Abstract

While research literature on democracy has established close correlation between democracy and prosperity, we further investigate the causal relation between democracy and prosperity. The causal links are tested in both directions from democracy to prosperity and the opposite direction. Adopting events study method, we check to see if a democratic event always leads to higher economic growth rate. Using data of 169 countries from year 1960 to 2003, we find that democracy has conflicting impact on economic growth, and on average the effect is zero. This suggests that democracy does not lead to prosperity. We then check whether an income event leads to change in democracy level. We find increasing income leads to higher freedom level, and more stable democracy. This suggests prosperity is a cause for democracy.  Our quantitative analysis leads to some specific predictions on countries’ future change toward democracy.

Keywords: Democracy, prosperity, growth, freedom.

I. Introduction

The world has seen widespread of democracy in the last two decades of the 20^th century. From the collapse of Soviet Union, to free election in Taiwan and Indonesia, to the democratic transition in Chile in South America and Benin in Africa, a wave of countries have embraced democracy. By 2003, there are 121 electoral democracies out of total 192 world’s governments, with ratio at 63%. In 1987, the ratio was 40% (Karatnycky et al., 2003). Such undeniable trend reflects universal human desire for freedom, and popular acceptance of democracy as a fundamental value and form of governance. However, there are still 71 countries that do not have elective governments, and 49 of them have very repressive regimes. Among them, China stands out as a glaring example. Systematically denying its people the basic rights of free speech, press and association, China continues to be ranked as one of the worst totalitarian states. The government maintains its repression with brutal police force: Each year, China executes 70% of all those people put to death in the world while its population is only 20% of the world (Amnesty International, 2003). On the other hand, China has one the world’s fastest growing economies. In 2003, its growth rate was 9%, and it has maintained an average 8% growth rate throughout the last two decades. The paradox between China’s political and economic performance poses a serious question: Do we need to give up democracy in exchange of prosperity?  

A more comprehensive look at the world tells another picture. The world’s wealthiest nations are almost all democracies. These include US, Canada, Australia, Japan and most west European countries. On the contrary, the world’s poorest nations such as North Korea, Burma, and Sudan are mostly under authoritarian regimes. Even though China has high growth rate, its GDP per capita is only $1,000, which makes it still one of the poor nations. Based on such observation, traditional theory of democracy states that democracy leads to prosperity (Dahl, 1998). This statement is worth careful scrutiny, because we cannot jump from correlation to causality. Opposite theory has been proposed: The causal relationship may be in the other way. Instead being facilitated by democracy, prosperity creates condition for democracy. A prosperous economy may create demand for more protection of property rights, and rule of law. Thus it leads to a movement toward more democracy. The link between democracy and wealth we see today may come from a long history of development. Democracy may be just a by-product of such development. South Korea and Taiwan fit into this theory quite well.

How can we tell which direction is the causality? Or is there a causal relationship between them? Roll and Talbott (2003) made the first attempt to address this question by testing the causality from democracy to prosperity. They use the events study method to isolate the impact the democracy event. By comparing the average growth rate in the years after the democracy event to the years before, they want to see whether the event leads to higher growth rate. Using data from 44 countries that transited into democracy from 1959 to 1999, they show that the average growth rate is 1.9% for the 10 years after the event[1], but only 0.8% in the 10 years before[2]. By this they conclude that democracy is a cause for prosperity.

We use the same events study method that Roll and Talbott used to test two separate causal effects:

1. The impact of a democracy event on economic growth (which leads to prosperity)
2. The impact of reaching certain prosperity (measured by GDP per capita) on democracy status.

Each bullet point addresses the causal relation in a different direction. Therefore we extend Roll and Talbott’s work of testing causality in one direction to testing in both directions.

Our findings are different from Roll and Talbott’s. We find that, about half of the countries have higher growth rate after becoming democracy, while another half have lower growth rate. This suggests democracy has conflicting effects on growth, and the average effect for all countries is close to zero. Therefore democracy is not a cause for prosperity.

For the test of the opposite direction, we check countries’ democracy status (measured by freedom rating) after they reach certain income level (measured by GDP per capita). Our data is limited to those reaching certain income after 1972 because it is the year when freedom index rating is available. We check 6 different income levels: $1,000, $2,000, $3,000, $5,000, $10,000, and $20,000. These lead to 6 different event tests.

