The Quality of Democracy and a “Kinder, Gentler” Democracy: Consensus Democracy Makes a Difference

(* This is the ch.16 of "Patterns of Democracy" by Arend Lijphart)

 

The conventional wisdom, cited in the previous chapter, argues—erroneously, as I have shown—that majoritarian democracy is better at governing, but admits that consen sus democracy is better at representing—in particular, representing minority groups and minority interests, representing everyone more accurately, and representing people and their interests more inclusively. In the first part of this chapter I examine several measures of the quality of democracy and democratic representation and the extent to which consensus democracies perform better than majoritarian democracies according to these measures. In the second part of the chapter I discuss differences between the two types of democracy in broad policy orientations. Here I show that consensus democracy tends to be the “kinder, gentler” form of democracy. I borrow these terms from President George H. W. Bush’s acceptance speech at the Republican presidential nominating convention in August 1988, in which he asserted: “I want a kinder, and gentler nation” (New York Times, August 19, 1988, A14). Consensus democracies demonstrate these kinder and gentler qualities in the following ways: they are more likely to be welfare states; they have a better record with regard to the protection of the environment; they put fewer people in prison and are less likely to use the death penalty; and the consensus democracies in the developed world are more generous with their economic assistance to the developing nations.

 

CONSENSUS DEMOCRACY AND DEMOCRATIC QUALITY

 

Table 16.1 presents the results of multivariate regression analyses of the effect of consensus democracy on six sets of indicators of the quality of democracy. The organization of the table is similar to that of Tables 15.1 and 15.2 in the previous chapter. The independent variable is the degree of consensus democracy on the executives-parties dimension in the period 1981–2010, and the control variables are the level of economic development and logged population size. The first indicator is the overall measure of democratic quality produced by Worldwide Governance Indicators project: “voice and accountability,” defined as the extent to which citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free press (Kaufmann, Kraay, and Mastruzzi 2010). Like the five WGI indicators used in the previous chapter, the scale ranges from −2.5 to +2.5, and the scores are the averages of the eleven scores assigned to each of our thirty-six countries between 1996 and 2009. All of our democracies receive positive scores, and their empirical range is much narrower than the theoretically possible five-point difference: from a low of 0.28 to a high of 1.58. Relatively low performers are Argentina (0.28) and India (0.37) and the best performers are Denmark (1.59) and New Zealand (1.58). The estimated regression coefficient is therefore a modest 0.086, but it is statistically significant at the 5 percent level. The score of the average consensus democracy is approximately one-sixth of a point (twice the regression coefficient) higher than that of the average majoritarian democracy. The level of development and population size have strong impacts as well (at the 1 and 5 percent levels, respectively): the more developed and smaller countries tend to receive the higher ratings.

 

Multivariate regression analyses of the effect of consensus democracy (executives-parties dimension) on nineteen indicators of the quality of democracy, with controls for the effects of the level of economic devel￾opment and logged population size, and with extreme outliers removed

 

  In order not to clutter the discussion with repeated references to the two control variables, which, however important, are not our main focus, let me briefly summarize the general pattern for all of the performance variables discussed in this chapter—which is very similar to the situation for “voice and accountability” reported in the previous paragraph. The level of development almost always has the greater impact, usually at the 1 or 5 percent level, and it almost always has a favorable influence (for instance, more voice and accountability, better women’s representation, and less inequality). Population size does not have as strong an impact; if significant, the effect is usually at the 5 or 10 percent level; and it usually has an unfavorable influence. Both variables are clearly influential to such an important extent that they must be used as controls in all of the regression analyses. When I report the effects of consensus democracy on the performance variables in this chapter, as in the previous chapter, these are always the effects with level of development and population size controlled for. Without these controls, the bivariate correlations between consensus democracy and the various performance variables would invariably be stronger—but deceptively strong and not at all meaningful.

