Do party systems matter? Funding agencies think so. In fact, they spend millions of dollars annually to create and cultivate democratic party systems in developing countries. They want competitive party systems with stable factions that avoid fragmentation. These party system traits—competitiveness, stability, and lack of fragmentation—are important in the world of party aid.


Thomas Carothers describes these funding efforts in Confronting the Weakest Link: Aiding Political Parties in New Democracies. International aid to parties blossomed in the 1970s, when German foundations aided democratic parties in Southern Europe and Latin America. In the mid-1980s, the United States created the National Endowment for Democracy, which funded the International Republican Institute (IRI) and the National Democratic Institute for International Affairs (NDI) to support party development in countries across the globe. Since then, foundations in other Western European countries developed and expanded their own programs of party aid.

International agencies, such as the Organization of American States, the Organization for Security and Cooperation in Europe, and the United Nations joined in funding party development in the context of democratic assistance. The UN Democratic Governance Group’s Handbook of Working with Political Parties set forth the rationale, saying that “political parties are an essential part of the apparatus of governance.” It cited many contributions of political parties: aggregating interests, mobilizing the electorate, shaping public policies, holding politicians accountable, and fostering future leaders.


Citizens across the world, however, do not share the UN’s positive view of political parties. When the World Values Survey asked people in 55 countries “how much confidence” they had in their parties, almost 75 percent on average said either “not much” or “very little.” Only in China, Vietnam, and Malaysia did a majority say they had “a great deal” or “quite a lot” of confidence in their parties.



Notwithstanding the public’s negative views, most scholars agree with the UN Democratic Governance Group and regard parties as necessary to democratic government and value their contributions to governance. Assuming that political parties contribute to the quality of country governance, international organizations and non-governmental organizations have poured millions of dollars into party development under the framework of democratic assistance. Carothers offers a “very rough estimate” of US$200 million spent in 2005 on total worldwide party aid. By 2011, the AP reported that the IRI and NDI alone spent over US$100 million to support the democratic movement in Egypt. Agencies providing the funds assume that their spending has positive effects, but there is little measurable evidence.

Case studies in different countries show mixed results. African and Asian leaders and scholars have accused Western donors of pressing multiparty politics on skeptical publics that think parties represent corrupt elites, while ruling elites fear that party reforms threaten their hold on power. Carothers described these national concerns and problematic results anecdotally in Confronting the Weakest Link. He offers an “up-to-date analytic treatment of party aid,” but he does not demonstrate that aiding political parties helps country governance in any measurable way.

Relying on a comprehensive cross-national survey of political parties and country governance in 212 countries, we addressed the underlying assumption of funding party aid in developing countries. In Party Systems and Country Governance, written with Korean scholar Jin-Young Kwak, we ask, “Does the quality of the party system affect the quality of governance?” We sought to explain variance in the World Bank’s six 2007 Worldwide Governance Indicators for 212 countries by traits of their party systems.

Using data on parliamentary representation of political parties, we found party system traits that significantly improved the quality of country governance. Countries with political parties have better governance than countries without parties, and those with competitive and stable party systems have better governance than those with less competitive and less stable systems.

Measuring Governance

“Governance” is a loose term associated with a messy concept, so it means different things to different people. We define governance simply as the extent to which a state delivers to its citizens the desired benefits of government, at acceptable costs. Government benefits may reflect specific values—such as high literacy, good roads, clean water, sanitation—or abstract meta-values, such as Rule of Law, Government Effectiveness, Control of Corruption, Regulatory Quality, Voice and Accountability, and Political Stability. In fact, the Worldwide Governance




Indicators (WGI) targeted precisely those six meta-values. We used the 2007 WGI data for 212 countries, which included all 192 members of the United Nations, some non-member nations (such as Taiwan), and some entities (such as Guam and Hong Kong) not normally regarded as independent nations. The WGI measures were all significantly intercorrelated, most above r = 0.70. This article focuses on one indicator, the Rule of Law. According to the WGI’s web site, “Rule of Law captures perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence.” It stands at the core of what many people regard as good governance. Relying on thirty-five sources from research organizations around the world, the WGI project scored 211 countries on Rule of Law (RL)—failing to do so only for Niue, an island nation in the South Pacific.Screen Shot 2015-01-19 at 10.21.57 PM

The histogram in Figure 1 illustrates the distribution of the WGI scores for RL, with selected countries identified to aid interpretation. Switzerland enjoyed the highest score and Somalia suffered the lowest. The United States also ranked relatively high, while South Korea was substantially above the mean of 0.0, and China and Russia somewhat below it.

The scores were normed to have a mean of 0 and a standard deviation of 1. Commonly called z-scores, they tell—in standard deviation units—where each country stands in relationship to all other countries.

