Making the World Safe for Partial ? Making the World Safe Arthur A. Goldsmith for Partial Democracy? Questioning the Premises of Democracy Promotion

Should the promote democracy abroad? Will moves toward democracy, particularly in the Muslim world, advance U.S. interests? To what extent can the United States bring about democratic change? What are the odds of success, and how great are the risks should democracy promotion fail? The answers to these questions are vital to U.S. foreign policy. Even after the debacle in , there remains broad agreement in Washington that the United States should work to foster a less autocratic world, which, in turn, is predicted to lead to less anti- Americanism and diminished threats to U.S. interests. This article calls attention to two errors in reasoning and evidence that al- most everyone in the debate over democracy promotion seems to have over- looked. First, if democracy enhances international security, that does not necessarily mean that “every step toward makes our country safer,” to quote President George W. Bush.1 Frequently, the suc- cessor regime to a is a partial democracy, which can pose an even greater security threat. Second, having the laudable purpose of furthering democracy is not a relevant reason for claiming that this goal is attainable. De- spite its signiªcant inºuence, the United States cannot consistently shape for- eign political systems to its liking, particularly in the short term. Democracy promotion’s limitations were brushed aside in the Bush admin- istration’s “forward strategy of freedom” or “freedom agenda,” which became the cornerstone of its foreign policy. The president prominently justiªed ad- ministration plans using strident neoconservative themes, asserting repeatedly that democracy promotion is both a normative prerogative and a pragmatic means to bolster the United States’ security and further its geopolitical preemi- nence.2 As he summarized these arguments in his 2006 State of the Union

Arthur A. Goldsmith is Professor of Management at the University of Massachusetts Boston.

1. George W. Bush, “President Bush Delivers State of the Union Address” (Washington, D.C.: White House, January 31, 2006), http://www.whitehouse.gov/news/releases/2006/01/20060131- 10.html. 2. Neoconservatives had long been conªdently advocating forceful steps by the United States to spread democracy around the world. See, for example, Joshua Muravchik, Exporting Democracy: Fulªlling America’s Destiny (Washington, D.C.: AEI Press, 1991); and the essays by William Bennett and in and William Kristol, eds., Present Dangers: Crisis and Opportu- nity in American Foreign and Defense Policy (New York: Encounter Books, 2000). On the inºuence of neoconservative thought on contemporary U.S. foreign policy, see Piki Ish-Shalom, “‘The Civiliza- tion of Clashes’: Misapplying the Democratic Peace in the Middle East,” Political Science Quarterly,

International Security, Vol. 33, No. 2 (Fall 2008), pp. 120–147 © 2008 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.

120

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 121

Address, one of the nation’s “deªning moral commitments” is to end tyranny around the world and replace it with democracy. Regime change is not solely a question of altruism, the president avowed, but also of national self-interest: “ replace resentment with hope, respect the rights of their citizens and their neighbors, and join the ªght against terror.”3 The United States spent billions of dollars in and Iraq to secure limited constitutional government in those countries. In addition, federal funding for other overseas democracy promotion activities jumped, starting in 2000, when it was about $500 million per year. The 2008 budget request raised foreign aid spending for democracy and human rights to nearly $1.5 billion, excluding Afghanistan and Iraq.4 The Bush administration initiated high- proªle efforts to improve public diplomacy toward areas with large Muslim populations, and engaged in pro-democratic lobbying of some of their leaders. In 2002 it launched the Middle East Partnership Initiative to support non- governmental organizations and government agencies with activities leading to democratic change in the Middle East.5 The freedom agenda never delivered. Five years later, the prospects for nonauthoritarian order in Afghanistan and Iraq seem more remote than ever. Competitive in Palestine, , , and other places have produced troublesome results for the United States. Human Rights Watch and both report that democracy is in retreat globally.6 Commenta- tors from across the political spectrum agree that the U.S. approach to democ- racy promotion since the September 11, 2001, terrorist attacks is in tatters, though they offer different diagnoses for what went wrong.7

Vol. 122, No. 4 (Winter 2007–08), pp. 533–554; Fred Kaplan, Daydream Believers: How a Few Grand Ideas Wrecked American Power (New York: Wiley, 2008); Jacob Heilbrunn, They Knew They Were Right: The Rise of the Neocons (New York: Doubleday, 2008); and Francis Fukuyama, America at the Crossroads: Democracy, Power, and the Neoconservative Legacy (New Haven, Conn.: Yale University Press, 2006). 3. Bush, “President Bush Delivers State of the Union Address.” 4. Thomas O. Melia, “The Democracy Bureaucracy: The Infrastructure of American Democracy Promotion,” Discussion Paper (Princeton, N.J.: Princeton Project on National Security, Princeton University, 2005), p. 14; and Susan B. Epstein, Nina M. Seraªno, and Francis T. Miko, Democracy Promotion: Cornerstone of U.S. Foreign Policy? CRS Report for Congress (Washington, D.C.: Library of Congress, December 26, 2007), Order Code RL34296, p. 21. 5. The practical steps taken by the United States, of course, never rose to the level implied by ad- ministration rhetoric. Thomas Carothers is one of many critics to note the inconsistencies in Bush’s policies—for example, the compromises made for certain authoritarian countries, such as Paki- stan, , and , that cooperate in the war on terrorism. See Carothers, “U.S. Democracy Promotion During and After Bush,” Carnegie Endowment Report (Washington, D.C.: Carnegie Endowment for Peace, September 2007), pp. 6–10. 6. Human Rights Watch, World Report, 2008 (New York: Human Rights Watch, 2008); and Freedom House, Freedom in the World, 2008 (Washington, D.C.: Freedom House, 2008). 7. See, for example, Steven Heydemann, “Upgrading Authoritarianism in the Arab World,” Saban

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 122

Yet bipartisan support lives on for the ends (and many of the means) of the freedom agenda. Liberal internationalism probably has as much inºuence on the Democratic Party as has on the Republicans.8 It was not so long ago, for example, that President Bill Clinton declared that U.S. foreign policy would be guided by the doctrine of “democratic enlargement,” aimed at expanding the community of democratic states.9 More recently, both Senators Barack Obama and John McCain were sponsors of the ADVANCE Democracy Act of 2005. Had it passed, the bill would have increased assistance to foreign countries to carry out democratic reforms of government.10 Public opinion has become more skeptical about democracy promotion, but remains favorable. A survey supported by the German Marshall Fund in 2007 found that about one- third of Democrats (and half of Republicans) agreed when asked whether the United States should help establish democracy in other countries.11 Americans are not alone in wanting to supply the beneªts of democracy to

Center Analysis Paper, No. 13 (Washington, D.C.: Brookings Institution, October 2007); Christo- pher J. Coyne and Tamara Cofman Wittes, “Can We Export Democracy?” Cato Policy Report, Vol. 30, No. 1 (January/February 2008), pp. 11–13; and Francis Fukuyama and Michael McFaul, “Should Democracy Be Promoted or Demoted?” Washington Quarterly, Vol. 31, No. 1 (Winter 2007– 08), pp. 23–45. 8. Michael C. Desch discusses this bipartisan tradition in Desch, “America’s Liberal Illiberalism: The Ideological Origins of Overreaction in U.S. Foreign Policy,” International Security, Vol. 32, No. 3 (Winter 2007/08), pp. 7–43. Charles A. Kupchan and Peter L. Trubowitz dispute whether the inter- nationalist consensus will long survive the Bush administration. See Kupchan and Trubowitz, “Dead Center: The Demise of Liberal Internationalism in the United States,” International Security, Vol. 32, No. 2 (Fall 2007), pp. 7–44. 9. S. 516 [109th]: ADVANCE Democracy Act of 2005, http://www.govtrack.us/congress/bill .xpd?bill=s109-516. 10. Making the world safe for democracy has been an undercurrent in U.S. foreign policy since the nineteenth century. See Jonathan Monten, “The Roots of the : Power, Nationalism, and Democracy Promotion in U.S. Strategy,” International Security, Vol. 29, No. 4 (Spring 2005), pp. 112–156. For additional discussion of democracy promotion in U.S. foreign policy, see Michael Cox, G. John Ikenberry, and Takashi Inoguchi, eds., American Democracy Promotion: Impulses, Strat- egies, and Impacts (New York: Oxford University Press, 2000); Thomas Carothers, Aiding Democracy Abroad: The Learning Curve (Washington, D.C.: Carnegie Endowment for International Peace, 1999); Peter J. Schraeder, ed., Exporting Democracy: Rhetoric vs. Reality (Boulder, Colo.: Lynne Rienner, 2002); and Tony Smith, America’s Mission: The United States and the Worldwide Struggle for Democracy in the Twentieth Century (Princeton, N.J.: Princeton University Press, 1994). 11. Transatlantic Trends: Key Findings (Washington, D.C., and Torino, : German Marshall Fund and Compagnia di San Paolo, 2007), http://www.transatlantictrends.org/trends/doc/ Transatlantic%20Trends_all_0920.pdf. Another poll, sponsored by the Program on International Policy Attitudes (PIPA), ªnds that a large majority favors the United States helping emerging de- mocracies with aid and technical assistance in conducting elections. But two-thirds of respondents reject the idea of threatening countries with military force if they fail to adopt democratic reforms, and most think that overthrowing a dictator to set up a democracy will do more harm than good. See “Americans on Promoting Democracy” (Chicago and College Park, Md.: Chicago Coun- cil on Foreign Relations and PIPA, September 29, 2005), http://65.109.167.118/pipa/pdf/sep05/ Democratization_Sep09_rpt_revised.pdf.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 123

