June 2000 Fraser Forum: Democracy & the Rule of Law

[Previous][Contents] [Next]

Robert Barro

In the long run, the difference between prosperity and poverty depends on how fast an economy grows. Much of the recent empirical research on the determinants of growth has concentrated on the role of government policy in fostering this long-term growth. Although the government's influence includes the fiscal and monetary instruments that have been the main focus of macroeonomists, an even more important dimension of policy concerns the character of a nation's political, legal, and economic institutions. Differences in institutions across countries have proven empirically to be among the most important determinants of differences in rates of economic growth and investment. Consequently, basic reforms that improve the quality of institutions provide one of the best routes for transforming a country in the long run from poverty to prosperity.

The question of which aspects of institutions matter for long-run economic performance has proven to be more controversial than the proposition that institutions are important overall. One strand of the recent research has focused on democracy, specifically on the strength of electoral rights and civil liberties. The second strand has emphasized property rights and legal structures that promote the rule of law. Some scholars, such as Milton Friedman in Capitalism and Freedom (1962), argue that these two aspects of liberal institutions are mutually reinforcing and that both are conducive to economic performance. Recent empirical research supports the idea that property rights and the rule of law are key determinants of economic growth and investment but delivers mixed results with respect to the contributions from democracy. Before we turn to this empirical evidence it is worthwhile to assess the situation theoretically.

Economic Effects of Property Rights and the Legal System

The economic effects of secure property rights and a well-functioning legal system are reasonably straightforward. Since people are, to a considerable degree, self-interested, they tend to undertake hard work and investments only if they have a reasonable probability of enjoying the fruits of their efforts. Thus, if property rights are insecure—for example, because of high crime rates or high rates of taxation or high chances of government expropriation—then people tend to work and invest little. The concept of high taxation can be extended from income taxes or other formal levies to include onerous government regulations and licensing requirements, as well as bribes required by corrupt officials.

Vigorous business activity also depends to a considerable extent on a legal system that allows for contracts to be clearly specified and enforced. This contracting potential influences relations of businesses with suppliers, creditors, workers, and customers. For example, if the legal system does not enforce the repayment of loans, then loans will be scarce, and many productive investments will therefore not occur. (The private sector can respond to some extent to the public vacuum by creating its own enforcement, such as that provided by the Mafia.)

One way that businesses can react to poorly defined property rights is to reduce their levels of operations. However, another possibility—especially in response to high rates of taxation and oppressive regulations—is to move from the formal part of the economy to the informal or black-market sector. This informality may be better for the economy than a cessation of operations, but it does entail costs. Informal operation tends to be less efficient because businesses have to expend resources to conceal their activities. In addition, black-market participants typically lose access to useful government services, such as contract enforcement. Another effect is that the government fails to raise much in taxes on black-market activities, and the amounts collected from legal enterprises must therefore rise to pay for a given level of public expenditures. (For a discussion of these issues, see Loayza [1996].)

The stress on property rights and the legal system does not imply that more government activity is necessarily good or bad for economic performance. Some actions—such as maintenance of internal and external security and enforcement of contracts—entail more government spending and tend to enhance economic activity. Others—such as burdensome regulations and nonproductive expenditures that require high tax rates—would hinder the economy.

Economic Effects of Democracy

What effects on the economy would we anticipate from an expansion of democracy, say in the form of an increase in electoral rights? One effect, characteristic of systems of one-person/one-vote majority voting, involves the pressure to enact redistributions of income from rich to poor. These redistributions may involve land reforms and various social-welfare programs. Although the direct effects on income distribution may be desirable (because they are equalizing), these programs tend to compromise property rights and reduce the incentives of people to work and invest. One kind of disincentive involves the transfers given to poor people. Since the amount received typically falls as the person earns more income, the recipient is motivated to remain on welfare or otherwise disengage from productive activity. The other adverse effect involves the income taxes or other levies that are needed to pay for the transfers. An increase in these taxes encourages the non-poor to work and invest less.

One offsetting effect is that an evening of income distribution may reduce the tendency for social unrest. Specifically, transfers to the poor may reduce incentives to engage in criminal activity, including riots and revolutions. Since social unrest reduces everyone's incentives to work and invest, some amount of publicly organized income redistribution would contribute to overall economic activity. However, even a dictator would be willing to engage in transfers to the extent that the decrease in social unrest were worth the cost of the transfers. Thus, the main point is that democracy will tend to generate “excessive” transfers purely from the standpoint of maximizing the economy's total output.

Although democracy has its down side, one cannot conclude that autocracy provides ideal economic incentives. One problem with dictators is that they have the power and, hence, the inclination to steal the nation's wealth. More specifically, an autocrat may find it difficult to convince people that their property will not be confiscated once investments have been made. This convincing can sometimes be accomplished through reputation—that is, from a history of good behavior—but also by relaxing to some degree the hold on power. In this respect, an expansion of democracy—viewed as a mechanism for checking the power of the central authority—may enhance property rights and, thereby, encourage economic activity. From this perspective, democracy would encompass not only electoral rights but also civil liberties that allow for freedom of expression, assembly, and so on.

