MOBASHER 4/18/2016 3:11 PM

UNDERSTANDING ETHNIC-ELECTORAL DYNAMICS: HOW ETHNIC POLITICS AFFECT ELECTORAL LAWS AND ELECTION OUTCOMES IN AFGHANISTAN

*Mohammad Bashir Mobasher

After the fall of the Taliban in 2001, Afghanistan began transitioning toward democracy. The democratization, which began with the Bonn Agreement, was very promising for all ethnic groups in Afghanistan. The Bonn Agreement provided that Afghanistan must establish “broad based, gender sensitive, multi- ethnic and fully representative government.”1 As a part of this agreement, Afghanistan adopted a new Constitution in 2004, which established elections as the foundation for representative government and guaranteed that the elections be inclusive, fair, and representative of all groups and regions. Despite this progress in the laws, Afghanistan has not been able to reduce ethnic tensions. In fact, this article shows that ethnic tensions tend to intensify during elections, a result that belies the vision of the Bonn Agreement. After presenting a statistical and qualitative analysis of voting practices across three presidential elections and 26 provinces, this article makes several observations about the role of ethnicity in Afghan presidential elections. These observations are: (1) people of an ethnic community are more likely to vote for a candidate of their own ethnic group; (2) voters from one ethnic group are likely to vote for a candidate from another ethnic group only when they do not have a candidate of their own; (3) cross ethnic voting is more likely to occur in blocs and in exchange for patronage to the elites of voting groups; and (4) candidates and ethno-political elites tend to prioritize their electoral campaigns by ethnic mobilization, manipulation of laws and political bargains rather than by policy development in order to generate votes. This article contributes to the ongoing legal and political discourse on reforming the Constitution and the electoral laws, revealing the danger of neglecting the role of ethnicity in Afghan politics.

* Ph.D. Candidate, University of Washington, School of Law. 1. Agreement on Provisional Arrangements in Afghanistan Pending the Re- Establishment of Permanent Government Institutions (Dec. 5, 2000), http://www.un.org/ News/dh/latest/afghan/afghan-agree.htm [hereinafter “Bonn Agreement”].

355 MOBASHER 4/18/2016 3:11 PM

356 GONZAGA LAW REVIEW Vol. 51:2

TABLE OF CONTENTS

INTRODUCTION ...... 357

I. IS ETHNICITY LOSING ITS INFLUENCES ON ELECTION OUTCOMES? CURRENT UNDERSTANDINGS AND SCHOLARSHIPS ...... 359 A. Electoral Laws, Electoral Strategies, and Cross-Ethnic Coalitions ...... 362 1. Electoral Strategies of Candidates, Elites, and Voters ...... 364 a) Candidates ...... 364 b) Elites ...... 365 c) Voters ...... 367 2. The Pattern of Cross-Ethnic Voting in Afghanistan ...... 369

II. THE EMPIRICS OF ETHNIC-ELECTORAL DYNAMICS ...... 370 A. Ethnic Demographics in Afghanistan ...... 371 B. Role of Ethnicity on Election Outcomes: Nationwide ...... 373 1. The Presidential Election of 2004 ...... 374 2. The Presidential Election of 2009 ...... 377 3. The Presidential Election of 2014 ...... 383 C. Province-Wide Analysis of Ethnic-Electoral Relations ...... 387 1. Electoral Politics in the 2004 and 2009 Presidential Elections ...... 388 a) Hamid Karzai, The Unparalleled Pashtun Candidate in the 2004 and 2009 Elections ...... 388 b) Qanooni and Abdullah, The Tajik Runner Ups...... 394 c) The Third Placed Hazaras: Mohaqiq and Dr. Bashardost ...... 398 d) Abdul Rashid Dostum: The Uzbek Powerbroker ...... 400 2. The Presidential Election of 2014 ...... 402 a) Dr. Abdullah: The Apparent Frontrunner ...... 402 b) Dr. Ashraf Ghani: The Emerging Elite ...... 405

III. ALTERNATIVE EXPLANATIONS TO ETHNIC-ELECTORAL DYNAMICS ... 409 A. Insecurity and Turnout Rate ...... 410 B. Electoral Irregularities ...... 410 C. Localism ...... 411

CONCLUSION ...... 412 APPENDIX ...... 414 MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 357

INTRODUCTION

After the fall of the Taliban in 2001, Afghanistan began transitioning toward democracy. This transition began with the Bonn Agreement, which provided that Afghanistan must establish “broad based, gender sensitive, multi-ethnic and fully representative government.”2 As a part of this agreement, Afghanistan adopted a new Constitution in 2004, under which sovereignty belongs to people and citizens exercise their sovereignty through their representatives.3 Their representatives include an elected president,4 elected legislators,5 elected mayors,6 and elected members of provincial councils, district councils, village councils,7 and municipality councils.8 Recognizing the ethnic plurality in Afghanistan, the Constitution guarantees that the elections be inclusive, fair, and representative of all groups and regions.9 This mandate is particularly important in Afghanistan, where political contests are characterized by group demands, struggles, and even violence.10 This and similar electoral rules have been expected to rearrange the field for political

2. Id. 3. Article 4 of the Constitutions provides that “[n]ational sovereignty in Afghanistan shall belong to the nation, manifested directly and through its elected representatives.” THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN [QANUN ASSASSI JUMHURI ISLAMAI AFGHANISTAN] Jan. 26, 2004, art. 4. 4. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 61; see also ELECTION LAW [QANUN INTIKHABAT] Jun. 8, 2014 [15/5/1392], S. 1112, arts. 1, 10 (Afg.). 5. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 83; see also ELECTION LAW Jun. 8, 2014, S. 1112, arts. 1, 10, 25, 26, 27. 6. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 141; see also ELECTION LAW Jun. 8, 2014, S. 1112, arts. 38-39. 7. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, arts. 138, 140; see also ELECTION LAW Jun. 8, 2014, S. 1112, arts. 1, 10, 29, 32, 35. 8. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 141; see also ELECTION LAW Jun. 8, 2014, S. 1112, arts. 40-42. 9. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 83; see also ELECTION LAW Jun. 8, 2014, S. 1112, arts. 2, 50, 59, 61 (These Constitutional mandates were also reflected in the Election Law of Afghanistan.). 10. RASUL BAKHSH RAIS, WAR WITHOUT WINNERS: AFGHANISTAN’S UNCERTAIN TRANSITION AFTER THE COLD WAR 158-163, 230 (1994). MOBASHER 4/18/2016 3:11 PM

358 GONZAGA LAW REVIEW Vol. 51:2 contestations11 and pave the way for conflicting groups “to renegotiate power . . . in a non-violent manner.”12 Despite all this progress in the laws, they have not been able to reduce ethnic tensions or cross-ethnic gaps. In fact, ethnic tensions tend to intensify during the elections.13 This article shows that candidates and elites have continued to mobilize ethnic groups; furthermore, ethnic groups have tended to vote collectively and on the basis of ethnicity. After presenting an analysis of data about voting practices, this article makes the following observations about the role of ethnicity in Afghan presidential elections: (1) people of an ethnic community are more likely to vote for a candidate of their own ethnic group; (2) voters are likely to vote for a candidate from another ethnic group only when they do not have a candidate of their own; (3) cross ethnic voting is more likely to occur in blocs and in exchange for patronage to the elites of voting groups; and (4) candidates and ethno-political elites tend to prioritize their electoral campaigns by ethnic mobilization and political bargains, rather than by policy development, in order to generate votes. In this situation, the extent to which an elected individual can adequately represent the whole nation as a president is debatable. This article offers an interdisciplinary analysis, examining the legal and political aspects of ethnic-electoral dynamics. The analysis reveals that, although some electoral rules do encourage cross-ethnic coalitions in the presidential elections, ethnic voting continues to dictate the course of elections. It further reveals that ethnic politics has consistently undermined electoral laws, including constitutional principles, in all presidential elections. Part I of this article establishes the theoretical framework of ethnic-electoral links, explaining current understandings and scholarship about ethnicity and elections in Afghanistan. It also discusses the relevant Afghan electoral laws and provides some background on why candidates, elites, and voters may be inclined to manipulate elections through ethnic mobilization and cross-ethnic coalitions. Next, in Part II, the article provides an empirical analysis of ethnic-electoral

11. CARINA PERELLI & SCOTT SMITH, ANTICIPATING AND RESPONDING TO FRAUD IN THE 2014 AFGHAN ELECTIONS, CHATHAM HOUSE 3-4 (Feb. 2014), http://www.chat hamhouse.org/sites/files/chathamhouse/home/chatham/public_html/sites/default/files/20140 200AfghanistanElectionsFraudPerelliSmith.pdf. 12. Id. at 4. 13. INT’L CRISIS GRP., Asia Report N*260: Afghanistan’s Political Transition 2 (OCT. 16, 2014), http://www.crisisgroup.org/~/media/Files/asia/south-asia/afghanistan/260-afghan istan-s-political-transition.pdf; NAT’L DEMOCRATIC INST., Political Parties in Afghanistan: A Review of The State of Political Parties After 2009 and 2010 elections 28 (June 2011), https://www.ndi.org/files/Afghanistan-political-parties-july-2011.pdf; Frud Bezhan, Ahead Of Presidential Vote, Afghan Political Forces Divide Along Ethnic Lines, RADIO FREE EUROPE, Nov. 15, 2015, http://www.rferl.org/content/afghan-politics-feature/25101500.html. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 359 dynamics, based on electoral and demographic data collected from three presidential elections and 26 provinces of Afghanistan. This part tests all the theories and conceptual frameworks set forth in the first part. The analysis includes statistical and qualitative methods. While the statistical analysis mainly examines the relations between ethnicity and election outcomes, cross-ethnic voting and the legal dimensions of ethnic-electoral dynamics are explained qualitatively.14 Part III analyzes the role of intervening variables, such as localism, electoral irregularities, turnout rates, and security threats in the elections. Finally, the article concludes by setting the stage for further research and recommendations.

I. IS ETHNICITY LOSING ITS INFLUENCES ON ELECTION OUTCOMES? CURRENT UNDERSTANDINGS AND SCHOLARSHIPS

In 2014, just prior to the most recent presidential election, a survey conducted by the Free & Fair Election Foundation of Afghanistan (FEFA) concluded that 64% of voters would not vote on the basis of ethnicity.15 A separate pre-election poll conducted by Glevum Associates suggested a similar finding: 84% of respondents said that they would not cast ethnic votes.16 In some interviews conducted by BBC Persian, Tolonews, and other media, respondents said that they would vote on account of the personality, leadership skills, or policies of the candidates.17 Likewise, Ahmad Yusif Nuristani, the head of Independent Election Commission (IEC), once stated that “the votes are totally

14. This article uses correlation, regression, and graphs to examine the association between the ethnicity of candidates and their votes and compare their votes with the votes of their ethnic counterparts in other elections. In cases of high level of cross-ethnic voting in some provinces, the article process traces the intervening causes, such as political bargains, electoral manipulation, and security threats. It also uses causal mechanisms and process tracing, to test whether the electoral laws and systems led to ethnic-oriented rather than policy- oriented elections. 15. Masood Korosh, Promising Findings of FEFA Survey, DAILY OUTLOOK AFGHANISTAN (Mar. 29, 2014), http://outlookafghanistan.net/topics.php?post_id=9754 #ixzz35nf7DiQF. 16. Afghanistan Presidential Election 2014 Poll Results (Wave 1), GLEVUM ASSOCIATES 17 (Dec. 2013), http://glevumassociates.com/doc/GlevumAfghanistan PresidentialElection2014WaveOneSurveyFindings%20.pdf. 17. Bashir Payman, Intikhabat-e-Afghanistan: Aya Meyar Hai Madum Taghir Karda? [Afghan Election: Whether People Have Changed the Selection Criteria?], BBC PERSIAN (Mar. 31, 2014), http://www.bbc.co.uk/persian/afghanistan/2014/03/140330_ k03_af2014_bloc_voting_in_afghanistan.shtml?s. MOBASHER 4/18/2016 3:11 PM

360 GONZAGA LAW REVIEW Vol. 51:2 mixed and national.”18 These pre-election polls and statements were supported by Ashok K. Mehta who, in Elections in Afghanistan: The First Democratic Transition, wrote that in 2014 elections “Afghans did not vote along ethnic lines . . . .”19 This mode of optimism existed in the previous elections also. Before the 2009 election, a report by the International Council on Security and Development concluded that its “field research has . . . revealed a remarkable absence of concern over the President’s ethnic background . . . .”20 Similarly, in their survey of voters in the district of Dasht-Barchi, the western part of Kabul, Cohen and Larson found that most respondents believed that the 2009 presidential election would be less influenced by ethnicity than the previous election because “there was more ‘awareness’ among the people.”21 Cohen and Larson suggested that urbanization prompted this positive social transition.22 Likewise, in his 2009 article, Ethnicity, Space, and Politics in Afghanistan, Benjamin Dubow surmised that there have been more cross-ethnic votes in 2009 than in 2004.23 While he posited that such cross-ethnic voting may be because of political bargains, his analysis did not go further to substantiate it with evidence. All these findings suggest a growing optimism about the reduced role of ethnicity in elections. I suspect that the ethnic tension following the runoff election in 2014, when Dr. Abdullah’s supporters marched on the streets of Kabul with gunfire,24 should have shattered these optimisms. If not, I present

18. Saleha Sadat, Election Shows Potential for Politics Beyond Ethnicity, TOLONEWS, Apr. 29, 2014, http://www.tolonews.com/en/afghanistan/14704-election-shows- potential-for-politics-beyond-ethnicity. 19. Ashok K. Mehta, Elections in Afghanistan: The First Democratic Transition, FRIDRICH EBERT STIFUNG 1 (2014), http://library.fes.de/pdf-files/bueros/indien/10741.pdf. 20. Afghanistan’s Presidential Election: Power To The People, Or The Powerful?, THE INT’L COUNCIL ON SEC. AND DEV. 5 (Apr. 2009), http://www.icosgroup.net /static/reports/power_to_the_people.pdf [hereinafter “Decision Point 2009”]. 21. Noah Coburn & Anna Larson, Voting Together: Why Afghanistan’s 2009 Elections were (and were not) a Disaster, AFGHANISTAN RESEARCH AND EVALUATION UNIT 10 (Nov. 2009), http://www.areu.org.af/Uploads/EditionPdfs/937E-Voting%20Together%20 Why%20Afghanistans%202009%20Elections%20were%20(and%20were%20not)%20a%20 Disaster-BP.pdf. 22. Id. at 8. 23. Benjamin Dubow, Ethnicity, Space, and Politics in Afghanistan 42 (Nov. 1, 2009) (Senior Seminar Paper, University of Pennsylvania Scholarly Commons), http://repository.upenn.edu/cgi/viewcontent.cgi?article=1012&context=senior_seminar. 24. KENNETH KATZMAN, CONG. RESEARCH SERV., RS21922, AFGHANISTAN: POLITICS, ELECTIONS, AND GOVERNMENT PERFORMANCE 28 (2015), https://www.fas.org/sgp/crs/ row/RS21922.pdf. After the preliminary result was released by IEC declaring Ghani as the winner, “Abdullah’s supporters gathered the next day at the Loya Jirga hall on the campus of the Polytechnic University of Kabul, many of them heavily armed. While his followers MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 361 substantial evidence in this article to reject these optimistic analyses. In fact, the data examined in this article suggests that these polls and findings were likely subject to response bias. It is more likely than not that survey respondents have not freely admitted that they vote on the basis of ethnicity. Typically, respondents hide their biases and censor their responses accordingly.25 Therefore, it is not surprising that a majority of the respondents in the FEFA survey reported that they vote on the basis of qualities and policies of candidates, 26 even though the evidence presented in this article suggests otherwise. Some analysts and authors like Kenneth Katzman, Waliullah Rahmani, and Martine Van Bijlert have recognized the persistent role of ethnic division on election outcomes. For instance, Kenneth Kartzman, noting ethnic tensions during 2014 elections, wrote that ethnic and familial ties often supersede ideological affiliations and political views.27 Waliullah Rahmani, the director of Center for Strategic Studies, posited that, “Ethnocentric politics has been a dominant discourse in Afghanistan. . .[and] that in the 2014 election ethnocentric alignments will play a major and important role.”28 Van Bijlert, who wrote several seminal articles about Afghan elections, seemed to agree with Katzman and Rahmani. Analyzing the presidential election of 2009 in his article titled, How to Win an Afghan Election: Perceptions and Practices, he suggested that voters vote for their elites as long as “it will help position their ethnic group as a force to be reckoned with.”29 This article takes a similar position but breaks away from them on two major points. First, this article, unlike Van Bijlert’s and Katzman’s, does not treat ethnic division as one among many variables but as a prevailing factor that influence the election outcomes. Second, the existing literature has not provided a theoretical framework for how ethnic division and mobilization work during the elections and when cross-ethnic bargains are effective. This article is an attempt to fill that gap by providing a comprehensive theory about ethnic- electoral dynamics in Afghanistan. shouted at him to declare a ‘parallel state’ . . . [s]ome of his supporters . . . allegedly made preparations to seize control of government buildings in at least three provinces and occupy the presidential palace in Kabul.” INT’L CRISIS GRP., supra note 13, at 18. 25. Richard B. Warnecke et al., Improving Question Wording in Surveys of Culturally Diverse Populations, 7 ELSEVIER 334, 336 (1997), http://www.tree4health.org/distance learning/sites/www.tree4health.org.distancelearning/files/readings/Warenke-wording- cultural.pdf. 26. Korosh, supra note 15. 27. KATZMAN, supra note 24, at 2. 28. Bezhan, supra note 13. 29. Martine van Bijlert, How to Win an Afghan Election: Perceptions and Practices, AFGHANISTAN ANALYSTS NETWORK 17 (Aug. 2009), http://www.afghanistan-analysts.org/wp- content/uploads/2013/07/20090809-how-to-win-an-afghan-election.pdf. MOBASHER 4/18/2016 3:11 PM

362 GONZAGA LAW REVIEW Vol. 51:2

The most important contribution of this article is its timely correspondence to the danger of the optimist perspective mentioned earlier. The danger of this perspective does not come from the findings themselves, but from the resulting recommendations. It is likely that these recommendations would disregard the crucial role of ethnicity on the elections; hence, they will be destructive if they strongly influence the ongoing political efforts of amending the Constitutions and reforming the election laws. Though this article does not provide any recommendations, it does inform the reformers about the danger of neglecting the significant role of ethnicity on election outcomes.