those who lack it. The Millennium Summit in 2000, for example, pledged United Nations members to spread democracy in the developing world. The European Union’s European Neighborhood Policy endeavors to encourage de- mocracy and human rights in nearby Arab and central European states. In the private sector in Europe, the Soros and Bertelsmann Foundations are working to expand open governance around the globe. On the demand side, grassroots desire for more participation and account- ability in politics is strong in the Muslim world and other postcolonial regions. Human rights groups are proliferating in countries such as and . Public opinion surveys in Muslim societies show broad domestic support for the establishment of democratic systems in their countries. Democ- racy is a contested concept, and what respondents have in mind is not exactly what Americans or Europeans mean.12 Still, that different groups deªne de- mocracy in different ways does not diminish the global appeal of democratic ideals. Thus, despite the setbacks experienced by the Bush administration, democ- racy promotion (stripped perhaps of President Bush’s unilateralism) will re- main on the international agenda, for at least the following reasons: (1) the United States has a hundred-year history of attempting to promote democracy; (2) neoconservatives and liberal internationalists concur that democratization should remain an important consideration in U.S. foreign policy; (3) interna- tional organizations and U.S. allies also support democracy promotion; and (4) Muslims (not to mention citizens of developing nations generally) are calling for their societies to become more democratic. The encouragement of democracy will be ineffective, however, if its propo- nents do not examine the implicit assumptions made about the beneªts and feasibility of democratic regime change, and tailor their strategies accordingly. This article tests a few of the key premises about pro-democracy policy against ªndings from the latest large-n studies in international relations and compara- tive politics. I augment the literature review with a new compilation of data on regime transitions in Muslim countries from 1970 to 2002. I show how the con- ventional wisdom narrows the likely ends and exaggerates the available means of generalized democracy promotion.

12. Pew Research Center, “Iraqi Vote Mirrors Desire for Democracy in Muslim World,” Pew Global Attitudes Project Commentary (Washington, D.C.: Pew Research Center, February 3, 2005); WorldPublicOpinion.org, “Muslim Public Opinion on U.S. Policy, Attacks on Civilians and al Qaeda” (College Park: Program on International Policy Attitudes, University of Maryland, April 24, 2007). See also Salwa Ismail, “Democracy in Contemporary Arab Discourse,” in Rex Brynen, Bahgat Korany, and Paul Noble, eds., Political Liberalization and Democratization in the Arab World, Vol. 1: Theoretical Perspectives (Boulder, Colo.: Lynne Rienner, 1995).

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 124

The argument proceeds in four steps. The next section demonstrates the widespread nature of partial regime transitions and semidemocratic rule in to- day’s world, especially in the Greater Middle East (a term of convenience I use to cover the broad band of predominantly Muslim countries stretching from West Africa to Southeast Asia). The subsequent section analyzes the unhelpful implications of semidemocratic rule for the theory of democratic peace. The third section reviews evidence showing the difªculty of outsiders in promot- ing regime change. The article concludes with a brief discussion of the cau- tious and targeted democratization strategy implied by these rather somber observations.

Mixed Political Systems and Partial Democratic Transitions

Democratization is a broad, irregular process of replacing authoritarian re- gimes with rule-bound competitive political systems. It is not a universal his- torical sequence. Many countries mix major elements of undemocratic and democratic practice. An example from the Muslim world is the Islamic Republic of , a religious state that abridges the freedom of worship and prohibits people from speaking and writing freely. Yet, within the bounds set by spiritual leaders, Iranians have lively political participation and vigorous competition for ofªce. Another mixed-regime Muslim country is Bangladesh. The opposition won an victory in 2001, but the period since then has seen growing levels of lawlessness and political violence associated with Islamist extremist groups. General elections were postponed in 2007 following the declaration of a national state of emergency. is a further example of semidemocracy from the Muslim world. It had a competitive national election with an opposition victory in 1999, but contests in 2003 and 2007 were marred by violence and charges of fraud and other voting irregularities. Several lead- ing reformist politicians have been assassinated in recent years, and political corruption continues to be a serious problem for Nigeria. All three of these countries have taken “steps toward freedom,” in Bush’s phrase, but they are far from having reliable and consistent democratic institutions. How many semidemocratic systems are there in total, and in the Muslim world in particular? The , a source for national regime typ- ologies used widely in political science research, offers a reasonably precise answer.13 The latest version, Polity IV, contains coded annual information on

13. The Polity IV Project is maintained by the Center for International Development and Conºict Management, University of Maryland, http://www.systemicpeace.org/polity/polity4.htm.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 125

political system characteristics for all independent states (with greater than 500,000 total population) and covers the years 1800 to 2006. Forty-four coun- tries in the Polity IV database have predominantly Muslim populations (henceforward referred to simply as Muslim countries). Of particular interest is the polity2 variable, a composite index for /democracy that is measured on a 21-point scale ranging from Ϫ10 to ϩ10. The maximum score would be given to a country in which the execu- tive is chosen in free and fair elections with universal suffrage and where sub- stantial checks and balances constrain the chief executive’s power. Lower scores reºect diminishing constraints on executive powers and lessening de- grees of competitiveness and political participation. From 1970 to 2006, the average country’s polity2 score rose more than ªve points, indicating a sub- stantial increase in average democratization around the world. But only some countries crossed the threshold to become institutionally consistent democratic systems represented at the high end of the polity2 scale; many others became less autocratic but not dependably democratic with open electoral competition and constitutional protections. To determine the number of mixed regimes versus democratic or authoritar- ian political systems, I follow the conventional coding system researchers use to sort nations into three groups: a country is classiªed as fully democratic if its polity2 score is ϩ6 or greater, and autocratic if its polity2 is Ϫ6 or less. Countries with scores between Ϫ5 and ϩ5 are categorized as semidemocratic. As of 2006, the latest year in the Polity IV time-series, the world contained seventy-seven full democracies, twenty-one , and ªfty-eight semi- democratic systems. That compares with thirty-three full democracies, sixty- ªve autocracies, and thirty semidemocracies in 1970.14 President Bush, citing the United Nations Development Programme’s Arab Development Report, frequently identiªed a “freedom deªcit” in the Greater Middle East.15 But the three-way coding of regimes also suggests a “partial de-

14. The Polity IV project uses the alternative designation “” for mixed or partially demo- cratic political systems. David L. Epstein, Robert Bates, Jack Goldstone, Ida Kristensen, and Sharyn O’Halloran, in “Democratic Transitions,” American Journal of Political Science, Vol. 50, No. 3 (July 2006), pp. 551–569, employ a more restrictive three-way coding system for regime types: full democracies (polity2 value ϩ8toϩ10), full autocracies (Ϫ10 to 0), and partial democracies (ϩ1to ϩ7). Applying their standard, there were sixty-seven full democracies, ªfty-one full autocracies, and thirty-nine partial democracies in 2006. On the upsurge in these mixed regimes, see also Marina Ottaway, Democracy Challenged: The Rise of Semi-Authoritarianism (Washington, D.C.: Carne- gie Endowment for International Peace, 2003). 15. Arab Human Development Report, 2004: Freedom and Good Governance (New York: United Na- tions Development Programme, 2004). The World Bank has also documented what it calls a “gov- ernance gap” in the Middle East and North Africa. See Better Governance for Development in the

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 126

mocracy surplus” in this expansive region, which is at least as signiªcant for international relations. Figure 1 illustrates the scale of the phenomenon. Semi- democracies are unusually prevalent among Muslim countries, and the differ- ence with the rest of the world increased during the 1990s. Sometimes semidemocracy is a necessary but brief transition stage before the adoption of full democracy, but that is not the typical pattern in Muslim countries. More often they settle into a long-term condition between autocracy and constitutional democratic rule. To determine how long these periods can last, I identify regime changes in the 1970s, 1980s, and 1990s using Polity IV codes and the tripartite division of regime types just mentioned. This allows me to estimate six distinct transitions: autocracy to full democracy, autocracy to partial democracy, and partial democracy to full democracy, as well as the reverse of each. I deªne a transition as a shift in regime category that lasts ªve years or more, and that puts the country in a different category than it was in during the nearest previous episode that also lasted a minimum of ªve years. The ªve-year cutoff is based on the assumption that the national leadership will usually be replaced or renewed within that time, giving some assurance that a regime lasting that long has established itself and has been stabilized. Regime interludes of fewer than ªve years are not counted as transitions. Be- cause the Polity IV database extends to 2006, the most recent regime changes that can be recorded using this methodology took place in 2002. To gain some historical perspective and have a reasonable sample of cases, I take 1970 as the base year for tabulating transitions (i.e., I go back to 1966 to start grouping po- litical systems using the ªve-year survival rule). From 1970 to 2002, the world witnessed 123 regime changes as deªned above, with 33 occurring in Muslim nations (see table 1).16 Many countries in the world underwent more than 1 transition; most had none. Among the tran- sitions in Muslim nations, only 5 brought about democratic systems, and 3 of them (, Nigeria, and Pakistan) barely lasted beyond the ªve-year baseline. The majority of these transitions resulted in semidemocratic regimes (table 2 has the full breakdown of the regime change outcomes for Greater Middle Eastern countries).

