Dangerous Drafts? A Time-Series,Cross-National Analysis of Conscription and the Use of Military Force,1946—2001

For example, bills calling for reinstating some form of the draft were introduced both by Republicans and Democrats in the U.S. House of Representatives from 2001 to 2003. The Universal Military Training and Service Act of 2001 (HR 3598) was introduced by Republicans Nick Smith and Curt Weldon and was cosponsored by Republican Roscoe Bartlett. The Universal National Service Act of 2003 (HR 163) was introduced by Democrat Rangel in the House and by Democrat Fritz Hollings in the Senate (S 89). Senator Hagel was an outspoken Republican favoring reinstating some form of conscription at this time, while Representative Murtha (one of two Representatives to vote for HR 163) was a notable Democrat making relatively similar arguments. In contrast, Republican Representatives John J. Duncan and Ron Paul joined Democrats John Conyers, Jr., Cynthia McKinney, and others in cosponsoring H.CON.RES 368 in 2002 that explicitly opposed conscription. The nonpartisan character of this issue had largely evaporated by the time Rangel reintroduced the Universal National Service Act in 2006 and again in 2007. By that time, the bill had scant support and tended to be seen as a symbolic measure in opposition to the war in Iraq. Among others, see "Legislative Background: Recent Action on Compulsory National Service," Congressional Digest 85 (2006): 193-224.

"Legislative Background," 206.

Seung-Whan Choi and Patrick James , "No Professional Soldiers, No Militarized Interstate Disputes? A New Question for Neo-Kantianism," Journal of Conflict Resolution 47 ( 2003): 796-816; Joseph Paul Vasquez III , "Shouldering the Soldiering: Democracy, Conscription, and Military Casualties," Journal of Conflict Resolution 49 (2003): 849-73.

Fordham and Sarver spell out the conceptual problems associated with equating military force used against other state governments (dyadic force) with the more general category of state use of force, which can include actions against nonstate actors and various "friendly" uses of force such as humanitarian missions. Benjamin O. Fordham and Christopher C. Sarver, "Militarized Interstate Disputes and US Uses of Force," International Studies Quarterly 45 (2001): 455-66.

Charles Moskos , "What Ails the All-Volunteer Force: An Institutional Perspective," Parameters (Summer 2001): 29-47; Michael E. O’Hanlon , Defense Strategy for the Post-Saddam Era (Washington, DC : Brookings Institution, 2005).

Immanuel Kant , Kant’s Political Writings, ed. Hans Reis , trans. H. B. Nesbit (Cambridge, UK: Cambridge University Press, 1963), 94.

Ibid., 95.

Mark J. Peceny and Caroline C. Beer , with Shannon Sanchez-Terry , "Dictatorial Peace?" American Political Science Review 96 (2002): 15-26.

See Barbara Geddes, "Stages of Development in Authoritarian Regimes," in World Order after Leninism, ed. Vladimir Tismaneanu, Marc M. Howard, and Rudra Sil (Seattle: University of Washington Press, 2006), 161-63; and Barbara Geddes, Paradigms and Sand Castles: Theory Building and Research Design in Comparative Politics (Ann Arbor: University of Michigan Press, 2003). Geddes observes that in personalist dictatorships, rulers tend to politicize the military by recruiting elite political supporters into leadership positions within the military. These supporters may represent important political or ethnic groups within society. Furthermore, she finds that from 1945 to 1999 nearly half of all authoritarian regimes created by militaries or political parties were fully or partially personalized within three years of the regime’s creation. "Stages of Development," 164. For an example of the personalization of an African military, see Kisangani N. F. Emizet, "Explaining the Rise and Fall of Military Regimes: Civil-Military Relations in the Congo," Armed Forces & Society 26(2000): 203-27.

Samuel P. Huntington, The Soldier and the State: The Theory and Politics of Civil-Military Relations (Cambridge, MA: Harvard University Press, 1957); Richard K. Betts, Soldiers, Statesmen, and Cold War Crises (Cambridge, MA: Harvard University Press, 1977); Morris Janowitz, "The Logic of National Service," in The Military Draft: Selected Readings on Conscription, ed. Martin Anderson and Barbara Honegger (Stanford, CA: Hoover Institution Press, 1982); Peter D. Feaver and Christopher Gelpi, Choosing Your Battles: American Civil-Military Relations and the Use of Force (Princeton, NJ: Princeton University Press, 2004).