A. Electoral Laws, Electoral Strategies, and Cross-Ethnic Coalitions

Some provisions of the Constitution and the Election Law of Afghanistan provide incentives for voting and coalitions across ethnicities. For example, Article 45 of the Election Law requires presidential nominees to collect signatures of “one hundred thousand voters, from a minimum of twenty provinces, two percent from each province.”30 Given the regional concentration of ethnic groups, these requirements suggest that presidential nominees need support from more than one ethnic group in order to qualify for candidacy.31 The other electoral rule that encourages cross-ethnic coalitions is Article 60 of the Constitution, addressing the number and appointment of vice presidents.32

30. ELECTION LAW Jun. 8, 2014, S. 1112, art. 45. 31. One might question the purpose and fairness of Article 45, which seems to raise the bar so high that candidates from only the two largest groups can qualify for candidacy. And in fact, the data shows that Pashtuns and Tajiks are the only ethnic groups that successively have at least 2% population in as many as 22 and 16 out of the 26 provinces that are subject to study in this article. See Table I-A in the Appendix. The next two largest ethnic groups, Hazaras and Uzbeks, have a minimum of 2% population in barely 10 of 26 provinces. The remaining 6 provinces that are not subject to this study are unable to increase the number up to 20 provinces. Notably, this provision did not exist during the presidential elections of 2004 and 2009; and perhaps that is the reason that Uzbek and Hazara candidates were able to run in those elections. By logical extension, it was because of the adoption of Article 45 in 2013 that only Pashtun and Tajik candidates were able to run in 2014 election. With this provision in place, it is hard to imagine that a non-Pashtun or non-Tajik presidential nominee would be able to amass support from ethnic groups across 20 provinces to qualify for candidacy. 32. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 60 (“The President shall have two Vice-Presidents, first and second. The Presidential candidate shall declare to the nation names of both vice presidential running mates.”); see also Maryam Safi, Afghanistan’s 2014 Presidential Elections: A Hope for Change, PAX POPULI, May 8, 2014, http://www.paxpopuli.org/2014elections-a-hope-for-change/#sthash.nHHI2J96. ca28hy3b.dpuf (Suggesting that, the choice of two running mates before the election is seen as a way of forming political coalitions with elites from different ethnic groups). MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 363

This article, which is also reflected in Article 45 of the Election Law, requires each president to introduce two vice presidents before the election.33 This requirement has corresponded well to the electoral system and ethnic demographics of Afghanistan. While a candidate has to win 50% of votes in order to win the election in the first round under a runoff system,34 no ethnic group alone can deliver that number of votes. The largest ethnic group in Afghanistan is Pashtun, which, according to some estimates, makes up only 42% of the population of Afghanistan.35 Therefore, in all three presidential elections, the leading candidates have had a great incentive to choose their first and second vice presidents from other ethnic groups.36 In general, each presidential running team has represented three ethnic groups, with the presidential candidate being from one ethnic group, and the first and second vice presidents being from two other, distinct ethnic groups.37 The following Table models the composition of leading presidential teams in the last three presidential elections. Nevertheless, while these electoral rules were meant to promote cross-ethnic coalitions, they have not been able to reduce the influence of ethnic divides and tensions. Instead, as this article shows, the political leverage that these rules provided for ethnic groups led to a complex set of electoral strategies among candidates, elites, and voters, strategies that have amplified ethnic divide and group voting. Part I.A.1. addresses the electoral behavior of candidates, elites, and voters. Then, Part I.A.2. reveals the real nature of cross-ethnic voting in Afghanistan. Together these sections provide the analytical framework for the empirical analysis of this article, which appears in Part II.

33. ELECTION LAW Jun. 8, 2014, S. 1112, art. 45(3). 34. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 61. 35. CENTRAL INTELLIGENCE AGENCY, WORLD FACTBOOK: AFGHANISTAN [hereinafter “CIA Factbook”], https://www.cia.gov/library/publications/the-world-factbook/geos/print/ country/countrypdf_af.pdf (last visited March 17, 2016). 36. On the surface, the legal reinforcement of the cross-ethnic composition of presidential teams seems optimal for promoting cross-ethnic voting. However, this kind of legislation has some drawbacks. Mainly, the last three presidential elections demonstrated that the composition of presidential teams represented only the largest ethnic groups. Likely, candidates chose running mates from the four larger ethnic groups because those candidates would draw greater numbers of ethically motivated support. As Table I shows, other smaller size ethnic groups have remained unrepresented in presidential teams. 37. Sean Mann, Are Ethnic Politics Afghanistan’s Great Hope?, RAND: THE RAND BLOG (Nov. 11, 2014), http://www.rand.org/blog/2014/11/are-ethnic-politics-afghanistans- great-hope.html. MOBASHER 4/18/2016 3:11 PM

364 GONZAGA LAW REVIEW Vol. 51:2

1. Electoral Strategies of Candidates, Elites, and Voters

a) Candidates

Candidates often pursue campaign strategies designed mainly to assemble ethnic votes.38 Since no one Afghan ethnic group is large enough to deliver 50% of the vote, the viability of candidates depends first on their own ethnic support, and second on their cross-ethnic political bargains. Hence, candidates first engage in strategic coordination with hopefuls and would-be candidates from their ethnic groups, persuading them not to run in the elections.39 This article will show that the first mission of candidates is to attempt to secure the support of non-candidate elites, barring the split of votes in their own constituencies. Next, candidates engage in a series of bargaining with elites from other ethnic groups in order to gain cross-ethnic support. Through these bargains, they not only seek to obtain the support of strong elites, but they also try to reorganize “the field[s] of [their] rivals” within their own constituencies.40 As Part II of this article will show, one of the first steps candidates take to secure cross-ethnic support is to choose their vice-presidential candidates from two (mainly large) ethnic groups.41 This practice has been reflected in the choices of leading candidates for the past three presidential elections. The second step is to draw the support of elites across ethnicities by the promises of power sharing, distribution of public funds, and other social and economic pledges. Students of electoral studies have found that not all presidential candidates enter the fray expecting to win in the first round.42 As Cox explains in a more general context, some candidates merely nominate themselves with the hope of acquiring “long-term payoffs.”43 For instance, some candidates enter only as spoilers, splitting the votes of viable candidates for different reasons.44 Many candidates run in the first round, hoping to finish as the second runner up.45 Others simply want to increase their political capital in the first round so they may bargain with runner-ups in the second round.46 Therefore, it is no wonder

38. KATZMAN, supra note 24, at 2. 39. Bijlert, supra note 29, at 9. 40. Id. at 6. 41. Mainly the candidates choose their running mates from the four largest ethnic groups. See infra Part II. 42. See Bijlert, supra note 29, at 9. 43. GARY W. COX, MAKING VOTES COUNT STRATEGIC COORDINATION IN THE WORLD’S ELECTORAL SYSTEMS 152 (1997). 44. See id. at 158; see also Bijlert, supra note 29, at 9. 45. See Cox, supra note 43, at 158. 46. Bijlert, supra note 29, at 9. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 365 why a long list of candidates ran in the 2004 (18 candidates),47 2009 (32 candidates),48 and 2014 (8 candidates)49 elections. The large number of candidates reduces expectations50 about the viability of candidates and increases uncertainty about the election results, which in turn discourages minimum winning coalitions.51 According to William H. Riker’s book The Theory of Political Coalitions, serious candidates tend to build a minimum winning coalition; however, the lack of perfect information encourages them to form larger coalitions.52 In the absence of such information, building a broader coalition is much more rational than building a minimum one.53 As such, it is not surprising that most electoral coalitions in Afghanistan are of a consociational form,54 rather than a minimal one (See Table I above).

b) Elites

Ethno-political elites are the real game changers in Afghanistan. They attempt to present themselves as mediators between candidates and their ethnic groups; however, they are increasingly referred to as ethnic Tikkadaran, a term defined generally as the ones who trade things—in this case constituencies—for some favor or fee.55 These elites are also referred to as commission-kars,56 power

47. The 2004 Afghanistan Presidential Election Results, INDEP. ELECTION COMM’N OF AFG., http://www.iec.org.af/public_html/Election%20Results%20Website/english/english .htm (last visited March) [hereinafter “IEC 2004”]. 48. The 2009 Afghanistan Presidential Election Results, INDEP. ELECTION COMM’N OF AFG., http://www.iec.org.af/results_2009/resultsProvince.html (last visited March 22, 2015) [hereinafter “IEC 2009”]. 49. The 2014 Afghanistan Presidential Election Results, INDEP. ELECTION COMM’N OF AFG., http://www.iec.org.af/results/en/elections (last visited Jan 15, 2015) [hereinafter “IEC 2014”]. 50. Different authors used different terms with slightly different meanings. For example, Cox refers to it “rational expectations.” Compare Cox, supra note 44, at 123-24, with ANTHONY DOWNS, AN ECONOMIC THEORY OF DEMOCRACY 212, 224, 244 (1957) (referring to it as “information’), and WILLIAM H. RIKER, THE THEORY OF POLITICAL COALITIONS 125 (1962) (using four categories of terms such as Complete information, systematically complete information, perfect information, and systematically perfect information in order to describe different levels of having expectations about candidates). 51. See RIKER, supra note 50, at 48, 81. 52. See id. at 47-48, 81. 53. See id. at 48. 54. See AREND Lijphart, DEMOCRACY IN PLURAL SOCIETIES: A COMPARATIVE EXPLORATION 26 (1977) (a consociational coalition, to which Arend Lijphart referred as ‘grand coalition,’ is the kind of coalition where all ethnic groups are included). 55. See Bijlert, supra note 29, at 11. 56. Id. at 11. MOBASHER 4/18/2016 3:11 PM

366 GONZAGA LAW REVIEW Vol. 51:2 brokers,57 favor-buyers,58 and political entrepreneurs59 since they promise vote- blocs in exchange for favors from candidates.60 More interestingly, some top Afghan ethnic leaders who can deliver the largest blocs of votes are known as kingmakers.61 Ethnic Tikkadaran mobilize their groups and encourage them to vote collectively for their endorsees, claiming it would best serve their ethnic interests.62 These elites may pursue a strategy of coordination or strategy of personal interest. Some elites—usually lesser-known elites—act as moderators in order to decrease the number of their ethnic candidates and to coordinate their support behind the most viable ones.63 This is the case in all presidential elections, but the most prominent example was in the presidential election of 2014, when Pashtun elites held a council to choose between Abdul Qayoom Karzai and Dr. Zalmai Rassul.64 Even though the council failed to unite around these candidates, Abdul Qayoom Karzai later dropped off in favor of Zalmai Rassul.65 Stronger elites trade their support even with their own ethnic candidates, knowing that other candidates are willing to pay for their support. Not surprisingly, powerful ethnic Tikkadaran appear more open to negotiations with candidates across ethnic groups.66 In fact, in some cases, elites have gone to the

57. AFGHAN PRESIDENTIAL ELECTION: POTENTIAL CANDIDATES AND POWERBROKERS, OPEN SOURCE CENTER (2009), http://fas.org/irp/dni/osc/afghan-election.pdf (providing a list of Afghan elites who were the potential powerbrokers in the 2009 presidential election). 58. DOWNS, supra note 50, at 88. 59. NIKOLAOS BIZIOURAS, THE POLITICAL ECONOMY OF ETHNIC CONFLICT IN SRI- LANKA: ECONOMIC LIBERALIZATION, MOBILIZATIONAL RESOURCES, AND ETHNIC COLLECTIVE ACTION 20 (2014). 60. See Bijlert, supra note 29, at 13. 61. See Brian Glyn Williams, The Return of the Kingmaker, FOREIGN POLICY, Aug. 17, 2009, http://Foreignpolicy.Com/2009/08/17/The-Return-Of-The-Kingmaker/; Another election, another cycle of fraud and violence?, FRANCE 24 (Sep. 17, 2010), http://www.france24.com/en/20100916-afghan-parliamentary-elections-fears-another-cycle- fraud-violence-karzai-taliban. 62. Bijlert, supra note 29, at 11. 63. Id. at 9. 64. Initially there was an attempt to unite four Pashtun candidates until Gul Agha Sherzai and Muhammad Nader Na’im declined to cooperate with the council. See Thomas Ruttig, Pluralistic within Limits, but Not Democratic: Afghanistan’s political landscape before the 2014 elections, Afghanistan Analysts Network, Oct. 24, 2013, https://www. afghanistan-analysts.org/a-joint-durrani-candidate-for-president-after-the-unity-jirga-in- kabul/. 65. Safi, supra note 32. 66. Bijlert, supra note 29, at 2. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 367 candidates for negotiations before candidates seek their endorsements.67 In other cases, Tikkadaran have publicly outlined their ethnic or personal demands long before the elections.68 The tendency of elites toward ethnic mobilization has become more revealing in every subsequent election. In 2004, although elites mobilized ethnic groups, they tended not to do so publicly. In 2009, however, Mohaqiq publicly outlined a joint charter of their demands for joining any coalitions.69 In the runoff of the 2014 election, during a campaign in Baghlan province, Dr. Ghani’s first candidate for vice president, Dr. Dostum, said that those Uzbeks and Turkmens who would not vote for his side were traitors.70 In another event, Juma Khan Hamdard, the governor of Paktia, called on Pashtuns to only vote for Dr. Ghani for their own protection.71

c) Voters

In plural societies, voters vote collectively along ethnic lines.72 Under this theory, voters merely follow instructions of their elites, rather than make decisions for themselves.73 Hence, strategic voting, where voters defect from their ethnic candidates in favor of a more viable candidate from another group, is less likely.74 While it may be true, this is a very narrow notion of strategic voting, since it disregards other aspects of strategic voting. For instance, as this article illustrates, voters may defect from one of their ethnic candidates in favor of another from the same group. Therefore, this article attempts to extend the

67. Id. 68. Mohaqiq Outlines Joint Charter of Demands with Junbish, HAZARISTAN TIMES, Apr. 24, 2009, https://hazaristantimes.wordpress.com/2009/04/26/Mohaqiq-Outlines-Joint- Charter-Of-Demands-With-Junbish/. 69. Id. 70. Karim Amin, Uzbek And Turkmen Traitors If They Don’t Vote Ghani: Dostum, TOLONEWS, June 9, 2014, http://www.tolonews.com/en/election-2014/15161-uzbek-and- turkmen-traitors-if-they-dont-vote-ghani-dostum. 71. Tahi Khan, Presidential election: Ethinic divide comes into play in Afghan runoff, THE EXPRESS TRIBUNE, June 9, 2014, http://tribune.com/pk/story/719279/presidential- election-ethnic-divide-comes-into-play-in-afghan-runoff/. 72. ALVIN RABUSHKA & KENNETH A. SHEPSLE, POLITICS IN PLURAL SOCIETIES: A THEORY OF DEMOCRATIC INSTABILITY 63 (1972). 73. Bijlert, supra note 29, at 2. 74. Robert G. Moser et al, Social Diversity Affects the Number of Parties Even under First-Past-the-Post Rules 6 (Sept. 2011) (unpublished paper presented at the 2011 Annual Meeting for the American Political Science Association). For the definition of ‘strategic voting’ refer to THE NEW PALGRAVE DICTIONARY OF ECONOMICS, www. dictionaryofeconomics.com/dictionary (last visited Feb. 5, 2016). MOBASHER 4/18/2016 3:11 PM

368 GONZAGA LAW REVIEW Vol. 51:2 definition of strategic voting to include intra-ethnic coordination and defection, and cross-ethnic bargains and patronage. The tendency for the intra-ethnic strategy of coordination and defection is considerably high among all ethnic groups and in all presidential elections. This article shows that ethnic groups tend to vote collectively for their most viable candidates and defect from the others. For example, in the 2004 election, while Karzai, Qanooni, Dostum and Mohaqiq dominated the polls in their constituencies, their ethnic counterparts won less than 1% of the votes.75 This electoral behavior is very similar to strategic voting in consolidated democracies where a left wing supporter is more likely to defect from a losing left candidate for a more viable one, but is less likely to defect from a losing left wing candidate to vote for a right wing candidate. However, as Cox posits in his extensive analysis of strategic voting, in the absence of an apparent viable candidate, the ethnic voters may encounter an expectation crisis that may lead to dispersion of their votes to different candidates.76 Pashtuns faced this kind of crisis in 2014 when Karzai was no longer running and there were a variety of opinions about who was the most viable Pashtun candidate.77 As a result, Pashtuns’ votes were split among a number of candidates, which led to a Tajik candidate topping the list for the first time in any presidential election.78 Contrary to the notion of instruction and loyalty,79 the voters do not always follow the instruction of their elites when voting. Ethnic groups sometimes vote collectively and independently from the instruction of their elites. This often occurs when they conclude that their best interest is not served by the elites, and it is particularly likely when they have a viable candidate of their own.80

75. IEC 2004, supra note 47. 76. See COX, supra note 43, at 9. 77. Nathan Hodge, Afghan Candidates Discuss Ways to Secure Pashtun Support, THE WALL STREET JOURNAL, Feb. 26, 2014, http://www.wsj.com/articles/SB1000142405270 2304709904579406813606798606; Thomas Ruttig, A Joint ‘Durrani’ Candidate for President? After the Unity Jirga in Kabul, Afghanistan Analysts Network, Mar. 5, 2014, at 1, 2, https://www.afghanistan-analysts.org/a-joint-durrani-candidate-for-president-after- the-unity-jirga-in-kabul/. 78. IEC 2014, supra note 49; Profile: Abdullah Abdullah, BBC NEWS, Sept. 29, 2014, http://www.bbc.com/news/world-asia-27138728. 79. See Bijlert, supra note 29, at 15-16 (“When following the principle of instruction voters simply adhere to a leader’s lead, based on the assumption that this will somehow further the group’s interest or, alternatively, based on a lack of appetite to challenge the leader’s authority. . .[t]he principle of loyalty describes often very personal claims that are made on Afghans to provide support to a certain person or cause, based on ties of friendship, duty, obligation and reciprocity (often amounting to emotional blackmail).”). 80. Moser et al, supra note 74, at 6. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 369

Therefore, it is important to note that the liaison between voters and their elites is a mutual interest relationship that either side can breach. Ethnic groups need elites for their collective interests as much as elites need followers for their personal interests. While elites are the foundation for collective bargaining and voting, they are not necessarily the ones dictating the outcome.81

2. The Pattern of Cross-Ethnic Voting in Afghanistan

Post-election statistics reveal that cross-ethnic voting happens frequently in Afghanistan. As Benjamin Dubow found in the 2004 and 2009 elections, cross- ethnic voting did happen in several provinces.82 Perhaps one reason that Dubow and others considered ethnic voting less significant in Afghanistan is the high level of cross-ethnic voting in some provinces. Indeed, the scale of cross-ethnic voting in the Afghan election of 2014 was extraordinary. Most Hazaras voted for a Tajik candidate, and most Uzbeks voted for a Pashtun candidate. See Table A- I in the appendix. But do these cross-ethnic votes suggest that people are not voting on the basis of ethnicity? To understand the nature of cross-ethnic voting in Afghanistan, one needs to distinguish between two kinds of cross-ethnic voting: (1) merit-based and (2) ethnic-based. In merit-based cross-ethnic voting, ethnicity has lesser or no importance, and voters tend to choose the candidates based on their policies and leadership qualities.83 With this kind of voting, candidates tend to mobilize voters by showing their party tags, selling their policies, and promising a better future. They may offer distinct promises to certain minorities, but they do not generally condition their promises in exchange for votes (i.e., patronage). This article does not deny the existence of merit-based cross-ethnic voting in Afghanistan. It does suggest, however, that voting because of policy, in general, is much less common than the existing literature would suggest. Those cross-ethnic votes that are cast merely in exchange for patronage for the voting group are still ethnic-based voting.84 In a survey in 2009, Cohen and Larson found that “despite the televised debates, campaign banners, and rallies that took place” most votes are cast on the basis of patronage.”85 The patronage might involve assigning some seats in the cabinet for the voting groups,

81. Bijlert, supra note 29, at 2. 82. Dubow, supra note 23, at 26. 83. This merit-based voting seems to encourage individualist voting since individuals from the same ethnic group do not necessarily have the same choice of policies and candidates. Merit-based, cross-ethnic voting occurs only in consolidated democracies, where all communities are, to a large degree, politically well off. 84. Coburn & Larson, supra note 21, at 11. 85. Id. at 7. MOBASHER 4/18/2016 3:11 PM

370 GONZAGA LAW REVIEW Vol. 51:2 appointing their elites as a vice-president, providing them public funds, or even creating and administering new provinces and districts.86 For instance, in the presidential election of 2009, Karzai promised Mohaqiq that, in exchange for Hazara votes, he would upgrade two Hazara districts into two provinces.87 Ethnic-based, cross-ethnic voting occurs at both macro- and micro-levels. Cross-ethnic voting at a macro level occurs when the bulk of an ethnic group across the country vote for a specific candidate.88 It appears that the main leaders of ethnic groups—otherwise known as kingmakers—tend to mobilize this macro-level voting. On the micro-level, cross-ethnic voting occurs at the local level. Candidates who fail to win the endorsement of an ethnic leader are more likely to negotiate with secondary and local elites of the same group.89 They bargain with local leaders, Mullahs, elders, commanders, and even influential youth who can promise blocs of votes from their localities.90

II. THE EMPIRICS OF ETHNIC-ELECTORAL DYNAMICS

This part of the article provides the first systematic statistical analysis of ethnic-electoral dynamics across three presidential elections—excluding the 2014 runoff election because of a lack of data91—and across 26 provinces of Afghanistan.