Theoretical reasoning suggests that enhanced property rights and the rule of law will likely encourage economic activity. The overall effects of expanded democracy, particularly in the sense of voting rights, are ambiguous. To sort out these relationships, we now turn to empirical evidence. But the first thing to consider is the measurement of democracy, the rule of law, and related concepts.

Measuring Democracy

A number of researchers have provided quantitative measures of democracy, and Alex Inkeles (1991, p. x) finds in an overview study a “high degree of agreement produced by the classification of nations as democratic or not, even when democracy is measured in somewhat different ways by different analysts.” One of the most useful measures—because it is available for almost all countries annually on a consistent basis since 1972—is the one provided by Raymond Gastil (1982-83 and other years) and his followers at Freedom House. This source provides separate indexes for electoral rights and civil liberties.

The Freedom House concept of electoral rights uses the following basic definition: “Political rights are rights to participate meaningfully in the political process. In a democracy this means the right of all adults to vote and compete for public office, and for elected representatives to have a decisive vote on public policies” (Gastil, 1986-87 edition, p. 7). In addition to the basic definition, the classification scheme rates countries (somewhat impressionistically) as less democratic if minority parties have little influence on policy.

Freedom House applies the concept of electoral rights on a subjective basis to classify countries annually into seven categories, where group one is the highest level of rights and group seven is the lowest. This classification was made by Gastil and his associates and followers based on an array of published and unpublished information about each country. The original ranking from one to seven was converted here to a scale from zero to one, where zero corresponds to the fewest rights (Freedom House rank seven) and one to the most rights (Freedom House rank one). The scale from zero to one corresponds to a classification made by Kenneth Bollen (1990) for 1960 and 1965. The Bollen index differs mainly in that its concept of democracy goes beyond electoral rights.

To fix ideas on the meaning of the zero-to-one subjective scale, note first that the United States and most other OECD countries in recent years received the value 1.0, thereby being designated as full representative democracies. Dictatorships that received the value 0.0 in 1995 included Indonesia, Iraq, Syria, Zaire, and several other countries in Africa. Places that were rated in 1995 at 0.5—halfway along between dictatorship and democracy—included Colombia, Dominican Republic, Ghana, Guatemala, Malaysia, Mexico, Nicaragua, Paraguay, Senegal, and Sri Lanka.

The solid line in Figure 1 shows the time path of the unweighted average of the electoral-rights index for the years 1960, 1965, and 1972-1995. The number of countries covered rises from 99 in 1960 to 109 in 1965 and 138 from 1972 to 1995. The figure shows that the mean of the index peaked at 0.66 at the start, 1960, fell to a low point of 0.44 in 1975, and rose subsequently to 0.59 in 1995. Thus, there has been noticeable democratization since the mid 1970s, but the level has not yet reattained the value for 1960. (This comparison may, however, not be fully accurate because the data for 1960 are not precisely comparable to those used since 1972.)

The figure also demonstrates that the main reason for the decline in the world average of electoral rights after 1960 is the experience in Sub Saharan Africa. The dotted line shows that the average of the index in Sub Saharan Africa peaked at 0.58 in 1960 (26 countries), then (for 43 countries) fell to low points of 0.19 in 1977 and 0.18 in 1989 before rising to 0.40 in 1995. This pattern emerges because many of the African countries began with ostensibly democratic institutions when they became independent in the early 1960s, but most evolved into one-party dictatorships by the early 1970s. The democratization in Africa since 1989 has been substantial, and some of this development has likely reflected pressures from international aid givers, including international agencies and the United States. Whether this democratization will be sustained—unlike in the 1960s—is not yet known.

For countries outside of Sub Saharan Africa, the dashed line in the figure shows that the average of the electoral-rights index fell from 0.68 in 1960 (73 countries) to 0.55 in 1975 (95 countries). The index then rose by 1995 to 0.68. Thus, outside of Sub Saharan Africa, the democratization since the mid 1970s has been sufficient to reattain the value for 1960.

The Freedom House index of civil liberties is constructed in a similar way. The definition here is “civil liberties are rights to free expression, to organize or demonstrate, as well as rights to a degree of autonomy such as is provided by freedom of religion, education, travel, and other personal rights” (Gastil, 1986-87 edition, p. 7). In practice, the indicator for civil liberties is extremely highly correlated with that for electoral rights. Figure 2 shows this correlation in terms of the unweighted averages of the two indexes for 138 countries from 1972 to 1995.

For practical purposes, it makes little difference in the subsequent analysis—for example, in looking at the relation between democracy and economic growth—whether one uses the index for electoral rights or the one for civil liberties. The rest of this study focuses on the index of electoral rights and sometimes refer to this indicator as simply a measure of democracy.

