyed higher pre-crisis growth while effectiveness had no impact, and regimes with more effective governments appear to have also been more locked into the economic networks responsible for the collapse. While the direct impact of institutional design in representation and policy-making on economic stability is mixed and weak, political institutions shape the demands that emerge from crisis as well as a regime s capacity to respond. The wave of regime collapse and upheaval that followed the global food crisis and Great Recession were concentrated among the least permissive, least congruent regimes in the world. Even though protest was widespread in the years following the crisis, regime permissiveness and proximity to simple majority rule did not amplify the fragmentation or decomposition of party systems or governing coalitions. Instead, they appear to have facilitated some of the global contraction in party systems. Overall, permissiveness and concurrence contribute to regime stability by facilitating transitions in electoral and governing coalitions. Beyond generalities, it is important to note how political components shape responsiveness as a whole; the regimes that collapsed in the Arab Spring appear to have combined a particularly volatile combination of moderately low permissiveness and effective governance with severe repression of political communication and freedom of the press. Press freedoms, like permissiveness and concurrence, are weakly and possibly negatively related to overall economic performance during the crisis, but the experience of the Arab Spring suggest that effective but exclusive governing institutions are susceptible to failure over the long term. Moreover, a partial explanation for the lack of linear results in the multivariate analysis of economic performance lies in the greater variation of economic outcomes in more permissive regimes and concurrent regimes. For example, an examination of average change in per capita GDP reveals that at the lowest levels of permissiveness, change in any direction was less likely, but highly permissive regimes included regimes like the Netherlands and Italy that were caught in the center of the European crisis, as well as regimes like Brazil, Russia and Indonesia, which weathered the storm relatively well. The policy relevance of representative institutions was more clearly seen in the analysis of economic inequality. Permissive regimes appear better able to constrain inequality and placing the burden of economic disruption onto the least powerful. Additional analysis (not shown) among only globalized economies suggests that simple majority rule regimes also reduced redistributive outcomes after 2007 compared to authoritarian and supermajoritarin regimes, further supporting the claim that rather than facilitating extraction, simple majority rule provides less political cover for established constituencies, which may in turn contribute to long-term fairness and regime stability. 13

14 Effective government and contingent economic factors may be the short-term drivers of economic performance, but overall survival of regimes depends on the mutually reinforcing contributions of political equality through effective governance, freedom of public communication, permissive representation, and political equality in policy design and implementation. 14

15 References Axelrod, R., & Cohen, M. D. (2001). Harnessing Complexity (Reprint edition.). New York: Basic Books. Axelrod, R., & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211(4489), doi: /science Besley, T., & Kudamatsu, M. (2007). Making Autocracy Work (STICERD - Development Economics Papers - From 2008 this series has been superseded by Economic Organisation and Public Policy Discussion Paper No. 48). Suntory and Toyota International Centres for Economics and Related Disciplines, LSE. Retrieved from Crozier, M. P. (2010). Rethinking Systems: Configurations of Politics and Policy in Contemporary Governance. Administration & Society, 42(5), doi: / De Melo, M., Denizer, C., Gelb, A., Tenev, S., Alderman, H., Collier, P., Van de Walle, D. (2001). World Bank Economic Review 15 (1), Jan S.l.: s.n.]. Retrieved from genta.com/content/wb/2151 Diamond, L. (1999). Developing Democracy: Toward Consolidation. Baltimore: Johns Hopkins University Press. Dreher, A. (2006). Does globalization affect growth? Evidence from a new index of globalization. Applied Economics, 38(10), doi: / Easton, D. (1979). A Systems Analysis of Political Life. Univ of Chicago Pr (Tx). Economist Intelligence Unit. (2014). The Economist. Retrieved July 16, 2014, from Elkins, Z., Ginsburg, T., & Melton, J. (2014). Comparative Constitutions Project. Retrieved from Evans, P., & Rauch, J. E. (1999). Bureaucracy and Growth: A Cross-National Analysis of the Effects of Weberian State Structures on Economic Growth. American Sociological Review, 64(5), 748. doi: / Freedom in the World. (2014). Retrieved July 16, 2014, from Fukuyama, F. (2014). Political Order and Political Decay: From the Industrial Revolution to the Globalization of Democracy. New York: Farrar, Straus and Giroux. Gwartney, J., Lawson, R., & Easterly, W. (2006). Economic Freedom of the World: 2006 Annual Report. Retrieved from Henisz, W. J. (2004). Political Institutions and Policy Volatility. Economics & Politics, 16(1), doi: /j x Humphreys, M., Sachs, J., & Stiglitz, J. (2007). Escaping the Resource Curse. New York: Columbia University Press. IMF. (2009). World Economic Outlook: Crisis and Recovery. Retrieved from 15