86. These patronages will be explained in detail throughout the paper. However, for knowing some instances of these patronage refer to: Richard A. Oppel Jr., Afghan Leader Courts the Warlord Vote, N.Y. TIMES, Aug. 7, 2009, http://www.nytimes.com/2009/08 /08/world/asia/08warlords.html?_r=0; Jon Boone, Afghan president Hamid Karzai picks ex-warlord as election running mate, THE GUARDIAN, May 4, 2009, http://www.theguard ian.com/world/2009/may/04/afghanistan-president-hamid-karzai-election; Makia Monir, Karzai Vow Draws Negative Reaction, PAJHWOK, Jun. 18, 2009, http://e-ariana.com /ariana/eariana.nsf/allDocs/7cb3e5dbec08cc41872575d90052e074!OpenDocument&Click=; Renee Montagne, Afghan Politicians Pursue Minority Hazara Voters, NPR, Aug. 07, 2009, http://www.npr.org/templates/story/story.php?storyId=111639248. 87. Montagne, supra note 86; HASEEB HOMAYOUN, INSITIUTE FOR THE STUDY OF WAR, AFG. REP. 4: THE REELECTION OF HAMID KARZAI 15-16 (Jan. 2010), http://www. understandingwar.org/report/re-election-hamid-karzai. 88. A good example of macro-level cross ethnic voting is the collective voting of Hazaras to Dr. Abdullah in 2014. In the same election majority of Uzbeks voted for Dr. Ghani. These instances of cross-ethnic voting are better illustrated in figure VII, infra p. 59, and figure IX, infra p. 66. 89. See Coburn & Larson, supra note 21, at 7. 90. Id. 91. Thomas Ruttig, Elections 2014 (52): The not yet officially announced results – electoral maths with unknowns, Afghanistan Analysts Network, SEPT. 28, 2014, AT 1, 2, https://www.afghanistan-analysts.org/2014-elections-52-the-not-yet-officially-announced- results-elections-maths-with-unknowns/?format=pdf. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 371

The data this article examines is the product of a long process of gathering and cataloging of information from numerous sources. Among this data, the most socially sensitive was collecting data about ethnic demographics.92 The Afghan government does not provide a census about ethnic populations, presumably because of the sensitivity and delicacy of this matter in the society. Table I-A in the Appendix shows the ethnic composition of different provinces, along with the sources of those data. While some credible and information-sensitive organizations have generated this data, this article refers to the data as an estimation, rather than as a census record, mainly due to the sensitivity of ethnic groups about their population and the suspicion that the organizations have not conducted all proper procedures of a census survey. In fact, there may be some differences between the votes of candidates and their ethnic populations that are are unaccounted for because of errors in the data. In this part, I examine all theoretical frameworks and subsidiary arguments set forth in the previous sections. Section A begins with an introduction to the ethnic demographics in Afghanistan. Section B examines ethnic-electoral dynamics in three presidential elections on the national level. It explains how ethnic populations and political bargains determine who wins the election in Afghanistan. This section also explains how candidates and elites circumvent electoral laws. Then, section C provides a comparative analysis of how ethnic patterns are reflected in election outcomes across provinces. While discussing the ethnic-electoral association in all 26 provinces, the provinces of special interest are the ones in which there has been a change in voting patterns from one election to another, or in which there has been a high level of cross-ethnic voting. This article will attempt to trace the causes of those changes and trends in those particular provinces.

A. Ethnic Demographics in Afghanistan

Afghanistan has a total population of 28.6 million.93 Many have considered Afghanistan to be a country of minorities, since, according to most estimations, no ethnic group comprises over 50 percent of the population.94 The Afghan

92. Emma Graham-Harrison, Afghan Census Dodges Questions of Ethnicity And Language, THE GUARDIAN, Jan. 3, 2013), http://www.theguardian.com/world/2013/jan/03 /afghan-census-questions-of-ethnicity. 93. It is the latest census released by the Central Statistic Office. See CENTRAL STATISTICS ORGANIZATION, http://cso.gov.af/en/page/demography-and-socile-statistics/demo graph-statistics/3897111 (last visited Feb. 22, 2016) [hereinafter “CSO”]. 94. Muhammad Saleem Mazhar et al, Ethnic Factor in Afghanistan, 19 J. POL. STUD. 99 (2012), http://pu.edu.pk/images/journal/pols/pdf-files/Naheed-winter2012.pdf; see also CIA Factbook, supra note 34; Nahid Suleman, Ethnic Discrimination in Afghanistan, MOBASHER 4/18/2016 3:11 PM

372 GONZAGA LAW REVIEW Vol. 51:2

Constitution acknowledges fourteen ethnic groups in Afghanistan, namely “Pashtun, Tajik, Hazara, Uzbek, Turkman, Baluch, Pachaie, Nuristani, Aymaq, Arab, Qirghiz, Qizilbash, Gujur, Brahwui and other tribes.”95 The first four ethnicities together are estimated to comprise the largest percentage of the population, at around 87%.96 The findings of any studies about the exact population of each ethnicity have always been controversial because no governmental institution has yet officially surveyed ethnic populations. According to CIA World Factbook, of the total population of Afghanistan, Pashtuns comprise 42%, Tajiks 27%, Hazaras 9%, and Uzbeks 9%.97 Pashtuns mostly live in the South, Southeast, and Southwest (see figure I). There are some small populations of Pashtuns in the North and West. Tajiks mostly live in the North, Northeast, Northwest, and central regions. Uzbeks live in some northern provinces and Hazaras in the central region called Hazarajat (see figure I). Nonetheless, this does not suggest that different regions are home to only one ethnicity.98

INTERMEDIA 1, http://www.intermedia.org.pk/pdf/pak_afghan/Naheed_Soleman_Ethnic_ Discrimination_in_Afghanistan.pdf. 95. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 4. Some surveys and authors, however, do not agree with the number of ethnic groups mentioned in the Constitution. Some surveys have counted up to 54 ethnicities in Afghanistan. Others suggest that there are around 200 ethnicities in Afghanistan based on how ethnicity is defined. See Conrad Schetter, Ethnicity and the political reconstruction in Afghanistan 3 (June 2010) (unpublished Conference Paper, London School of Economics and Political Science), http://eprints.lse.ac.uk/28376/1/Schetter_LSERO_version.pdf. 96. CIA Factbook, supra note 35. 97. Id. 98. NOAH COBURN & ANNA LARSON, DERAILING DEMOCRACY IN AFGHANISTAN 37 (2014). MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 373

FIGURE I. This map was developed on the basis of ethnic census estimations by my account99

B. Role of Ethnicity on Election Outcomes: Nationwide

The presidential elections of 2004 and 2009 reflected ethnic populations in the votes of the leading candidates: in both elections, Pashtun candidates ranked at the top, Tajik candidates ranked second, and Hazara and Uzbek candidates placed third and fourth, respectively.100 While the 2014 presidential election suggested a possible shift away from this chronicle—as a Tajik led the tallies and a Pashtun candidate followed101—a close analysis of the election shows that ethnicity is still a strong determiner of the election outcome. The findings in this section suggest that political bargains and ethnic-based cross ethnic voting had significant roles in the presidential election of 2014. Each election is analyzed

99. AFGHANISTAN ELECTION DATA, http://afghanistanelectiondata.org/election/2009/ report/karzai-abdullah-ethnicity#zoom=0&lat=33.9&lon=66.5&layers=district_map_open layers_7,district_map_openlayers_8,district_map_openlayers_9,afghanistan-ethnic&base Layers=afghanistan-grey (last visited Feb. 6, 2016). 100. IEC 2004, supra note 47; IEC 2009, supra note 48; Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016). 101. IEC 2014, supra note 49; Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016). MOBASHER 4/18/2016 3:11 PM

374 GONZAGA LAW REVIEW Vol. 51:2 separately; and, within each election, electoral strategies of candidates, elites, and voters and the resulting irregularities are analyzed.

1. The Presidential Election of 2004

In the presidential election of 2004, Hamid Karzai, a Pashtun candidate, won the election with the majority votes.102 Yunis Qanooni, a Tajik candidate, got the second largest number of votes.103 Mohammad Mohaqiq, a Hazara, and Abdul Rashid Dostum, an Uzbek, obtained respectively the third and fourth highest votes.104 Figure II illustrates that the ranking of each candidate corresponds to the size of his ethnic group.

Candidates' Votes & Their Ethnic Populations: Presidential Election of 2004

Pashtun 42.00% Tajik 25.00% Hazara 9.00% Uzbek 9.00%

55.4%

16.3% 11.7% 10%

Karzai Qanooni Mohaqiq Dostum

FIGURE II. This chart compares the votes of the leading candidates with the population of their ethnic groups.105

102. IEC 2004, supra note 47; Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016). 103. IEC 2004, supra note 47; Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016). 104. Wesal Zaman, With Vote Ruled Valid, Karzai’s the Official Winner, LOS ANGELES TIMES, Nov. 4, 2004, http://articles.latimes.com/2004/nov/04/world/fg-karzai4. 105. See IEC 2004, supra note 47 (for information about the votes of candidates). See also CIA Factbook, supra note 35 (for information about the population of ethnic groups); Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016) (for the ethnicity of each candidate). MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 375

There are, however, some discrepancies between the tallies of candidates and the estimated population of their ethnic groups. For instance, Karzai won 55.6% votes,106 while Pashtuns are estimated at only 42%.107 The additional 13.6% of votes was due to Karzai’s deals with non-Pashtun elites and his ability to coordinate Pashtun votes. For example, in order to win support from some Tajiks and Hazaras, Karzai added a Tajik elite, Ahmad Zia Masoud, and a Hazara leader, Karim Khalili, to his ticket.108 In addition, he sealed the endorsement of Atta Mohammad Noor, an influential Tajik commander in the north, in exchange for his appointment as the governor of Balkh in 2004.109 It was because of Atta Mohammad Noor that Karzai garnered more votes than Tajik candidates did in Balkh, a Tajik majority province.110 Moreover, Karzai, as the incumbent since the collapse of the Taliban in 2001, was expectedly the most viable Pashtun candidate, so he dominated the ballots in the regions with a Pashtun majority. The voters defected from all of his Pashtun rivals, as their tallies remained below 1% nationwide,111 even while some very influential Pashtun elites like Sayed Ashaq Gilani and Ahmadshah Ahmadzai, leaders of two Mujahidin factions, were also in the race.112 The defection from these elites by voters shows a strong voting coordination among Pashtun voters when they have a clear expectation about the viability of candidates.

106. IEC 2004, supra note 47. 107. CIA Factbook, supra note 35. 108. Boone, supra note 86. 109. Karzai appointed Atta Mohammad Noor as the governor of Balkh in 2004 despite General Dostum’s criticism. Atta was a rival of Dostum and Haji Mohaqiq for controlling the Balkh province. He eventually marginalized both and gained a lot of popular support in Balkh. See Dipali Mukhopadhyay, Warlords As Bureaucrats: The Afghan Experience (Aug. 2009) (unpublished Ph.D. dissertation, Carnegie Endowment for International Peace) (on file with author); see also HOMAYOUN, supra note 87, at 14. 110. HOMAYOUN, supra note 87, at 15. 111. IEC 2004, supra note 47. 112. Sayed Ishaq Gilani was a member of Mahaz Mili faction that was fighting Soviet occupation in Afghanistan. Mahaz Mili faction has been led by his uncle Sayed Pir Ahmad Gilani. He, however, late left the faction and established his own party entitled, the National Solidarity Movement. See Suhaila Muhsini, Sayed Ishaq Gailani: Ex-Mujahedin With Moderate Views, INSTITUTE FOR WAR AND PEACE REPORTING (Feb. 22, 2016), https://iwpr.net /global-voices/sayed-ishaq-gailani-ex-mujahedin-moderate-views. Ahmad Shah Ahmadzai was also leading a fighting faction, Itihad Islami, against Soviet occupation as the deputy head. He later claimed that he has loosened all ties with militant groups. See abdulbasir saeed, Ahmad Shah Ahmadzai: Islamist Insists He’s Broken Armed Faction Ties, INSTITUTE FOR WAR AND PEACE REPORTING (Feb. 22, 2016), https://iwpr.net/global-voices/ahmad-shah-ahmadzai- islamist-insists-hes-broken-armed-faction-ties. MOBASHER 4/18/2016 3:11 PM

376 GONZAGA LAW REVIEW Vol. 51:2

Perhaps one other factor that helped Karzai with 13.6% higher votes is the refugees’ votes. Unlike in the 2009 and 2014 elections, in 2004 Afghan refugees in Iran and Pakistan were allowed to vote, resulting in over 846,776 refugees in these countries casting their votes.113 Karzai won 80.2% votes from refugees in Pakistan,114 in which 85.1% of refugees are ethnic Pashtuns.115 He also secured 44.4% votes of refugees in Iran, which hosts 15% Pashtun refugees.116 Mohammad Yunis Qanooni, a prominent Tajik elite, won 16.3% votes.117 Tajiks are the second largest ethnic group, comprising an estimated 27% of the population.118 This suggests that Qanooni received 9.3% votes lower than the estimated Tajik population. Since other Tajik candidates received only 3.4% of the votes in total, Karzai’s coalition with some Tajik powerbrokers is the most plausible explanation for the remaining discrepancy.119 Haji Mohaqiq and Dostom secured respectively 11.7% and 10% votes nationwide.120 According to estimations by CIA’s Factbook, Hazaras and Uzbeks each make 9% of the Afghan population, comprising 18% of the population together.121 The presidential election of 2004 was less notorious for electoral irregularities than the 2009 and 2014 elections. However, electoral fraud and the use of intimidation was damaging enough that it led 15 candidates to boycott the election.122 Mainly, ballot stuffing, voter intimidation, and multiple registrations

113. Catinca Slavu, External Voting for Afghanistan’s 2004 Presidential Election, AFGHANISTAN ANALYSTS NETWORK, in SNAPSHOTS OF AN INTERVENTION: THE UNLEARNED LESSONS OF AFGHANISTAN’S DECADE OF ASSISTANCE (2001-11) 27, 30-31 (Martine van Bijlert & Sari Kouvo eds., 2012) (ebook), http://www.afghanistan-analysts.org/wp-content/ uploads/downloads/2012/09/Snapshots_of_an_Intervention.pdf. 114. Id. at 32. 115. UNITED NATION HIGH COMMISSIONER FOR REFUGEES, ASSESSMENT REPORT: INTERNATIONAL CONFERENCE ON SOLUTION STRATEGIES FOR AFGHAN REFUGEES TO SUPPORT VOLUNTARY REPATRIATION, SUSTAINABLE REINTEGRATION, AND ASSISTANCE TO HOST COUNTRIES 42 (May 2012), http://www.unhcr.org/4fa7e0ea9.pdf. 116. SLAVU, supra note 113, at 28, 32. 117. IEC 2004, supra note 47; Who Is Who In Afghanistan: Qanooni, Mohammad Younis Younus Qanuni Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php? option=com_afghanbios&id=1392&task=view&total=3240&start=2315&Itemid=2, (last vis- ited Jan. 10, 2016). 118. CIA Factbook, supra note 35. 119. IEC 2004, supra note 47; Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016). 120. IEC 2004, supra note 47; Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 5, 2016). 121. CIA Factbook, supra note 35. 122. Carlotta Gall, Election of Karzai Is Declared Official, N.Y. TIMES, Nov. 4, 2004, http://www.nytimes.com/2004/11/04/international/asia/04afghan.html?_r=0. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 377 took place in 2004.123 In one case, a tribal council in Khost declared “that they would burn the houses of anyone not voting for Karzai.”124

2. The Presidential Election of 2009

The results of the 2009 election exhibited a pattern of ethnic-election association, similar to the presidential election of 2004 (see figure II-5). In that race, Karzai won the highest number of votes.125 Dr. Abdullah Abdullah, a Tajik elite, 126 won the second highest votes.127 Dr. Ramazan Bashardost, a Hazara technocrat, stood as the third leading candidate.128 Uzbeks did not have any candidate in the presidential election of 2009.129 Candidates' Votes & Their Ethnic Populations: Presidential Election of 2009

Pashtun 42% Tajik 27% Hazara 9% 49.7% 30.6% 10.5

Karzai Abdullah Bashardost

FIGURE III. This chart compares the votes of the leading candidates in 2009 with the population of their ethnic groups.130

123. SWISSPEACE, FAST UPDATE: AFGHANISTAN QUARTERLY RISK ASSESSMENT (SEP- DEC 2004) 4-5 (2005), http://www.swisspeace.ch/fileadmin/user_upload/Media/Projects/ FAST/Asia/Afghanistan/FAST_Update_Afg_04_04.pdf. 124. Bijlert, supra note 29, at 16. 125. IEC 2009, supra note 48. 126. Dr. Abdullah is half-Pashtun. However, he has always been regarded as a Tajik leader inside and outside Afghanistan. Who Is Who In Afghanistan: Dr. Abdullah Abdullah Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios &id=14&task=view&total=2837&start=8&Itemid=2 (last visited Oct. 4, 2015); see also Abdullah Abdullah to Contest, THE HINDU (Oct. 2, 2013, 00:03 IST), http://www.thehindu .com/news/international/south-asia/abdullah-abdullah-to-contest/article5189613.ece. 127. IEC 2009, supra note 48. 128. Id.; KATZMAN, supra note 24, at 4. 129. IEC 2009, supra note 48. 130. Id. (for information about the votes of candidates); CIA Factbook, supra note 35 (for information about the population of ethnic groups); Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan-bios.info/database.html (last visited Feb. 22, 2016). MOBASHER 4/18/2016 3:11 PM