Measurement of the Rule of Law

Many analysts believe that maintenance of property rights and the rule of law are central for investment and other aspects of economic activity. The empirical challenge has been to measure these concepts in a reliable way across countries and over time. Probably the best indicators available come from international consulting firms that advise clients on the attractiveness of countries as places for investments. These investors are concerned about institutional matters such as the prevalence of law and order, the capacity of the legal system to enforce contracts, the efficiency of the bureaucracy, the likelihood of government expropriation, and the extent of official corruption. These kinds of factors have been assessed by a number of consulting companies, including Political Risk Services in its publication International Country Risk Guide. This source is especially useful because it covers over 100 countries since the early 1980s. Although the data are subjective, they have the virtue of being prepared contemporaneously by local experts. Moreover, the willingness of customers to pay substantial fees for this information is perhaps some testament to their validity.

Among the various indicators available, the index for overall maintenance of the rule of law (also referred to as “law and order tradition”) turns out to have the most explanatory power for investment and economic growth. This index was initially measured by Political Risk Services in seven categories on a zero to six scale, with six the most favorable. To make the index comparable with the ones discussed before for electoral rights and civil liberties, the rule-of-law variable was converted to a zero-to-one scale, with zero indicating the poorest maintenance of the rule of law and one the best.

To understand the scale, note that the United States and most of the OECD countries (not counting Mexico and Turkey) had values of 1.0 for the rule-of-law index in recent years. However, Belgium, France, Greece, Portugal, and Spain were downgraded from 1.0 in 1996 to 0.83 in 1997. Also rated at 1.0 in 1997 were Hungary, Kuwait, Malta, Morocco, and Singapore. (Hong Kong was downgraded upon its return to China from 1.0 in 1996 to 0.83 in 1997.)

No country had a rating of 0.0 for the rule of law in 1997, but countries rated at 0.0 in some earlier years included Ethiopia, Guyana, Haiti, Sri Lanka, Yugoslavia, and Zaire. Countries rated at 0.5 in 1997 included Algeria, Brazil, Mexico, Peru, Uruguay, South Africa, several other countries in Sub Saharan Africa, and much of Central America.

The solid line in Figure 3 shows how the unweighted average of the rule-of-law indicator evolved from 1982 to 1997. (The number of countries covered was 88 in 1982-83, 100 in 1984, and 114 since 1985.) The average began at 0.50 in 1982 and showed little change from there to 1991. Then the average rose to 0.71 in 1996 and 0.70 in 1997. Hence, the 1990s have been a period in which the world average of maintenance of the rule of law has expanded.

Figure 3 also shows the behavior of two other series constructed by Political Risk Services, one for political corruption and the other for the efficiency of the bureaucracy. In these cases, zero indicates the most corruption and the least efficiency, whereas one indicates the least corruption and greatest efficiency. These two indicators turn out to be closely correlated with the rule-of-law indicator in terms of variations over time and across countries. However, once the rule-of-law measure is held constant, the indicators for corruption and bureaucratic efficiency do not contribute much to explaining economic growth and investment.

The last findings may reflect the two-sided nature of political corruption and bureaucratic efficiency. In some circumstances, corruption may be preferable to honest enforcement of bad rules. For example, outcomes may be worse if a regulation that prohibits some useful economic activity is thoroughly enforced rather than circumvented through bribes. However, the economy is hampered when few legitimate activities can be undertaken without bribes. Thus, the overall impact of more corruption may be ambiguous.

Similarly, enhanced bureaucratic efficiency has obvious advantages. However, if bureaucrats are carrying out activities in which they ought to be absent, then the economy may suffer from more bureaucratic efficiency. Moreover, there may be a tendency for the bureaucracy to grow larger when it is more efficient. Thus, the predicted net effect of bureaucratic efficiency is also uncertain.

As an overall tendency, countries that are strong in terms of rule of law (and also in terms of low corruption and high bureaucratic efficiency) tend also to be strong in terms of the democracy indicators, electoral rights and civil liberties. However, the correlation between any of the property-rights/legal-structure indicators, such as the rule-of-law index, and any of the democracy indicators, such as the electoral-rights index, is much weaker than that within either of the two categories. In particular, there are many cases in which the rule-of-law index is high while the electoral-rights index is low, and vice versa. These cross-country differences between rule of law and electoral rights make it possible to distinguish empirically the effects of these institutional characteristics on economic growth and other variables.

Table 1 shows countries with large gaps (magnitude of at least 0.5) between the rule-of-law and electoral-rights indexes. Cases in which the rule of law was high in relation to electoral rights included Chile, Hong Kong, Hungary, Poland, Singapore, and Taiwan in 1982 or 1985 and China, Hong Kong, Indonesia, Iran, Saudi Arabia, Singapore, and Syria in 1995. These countries maintained strong law and order but had relatively little democracy. In the typical case (with Hong Kong as an exception as a British colony), the country was run by a dictator who nevertheless promoted property rights and a reliable legal system. Prototypes of this kind of dictator are Pinochet in Chile, Lee in Singapore, and the Shah in Iran.