16 Inglehart, R. (2005). Modernization, Cultural Change, and Democracy: The Human Development Sequence. Cambridge, UK ; New York: Cambridge University Press. John M. Carey, M. S. S. (1995). Incentives to Cultivate a Personal Vote: A Rank Ordering of Electoral Systems. Electoral Studies, 14(4). doi: / (94) Johnson, J., & Wallack, J. (2012). ELECTORAL SYSTEMS AND THE PERSONAL VOTE. Lijphart, A. (1984). Democracies: Patterns Of Majoritarian And Consensus Government In Twenty-One Countries. New Haven: Yale University Press. Marshall, M. G., Jaggers, K., & Gurr, T. (2011). Polity IV Project: Political Regime Characteristics and Transitions, Retrieved from Norris, P. (2012). Making Democratic Governance Work: How Regimes Shape Prosperity, Welfare, and Peace (1 edition.). Cambridge: Cambridge University Press. Powell, G. Bingham Jr. (2000). Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven: Yale University Press. Regan, P., & Clark, D. (2014). The Institutions and Elections Project. Retrieved from Rothstein, B. (2011). The Quality of Government: Corruption, Social Trust, and Inequality in International Perspective. Chicago ; London: University Of Chicago Press. Solt, F. (2009). Standardizing the World Income Inequality Database*. Social Science Quarterly, 90(2), doi: /j x Stoelhorst, J. W., & Richerson, P. J. (2013). A naturalistic theory of economic organization. Journal of Economic Behavior & Organization, 90(S), S45 S56. Teorell, J., Charron, N., Dahlberg, S., Holmberg, S., Rothstein, B., Sundin, P., & Svensson, R. (2014). The Quality of Government Dataset, version 20Dec13. University of Gothenburg: The Quality of Government Institute. Retrieved from World Development Indicators Data. (2013, October 16). Retrieved July 16, 2014, from 16

17 Figure 1 Political System Components 17

18 Table 2 Regime traits and overall economic performance Pre- recession gdp Pre- recession gdppc Pre- recession growth Democratic culture * Govt effectiveness 0.656*** 0.498*** 0.01 Press freedom * *** Permissiveness *** Majority rule Size of government Economic globalization 0.389*** 0.262*** 0.355** n, adj. R2 125, , ,.15 growth gdp gdppc Democratic culture Govt effectiveness 0.649*** 0.473*** ** Press freedom ** Permissiveness Majority rule Size of government Economic globalization 0.413*** 0.299*** n, adj. R2 125, , ,.22 18

19 Figure 2 System Permissiveness, Concurrence and Political Stability 19

20 Table 3 Political Stability GDELT citizen protest 2011 Regime change left Regime change right Permissiveness Lower Higher Lower Higher Lower Higher Mean Std Dev (0.39) (0.48) (0.29) (0.40) (0.21) (0.31) n Concurrence Lower Higher Lower Higher Lower Higher Mean Std Dev (0.99) (0.30) (0.30) (0.42) (0.25) (0.33) n

21 Table 4 Institutional Linkages to Political Stability Protest Party Frac Govt Frac (Constant) Democratic culture -.630** 0.522** -.420* Govt effectiveness Press freedom ** Permissiveness ** Majority rule ** Size of government ** Economic globalization.419* 0.566*** -.505** n, adj. R2 59,.08 60,.30 60,.26 21

22 Table 5 Post Economic Performance Δ avg gdppc growth Δ avg growth gdp Δ avg gdp gdppc (Constant) Economist dem pol culture press freedom 06 (FH) ** Permissiveness Majority rule * govt effectiveness ** size of gov *** economic globalization ** *** poplog 0.57*** ** 0.344* 0.435* 0.659*** gdppc1log * * *** gdpg1log *** 0.177* *** Lagged variable *** ** *** 119, , , , , ,.47 22

23 Table 6 Post Economic Equality gross gini gross gini avg domestic gini dom. avg. redist. redist avg Economist dem pol culture press freedom 06 (FH) Permissiveness * ** Majority rule govt effectiveness * size of gov ** ** economic globalization * poplog gdppc1log gdpg1log Lagged variable *** 0.295** 0.821*** *** 56,.04 59,.58 56,.16 59,.68 50,.13 52,.61 23