378 GONZAGA LAW REVIEW Vol. 51:2

Karzai won 49.7% of the votes, winning around 5.7% lower than his previous total of votes.131 He still received around 7.7% higher votes than the estimated Pashtun population. Again, this discrepancy is likely attributable to inter-ethnic and intra-ethnic bargains that Karzai sealed with ethnic Tikadarans. Karzai, whose popularity was damaged by 2009, first attempted to establish himself as the most viable candidate among Pashtuns.132 He persuaded some strong Pashtun contenders, such as Gul Agha Shirzai (then governor of Nangarhar) and Anwar-ul-Haq Ahadi (the head of a Pashtun nationalist party)133 not to run against him.134 Then, Karzai began negotiations and patronage to elites of different regions, to undermine other Pashtun counterparts in their local constituencies. For instance, Karzai appointed Haji Din Mohammad, a relative to two other candidates, Nasrullah Baryalay and Amin Arsala, as his campaign manager.135 In addition, he established an electoral network of local elites and key Mujahiden leaders like Assadullah Khaled (then acting Minister for Border affairs), Abdulrab Sayyaf (leader of Dawat Islami fraction), Sayed Ahmad Gailani (leader of the Mahaz-i Milli party), and Sher Mohammad Akhundzada (the former governor of Hilmand).136 These networks were strong enough to discourage some well-known figures, such as Ali Ahmad Jalali and Zalmai Khalilzad, from running.137 These intra-ethnic coalitions and coordination had a substantial impact in Karzai’s winning of votes from Pashtuns. In the meantime, by introducing two prominent Tajik and Hazara strongmen—Marshal Qasim Fahim and Karim Khalili— as his vice president candidates, Karzai attempted to draw cross-ethnic votes.138 He also secured the backing of some influential Tajik leaders such as Burhanuddin Rabbani (the former president), Ismail Khan (the former governor of Herat), Ahmad Wali

131. IEC 2009, supra note 48. 132. Andrew Reynolds & John Carey, Fixing Afghanistan’ s Electoral System: Arguments And Options for Reform, AFG. RES. AND EVALUATION UNIT 9 (July 2012), http://www.areu.org.af/Uploads/EditionPdfs/1211EFixing%20Afghanistans%20Electoral%2 0System%20BP%202012.pdf; see also Bijlert, supra note 29, at 9. 133. Decision Point 2009, supra note 20, at 5. 134. Reynolds et al, supra note 132, at 9 n.21; see also HOMAYOUN, supra note 87, at 12; Bijlert, supra note 29, at 9. 135. See Bijlert, supra note 29. 136. Id. at 10; Reynolds et al, supra note 132, at 9; see also KATZMAN, supra note 24, at 1. 137. See Bijlert, supra note 29 at 9. 138. Oppel Jr., supra note 86. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 379

Massoud (brother of the late Ahmad Shah Massoud, the Northern Alliance figure),139 and the former Minister of Interior, Moqbel Zarar.140 Karzai’s most electorally beneficial alliance was the one with Mohaqiq and Dostom, both former presidential candidates.141 Karzai succeeded in obtaining their support in exchange for several seats in his cabinet and the promise of upgrading two Hazara populated districts to provinces.142 Karzai’s coalition with Mohaqiq, however, did not bring him as many Hazara votes. Hazaras, unlike Uzbeks, had a candidate of their own ethnic group: Dr. Ramazan Bashardost, for whom they mostly voted (See figure III). Considering his personality and past, Dr. Bashardost’s appeal to Hazara voters adds more to our understanding about the dynamics of ethnicity in relation to elections.143 Even though he is an ethnic Hazara, Dr. Bashardost explicitly campaigned for national loyalty at the expense of ethnic loyalties.144 He vocally attacked ethno-political elites, including those of Hazaras.145 Considering Dr. Bashardost’s loose ethnic ties with the Hazaras, Karzai’s team did not expect Dr. Bashardost to garner much of the Hazara votes, bearing in mind Mohaqiq’s alliance with Karzai, who is still considered the most powerful leader of Hazaras.146 Nevertheless, the data shows that most Hazaras choose to vote for Dr. Bashardost. The two most likely reasons for Hazaras choosing Dr. Bashardost over Mohaqiq’s endorsement were Dr. Bashardost’s viability and his charisma.147 Dr.

139. Reynolds et al, supra note 132, at 9 n.21. 140. Bijlert, supra note 29, at 10. 141. Montagne, supra note 86; see also HOMAYOUN, supra note 87, at 14. 142. See Bijlert, supra note 29, at 10. 143. Dubow, supra note 23, at 24. 144. Montagne, supra note 86. 145. Karzai’s warlord links challenged, BBC NEWS, Aug. 16, 2009, http://news.bbc.co.uk/2/hi/south_asia/8204350.stm. 146. Montagne, supra note 86. 147. Some argue that Mohaqiq’s inability to mobilize all Hazaras to vote for Karzai was also due to some incidents that eroded Karzai’s popularity among Hazaras—as well as Mohaqiq’s. A year prior to the presidential election in 2009, a historical land dispute between Kuchis (Nomads) ultimately led to violence and caused tensions between Hazaras and Karzai’s government. See COBURN & LARSON, supra note 98, at 92. They accused Karzai’s government of failing to protect Hazaras against Kuchis’ incursion. See Decision Point 2009, supra note 20, at 26. Ironically, their protest was led by Mohaqiq. See Ethnic Hazaras Protest In Feud With Afghan Nomads, REUTERS, Jul. 2, 2008, http://mobile.reuters.com/ article/idUSISL116163?irpc=932. Mohaqiq’s alliance with Karzai, in a very short period of time, was considered hypocritical and generated criticism among Hazaras and non-Hazara intellectuals. Moreover, Mohaqiq’s endorsement of Sayyaf for chairmanship of parliament in 2005 also angered many Hazaras. Sayyaf, the head of Itehad Islami fraction engaged in a ruthless war against Hazaras during the civil war and has been accused of committing the act MOBASHER 4/18/2016 3:11 PM

380 GONZAGA LAW REVIEW Vol. 51:2

Bashardost was not just the only Hazara candidate; he was also one of the viable candidates across the board, according to pre-election polls.148 His viability would normally have been magnified by the ethnic sentiments of his supporters, who would want to him to at least finish second to Karzai in the first round election—in order to compete in the second round. The second most likely reason was Dr. Bashardost’s popularity among other ethnic groups, due to his anti- corruption commitment as the Minister of Planning and his campaign against poverty.149 Indeed, his viability increased by the expectation that he would mobilize a considerable number of protest votes from other ethnic groups.150 And, he did receive some votes in non-Hazara constituencies (See Table I-A in the Appendix). Overall, even though Dr. Bashardost won a majority of the Hazara votes, Hazaras were not able to coordinate effectively since their main leaders, Mohaqiq and Khalili, supported Karzai’s campaign. Dr. Abdullah introduced Dr. Cheragh Ali Cheragh, an ethnic Hazara, and Humayon Shah Asifi, an ethnic Pashtun, as his candidates for vice president.151 These figures were not known as prominent elites, but rather as intellectuals. As such, it seemed that Dr. Abdullah was attempting to appeal to educated Pashtun

of genocide in Afshar war. See HAFIZULLAH EMADI, DYNAMICS OF POLITICAL DEVELOPMENT IN AFGHANISTAN: THE BRITISH, RUSSIAN, AND AMERICAN INVASION 221 (2010); see also Afshar Massacre 1993, HAZARA PEOPLE INTERNATIONAL NETWORK (Feb. 9, 2011) http://www.hazarapeople.com/2011/02/09/afshar-massacre-1993/. 148. Three pre-election polls were conducted in 2009. See AFGHANISTAN PUBLIC OPINION SURVEY, THE INTERNATIONAL REPUBLIC INSTITUTE (IRI) (May 2009), http://graphics8.nytimes.com/packages/pdf/world/afghanpoll.pdf; AFGHANISTAN PUBLIC OPINION SURVEY, THE INTERNATIONAL REPUBLIC INSTITUTE (IRI) 13 (Jul. 2009), http://www.iri.org/sites/default/files/2009-News-Release-Files/2009%20August%2014%20 Survey%20of%20Afghan%20Public%20Opinion%20July%2016-26,%202009.pdf; Glevum Associates, Afghan Election Survey Executive Summary, E-ARIANA (Jul. 11, 2009), http://e- ariana.com/ariana/eariana.nsf/allDocs/85427abc66ff10868725760f004cd43c!OpenDocument &Click=. 149. “Bashardost is well known for his struggle for human rights and against corruption. In 2004, while acting as the Minister of Planning, he conducted a study of the effectiveness of foreign and local NGOs working in Afghanistan. Based on his findings he recommended the closure of 1,900 NGOs that he claimed were not providing aid to the people. When his recommendations were not acted upon, he quit cabinet and was later elected as an independent MP.” Decision Point 2009, supra note 20; see also Afghanistan’s Presidential Election: Power To The People, Or The Powerful?, THE INT’L COUNCIL ON SEC. AND DEV. 27 (Apr. 2009), http://www.icosgroup.net/static/reports/power_to_the_people.pdf; AFGHANISTAN’S GANDHI: DR. RAMAZAN BASHARDOST, http://ramazanbashardost.com/ spip.php?rubrique2 (last visited Nov. 13, 2014). 150. See Decision Point 2009, supra note 20. 151. See Who Is Who In Afghanistan, AFGHAN BIOS, http://www.afghan- bios.info/database.html (last visited Jan. 11, 2016). MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 381 and Hazara voters. The elections results, however, suggested that his strategy was a miscalculation. Dr. Abdullah, who also failed to gain the endorsement of high-ranking Hazara and Uzbek elites, attempted to undercut their influence by getting the support of local and low ranking elites. These elites included Akbar Qasemi, an influential Hazara parliamentarian from Ghazni, and a series of local elites from Jombesh party, an Uzbek faction.152 His votes from small pockets of Turkmens and Uzbeks also likely came due to the support of Gen. Abdul Malik, who is Dostum’s main Uzbek rival.153 Even though Dr. Abdullah was not able to coordinate the support of all Tajik elites, he did seal the endorsement of some prominent ones. Most notably, he was able to secure the backing of the governor of Balkh, Atta Mohammad Nur, who in the 2004 election delivered a large number of votes to Karzai from Balkh province.154 As a result, Dr. Abdullah received 14.3% higher votes than Yunis Qanooni did in the 2004 election.155 In the presidential election of 2009, candidates, elites, and even electoral commissions undermined the Election Laws.156 The violation of laws began with that of the Constitution, when the IEC postponed the election for four months.157 According to the Constitution, the presidential election must be held no later than April 22;158 the IEC, however, postponed it to August 20.159 Furthermore, electoral fraud was so obvious that it led to the invalidation of almost 23% of the votes.160 Many high-ranking officials involved themselves in elections using their position and governmental resources, while their

152. Bijlert, supra note 29, at 11 n.32. 153. Biography: Gen. Abdula Malik Pahlawan, AFGHANISTAN ONLINE, http://www.afghan-web.com/bios/today/others.html#Malik (last visited Jan. 11, 2016). 154. Bijlert, supra note 29, at 11 n.32. 155. Qanooni won 16.3% votes in 2004 and Abdullah won 30.6% votes in 2009. For their election result see IEC 2004, supra note 47 and IEC 2009, supra note 48. 156. For full knowledge about electoral fraud during 2009 presidential election see generally Martine van Bijlert, Who Controls the Vote? Afghanistan’s Evolving Elections, AFGHANISTAN ANALYSTS NETWORK, Sept. 15, 2010, https://www.afghanistan-analysts.org /publication/aan-papers/who-controls-the-vote-afghanistans-evolving-elections/. 157. Dexter Filkins, Afghan Presidential Election Delayed, N.Y. TIMES, Jan. 29, 2009, http://www.nytimes.com/2009/01/30/world/asia/30afghan.html?_r=0. 158. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, art. 61. 159. Filkins, supra note 157. 160. Archie Tse, Audit Finds Almost a Quarter of Afghan Vote Is Fraudulent, N.Y. TIMES,http://www.nytimes.com/interactive/2009/10/20/world/asia/1020-afghan-recount- analysis.html (last visited Jan. 12, 2016). MOBASHER 4/18/2016 3:11 PM

382 GONZAGA LAW REVIEW Vol. 51:2 involvement was clearly prohibited by electoral laws.161 Some of these officials, including Omar Zakhilwal (then acting Minister of Finance),162 Hamdard (governor of Paktia),163 Asadullah Khalid (acting Minister of Border affairs),164 Ghulam Farooq Wardak (then Minister of Education); Hanif Atmar (Minister of Interior), and Omar Daudzai (Chief of Staff), campaigned for the incumbent Karzai.165 Others, like Atta Mohammad Noor, campaigned for Dr. Abdullah.166 The most notorious violation was Karzai’s release of drug traffickers and his pardon of Dostum in a bid for his reelection.167 In 2009, prior to the election, Karzai issued a decree ordering the release of five convicted drug traffickers, one of which was the nephew of his campaign manager, Din Mohammad.168 The decree justified their release as an act made “‘out of respect’ to their well-known family.”169 Karzai’s second notorious compromise was his reconciliation with Dostum after Dostum’s widely publicized criminal assault of Akbar Bai, Dostum’s

161. ELECTION LAW Jun. 8, 2014, S. 1112, art. 66. 162. Later, Zakhilwal was accused of spending the government’s money and bribes on Karzai’s campaign, but this allegation was never investigated by the attorney general’s office. See Dexter Filkins, The Afghan Bank Heist: A secret investigation may implicate dozens of high-ranking government officials, THE NEW YORKER, Feb. 14, 2011, http://www. newyorker.com/magazine/2011/02/14/the-afghan-bank-heist. 163. Juma Khan Hamdard, the governor of Paktia, called on Pashtuns of the North to vote for Dr. Ghani only. See Zafar Shah Rouyee, Tensions rise before Afghan run-off, AFGHANISTAN TODAY, (Oct. 6, 2014), http://www.afghanistan-today.org/en/articles/ politics/716/; Amini, supra note 70. 164. Reynolds et al, supra note 132, at 9; see also Bijlert, supra note 29, at 10. 165. HOMAYOUN, supra note 87, at 20. 166. Jon Boone, Hamid Karzai pulls out of historic TV debate just hours before broadcast, THE GUARDIAN, Jul 22, 2009, http://www.theguardian.com/world/2009/ jul/22/hamid-karzai-tv-debate-afghanistan; see also, Profiles of Afghan Power Brokers, CENTER FOR AMERICAN PROGRESS, (Oct. 26, 2009), https://www.americanprogress.org/ issues/security/news/2009/10/26/6734/profiles-of-afghan-power-brokers/#8. 167. See Farah Stockman, Karzai’s Pardons Nullify Drug Court Gains: Well-known traffickers set free ahead of election, BOSTON GLOBE, (Jul. 4, 2009), http://www. rawa.org/temp/runews/2009/07/04/karzai-s-pardons-nullify-drug-court-gains-2.phtml (dis- cussing Karzai’s pardoning of drug traffickers); see also Sayyad Salahuddin, Karzai Pardons Five Afghan Heroin Traffickers, REUTERS, Jul. 9, 2009, http://www.reuters.com/article/2009/ 07/09/us-afghanistan-pardons-idUSTRE5684FW20090709. See also Williams, supra note 61 (discussing Karzai’s Pardon of Dostum). 168. Bijlert, supra note 29, at 11. 169. Farah Stockman, Karzai’s Pardons Nullify Drug Court Gains: Well-known traffickers set free ahead of election, BOSTON GLOBE, Jul. 4, 2009, http://www. rawa.org/temp/runews/2009/07/04/karzai-s-pardons-nullify-drug-court-gains-2.phtml. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 383 former ally and a Turkmen elite.170 When the attorney general insisted on prosecuting Dostom, Dostum fled to Turkey.171 However, just prior to the election, in order to access Uzbek vote banks, Karzai not only pardoned Dostum, he also allowed him back to the country and promised his party, Junbish-i-Mili, some cabinet seats and other high offices.172 Both moves were highly criticized at a national and an international level.173

3. The Presidential Election of 2014

The presidential election of 2014 broke from the pattern seen in the 2004 and 2009 elections, where a Pashtun candidate led and a Tajik candidate followed.174 This time, a Tajik candidate, Dr. Abdullah, led the race, whereas an ethnic Pashtun, Dr. Ghani, came in second.175 This dramatic change from previous presidential elections raised three important questions: First, if ethnicity determines the election outcome, how could a Pashtun candidate lag behind a Tajik candidate? Second, how could a Tajik candidate win 45% of the votes, while the Tajik population was estimated at a mere 27%? Conversely, how could a Pashtun candidate obtain only 31.6% of the votes, while the Pashtun population

170. Brian Glyn Williams, Dostum: Afghanistan’s Embattled Warlord, TERRORISM MONITOR, April 17, 2008, at 1, http://www.brianglynwilliams.com/pdfs/dostum_embattled_ warlord.pdf?tx_ttnews[tt_news]=4861&tx_ttnews[backOid]=167&no_cache=1. 171. Id. at 2. 172. HOMAYOUN, supra note 87, at 15. 173. Williams, supra note 61. 174. The 2014 election also had two other distinctive features from 2004 and 2009 elections: First, Afghans for the first time experienced a runoff election in 2014. Although a runoff was also possible in 2009 when Karzai’s votes decreased to below 50% after a reevaluation of ballots. Dr. Abdullah stepped down and, thus, the runoff was called off. Laura Rozin, Abdullah withdraws from runoff, POLITICO, Nov. 1, 2009, http://www.politico.com /story/2009/11/abdullah-withdraws-from-runoff-028986. However, in the presidential election 2014 none of the two contesting candidates stepped down. On the contrary, the runoff was held with more heated campaigns from both sides. Mujib Mashal, Heated rhetoric, but few real differences between Afghan front-runners, AL JAZEERA AMERICA, Jun. 12, 2014, http://america.aljazeera.com/articles/2014/6/12/afghan-election-runoffghaniabdullah.html. Second, the covert ethnic mobilization by campaigners became more blatant in the presidential election of 2014. The rival camps not only blatantly accused each other of ethnic mobilization but they also called on their groups to support a particular candidate. See Rouyee, supra note 163; Amini, supra note 70. For instance, Dostum, the first vice-president candidate of Ashraf Ghani, during a campaign in Baghlan province said that those Uzbeks and Turkmens who do not vote for his side are traitors. Juma Khan Hamdard, the governor of Paktia, called on Pashtuns to vote for Dr. Ghani only. See Amini, supra note 70. These statements, made in a huge crowd of supporters during the campaigns, caused serious dissatisfaction, debate, and accusation among Afghans. Rouyee, supra note 163. 175. IEC 2014, supra note 49. MOBASHER 4/18/2016 3:11 PM

384 GONZAGA LAW REVIEW Vol. 51:2 was estimated around 42%? The answer to each of these questions points to the failure of strategic coordination among Pashtun candidates, a lack of information about viable candidates among Pashtun voters, and Dr. Abdullah’s success in cross-ethnic coalitions. In the initial list of candidates, 10 out of 11 were Pashtun, and only one was an ethnic Tajik.176 Just prior to the election, three Pashtun candidates pulled out, which only increased the chances of the other two Pashtun candidates.177 The seven remaining candidates held several meetings behind closed doors to decide whether one should step aside in favor of the other.178 None, however, yielded to another;179 as a result, Pashtun voters experienced an expectation crisis as to who would be the most viable candidate. The frustration of Pashtun voters became more obvious when Pashtun tribal elders from 26 provinces held a Jirga (council) in Kabul to reduce the number of their candidates and unite behind one.180 Considering the strong performances of different Pashtun contenders in different regions, it is not surprising that they did not back down. For instance, Dr. Rassul received a majority of votes in the Southern Provinces.181 Sayyaf, who ranked fourth among the candidates, was able to obtain 7% nationwide.182 Dr. Ghani scored 26%, mostly from Pashtun dominated provinces in the East and