Countries in the reverse situation—in which electoral rights were high in relation to the rule of law—included Bolivia, Colombia, Cyprus, Dominican Republic, Greece, Honduras, Israel, Jamaica, Peru, Sri Lanka, Trinidad, and Venezuela in 1982 or 1985. In contrast, in 1995, only Burkina Faso was in this situation. (Countries with gaps of 0.33 in 1995 included Bolivia, Brazil, Costa Rica, Jamaica, Panama, South Africa, Trinidad, and Uruguay.) Countries in this group maintained a lot of democracy but were relatively weak in terms of property rights and legal protections. Interestingly, the substantial rise in the world average of the rule-of-law index in the 1990s (see Figure 3) has sharply curtailed the number of countries that have large positive gaps between electoral rights and the rule of law.

Effects of the Rule of Law and Electoral Rights on Economic Growth

Barro (1997) describes an empirical framework for assessing the effects of various factors on the rate of growth of real per capita gross domestic product (GDP). The growth rate is determined from an equation of the form

growth rate = F(y, y*)

The variable y represents the starting position of the economy, including the initial level of per capita GDP and the amounts of human capital in the forms of education and health. The variable y* represents the long-run position toward which the economy is heading. This position depends on government policies and other factors. For example, improved maintenance of property rights raises y*.

For given y*, the growth rate falls with y because of diminishing returns to the accumulation of physical and human capital. This force tends to generate a convergence pattern, whereby poor countries catch up to rich ones. For given y, the growth rate rises with y*. Therefore, improved policies—such as better maintenance of property rights—increase the growth rate.

The sample consists of roughly 100 countries observed over three decades—the 1960s, 1970s, and 1980s. This sample includes countries at vastly different levels of economic development, and places are excluded only because of missing data.

For the rule-of-law index, the empirical relation with the rate of economic growth is shown in Figure 4. This type of diagram is constructed so that the effect of the variable of interest—here the rule-of-law index—is measured after holding constant the influences of an array of other explanatory factors, including the initial level of real per capita GDP, initial amounts of schooling and health, and other variables. The important result shown in the figure is that a higher level of the rule-of-law index generates a higher rate of economic growth. This result turns out to be significant in a statistical sense.

The estimated effect of improved rule of law on growth is quantitatively large. Specifically, a rise by one category (among the seven used) in the Political Risk Services index is estimated to raise the growth rate on impact by 0.5 percent per year. A change from the worst rule of law (0.0) to the best (1.0) would contribute an enormous 3.0 percent per year to the growth rate. Of course, this kind of growth dividend from legal reform could arise only for cases, such as Haiti and Zaire, that began as total institutional disasters and were then able to transform themselves overnight into institutional marvels. For countries that already have reasonably well-functioning legal systems, such as the United States and most other OECD countries in the 1990s, the potential for this kind of growth enhancement through institutional improvement does not exist.

If we graphed the ratio of investment to GDP versus the rule-of-law index, then we would get a relation similar to that shown in Figure 4. That is, one route by which better rule of law promotes growth is by encouraging investment. However, a positive effect of the rule-of-law index on economic growth still appears even if the ratio of investment to GDP is held constant. This effect likely involves improved productivity of resources and the encouragement of investments that are not measured by the standard national-accounts variable.

Figure 5 shows the relation between the growth rate and the extent of democracy, as measured by the electoral-rights index. The overall relation between economic growth and democracy is weak. In particular, there are examples of dictatorships (values of electoral rights near zero) with high and low rates of growth and similarly for democracies (values of democracy near one). However, there is a suggestion of a nonlinear relation—an inverted u-shape—in which growth rises initially with democracy, reaches a peak at a value for the electoral-rights index of around 0.5, and then declines subsequently with further rises in democracy. This relationship, shown by the curve in the figure, is statistically significant. Moreover, the same kind of nonlinear relation emerges if we replace the growth rate on the vertical axis by the ratio of investment to GDP. That is, democracy also has a nonlinear (inverted-u) effect on a country's propensity to invest.

One way to interpret the results is that, in the worst dictatorships, an increase in electoral rights tends to increase growth and investment because the benefit from the limitations on governmental power is the key matter. But in places that have already attained a moderate amount of democracy, a further increase in electoral rights impairs growth and investment because the dominant effect comes from the intensified concern with social programs and income redistribution.

Specifically, the results suggest that growth would be somewhat reduced by further democratization beyond the levels attained in 1995 in countries such as Malaysia and Mexico (which had index values of 0.5 in 1995). Moreover, political liberalization probably already went by 1995 beyond the point of growth maximization in places such as Chile, South Korea, and Taiwan. These countries went from levels of the electoral-rights index of 0.17, 0.33, and 0.33, respectively, in the early 1980s to 0.83, 0.83, and 0.67, respectively, in 1995.

