Economic freedom and conflict reduction: evidence from the 1970s, 1980s, and 1990s.
Proponents of the statist approach might note that a strong government can bend the market to its will, directing activity toward policies necessary to achieve greater levels of gross domestic product and growth. By extracting more resources for the economy, a powerful state can redistribute benefits to keep the populace happy. Higher taxes can also pay for an army and police force that intimidate people. Such governments range from command economies of totalitarian systems to autocratic dictators and military juntas. Other economically unfree systems include some of the authoritarian "Asian tigers."
A combination of historical evidence, modern theorists, and statistical findings, however, has indicated that a reduced role for the state in regulating economic transactions is associated with a decrease in internal conflicts. Countries where the government dominates the commercial realm experience an increase in the level of domestic violence. Scholars have traced the history of revolutions to explain the relationship between statism and internal upheavals. Contemporary authors also posit a relationship between economic liberty and peace. Statistical tests show a strong connection between economic freedom and conflict reduction during the past three decades.
Historical analysis of regime changes reveals that internal violence, whether it takes the form of a coup d'etat, revolution, rebellion, or war of liberation, occurs as a result of financial failures and administrative incompetence (Eckstein 1965; Kamenka 1970: 117). But as Leiden and Schmitt (1968: 42) point out, economic deterioration and decline alone do not produce a revolution. Analysis of the English, American, and French revolutions, in fact, indicates that those upheavals occurred largely because governments attenuated economic liberties (Davies 1970).
Current critics of economic statism find a connection between a government stranglehold on the economy and internal political violence. Amartya Sen (1996: 139) uncovers evidence that free countries do not experience famines. Unfree governments, on the other hand, have used famines to weaken opponents, as the Sudanese government of Lt. General Umar al-Bashir has done to the Christian Animists.
In addition to critiquing the ills of economically unfree governing systems, modern writers also praise the benefits of commercial freedom. Sen (1999: 16) contends that economic growth emerges more from a friendlier economic climate than a harsh political system. Barbara Crossette (1997: 200) finds a relationship between the economic growth attributable to globalization and peace. During times of peace, people are free to plant crops, develop crafts, and otherwise channel their energies into economic activities. She cites Stanford economist Gavin Wright, who also finds a connection between globalization (in the form of unregulated transfers of technology) and peace. Michael Klare (2001: 23) also argues that global market forces have encouraged the nonviolent resolution of conflicts. People are realizing that the economic benefits of compromise are greater than the likely costs of war.
To test the association between economic freedom and conflict, I begin by operationally defining those terms. To represent the dependent variable "conflict," I employ the KOSIMO data set from the Heidelberg Institute for International Conflict Research (1998-99). The Heidelberg Institute defines conflict as "the clashing of overlapping interests (positional differences) around national values and issues (independence, self-determination, borders and territory, access to or distribution of domestic or international power)." Moreover, "the conflict has to be of some duration or magnitude of at least two parties (states, groups of states, organizations or organized groups) that are determined to pursue their interests and win their case." The institute also points out that "at least one party is the organized state," and that "possible instruments used in the course of a conflict are negotiations, authoritative decisions, threat pressure, passive or active withdrawals, or the use of physical violence and war." I include all cases listed by the institute as either most severe crises or wars that involve only a single state as cases of internal conflict, and also analyze all KOSIMO cases of internal war separately.
The independent or explanatory variable is represented by data from the economic freedom of the world index (EFW), which reflects the size of government, price stability, trade openness, the quality of legal structures, and other variables (Gwartney, Lawson, and Samida 2000). The data set provides scores (on a scale of 0 through 10) for countries based on the ability of citizens to choose for themselves, engage in market activities, and keep what they earn. Higher EFW scores indicate more economic freedom, while lower numbers indicate increased state domination of commercial transactions. I break the EFW scores into three categories for ease of reporting the results: "Free" states have EFW scores ranging from 7 through 10; "Partly Free" countries have EFW scores as low as 5 and as high as 6.999; and "Not Free" nations have EFW scores at 4.999 or lower.
I examine EFW scores for a series of seven cross sections: 1970, 1975, 1980, 1985, 1990, 1995, and 1997. In matching those scores with the KOSIMO data, I examine whether an internal conflict (or an internal war) occurred in each cross section for each country. This approach generated 712 country cases or units of analysis.