Lawrence J. Korb, "Military Manpower Training Achievements and Challenges for the 1980s," in The All-Volunteer Force after a Decade: Retrospect and Prospect, ed. William Bowman, Roger Little, and G. Thomas Sicilia (New York: Pergamon-Brassey’s, 1986); Seung-Whan Choi and Patrick James, Civil-Military Dynamics, Democracy, and International Conflict (New York: Palgrave, 2005); John Byron, "The Failure of the All-Volunteer Force," Proceedings of the US Naval Institute 132 (2006): 12-14. The bulk of the literature on the U.S. all-volunteer force claims it is overwhelmingly made up of individuals from lower socioeconomic classes. See Moskos, "What Ails"; Vasquez, "Shouldering." Bandow and O’Hanlon, however, argue that it may be more representative of U.S. society than is typically claimed. See O’Hanlon, Defense Strategy; Doug Bandow, "Mend, Never End, the All-Volunteer Force," Orbis 43 (2000): 463-75.

Bandow, "Mend, Never End."

Ibid., 470; Choi and James, Civil-Military Dynamics. White finds that conscript armies are on average 34 percent larger than volunteer armies after controlling for a number of factors that affect military size. Michael D. White, "Conscription and the Size of Armed Forces," Social Science Quarterly 70 (1989): 772-81.

Denis Hayes , Conscription Conflict: The Conflict of Ideas in the Struggle for and against Military Conscription in Britain between 1901 and 1939 ( New York: Garland, 1973), 346.

Steven Morse , "Should the All-Volunteer Force Be Replaced by Universal, Mandatory National Service?" Congressional Digest 85 (2006): 219.

Leonard Berkowitz, Aggression: Its Causes, Consequences, and Control (Philadelphia: Temple University Press, 1993). I thank an anonymous reviewer for noting the relevance of Berkowitz’s ideas to this argument.

Roger W. Benjamin and Lewis J. Edinger, "Conditions for Military Control over Foreign Policy Decisions in Major States: A Historical Explanation," Journal of Conflict Resolution 15 (1971): 5-31; Graham T. Allison, Albert Canesale, and Joseph S. Nye, Jr., Hawks, Doves, and Owls: An Agenda for Avoiding Nuclear War (New York: Norton, 1985); Guy D. Whit-ten and Henry S. Bienen, "Political Violence and Time in Power," Armed Forces & Society 23 (1996): 209-34; Julian Schofield, "Militarized Decision-Making for War in Pakistan, 1947-1971," Armed Forces & Society 27 (2000): 131-48.

Seung-Whan Choi and Patrick James , "Civil-Military Relations in a Neo-Kantian World, 1886-1992," Armed Forces & Society 30 (2004): 229.

David H. Marlowe , "The Manning of the Force and the Structure of Battle," in ed., Conscripts and Volunteers, ed. Robert K. Fullinwider (Totowa, NJ: Rowman and Allan-held ), 50.

Byron , "Failure"; Charles Moskos, "Should the AVF be Replaced by Universal, Mandatory National Service?" Congressional Digest 85 ( 2006): 220-22.

Bandow, "Mend, Never End"; James Fallows, "Why the Country Needs It," in The Military Draft, ed. Jason Berger (New York: H. W. Wilson, 1981), 75-82; John G. Kester, "The Reasons to Draft," in The All-Volunteer Force After a Decade: Retrospect and Prospect, ed. William Bowman, Roger Little, and G. Thomas Sicilia (New York: Pergamon-Brassey’s, 1986), 286-315. Moskos maintains that "there is clear relationship between socio-educational background and soldierly performance." Charles Moskos, "National Service and the All-Volunteer Force," in The Military Draft, ed. Jason Berger (New York: H. W. Wilson Company, 1981), 36. Also see Marlowe, "The Manning," 50; Kester, "Reasons," 295.

As measured by Armed Forces Qualification Test scores and high school graduation rates. Byron, "Failure"; Moskos, "Should the AVF be Replaced"; Ann Scott Tyson, "A Historic Success in Military Recruiting: In the Midst of Downturn, All Targets Are Met," Washington Post, October 14, 2009.

Choi and James, "No Professional Soldiers," 802. Fick takes the opposite view. Nathaniel Fick, "A Draft Will Lower the Quality of the Military," in Military Draft, ed. George Milite (New York: Greenhaven, 2007), 36-42.

Charles Moskos, "The Marketplace All Volunteer Force: A Critique," in The All-Volunteer Force after a Decade: Retrospect and Prospect, ed. William Bowman, Roger Little, and G. Thomas Sicilia (New York: Pergamon-Brassey’s, 1986), 17; Choi and James, "No Professional Soldiers."