 

More detailed measures of democratic quality than the above WGI index have been constructed by the Economist Intelligence Unit (EIU) in 2006, 2008, and 2010. The EIU’s overall index of democracy is an average of the scores in the five categories shown in Table 16.1. Each category is composed of an average of twelve subcategories. Most of the countries in the world are covered by the EIU surveys, including thirty-four of our thirty-six countries; only the Bahamas and Barbados are missing. Let me give a few examples of the questions that the EIU asks about each country. In the first category, electoral process and pluralism: “Are elections for the national legislature and head of government free and fair?”; “Are municipal elections both free and fair?”; “Do laws provide for broadly equal campaigning opportunities?”; and “Do opposition parties have a realistic prospect of achieving government?” Questions for the second category, the functioning of government, include: “Do freely elected representatives determine government policy?”; “Do special economic, religious or other powerful domestic groups . . . exercise significant political power, parallel to democratic institutions?”; “Are sufficient mechanisms and institutions in place for assuring government accountability to the electorate in between elections?”; and “Is the functioning of government open and transparent, with sufficient public access to information?” The third category, political participation, has questions about interest and participation in elections, political parties, other organizations, and lawful demonstrations, and women’s legislative representation. The fourth category, political culture, focuses on the degree to which citizens express faith in and support for democracy. The fifth category, civil liberties, looks at the traditional freedoms of expression, association, and religion, a free and robust press and other media of communication, equal treatment under the law, and an independent judiciary (Economist Intelligence Unit 2010, 33–42).

 

Table 16.1 shows the effect of consensus democracy on the five categories of democratic quality and on the overall EIU democracy index (averaged over the years 2006, 2008, and 2010), which are measured on a ten-point scale, after the effects of level of development and population size have been taken into account. Consensus democracy has very strong effects on four of the performance variables (at the 1 percent level) and somewhat weaker but still significant effects on the first and fourth categories. Israel is an extreme outlier on the civil liberties variable and was therefore removed from the analysis; its score of 5.29 is far below those of all of the other countries that are in a narrow range between 8.04 and 10.00. The highest scores on the overall EIU democracy index are Sweden’s (9.75) and Norway’s (9.68), and the lowest are Argentina’s (6.70) and Trinidad’s (7.18). The average consensus democracy scores more than half a point higher than the average majoritarian democracy.

 

Both the Worldwide Governance Indicators project and the Economist Intelligence Unit use accountability as one of their criteria for high-quality democracy. This is indeed a crucial democratic desideratum, and a frequent claim in favor of majoritarian democracy is that its typically one-party majority governments offer clearer responsibility for policy-making and hence better accountability of the government to the citizens—who can use elections either to “renew the term of the incumbent government” or to “throw the rascals out” (Powell 1989, 119). The claim is undoubtedly valid for majoritarian systems with pure or almost pure two-party competition like the Barbadian prototype discussed in Chapter 2. However, in two-party systems with significant third parties, “rascals” may be repeatedly returned to office in spite of clear majorities of the voters voting for other parties and hence against the incumbent government. All reelected British cabinets since 1945 fit this description; in 2005 the negative vote of almost two-thirds (64.8 percent) of the voters against the incumbent Labour party was insufficient to dislodge it from power. Moreover, it is actually easier to change governments in consensus democracies than in majoritarian democracies, as shown by the shorter duration of cabinets in consensus systems (see the first column of Table 7.1). Admittedly, of course, changes in consensus democracies tend to be partial changes in the composition of cabinets, in contrast with the more frequent complete turnovers in majoritarian democracies.

 

WOMEN’S REPRESENTATION

 

The next five performance variables in Table 16.1 measure women’s political representation and the inequality between women and men. The representation of women in parliaments and cabinets is an important measure of the quality of democratic representation in their own right, and it can also serve as an indirect proxy of how well minorities are represented generally. That there are so many kinds of ethnic and religious minorities in different countries makes comparisons extremely difficult, and it therefore makes sense to focus on the “minority” of women—a political rather than a numerical minority—that is found everywhere and that can be compared systematically across countries. As Rein Taagepera (1994, 244) states, “What we know about women’s representation should also be applicable to ethnoracial minorities.”