Figure 1 helps us appreciate the variability of governance across countries. How much of this variance is due to their party systems? Governance of varying quality occurs in both democratic and autocratic governments—under multiple political parties, single parties, and no parties. We attempted to explain variations in Rule of Law over all 211 countries by variations in the countries’ party systems. In statistical terms, RL became the dependent variable, while aspects of the party systems were the independent variables.

Other Effects on Country Governance

In social research, no explanatory model includes all possible causes for the dependent variable. A complete explanation of country governance cannot be limited to the effects of party systems. Many other factors—certainly including intelligent, honest, courageous governmental leadership—affect the quality of governance. Studying the quality of government leadership was beyond our capabilities. We could not include that critical factor in our explanatory model, leaving it knowingly incomplete.

But we could and did control for two other factors that theory and data showed to explain country governance. One was country size ,and the other country wealth. Early Greek and modern European philosophers believed that small countries, but not large ones, could maintain a democratic form of government. The effect of country wealth on country governance is even more obvious than that of country size: delivering the benefits of government costs money, so poor countries are inherently disadvantaged.


Variables for country size and wealth are included in our model as control variables: the larger the country (in either population or area), the more difficult to govern; and the poorer the country, the more difficult to govern. The statistical method of regression analysis adjusts for the effects of country size and wealth while assessing the effects of the party system variables.

To measure country size, we tried both population and land areas—expressed as logarithms to deal with the few very large countries that dwarfed the many small ones. To measure country wealth, we used the logarithm of GDP per capita—adjusting for the few very wealthy countries far richer than the very many poor ones. Variations in country size (land area log did better than population log) alone explained 16 percent of the variation in Rule of Law. Variation in GDP per capita (log) explained 61 percent of the variation. Entering both country size and wealth into the analysis explained 66 percent of the variance in the WGI measures of Rule of Law.

Effects of Party Systems



Space available for this article does not allow thorough description of our data on party systems. Two points must suffice: (1) We collected data only on the percentage of seats held by the three largest parties in the legislative body or parliament and the number of parties seated; and (2) the seat data represented two points in time: after a stimulus election prior to 2007 and after a referent election adjacent to the stimulus election. The stimulus election captured the party system that was positioned to affect governance in 2007 (measuring competitiveness), while seat differences between the stimulus and referent elections measured the party system’s stability. The number of parties seated measured fragmentation.

The data in Table 1 on the status of parliamentary parties in 212 countries were derived mostly from the 2006 CIA World Factbook (available on the Internet), and they apply to unicameral parliaments or to the lower chambers of bicameral parliaments. The table cross-classifies countries by two criteria: do the deputies represent parties and were deputies popularly elected?

Column 1 shows that 185 of the parliaments in 2006 seated deputies by publicly identified political parties. Only 152 countries popularly elected all parliamentary seats. In another 28 most seats were elected, but some were indirectly elected or appointed, and in one country fewer than half were directly elected. Only 181 chose at least some of their deputies through popular elections—using the phrase generously to mean direct selection by voters, regardless of the quality of the process. Four countries did not select deputies through popular elections, yet seated them by parties.



Column 2 classifies 9 countries with “shadowy” parties (unofficial or underground) by which deputies were selected. Seat data was obtained for only four (identified in boldface). Column 3 shows 16 parliaments without party deputies, half of which (mostly small island nations) were elected through nonpartisan elections. Two nations in 2006 (Nepal and Myanmar) had no parliament or legislative council.

Given that country wealth and size together explain two-thirds of the variation in country ratings on Rule of Law, we should not expect variations in party systems to explain much more.Screen Shot 2015-01-19 at 10.22.13 PM

Data Analysis

We tested five theoretical propositions concerning the effects of party systems on Rule of Law as a key indicator of governance. The first two propositions should be separated from the other three, for they apply to a small subset of the countries.

1. Without parties and elections, the Rule of Law is lower.

2. Without parties but with elections, the Rule of Law




is higher.
Only 26 countries had no parliamentary parties. Of

these, 16 (e.g., Swaziland, Togo, Myanmar, Somalia) also had no elections, and two (Myanmar and Nepal) had no parliament. The other eight were mostly island nations (e.g.. Tuvalu, American Samoa, Palau, Micronesia) that held nonpartisan elections. Controlling for country size and wealth, the regression analysis showed that these two groups of countries differed significantly compared with all 211 countries scored for Rule of Law. Having no parties and no elections had a significant negative effect on RL, while having no parties but nonpartisan elections had a significant positive effect on RL.

The analysis supported Propositions 1 and 2. The absence of both parties and elections depressed RL, but RL was enhanced if parliamentary deputies were popularly elected. At least in small island nations, popularly elected deputies need not represent political parties in order to promote the quality of governance.