We find that 83% of non-democratic countries gained higher freedom in the 10 years after reaching $1,000. At $2,000 and $3,000, the ratio of non-democratic countries gaining higher freedom is close to 60%, which is not significant (relative to the previous 83%). But after reaching $5,000, 100% of non-democratic countries will gain higher freedom. This reveals significant impact of income level on a country’s freedom level. For countries that have already become a democracy before the reaching an income level, we check whether they will slip back into a non-democratic state. We find that after reaching $3,000, a democracy will never go back to be a non-democracy. This finding is both interesting and astonishing. In summary, we find a strong causal direction from prosperity to democracy.

Our results suggest that democracy does not cause prosperity. Instead, prosperity causes   democracy. This finding is a little surprising, but it is consistent with many casual observations of East Asian countries and some South American nations.

Even though democracy is not a cause for prosperity, it is not a hindrance for prosperity either. Our results do not support the argument that growth has to come at the price of freedom. A country can maintain high growth rate after transiting to democracy. In fact, Our analysis (see Figure 3) show that a country can grow faster than before, regardless of its initial income level. The belief that a low-income country has to maintain authoritarian rule in order to grow is false.

Democracy as a standalone value is worth pursuing. Combined with a sound economic strategy, democracy provides the basic freedom, justice and equality to people. It’s a safe guard against the tyranny of the state, which causes most human sufferings.

Our results also provide good news: today’s totalitarian countries will become free as their income grows. Another good news is that, today’s democratic countries of middle or high income will maintain their freedom regardless external events.

II. Related Literature and Methodology

There is a vast literature that try to establish the link between democracy and economic growth through regression (Barro, 1996; Dawson, 1998; Perotti, 1996; Aron, 2000).

Some of these studies find no correlation, while others find positive correlation. Tavares and Wacziarg (2001) find that, by taking into account of different channels through which democracy influence growth, the overall effect of democracy on growth is moderately negative. Roll and Talbott (2003) ran regressions on prosperity and political freedom. They show that there is clear positive correlation between political freedom and income level, when holding all other factors, such as economic freedom and regulation, constant.

1. Linear Regression and the Test of Causality

Regression method is useful in identifying correlations, but it cannot identify causality.

Of course one can run regression of variables at lagged periods on the independent variables to test the latter’s impact. This is the idea behind Granger Causality test. But so far, this has not been done in democracy and prosperity context. Furthermore, the Granger Causality test does not address the relationship revealed by events study method. Granger Causality test requires time-series data for both dependent and independent variables. It cannot address the impact of an event that happens at one point of time. Of course, we can probably use time-series data of freedom ratings to replace the one-point event. But Granger Causality test has another limitation: It assumes that the underlying relationship is linear. In events study, no such assumption is made.

Regression method used by most studies on democracy and growth (or prosperity) makes implicit assumption of linear relationship. In many cases, this seems a harmless assumption, if we are only concerned with a positive or negative sign of the coefficient. But if we want to make precise prediction based on the coefficients given by regression, we will run into problems. Based on regression results, Barro (1996) made forecast for 36 countries’ change of democracy level from 1994 to year 2000.  It turns out, for the 20 countries that were projected to be more democratic, 10 of them actually became less democratic or the same as before[3]. For the 16 countries that are projected to be less democratic, 9 of them turned out to be more democratic[4]. Therefore the derived regression model gives us wrong prediction 50% of the time. Does this suggest the model is not complete enough, meaning we should throw in more variables, or does it mean the underlying linear relationship in fact does not exist at all?

Another problem of using regression on democracy study is the discrete nature of the variables. For example, even though GDP per capital is a continuous number, it does influence the freedom level in a continuous fashion. The impact of change in income on democracy in the [$10,000, $20,000] range may be very different from the change in [$0, $10,000] range. This difference can be characterized as non-linear relationship, but can also be characterized as treating a continuous variable as discrete one. Regression results smooth away the discrete effects, and tend to obscure the true relationship. 

Of course, regression provides us with a nice way of quantifying results and deriving a functional (linear) relationship. In many cases, it allows us to analyze impact of multiple factors at the same time. With regression equation, we can incorporate many variables at once and test the impact of one variable by holding all others constant at the same time.