 

I chose years in the 1990s and in the first decade of the twentyfirst century for the measurement of the percentages of women elected to the lower or only houses of parliament and the percentages of women’s participation in cabinets. Women’s parliamentary representation has increased at a rapid rate according to several studies (Sawer, Tremblay, and Trimble 2006, Tremblay 2008); I deliberately selected 1990 and 2010 in order to discover the exact extent to which women succeeded in improving their share of representation in our long-term democracies over these twenty years.[1] They more than doubled their representation: the respective percentages are 12.0 and 24.9. For cabinet representation, the same long time span was not available, but from 1995 to 2008 women also improved their participation in cabinets from 15.5 to 26.5 percent. In spite of these overall improvements, major differences have persisted between consensus and majoritarian democracies. The effect of consensus democracy on women’s legislative representation in both years is strong and highly significant (at the 1 percent level). In 2010, the highest percentages were those of Sweden (45.0 percent), Iceland (42.9 percent), and the Netherlands (40.7 percent). The lowest were Botswana’s (7.9 percent) and Malta’s (8.7 percent). In both years, there were more than 9 percentage points more women in the first or only chambers in the average consensus than in the average majoritarian democracy. The results for women’s cabinet representation are similar although less strong (at the 5 instead of the 1 percent level of significance). Women were better represented in the average consensus democracy than in the average majoritarian democracy by about 8 percentage points.

 

The tables also shows the gender inequality index devised by the United Nations Development Programme (2010, 219). It  “reflects women’s disadvantage in three dimensions—reproductive health, empowerment and the labour market—for as many countries as data of reasonable quality allow.” It is a good overall measure of the status of women and is available for all of our democracies except the Bahamas. For the other thirty-five countries the scale ranges from 0.174 (indicating low inequality) to 0.748 (indicating high inequality). At the high end are India (0.748). Botswana (0.663), and Jamaica (0.638); at the low end are the Netherlands (0.174), Denmark (0.209), and Sweden (0.212). The effect of consensus democracy on the gender inequality index is strongly negative and highly significant (at the 1 percent level). The average consensus democracy has an index score that is about 0.075 lower than the average majoritarian system.

 

[1]. The increase in women’s legislative representation is partly due to the introduction of gender quotas by political parties and legislatures. Mona Lena Krook’s (2009) comparative study of this subject presents both a global perspective and detailed case studies of legislative and party quotas in four of our democracies: Argentina, France, Sweden, and the United Kingdom.

 

POLITICAL EQUALITY

 

Political equality is a basic goal of democracy, and the degree of political equality is therefore an important indicator of democratic quality. Political equality is difficult to measure directly, but economic equality can serve as a valid proxy, since political equality is more likely to prevail in the absence of great economic inequalities: “Many resources that flow directly or indirectly from one’s position in the economic order can be converted into political resources” (Dahl 1996, 645). Table 16.1 shows three measures of income inequality for varying years around 2000 provided by the United Nations Development Programme (2007, 281–84). The first compares the income share of the richest 10 percent to the poorest 10 percent of the population. The second is a similar measure comparing the richest to the poorest 20 percent. These data are available for all of our democracies except the five smallest countries and Mauritius. Botswana is an outlier with extremely high inequality and was removed from the analysis. The 10/10 ratio ranges from a high of 31.8 for Argentina to a low of 4.5 for Japan; the 20/20 ratio ranges from 16.3 to a low of 3.4, with Argentina and Japan again at opposite ends of the scale. The effect of consensus democracy on both variables is very strong and significant (at the 1 percent level). An even better and more comprehensive measure is the Gini index of inequality which has a theoretical range of 100, indicating extreme inequality (with one person receiving all of the country’s income) to zero, indicating complete equality. Botswana has the highest inequality (60.5) but cannot be considered an extreme outlier and is therefore included in this part of the analysis. After Botswana the highest inequalities, above 40.0, occur in the Western hemisphere: Argentina (51.3), Costa Rica (49.8), Jamaica (45.5), Uruguay (44.9), and the United States (40.8). The lowest inequalities are found in Denmark (24.7), Japan (24.9), and Sweden (25.0). The effect of consensus democracy on this measure of inequality is even stronger and more highly significant than on the two ratio measures. The average consensus democracy has a Gini index that is more than 9 points lower than the average majoritarian democracy.