This article does not directly address the possibility that there may be a reciprocal relationship between conscription and interstate conflict. In other words, it is theoretically possible that conscription leads to a higher propensity to become involved in interstate conflict and that involvement in interstate conflict simultaneously tends to increase decision makers’ proclivity to implement conscription. To increase confidence that causality does in fact run from military manpower systems (MMSs) to the use of force, I ran all of my statistical estimates after lagging my independent variable conscription two and then five years. The substance of my findings remains the same.

W. Lee Hansen and Burton A. Weisbrod , "Economics of the Military Draft," Quarterly Journal of Economics 81 (1967): 395-421; Milton Friedman , An Economist’s Protest: Columns in Political Economy ( Glen Ridge, NJ: Thomas Horton, 1972 ).

Dwight R. Lee and Richard B. McKenzie , "Re-examination of the Relative Efficiency of the Draft and the All-Volunteer Force," Southern Economic Journal 58 (1992): 644-54; John T. Warner and Beth J. Asch , "The Economics of Military Manpower," in Handbook of Defense Economics, ed. Keith Hartley and Todd Sandler (Amsterdam: Elsevier , 1995), 347-98.

Choi and James , "No Professional Soldiers."

Vasquez , "Shouldering"; Choi and James , "No Professional Soldiers."

White , "Conscription."

Vasquez, "Shouldering," 870. Research on the long-term impact that MMSs have on the use of force is similarly divided. Feaver and Gelpi find that the U.S. propensity to initiate militarized interstate disputes (MIDs) declined significantly when the number of veterans in the U.S. Congress increased during the period 1816 to 1992. Choi and James, however, reach the opposite conclusion in a cross-national study of MID involvement from 1886 to 1992. Feaver and Gelpi, Choosing Your Battles; Choi and James, "Civil-Military Relations."

Choi and James have limited utility for understanding state use of force for a second reason. They analyze MID involvement rather than MID initiation as their dependent variable. Decision makers can become involved with MIDs by being drawn into ongoing conflicts rather than consciously choosing to initiate armed force. Choi and James, "No Professional Soldiers." Choi and James look specifically at MID initiation in a 2008 article, but again analyze only dyadic disputes. Seung-Whan Choi and Patrick James, "Civil-Military Structure, Political Communication, and the Democratic Peace," Journal of Peace Research 45 (2008): 37-53.

Fordham and Sarver , "Militarized Interstate Disputes."

A host of other examples could be given, ranging from Chinese intervention into Burma to attack Kuomintang troops active on the Chinese-Burmese border in 1951 to Rwandan military missions against Hutu militias operating in eastern Congo after 1994.

For further information on the International Military Intervention (IMI) database, see note 37.

Note that many former colonies enter my sample decades after 1946. After accounting for these former colonies, there are still 1,096 missing observations in my sample. The bulk of these (726) are a result of missing information in my control variable troop quality. As a robustness check on my results, I removed this control variable and reran my estimates. The results were substantively the same.

On IMI see Jeffrey Pickering and Emizet F. Kisangani, "The International Military Intervention Data Set: An Updated Resource for Conflict Scholars," Journal of Peace Research 46 (2009): 589-99; Frederic S. Pearson and Robert A. Baumann, "International Military Intervention, 1946-1988" (Inter-University Consortium for Political and Social Research [ICPSR], Data Collection 6035, University of Michigan, 1993); Emizet F. Kisangani and Jeffrey Pickering, "International Military Intervention, 1989-2005" (ICPSR, Data Collection 21282, 2008). Unilateral force is defined as a mission under the control of the nation’s military high command. Unilateral actions may be taken in coordination with other states’ armed forces. These operations are termed multilateral actions. In contrast, multinational operations are missions where two or more states operate under the command of a single nation’s military high command or under the command of an international or regional organization. See Pickering and Kisangani, "International Military Intervention Data Set," 593; Patrick Regan, Civil Wars and Foreign Powers: Outside Intervention in Interstate Conflict (Ann Arbor: University of Michigan Press, 2002), 102)

Pearson and Baumann , "International Military Intervention," iii.

IMI is an event data set on military intervention that records a wide range of information on these incidents, from actors, targets, and initiation and termination dates to specific characteristics of military operations. As is common in international conflict events data, IMI codes a new event for each country that is targeted and for each actor intervening in the country, including multinational actors. For example, the U.S. use of force against Iraq in Kuwait in 1991 is coded as a separate event than the U.S. invasion of Iraq the same year. IMI also includes multinational operations against states, but these are not included in the present analysis. The data set does not aggregate multilateral uses of force. IMI includes 1,115 military interventions from 1946 to 2005. The present study examines large-scale military interventions with over one thousand intervening troops from 1946 to 2001.