2. Definition of Democracy and Other Terms

There is a general consensus among researchers that any definition of democracy should include the basic elements of free election, free speech, free press and association. But in practice there are many variations of such freedom. For example, election may not be fair or clean in some countries, and some countries do not have a real oppositional party. Do we consider them as democracy or not?

Therefore some people propose that there is continuum spectrum of democracy (Diamond, 1999). Some democracy is “mature democracy”, while others are “developing democracy” or “pseudo-democracy”. The best quantitative measure of democracy is provided by Freedom House Annual Survey (Karatnycky et al., 2003). This survey ranks a country’s freedom in terms of the following criteria: electoral process (free and fair), political pluralism (with oppositional party), accountability of government, freedom of expression, freedom of association, rule of law, and individual rights. Countries are given freedom index ranging from 1 to 7, where 1 represents the most free, and 7 represents the least free (or most repressive).

If we have to choose a dichotomous criterion for democracy, then 3 is a cutoff point. Perotti (1996) show that the freedom index less than or equal to 3 is equivalent to[5] the “democracy” category by Jodice and Taylor (1988), which assign 1 to a democracy, .5 to a semidemocracy, and 0 to dictatorships.

In this paper, we will use the term democracy as a dichotomous term. Therefore a country is a democracy if its freedom index is less than or equal to 3.

Freedom House has separate indexes (both ranging from 1 to 7) for political rights and civil liberties. The freedom index we refer to is the average of index of political rights and civil liberties. We will also use the word rating interchangeably with index.

Here is the list of definitions of terms in this paper:

A democracy:

A country with political freedom rating <= 3.0

A stable democracy:

A country that has had political freedom rating <=3.0 for at least 10 years and has never had rating >3.0 since then (until 2003).

A democracy event:

The year when a country starts the transition to a stable democracy. This requires

checking this country’s political rating in the next 10 years and until 2003.

Slip back into non-democracy:

After being a democracy for at least a year, a country’s freedom rating rises above

3.0.

Income level of a country:

GDP per capita (constant 1995 $)

Reaching an income level X:

When GDP per capital is >= X for the first time in the country’s history. Some country’s GDP level falls later and reaches back to level X again. We will always use the year when for the first time X is reached.

For simplicity, in some cases we will use the term GDP interchangeably with GDP per capita.

3. Empirical scope of this paper and the related papers

The Freedom Index data are available from 1972 to 2003. This is the range that we will check for democracy event. Our GDP per capita and growth data are all from World Bank, and the data is available from 1960 to 2002.

We adopt events study method used by Roll and Talbott (2003). In testing the causality from democracy to prosperity, our conclusions are different from Roll and Talbott’s. This may be due to different definitions and datasets, as shown in Table 1.

Table 1. Difference between Roll and Talbott (2003) and this paper

Our finding this paper is consistent with findings by Huntington (1995) and Przeworski et al. (1996). We find that a democracy will never go back to be non-democracy when reaching above income level $3,000 per capital GDP (constant 1995 $). Similarly, Huntington finds a “ceiling” for successful coups at the income level of about $1,000 per capita GNP. Przeworski et al. find that democracy to be “impregnable” at $6,000 or above, which is measured by PPP adjusted GNP (1985 $). The difference in GDP and GNP, and the difference between constant dollar and PPP adjusted dollar are probably the quantitative difference here.

III. The Correlation between Democracy and Prosperity

Before investigating the causal link, we want to show that there exists a link between democracy and economic wealth.

Figure 1 shows 175 countries with GDP per capita plotted against political freedom. The graph reveals a very clear pattern: All the countries that have income above $30,000 are very free, whose freedom indexes are equal to 1 or 1.5.  Between $10,000 and $30,000, most countries have freedom index between 1 and 3, and only a few (all of which turn out to be oil export countries) have index higher than 3.

Figure 1. The Correlation between Freedom and Economic Wealth

Data source: Freedom House 2003 survey (for year 2002)

                                      World Bank Development Indicator of year 2002

                 Total number of countries: 175 (including the 6 oil export nations listed in Table 3)

We then remove the 6 outliers in the data (all oil export nations), and investigate the correlation between income and freedom level for the rest 169 countries. We break all countries’ income level into 6 brackets, and calculate the ratio of democracies inside each bracket. Table 2 shows the results.