 

ELECTORAL PARTICIPATION

 

Voter turnout is an excellent indicator of democratic quality for two reasons. First, it shows the extent to which citizens are actually interested in being represented. Second, turnout is strongly correlated with socioeconomic status and can therefore also serve as an indirect indicator of political equality: high turnout means more equal participation and hence greater political equality; low turnout spells unequal participation and hence more inequality (Lijphart 1997). Table 16.1 uses the turnout percentages in legislative elections in parliamentary democracies and the average turnout percentages in presidential and legislative elections in presidential systems. The percentage for each country is the mean turnout in all elections between 1981 and 2010. The basic measure is the number of voters as a percentage of voting-age population.[2]

 

[2]. This is a more accurate measure of turnout than actual voters as a percent of registered voters, because voter registration procedures and reliability differ greatly from country to country. The only problem with the voting-age measure is that it includes noncitizens and hence tends to depress the turnout percentages of countries with large noncitizen populations. Because this problem assumes extreme proportions in Luxembourg with its small citizen and relatively very large noncitizen population, I made an exception in this case and used the turnout percentage based on registered voters.

 

Average voter turnout varies a great deal from country to country in our thirty-six democracies—from a low of 38.3 to a high of 95.0 percent. The countries with the highest voter turnout are Malta (95.0 percent), Uruguay (94.5 percent), and Luxembourg (88.5 percent). At the low end of the range are Switzerland (38.3 percent), Botswana (46.5 percent), Jamaica (50.6 percent), and the United States (51.3 percent). Consensus democracy has a significant positive effect on voter turnout, but the effect is relatively weak and significant only at the 10 percent level. One possible explanation for the weak relationship is that turnout is also affected by the presence or absence of compulsory voting, which tends to increase turnout. Of the three highest-turnout countries above, Uruguay and Luxembourg have mandatory voting laws with sanctions—usually modest fines—that are actually enforced. Three other countries have such laws: Argentina, Australia, and Belgium (International IDEA 2010, Birch 2009). In order to check whether compulsory voting changes the effect of consensus democracy on voter turnout, it can be entered as a dummy control variable (in addition to the two standard controls of level of development and population size) in the multivariate regression analysis. The result is an estimated regression coefficient of 3.178—almost identical to the 3.185 without mandatory voting as a control—and the level of significance is barely changed. A second check is to run the regression analysis without the five countries with compulsory voting. Table 16.1 shows that for the thirty-one countries with voluntary voting, the results are again almost the same: all three regression coefficients are remarkably close to each other, and they are all statistically significant only at the 10 percent level.

 

Another potential disturbing influence is suggested by the fact that in two countries with the lowest turnouts—Switzerland and the United States—turnout is severely depressed by the high frequency of elections and the multitude of electoral choices to be made. When the frequency of elections as well as compulsory voting and the two standard control variables are controlled for, the effect of consensus democracy on total turnout becomes much stronger and is now significant at the 5 percent level. The effect on nonmandatory turnout when the frequency of elections is controlled for is about the same. The estimated regression coefficients are 3.719 and 3.634, respectively, both significant at the 5 percent level. The two coefficients show that the average consensus democracy has a voter turnout that is more than 7 percentage points higher than the turnout in the average majoritarian democracy.

 

SATISFACTION WITH DEMOCRACY

 

Does type of democracy affect citizens’ satisfaction with democracy? Hans-Dieter Klingemann (1999) reports the responses to the following survey question asked in many countries, including eighteen of our democracies, in 1995 and 1996: “On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the way democracy works in (your country)?” The Danes and Norwegians expressed the highest percentage of satisfaction with their democracies: 83 and 82 percent, respectively, said that they were very or fairly satisfied. The Italians and Greeks were the least satisfied: only 19 and 28 percent, respectively, expressed satisfaction. The low percentage in the Italian survey conducted in 1995 is due at least in part to the political turbulence in Italy following the first election after Italy’s drastic electoral reform. Table 16.1 reports the effect of consensus democracy on satisfaction with democracy after Italy is removed as an outlier. The correlation is positive, but only at the 10 percent level. When Italy is included in the analysis, the statistical significance falls below 10 percent, but the effect of consensus democracy is still clearly positive.