176. Salih Dogan, 2014 Afghanistan Presidential Election: An Ethnic Analysis, CAUCASUS INTERNATIONAL 94 (Winter 2014), http://www.academia.edu/7214404/2014_ Afghanistan_Presidential_Election_An_Ethnic_Analysis. 177. The initial list of candidates included Abdul Qayum Karzai, Abdulrab Rassoul Sayyaf, Abdul Rahim Wardak, Abdullah Abdullah, Ashraf Ghani Ahmadzai, Daud Sultanzoy, Hedayat Amin Arsala, Mohammad Shafiq Gul Agha Sherzai, Qutbuddin Helal, Sardar Mohammad Nader Naeem, and Zalmai Rassoul. Emma Graham-Harrison, Afghanistan election guide: everything you need to know, THE GUARDIAN (Feb. 3, 2014), http://www.theguardian.com/world/2014/feb/03/afghanistan-election-guide-candidates-list. Three candidates pulled out prior to the election only to endorse other candidates. These candidates include Abdul Qayum Karzai, President Karzai’s brother, Sardar Mohammad Nader Naeem, and General Abdul Rahim Wardak. See Safi, supra note 32. 178. Hodge, supra note 77. 179. Id. 180. Thomas Ruttig, A Joint ‘Durrani’ Candidate for President? After the Unity Jirga in Kabul, AFGHANISTAN ANALYSTS NETWORK, Mar. 5, 2014, https://www.afghanistan- analysts.org/a-joint-durrani-candidate-for-president-after-the-unity-jirga-in-kabul/. 181. Frud Bezhan, Surprises in Preliminary Afghan Election Results, RADIO FREE EUROPE RADIO LIBERTY, Apr. 28, 2014, http://www.rferl.org/content/surprises-in-preliminary- afghan-election-results/25365544.html. 182. IEC 2014, supra note 49. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 385

Uzbek dominated provinces in the North.183 Indeed, during the election, Pashtun candidates campaigned against one another, rather than against Dr. Abdullah.184 In addition, there are different kinds of long-term payoffs that might motivate candidates not to pull out of the election.185 Some candidates wanted to raise their profile or establish themselves as the main leaders of their constituencies.186 Others simply wanted to increase their political capital through their votes for bargaining in a runoff election. Should that have been their strategy, it worked well; in the second round, both candidates were competing to gain the endorsement of the losers.187 Dr. Abdullah, unlike Dr. Ghani, did not have to worry about electoral coordination, since he did not have any Tajik opponents who would have potentially split the Tajik votes with him. He only had to focus on coordinating votes from other ethnic groups through bargaining with their elites. Adding Mohaqiq to his ticket was, for the most part, the source of his extra ethnic votes.188 Unlike the two previous elections, Hazaras and Uzbeks did not have any candidates in the presidential election of 2014. Perhaps Hazara and Uzbek elites have realized that their interest would be better served by entering into political bargaining, rather than by running for election themselves. It is also likely that the requirement of Article 45 of Election Law might have impeded the candidacy of elites from these ethnic groups.189 Dostum and Mohaqiq, for example, were unlikely to collect the required 2% of voting cards from 20 provinces, since Hazaras and Uzbeks live in a very limited number of provinces. In addition, as

183. Bezhan, supra note 181. 184. GEETI MOHSINI, Blame Game Between Rassoul & Ashraf Ghani Campaigns, TOLONEWS, Mar. 20, 2014, http://www.tolonews.com/en/election-2014/14271-blame-game- between-rassoul-a-ashraf-ghani-campaigns; see also Muhammad Hassan Khetab, Electoral watchdog fines Sherzai, Rassoul, PAZHWAK AFGHAN NEWS, Feb. 4, 2014, http://elections.pajhwok.com/en/2014/04/02/electoral-watchdog-fines-sherzai-rassoul. 185. COX, supra note 43, at 152. 186. Some candidates “intend to use their new found fame to negotiate a settlement with other candidates (and to step back in their favor), to be offered a high-level position after the elections, or to successfully run for parliament.” Bijlert, supra note 29, at 9. 187. By Najmuddin A Shaikh, Will Dr Abdullah Be president in Afghanistan?, INT’L N.Y. TIMES: THE EXPRESS TRIBUNE, May 12, 2014, http://tribune.com.pk/story/707135/will- dr-abdullah-be-president-in-afghanistan/. 188. Amanullah Ghilzai, Ethnic Divisions and Alliances in the Afghan Elections, MIDDLE EAST INSTITUTE (Apr. 1, 2014), http://www.mei.edu/content/ethnic-divisions-and- alliances-afghan-elections. 189. According to Article 45 of the Election Law, a presidential candidate has to provide the Independent Election Commission of Afghanistan with “one hundred thousand voters, from a minimum of twenty provinces, two percent from each province.” See ELECTION LAW Jun. 8, 2014, S. 1112, art. 45. MOBASHER 4/18/2016 3:11 PM

386 GONZAGA LAW REVIEW Vol. 51:2 the elite of smaller groups, they were less likely able to attract the endorsement of elites from larger groups. Mohaqiq and Dostum have also learned from the setbacks of their candidacy in the 2004 election. In the presidential election of 2004, by running for presidency against Karzai, they eventually sidelined themselves in his forthcoming government. In his transitional and first term presidency, Karzai was pursuing any opportunity to marginalize these ethno-political elites.190 Having learned from experience, Mohaqiq and Dostum likely chose to bargain with viable candidates. Eventually, their partnerships with candidates promised a flow of votes from their supporters to their endorsees.191 The presidential election of 2014 was characterized by an enormous violation of electoral laws—even more so than the 2009 election.192 The fraud in the second round was so massive that it resulted in reevaluation of all ballot boxes.193 Ziaulhaq Amarkhil, Chief Electoral Officer of IEC, who allegedly led most of the electoral fraud in the second round, walked out of the office without being prosecuted.194 The resulting electoral deadlock led to another violation of law by IEC and runoff candidates. The Election Law requires the Independent Election Commission of Afghanistan “to announce and publish the final election results after the completion of the counting process . . . .”195 Several articles of the Constitution also articulate the importance of announcing election results.196 However, Dr.

190. HAFIZULLAH EMADI, DYNAMICS OF POLITICAL DEVELOPMENT IN AFGHANISTAN: THE BRITISH, RUSSIAN AND AMERICAN INVASION 214, 216 (2010). 191. If we combine the votes of all Pashtun candidates and then compare that total with the votes of Abdullah, we see a clear correlation between ethnicity and votes. The total votes of Pashtun candidates are 55%, which is 10% higher than Abdullah’s total votes. This means that if Dr. Ghani was the only Pashtun candidate, Dr. Ghani might have outperformed Dr. Abdullah, as he did in the second round. Meanwhile, the 55% votes of Pashtun candidates, with a little difference, reflect the sum of Pashtun and Uzbek populations, which is 51%. However, 45% votes of Dr. Abdullah do not exactly reflect the combined population of Tajiks and Hazaras, which is estimated 36%. This slight difference is due to some limited Pashtun votes for Abdullah, and votes from other minority groups. Possible error in the estimation of ethnic populations may also be the cause of this difference. 192. Noah Coburn, Afghanistan: The 2014 Vote and the Troubled Future of Elections, AFGHANISTAN: OPPORTUNITY IN CRISIS SERIES NO. 8, Mar. 2015, at 1, 4, http://www. chathamhouse.org/sites/files/chathamhouse/home/chatham/public_html/sites/default/files/20 140200AfghanistanElectionsFraudPerelliSmith.pdf. 193. Id. 194. Id. 195. ELECTION LAW Jun. 8, 2014, S. 1112, art. 59. 196. THE CONSTITUTION OF THE ISLAMIC REPUBLIC OF AFGHANISTAN Jan. 26, 2004, arts. 61, 83, and 160. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 387

Ghani and Dr. Abdullah entered into an agreement that instructed the IEC not to disclose the tally in order to relieve the electoral crisis.197 In another case, a supporter of Dr. Abdullah allegedly shot and killed a child in Jalalabad.198 Despite many examples of election violations like this one, some violations did not go overlooked. For example, the IEC fined Dr. Ghani and Dr. Rassul on multiple occasions for using governmental resources for their own campaign purposes.199 IEC also fined Shirazai for using abusive language against Dr. Rassul.200

C. Province-Wide Analysis of Ethnic-Electoral Relations

A cross-provincial analysis is helpful for understanding why a candidate may get higher votes in some provinces and not in others. My findings suggest that tallies across provinces, to a large extent, reflect ethnic divisions. Some findings also suggest cross-ethnic political bargains.201 These findings indicate a higher level of cross-ethnic voting in some provinces over others. Hence, for the purpose of qualitative analysis, the provinces of special interest are the ones that show a higher level of cross-ethnic voting or the ones in which voting pattern changes dramatically from one election to another. This article incorporates analysis of data from both ethnically diverse and ethnically homogenous provinces. For instance, Kunduz, Ghazni, Baghlan and Kabul are the most ethnically heterogeneous provinces of Afghanistan.202 Zabul, Panjshir, and Daykundi are the most ethnically homogenous—or nearly homogenous—provinces, composed of Pashtuns, Tajiks, and Hazaras

197. Ruttig, supra note 91, at 1. 198. Ghanizada, More attacks reported by Abdullah backers on Ghani’s supporters, KHAAMA PRESS, Jun. 15, 2014, http://www.khaama.com/more-attacks-reported-by-abdullah- backers-on-ghanis-supporters-618. 199. Muhammad Hassan Khitab, Ghani Fined For Failing To Return Govt. Vehicles, PAZHWAK AFGHAN NEWS, Feb. 15, 2014, http://www.elections.pajhwok.com/en/ 2014/02/16/ghani-fined-failing-return-govt-vehicles; see also Khetab, supra note 184. 200. Khetab, supra note 184. 201. Coburn and Larson, supra note 21, at 1. 202. Table I-A in the Appendix shows that these Kunduz, Baghlan, Kabul and Ghazni are very diverse having considerable number of people from different ethnic groups. In addition to the four major ethnic groups shown in the Appendix, some other ethnic groups also live in these provinces. For example, Kuduz has a large number of Turkmens. Kabul has a considerable number of people Quzelbash, Arab, Sikhs, Hindus and other ethnic groups. See AFGHANISTAN PRT HANDBOOK: OBSERVATIONS, INSIGHTS, AND LESSONS, CENTER FOR ARMY LESSONS LEARNED, AFGHANISTAN PROVINCIAL RECONSTRUCTION TEAM HANDBOOK: ANNEX A - NATIONAL AND PROVINCIAL DATA, UNITED STATES ARMY COMBINED ARMED CENTER, 63, 75, 85, 95(Feb. 2008), http://usacac.army.mil/sites/default/files/publications/11-16.pdf (last visited Dec. 2, 2014) [hereinafter “CAC”]. MOBASHER 4/18/2016 3:11 PM

388 GONZAGA LAW REVIEW Vol. 51:2 respectively (See Table I-A in the Appendix). No province is exclusively populated by Uzbeks. Nonetheless, the sample includes Faryab, Samangan, and Jauzjan, which represent the largest population of Uzbeks in the region.203 In addition, this article includes analysis of data from Nooristan and Nimroz, in which none of the above groups is the majority.

1. Electoral Politics in the 2004 and 2009 Presidential Elections

This section compares ethnic-electoral patterns across provinces and elections. Here, the votes of the leading candidates of the 2004 election are compared with those of 2009 election. The presidential election of 2014 is analyzed in a separate subsection, because of its distinctive features. The provinces of special interest are the ones that show a high level of cross-ethnic voting or the ones in which voting pattern changes dramatically from one election to another.

a) Hamid Karzai, The Unparalleled Pashtun Candidate in the 2004 and 2009 Elections

In both 2004 and 2009, Karzai had been the only leading Pashtun candidate in the elections.204 As figure IV indicates, Karzai’s votes, largely, were proportional to the Pashtun population across provinces: In both presidential elections, he won a higher percentage of votes in the provinces that had a Pashtun majority. Conversely, he won fewer votes in those provinces in which Pashtuns were not the majority (see figure V). In 2004, Karzai’s best poll results came from the two southeastern provinces, Khost and Paktika, where he earned 96% in each—the Khost and Paktika populations are comprised of 99% and 96% Pashtuns, respectively.205 On the other hand, Karzai’s worst poll performance was in Panjshir, where he won below 1% in 2004.206 The estimated Pashtun population in this province is nearly 0%.207 Similarly, in the 2009 election, Karzai’s highest scores were 81% in

203. Its based on estimations by AIMS, NPS, and CAC Handbook. For the estimated population of Uzbeks and other ethnic groups in Faryab, Samangan and Jauzjan, see Table I- A in the Appendix. 204. IEC 2004, supra note 47; IEC 2009, supra note 48. 205. See IEC 2004, supra note 47. Karzai also scored 90% to 95% in Kunarha, Nangerhar, Kandahar, and Helmand. Id. Pashtun population ranges about the same in these provinces. Id. 206. See id. 207. IEC 2009, supra note 48. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 389

Kunar and 75% in Nangarhar, which are Pashtun dominated provinces; his lowest votes in the polls were 24% in Ghazni and 32% in Badakhshan.208 Notably, there were some gaps between Karzai’s votes and his ethnic population across provinces. Interestingly, these gaps indicate that on average Karzai secured more votes per capita in Pashtun minority provinces, whereas he gained lower votes per capita in Pashtun dominated provinces (see figure V). The significance of these discrepancies lies in Karzai’s cross-ethnic coalitions and his intra-ethnic competitions. For instance, as figure IV illustrates, Bamyan, Panjshir, Badakhshan, and Nooristan are the provinces in which the population of Pashtuns is minimal. Nonetheless, in the polls, Karzai’s numbers were significantly high in these provinces, especially in 2009.209

Pashtun Population & Hamid Karzai's Votes in 2004 & 2009 1.20 1.00 0.80 0.60 0.40 0.20 0.00 Ghor Zabul Logar Kabul Farah Khost Kunar Paktya Faryab Takhar Ghazni Nimroz Paktika Jauzjan Kunduz Badghis Baghlan Panjshir Bamyan Hilmand Noristan Urezgan Laghman Daykundi Kandahar Samangan Nangarhar Badakhshan

Pashtun Population Karzai In 2004 Karzai In 2009

FIGURE IV. This graph compares the votes of Karzai from 2004 and 2009 elections with the population of Pashtuns in 26 provinces.210

208. Id. 209. Id. 210. The data about the votes of candidates are collected from: IEC 2004, supra note 47; also IEC 2009, supra note 48. The data about the estimated population of Pashtuns is collected from: PROGRAM FOR CULTURAL AND CONFLICT STUDIES, http://www. nps.edu/Programs/CCS/ExecSumm.html (last visited, Jan. 6, 2015) [hereinafter “NPS”]. MOBASHER 4/18/2016 3:11 PM

390 GONZAGA LAW REVIEW Vol. 51:2

In 2004, Karzai won 9% of the total votes in Bamyan, which has an estimated 1% Pashtun population. See Table I-A in the Appendix. In the same province, he garnered 39% of the vote in the presidential election of 2009.211 The disproportionally higher number of votes for Karzai in Bamyan is due to his strategic coalitions with the Hazaras’ prominent leaders. Bamyan is mostly populated by Hazaras, who comprise an estimated 67% of the population there. See Table I-A in the Appendix. The two Hazara kingmakers who claim to have the support of Hazaras nationwide are Karim Khalili and Haji Mohammad Mohaqiq.212 In the 2004 election, Karzai introduced Khalili as his second candidate for vice president;213 Khalili’s endorsement was likely to increase Karzai’s vote in Bamyan.214 In that race, however, Karzai had to compete with Mohaqiq for Hazara votes. In the end, Mohaqiq defeated him by claiming 76% of the votes in Bamyan.215 Later, in the presidential election of 2009, Karzai gained the endorsement of both Khalili and Mohaqiq. In exchange for their endorsements, Khalili was introduced as his second vice president candidate, and Mohaqiq was promised 5 seats in the cabinet and other governmental positions.216 Karzai also promised to upgrade two districts of Hazara residents to provinces.217 In addition to these elites, Karzai secured the support of the most prominent Shiite clerk, Ayatollah Asef Mohseni, reportedly in exchange for signing the controversial Shiite Personal Status Law.218 Hazaras are predominantly Shiites.219 With endorsements of these elites, it was not surprising that Karzai topped the votes in Bamyan, although with only 0.5% of a separation between him and Dr.