Table 2: Income level and Political Freedom rating around the world[6]

     Note: Total number of countries are 169 (excluding the 6 oil export nations in Table 3)

What we discover is some surprisingly consistent patterns:

1. The higher an income groups is, the higher percentage of its members are completely free (rating=1).
2. All the repressive regimes (defined as having freedom index >=5.5) have per capita GDP lower than $3,000. .

These two trends are solidly established considering the fact that Singapore is in the dataset. Singapore, whose GDP is $27,000, is the only country what has freedom index worse than 3 in all countries above $10,000.

We list the 6 oil export nations in Table 3. These countries (5 in Middle East and 1 in Southeast Asia) do not fall into the above pattern at all. Deriving their high income from oil export, these countries manage to maintain very authoritarian regimes. These countries do not have to build economic infrastructure or sustainable economic growth. Since the government does not derive its major revenue from taxes of normal business, it does not have to bow to pressure from business group or middle class that demand change. This is so-called “no taxation, no representation” problem, and it is clearly reflected here in Table 3, where the link between democracy and income is cut off.

In this paper, we focus our attention on the 169 countries in Table 2. The astonishingly consistent pattern in Table 2 suggests close link between democracies and economic prosperity.

Table 3. Oil Export Nations and their political rating

While the general pattern for all income levels are clear, it is not clear in the low-income level. Here we plot out (see Figure 2) the relation between freedom rating and income level for countries that have less than $3,000 per capita GDP. A regression on this group would yield very weak link between democracy and economic prosperity.

                        Figure 2. GDP and Political Freedom of Low Income countries

In general, we can summarize that all the prosperous countries are democracies, but not all democracies are prosperous. Does this suggest some causal link between democracy and prosperity? But which way is the causality? We start by testing causality from democracy to prosperity.

IV. Causality: From Democracy to Prosperity

We want to find out whether democracy is a cause for prosperity. Using the events study method, we check for a country’s growth rate after becoming democracy (reaching freedom index <=3.0). Note that we require a country stays a democracy (freedom index <=3) for at least 10 years until the current year (2003). In other words, the countries we select are stable democracies. We compare the average growth rate in the 10 years after becoming a democracy to the 10 years before. Below is our finding.

Since we require a country to move from non-democracy to democracy, its freedom rating has to be greater than 3 in the year before. Our freedom index data starts from 1972, so we only choose countries that became a democracy at or after 1973. We also ignore countries that are not democracies by 2003. A stable democracy has to stay a democracy until today. Since we need 10 years (or 8) average data, we ignore countries that become democracy after 1994. Our sample country has to reach 3.0 or lower by year 1994.

We end up with 28 countries, which are listed in Appendix Table A2. A general plot of these countries’ income level and freedom index is shown in Figure 3.

Figure 3. Change in Growth rate after becoming democracy

Figure 3 suggests that transition to democracy happens at all income level. However, a country may experience higher or lower growth rate than before the transition. There is no visible trend toward higher growth. The overall effect of initial income level on growth rate is almost zero.

We further calculate the ratio of countries that experience higher or lower growth after becoming democracy.

Our table shows that democracy has conflicting impact on economic growth. For a little less than 50% of countries, it has positive impact; but for the other 50% of countries, it has negative impact. Overall, the average effect is zero.

Table 3. Ratio of countries with increasing or decreasing growth rate

        Note: The Raw data is available in Appendix Table A2.

In terms of absolute performance, 40% of the countries have growth rate higher than 2%, while 60% have growth rate below that threshold. The number 2% is used because it is close to world average, and it is the minimum growth rate for a developing country to catch up with wealthier nations. The ratio here does not mean much. It is possible that a low-growth economy still has low growth rate after becoming democracy. But the ratio does tell us that democracy does not lead to high absolute performance either.

Therefore both in terms of relative impact or absolute impact, democracy has no effect on economic growth.

Our results are a little surprising. What we discover is that democracy has little impact on economic growth. This contradicts conventional theory that democracy leads to economic prosperity. The convention theory is that democracy would promote rule of law, and the protection of property rights. But democracy also has its impact of income redistribution, policy that may harm economic growth.

V. Causality: From Prosperity to Democracy

If democracy has no impact on economic prosperity, then why do we observe the close link between national wealth and democracy level? We hypothesize that economic growth and the resulting economic wealth, will create demand for better judicial system, protection of property rights. The social interaction becomes more sophisticated. The speed of economic activities also requires more government transparency and lower regulation. Overall, there is increasing demand for a freer and fairer society, where the rights of average citizens are protected. In other words, our theory is that economic prosperity will ultimately lead to democracy.