Middle East and North Africa: Enhancing Inclusiveness and Accountability (Washington, D.C.: World Bank, 2003). These deªcits are put into perspective by Arthur A. Goldsmith, “Muslim Exceptionalism? Measuring the ‘Democracy Gap,’” Middle East Policy, Vol. 14, No. 3 (Fall 2007), pp. 86–96. 16. Applying the stricter alternative thresholds proposed by Epstein, Bates, Goldstone, Kristensen, and O’Halloran, “Democratic Transitions,” there were ninety-two regime changes worldwide from 1970 to 2002, with twenty-two occurring in Muslim countries. Forty-two of the transitions during the period produced a partial democracy, according to this coding system.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 127

Figure 1. Partial Democracy Surplus: Muslim Countries versus Rest of the World, 1970– 2006

SOURCE: Author’s calculations based on the Polity IV Project, “Political Regime Characteris- tics and Transitions, 1800–2006” (College Park: Center for International Development and Conflict Management, University of Maryland, 2008). The composite index for regime type (polity2) is measured on a 21-point scale ranging from Ϫ10 to ϩ10. A country is classified as a semidemocratic regime if the polity2 score is Ϫ5 and ϩ5.

Table 1. Regime Transitions, 1970–2002

Non-Muslim Muslim All Countries Countries Countries Autocracy ® Democracy 17 3 20 Democracy ® Autocracy 1 1 2 Autocracy ® Semidemocracy 36 19 55 Semidemocracy ® Autocracy 9 5 14 Semidemocracy ® Democracy 21 2 23 Democracy ® Semidemocracy 6 3 9 All regime transitions 90 33 123

SOURCE: Polity IV Project, “Political Regime Characteristics and Transitions, 1800–2006” (College Park: Center for International Development and Conflict Management, University of Maryland, 2008). NOTE: A regime transition is a shift in regime category that lasts five years or more, and that puts the country in a different category than it was in during the nearest previous episode that also lasted a minimum of five years. The categories are based on annual national rank- ing on the polity2 variable: democracy (ϩ6toϩ10), autocracy (Ϫ10 to Ϫ6), and semidemocracy (Ϫ5toϩ5).

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 byguest on 26 September 2021

Table 2. Regime Transitions and Foreign Interactions with Muslim Countries, 1970–2004

Official Development Governance and Target of Politically Assistance as Civil Society Aid, USAID Democracy/ Target of Democratic Oriented Economic Percentage of Gross Per Capita (annual Governance Military Intervention, Sanctions, 1970– National Income (annual average 1999– Assistance, 2003 Country Regime Transition 1970–96 2004 average 1995–2004) 2003) (million) (1) (2) (3) (4) (5) (6) (7) Afghanistan 1988–89 30.0% — $4.4 Autocracy ® Semidemocracy 1990 7.7% $16.71 $6.0 Semidemocracy ® Democracy 2002 Autocracy ® Semidemocracy 1995 1992–94 0.5% $0.23 $0.7 Semidemocracy ® Autocracy 1998 1992 3.8% $4.37 $6.9 0.8% $0.14 $0.6 Bangladesh Autocracy ® Semidemocracy 1996 1975–76, 1991–93 2.5% $0.31 $3.4 Brunei 1984–88 — $4.42 — Semidemocracy ® Autocracy 1980 14.3% $2.01 — Autocracy ® Semidemocracy 1991 1969–75, 1977–84, 12.4% $1.14 — 1986–87, 1990– 92, 1994 Semidemocracy ® Autocracy 1985 1989, 1995 12.2% $3.47 — Autocracy ® Semidemocracy 1990 Autocracy ® Semidemocracy 1996 1992 14.1% $3.14 — Autocracy ® Semidemocracy 1976 1960–88 2.0% $0.80 $14.1 31.4% $4.77 $0.01 Gambia Democracy ® Semidemocracy 1996 1981 1994–98 12.4% $2.80 — -Bissau Autocracy ® Semidemocracy 1991 8.3% $1.74 $2.3 Indonesia Autocracy ® Semidemocracy 1998 1991–97 0.8% $0.45 $29.6 Iran Autocracy ® Semidemocracy 1979 1980, 1988, 1994, 0.1% $0.04 — 1996 Iraq 1981, 1983–88, 1980–2003 — $0.95 $78.5 1991–93 Autocracy ® Semidemocracy 1989 1960–88 6.7% $7.03 $1.7 1991, 1993–94, 0.0% $0.03 — 1996 Kyrgyz Republic 14.8% $3.23 $6.2 Lebanon 1960–88 1.5% $3.24 $2.4 Libya 1986, 1994 1992–2003 — $0.36 — 0.0% $0.60 $0.1 Maldives 1988 5.7% $16.60 — Mali Autocracy ® Democracy 1992 15.6% $2.02 $1.8 Democracy ® Semidemocracy 1997 Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 byguest on 26 September 2021

Mauritania Autocracy ® Semidemocracy 1991 1977–80 18.3% $2.49 — Autocracy ® Semidemocracy 1998 1976–78, 1991 1.5% $0.18 $0.4 Autocracy ® Semidemocracy 1996 14.3% $0.66 $0.2 Nigeria Autocracy ® Democracy 1979 1993–98 0.6% $0.29 $12.9 Democracy ® Autocracy 1984 Autocracy ® Semidemocracy 1995 1966–77 0.3% $3.31 $0.1 Pakistan Democracy ® Autocracy 1977 1960–88, 1990, 1991–2001 1.6% $0.22 $5.2 Autocracy ® Democracy 1988 1996 Democracy ® Semidemocracy 1999 0.0% $8.77 $0.4 1984, 1988, 1990– 0.0% $0.01 $0.2 91 Autocracy ® Semidemocracy 1974 10.9% $8.49 $2.0 Semidemocracy ® Democracy 2000 Semidemocracy ® Autocracy 1978 1997–2003 26.4% $5.91 $10.5 Autocracy ® Semidemocracy 1996 Autocracy ® Semidemocracy 1991 1976, 1992–95 — $1.61 $0.7 1984–85 2.7% $0.21 $9.0 1960–88 1.2% $0.19 — 1992, 1994 11.7% $2.60 $4.4 Autocracy ® Semidemocracy 1987 1.1% $0.43 $0.3 1986 1981–86 0.1% $0.02 — Turkmenistan 1.0% $0.50 $1.2 0.0% $0.04 — Uzbekistan 1.3% $0.44 $6.2 West Bank & Gaza 20.9% — $4.5 Yemen 4.6% $1.58 $1.1

SOURCES: col. 2: Polity IV Project, “Political Regime Characteristics and Transitions, 1800–2006” (College Park: Center for International Development and Conflict Management, University of Maryland, 2008); col. 3: Nils Petter Gleditsch, Lene Siljeholm Christiansen, and Håvard Hegre, “Democratic Jihad? Military Intervention and Democracy,” World Bank Policy Research Working Paper, No. 4242 (Washington, D.C.: World Bank, June 2007); col. 4: Kimberly Ann Elliott, Jeffrey J. Schott, and Gary Clyde Hufbauer, Economic Sanctions Reconsidered: History and Current Policy, 3d ed. (Washington, D.C.: Peterson Institute for International Economics, 2007); col. 5: OECD.Stat Extracts, http:// stats.oecd.org/wbos/default.aspx?DatasetCodeϭREF_TOTALODA; col. 6: OECD Development Assistance Committee (data kindly provided by Irene Vlachaki, Athens University of Economics and Business); and col. 7: Democracy Assistance Project website, http://www.pitt.edu/~politics/democracy/democracy.html.http://www.pitt.edu/~politics/ democracy/democracy.html. NOTE: A regime transition is a shift in regime category that lasts five years or more, and that puts the country in a different category than it was in during the nearest previous epi- sode that also lasted a minimum of five years. The categories are based on annual national ranking on the polity2 variable: democracy (ϩ6toϩ10), autocracy (Ϫ10 to Ϫ6), and semidemocracy (Ϫ5toϩ5). A democratic military intervention is a military intervention in which one of the intervening states is a democracy. Governance and civil-society aid is from all donors. Figures for USAID (U.S. Agency for International Development) democracy/governance assistance are in 1995 prices, and refer to the two-year average in aid. International Security 33:2 130