211. IEC 2009, supra note 48. 212. Decision Point 2009, supra note 20, at 26. 213. Antonio Giustozzi, Process of Political Ethnicization in Today’s Afghanistan, in, AFGHANISTAN: IDENTITY, SOCIETY AND POLITICS SINCE 1980, 211 (Micheline Centlivres- Demont ed., 2015). 214. Who Is Who In Afghanistan: Khalili, Mohammad Karim Chalili Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=1571 &task=view&total=3158&start=2523&Itemid=2 (last visited Feb. 4, 2016). 215. IEC 2004, supra note 47. 216. Makia Monir, Karzai Vow Draws Negative Reaction, E-ARIANA, Jun. 16, 2009, http://e-ariana.com/ariana/eariana.nsf/allDocs/7cb3e5dbec08cc41872575d90052e074!Open Document&Click=; see also Montagne, supra note 86; Bijlert, supra note 29, at 10. 217. Montagne, supra note 86; see also Bijlert, supra note 29, at 10; HOMAYOON, supra note 87, at 15; Oppel Jr., supra note 86. 218. Robert Mackey, Afghan Husbands Win Right to Starve Wives, N.Y. TIMES, Aug. 17, 2009, http://thelede.blogs.nytimes.com/2009/08/17/afghan-husbands-win-right-to- starve-wives/; see also Bijlert, supra note 29, at 10. 219. THOMAS BARFIELD, AFGHANISTAN: A CULTURAL AND POLITICAL HISTORY 26 (2010). MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 391

Bashardost—who defeated Karzai in all other constituencies dominated by Hazaras.220 Panjshir is another province in which Karzai performed astonishingly well in the 2009 election.221 Panjshir was the home province of Karzai’s leading opponents in both the 2004 and 2009 elections.222 In order to challenge these opponents in their home province, Karzai choose his vice presidential candidates from Panjshir.223 Moreover, Karzai offered cabinet seats and other high positions to several other elites from this province: for example, General Bismellah Khan was appointed as the minister of interior and later as minister of defense in both of Karzai’s administrations.224 Further, Amerullah Saleh was the chief of the Intelligence Service until 2010.225

220. Nonetheless, Karzai’s poll numbers increased in other Hazara constituencies also in 2009. For instance, in Daykundi, Karzai secured 29.1% votes in 2009 compared to 13.1% of the votes in 2004. In Ghor, his poll increased from 28% to 36%. The only exception is Ghazni, the home province of Bashardost, in which Karzai’s poll number declined to even less than the Pashtun Population: from 52% in 2004 to 24% in 2009. What can be concluded from Karzai’s poll in Bamyan and other Hazara constituencies is that the increase in his votes was likely a direct result of the political bargains that he sealed with Hazara Tikadaran. See IEC 2004, supra note 47; IEC 2009, supra note 48. 221. IEC 2009, supra note 48. 222. Dr. Abdullah Abdullah, the runner up in both 2009 and 2014, though from a Kandahari father, is predominantly considered a Panjshiri because of his mom being from this province and also because he fought alongside Ahmad Shah Masoud during soviet war and resistance against Taliban. See Who Is Who In Afghanistan, supra note 130. Mohammad Yunis Qanooni, the runner up in 2004 election, is also from Panjshir. See Biography: Mohammad Yunis Qanuni, AFGHANISTAN ONLINE, http://www.afghan-web.com/bios/today/qanuni.html (last visited Nov. 10, 2015). 223. These negotiations and appointments likely helped Karzai primarily with several goals. First, he challenged his leading opponents on their home turf. Second, he gained the support of Panjshiri elites who were many and powerful and who could challenge his government even after the elections. Third, he used the popularity and influence of these elites to attract Tajiks’ votes from across provinces. Some Panjshiri elites like Marshal Qasim Fahim, who endorsed Karzai in 2009, Bismullah Khan, Amrullah Saleh as well as Yunis Qanooni, and Dr. Abdullah, who were Karzai’s rivals in 2004 and 2009 have influences beyond Panjshir and among all Tajik population. 224. Who Is Who In Afghanistan: Bismullah Khan Mohammadi Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=1068&task=view &total=3158&start=1755&Itemid=2 (last visited Nov. 10, 2015). 225. Who Is Who In Afghanistan: Amrullah Saleh Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=1571&task=view& total=3158&start=2523&Itemid=2 (last visited Nov. 10, 2015). MOBASHER 4/18/2016 3:11 PM

392 GONZAGA LAW REVIEW Vol. 51:2

In addition, Karzai introduced Ahmad Zia Masoud, a Panjshiri elite, as his first vice president candidate in the 2004 election.226 Ahmad Zia Masoud and Ahmad Wali Masoud, who also endorsed Karzai,227 are brothers of the late Ahmad Shah Masoud, who ruthlessly fought against the Soviet invasion and, later, against the Taliban.228 He was assassinated by Al-Qaeda just two days before the 9/11 attacks.229 Ahmad Shah Masoud is regarded as a national hero among Tajiks. However, selecting Ahmadzia Masoud as a running mate was a miscalculation, since Ahmad Zia Masoud, unlike his late brother, was not popular among Tajiks or even Panjshiris.230 Together they only managed to earn less than 1% of the votes in Panjshir in 2004.231 Yunis Qanooni, the Panjshiri contender, won 95% of the votes in that province, leaving the rest to another Panjshiri candidate.232 Marshal Qasim Fahim’s support of Karzai as his running mate in 2009, however, increased Karzai’s votes to 29% in Panjshir.233 Fahim, born in Panjshir, is one of the influential Tajik ethno-political elites, who is popular for fighting against the Soviets and the Taliban—shoulder to shoulder with late Ahmad Shah Masoud.234 The coalition of Karzai and Fahim, however, was not able to defeat Dr. Abdullah, who won 68% of the votes.235 Dr. Abdullah, despite being from a Kandahari father and a Panjshiri mother, is popularly known as a Tajik leader and a Panjshiri son.236 Ultimately, Dr. Abdullah outperformed the Karzai- Fahims’ coalition in Panjshir, as was expected.

226. Amin Tarzi, Dispute Within Political Party Ranks Over Support For Afghan Leader, As Claims of Support Are Conflicting, AFGHANISTAN NEWS CENTER, Oct. 8, 2004, http://www.afghanistannewscenter.com/news/2004/october/oct92004.html. 227. Id. 228. Peter Bayon DeNeufville, Ahmad Shah Massoud and the Genesis of the Nationalist Anti-Communist Movement in Northeastern Afghanistan, 1969-1979 16-18 (May 2006) (unpublished Ph.D. dissertation, University of London) (on file with author). 229. Erin Friar, Afghanistan Election: The Crucial Choice of VP by Ghani (Series), THE GUARDIAN, Apr. 3, 2014, http://guardianlv.com/2014/04/afghanistan-election-the- crucial-choice-of-vp-by-ghani-series/#EXwAlRukfvhx1sTy.99. 230. His unpopularity is probably because of his absence from the struggle and political arena for over 12 years since the fall of Soviet backed government in Afghanistan. See Who Is Who In Afghanistan: Ahmad Zia Massoud Biography, AFGHAN BIOS, http://www.afghan- bios.info/index.php?option=com_afghanbios&id=1032&task=view&total=3162&start=1680 &Itemid=2 (last visited Nov. 10, 2015); see also Decision Point 2009, supra note 20, at 21. 231. IEC 2004, supra note 47. 232. IEC 2004, supra note 47. 233. IEC 2009, supra note 48. 234. Biography: Mohammad Qasim Fahim, AFGHANISTAN ONLINE, http://www. afghan-web.com/bios/today/fahim.html (last visited Nov. 10, 2015). 235. IEC 2009, supra note 48. 236. Dubow, supra note 23, at 24. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 393

Badakhshan provides another example of cross-ethnic voting, where Karzai garnered 30% votes in 2004237 and 34% of the votes in 2009.238 The estimated population of Pashtuns in Badakhshan is below 1%, which suggests a 29% and 33% difference between Karzai’s votes and the total Pashtun population in that province. See Table I-A in the Appendix. Karzai likely secured this lead by forming an alliance with Burhanuddin Rabani. Born in Badakhshan, Rabani was the leader of Jamiat Islami Party and the president of Afghanistan during the civil war.239 Karzai obtained his endorsement in exchange for cabinet seats and high offices for his party and selected figures in both the 2004 and 2009 elections.240 Karzai’s coalition with prominent Tajik figures like Rabani, Fahim, and Ismael Khan (a Tajik elite who helped Karzai win the relative majority in Herat in both elections)241 likely secured him a handful of votes in other Tajik constituencies, such as Takhar, Kunduz, and Ghor. Nooristan presents an interesting case of cross-ethnic voting because it predominantly hosts a minority group called Noristani, which is distinct from the four large ethnic groups mentioned in this article.242 Nevertheless, the province returned a high percentage of votes for Karzai and his Tajik rivals. Karzai received 57.7% of the votes in 2004243 and 46.2% of the votes in 2009.244 Karzai was followed by Yunis Qanooni, who obtained 31% of the vote in 2004,245 and Dr. Abdullah, who garnered 41.2% of the vote in 2009.246 This split of votes between Karzai and the Tajik candidates reflects that Nooristanis are geographically in close contact with Pashtun dominated provinces in the South— Kunar and Laghman—and the Tajik dominated provinces in the North—

237. IEC 2004, supra note 47. 238. IEC 2009, supra note 48. 239. Biography: Burhanuddin Rabani AFGHANISTAN ONLINE, http://www.afghan- web.com/bios/today/brabbani.html (last visited Nov. 16, 2015). 240. Influential ex-president backs Karzai in election, CHICAGO TIMES, Oct. 4, 2014, http://articles.chicagotribune.com/2004-10-04/news/0410040253_1_president-hamid-karzai- presidential-election-president-burhanuddin-rabbani; MARA TCHALAKOV, INSTITUTE FOR THE STUDY OF WAR, AFG. REP. 10: THE NORTHERN ALLIANCE PREPARES FOR AFGHAN ELECTIONS IN 2014 21, 23 (Aug. 2013), http://www.understandingwar.org/report/northern-alliance- prepares-afghan-elections-2014. Some authors, however, suggest that Rabbani endorsed Dr. Abdullah in 2009. See HOMAYOUN, supra note 87, at 21. 241. KATZMAN, supra note 24, at 16. 242. Some suggest that Nuristanis are the descendants of Greek solders who came with Alexander to conquer the region. See NICHOLAS BARRINGTON ET AL, A PASSAGE TO NURISTAN: EXPLORING THE MYSTERIOUS AFGHAN HINTERLAND 123-25 (2006). 243. IEC 2004, supra note 47. 244. IEC 2009, supra note 48. 245. IEC 2004, supra note 47. 246. IEC 2009, supra note 48. MOBASHER 4/18/2016 3:11 PM

394 GONZAGA LAW REVIEW Vol. 51:2

Badakhshan and Panjshir; this close proximity is likely reflected in voting practices. Nooristanis are, however, geographically distant from Hazaras and Uzbeks, which may also explain why Hazara and Uzbek candidates received few votes in this province in both elections. Karzai’s 2004 performance in Faryab, Jauzjan, and Samangan demonstrates a normal pattern of association between the Pashtun population and his total votes—with only marginal discrepancies. He polled nearly 10% in Faryab, 12% in Jauzjan, and 7% in Samangan in 2004.247 In 2009, however, his votes increased by 60% in Faryab, 58% in Jauzjan, and 43% in Samangan.248 This disparity is likely reflective of Karzai’s coalition with Dostum in the absence of a viable Uzbek candidate in 2009. Significant portions of the population in all three provinces are Uzbeks: an estimated population of over 50% in Faryab, 40% in Jauzjan, and 30% in Samangan. See Table I-A in the appendix. Notably, combining Pashtun and Uzbek populations in these provinces produces results that were only slightly different from Karzai’s returns in 2009.249 On the other hand, as mentioned earlier, Karzai garnered slightly fewer votes than the estimated Pashtun population in Pashtun Majority provinces. This trend likely stems from Pashtun candidates who took away some Pashtun votes, contenders who were weaker in 2004 and stronger in 2009. For example, in 2009, Dr. Ghani Ahmadzai won 7% of total votes in the Pashtun dominated Helmand along with other Pashtun candidates receiving 11% votes.250 These splits of votes resulted in Karzai winning 73% votes, which was 17% lower than his votes in 2004.251 The same was true in Pashtun dominated Kandahar, the home province of Karzai. There, Dr. Ghani and other Pashtun candidates secured 14.5% of the total votes.252 Accordingly, Karzai’s votes decreased from 91% in 2004253 to 73.8% in 2009.254

b) Qanooni and Abdullah, The Tajik Runner Ups

The estimated population of Tajiks ranges from 0% in some provinces to nearly 100% in others. According to estimations, Daykundi, Nooristan, and

247. IEC 2004, supra note 47. 248. IEC 2009, supra note 48. 249. The slight difference between Karzai’s votes in 2009 election and the combined population of Pashtuns and Uzbeks is only 4% in Faryab and 1%. In Jauzjan, however, the difference is 11%, which is the highest. For a comparison, see Table I-A in the Appendix. 250. IEC 2009, supra note 48. 251. IEC 2004, supra note 47. 252. IEC 2009, supra note 48. 253. IEC 2004, supra note 47. 254. IEC 2009, supra note 48. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 395

Kunar have the lowest population of Tajiks, while Panjshir has the highest.255 In between, there are many provinces with Tajik majorities, and still others with Tajik minorities. See Table I-A in the appendix. A comparison of Tajik populations across provinces with the votes of Tajik leading candidates in the 2004 and 2009 elections suggests a strong ethnic voting in Tajik constituencies. As figure V illustrates, the smaller the Tajik Population in a province, the fewer the votes their candidates get in that province. Correspondingly, the larger the Tajik population, the more votes their leading candidates receive.

Tajik Population & Their Candidates' Votes in 2004 & 2009 1.20

1.00

0.80

0.60

0.40

0.20

0.00 Ghor Zabul Logar Kabul Farah Khost Kunar Paktya Faryab Takhar Ghazni Nimroz Paktika Jauzjan Kunduz Badghis Baghlan Panjshir Bamyan Hilmand Noristan Urezgan Laghman Daykundi Kandahar Samangan Nangarhar Badakhshan Tajik Population Qanooni In 2004 Abdullah In 2009

FIGURE V. This graph compares the votes of Tajik candidates from 2004 and 2009 elections with the population of Tajiks in 26 provinces.256

Cross-ethnic coalitions and intra-ethnic competitions explain some dis- crepancies between the Tajik population and their candidates’ votes in the provinces. The provinces with the highest discrepancies are Nooristan, Paktika, Kandahar, and Kunduz. Whereas cross-ethnic coalitions explain the relatively higher number of votes for candidates in Tajik minority provinces, intra-ethnic

255. The population of Tajiks is estimated less than 1% in Daykundi, Nooristan and Kunar and almost 100% in Panjshir. See Table I-A in the Appendix. 256. The data about the votes of candidates are collected from: IEC 2004, supra note 47; IEC 2009, supra note 48. The data about the estimated population of Tajiks is collected from: NPS, supra note 210. MOBASHER 4/18/2016 3:11 PM

396 GONZAGA LAW REVIEW Vol. 51:2 competitions and the lack of coordination reflect the split of votes in Tajik majority provinces. Paktika, a Southeastern province, is predominantly populated by Pashtuns.257 The Tajik population is estimated at less than 1% in this province,258 which is reflected in Qanooni’s votes, as he received just above 3% in 2004.259 Dr. Abdullah, however, netted 23.6% of the votes, significantly larger than the size of the Tajik population in Paktika.260 The higher votes of Dr. Abdullah were more likely due to poor security between Pashtun and Tajik areas, which influenced the turnout rates in those regions. This relationship between lack of security and lower voter turnout is better illustrated by examining the votes in the Urgon district of Paktitka. The Urgon district, which has an estimated population of 88,000, comprises 88% Pashtuns and less than 12% Tajiks.261 This difference would normally lead to the expectation that Abdullah would have won one eighth of Karzai’s votes, as was true for Qanooni in the 2004 election. However, Dr. Abdullah secured more than half of Karzai’s votes in Urgon, receiving 8,321 votes, compared to Karzai’s 16,093 votes.262 Dr. Abdullah’s unexpectedly higher number of votes in Paktika was likely due to insecurity in the Pashtun areas, leading to a substantial shift of turnout rates in his favor. Kandahar’s population is estimated at 1% Tajik and 98% Pashtun.263 Accordingly, Yunis Qanooni gained 3% of the votes in Kandahar in 2004.264 Dr. Abdullah, on the other hand, won 9.1% of the votes in 2009,265 which is 8.1% higher than the estimated Tajik population. Several ethnicity-related factors probably contributed to this outcome. First, Dr. Abdullah’s father was a Pashtun from Kandahar.266 Second, Kandahar was one of the first few provinces to which

257. CAC, supra note 202, at 107-08. 258. Id. 259. IEC 2004, supra note 47. 260. IEC 2009, supra note 48. 261. USAID & ALTAI CONSULTING, PRIORITY DISTRICT REPORT: URGON (PAKTIKA) 3 (Oct. 13, 2010), http://www.altaiconsulting.com/docs/media/2010/15.%20Paktika%20- %20Urgun.pdf 262. Afghanistan Election Data, PAKTIKA, http://afghanistanelectiondata.org/ election/2009/province/24#layers=district_map_openlayers_1&baseLayers=afghanistan- grey (last visited Nov. 16, 2015). 263. CAC, supra note 202, at 87-88. 264. IEC 2004, supra note 47. 265. IEC 2009, supra note 48. 266. Who Is Who In Afghanistan, supra note 130. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 397

Dr. Abdullah traveled during his campaign in 2009.267 Third, and most importantly, he sealed the support of some Kandahari figures, such as Noorul Haq Ulumi, a prominent former leftist, and Malalai Ishaqzai a female MP from Kandahar.268 More importantly, he appointed Homayoun Shah Assefi, a member of the royal family, as his first vice-presidential candidate.269 Kunduz is one of the most diverse provinces in Afghanistan. The Kunduz population is approximately 31.2% Pashtun, 22.5% Tajik, 6.7% Hazara, 20% Uzbek, with 19.7% of the population coming from other ethnic groups.270 In the presidential election of 2009, Dr. Abdullah’s number of votes surpassed 54%,271 which is two-and-one-half times the estimated Tajik population and 36% higher than Qanooni’s votes. Dr. Abdullah’s higher performance was likely due to the support of many local strongmen and Kunduz’s high level of instability, which reduced turnouts in Pashtun areas. One of Dr. Abdullah’s endorsers was Commander Mir Alam, who continued to lead a group of militia even after the Disarmament Demobilization and Reintegration (DRR) period from 2003-2006.272 Both Mir Alam and Dr. Abdullah are from Hizb-e-Jamiati Islami (Islamic Society Party).273 During the electoral deadlock in 2014, Mir Alam threatened to destabilize Kunduz if he was called upon by Dr. Abdullah.274 In addition to Mir Alam, an ethnic Tajik, Dr. Abdullah acquired the support of other local elites; this support included Nazari Turkman, an ethnic Turkman, and Qari Rahmatullah, an ethnic Aimaq, as well as General Malik, Gol Mohammad Pahlawan, and Ahmad Khan, all Uzbek commanders from surrounding provinces.275

267. Carlotta Gall & Ruhullah Khapalwak, As Afghan Vote Nears, Taliban Step Up Intimidation Campaign, N.Y. TIMES, Aug. 12, 2009, http://www.nytimes.com/2009/08/13/ world/asia/13kandahar.html. 268. Matthieu Aikins, 2010 Elections 29: Losing legitimacy–Kandahar’s prelimin- ary winners, AFGHANISTAN ANALYSTS NETWORK, Oct. 25, 2010, at 2, https://www. afghanistan-analysts.org/2010-elections-29-losing-legitimacy-kandahars-preliminary- winners/?format=pdf. 269. Tanya Goudsouzian, Afghan royal eyes poll success: Will former royal emerge as a key player in Afghanistan’s presidential election?, AL JAZEERA, Aug. 19, 2009, http://www.aljazeera.com/focus/afghanistanelections2009/2009/08/200981910755438497 .html. 270. AIMS, supra note 202. 271. IEC 2009, supra note 48. 272. Who Is Who In Afghanistan: Mir Alam Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=903&task=view&total =3079&start=1681&Itemid=2 (last visited Aug. 4, 2014). 273. Id. 274. Id. 275. This information was gathered through interview with few residents of Kunduz and other northern provinces while having the opportunity of meeting them in Seattle. MOBASHER 4/18/2016 3:11 PM

398 GONZAGA LAW REVIEW Vol. 51:2

A high level of insecurity in Kunduz also worked in Dr. Abdullah’s favor. Kunduz had a bloody year of Taliban-NATO conflict in 2009, resulting in turnouts decreasing from 283,987 in the 2004 election to merely 93,747 in the 2009.276 The Taliban threatened to disturb the election through violence.277 Their threats particularly affected the turnouts in Pashtun areas, which in turn lowered Karzai’s share of votes.

c) The Third Placed Hazaras: Mohaqiq and Dr. Bashardost

Hazaras, the third largest ethnic group in Afghanistan, mostly live in the central regions of Afghanistan, which is known as Hazarajat.278 Only a few provinces, like Bamyan and Daykundi, are dominated by Hazaras.279 A number of provinces have a sizable population of Hazaras, while others have only a few or no Hazaras. See figure VI. Hazara Population & Their Candidates' Votes in 2004 & 2009 1.00 0.80 0.60 0.40 0.20 0.00 Ghor Zabul Logar Kabul Farah Khost Kunar Paktya Faryab Takhar Ghazni Nimroz Paktika Jauzjan Kunduz Badghis Baghlan Panjshir Bamyan Hilmand Noristan Urezgan Laghman Daykundi Kandahar Samangan Nangarhar Badakhshan Hazara Population Mohaqiq In2004 Bashardost In 2009