The results should not be taken as saying that dictatorship is desirable from the standpoint of economic performance. There are examples of autocrats—such as Pinochet in Chile, Fujimori in Peru, the Shah in Iran, and Lee and several others in East Asia—that produced good growth outcomes. There are, however, even more examples—including Marcos in the Philippines, Mao in China, Mobutu and numerous other despots in Africa, and many others in South America and eastern Europe—that delivered poor growth outcomes.

History suggests that dictators come in two types: one that focuses on theft and another that emphasizes economic development (and relies to a considerable degree on free markets and property rights to accomplish this objective). More fundamentally, however, a dictator's quest for individual wealth would not imply an indifference to the size of the national pie that can be taxed. Hence, a dictator's self-interest would be consistent with a desire for national economic development, at least if this development did not threaten the leader's hold on power. The statement that autocrats come in two types is therefore really an admission that the theory of dictatorial behavior has not been worked out very well. Possibly the unpredictability of a dictator's behavior is fundamental—in this case, the choice of an autocratic political structure can be viewed as a risky investment at the national level. In this case, autocracy would always be an unreliable way to generate economic growth, even if a nation were willing to accept dictatorship on other grounds.

The findings do not support the idea that democracy is necessary for growth—or even that democracy beyond an intermediate range is likely to raise the rate of economic growth. Although it may be unpleasant, the data suggest that, once an intermediate level of democratization has been attained, a further expansion of democracy comes at the expense of economic performance.

Determinants of Democracy and the Rule of Law

The analysis, thus far, has dealt with the impact of alternative institutional arrangements—specifically, more or less democracy and the rule of law—on the economy. But nothing has yet been said about how the different institutional arrangements come about and, particularly, how these arrangements are influenced by the state of the economy.

A common view since the research of Seymour Martin Lipset (1959) is that prosperity, measured in various ways, stimulates democracy. This idea is often called the Lipset hypothesis, although Lipset (1959, p. 75) appropriately credits the basic notion to Aristotle: “From Aristotle down to the present, men have argued that only in a wealthy society in which relatively few citizens lived in real poverty could a situation exist in which the mass of the population could intelligently participate in politics and could develop the self-restraint necessary to avoid succumbing to the appeals of irresponsible demagogues.” (For a statement of Aristotle's views, see Aristotle [1932, book VI].) The fair thing seems to be to refer to this idea as the Aristotle-Lipset hypothesis.

Theories of the effect of economic conditions on the extent of democracy or property rights are not well developed. Lipset (1959, pp. 83-84) emphasized increased education and an enlarged middle class as key elements, and he also stressed Tocqueville's (1835) idea that private organizations and institutions are important as checks on centralized government power. In some models, an autocrat would voluntarily relinquish authority—for example, by establishing a constitution, allowing power to a legislative body, expanding voting rights, and extending civil liberties—in order to deter revolutions and to encourage the private sector to invest (and, thereby, to expand the pie that the government can tax).

Despite the lack of strong theoretical underpinnings, the cross-country data confirm that the Aristotle-Lipset hypothesis is a strong empirical regularity. In particular, increases in various measures of the standard of living forecast a gradual rise in democracy. In contrast, democracies that arise without prior economic development—sometimes because they are imposed by former colonial powers or by international agencies—tend not to last. Given the strength of this empirical regularity, one would think that clear-cut theoretical analyses ought also to be attainable. (This seems to be a case where the analysis works better in practice than in theory.)

Table 2 summarizes empirical findings on the determinants of democracy, as measured by the electoral-rights index. The system attempts to explain the level of this index for each of the roughly 100 countries at six points in time—1972, 1975, 1980, 1985, 1990, and 1995. The first part of the table uses a basic system that includes as explanatory variables an array of economic and social factors. The nine variables included are described in the table. The results in the right-hand column show the effect of each explanatory variable on the electoral-rights index when all of the other explanatory variables are held constant.

From the standpoint of the Aristotle-Lipset hypothesis, the key findings from the basic system are the positive and statistically significant effects on electoral rights from real per capita GDP and primary schooling. These results strongly confirm the idea that a

higher standard of living goes along with more democracy. Moreover, the effects are predictive, in the sense that, for a given current value of the electoral-rights index, higher values of per capita GDP and primary schooling predict that future values of the electoral-rights index will be greater.

The result shown for reliance on oil production also fits in with the spirit of the Aristotle-Lipset framework. The negative effect here means that GDP based primarily on natural resources—in this case, oil—does not provide the same pressure for democratization as that derived from the accumulation of physical and human capital.

Another variable with considerable explanatory power for democracy is the gap in years of education at the primary level between males and females. The negative effect here means that, if the education system treats males and females more equally, then it is more likely that electoral rights will be greater (which includes, in part, the extension of the voting franchise to women). The causation here is, however, not clear cut, because an expansion of electoral rights also makes it more likely that the school system will treat males and females more equally.