The median age in 2006 of the semidemocratic regimes that emerged among Muslim nations over the 1970 to 2002 period was sixteen years, not including two long-lived regimes that had shifted into semidemocratic status before 1970. These are not passing interregnums. The only fully democratic countries by my designation in the sample in 2006 are Albania, Senegal, and Turkey, pos- sibly to be joined by Indonesia if it makes it past the ªve-year mark in 2008. Scholars have offered several explanations for why regime transitions in the Greater Middle East and other developing regions stop short of full democrati- zation or are reversed and followed by extended periods of political instability. The most widely accepted accounts are associated with modernization theory. Many variants of this theory exist, but the common theme is that certain eco- nomic and social preconditions favor democratic self-government. Countries with extensive poverty (which also implies limited education, lack of social differentiation, and, more controversially, conservative or “traditional” values) are not welcome environments for full democratization. Modernization theory has been challenged over the years for being too deterministic, but the correla- tion between per capita income and regime type holds up well in study after study.17 In 2003 the median gross domestic product per capita in Muslim na- tions was only about $3,600, compared with $7,800 for the rest of the world.18 All other things being equal, only the achievement of limited democracy in the Muslim countries as a group is to be expected, according to modernization theory. Another factor that can inºuence democratization is the neighborhood ef- fect: democracy is more likely to emerge in countries whose neighbors are democratic, presumably due to the cross-border exchanges and peer-to-peer learning that close proximity allows between nations. The relative absence of pluralistic political systems in the Greater Middle East thus also works against the emergence of additional institutionally consistent democratic regimes across that region.19

17. One of the latest studies to demonstrate this pattern is Epstein, Bates, Goldstone, Kristensen, and O’Halloran, “Democratic Transitions.” The seminal paper on modernization is Seymour Mar- tin Lipset, “Some Social Requisites of Democracy: Economic Development and Political Legiti- macy,” American Political Science Review, Vol. 53, No. 1 (March 1959), pp. 69–105. 18. Calculated from Alan Heston, Robert Summers, and Bettina Aten, Penn World Table Version 6.2 (Philadelphia: Center for International Comparisons, University of Pennsylvania, September 2006). 19. See, for example, Renske Doorenspleet, “The Structural Context of Recent Transitions to De- mocracy,” European Journal of Political Research, Vol. 43, No. 3 (May 2004), p. 328; and Daniel Brinks and Michael Coppedge, “Diffusion Is No Illusion: Neighbor Emulation in the Third Wave of De- mocracy,” Comparative Political Studies, Vol. 39, No. 4 (May 2006), pp. 463–489. One false hope of

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 131

Probably the most controversial explanation offered for regime patterns in this area is cultural. In the words of Samuel Huntington, “Islamic culture ex- plains in large part the failure of democracy to emerge in much of the Muslim world.”20 Advocates of democracy promotion obviously reject the idea that Muslim norms are particularly adverse to democracy and rule of law. Larry Diamond, for example, attributes the regional democracy gap to entrenched leadership and lack of opportunity for participation by ordinary people, not retrograde values or beliefs.21 Hence, several deep structural factors could account for the prevalence of semidemocracy in Muslim countries. Policymakers should not confound these mixed regimes, however, with full or institutionally consistent democracies.

Misapplication of the Democratic Peace

The intellectual foundation for democracy promotion is the theory of demo- cratic peace. (Note that it is not the theory of “semidemocratic peace.”) The theory points to two well-established empirical relationships: democratic po- litical systems seldom or never ªght wars against each other; and (more con- troversially) democracies have a greater propensity to avoid serious disputes with other countries that could boil over to war.22 Therefore, if democracies are peace loving and nonthreatening, the elimination of must be a clear way to reduce potential conºict among nations and enhance U.S. secu- rity. This is the underlying rationale for the forward strategy of freedom and democracy promotion generally. The empirical regularities observed for peaceful regimes, however, apply to established constitutional political systems, not to the mixed regimes common in the Greater Middle East. Any expansion in the group of partially democratic

the forward strategy of freedom was that the establishment of a democratic regime in Baghdad would start a regional democratic chain reaction. Recent cross-national studies, however, suggest there is much greater likelihood of spatial contagion for internal conºicts, which partially demo- cratic conditions may both facilitate and make more likely to spread. See, for example, Clionadh Raleigh, “Civil War Risk in Democratic and Non-Democratic Neighborhoods,” World Bank Policy Research Working Paper, No. 4260 (Washington, D.C.: World Bank, June 2007). 20. Samuel P. Huntington, The Clash of Civilizations and the Remaking of World Order (New York: Si- mon and Schuster, 1996), p. 29. 21. Larry Diamond, “Can the Whole World Become Democratic? Democracy, Development, and International Policies,” Center for the Study of Democracy, Paper No. 03–05 (Irvine: University of California, April 17, 2003), p. 13. 22. The democratic peace literature is too vast to cite, but a good place to start is the reader edited by Michael E. Brown, Sean M. Lynn-Jones, and Steven E. Miller, eds., Debating the Democratic Peace (Cambridge, Mass.: MIT Press, 1996).

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 132

nation-states seems more likely to hinder international security than to help it. Stalled transitions and semidemocratic rule are each correlated in some empir- ical studies with (1) strife within nations that can spread to other countries, (2) large and small wars between nations, and (3) rising incidence of terrorism. The cross-national evidence of danger is strongest regarding intrastate conºicts. Partial democracies, no matter the region, represent a disproportion- ate amount of the world’s political turmoil, according to a report from the Political Instability Task Force (formerly the State Failure Task Force), a U.S. government–funded interdisciplinary research program. In the period 1955 to 2001, mixed regimes by its tally accounted for more than one-third of all major “political instability events” (i.e., adverse shifts in patterns of governance, eth- nic wars, revolutionary wars, and genocides).23 Mass killings of civilians, for example, are most common at intermediate levels of democracy, with a decline at higher levels.24 Other studies conªrm that more political repression and hu- man rights violations occur in semidemocratic states compared to full democ- racies.25 These difªculties are plausibly explained as the consequence of social mobilization and expanded political participation in the absence of self- restraining governing institutions. Social disorder in an intermediate regime stands a good chance of spiraling into civil war, as implied by James Fearon and David Laitin’s empirical work. The likelihood of armed conºicts breaking out within partial democracies is two-thirds greater than in full autocracies for the period 1945 to 1999. This may be the consequence of such regimes lacking the resources to either crush or placate insurgents, the authors speculate.26 Extending the analysis to 1816, an- other study by Håvard Hegre, Tanja Ellingsen, Scott Gates, and Nils Gleditsch

23. Jack A. Goldstone, Robert Bates, Ted Robert Gurr, Michael Lustik, Monty G. Marshall, Jay Ulfelder, and Mark Woodward, “A Global Forecasting Model of Political Instability,” paper pre- sented at the annual meeting of the American Political Science Association, Washington, D.C., Sep- tember 3, 2005, http://globalpolicy.gmu.edu/pitf/pitfp5.htm. 24. William Easterly, Roberta Gatti, and Sergio Kurlat, “Development, Democracy, and Mass Kill- ings,” Journal of Economic Growth, Vol. 11, No. 2 (June 2006), p. 135. 25. Patrick M. Regan and Errol A. Henderson, “Democracy, Threats, and Political Repression in Developing Countries: Are Democracies Internally Less Violent?” Third World Quarterly, Vol. 23, No. 1 (February 2002), pp. 119–136; and Christian Davenport and David A. Armstrong II, “Democ- racy and the Violation of Human Rights: A Statistical Analysis from 1976 to 1996,” American Jour- nal of Political Science, Vol. 48, No. 3 (July 2004), pp. 538–554. 26. James D. Fearon and David D. Laitin, “Ethnicity, Insurgency, and Civil War,” American Political Science Review, Vol. 97, No. 1 (February 2003), p. 85. Looking at a similar period, Errol A. Henderson and J. David Singer also show that semidemocratic regimes are signiªcantly associated with an increased likelihood of civil war. See Henderson and Singer, “Civil War in the Post- Colonial World, 1946–92,” Journal of Peace Research, Vol. 37, No. 3 (May 2000), p. 289.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 133