Of the 712 country cases, 572 (or 80.3 percent) experienced no conflict (most severe crises or war). Of the remaining 140 country cases, where a country experienced one or more conflicts, 112 witnessed only a single internal conflict. The remaining 28 country cases had multiple instances of internal violence (one country case even had four in a single year). These 140 country cases accounted for 177 conflicts from all of the cross sections or years of analysis. When we limit the internal conflicts to wars, we find 667 country cases (or 93.7 percent) with no war. Therefore, 45 of the 140 conflict country cases involve the most severe type of hostilities. Of the 45 cases of war, only two cases had two conflicts, producing a total of 47 wars in the seven years of analysis. Breaking down the economic freedom classifications reveals that 198 of the 712 cases (or 27.8 percent) are economically Free. The modal category was Partly Free, with 283 of 712 cases (or 39.7 percent). The remaining 230 cases (or 32.3 percent of the total) received EFW scores indicating that they were Not Free.
In this section, I provide results of the test where the level of economic freedom is used to explain the level of internal conflict. I provide a comparison of the conflicts for each category, and I also compare the observations to an "expected model" randomly generated by the amount of conflict and economic freedom in the sample during the time frame.
Results show that Free states accounted for 17 country cases of at least one internal conflict, or only 12.1 percent of the 140 conflict cases. The 22 total conflicts generated by these country cases also produced only 12.4 percent of all 177 internal conflicts. Furthermore, the chance that any Free state in a country case will experience a most severe crisis or war is only 8.6 percent. How do those findings compare with data from Partly Free and Not Free economic systems? Partly Free countries suffered 57 country cases of one or more internal conflicts (resulting in a total of 77 conflicts). This means that Partly Free country cases make up 40.7 percent of all conflict cases ("country years" with a conflict), and 43.5 percent of all conflicts. This also indicates that the chance that a country case with a Partly Free economic administration will have at least one conflict in a given year is 20.1 percent, or 2.34 times greater than for a Free state. Country cases with Not Free states accounted for 66 internal conflict cases, or almost half (47.1 percent) of all 140 cases with at least one conflict. These 66 internal conflict country cases involving unfree governments produced a total of 78 conflicts. This means that such unfree states, which make up just less than a third of all country cases, generate 44.1 percent of all conflicts. Even more disturbing are the findings that show that the likelihood that a country case with an economically authoritarian state suffers internal violence is 28.6 percent, or 3.3 times greater than for a Free state. These percentages are summarized in Table 1.
In addition to comparing categories, we can compare observations for an expected model, which is created by the distribution of conflict and economic freedom in the sample. Expected numbers might indicate, for example, how many conflicts would exist if there were no connection between economic freedom and conflict. As Table 2 shows, there were 13 country cases where a Free state experienced a single conflict. The expected number indicates that if there were no relationship between the variables (or if the number of conflicts mirrored the amount of Free states in our sample), there would be 31.2 cases where a Free state had an internal war or mostly violent crisis. That allows us to state Free country cases had fewer than expected conflicts during the 1970s, 1980s and 1990s. (1)
Results from Table 2 show that Free country cases generally have more peace than expected, while also experiencing fewer conflicts than expected. There is little difference between observed country cases of Partly Free state conflicts and their expected numbers. But data from Not Free country cases show the opposite results. Economically unfree states not only have fewer country cases of peace than expected, but also more cases of internal conflict than expected. The chi-square statistic is 31.039 and is statistically significant at a high level. Indeed, there is less than a .001 percent chance of our results being in error, which means there is a high probability that economic freedom is connected to conflict reduction.
The preceding section focused on all forms of internal conflict. What would the results show if we restricted our analysis to wars, or the most severe forms of internal violence? I begin by comparing data from the Free, Partly Free, and Not Free country cases, including number and percentages of war cases. I also evaluate the chances each type of economic system faces an internal war. As with internal conflicts, I compare the results of observed internal wars with an expected model to determine whether extremely severe conflicts are more or less likely to occur, on average.
The findings of the distributions can be found in Table 3. Of all 198 economically Free country cases, only five (or 2.5 percent) suffered a domestic war. These Free country cases were only responsible for 11.1 percent of all wars of an internal nature. Country cases with Partly Free states, on the other hand, fought 19 internal wars (17 cases of a single war, and one case where a country fought two wars in a given year), constituting 40 percent of all wars (and war cases). Although only 6 percent of all Partly Free country cases experienced a war, the odds of such an event were more than twice as high as the odds that a Free country would go to war. Countries with little economic freedom (Not Free country cases) waged the most wars (23) and accounted for nearly half of all wars and war cases (where at least one war occurs in a given year). The chance that a Not Free country fought at least one war in a given year was 9.5 percent, which was almost four times more likely than for a Free country.