FIGURE VI. This graph compares the votes of Hazara candidates from 2004 and 2009 elections with the population of Hazaras in 26 provinces.280

276. The 93,747 votes include the invalid ones that were removed by ECC. For knowing the valid and invalid votes see IEC 2009, supra note 48; IEC 2004, supra note 47. 277. Gall & Khapalwak, supra note 267. 278. Dubow, supra note 23, at 4. 279. According to estimation by NPS, 67% of Bamyan and 86% of Daykundi are Hasaras. See CAC, supra note 202, at 67, 69. 280. The data about the votes of candidates are collected from: IEC 2004, supra note 47; IEC 2009, supra note 48. The data about the estimated population of Hazaras is collected from: NPS, supra note 210. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 399

Similar to Pashtuns and Tajiks, there appears to be a direct association between the size of Hazara populations across provinces and the votes of Hazara candidates. As indicated in figure VI, both Mohaqiq (2004) and Dr. Bashardost (2009) won their lowest provincial votes, between 0% and 1%, in non-Hazara provinces such as Kunar, Khost, Nangarhar, Nooristan, Zabul, Paktya, Panjshir, Badghis, and Badakhshan. On the other hand, Mohaqiq and Dr. Bashardost had their greatest number of provincial votes in Bamyan and Daikundy. In Daykundi, which was demographically 88% Hazara, Mohaqiq scored 84% in 2004, and Dr. Bashardost secured 61% in 2009;281 similarly, in Bamyan, which is estimated at 67% Hazara, Mohaqiq dominated the votes, earning 76%, and Dr. Bashardost received just over 39%. See Table I-A in appendix. As discussed earlier, the relatively low number of votes Dr. Bashardost received in Daykundi and Bamyan was likely due to Karzai’s coalitions with Mohaqiq, Khalili, and Mohseni. From among the provinces, Ghazni and Samangan deserve special attention because their numbers suggest some cross-ethnic voting in favor of Dr. Bashardost and/or Mohaqiq. In Ghazni, Mohaqiq won 39% of the votes, which is lower than the estimated 45% of the Hazara population.282 In the same province, however, Dr. Bashardost scored a disproportionate vote of 61%.283 There are three likely reasons for Dr. Bashardost’s considerably higher percentage of votes than the estimated Hazara population in Ghazni and Farah. First, his higher number of votes was probably due to the cross-ethnic appeal of his anti-establishment, anti-corruption, and anti-poverty campaigns. Second, Ghazni is his home province.284 Nevertheless, the data suggests that the voting in Ghazni ran along ethnic lines, with Dr. Bashardost prevailing in the predominantly Hazara regions in the North and West of Ghazni, and Karzai dominating the predominantly Pashtun districts in the South and East.285 The third and most likely reason was a high lack of security in the Pashtun populated regions of Ghazni, which denied Karzai proportional Pashtun votes. Mohaqiq and Dr. Bashardost both won a considerably higher number of votes than one would have projected, given the estimated Hazara population in Samangan. In 2004 in Samangan, Mohaqiq received 20% of vote, and in 2009, Dr. Bashardost received around 9%.286 According to estimations provided by the

281. IEC 2004, supra note 47; IEC 2009, supra note 48. 282. NPS, supra note 210; IEC 2014, supra note 49. 283. IEC 2009, supra note 48. 284. Profile of Ramazan Bashardost, WAR NEWS RADIO (Jul. 28, 2015), http://warnewsradio.org/read-more/profile-of-ramazan-bashardost/. 285. Ghazni, AFGHANISTAN ELECTION DATA, http://afghanistanelectiondata.org/ election/2009/province/11#layers=district_map_openlayers_1&baseLayers=afghanistan- grey (last visited Nov. 29, 2015). 286. IEC 2009, supra note 48; IEC 2004, supra note 47. MOBASHER 4/18/2016 3:11 PM

400 GONZAGA LAW REVIEW Vol. 51:2

CAC handbook, Samangan is primarily Tajik and Uzbek.287 Assuming Hazaras and Pashtuns share the remaining portions of the ethnicities equally, there is a huge difference between Hazara population and their candidates’ votes. This article suggests that the difference is likely due to an error in the estimation of ethnic populations of Samangan. The Dari Suf Bala District Development Assembly estimated that 99% of the district of Dara Suf Bala are Hazara.288 Using the most recent census by CSO, nearly 16% of the total population of Samangan live in this district.289 These numbers suggest that the Hazara population is much larger than estimations provided by NPS.290 The data suggests that Dr. Bashardost received just under half of the votes that Mohaqiq received because of Mohaqiq and Khalilis’ endorsement of Karzai in 2009. The data indicates that Mohaqiq dominated the votes in Dara Suf Bala in 2004, while Karzai and Dr. Bashardost shared the votes in the same district in 2009.291

d) Abdul Rashid Dostum: The Uzbek Powerbroker

The Uzbeks did not have a candidate in the 2009 election. In 2004, in addition to Dostum, Abdul Satar Sirat, another Uzbek elite, also ran for presidency; however, he won less than 1% of the votes.292 Examining Dostum’s votes in the presidential election of 2004 opens a new window to our understanding of ethnic-electoral dynamics, since race likely played an intervening role in this election. Dostum garnered votes not only from Uzbeks, but also from Turkmens, because Uzbek and Turkmens are racially Turkic, even though they are considered ethnically distinct groups. 293

287. CAC, supra note 202, at 115. 288. ISLAMIC REPUBLIC OF AFGHANISTAN, MINISTRY OF RURAL REHABILITATION AND DEVELOPMENT, NATIONAL AREA BASED DEVELOPMENT PROGRAMME, SUMMARY OF DISTRICT DEVELOPMENT PLAN: DARI SUF BALA DISTRICT, SAMANGAN PROVINCE 3 (Oct. 2007), http://r.search.yahoo.com/_ylt=AwrTceHl9sBWfFsAPDInnIlQ;_ylu=X3oDMTEyaWVndnF zBGNvbG8DZ3ExBHBvcwMxBHZ0aWQDQjExMTVfMQRzZWMDc3I-/RV=2/RE=1455 515494/RO=10/RU=http%3a%2f%2fwww.mrrd-nabdp.org%2fattachments%2farticle%2f13 3%2fSamangan_Dari%2520Suf%2520Bala_DDP%2520Summary-tanslated_finalized.pdf/ RK=0/RS=tFAOME8FIc608PYA5jkjXYdnuwY-. 289. See CSO, supra note 93. 290. NPS, supra note 210. 291. Samangan, AFGHANISTAN ELECTION DATA, http://afghanistanelectiondata.org/ election/2009/province/18#layers=district_map_openlayers_1&baseLayers=afghanistan- grey (last visited Feb. 22, 2016). 292. IEC 2004, supra note 47. 293. HOMAYOUN, supra note 87, at 14. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 401

Turkmens are one of the minority groups in Afghanistan.294 Meanwhile, with Uzbeks, they make an estimated 74% of population in Faryab and 75% in Jauzjan. See Appendix, Table I-A. Dostum, during his campaign in 2004, successfully played his Turkic race card in order to win Turkmens’ votes. Like Uzbeks, Turkmens call him Baba (Father).295 Predictably, Dostum championed Faryab by 72.9% and Jauzjan by 78%,296 while Uzbeks represent only 50% of the population in Faryab and 40% of the population in Jauzjan. See figure VII.

Uzbek Population & Dostum's Votes In 2004

0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Ghor Zabul Logar Kabul Farah Khost Kunar Paktya Faryab Takhar Ghazni Nimroz Paktika Jauzjan Kunduz Badghis Baghlan Panjshir Bamyan Hilmand Noristan Urezgan Laghman Daykundi Kandahar Samangan Nangarhar Badakhshan Uzbek Population Dostum In 2004

FIGURE VII. This graph compares the votes of Dostum with the population of Uzbeks in 26 provinces.297

In 2009, however, Turkmens did not appear to fully support the alliance of Dostum and Karzai. This change of voting behavior probably stemmed from two incidents. The first was in 2008, when Dostum assaulted Akbar Bai, a Turkmen

294. According to some estimations, Turkmens comprise 3% of the population in Afghanistan. See CAC, supra note 202, at 56. 295. Williams, supra note 61. 296. IEC 2004, supra note 47. 297. The data about the votes of Dostum is collected from: IEC 2004, supra note 47. The data about the estimated population of Uzbeks is collected from: NPS, supra note 210. MOBASHER 4/18/2016 3:11 PM

402 GONZAGA LAW REVIEW Vol. 51:2 elite.298 The other incident was the disqualification of Akbar Bai’s candidacy upon the request of Dostum and decision of Karzai.299 The noteworthy fact about the 2004 election was Uzbeks’ desertion of Abdul Satar Sirat, who won less than 1% of the votes.300 The defection from Sirat was notable because he was the initial choice of the Bonn Conference to lead the transitional government after the collapse of the Taliban in 2001.301 Hence, even though Sirat was not a strong military leader like Dostum, he was one of the well- known Uzbek intellectuals—at least since the Bonn Conference. In fact, Sirat most likely ran for the presidency expecting that his popularity would bring him votes.

2. The Presidential Election of 2014

Previous sections showed that voters from different ethnic groups appeared to vote for a viable candidate of their own, even if that candidate was not an elite, and even if their prominent elites endorsed a different candidate. This strategic voting behavior resurfaced in the 2014 election when Dr. Ghani, once an insignificant candidate in 2009, emerged as a prominent presidential hopeful among Pashtuns. This election probably tells us more about electoral dynamics and strategic coordination. In addition, results from the presidential election of 2014 provided some strong evidence about how in the absence of an ethnic candidate, political bargain determine cross-ethnic votes in Afghanistan.

a) Dr. Abdullah: The Apparent Frontrunner

In the first round of the 2014 election, Dr. Abdullah’s polls reflect the population of either Tajik or Hazara or both. See figure VII. In fact, after Panjshir, which is his home province, Dr. Abdullah obtained his highest proportion of provincial votes in Daykundi (75%) and Bamyan (68%), which are predominantly Hazara provinces.302 In the same provinces, respectively, Dr. Abdullah received 6.6% and 18.1% in 2009, which to some extent reflected the

298. Williams, supra note 170, at 4. 299. Sayed Yaqub Ibrahimi, Candidates Controversy as Presidential Race Kicks Off, INSTITUTE FOR WAR AND PEACE REPORTING (July 31, 2009), https://iwpr.net/global- voices/candidates-controversy-presidential-race-kicks; see also Bijlert, supra note 29, at 10. 300. IEC 2004, supra note 47. 301. Who Is Who In Afghanistan: Abdul Satar Sirat Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=1728&task=view& total=3158&start=2731&Itemid=2 (last visited Feb. 22, 2016). 302. IEC 2014, supra note 49. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 403

Tajik populations.303 The extraordinary change of voting patterns in favor of Dr. Abdullah in these provinces likely took place for two reasons: First, Dr. Abdullah sealed the endorsement of Mohaqiq by introducing him as his second vice presidential candidate.304 Second, and more importantly, there were no Hazara candidates running for the presidency in the 2014 election. It is more likely than not that most Hazaras would not have voted for Dr. Abdullah if there were any viable Hazara candidate, as they had not voted for Karzai in 2009, when Bashardost ran for presidency.305

Tajik & Hazara Populations & Dr. Abdullah's Votes

2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 Ghor Zabul Kabul Farah Khost Kunar Paktya Faryab Ghazni Nimroz Paktika Jauzjan Kunduz Badghis Baghlan Panjshir Bamyan Hilmand Noristan Urezgan Laghman Daykundi Kandahar Samangan Nangarhar Badakhshan

Abdullah's Votes Tajik Population Hazara Population

FIGURE VIII. This graph compares the votes of Dr. Abdullah with the populations of Tajiks and Hazaras in 26 provinces.306

303. IEC 2009, supra note 48. 304. Ghanizada, Mohaqiq and Hezb-e-Islami deputy named Dr. Abdullah’s vice- president, KHAAMA PRESS, Oct. 1, 2013, http://www.khaama.com/mohaqiq-and-hezb-e- islami-deputy-named-dr-abdullahs-vice-president-1918. 305. See Table I-A in the Appendix. 306. The data about the votes of candidates are collected from: IEC 2014, supra note 49. The data about the estimated population of ethnic groups is collected from: NPS, supra note 210. MOBASHER 4/18/2016 3:11 PM

404 GONZAGA LAW REVIEW Vol. 51:2

Nonetheless, Dr. Abdullah had to share some Hazara votes with his main rivals. For instance, both the Daykundi and Bamyan provinces, Dr. Zalmai Rassul, the third place candidate, gained nearly 12% of the vote, and Dr. Ghani received just over 11% of the vote.307 Ironically, this sharing was not a coincidence, as both Dr. Rassul and Dr. Ghani selected Hazaras from these provinces as their second vice presidential candidate: Zalmai Rassul’s second vice president, Dr. Habiba Sorabi, was the former governor of Bamyan (2005 to 2013);308 similarly, Dr. Ghani’s second vice president, Mohammad Sarwar Danish, the former Minister of Justice, is from Daikundi.309 Predictably, Dr. Abdullah won fewer votes in Pashtun dominated provinces, even though he performed much better in those provinces than he did in 2009.310 In particular, he garnered close to 20% of the votes in Nangarhar, Hilmand, and Zabul.311 He also had the lead in Ghazni (54%) and Kunduz (47%), where Pashtuns are the majority. See Table I-A in the Appendix.312 Choosing Mohammad Khan as his running mate, bargaining with Pashtun Jihadi elites, and emphasizing his half-Pashtun identity in Southern provinces were keen strategies that likely assured him voter blocs from Pashtun communities. More importantly, Dr. Abdullah was pursuing a strategy of coordination through establishing a large coalition of opposition to government. Between 2006 and 2014, Dr. Abdullah and his circles attempted several times to create an opposition coalition.313 Even though the coalitions were doomed to break apart, some members stayed with Dr. Abdullah and endorsed his candidacy in 2014.314 Some influential individuals who upheld their endorsements of Dr. Abdullah included some Pashtun elites, such as Nurulhaq Ulumi, an elite from

307. IEC 2014, supra note 49. 308. Who Is Who In Afghanistan: Dr. Habiba Sarabi Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=1733&task=view& total=3162&start=2742&Itemid=2 (last visited Feb. 22, 2016). 309. Who Is Who In Afghanistan: Mohammad Sarwar Danish Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=402& task=view&total=3162&start=672&Itemid=2 (last visited Feb. 22, 2016). 310. On average Abdullah won 11% in Pashtun majority provinces (provinces with over 50% Pashtun population—this excludes Kunduz, Laghman and Ghazni) in 2009. See IEC 2009, supra note 48. In 2014, he won around 17.4% votes on average in Pashtun majority provinces. See IEC 2014, supra note 49. 311. IEC 2014, supra note 49. 312. Id. 313. For the full account of coalition development by Dr. Abdullah and his cohort in Afghanistan read TCHALAKOV, supra note 240. 314. Thomas Ruttig & Gran Hewad, Narrower Than Expected: Political opposition presents “Electoral Union of Afghanistan, AFGHANISTAN ANALYSTS NETWORK, Aug. 29, 2013, at 1, https://www.afghanistan-analysts.org/narrower-than-expected-political-opposition -presents-electoral-union-of-afghanistan/. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 405

Kandahar; Kalimullah Naqibi, Chief of a Pashtun tribe;315 and Shah Wali Karizai, brother of Karzai. Some Hazara and Uzbek elites also upheld their endorsements of Dr. Abdullah; these included Mohammad Akbari, a Hazara elite; Abdul Malik Pahlawan, the leader of Liberation Party and main Uzbek rival of Dostum;316 and Piram Qul, an influential Uzbek elite from Takhar.317 Moreover, during his campaign, and particularly in Pashtun dominated areas, Dr. Abdullah mostly appeared in traditional Afghan clothes, defended Mujahidin against criticism, and gave speeches in Kandahari Style Pashtu language,318 thereby invoking his half-Pashtun identity.

b) Dr. Ashraf Ghani: The Emerging Elite

The first time that Dr. Ghani ran for the presidency was in the presidential election of 2009.319 At that time, he was one of many insignificant candidates who received slightly over 2.9% votes nationwide.320 He was considered an outsider and a puppet who knew little about the realities of Afghanistan.321 He was also popularly labeled as Zan-e-Bush (Wife of Bush).322 Even in his home-

315. Who Is Who In Afghanistan: Kalimullah Naqibi Biography, AFGHAN BIOS, http://www.afghan-bios.info/index.php?option=com_afghanbios&id=2900&task=view& total=3096&start=1902&itemid=2 (last visited Feb. 22, 2016). 316. General Malik, a prominent Uzbek politician, has endorsed Dr. Abdullah for the 2014 presidential election!, NATIONAL COALITION OF AFGHANISTAN, http://www. nca.af/general-malik-dr-abdullahf2014010852.html (last visited June 26, 2015). 317. Ruttig et al, supra note 313, at 1-2. Zalmai Khalilzad, a speculated presidential hopeful, accused the National Front Coalition of Dr. Abdullah for ethnocentrism by having few Pashtuns in their ranking. The coalition ranking was allegedly filled with Tajik prominent elites such as Qanooni, the former candidate in the 2004 election; Atta Mohammad Noor, governor of Balkh; Salhuddin Rabbani, the head of High Peace Council. See id. at 2. 318. Yaroslav Trofimov, Margherita Stancati and Nathan Hodge, Afghan Rivals Abdullah Abdullah and Ashraf Ghani Race to Presidential Runoff, THE WALL STREET JOURNAL, June 14, 2014, http://www.wsj.com/articles/afghan-rivals-abdullah-abdullah-and- ashraf-ghani-race-to-presidential-runoff-1402105898. 319. See Carlotta Gall, A Technocrat Shakes Up the Afghan Campaign, N.Y. TIMES, Aug. 13, 2009, http://www.nytimes.com/2009/08/14/world/asia/14ghani.html; IEC 2009, supra note 48. 320. IEC 2009, supra note 48. 321. See Coburn & Larson, supra note 21, at 5. 322. Rob Crilly, Ashraf Ghani: From the World Bank to Afghanistan’s next president?, THE TELEGRAPH, Jul. 7, 2014, http://www.telegraph.co.uk/news/worldnews/asia/afghanistan/ 10952284/Ashraf-Ghani-From-the-World-Bank-to-Afghanistans-next-president.html; see al- so Jeffery Stern, The Making of an Afghan Politician, SLATE, Aug. 19, 2009, http://www.slate.com/articles/news_and_politics/dispatches/2009/08/the_making_of_an_ afghan_politician.html. MOBASHER 4/18/2016 3:11 PM

406 GONZAGA LAW REVIEW Vol. 51:2 province, Logar,323 he received less than 20% of the votes.324 Across the 26 provinces considered in this article, Ghani won 1% or less in 11 provinces and 5% or less in 19 provinces.325 In 2014, however, he became known as a patriot, and he was labeled as the Gandhi of Afghanistan.326 His votes rose to more than 10% in all provinces, except for in Panjshir. In more than half of these provinces, he received over 30% of the votes.327 He scored 50% to 69% of the votes in 7 provinces that have either Pashtun or Uzbek Majorities. See figure IX. Dr. Ghani’s transition in the eyes of the voters from Zan-e-Bush in 2009 to a charismatic and patriotic candidate in 2014 is the most telling case of electoral strategy and voting behavior in Afghanistan.328 In the absence of an already established candidate, Dr. Ghani sought the opportunity to first portray himself as the only viable candidate, hoping that if everyone in his primary constituency entertained the same expectations about his viability, he eventually would become the actual winner.329 In order to create such an image amongst Pashtun