We use the same “event” driven method to test this hypothesis. The event here is reaching a certain GDP per capita level. We check a country’s freedom index before and after reaching an income level. We investigated 6 different events corresponding to 6 different income levels: $1,000, $2,000, $3,000, $5,000, $10,000, and $20,000. For each event, we separate two groups, the group that were democratic before the event, and the group that were non-democratic before the event. The results for the non-democratic group is listed in Table 4.

Table 4. Change in freedom for countries that were originally not democratic.

              (at the event year)

   Note: A country is a democracy if its freedom rating is <=3.

             A country is a non-democracy if its freedom rating >3,

                Freedom status is measured by average rating of the next 10 years

Table 4 shows that a non-democratic country will always move toward more freedom after its income reaches $5,000. At all income level (starting from $1,000), the likelihood for a non-democratic nation to become freer is higher than becoming less free. This reveals a tremendous force toward more freedom in non-free countries. Moreover, starting from $3,000, a country will have higher probability of becoming freer as its income grows.  

For countries that are already democracies (when they reach certain income level), the movement toward more freedom is a little different from above. As shown in Table 5,

a democracy becomes stabilized after it reaches income level $3,000. After this level, no country that is a democracy will ever go back to be a non-democracy. Moreover, starting from $1,000, a democracy will more likely to be freer as its income grows. It reaches 100% at $10,000, then gets down to about 88% at $20,000.  This reflects the consolidation effect of income on democracy.

Overall, we see a significant impact of income level on a country’s freedom rating. Particularly, the income level $3,000 serves an important threshold. Above this income level, a democracy will consolidate, and a non-democracy will experience higher probability of freedom as its income grows. Once crossing $5,000, a non-democracy will definitely become freer over time. Therefore we conclude that there exists a causal direction from economic prosperity to democracy.

Table 5. Change in freedom for countries that were originally democracy

                         (at the event year)

VI. Prediction from this research and Policy Implication

The general prediction from our research is that: today’s totalitarian countries will become freer as their income grows, and today’s democratic countries of middle or high income will maintain their freedom regardless external events.

Our findings can also be used to make specific prediction for a country’s change of freedom. The prediction will be based on Table 4 and 5, a country’s current freedom rating and its future GDP per capita. For example, China’s GDP per capita just reached over a little over $1,000 in 2003. Based on our analysis, China will very likely become freer in the next few years (with more than 80% probability), but not likely to become a democracy soon. With 8% growth rate, China will reach $2,000 in 9 years, $3,000 in 15 years, and $5,000 in 21 years from today. Given the favorable condition for democracy around the world today, China will likely transit to a democracy in 21 years from now (with more than 50% probability). 

Our results do not diminish the value of democracy. We show that even though democracy is not a cause for prosperity, it is also not a hindrance either. As a standalone value, democracy or political freedom is always an aspiration for human society. But as a tool for economic development, our research showed that democracy has to be combined with other economic factors to create high growth.

Our analysis also reveals the unique pattern of 6 oil export nations. These six countries drive more than 30% of GDP from oil exports, and were able to maintain very repressive regime even at high national income. They do not follow the same pattern as the other countries. The unique pattern of these 6 countries suggests a disturbing fact: the world’s reliance on oil creates governments that resist change. An oil rich country has little incentive to develop its own economy, which could have created its middle class and demand for democracy. The road to democracy for the oil export nations presents a special daunting task.

VII. Conclusion

We investigated the causality between democracy and economic prosperity. We tested the causal relation in both directions.

Adopting an event driven method, we test a country’s economic performance before and after it becomes a democracy. We find that half of the new democracies enjoy higher growth, while another half has lower growth rate. In terms of absolute performance, we find half of the countries grow above world average (3%), while the other half grow very slow or even negatively. In summary, democracy is neutral in its impact on economic growth. We conclude that democracy does not lead to economic prosperity.

We then test the causal relation in the opposite direction. Treating reaching an income level as an event, we test a country’s democracy level (freedom rating) before and after the event. We find a democracy becomes stabilized after it reaches income level (GDP per capita) $3,000. After this level, no country that is a democracy will ever go back to be a non-democracy. Furthermore, after reaching $3,000, a non-democracy will have higher probability of becoming freer as its income grows. Once crossing $5,000, a non-democracy will definitely become freer over time. We conclude that economic prosperity leads to consolidation and movement toward democracy.