Although the simple relation (when per capita GDP, education, and the other explanatory variables are not held constant) between electoral rights and urbanization is positive, the effect shown for urbanization in Table 2 is negative. (Urbanization has a strong positive correlation with per capita GDP and schooling, but these variables are held constant for the findings presented in the table.) This result may indicate that it is easier to maintain dictatorial powers when the populace is concentrated in cities. However, as a demonstration of the weakness of the underlying theory, some political scientists have argued in reverse that the urban population is harder to suppress because of its better access to communications and transport.

Country size, measured by population, turns out not to be systematically related to democracy. On a theoretical level, the expected relation could go either way, depending on how the technology of suppression relates to the overall scale of operations, that is, to the size of the country.

Finally, the basic system shown in the table includes five-year and ten-year lags of the electoral-rights index—these effects are positive (and highly significant) and thereby indicate substantial inertia in changing democratic institutions in response to changes in the economic and social variables that are included in the system. In practice, the changes in democracy (as measured and also in reality) are often discrete, with either no change or a substantial shift occurring in a particular year. If, for example, per capita GDP or education rises, then an increase in democracy becomes more likely, and the probability of this increase occurring becomes greater as time passes. Quantitatively, the results imply that (on average) about 25% of the full adjustment of democracy to a change in an explanatory variable occurs over 5 years, and nearly 70% occurs over 20 years. Thus, after 20 or more years, the level of democracy is nearly fully determined by the economic and social variables—the kinds of influences stressed by Aristotle and Lipset—and very little by the country's longer term history of democracy. (There must be a Marxist flavor to these results, because they reflect a lot of economic determinism.)

Table 2 also shows some additional variables that were added, one at a time, to the basic system. One result is that years of educational attainment at the secondary and higher levels has no predictive power for the electoral-rights index. Hence, it appears to be early education that interacts with a country's propensity for democracy.

Greater income inequality, measured in some alternative ways, turns out to be negatively related to democracy—this finding was also predicted by Aristotle and Lipset. (The data on inequality are compiled by Deininger and Squire [1996].) The main difficulty here is that the limited availability and poor quality of the data on income distribution causes the sample size to contract substantially when an income-distribution variable is included in the system. It seems likely that, with more and better data, the negative impact of income inequality on democracy would show up more clearly in a statistical sense.

One important idea stressed by Milton Friedman and others is that political and economic freedoms are reinforcing. Our interpretation of this idea is that greater rule of law should predict more democracy in the future, and, similarly, greater democracy should predict more rule of law in the future. With respect to the former channel, the results shown in Table 2 fail to confirm any reliable effect of the rule-of-law index on the electoral-rights index. That is, given the current state of democracy and the values for the economic and social variables, a higher level for rule of law does not predict future increases in the extent of democracy.

Colonial heritage would be important for democracy if countries inherit a tendency for more or less electoral rights from their previous rulers. For example, Lipset has argued that British rule provided a crucial learning experience for subsequent democracy. It is true that former colonies are less likely than non-colonies to be democratic—the average of the electoral-rights index from 1975 to 1995 was 0.69 for 32 non-colonies and 0.46 for 106 former colonies. Moreover, within the colonies, electoral rights were higher, on average from 1975 to 1995, among former possessions of Britain and Spain than among those of France, Portugal, and other countries (in the present sample, Australia, Belgium, the Netherlands, New Zealand, and the United States).

However, the empirical results summarized in Table 2 fail to find these kinds of linkages. Given a country's economic and social conditions—as reflected in the explanatory variables shown in the table—differences in colonial history have no predictive content for democracy. Hence, if colonial status matters for democracy, then this influence must operate indirectly by influencing the pace of economic development. This indirect channel is worth exploring but has not yet been documented.

Religious affiliation has also been stressed by some political theorists, such as Samuel Huntington (1991), as an important determinant of democracy. This connection was assessed here by using data on professed religious affiliation among nine major groups: Catholic (including eastern orthodox), Protestant (including Anglican and some other Christian denominations), Hindu (including Jains and Sikhs), Buddhist, miscellaneous eastern religions (Chinese folk religions, Shinto, Confucianism, and new-religionists), Jewish, no professed religion (including atheists), and other religions (such as Parsis, Bahais, Spiritists, tribal religions, and indigenous third-world Christians). These data are from Barrett (1982).

The simple relation between religion and democracy is strong—if one classifies countries in accordance with their largest religious affiliation, then the averages for democracy from 1975 to 1995 were 0.9 for Jewish countries (1 country), 0.8 for Protestant (24), 0.7 for Hindu (5), 0.6 for Catholic (49) and Buddhist (4), 0.5 for miscellaneous eastern religions (3), 0.3 for other religions (17) and Muslim (32), and 0.1 for non-religion (1 country, China). However, the estimated effect of religion on democracy becomes much weaker once one holds fixed a country's economic and social variables, as in Table 2. In this context, the significant results are—using Catholic affiliation as a benchmark—negative effects from Muslim, non-religion, and other religion. The Muslim effect may reflect the strong connection between church and state in Muslim countries. The non-religion effect becomes less clear cut if China is eliminated from the sample.