 

A similar question was asked in the World Values Survey in a large number of countries, including nineteen of our democracies, in 2005–7: “How democratically is your country being governed today? . . . Using a scale from 1 to 10, where 1 means that it is ‘not at all democratic’ and 10 means that it is ‘completely democratic,’ what position would you choose?” Respondents who chose the high numbers 8 to 10 can be counted as being satisfied with their democratic system. The highest percentage is Norway’s 74.1 percent, and the lowest—again—is Italy’s 24.5 percent. Italy is no longer an outlier in this respect, however, because several other countries have only slightly higher percentages of satisfaction: Korea (29.5 percent), the Netherlands (30.1 percent), Trinidad (32.1 percent), the United Kingdom (33.3 percent), and the United States (35.5 percent). With all nineteen countries included in the multivariate analysis, democracy has a positive effect on satisfaction with democracy, but only at the 10 percent level of significance. Pippa Norris (2011, 214) reports a similar positive but small impact of proportional representation.

 

These results should be treated with caution because they are based on only seventeen to nineteen countries. Moreover, the results of the 1995–96 and 2005–7 surveys are not strictly comparable, because the questions about democratic satisfaction were phrased differently and also because the surveys were conducted in different countries: only eleven of our democracies were included in both surveys. The average percentage of respondents expressing approval in 2005–7 is also considerably lower than in 1995–96: 44.3 percent versus 54.6 percent (including Italy’s low percentage). However, these differences can also be interpreted as strengthening the conclusion in favor of consensus democracy: in two surveys held ten years apart, in different sets of countries, with different questions, and with different overall levels of approval, consensus democracy still has roughly the same positive and statistically significant effect on citizens’ satisfaction with the operation of their democratic systems.

 

The general conclusion is that consensus democracies have a better record than majoritarian democracy on all of the measures of democratic quality in Table 16.1, that all of the favorable effects of consensus democracy are statistically significant, and that more than half are significant at the most demanding 1 percent level. This conclusion applies to the effect of consensus democracy on the executives-parties dimension. In order to test the effect of consensus democracy on the federal-unitary dimension, I repeated the nineteen regression analyses reported in Table 16.1 with consensus-federalist democracy as the independent variable— with the same controls and with the same outliers removed from the analysis. Without only a few slight exceptions, the relationships are extremely weak and statistically insignificant even at the 10 percent level. Consensus-federalist democracy has more unfavorable than favorable effects, but this finding is counterbalanced by the positive—but far from statistically significant— effects it has on the WGI indicator of voice and accountability and the overall EIU democracy index, which are the broadest and most comprehensive indicators of the quality of democracy. As in the previous chapter, I should emphasize that the effects are so weak that they do not allow any substantive conclusions in favor of one or the other type of democracy—and that they are not worth reporting in detail.

 

CONSENSUS DEMOCRACY AND ITS KINDER, GENTLER QUALITIES

 

The democratic qualities discussed so far in this chapter should appeal to all democrats: it is hard to find fault with better performance on the fundamental criteria of democracy used by the Worldwide Governance Indicators and the Economist Intelligence Unit, and with better results for women’s representation, political equality, and participation in elections. In addition, consensus democracy (on the executives-parties dimension) is associated with some other attributes that I believe most, though not necessarily all, democrats will also find attractive: a strong community orientation and social consciousness—the kinder, gentler qualities mentioned at the beginning of this chapter. These characteristics are also consonant with feminist conceptions of democracy that emphasize, in Jane Mansbridge’s (1996, 123) words, “connectedness” and “mutual persuasion” instead of self-interest and power politics: “The processes of persuasion may be related to a more consultative, participatory style that seems to characterize women more than men.” Mansbridge further relates these differences to her distinction between “adversary” and “unitary” democracy, which is similar to the majoritarian-consensus contrast. Accordingly, consensus democracy may also be thought of as the more feminine model and majoritarian democracy as the more masculine model of democracy.

 

There are four areas of government activity in which the kinder and gentler qualities of consensus democracy are likely to manifest themselves: social welfare, the protection of the environment, criminal justice, and foreign aid. My hypothesis is that consensus democracy will be associated with kinder, gentler, and more generous policies. Table 16.2 presents the results of the multivariate regression analyses of the effect of consensus democracy on eight indicators of the policy orientations in these four areas. The independent variable in all cases is the degree of consensus democracy on the executives-parties dimension in the period 1981–2010. The control variables are again the level of economic development and logged population size.