ªnds much the same pattern: democratic regime change is strongly correlated with internal military conºict, though the prospects for civil peace improve with time if the country evolves into an established democratic system.27 Refugee ºows, border incursions, rising irredentist sentiment in neighboring states, and other dangers make civil wars difªcult to contain.28 A contempo- rary example from the Muslim world of uncontained intrastate conºict is the ethnic war in the Darfur region of Sudan (ranked a semidemocratic country when that conºict erupted in 2003). Irregular forces soon began to attack vil- lages across the frontier with Chad, where civilians and resistance ªghters had ºed. These events escalated into coup attempts in Chad, which were alleged to have been coordinated with the Sudanese government operating through local proxy groups. The has also been pulled into the tur- moil, with rebel militias from its territory taking refuge in Sudan. The second way partial democracies may imperil international peace is more direct: a mixed regime may have a greater propensity to use its regular military forces against another nation-state (or inadvertently to invite an attack against itself). A chief reason fully established democracies are thought to re- solve foreign problems peacefully is because institutional checks and balances make them attentive to the demands of the citizenry—and hence cautious in embarking on policies that might be detrimental to the majority interest, such as starting wars. Where the representative institutions are less rule-bound or only partly competitive, the opposite may happen because the regime is not as beholden to popular sentiment. On the other hand, a mitigating factor might be that semidemocracies are too disorganized to mobilize resources to assault a neighboring nation-state, though internal disorganization also could send confused signals to potential adversaries and raise the risk of invasion. Recent examples of partial democracies becoming involved in overt conºict

27. Håvard Hegre, Tanja Ellingsen, Scott Gates, and Nils Petter Gleditsch, “Toward a Democratic Civil Peace? Democracy, Political Change, and Civil War, 1816–1992,” American Political Science Re- view, Vol. 95, No. 1 (March 2001), p. 44. This study’s ªndings regarding partial democracy are dis- puted in replication analysis by Shawn Treier and Simon Jackman, “Democracy as a Latent Variable,” American Journal of Political Science, Vol. 52, No. 1 (January 2008), pp. 201–217, who ªnd no signiªcant relationship between the level of democracy and civil war onset. James Raymond Vreeland shows that the relationship between partial democracy and civil war may actually be due to underlying ethnic conºicts. Vreeland, “The Effect of Political Regime on Civil War: Un- packing Anocracy,” Journal of Conºict Resolution, Vol. 52, No. 3 (June 2008), pp. 401–425. 28. See, for example, Idean Salehyan and Kristian Skrede Gleditsch, “Refugees and the Spread of Civil War,” International Organization, Vol. 60, No. 2 (April 2006), pp. 335–366; and Håvard Hegre and Nicholas Sambanis, “Sensitivity Analysis of Empirical Results on Civil War Onset,” Journal of Conºict Resolution, Vol. 50, No. 4 (August 2006), p. 529.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 134

with each other are easy to ªnd—for example, the war between Azerbaijan and that broke out in 1992 or Senegal’s intervention in Guinea-Bissau in 1998—but the large-n research is contradictory on whether such incidents are part of a pattern of international violence. Some studies suggest that coun- tries undergoing democratic transitions are not unusually likely to join in wars with neighbors, and that truncated transitions do not contribute to the prob- ability of participation in interstate hostilities.29 Other statistical evidence, however, shows that states experiencing incomplete transitions are apt to start wars. Limited democracies appear to be inherently more aggressive than other regime types even beyond the period of regime transition, according to an- other study. Swings between democracy and autocracy are also found to in- crease a country’s inclination to ªght.30 The net effects of democratization and semidemocracy on interstate warfare are thus questions that political science has yet to settle, but these factors do not appear likely to diminish the risks of external war in the short run. At best, they may not add to the risks. Terrorism is the third threat to world peace that the spread of democracy is supposed to alleviate. The ofªcial view in Washington is that democratic gov- ernment reduces the likelihood of nonstate actors committing politically moti- vated acts of unlawful violence.31 Democratic institutions offer aggrieved citizens avenues for redress, thereby eliminating the main motivation to use force as a means to publicize or avenge grievances. It is also plausible, how- ever, that democratic institutions create an optimal breeding ground for vio- lent radicalism. Every society has its disaffected elements. The openness of democratic societies reduces the cost to extremists of becoming organized and trying to create a climate of fear and intimidation. One reason to be particu- larly suspicious about partial democracies is that they have incentives to toler- ate certain international terrorists on their territory. As noted, such regimes are

29. William R. Thompson and Richard Tucker, “A Tale of Two Democratic Peace Critiques,” Jour- nal of Conºict Resolution, Vol. 41, No. 3 (June 1997), p. 445; and David L. Rousseau, Democracy and War: Institutions, Norms, and the Evolution of International Conºict (Stanford, Calif.: Stanford Univer- sity Press, 2005), p. 279. 30. Edward D. Mansªeld and Jack Snyder, Electing to Fight: Why Emerging Democracies Go to War (Cambridge, Mass.: MIT Press, 2005), p. 134; Sandeep Baliga, David O. Lucca, and Tomas Sjöström, “Domestic Political Survival and International Conºict: Is Democracy Good for Peace?” Working Paper (Evanston, Ill.: Kellogg School of Management, Northwestern University, April 2007); and Kristian Gleditsch and Michael D. Ward, “War and Peace in Space and Time: The Role of Democra- tization,” International Studies Quarterly, Vol. 44, No. 1 (March 2000), p. 26. 31. George W. Bush, National Strategy for Combating Terrorism (Washington, D.C.: White House, September 2006), p. 10.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 135

already vulnerable to domestic rebellion, so they may want to offer passive support to foreign terrorist groups to avoid becoming targets themselves and adding to their security headaches. They may even sponsor international ter- rorists as a low-cost alternative to projecting ofªcial military force. By my count, nine of the twenty-ªve countries most at risk for terrorist at- tacks in 2003 were semidemocracies which looks as if it is a statistically sig- niªcant correlation according to other research.32 Alberto Abadie, for example, ªnds a nonmonotonic relationship between political freedom and the risk of terrorism for that year, indicating that countries with intermediate lev- els of democracy are more prone to this form of violence than countries with high or low levels.33 Peter Kurrild-Klitgaard, Mogens Justesen, and Robert Klemmensen developed an alternative model to explain the probability of ter- rorist attacks in different settings. They show that countries at a midrange of democracy were more liable to experience terrorism than were authoritarian or democratic countries for the period 1996 to 2002.34 These ªndings suggest that semidemocracy may be the worst of both worlds—unable either to repress violent dissent or to route it into socially productive channels. If terrorism does tend to ºourish in mixed regimes, the impact on the United States may be limited unless it has numerous citizens or economic assets in those countries. Cross-border terrorism is the bigger concern. An important question therefore is whether partial democracies are more or less likely to serve as sanctuaries for exporting terrorism. Alan Krueger and David Laitin have published a piece on the national origins of transnational terrorism from 1981 to 2002 using a three-way split among regimes. They separate countries with a middle level of civil liberties—a classiªcation similar to the concept of partial democracy used here. According to Krueger and Laitin, these midlevel

32. The list of at-risk countries is from David A. Rice, “WMRC’s Global Terrorism Index in 2003,” in “Global Threat Analysis,” presentation, August 14, 2003, http://www.policyconcepts.com/ 220054.ppt, slides 7–11. 33. Alberto Abadie, “Poverty, Political Freedom, and the Roots of Terrorism,” American Economic Review, Vol. 96, No. 2 (May 2006), p. 51. 34. Peter Kurrild-Klitgaard, Mogens K. Justesen, and Robert Klemmensen, “The Political Econ- omy of Terrorism, Democracy and Freedom,” Public Choice, Vol. 128, Nos. 1–2 (July 2006), p. 304. See also Erica Chenoweth, “Democratic Pieces: Democratization and the Origins of Terrorism,” in William R. Thompson, and Rafael Reuveny, ed., Coping with Contemporary Terrorism (Baltimore, Md.: Johns Hopkins University Press, forthcoming). Looking just at suicide terrorism, however, it appears there is no systematic correlation with regime type. See Sara Jackson Wade and Dan Reiter, “Does Democracy Matter? Regime Type and Suicide Terrorism,” Journal of Conºict Resolu- tion, Vol. 51, No. 2 (April 2007), pp. 329–348.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 136

countries are a more common base for attacks than are countries with either high or low levels of civil liberties. The targets, when these terrorists go inter- national, are usually high-income nations such as the United States.35 In brief, even if one accepts the proposition that democracy is historically a peaceable type of political system, there are several reasons to doubt whether that depiction applies to the mixed regimes characteristic of the Greater Middle East and other parts of today’s postcolonial world. Measured in a vari- ety of ways, partial democratization represents little or no improvement for in- ternational security. More likely outcomes from halfway democratization are increases in regionally destabilizing internal disorder and civil war, a steady or perhaps growing number of cross-border military disputes, and possibly the further spread of terrorist activities. The neoconservative mind-set, with its emphasis on “moral clarity” and simplistic binary perspective on geopolitics, has failed to grapple with the security implications of partial democracy.