Table 3 shows the results from the application of the same kind of analysis to the determination of the rule-of-law index. Because the data on the rule of law begin only the early 1980s, this system seeks to explain the rule-of-law index at only three points in time—1985, 1990, and 1995. Many of the results are similar to those shown for the electoral-rights index in Table 2. In particular, the Aristotle-Lipset hypothesis works also for the rule of law, in the sense of the significantly positive effects shown in Table 3 for per capita GDP and years of primary schooling. Also similar are the negative effect of urbanization, insignificant effect of country size, and insignificant effect from average years of schooling at the secondary and higher levels. Greater income inequality again has a negative impact, and this effect is statistically more reliable for the rule of rule than for democracy. Also as before, the positive effect from the lagged value of the rule-of-law index indicates the gradual adjustment of this institutional feature to changes in the economic and social variables.

Some differences in the results are that the gap between male and female primary schooling, the oil variable, and the array of religious affiliations have no influence on the rule-of-law index. Also, colonial heritage is much more important for the rule of law than for democracy. Even with the economic and social variables held constant, former colonies have significantly lower values of the rule-of-law index than do non-colonies. Within the colonies, the ordering of the size of the negative effect goes from former possessions of Portugal to those of the group of other countries, to those of Spain, to those of France, and to those of Britain. Hence, colonial powers seems to bequeath propensities for the rule of law more than they bequeath propensities for democracy.

Probably Britain and France were best at making these bequests because they themselves had better established legal systems than did Spain, Portugal, and some of the other colonial rulers.

Finally, Table 3 indicates that the electoral-rights index has no predictive content for the rule-of-law index. This result parallels the finding from Table 2 that the rule-of-law index does not predict the electoral-rights index. Putting the two results together, we conclude that—given economic and social conditions—the evolution of electoral rights and the rule of law are largely independent. That is why we found numerous instances of substantial gaps between the two indexes in Table 1.

Recall, however, that the electoral-rights and rule-of-law indexes exhibit a substantial positive correlation overall across countries and over time. We can explain this relation by noting that movements in the economic and social variables—for example, increases in per capita GDP and years of primary schooling—will tend to shift the two institutional measures in the same direction. Specifically, economic development tends to raise the levels of both indexes. Moreover, an independent improvement in the rule of law tends to stimulate economic growth and, therefore, to set in motion the kinds of increases in the economic variables that lead to expansions of electoral rights (as well as to further rises in the rule of law). Thus, the findings document this indirect channel whereby an expansion of the rule of law leads to an increase of democracy. The reverse channel is weaker because an expansion of democracy does not necessarily lead to a higher rate of economic growth.

Concluding observations

Perhaps the findings about democracy and the rule of law can best be summarized by relating them to U.S. foreign policy toward developing countries. The U.S. focus for many years has been toward promoting democracy—notably free elections with multiple parties—in all times and places. For example, when Haiti was run by a military dictator, the United States intervened to restore the previously elected president, despite his doubtful credentials. When President Fujimori of Peru disbanded the legislature and assumed dictatorial powers—ostensibly in a temporary way to counter a terrorist threat and to enact drastic economic and political reforms—the United States complained bitterly. When President Mobutu of Zaire was finally toppled by a revolution—after more than 30 years of mismanagement and corruption—the United States called immediately for the new leader, Kabila, to organize free elections. Moreover, China was continually attacked for its lack of democracy, although it had made major strides in enhancing the rule of law, whereas Russia was applauded for its free elections despite its difficulties in maintaining law and order.

The current U.S. Secretary of State, Madeleine Albright, was asked recently whether it was sometimes necessary to sacrifice democracy in the short run in order to promote economic growth. She replied to the effect that there was no such tradeoff because democracy was a prerequisite for economic growth. This response sounds pleasant but is simply false. The idea that democracy is necessary for growth is just as false as the proposition that dictatorship is essential for poor countries to escape poverty. The more nearly correct statement is that, on average, the extent of democracy has only a weak relation with subsequent economic performance.

When President Fujimori initiated his self-coup in 1992, the then U.S. Secretary of State, James Baker, said “you cannot destroy democracy in order to save it.” This proposition turned out to be false for Peru. Fujimori's curtailment of democracy led, in fact, to important economic and legal reforms, to success against a severe terrorist threat, and in a few years to a restoration of democracy. Apparently it is sometimes possible for dictators to accomplish things, including sustainable democracy, although the serious challenge is to predict the behavior of autocrats in advance.