The next three propositions apply to most of the 212 countries that had parliamentary parties, regardless of whether the deputies were popularly elected.

3. The more competitive the party system, the greater the Rule of Law.

4. The more stable the party system, the greater the Rule of Law.

5. The more fragmented the party system, the lower the Rule of Law.
After trying out several measures of party system

competition, volatility, and fragmentation, we settled on single measures for each. To measure party system competition, we used the strength of the largest opposition party, measured by the percentage of parliamentary seats held by the second largest party after the stimulus election. Controlling for country size and wealth, as well as the absence of political parties and nonpartisan elections, we found a significant positive effect for party system competition. The data supported Proposition 3. More competitive party systems had higher RL scores.



To measure party system volatility, we used a standard formula based on the percentage point differences in votes cast for all parties in two adjacent elections (we only had data for the three largest parties, but they accounted for almost 80 percent of the seats after the stimulus and referent elections). Because one-party systems, like China, emerged as the most stable party systems, we tested proposition 4 for only the 137 nations that qualified as “electoral democracies” according to Freedom House criteria. For these electoral democracies, party system stability had a significant effect on Rule of Law. The more stable the party system, the higher the RL.

Measuring fragmentation by the number of parties represented in parliament, we found no significant effects on Rule of Law. Other measures worked no better. (There was no correlation between our measures of fragmentation and competitiveness.) We suspect that Proposition 5 was not supported because it conflated two different theories of democratic government. The majoritarian model favors popular control of government through two rival groups struggling to gain control of government. The pluralist model (or better, the consociational model) favors having government shared by politically salient social groups.

The model—majoritarian or consociational—that operates in a given country depends on its governmental and social structure. Having multiple parties in parliament— which produces high party system fragmentation—works against the majoritarian model of government but tends to support the consociational model. Our test may have failed because it did not separate the countries according to the presumed operative model of democracy.

Notwithstanding the failure to support Proposition 5, we found support for all other proposed effects of party systems on country governance as measured by Rule of Law. Countries with party systems rate higher on RL than those without parties, excepting the few island nations that hold nonpartisan elections. More competitive party systems are significantly related to higher RL scores. In electoral democracies, more stable party systems are significantly related to higher RL scores. Although the percentages of explained variance increased only marginally, the party system coefficients were all statistically significant at the standard level.

Party System Properties: Cause or Effect?

To this point, we have not specified why greater party system competitiveness and stability should be related to higher Rule of Law scores for countries. Does the party




system contribute to their higher scores—as assumed in the propositions we tested—or does a more competitive and stable party system merely reflect the extent to which countries enforce the governmental value, rule of law? One can argue that party system competitiveness and stability are simply the effects of rule of law as a generic value. When countries observe the rule of law, opposition parties are freer to compete with governmental parties for political power in multiple elections. Thus, positive properties of party systems become the effect, not the cause, of rule of law.

We argue the contrary: that party system competitiveness and stability cause countries to promote the rule of law. While non-competitive parties have little incentive to promote it, competitive parties do have an incentive. They promote it because voters prefer government by rule of law to government by the whim of the rulers. Promoting the rule of law meshes with competitive parties’ strategic goals: to win votes and seats. Put more generally, a competitive party system has an incentive to promote country governance—of which the WGI variable, Rule of Law, is just one manifestation.

It is time to consider the other Worldwide Governance Indicators: Government Effectiveness, Control of Corruption, and Regulatory Quality. (We exclude Voice and Accountability because it includes party competition as a component.). Essentially, all our findings of party system effects hold for those indicators as well. The more competitive and stable the party system, the higher the Government Effectiveness, the Control of Corruption, and the Regulatory Quality.

The argument that rule of law causes party competitiveness by freeing parties to compete in elections does not apply to Government Effectiveness. How could one argue that effective government produces more competitive parties? How could one argue that party competitiveness is a reasonable consequence of Control of Corruption—or of Regulatory Quality?

That a competitive party system is significantly related to all four indicators of country governance suggests that the nature of the party system is causal, not consequential. Recall our earlier definition of governance as the extent to which a state delivers to its citizens the desired benefits of government at acceptable costs. Competitive political parties propose government benefits in order to win votes and seats. Hence, they promote the Rule of Law, Government Effectiveness, Control of Corruption, and Regulatory Quality—all meta-values that voters favor. That argues for the positive effect of party systems on country governance.

Party Systems and Country Governance discusses the many limitations of this cross-national study of one set of governance indicators in 2007. While not a definitive study, it did produce mostly strong and consistent results that support the underlying assumption of those who fund programs to develop party systems abroad: party systems do matter.