How do the observed cases of war compare with the expected model discussed in the previous section? The results are displayed in Table 4. As with internal conflicts, Free country cases had more peace than expected and less than half as many wars as expected. Country cases with Partly Free states experienced little deviation between the observations and the numbers generated by the expected model. Country cases with Not Free states, on the other hand, had fewer cases of peace than predicted, but higher levels of war than expected. Furthermore, though not as strong as the results from internal conflict, the chi-square statistic of 8.967 is statistically significant at the .062 level. This means that we can reject the null hypothesis that war is unrelated to the amount of economic freedom in a state with a 93.8 percent degree of confidence (that our results cannot be written off to error or chance).
Controlling for the Level of Development
The research presented is a bivariate analysis connecting economic freedom with levels of internal conflict. However, without controlling for additional factors, critics could charge that the relationship is attributable to other factors. They could claim that the results can be credited to the well-established relationship between economic freedom and income. According to this logic, developed countries are less likely to experience conflict because they have achieved a degree of wealth sufficient to satisfy their domestic population, and hypotheses concerning economic freedom are somewhat irrelevant. To test that argument, I examine whether the results hold for countries that have achieved a significant level of development (as opposed to societies with high incomes, but also high disparities between haves and have-nots) by controlling for membership in the Organization for Economic Cooperation and Development (OECD). These results, which are available in Tables 5 and 6, show relatively similar patterns for OECD countries: the most economically Free countries are less likely to experience an internal conflict or war, while other OECD country cases are slightly more likely to wage an internal conflict or war (especially Partly Free states). These results, which are also statistically significant, support the connection between economic freedom and lower levels of internal conflict among developed countries. The absence of such statistical significance would have aided critics' contentions that the results are attributable to level of development.
A separate analysis of non-OECD countries shows that they generally mimic their more developed counterparts. Economically Free non-OECD countries have less internal conflict than is expected from a random model, while Partly Free states have a mixed record. Not Free developing states, however, are less likely to experience peace and are more likely to engage in conflict.
Clearly, economic freedom is strongly connected with conflict reduction. Free states right fewer conflicts and wars in general. They also engage in such battles at levels well below their distribution in the sample, as well as levels below what an "expected model" predicts. Partly Free states had more domestic battles than their Free counterparts, and constituted sizable portions of conflicts and wars in the sample, yet these cases differed little from what our model predicted. Not Free states were the most prone to conflict and to wage war against their own people; they also had more cases of conflict and war than expected (while experiencing less peace than expected).
The major lesson from our study is that economic freedom promotes peace. Consequently, the more a government erodes economic liberties, the greater the likelihood that a severe crisis or war will erupt.