 

Multivariate regression analyses of the effect of consensus democracy (executives-parties dimension) on eight indicators of social welfare ex penditures, environmental performance, criminal justice, and foreign aid, with controls for the effects of the level of economic development and logged population size, and with extreme outliers removed

 

Determining the degree to which democracies are welfare states is an extremely difficult task (Castles, Leibfried, Lewis, Obinger, and Pierson 2010). In particular, it is not sufficient simply to count the total amount of direct public social expenditure as a percentage of gross domestic product, because this amount is invariably reduced by direct and/or indirect taxes paid by the recipients of social benefits. The most careful analysis of the funds that should be included and that should be subtracted to arrive at the net expenditure on social welfare is the study “How Expensive Is the Welfare State?” by Willem Adema and Maxime Ladaique (2009), which covers the member countries of the Organization for Economic Development and Cooperation (OECD), including twenty-two of our democracies, in the year 2005. The first two rows of Table 16.2 are based on their calculations. Net public social expenditure consists of all direct public social expenses plus “tax breaks for social purposes that mirror cash benefits,” minus all direct and indirect taxes and social contributions paid by beneficiaries. The second row in the table, net publicly mandated social expenditure, adds private social expenditure that is mandated by the state, again minus direct and indirect taxes and social contributions.[3] Although the second total is only slightly higher than the first in most countries, it is worth testing the effect of consensus democracy on both percentages. France has the highest social expenditures as percentages of GDP (30.4 and 30.7 percent), followed by Germany (28.1 and 28.8 percent), and Sweden (27.3 and 27.5 percent). Korea has by far the lowest percentages (8.0 and 8.6 percent); the next lowest are Ireland (twice 17.2 percent), New Zealand (twice 18.4 percent), the United States (18.4 and 18.8 percent) and Iceland (18.1 and 19.3 percent). The effect of consensus democracy on both of the social expenditure totals is strongly positive and statistically significant at the 5 percent level. The social expenditures of the average consensus democracy are about 4.75 percentage points higher than those of the typical majoritarian democracy.

 

The best indicator of how well countries do with regard to protecting the environment is the Environmental Performance Index, produced by a team of environmental experts at Yale University and Columbia University. It is a broad and comprehensive index that rates the performance of most of the countries in the world on twenty-five indicators in ten policy areas, including environmental health, air quality, water resource management, biodiversity and habitat, forestry, fisheries, agriculture, and climate change (Yale Center for Environmental Law and Policy 2010). The first report was based on a pilot project and was published in 2006. Updates were released in 2008 and 2010. I used the ratings in the 2010 report in Table 16.2 because it includes the largest number of the world’s countries and thirty-four of our democracies; only the Bahamas and Barbados are missing.

 

[3]. Adema and Ladaique present a third total that also includes voluntary private social expenditure which, in my opinion, is at odds with the basic concept of the welfare state in which it is the state that directly or indirectly serves as the provider of social protection.

 

Countries are rated on a scale from 100, indicating the best performance, to zero, indicating the poorest performance, although in practice no country is rated even close to zero; the worst performer, Sierra Leone, ranked number 163, still has a score of 32.1. Among our thirty-four democracies, Iceland receives the highest score (93.5), followed by Switzerland (89.1), Costa Rica (86.4), Sweden (86.0), Norway (81.1), and Mauritius (80.6). The poorest performers are Botswana (41.3), India (48.3), Trinidad (54.2), Korea (57.0), Jamaica (58.0), and Belgium (58.1). Table 16.2 shows that consensus democracy has a positive and statistically significant effect (at the 5 percent level) on environmental performance. Consensus democracies score more than six points higher than majoritarian democracies. As in all of the tables in Chapters 15 and 16, levels of development and population size are controlled for, and the former has a significant positive effect on environmental performance, too. The above examples show, however, that it is not always the most developed countries that receive the highest scores: Costa Rica and Mauritius are among the better and Korea and Belgium are among the poorer protectors of the environment.