The fact is that democracy is a tricky matter. It is desirable for its own sake, and it tends to arise more often and to a greater extent when countries become richer. History indicates that poor countries have difficulty sustaining democracy—when a democracy arises without prior economic development it tends not to last. Even if a poor country could beat the odds and sustain democratic institutions, there is no reason to believe that this accomplishment would help much in the quest to escape poverty. The movement from dictatorship to a moderate degree of democracy seems to contribute a little to economic performance, but this positive effect does not continue once a moderate degree of democracy has been attained. In this sense, a full representative democracy is something of a luxury good—it costs something but rich countries can well afford it.

For a country that starts with weak institutions—little democracy and little rule of law—an increase in democracy is less important than an expansion of the rule of law as a stimulus for economic growth and investment. In addition, democracy does not seem to have a strong role in fostering the rule of law. Thus, one cannot argue that democracy is critical for growth because democracy is a prerequisite for the rule of law.

The problem in the United States recommending democracy to a country such as Zaire (now the Republic of Congo) is not that democracy would harm economic performance but, rather, that it would have little impact. If there is a limited amount of energy that can be used to accomplish institutional reforms, then it is much better spent in a poor country by attempting to implement the rule of law—or, more generally, property rights and free markets. These institutional features are the ones that matter most for economic growth, and these features are not the same thing as democracy. Moreover, in the long run, the rule of law tends to generate sustainable democracy by first promoting economic development. Thus, even if democracy is the principal objective in the long run, the best way to accomplish it may be to encourage the rule of law in the short run. Perhaps U.S. advice to poor countries should focus more on the rule of law, property rights, and free markets, and less on the romance of democracy.

References

Aristotle (1932). Politics, translated by H. Rackham Cambridge MA, Harvard University

Press.

Barrett, David B., editor (1982). World Christian Encyclopedia, Oxford, Oxford

University Press.

Barro, Robert J. (1997). Determinants of Economic Growth: A Cross-Country

Empirical Study, Cambridge MA, MIT Press.

Bollen, Kenneth A. (1990). “Political Democracy: Conceptual and Measurement Traps,”

Studies in Comparative International Development, Spring, 7-24.

Deininger, Klaus and Lyn Squire (1996). “Measuring Income Inequality: A New Data-

Base,” unpublished, The World Bank.

Friedman, Milton (1962). Capitalism and Freedom, Chicago, University of Chicago

Press.

Gastil, Raymond D. (1982-83 and subsequent years). Freedom in the World, Westport

CT, Greenwood Press.

Huntington, Samuel P. (1991). The Third Wave: Democractization in the Late Twentieth

Century, Norman OK, University of Oklahoma Press.

Inkeles, Alex (1991). On Measuring Democracy, New Brunswick NJ, Transaction

Publishers.

Knack, Stephen and Philip Keefer (1995). “Institutions and Economic Performance:

Cross-Country Tests Using Alternative Institutional Measures,” Economics and

Politics, 7, 207-227.

Lipset, Seymour M. (1959). “Some Social Requisites of Democracy: Economic

Development and Political Legitimacy,” American Political Science Review, 53,

69-105.

Loayza, Norman V. (1996). “The Economics of the Informal Sector: A Simple Model

and Some Empirical Evidence from Latin America,” Carnegie-Rochester

Conference Series on Public Policy, 129-162.

Tocqueville, Alexis de (1835). Democracy in America, translated by Henry Reeve,

London, Saunders & Otley.

The Freedom House data cover more than 138 countries, but some countries were omitted here because of the lack of other historical data, especially on national accounts. The main countries left out are from the formerly centrally planned group (aside from China, Hungary, Poland, and Yugoslavia, which are included in the sample).

These data were introduced to economists by Knack and Keefer (1995). Two other consulting services that construct these type of data are BERI (Business Environmental Risk Intelligence) and Business International (now a part of the Economist Intelligence Unit).

The values shown for the rule-of-law index apply for each country in the early 1980s. Hence, this variable takes on only one of seven possible values, corresponding to the seven categories used by Political Risk Services.

In the framework used, this growth-rate effect persists for a long time. However, the magnitude of the effect diminishes slowly as the economy develops, and the influence in the very long run is on the level of economic activity, not its rate of growth.

The values shown for the electoral-rights index refer to averages over several years. Therefore, these values are not confined to the seven possible values that correspond to the categories used by Freedom House.

If one examines the residual errors in the equations for the electoral-rights and rule-of-law indexes, then the contemporaneous correlation of these errors turns out to be positive but small.

Robert J. Barro is a Senior Fellow at the Hoover Institution and the Robert C. Waggoner Professor of Economics at Harvard University.  His expertise is in the areas of macroeconomics, economic growth, and monetary theory. He recently completed a volume titled Determinants of Economic Growth: A Cross- Country Empirical Study (MIT Press, 1997).

[Previous][Contents] [Next]

info@fraserinstitute.ca
4th Floor, 1770 Burrard Street, Vancouver, BC, Canada, V6J 3G7
Tel: (604) 688-0221 Fax: (604) 688-8539 Book Orders: 1-800-665-3558 ext. 580

You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.