Molding Democracy from the Outside

In an ideal world, the United States and its allies would take the steps neces- sary to reduce the number of mixed regimes and replace them with full- ºedged democracies. The evidence, however, does not support broad claims that conscious action by outside actors makes a consistent difference in pro- ducing any type of regime. Scholars have debated the external versus the in- ternal sources of regime change and democratization, with international relations specialists often emphasizing external elements such as membership in international organizations and trading relationships, and comparativists stressing indigenous social and economic preconditions and domestic political bargaining and coalitions.36 No one denies, however, that regime change is

35. Alan B. Krueger and David D. Laitin, “Kto Kogo? A Cross-Country Study of the Origins and Targets of Terrorism,” in Philip Keefer and Norman Loayza, eds., Terrorism, Economic Development, and Political Openness (New York: Cambridge University Press, 2008), pp. 148–173. A piece by S. Brock Blomberg and B. Peter Rosendorff ªnds that the advent of democratic institutions does signiªcantly reduce the likelihood of citizens from one country engaging in unauthorized violent conduct against another country. The model has a dichotomous regime variable, however, and thus may have failed to pick up terrorism exported from semidemocratic countries. See Blomberg and Rosendorf, “A Gravity Model of Globalization, Democracy, and Transnational Terrorism,” in Gregory D. Hess, ed., Guns and Butter (Cambridge, Mass.: MIT Press, forthcoming). 36. See, for example, Michael Mandelbaum, “Democracy without America,” Foreign Affairs, Vol. 86, No. 5 (September/October 2007), pp. 119–130; Hans Peter Schmitz, “Domestic and Transna- tional Perspectives on Democratization,” International Studies Review, Vol. 6, No. 3 (September 2004), pp. 403–426; Jon C. Pevehouse, “Democracy from the Outside-In? International Organiza- tions and Democratization,” International Organization, Vol. 56, No. 3 (Summer 2002), pp. 515–549;

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 137

multicausal and that the international and national factors interrelate in com- plex and varied patterns. Consequently, external powers have an uphill task to promote democracy in a systematic way. Steven Levitsky and Lucan Way propose a useful way of conceptualizing the international dimension of democratization. They posit two basic sources of international inºuence: leverage, or the target governments’ exposure to ex- ternal pressure; and linkage, or the density of external economic and political ties. Both leverage and linkage raise the cost of authoritarianism, but linkage contributes more consistently to democratization, they say. Where linkage is low, Levitsky and Way speculate that external democratizing pressure is prob- ably going to be ineffective. Expanding linkage, however, seems likely to act slowly and incrementally in producing regime transitions. More aggressive ef- forts to accelerate the process through leverage, they say, seem certain to prove counterproductive in the absence of favorable internal conditions.37

leverage The most obvious form of leverage is the use of military power to change regimes and extend democracy. Neoconservatives have a cavalier attitude about this approach. and , for instance, casually pro- claim that “when it is in our power and our interest, we [the United States] should toss dictators aside with no more compunction than a police sharp- shooter feels when he downs a hostage-taker.”38 Offhandedness about “gun barrel” democratization is deeply misplaced, ac- cording to many empirical studies. An example is an article by Charles Kegley and Margaret Hermann of military interventions by all democratic states from 1974 to 1988. They report that on those rare occasions when a regime change does ensue, the result is most often a “partly free” state.39 Looking at a longer period (1960–96) of nearly 700 military interventions, Nils Gleditsch, Lene Christiansen, and Håvard Hegre again show that these are largely ineffectual in changing regimes. When a democracy introduces military forces to a non-

and Jean Grugel, ed., Democracy without Borders: Transnationalization and Conditionality in New De- mocracies (London: Routledge, 1999). 37. Steven Levitsky and Lucan A. Way, “Linkage versus Leverage: Rethinking the International Dimension of Regime Change,” Comparative Politics, Vol. 38, No. 4 (October 2006), pp. 379–400. 38. David Frum and Richard Perle, An End to Evil: How to Win the War on Terror (New York: Ran- dom House, 2003), p. 114. 39. Charles W. Kegley Jr. and Margaret G. Hermann, “A Peace Dividend? Democracies’ Military Interventions and Their External Political Consequences,” Cooperation and Conºict, Vol. 32, No. 4 (December 1997), pp. 357–359.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 138

democratic nation, that action is seldom followed by an advance in democrati- zation; when forward movement does occur, it frequently stops at the volatile semidemocratic type of regime.40 If the regime being imposed is only partially democratic, it will probably undermine regional peace, according to a third empirical study going back to 1909.41 A few researchers do report more favorable results, for instance, James Meernick in a study of the political outcomes of postwar U.S. military inter- ventions. He estimates that the target states are slightly more likely to make moves toward democracy. Notably, however, Meernick measures only shifts in national Polity index ratings, and does not code for major regime discontinu- ities as I have done. Some of the movement he picks up, therefore, may simply be minor changes within a semidemocratic regime or perhaps represents au- thoritarian states shifting to mixed political systems.42 Why democratic military interventions lack signiªcant democratizing effects on target governments is not entirely clear, but a common explanation is the intrinsic clash between the interests of the intervening power and the desire of the local population for self-determination. Regime change leads to political uncertainty and the possibility of “wrong” (e.g., anti-Western or pro-Islamist) outcomes. To minimize this risk, the intervener tries to sway postconºict deci- sions in the target country, sometimes obtaining the results it wants but under- cutting majority rule and national independence.43 There have been many

40. Nils Petter Gleditsch, Lene Siljeholm Christiansen, and Håvard Hegre, “Democratic Jihad? Military Intervention and Democracy,” World Bank Policy Research Working Paper, No. 4242 (Washington, D.C.: World Bank, June 2007), p. 39. 41. Andrew J. Enterline and J. Michael Greig, “Beacons of Hope? The Impact of Imposed Democ- racy on Regional Peace, Democracy, and Prosperity,” Journal of Politics, Vol. 67, No. 4 (November 2005), p. 1089. 42. James Meernik, “United States Military Intervention and the Promotion of Democracy,” Jour- nal of Peace Research, Vol. 33, No. 4 (November 1996), p. 395. Another favorable report is by Mark Peceny, “Forcing Them to Be Free,” Political Research Quarterly, Vol. 52, No. 3 (September 1999), pp. 549–582. Analyzing the historical relationship between U.S. military intervention and the de- mocratization of target states over the last century, Peceny suggests that active support for free and fair elections during those interventions often had a major positive impact. Subsequent repli- cation by Scott Walker and Frederic S. Pearson, however, challenges Peceny’s ªndings. Using al- ternative measures of democratic performance and different statistical analyses, they ªnd that attempts to compel countries to hold legitimate elections do not enhance long-term democratic outcomes. See Walker and Pearson, “Should We Really ‘Force Them to Be Free?’ An Empirical Ex- amination of Peceny’s Liberalizing Intervention Thesis,” Conºict Management and Peace Science, Vol. 24, No. 1 (March 2007), pp. 37–53. See also William Easterly, Shanker Satyanath, and Daniel Berger, “Superpower Interventions and Their Consequences for Democracy,” Brookings Institu- tion Global Economy and Development Working Paper, No. 17 (Washington, D.C.: Brookings In- stitution, January 2008), who show that covert operations and other U.S. interventions falling short of military involvement have no perceptible beneªt for democratization. 43. See, for example, Bruce Bueno de Mesquita and George W. Downs, “Intervention and Democ- racy,” International Organization, Vol. 60, No. 3 (Summer 2006), pp. 627–649.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 139

examples of this type of meddling in Iraq, for instance, when the Bush admin- istration demanded that Prime Minister Ibrahim al-Jaafari step aside in 2006. Regarding the regime changes for Muslim countries (table 2, column 2), it seems clear that democratic military interventions have not been a major factor in most transitions, no matter the direction of change.44 Forty-seven separate interventions (as compiled by Gleditsch, Christiansen, and Hegre) took place in twenty-six Muslim countries from 1970 to 1996; only twelve of these coun- tries experienced a transition. And of those, only the U.S. special forces opera- tion in Iran (1980) to rescue American hostages, the French incursion in Comoros (1989) to arrest the mercenary coup leader, Bob Denard, and the hu- manitarian mission in Somalia (1992) correspond closely in time to a regime change—and in none of these cases would it be plausible to argue that the mil- itary action helped spread democracy. In contrast, eleven other Muslim coun- tries underwent regime changes (three of which were democratic transitions, albeit short-lived ones in Mali and Nigeria) without having any external mili- tary involvement by democratic states or international organizations over the 1970 to 1996 period. These data imply that coercion is an unpromising means to spread political freedom. A less punitive form of leverage used by the United States and other powers is to impose international trade and ªnance restrictions on the target country until it makes democratic concessions. Many in the foreign policy ªeld are skeptical about the efªcacy of economic sanctions, especially on an autocracy where the leadership can shift the cost from itself to the larger society.45 The most notable and comprehensive study that indicates economic sanctions sometimes do work is by Kimberly Ann Elliott, Jeffrey Schott, and Gary Hufbauer.46 I use their data to explore the relationship between sanctions and regime change. Of 183 cases of economic sections they catalogue for the post–World War II period, 85 had primarily political aims, as opposed to military or economic ob- jectives. These were cases where Elliott, Schott, and Hufbauer list policy goals

44. A military intervention includes any episode when national military personnel are purpose- fully inserted into another state, and can include peacekeeping missions, evacuation of foreign cit- izens, and attempts to shore up a faltering government; a democratic intervention is one where the sending regime is democratic, and includes troops sent under the auspices of international organi- zations. See Gleditsch, Christiansen, and Hegre, “Democratic Jihad?” 45. Robert A. Pape, “Why Economic Sanctions Do Not Work,” International Security, Vol. 22, No. 2 (Fall 1997), pp. 90–136; and Robert A. Pape, “Why Economic Sanctions Still Do Not Work,” Interna- tional Security, Vol. 23, No. 1 (Summer 1998), pp. 66–77. 46. Kimberly Ann Elliott, Jeffrey J. Schott, and Gary Clyde Hufbauer, Economic Sanctions Recon- sidered: History and Current Policy, 3d ed. (Washington, D.C.: Peterson Institute for International Economics, 2007).