For studies using IMI, among others see Mark J. Peceny, "Forcing Them to Be Free?" Political Research Quarterly 52 (1999): 549-82; Dan Reiter, "Does Peace Nurture Democracy?" Journal of Politics 63 (2001): 935-48; Matthew Krain, "International Intervention and the Severity of Genocides," International Studies Quarterly 49 (2005): 363-87; Emizet Kisangani and Jeffrey Pickering, "Diverting with Benevolent Military Force," International Studies Quarterly 51 (2007): 277-300.

On the conceptual breadth of IMI, see Charles W. Kegley, Jr. and Margaret G. Hermann, "Putting Military Intervention into the Democratic Peace: A Research Note," Comparative Political Studies 30 (1997): 81; Peceny, "Forcing Them," 560.

Countries with both types of MMSs may also have highly trained "special" forces available for smaller missions. On the one-thousand-soldier threshold for large-scale interventions, see Jeffrey Pickering and William R. Thompson, "Stability in a Fragmenting World: Interstate Military Force, 1946-1988," Political Research Quarterly 51 (1998): 241-63.

Among others, Krain, "International Intervention," and Pickering and Thompson, "Stability in a Fragmenting World," use the same operationalization of belligerent military force.

Diehl observes that IMI is relatively unique among quantitative conflict data sets in coding contentious issues. His article on this topic, of course, predates Issue Correlates of War database. Paul F. Diehl, "What Are They Fighting For? The Importance of Issues in International Conflict Research," Journal of Peace Research 29 (1992): 333-44.

The number of humanitarian missions is small because a large proportion of the humanitarian operations in IMI are undertaken by international organizations such as the United Nations or the African Union. Also, many unilateral missions of this type include fewer than one thousand troops.

See Choi and James, "No Professional Soldiers," 803. Bart Horeman and Marc Stolwijk, eds., Refusing to Bear Arms: A World Survey of Conscription and Conscientious Objection to Military Service (London: War Resister’s, 1998); Devi Prasad and Tony Smythe, eds., Conscription: A World Survey (London: War Resister’s, 1968). The annual Military Balance publication from the International Institute for Strategic Studies (Oxford, UK: Oxford University Press, 1970-2003) was used to cross-check these sources.

See J. David Singer, "Reconstructing the Correlates of War Dataset on Material Capabilities of States, 1981-1985," International Interactions 14 (1987): 115-32; Scott D. Bennett and Allan Stam, "EUGene: A Conceptual Manual," International Interactions 26 (2000): 179-204. Bennett and Stam describe a prominent computer program used to generate data sets for quantitative studies of international conflict. In doing so, they describe the most common operationalizations of control variables used in the quantitative international conflict literature, including the operationalizations of capabilities, polity score, war, and civil war employed here.

Monty Marshall and Keith Jaggers , "Polity IV Project," www.cidcm.umd.edu/polity (accessed November 12, 2007).

On the common practice of subtracting the Polity IV autocracy score from the democracy score in the quantitative literature, see, among others, Gretchen Casper and Claudiu Tufis, "Correlation versus Interchangeability: Empirical Findings on Democracy Using Highly Correlated Data Sets," Political Analysis 11 (2003): 197-98. As a robustness check, I reran all of my statistical analyses with the 11-point Polity IV democracy index instead of polity2. The results were substantively the same.

Among others, see John MacMillan, "Beyond the Separate Democratic Peace," Journal of Peace Research 40 (2003): 233-43.

Vasquez , "Shouldering the Soldiering," 860; Dan Reiter and Allan C. Stam , Democracies at War (Princeton, NJ: Princeton University Press, 2002).

Among others, see Benjamin O. Fordham, "Another Look at Parties, Voters, and the Use of Force Abroad," Journal of Conflict Resolution 46 (2002): 572-96.

See Harvey Starr and Randolph M. Siverson, "Cumulation, Evaluation, and the Research Process: Investigating the Diffusion of Conflict," Journal of Peace Research 35 (1998): 231-37.

ACD version 3.0 (2004) was used since version 4.0 (2007) does not include war as a separate category in the intensity variable. Håvard Strand, Lars Willhelmsen, and Nils Peter Gleditsch, "Armed Conflict Database," version 3.0 (Oslo: International Peace Research Institute, 2004).

