Photo: U.S. Army Garrison

Is Afghanistan becoming a failed state again? Despite billions of dollars in aid and more than a decade of support in military and administrative capacity building, recent news is not encouraging: with more than 10,000 civilian casualties (deaths and injuries) due to insurgency-related fighting, 2014 was the deadliest year since the UN started keeping records. In a possible precursor to a renewed era of local warlords, local militia increasingly emerge in areas where the state is absent. Afghanistan seems to stand once more at the crossroads. What are we to make of this latest attempt to pacify and rebuild a failed state? Are there some general lessons to be learned? What type of institutions may facilitate state survival, or should international organizations and “the West” refrain in the future from trying to promote democracy in failed states? With state-threatening turmoil in the Middle East and several African states, these questions are likely to stay at the forefront of foreign policy for years to come. In an effort to learn more about some of these questions, two colleagues and I implemented several survey experiments in Afghanistan.

At first glance, the three rounds of successful national elections in 2004, 2009 and 2014 look promising. A record number of brave Afghans have repeatedly faced down Taliban threats to cut off purple-inked voter fingers and to bomb polling stations. A peaceful transition of power also took place in 2014. Still, criticism about Afghanistan’s electoral system (where each voter has a single vote in multi-seat districts) is widespread: it is said to frustrate the political ambitions and representation of marginalized groups. The voting system also results in very large numbers of wasted votes, i.e. cast votes that receive no representation as the supported candidate receives no seat. In 2009, for example, 63% of all votes (and over 80% in some provinces) were cast for losing candidates leading to no political representation of those votes – a worldwide record.[1] For historic reasons, Afghanistan’s new constitution also deemphasizes the role of parties leading to a highly fragmented legislature: Kabul province, for example, which has 33 MPs, fielded in 2010 over 600 independent candidates with some winning candidates receiving just 0.5% of the vote. Additionally, though long promised by Afghanistan’s 2004 constitution, representation at the local level remains an elusive concept: district governors are appointed by the distant President and village life remains dominated by non-recognized local shura councils in which traditional norms limit female participation. As a result, it may not come as a surprise that 29% of Afghans in our survey were “very dissatisfied” with the way democracy works while 59%, 61% and 47% respectively thought that religious rule, tribal rule or a strong leader without elections were either “fairly good” or “very good” modes of governance for Afghanistan (democracy received 54%).

Could a governance change in electoral institutions alter some of these dynamics? Fewer wasted votes and a greater representation of minority groups may pull conservative elements of society (those that may sympathize most with radical groups such as the Taliban) into the political process, thus invoking the ancient “trading bullets for ballots” rationale that might be democracy’s greatest strength. Alternatively, village elections that lead to a state-recognized and tax-endowed local council may expose locals to participatory democracy and empower them rather than far-off bureaucrats to solve local problems with government resources. But how many votes would Taliban-leaning candidates capture? How would locals in rural areas react to the state’s attempt to substitute traditional shuras with formal Western-type elections? Or, should the state instead recognize this centuries-old tradition with its successful track record in conflict resolution despite its limited franchise? Electoral reform may increase society’s support for democracy, a crucial input in an environment where the cost of voting may be a cut-off finger or worse.

In an effort to learn about some of the benefits that governance and electoral reform may hold for Afghanistan, two colleagues and I ran several nationally representative survey experiments that covered over 8,000 households in Afghanistan with generous funding from USAID and Democracy International.[2] In one experiment we randomly assigned subjects to different electoral voting methods for hypothetical national elections. We were interested in learning how different electoral voting systems could change individuals’ support for democracy, their beliefs that politicians would keep their promises and how such changes would affect the vote shares of minority candidates. A surprise finding was that a generic “Taliban-related candidate” increased his vote share from 1.4% under the current voting system where only one vote can be cast, to 16% in a voting system where subjects award votes by ranking all the candidates. A female candidate also increased her vote share from 13% to 18% while a recently returning expat technocratic candidate (in some sense an “outside option”) incurred most of the losses. Interestingly, a generic incumbent candidate did not lose any vote share, making electoral change politically feasible. We also found that a rank-based voting system not only reduced vote wastage and strengthened minority candidates but also increased (statistically significantly) voter satisfaction, voter turnout and the belief that politicians would keep their promises after elections. While no panacea, electoral reform may thus help to improve voters’ perceptions about democracy and in turn increase support among the population for democracy and a democratically-elected government.

In a second experiment, my colleagues and I randomly assigned subjects to hypothetical scenarios where the government formalized local councils in various ways. In one treatment, formal elections would be held to vote for a new local council. In a second treatment, the local customary council (shura) would be officially recognized by the state, while a third treatment had local development councils (a village body introduced in thousands of Afghan villages by the World Bank and international donors) become the state-recognized village council. We were interested in what Afghans thought about the benefits and costs to those alternative local institutions. Formal local elections for a local council again increased voter satisfaction as well as the belief that elected local officials would keep their promises. Respondents however also believed that formal elections would reduce the ability of elected locals to defend community interests from higher levels of the government. The largest effect occurred in the expectations on security: respondents believed that both formal elections as well as a formalized development council would reduce the likelihood of future reconciliation with the Taliban by 40-45%. This finding is in line with the ability of shuras to effectively mediate conflicts; a crucial finding in a country fighting a neo-Taliban uprising.

Could electoral reforms strengthen the democratic “buy-in” among Afghan voters and strengthen its young institutions that are under assault by radical groups? Our experiments suggest so. Voter support for democracy is likely to increase under a rank-based voting system, and local shuras may play a pivotal role in conflict mediation and reconciliation. Although electoral engineering cannot prevent the Taliban from brutalizing voters, jointly, these changes may lure back some alienated groups to the political table while Afghan democracy may be left more resilient against those that continue to attack the state. These findings may be Afghanistan-specific and call for more research in other fragile states, but one conclusion seems clear: institutions do matter.


[1] To compare, Iraq had about 5% of vote wastage in its first national elections in 2005.

[2] Jochem, Murtazashvili and Murtazashvili (2015), Electoral Institutions and Perceptions of Democracy: Evidence from a Survey Experiment in Afghanistan, working paper; and Jochem, Murtazashvili and Murtazashvili (2015), Establishing Local Government in Fragile States: Evidence from an Experiment in Afghanistan, working paper.