323. Biography: Ashraf Ghani, AFGHANISTAN ONLINE, http://www.afghan-web.com /bios/ghani.html (last visited Feb. 22, 2016). 324. IEC 2009, supra note 48. 325. Id. 326. In the social media he was popularly called the Gandhi of Afghanistan. For instance, several Facebook pages were created with the title of Gandhi of Afghanistan. See Gandhi of Afghanistan, FACEBOOK, https://www.facebook.com/pages/DrAshraf-Ghani- Afghanistans-Gandhi/486382698141057 (last visited Feb. 22, 2016). 327. IEC 2014, supra note 49. 328. Some argued that Dr. Ghani’s emerging from an unpopular candidate in 2009 to a runner up in 2014 was due to his leadership qualities and his vast experience in developmental programs. However, if it was his leadership qualities that earned him votes in 2014 election, it seems likely that he would have garnered the same support in the 2009 election. He was well known for carrying out several reforms successfully while leading Ministry of Finance (2002-2004) and other offices prior to 2009 elections. See Decision Point 2009, supra note 20, at 13. But in 2009, Dr. Ghani was considered an outsider and a puppet that knew little about the realities of Afghanistan. He was even labeled Zana-e-Bush (George W. Bush’s wife) among electorates. See Jeffery Stern, supra note 321. Others might argue that his policies probably made him more appealing than other candidates. Nonetheless, even though his policies seemed more calculated, he was not distinct from other Pashtun candidates on the major policies. On issues, such as peace negotiations with the Taliban, security, relations with Pakistan and other neighboring countries, and signing the Bilateral Security Agreement with the United States, Dr. Ghani shared views with the main Pashtun contenders such as Zalmai Rassul and Gul Agha Shirzai, the third and fourth placed candidates respectively. Moreover, his policies barely changed from 2009 to 2014 elections. Ali M Latifi, Afghan Presidential Hopefuls Take To TV, AL JAZEERA, Feb. 5, 2014, http:// blogs.aljazeera.com/blog/asia/afghan-presidential-hopefuls-take-tv; see also RAFI SEDIQI, Gul Agha Sherzai Asks for BSA Signature, TOLONEWS, Feb. 9, 2014, http:// www.afghanistannewscenter.com/news/2014/february/feb102014.html#a12. 329. This electoral calculation was extensively explained by Cox suggesting that voters expectation is crucial for how they vote. See COX, supra note 43, at 127. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 407 voters, Dr. Ghani pursued a series of strategies that closely resembled Karzai’s approach. First, Dr. Ghani added to his ticket Dostum and Sarwar Danish, a loyal supporter of Khalili.330 Particularly, his coalition with Dostum both guaranteed Uzbek votes and sent a strong message to Pashtun voters about his increasing electoral edge over his rivals. Since the voting decisions of others typically influence expectations of voters,331 Uzbek support of Dr. Ghani likely had a psychological effect on the Pashtun expectation about Dr. Ghani’s viability. Yet, Dr. Ghani had to deal with the criticisms and setbacks of his coalition with Dostum. Soon after joining Dr. Ghani’s campaign, Dostum made an unprecedented apology for his role in the past wars that caused “pain and suffering of the people.”332 This apology prior to the 2014 election was likely to address two issues: First, the apology was a response to criticisms of the Ghani-Dostum coalition that portrayed Dr. Ghani as a hypocrite for his previous accusations against Dostum.333 Second, the apology was likely an attempt to please those Pashtun voters—especially those of Kandahar—who had had experience with Dostum’s militia during the past wars.334 Moreover, among Pashtun candidates, Dr. Ghani was quicker to persuade his primary constituencies that he was one of them. He replaced his Western- style suits with Afghan traditional dress, including a Southeastern style “scarf” on his shoulder, and often a Turban on his head,335 which differed slightly from

330. In addition to Dostum and Danish, Ghani also won the endorsement of Sayed Mansour Naderi, the leader of Esmailis, a Hazara segment. See Ghanizada, Syed Mansoor Naderi supports Ashraf Ghani in presidential elections, KHAAMA PRESS, Feb. 13, 2014, http://www.khaama.com/syed-mansoor-naderi-supports-ashraf-ghani-in-presidential- elections-2782. 331. COX, supra note 43, at 71. 332. Ruttig, supra note 64, at 2; see also Wakil Kohsar, From Ex-Warlords to Scholars: Who Wants to Rule Afghanistan, NBC NEWS, Apr. 5, 2014, http://www. nbcnews.com/storyline/afghanistan-election/ex-warlords-scholars-who-wants-rule-afghanis tan-n24756. 333. In the presidential election of 2009, Dr. Ghani called Dostum a “known killer” and criticized Karzai for making alliances with Dostum. Choosing Dostum as his own vice- president candidate led to his rivals and many others calling him a hypocrite. See Erin Friar, supra note 229; also see Masoud, Hypocrite Politician Ashraf Ghani Ahmadzai and Known Killer Abdul Rashid Dostum, KABUL PRESS, Mar. 7, 2014, http://kabulpress.org/my/ spip.php?article177004. 334. See Erin Friar, supra note 229; also see Masoud, supra note 332. 335. See Harun Najafizada, Afghan notebook: Wowing the crowds, BBC NEWS, Mar. 18, 2009, http://www.bbc.com/news/world-asia-26576308; see also Renee Montagne, To Broaden Appeal, Afghan Candidates Make Surprising Choices, NPR, Apr. 3, 2014, MOBASHER 4/18/2016 3:11 PM

408 GONZAGA LAW REVIEW Vol. 51:2 one province to another.336 This was exactly what Karzai had been doing during his campaign in the 2004 and 2009 elections. Dr. Ghani even went further and criticized the U.S. and Afghan governments for detaining just Pashtuns in prisons.337 At one point in Kandahar, the base of the Taliban during their regime, Dr. Ghani declared that he would release all insurgent suspects.338 In Kandahar, several local Pashtun candidates did better than Dr. Ghani. Dr. Zalmai Rassul dominated the ballots, with 53.96% of the votes, and Gul Agha Sherzai garnered 16.02% of the votes.339 Dr. Ghani only managed to win 13.90% of the votes.340 Ghani also underperformed in Wardak, where Sayyaf led the Pashtun candidates with 24.64%, followed by Zalmai Rassul with 16.74% votes.341 There, Dr. Ghani’s votes amounted to 15.11%.342 In Nooristan, Sayyaf garnered twice as much as Ghani receiving 28.93% to Ghani’s 14.58%.343 The two provinces in which Ghani’s votes were significantly higher than the population of Pashtuns were Faryab (65%), and Jauzjan (69%).344 Not surprisingly, Dostum’s presence in Dr. Ghani’s campaign guaranteed him Uzbek votes in these provinces. Overall, as figure VIII illustrates, Dr. Ghani’s votes are comparable with the population of Pashtuns and Uzbeks in most provinces.

http://www.npr.org/sections/parallels/2014/04/03/298579270/to-broaden-appeal-afghan- candidates-make-surprising-choices. For his style in 2009 election, see Stern, supra note 321. 336. See Najafizada, supra note 334; see also Montagne, supra note 334; Stern, supra note 321. 337. Matthew Rosenberg, Technocrat to Afghan Populist, Ashraf Ghani Is Transformed, N.Y. TIMES, June 11, 2014, http://www.nytimes.com/2014/06/12/world/asia/ technocrat-to-populist-an-afghan-transformed.html. 338. Id. 339. IEC 2014, supra note 49. 340. Id. 341. Id. 342. Dr. Ghani also had to follow Sayyaf’s lead in some other provinces such as Hirat (13.78% to 11.08%), Ghor (16.02% to 13.09%), Parwan (6.14% to 5.61%), Badghis (12.98% to 9.80%), and Kapisa (5.74% to 4.12%). See IEC 2014, supra note 49. In addition to Sayyaf, Dr. Zalmai Rassul also performed better than Ghani in Parwan (10.56% to 5.61%), Hirat (11.41% to 11.08%), and Kapisa (5.62% to 4.12%). Id. Rassul almost split the votes with Ghani in Urezgan (23.86% to 26.95%), Nimroz (30.46% to 33.59%), and Hilmand (27.46% to 32.94%). Id. 343. IEC 2014, supra note 49. 344. Id. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 409

Pashtuns & Uzbeks Population, & Dr. Ghani's Votes

2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 Ghor Zabul Kabul Farah Khost Kunar Paktya Faryab Ghazni Nimroz Paktika Jauzjan Kunduz Badghis Baghlan Panjshir Bamyan Hilmand Noristan Urezgan Laghman Daykundi Kandahar Samangan Nangarhar Badakhshan

Ghani's Votes Pashtun Population Uzbek Population

FIGURE IX. This graph compares the votes of Dr. Ghani with the populations of Pashtuns and Uzbeks in 26 provinces.345

III. ALTERNATIVE EXPLANATIONS TO ETHNIC-ELECTORAL DYNAMICS

Both ethnicity and elections as social phenomena interact with other social dimensions, such as education, urbanization, security, localism, electoral fraud, and clientalism.346 In this section, I address the influence of some of these intervening variables, including insecurity, turnouts, localism, and electoral irregularities, looking to their impact on ethnic-electoral dynamics.

345. The data about the votes of candidates are collected from: IEC 2014, supra note 49. The data about the estimated population of Pashtuns is collected from: NPS, supra note 210. 346. While education, gender, income and area of residence have a great impact on elections in many other countries, they are less influential on elections in Afghanistan because of their insignificant influence on social life in general. These variables are, therefore, excluded from the text of this article. MOBASHER 4/18/2016 3:11 PM

410 GONZAGA LAW REVIEW Vol. 51:2

A. Insecurity and Turnout Rate

Security has a direct impact on turnout rates. The lower the security in a region, the fewer voters show up in polling centers.347 In fact, the Taliban threat was one of the main factors that reduced Karzai’s votes in the Southern and Eastern provinces in 2009. A survey conducted by ASIA in 2013 reported that the confidence in the Afghan National Police in the South, Southeast, and Southwest was much lower (between 25% and 65%) than in other regions of the country.348 Similarly, voters in the Pashtun regions experienced more fear (67%) than participants in other regions did.349

B. Electoral Irregularities350

In both the 2009 and 2014 runoff elections, large-scale electoral irregularities led to electoral deadlock.351 In the presidential election of 2009, around 20% of the votes were invalidated after a reevaluation.352 In the 2014

347. See Security fears shutter Afghan polls, AL JAZEERA, Aug. 18, 2010, http://www.aljazeera.com/news/asia /2010/08/201081871831375811. 348. ASIA FOUNDATION, AFGHANISTAN 2013: A SURVEY OF THE AFGHAN PEOPLE 5-7 (2013), http://asiafoundation.org/resources/pdfs/English2013ExecutiveSummary.pdf. 349. Compare this to the confidence of the people of other regions (90-98%). See ASIA FOUNDATION, supra note 348, at 7; Dexter Filkins, Threats by Taliban May Sway Vote in Afghanistan, N.Y. TIMES, Aug. 16, 2009, http://www.nytimes.com/2009/08/17/world/asia/ 17taliban.html; Recap: Afghanistan Votes 2014, THE WALL STREET JOURNAL, June 13, 2014, http://blogs.wsj.com/dispatch/2014/06/13/as-it-happens-afghanistan-votes-2014-3/. 350. Electoral Irregularities occurs in a variety of forms such as ballot stuffing, vote count tampering, bribery of electoral officials, vote buying, intimidations of voters, destruction of ballot boxes, closing polling centers, and ‘remedial fraud.’ See ASIA FOUNDATION, supra note 348, at 1, 87; see also Carina Perelli & Scott Smith, Anticipating and Responding to Fraud in the 2014 Afghan Elections, AFGHANISTAN: OPPORTUNITY IN CRISIS SERIES NO. 2, Feb. 2014, at 2, http://www.chathamhouse.org/sites/files/chathamhouse/home/chatham/public_ html/sites/default/files/20140200AfghanistanElectionsFraudPerelliSmith.pdf. 351. Electoral deadlock is the natural result of the violation of the electoral golden rule. The golden rule suggests “that the sum total of irregularities should not exceed the difference of votes between two candidates necessary to determine who wins the seat.” Perelli & Smith, supra note 11, at 9. 352. The final Report of EEC excludes the following number of polling stations in each provinces: Kandahar 84, Nangarhar 84, Paktika 77, Nooristan 67, Faryab 66, Paktia 65, Herat 42, Laghman 36, Kunar 33, Ghazni 33, Ghor 26, Baghlan 19, Kabul 12, Logar 12, Parwan 11, Kapisa 10, Badghis 10, Takhar 6, Zabul 6, Panjshir 5, Helmand 5, Kunduz 3, Nimruz 3, Badakhshan 2, Daikondi 2, Jowzjan 1, Samangan 1, Sar- e Pol 1, Wardak 0, Bamyan 0, Balkh 0, Farah 0, Khost 0, Uruzgan 0. See Bijlert, supra note 156, at 5. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 411 runoff, in order to unblock the electoral brinkmanship, a full recount of the entire ballot boxes was conducted.353 In Afghanistan, electoral fraud usually occurs because of irregularities in voter registration, anti-system electoral threats, and the presence of unlawful armed groups.354 Registration irregularities have paved the ground for opportunist voters to register for voting multiple times.355 While the total population of qualified voters in Afghanistan is estimated around 13.5 million,356 the distributed voting cards reached over 17.5 million by the 2010 election.357 Electoral violence has been another tool used by candidates and their supporting elites in order to intimidate potential voters, distort security assessments, and even eliminate key opponent figures.358 In the presidential election of 2004, a tribal council in Khost declared that they would burn the house of any voter not voting for Karzai.359

C. Localism

Localism may be another influence on election outcomes. Here, localism refers to the feelings of attachment to provincial values and pride among residents of provinces .360 For example, in the 2004 election, Karzai and Yunis

353. Sayyad Salahuddin, Karzai Pardons Five Afghan Heroin Traffickers, REUTERS, July 12, 2014, http://www.reuters.com/article/us-usa-afghanistan-kerry-recount-idUSKBN0F H0PJ20140712; Vishal Chandra, Total Recount in Afghanistan: What Next?, INSTITUTE FOR DEFENSE STUDIES AND ANALYSES 1 (July 26, 2014), http://www.idsa.in/issuebrief/Total RecountinAfghanistan_vchandra_260714. 354. See Perelli & Smith, supra note 11, at 2. 355. In most conservative areas where men register for their females use of multiple cards has been way obvious. For example, voting registry for the 2009 election has recorded higher proportions of female registration in Nuristan (71%), Khost (68%), Logar (66%) and Paktia (62%). These numbers are beyond any expectations in these provinces. See Bijlert, supra note 29, at 21. 356. The number for adult age—by which the population of qualified voters is determined—was taken from UNICEF website. See Afghanistan Statistics, UNITED NATIONS INTERNATIONAL CHILDREN’S EMERGENCY FUND [UNICEF], http://www.unicef.org/info bycountry/afghanistan_statistics.html (last visited Feb. 22, 2016). 357. Bijlert, supra note 156, at 21, 22. 358. Perelli & Smith, supra note 11, at 10. 359. Bijlert, supra note 29, at 16; see also SWISSPEACE, supra note 123, at 4. 360. Localism particularly becomes crucial in the absence of sufficient information about who is the leading candidate. This became obvious in the presidential election of 2014. Some Pashtun elders from Southeastern Afghanistan, mainly from Kandahar, held a council to support the candidacy of either Qayyum or Dr. Zalmai, both from Kandahar. See Ruttig, supra note 77. Kandahar was the only province that voted largely for Dr. Zalmai Rasul in the first round. See IEC 2014, supra note 49. MOBASHER 4/18/2016 3:11 PM

412 GONZAGA LAW REVIEW Vol. 51:2

Qanooni won over 90% of the votes in their home provinces.361 Dr. Bashardost did exceptionally well in his home province of Ghazni. See Table I-A in appendix. Dr. Ghani also polled his highest provincial votes in both elections in his home province, Logar.362 Nonetheless, the influence of localism seems overshadowed by ethnicity when there is a prominent candidate. For instance, Karzai defeated all Pashtun candidates in their home provinces, defeating even Dr. Ghani in both the 2004 and 2009 elections.363

CONCLUSION

Democratization has carried with it the promise of stabilization and solidarity for diverse ethnic groups in Afghanistan.364 Indeed, elections were expected to pave the way for conflicting groups to renegotiate power in a non- violent manner.365 In the context of Afghanistan, however, these processes are consistently complicated by ethnic tensions. As recently as the 2014 presidential elections, ethnic voting has continued to affect election outcomes. When voters are swayed by ethnicity, the extent to which one elected individual can adequately represent the whole nation is debatable.366 The elections in Afghanistan have shown a complex set of competing democratic, non-democratic and anti-democratic values.367 As such, Afghanistan is more of a hybrid state than a true democratic state. Not surprisingly, Perelli and Smith referred to the political system as Limited pluralism. 368 In their work, they explain that despite free elections in Afghanistan, the winners are consistently elites.369 This article has shown that ethnic tensions rise to their peak during Afghan elections. While candidates and elites tend to mobilize ethnic groups during the elections, ethnic groups tend to vote collectively, and, most significantly, on the basis of ethnicity. This article aims to provide important background material and an understanding of the persistence of ethnic voting and mobilization in

361. IEC 2004, supra note 47. 362. IEC 2009, supra note 48. 363. IEC 2004, supra note 47; IEC 2009, supra note 48. 364. Dubow, supra note 23, at 5. 365. See Perelli & Smith, supra note 11, at 3. 366. See NOAH COBURN & ANNA LARSON, DERAILING DEMOCRACY IN AFGHANISTAN: ELECTIONS IN AN UNSTABLE POLITICAL LANDSCAPE 23 (2014). 367. SCOTT RADNITZ, WEAPONS OF THE WEALTHY: PREDATORY REGIMES AND ELITE- LED PROTESTS IN CENTRAL ASIA 194 (2010). 368. Perelli & Smith, supra note 11, at 6. 369. Id. MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 413

Afghan elections; this information should be valuable to the current scholarship about elections and ethnic heterogeneity. My hope is that it contributes not only to the literature, but also to the ongoing legal and political discourse surrounding amending the Constitution and reforming the election laws

MOBASHER 4/18/2016 3:11 PM

414 GONZAGA LAW REVIEW Vol. 51:2

APPENDIX

TABLE I-A

Table on the following page shows ethnic composition of 26 provinces and candidates’ shares of votes in three presidential elections in those provinces.

Sources of Information About Ethnic Populations

1. Naval Post Graduate School (NPS):http://www.nps.edu/About/index.html 2. For information regarding the election results for each province, see IEC 2004, supra note 47; IEC 2009, supra note 48; IEC 2014, supra note 49.

ADDITIONALLY – THE PERCENTAGE FOR ABDULLAH AT KHOST IN 2014 IS WRONG – IT SHOULD BE 4%, NOT 36%.

MOBASHER 4/18/2016 3:11 PM

2015/16 ELECTION OUTCOMES IN AFGHANISTAN 415