I employ Correlates of War’s (COW) standard demarcation of six regions: the Americas, country codes 1-199; Europe, country codes 200-399; Africa, country codes 400-599; the Middle East, country codes 700-899; and Oceania, country codes 900-999. A more narrowly defined categorization of subsystems based on regions in the International Crisis Behavior dataset was also used in supplemental analyses. The substantive results were similar. Michael Brecher and Jonathan Wilkenfeld, A Study of Crisis (Ann Arbor: University of Michigan Press, 2000).

The following descriptive statistics summarizing my control variables include the mean, standard deviation, minimum, and maximum for each variable, respectively: capabilities: 0.008, 0.023, 1.00e-05, 0.364; polity2: -334, 7.542, -10, +10; troop quality: 8.435, 1.542, 0, 14.700; war: 0.0339, 0.180, 0, 1; civil war: 0.107, 0.310, 0, 1; regional crisis: 1.750, 2.041, 0, 10.

Since the use of military force is an uncommon phenomenon, I also employ zero-inflated negative binomial (ZINB) regressions to estimate my first two models (in Table 1). Since the substantive results for negative binomial and ZINB estimates are the same and various bayesian information criterion scores strongly favor the standard negative binomial regressions, I present results from my standard negative binomial estimates akaike information criterion scores comparing the two estimators were indeterminate. ZINB regressions would not converge for my last two models (in Table 2), presumably because of limited variation in the dependent variables. I converted these dependent variables to binary variables and estimated the models with relogit. Since my original negative binomial estimates, logit estimates, and relogit estimates are substantively the same, I present negative binomial regressions below.

Thomas Plümper , Vera E. Troeger , and Philp Manow , "Panel Data Analysis in Comparative Politics: Linking Theory to Method," European Journal of Political Research 44 ( 2005): 327-54.

Since the COW military personnel variable is a component of both capabilities and troop quality, it cannot be included in the full estimate without generating high levels of multicollinearity. In addition, since capabilities and troop quality are created with overlapping COW information, I ran them separately in each of my estimates in supplemental analyses with no change in the results. As noted above, variance inflation factor scores are uniformly low in my models even with the inclusion of both capabilities and troop quality.

Choi and James , "Civil-Military Relations," 236.

I tested the robustness of each of the models in Tables 1 and 2 by identifying influential cases in each model, removing them from my data, and rerunning the estimates. Pregibon D-Beta, deviance, Hosmer and Lemeshow Delta chi-squared, and Hosmer and Lemeshow Delta-D influence statistics were used to identify influential cases. For example, in my belligerent intervention model in Table 1, Libya’s intervention into Chad in 1979, Pakistan’s use of force in Afghanistan in 1989, and Somalia’s incursion into Ethiopia in 1977 were identified as influential cases. My independent variable conscription remains statistically significant after these cases are omitted from the estimate. Conscription’s z score actually increases from 1.77 to 2.03 when these variables are removed. Conscription retains its statistical significance when influential cases are removed from the operations other than war (Table 1) and nonstate operations (Table 2) estimates as well.

See Nicholas J. Wheeler, "Introduction: The Political and Moral Limits of Western Military Intervention to Protect Civilians," in The Dimensions of Western Military Intervention, ed. Colin McInnes and Nicholas J. Wheeler (Portland, OR: Frank Cass, 2002); William A. Boettcher III, "Military Intervention Decisions Regarding Humanitarian Crises," Journal of Conflict Resolution 48 (2004): 331-55.

Wheeler , "Introduction," 7.

Andrea Kathryn Talentino, Military Intervention after the Cold War (Athens: Ohio University Press, 2005), 282; Dominick Donald, "The Doctrine Gap: The Enduring Problem of Contemporary Peace Support Operations Thinking," in The Dimensions of Western Military Intervention, ed. Colin McInnes and Nicholas J. Wheeler (Portland, OR: Frank Cass, 2002), 107-39.

Pickering and Kisangani , "International Military Intervention Data Set," 595.

See Boettcher, "Military Intervention Decisions"; and Joseph Lepgold, "NATO’s Post-Cold War Collective Action Problem," International Security 23 (1998): 78-106.

Stephen Biddle , Military Power: Explaining Victory and Defeat in Modern Battle ( Princeton, NJ: Princeton University Press , 2004).

Vasquez , "Shouldering," 870.

Choi and James , "No Professional Soldiers" ; Vasquez , "Shouldering."