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 140

such as achieving democracy, human rights, or the destabilization of a dicta- torship. The objectives were pursued through both multilateral sanctions and unilateral embargos and boycotts. I separated a subset of cases where the initi- ating party was a democratic state in the terminal year, according to the crite- ria used earlier for coding countries with the polity2 variable. (International bodies were coded as democracies for this purpose.) I dropped cases if the eco- nomic action ended before 1970 or was ongoing in 2004. This sorting results in a sample of 61 cases in the data set where a demo- cratic state or states imposed politically oriented sanctions on another state. In only 9 cases was a Muslim country the object of foreign economic pressure that terminated during the period under consideration. What happened to them? Table 2 (column 4) shows that just 6 of the Muslim target countries underwent a regime change around the period of the sanctions—and all of those produced a semidemocratic system of government. Although many factors account for whether countries take up full demo- cratic rule, the application of economic pressure cannot be said to have pro- duced any more democratic headway than military pressure has in the Muslim world, judging from these limited observations. Certainty about the effectiveness of sanctions will remain elusive, because one cannot ob- serve the threat of imposing them. Threatened economic penalties may regu- larly elicit the intended response without the need to resort to actual sanctions, but that did not happen with democratization in these cases.47 Anecdotal evidence instead suggests that efforts to apply economic or mili- tary leverage are counterproductive in the Greater Middle East. Arab publics are deeply cynical about U.S. policy aims, according to opinion surveys. What are perceived as high-handed interventions by the United States have in- creased popular support for Islamists and nationalists with anti-American views. Under the circumstances, elites have grown even less likely to cooper- ate overtly with the United States or to open up their political systems.48

linkage A form of linkage (which also has elements of leverage because it can be with- held as punishment) that may promote democracy is development assistance. Such resources are the basis for a web of organizational connections between

47. For a more favorable view of economic sanctions helping promote representative government around the world, see Nikolay Marinov, “Foreign Pressure and the Political Foundations of De- mocracy” (Los Angeles: International Institute, University of California, Los Angeles, November 27, 2004). 48. See, for example, Carothers, “U.S. Democracy Promotion During and After Bush,” pp. 14–15.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 141

rich and poor countries, which could spur regime changes as aid recipients learn about and rework practices introduced and encouraged by aid donors. Over time, the international ºow of concessionary capital, modern technology, and trained personnel may induce national economic and political systems to converge on Western models. Aid donors sometimes try to hasten the diffusion of democracy through aid conditionalities requiring the aid recipient to undertake speciªc govern- ment reforms to increase transparency and accountability to its own citizens. Governance-related conditionalities represent the bulk of conditionalities that international ªnancial institutions impose on aid recipients.49 The United States’ Millennium Challenge Account reserves its funds for needy countries that also meet certain benchmarks for human rights and democratic governance. Intentions aside, the amount of ofªcial development assistance is seldom large enough to be a promising incentive for far-reaching democratization, un- less the recipient is already moving in that direction on its own. Of the Muslim countries in table 2, twelve received aid that was the equivalent of less than 1 percent of their gross national income from 1995 to 2004 (see column 5). Among the rest, there were eight countries whose aid equaled from 1 to 2 per- cent of gross national income. For many aid-receiving Muslim countries, there- fore, donors lack carrots with which to coax political and institutional reforms from an unwilling partner. It would be surprising, therefore, if the quantitative literature found a major impact from foreign aid on democratic transitions. If donors are merely blufªng about withholding aid to dictatorships, their assistance might plausibly have the perverse effect of creating “moral hazard” for leaders, allowing or encouraging them to pursue nondemocratic goals at odds with what donors want. For aid recipients, unenforced conditionalities may also signal a lack of “donor credibility” regarding domestic government reforms, which may lead aid recipients to carry on with sham improvements and just enough half-measures to keep the aid ºowing. Many countries in the Muslim world, for example, have been called “pseudodemocratic,” a pose adopted by elites who do not want to displease the international community or relinquish their hold on power.50 The limited quantitative research on total aid ºows and democratization

49. Devesh Kapur and Richard Webb, “Governance-related Conditionalities of the International Financial Institutions,” G24 Discussion Paper Series, No. 6 (New York: United Nations Conference on Trade and Development, August 2000). 50. Frédéric Volpi, “Pseudo-democracy in the Muslim World,” Third World Quarterly, Vol. 25, No 6 (November 2004), pp. 1061–1078.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 142

tends to back up the doubters and detractors. Stephen Knack performs a multivariate analysis of the impact of ofªcial development assistance on re- gime change in a large sample of recipient nations from 1975 to 2000. He ªnds no support for the proposition that aid promotes democracy.51 Using alterna- tive data for 108 recipient countries from 1960 to 1999, Simeon Djankov, José Montalvo, and Marta Reynal-Querol show that foreign aid has a negative im- pact on democracy. They argue that this is because too much foreign aid re- duces the incentives for domestic accountability, as recipient governments do not have to depend on the taxes raised from citizens and business.52 On the other hand, aid shows up as a small or contingent positive factor in some other studies.53 There is no research, however, indicating that all-purpose ªnancial and development aid is a major beneªcial element in jump-starting democrati- zation. The data in table 2 demonstrate no obvious correlation between over- all aid levels and democratic transitions; indeed, some of the largest recipients of aid on a proportional basis underwent no regime changes during the period under review. For the full picture of aid’s linkage effects, however, it is necessary to divide targeted democracy promotion aid from overall ºows of resources. There has been a rise since 1990 in technical assistance focusing on electoral processes, adherence to the rule of law, and related activities. Only a fraction of develop- ment aid is expressly set aside for such purposes, though democracy assis- tance is difªcult to identify with much precision. One approximation is the aid listed under the Organisation for Economic Co-operation and Development’s purpose code for government and civil society. Bilateral and multilateral dis-

51. Stephen Knack, “Does Foreign Aid Promote Democracy?” International Studies Quarterly, Vol. 48, No. 1 (March 2004), pp. 251–266. 52. Simeon Djankov, José García Montalvo, and Marta Reynal-Querol, “The Curse of Aid,” Eco- nomics Working Paper, No. 870 (Barcelona: Universitat Pompeu Fabra, March 2006, updated De- cember 2007). 53. See, for example, Arthur A. Goldsmith, “Foreign Aid and Statehood in Africa,” International Organization, Vol. 55, No. 1 (2001), pp. 123–148; and Thad Dunning, “Conditioning the Effects of Aid: Cold War Politics, Donor Credibility, and Democracy in Africa,” International Organization, Vol. 58, No. 2 (April 2004), pp. 409–423. Another study suggests donors can have positive political impact depending on the beneªciary’s existing regime type: foreign aid makes already democratic countries (including partial democracies) more democratic, but it has no effect on dictatorships, according to Nabamita Dutta and Peter T. Leeson, “The Ampliªcation Effect: Foreign Aid’s Impact on Political Institutions,” Working Paper (Fairfax, Va.: Department of Economics, George Mason University, 2007). Gabriella Montinola, “Foreign Aid and Democratization Revisited,” paper pre- sented at the annual meeting of the International Studies Association, Honolulu, Hawaii, March 5, 2005, also offers evidence that aid is effective for promoting democracy—but only after the instal- lation of democratic practices.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 143

bursements for this sector came to $9.1 billion in 2006. That was about seven times the amount disbursed for government and civil society in 2000.54 Again, this translates into tiny amounts at the country level in most nations, as column 6 in table 2 makes clear: the median value of governance assistance given to Muslim nations was just $1.14 per capita per year for the period 1999 to 2003. Four of the countries receiving the largest amounts (still only about $10 per capita) are a jumble: one semidemocracy (Jordan), one autocracy (Qatar), and two democracies (Albania and Senegal) in 2002. Yet, despite the small volumes of targeted democracy assistance, several recent studies ªnd that the strategy works, presumably because there is some democratic multi- plier effect from investing in public institutions and civil-society organizations. These studies do not confront the mixed regime problem being raised here, however, so they may be exaggerating the beneªcial impact. Steven Finkel, Aníbal Pérez-Liñán, and Mitchell Seligson, for example, iso- late U.S. Agency for International Development (USAID) spending on democ- racy and governance activities from 1990 to 2003. The level of specialized aid (see column 7 in table 2) is signiªcantly related to several metrics of rights and good governance, according to their work. The authors conclude that USAID democracy assistance has played a positive, though minor role in promoting democratic rule (measured by Polity IV and Freedom House indices) in recipi- ent countries.55 An incremental improvement in a country’s democracy score, however, may not represent a regime change, and even if it does, the country may have attained only partial democratic standing. Finkel, Pérez-Liñán, and Seligson also considered regional inºuences in their model. The coefªcients suggest that democracy and governance assis- tance lack a discernible effect in the Middle East, for reasons the authors do not explore.56 This stipulation makes it hard to conclude that USAID’s efforts have been effective as a short- to medium-term foreign policy tool in that region. Finkel, Pérez-Liñán, and Seligson’s qualiªed positive results are conªrmed