Our findings are a little surprising. Based on conventional political theory, democracy creates favorable condition for economic growth. Instead, we find that democracy has both positive and negative effect on economic growth. It is up to the policy maker to choose a sound economic policy that eventually leads to growth. Many people observed that economic growth brought democracy change. Our paper established empirical support for that argument. We showed a clear causal link from economic prosperity to democracy.

For future work in this subject, one thing is to expand the data range. The freedom rating data starts after 1972. If we have freedom ratings for earlier years, say from 1952 to 1972, we can enlarge the data sets and draw more statistically conclusive results. 

We also did not study the income effect on democracy below $1,000. Is there a threshold below $1,000 that links income to democracy? We don’t know the answer. It is an interesting question to investigate because there are about 70 countries still in this range.

Finally, the study of causality between democracy and prosperity only reveals favorable condition for democracy and its consolidation. It does not tell us the initial rising of democracy. For example, what is the trigger event? Given that democracy can happen in both low income or high income countries, we can help to initiate this process in all countries under repressive regimes today. In the meantime, a sound economic policy has to be pursued. A healthy economy not only improves people’s material welfare, but also their likelihood of living in democracy, which provides a free and just society.

References;

Alesina, Alberto and Ozler, Sule and Roubini, Nouriel and Swagel, Phillip, “Political

Instability and Economic Growth”, Journal of Economic Growth 1, 2: 189-211

(June, 1996)

Amnesty International, 2003, Annual Report

Aron, Janine. 2000. “Growth and Institutions: A Review of the Evidence.” The World

Bank Research Observer 15, 1 (February) 99-135

Banks, A. (1975-2002), A Political Handbook of the World. Binghamton,

NY: CSA Publications.

Barro, Robert J. 1996. “Democracy and Growth.” Journal of Economic Growth 1(1), 1-

27

Robert A. Dahl, On Democracy, 1998, by Yale University Press, Chapter 5.

Diamond, Larry and Byung-Kook Kim, editor, Consolidating democracy in South Korea,

Boulder : Lynne Rienner Publishers, 2000

Diamond, Larry, Developing Democracy: Toward Consolidation, 1999, John Hopkins

University Press

Dawson, John W. 1998. “Institution, Investment and Growth: New Cross-country and

Panel Data Evidence.” Economic Inquiry 36, 4, 603-619

Gwartney, James and Lawson, Robert, Economic Freedom in the World: 2003 Annual

Report, The Fraser Institute, 2003

Huntington, Samuel P., “Reforming Civil-Military Relations,” Journal of Democracy

6(4), 1995: 9-17.

Karatnycky, Adrian and Piano, Aili and Puddington, Arch, Editors, Freedom in the

World 2003: The Annual Survey of Political Rights and Civil Liberties, Rowman

& Littlefield Publishers, 2003

Taylor, Charles Lewis and Jodice, David A., 1983, World Handbook of Social and

Political Indicators, New Haven: Yale University Press

Perotti, Robert, “Growth, Income Distribution, and Democracy: What the Data Say”,

Journal of Economic Growth 1, 2, 149-187, June 1996

Przeworski, Adam and Michael Alvarez and Jose A. Cheibub and Fernando Limongi,

“What makes Democracies Endure?” Journal of Democracy 7 (1), 41-42

Robinson, Thomas W., editor, Democracy and development in East Asia : Taiwan, South

Korea, and the Philippines Washington, D.C. : AEI Press, 1991

Richard Roll and John R. Talbott, “Political Freedom, Economic Liberty, and Prosperity”,

Journal of Democracy, Volume 14, Number 3, July 2003

Full paper at: www.anderson.ucla.edu/acad_unit/finance/wp/2001/19-01.pdf

Tavares, José and Wacziarg, Romain, “How democracy affects growth”, European

Economic Review, Volume 45, Issue 8, August 2001, Pages 1341-1378.

Temple, Jonathan. 1999. “The New Growth Evidence.” Journal of Economic Literature

37(1): 112-156

World Bank, World Development Indicators 2003

 Appendix A. Political Freedom Checklist used by Freedom House

                                    (see Karatnycky et al.)