TABLE 1 ECONOMIC FREEDOM AND INTERNAL INTERNAL CONFLICT Economic % Chance of Freedom % of Internal % of Internal Experiencing an Levels Conflict Cases Conflicts Internal War Free 12.1 12.4 8.6 Partly Free 40.7 43.5 20.1 Not Free 47.1 44.1 28.6 TABLE 2 NUMBER OF INTERNAL CONFLICTS IN A GIVEN YEAR (1970, 1975, 1980, 1985, 1990, 1995, 1997) Economic Freedom Levels 0 1 2 Free O = 181 O = 13 O = 3 E = 159 E = 31.2 E = 5.6 Partly Free O = 226 O = 43 O = 9 E = 227.3 E = 44.6 E = 8 Not Free O = 165 O = 56 O = 8 E = 184.7 E = 36.2 E = 6.5 Totals 572 112 20 Economic Freedom Levels 3 4 Totals Free O = 1 O = 0 198 E = 1.9 E = 0.3 Partly Free O = 4 O = 1 283 E = 2.8 E = 0.4 Not Free O = 2 O = 0 231 E = 2.3 E = 0.3 Totals 7 1 712 Chi-square statistic: 31.039, Df = 8, X2 < .001 NOTES: O = observed, E = expected, Df = degrees of freedom (parameters), X2 = level of statistical significance. TABLE 3 ECONOMIC FREEDOM AND INTERNAL WAR Economic % Chance of Freedom % of Internal % of Internal Experiencing an Levels War Cases Wars Internal War Free 11.1 11.1 2.5 Partly Free 40.0 40.0 6.0 Not Free 48.9 48.9 9.5 TABLE 4 NUMBER OF INTERNAL WARS IN A GIVEN YEAR (1970, 1975, 1980, 1985, 1990, 1995, 1997) Economic Freedom Levels 0 1 2 Totals Free O = 193 O = 5 O = 0 198 E = 185.5 E = 12 E = 0.6 Partly Free O = 265 O = 17 O = 1 283 E = 265.1 E = 17.1 E = 0.8 Not Free O = 209 O = 21 O = 1 231 E = 216.4 E = 14 E = 0.6 Totals 667 43 2 712 Chi-square statistic: 8.967, Df = 4, X2 < .062 NOTES: O = observed, E = expected, Df = degrees of freedom (parameters), (X2) = level of statistical significance. TABLE 5 NUMBER OF INTERNAL CONFLICTS IN A GIVEN YEAR OECD COUNTRIES (1970, 1975, 1980, 1985, 1990, 1995, 1997) Economic Freedom Levels 0 1 Totals Free O = 108 O = 9 117 E = 103.7 E = 13.3 Partly Free O = 43 O = 10 53 E = 47 E = 6 Not Free O = 5 O = 1 6 E = 5.3 E = 0.7 Totals 156 20 176 Chi-square statistic: 4.696, Df = 2, X2 < .10 NOTES: O = observed, E: expected, Df = degrees of freedom (parameters), X2 = level of statistic significance. TABLE 6 NUMBER OF INTERNAL WARS IN A GIVEN YEAR OECD COUNTRIES (1970, 1975, 1980, 1985, 1990, 1995, 1997) Economic Freedom Levels 0 1 Totals Free O = 117 O = 0 117 E = 115 E = 2 Partly Free O = 50 O = 3 53 E = 52.1 E = 0.9 Not Free O = 6 O = 0 6 E = 5.9 E = 0.1 Totals 173 3 176 Chi-square statistic: 7.083, Df = 2, X2 , .05 NOTES: O = observed, E: expected, Df = degrees of freedom (parameters), X2 = level of statistic significance.
(1) If the chi-square statistic, which combines all the deviations between observed and expected numbers, is statistically significant, we reject the argument that no relationship exists among the variables.
Crossett, B. (1997) "A New Credo: Make Money, Not War." In J. M. Lindsay (ed.) Perspectives: Global Issues. Boulder, Colo.: Coursewise.
Davies, J. C. (1970) "Toward a Theory of Revolution." In G. A. Kelly and C. W. Brown (eds.) Struggles in the State: Sources and Patterns of World Revolution. New York: John Wiley.
Eckstein, H. (1965) "On the Etiology of Internal Wars." History and Theory 4: 133-63.
Gwartney, J.; Lawson, R.; and Samida, D. (2000) Economic Freedom of the World: 2000 Annual Report. Vancouver, B.C.: Fraser Institute.
Heidelberg Institute for International Conflict Research (1998-99) KOSIMO Manual. Heidelberg, Germany: HIIK.
Kamenka, E. (1970) "The Concepts of a Political Revolution." In G. A. Kelly and C. W. Brown (eds.) Struggles in the State: Sources and Patterns of World Revolution. New York: John Wiley.
Klare, M. T. (2001) Resource Wars: The New Landscape of Global Conflict. New York: Metropolitan Books.
Leiden, C., and Schmitt, K. M. (1968) The Politics of Violence: Revolution in the Modern World. Englewood Cliffs, N.J.: Prentice-Hall.
Sen, A. (1996)"Freedom Favors Development." New Perspectives Quarterly (Fall).
--(1999) Development as Freedom. New York: Anchor Books.
Cato Journal, Vol. 22, No. 3 (Winter 2003). Copyright [c] Cato Institute. All fights reserved.
John A. Tures is Assistant Professor of Political Science at LaGrange College.
|Printer friendly Cite/link Email Feedback|
|Author:||Tures, John A.|
|Publication:||The Cato Journal|
|Date:||Jan 1, 2003|
|Previous Article:||The 1992-96 Bulgarian trade data puzzle: a case of sanctions breaking?|
|Next Article:||Creative destruction and globalization.|