 

One would also expect the qualities of kindness and gentleness in consensus democracies to show up in criminal justice systems that are less punitive than those of majoritarian democracies, with fewer people in prison and with less or no use of capital punishment. To test the hypothesis with regard to incarceration rates, I used the numbers collected by the International Centre for Prison Studies (2011), available for all of our democracies. These rates represent the number of inmates per hundred thousand population. The highest and lowest rates are those of the United States and India: 743 and 32 inmates per hundred thousand population, respectively. In fact, the United States is an extreme outlier: its 743 prisoners per hundred thousand people is about twice as many as the 376 inmates in the next most punitive country, the Bahamas. After the United States and the Bahamas, the next most punitive countries are Barbados (326), Israel (325), and Trinidad (276). The least punitive countries after India are Japan (59), and Finland and Iceland (both with 60 inmates per hundred thousand population). When the United States is removed from the analysis, the effect of consensus democracy on incarceration rates is strongly negative and significant at the 1 percent level. The consensus democracies put almost 60 fewer people per hundred thousand population in prison than the majoritarian democracies.

 

As of the end of 2010, according to the data collected by Amnesty International, eight of our thirty-six democracies retained and used the death penalty: the Bahamas, Barbados, Botswana, India, Jamaica, Japan, Trinidad, and the United States. The laws of twenty-six countries did not provide for the death penalty for any crime. The remaining two countries—Israel and Korea— were in the intermediate category of countries with the death penalty only for exceptional crimes, such as crimes under military law, or having a policy of not carrying out executions. On the basis of these differences, I constructed a three-point scale with a score of two for the active use of the death penalty, zero for the absence of the death penalty, and one for the intermediate cases. The effect of consensus democracy on the use of capital punishment is strongly negative and significant at the 5 percent level.

 

In the field of foreign policy, one might plausibly expect the kind and gentle characteristics of consensus democracy to be manifested by generosity with foreign aid and a reluctance to rely on military power.[4] Table 16.2 uses three indicators for more than twenty OECD countries: foreign aid—that is, economic development assistance, not military aid—as a percentage of gross national product at the end of the Cold War in 1990; foreign aid in 2005, fifteen years later; and foreign aid in 2005 as a percent of defense expenditures. In 1990, foreign aid ranged from a high of 1.17 percent of gross national product (Norway) to a low of 0.11 percent (Austria); in 2005, the highest percentage was 0.98 percent (Sweden) and the lowest 0.19 (Japan and the United States). The highest foreign aid as a percent of defense expenditure was Ireland’s 70 percent, and the lowest was that of the United States, 5 percent.

 

[4]. This hypothesis can also be derived from the “democratic peace” literature (Lijphart and Bowman 1999). The fact that democracies are more peaceful, especially in their relationships with each other, than nondemocracies is often attributed to their stronger compromise-oriented political cultures and their institutional checks and balances. If this explanation is correct, one should expect consensus democracies to be even more peace-loving than majoritarian democracies.

 

In the analysis of the effect of consensus democracy on these three performance variables it is especially important to use the standard controls for level of development and population size: wealthier countries can better afford to give foreign aid than less wealthy countries, and large countries tend to assume greater military responsibilities and hence tend to have larger defense expenditures—which can be expected to limit their ability and willingness to provide foreign aid. In the multivariate analyses, consensus democracy has a positive effect on giving foreign aid and on foreign aid as a percentage of military expenditures, which is statistically significant at the 5 percent level for two of the performance variables and at the 10 percent level for the third. The average consensus democracy gave about 0.27 percent more of its gross national product in foreign aid than the average majoritarian democracy in 1990 and about 0.17 percent more in 2005. Its aid as a percent of defense spending was more than 16 percentage points higher.

 

Similar regression analyses can be performed to test the effects of the other (federal-unitary) dimension of consensus democracy on the above eight indicators, with the same controls in place and with the United States removed from the analysis of imprisonment rates. These analyses yield no interesting results. Consensus-federalist democracy has a favorable effect on five of the performance variables and an unfavorable effect on three—but the effects are all small and not statistically significant.

 

As the subtitle of this chapter states: consensus democracy makes a difference. Indeed, the results could hardly be clearer: consensus democracy—on the executives-parties dimension—makes a big and highly favorable difference with regard to almost all of the indicators of democratic quality and with regard to all of the kinder and gentler qualities.