54. Data taken from Organisation for Economic Co-operation and Development, OECD.StatsExtracts, Creditor Reporting System, http://stats.oecd.org/WBOS/Index.aspx ?DatasetCode?CRSNEW. 55. Steven E. Finkel, Aníbal Pérez-Liñán, and Mitchell A. Seligson, “The Effects of U.S. Foreign Assistance on Democracy Building, 1990–2003,” World Politics, Vol. 59, No. 3 (April 2007), pp. 404– 439. 56. Steven E. Finkel, Aníbal Pérez-Liñán, and Mitchell A. Seligson, with the assistance of Dinorah Azpuru, “Effects of U.S. Foreign Assistance on Democracy Building: Results of a Cross-National Quantitative Study,” Final Report to U.S. Agency for International Development (Washington, D.C.: USAID, January 12, 2006), p. 80.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 144

in another recent empirical inquiry that extends the analysis to include gov- ernment and civil-society aid provided by all donor countries over an even longer period of three decades. Looking at ªve-year time horizons, Sarantis Kalyvitis and Irene Vlachaki likewise report that democracy assistance is posi- tively associated with democratic transitions (using criteria established by Freedom House) in recipient countries. But the aid in their model enhances the occurrence of “free” and “partly free” regimes alike—as I have noted, having more of the latter is of uncertain advantage for U.S. foreign policy goals.57 Similar to the ªndings of Finkel, Pérez-Liñán, and Seligson, Kalyvitis and Vlachaki’s equations show that being a Muslim country has a consistently neg- ative coefªcient.58 Accordingly, it does not seem from their analysis, either, that government and civil-society aid has been productive in the Greater Mid- dle East, where the United States and its allies most want to see democratic progress. The negative outcome may be due to suspicion in that region that the West has ulterior motives for offering aid, which compromises any civic orga- nization, political party, or electoral institution that accepts outside help. Thus, leverage and linkage have both been widely employed to promote democracy, but with less success than is generally acknowledged. Linkage activities based on low-key technical assistance and institution-building relationships appear to be the most promising avenue for encouraging democ- ratization, but targeted aid does not seem to function in xenophobic or nation- alistic political environments as exist in some Muslim countries. Ironically, high-proªle rhetoric such as President Bush used to sell his freedom agenda probably exacerbates this problem. Even lower prospects of success can be projected for democracy promotion that relies mainly on the leverage of mili- tary force or economic sanctions—barring special advantageous local circum- stances, such as signiªcant recent experience with democratic government or the presence of democratic neighbors, neither of which pertain to most Muslim countries. It is little wonder that Bush’s forward strategy of freedom has turned out so badly.

57. Sarantis C. Kalyvitis and Irene Vlachaki, “Democracy Assistance and the Democratization of Recipients,” Working Paper (Athens, : Athens University of Economics and Business, Octo- ber 2007). 58. Ibid. Comparable marginal positive results are also given in Nicholas Burger, “Targeted Aid Effectiveness: Evidence from Democracy Promotion Organizations,” Working Paper (Santa Barbara: University of California, Santa Barbara, November 2007); and in Pamela Paxton and Rumi Morishima, “Does Democracy Aid Promote Democracy?” unpublished manuscript, Ohio State University, 2005.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 145

Conclusion

Neoconservatives tout the transformative power of democratization to protect U.S. interests by establishing zones of peace around the world. To anyone who suspects that the global democracy promotion project is overreaching or uto- pian, Robert Kagan and William Kristol posed the following question in 2000: “How utopian is it to imagine a change of regime in a place like Iraq?” Based on the growth in the ranks of democratic countries in the 1980s and 1990s, they went on to say, “We ought to be fairly optimistic that such change can be has- tened by the right blend of American policies.”59 Many liberal internationalists have been in basic agreement with these hopeful premises of neoconservatism, much as they may try to distance themselves from the particular blend of poli- cies represented by President George W. Bush’s freedom agenda.60 Whether the generic optimistic case for democracy promotion seemed plau- sible in 2000, as Kagan and Kristol asserted, the contemporary quantitative re- search cited in this article suggests a more skeptical assessment for the future. The rationale for blanket democratization is mistaken on two counts: it fails to differentiate sufªciently between partial and full democracy, and it glosses over the challenge of helping authoritarian countries avoid the ªrst and obtain the latter. At issue is not the goal of expanding the number of constitutional representative political systems in the world. Such systems are ªne in concept, but the preponderance of empirical evidence shows that means do not exist to produce more of this type of government consistently from outside. Aware- ness about the alternative likelihood of harmful consequences, especially in the short and medium term, is critical. Democracy promotion will probably always remain an element of U.S. for- eign policy for the historical and political reasons stated at the beginning of this article. But in the future, policymakers should focus less on to design their pro-democracy strategy, and pay greater attention to the literature on democratic transitions, which clearly underscores the difªculty of democratic regime change and the risks of halfway democratiza- tion, such as sectarian conºict, local human rights abuse, dislocated popu- lations, territorial disputes, and transnational terrorism. In particular, more

59. Kagan and Kristol, Present Dangers, p. 20. 60. For an assessment of how liberal internationalism helped to enable Bush’s foreign policy, see Tony Smith, A Pact with the Devil: Washington’s Bid for World Supremacy and the Betrayal of the Ameri- can Promise (New York: Routledge, 2007).

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 International Security 33:2 146

thought needs to be given to how to deal with the prevalence of mixed regimes in the Greater Middle East and to the security problems that this creates, with less reliance on a universal remedy of more democracy to treat these ills. The quantitative studies reviewed here suggest three broad lessons for policymakers. First, only under the rarest of circumstances should military pressure be employed preemptively to advance democracy. In some situations military intervention may be unavoidable, leaving the United States and its al- lies little choice except to try to help another country construct or reconstruct its public institutions. But it would be a fallacy to assume that the result will usually be a moderate pluralistic political system. Second, U.S. foreign policy needs to be adapted better to particular coun- tries’ individual circumstances. This is already being done in the Middle East, according to a recent Congressional Research Service study.61 But rather than an ad hoc approach, which is at odds with leaders’ rhetoric about democ- racy and exposes the United States to charges of hypocrisy and double- dealing, it would be best to confront the issue of mixed regimes openly. Orga- nizational support and electoral assistance could help to consolidate a new democracy, for instance, but be wasted effort or counterproductive in a semi- democracy, where a more effective approach could be to stress the estab- lishment of stronger international linkages that could serve as the base for democratization over the long term. Putting the emphasis on cultural and eco- nomic ties is also a more promising way to engage authoritarian regimes com- pared to menacing them with regime change. Again, this sort of constructive engagement does happen on an improvised basis, but it could be done better with coordination and an acknowledgment of the theoretical foundation for doing so. In general, this approach will not produce quick payoffs, but because potentially productive regime transitions can occur suddenly and unpredict- ably, the United States still must be prepared to adjust its bilateral strategies as circumstances dictate. Third, the United States should adopt a lower proªle. This means limiting the self-righteous oratory about freedom, because it triggers a defensive re- sponse in many corners of the globe that damages U.S. standing and inºuence. There should be a subtle shift in orientation, from campaigning for democracy to supporting it, taking cues from local democratic forces and avoiding one-

61. Jeremy M. Sharp, U.S. Democracy Promotion Policy in the Middle East: The Islamist Dilemma, CRS Report for Congress (Washington, D.C.: Library of Congress, June 15, 2006), Order Code RL 33486, pp. 6–7.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021 Making the World Safe for Partial Democracy? 147

sided efforts to push democratization in directions a foreign country is unpre- pared to go. A lower proªle also means reducing expectations for U.S. citizens so they do not turn against democracy promotion programs that may work at the margins, such as technical assistance for governance reform in certain countries. In the end, rule-bound democracy is largely produced from within, not spread from the outside in a standardized manner. Blustering, all-embracing democracy promotion is not a way to enhance national security because it wastes U.S. resources and can prove counterproductive in furthering the ulti- mate goal, which is to add to the world population of pluralistic majoritarian states. The empirical research on this issue demonstrates that textured support for government reform has a much better chance of serving U.S. national inter- ests than does an all-inclusive freedom agenda.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/isec.2008.33.2.120 by guest on 26 September 2021