As bad as endless debates among divergent political factions in Afghanistan are for efficient governance, the consequences of political exclusion are arguably worse. Ethnic and regional tensions, exacerbated by contrasting liberal versus conservative visions of governance, ensure constant turmoil in the political arena. On the other hand, violence frequently ensues whenever groups are excluded from the political mix.

 

The Taliban insurgency since the 2001 Bonn Agreement began in earnest only after attempts by more moderate former Taliban leaders to reconcile with the new government were rebuffed and the Taliban played no role in forming the constitution or participating in early elections. Later, the 2014 election crisis entered critical mode when President-elect Ashraf Ghani declared that all sitting governors would be dismissed as soon as he took office. This led Governor Atta Muhammad Nur in Balkh Province to threaten to form a ‘parallel’ government, which carried the implicit threat of civil war. Most recently, the Islamic State has been able to gain a foothold in Afghanistan when disgruntled factions within the Taliban or under-funded warlord militias decide that switching allegiances would enable greater recognition or resources.

 

For many, giving in to secessionist demands is equivalent to capitulating to blackmail, and those left out of a political process in Afghanistan are in fact often behaving irresponsibly or illegally. But, for a variety of reasons, the Afghan state has yet to build up the political, military or legal strength to impose its will over major factions that are willing to fight to gain a decision-making role in the political system. The way forward is to find mechanisms that can include everyone and still manage conflict in a way that does not produce total gridlock.

 

Rather than continuing to concentrate political power at the centre, other options should be considered to deconcentrate power to provinces, but in a way that still maintains national cohesion and adheres to the principles of the constitution. Expanding political participation and decision making so that fewer groups have political grievances against the central government could create a broader landscape for political compromise. This might slow the ideal path of reforms but would reduce conflict and violence in the process. Increasing local political autonomy could also benefit the peace process. If Taliban factions join local political processes, it would weaken the movement’s overall fighting strength. But even if the Taliban choose not to participate, a deconcentration of political power that better addresses the demands of political inclusion by non-Taliban factions will remove grievances that the Taliban have exploited to gain support and control territory.

The Bonn Agreement in 2001 was a momentous political milestone for Afghanistan and its successful implementation between 2002 and 2005 was a signature achievement. The agreement established an interim administration led by Hamid Karzai, an anti-Taliban Pashtun, with an ethnically and politically diverse interim cabinet that consisted of many Northern Alliance factional leaders. It also laid out a framework for establishing a constitution, a democratic system of government and respect for international human rights norms as the foundation of the state. The 2004 Constitutional Loya Jirga, which was attended by delegates selected in a democratic process, affirmed the Bonn Agreement’s democratic governance framework. It also established a highly centralised presidential system of government, with a directly elected president having vast powers of appointment – of one-third of the members of the upper legislature, of all provincial governors, of cabinet ministers and deputy ministers, and of district officials.

 

The main missing ingredient from the Bonn process – comprising the Bonn Agreement, the Constitutional Loya Jirga and the first presidential and parliamentary elections through 2005 – was the Taliban. This was for good reasons at the time: the Taliban had harboured Osama bin Laden while he planned and conducted the 9/11 attacks and refused to turn him over to US or to international authorities after bin Laden’s role in the attacks was clear. In addition, the Taliban had ruled much of Afghanistan since 1996 with extreme contempt for women’s rights and human rights, committing massacres against rival Afghan groups and destroying cultural heritage such as the Bamiyan Buddhas. The Taliban were culpable for terrorist acts and were reviled by many Afghans as persecutors and murderers. This made it politically very difficult to give them a seat at the negotiating table for a debate over the future of Afghanistan.

 

Nevertheless, the complete exclusion of the Taliban from the political and constitutional process, and the largely rural, largely Pashtun populations the Taliban derive support from and claim to represent, has come to be seen as a significant flaw in the Bonn framework and as a source of continuing instability. As Lakhdar Brahimi, the UN special envoy who convened the Bonn Conference, described in a 2008 Washington Post article, ‘I regret bitterly not having advocated even more forcefully’ after Bonn ‘to reach out to those members of the Taliban potentially willing to join the political process’. Steve Coll’s new history of the post-Bonn Afghan conflict, Directorate S: The CIA and America’s Secret Wars in Afghanistan and Pakistan, has a revealing description of Taliban overtures to the US to surrender on favourable terms in 2002 before the Constitutional Loya Jirga was announced. And President Karzai, who has an ambivalent relationship with the Taliban, struck a surprisingly conciliatory note in a speech just after his re-election in 2009, to ‘call on our Taliban brothers to come home and embrace their land’.

One of the most significant debates among the delegates at the Constitutional Loya Jirga was whether a presidential or parliamentary system of government was most relevant for Afghanistan. A presidential system was most analogous to the constitutional monarchy that was established by the 1964 constitution, the starting point for the new document. A strong presidential system was also better suited for quick executive actions deemed necessary to jump-start reconstruction in a country devastated by war. It was attractive to the interim government, led by Hamid Karzai, who, as the presumed president-to-be, had strong incentives to give maximum power to the new post. Moreover, a parliamentary system would empower political parties that tended to be ethnically divisive, led by the same warlords who were largely responsible for tearing the country apart over the previous decades.

 

On the other hand, putting such vast powers of patronage into the hands of a single chief executive was a significant risk in a country where the central government historically had few resources to exercise its writ beyond a handful of urban centres. Afghanistan has also suffered from recurrent tensions between centrist reformers and rural conservatives who neither asked for nor received significant services from the state and who have fiercely defended their local autonomy. A president who could choose representatives down to the local level would have to constantly perform a precarious political balancing act to appoint people with both loyalty to the state and local legitimacy. A parliamentary system would theoretically extend power to different regional and political groups, diversifying accountability beyond the presidential palace. A parliamentary system risks political gridlock but might insulate against an ill-advised or incapable president making sweeping decisions that alienate key constituencies.

 

The debate between presidential and parliamentary systems of governance re-emerged in the aftermath of the 2014 presidential elections when Ashraf Ghani, a Pashtun candidate running as an independent, won a controversial victory over Abdullah Abdullah, a member of the Jamiat-e Islami party that had its main support base among Tajiks in the north. There were widespread indications of fraud across the country during the election and the core of the Northern Alliance threatened to form a ‘parallel government’ if the results were not reviewed.

 

This led to a political crisis that the US and UN mediated, forging the NUG Agreement. According to the deal, the presidency would go to whoever received the most votes after a complete audit of the election results. The runner-up would be a Chief Executive Officer, who would have a prominent role in government decision-making, including ‘parity’ of appointments to national positions. The agreement also called for a Constitutional Loya Jirga within two years to ratify whether such a CEO position, or ‘executive prime minister’, should be enshrined into the constitution. An internationally supervised audit of the votes found 11 per cent of the ballots cast were invalid, but also confirmed that Ghani won a clear majority.

 

Four years later, the Constitutional Loya Jirga has not been held. But the demands by prominent Northern Alliance members for a system of government that more resembles parliamentary democracy remain strong. One reason appears to stem from doubt that the current electoral and constitutional system would enable a non-Pashtun to be elected president, combined with a belief that when not in power they will be denied what they consider to be a fair share of presidential patronage. Having a prime minister and president, with explicit shared duties of governance, would be one way to protect against exclusion. The Taliban have not engaged in the discussion over systems of government. But a parliamentary system may hold advantages for them because while they are not likely to win a national election, having a small voting bloc in parliament could provide influence over choosing a prime minister or in deciding national legislation.

 

Of course, there is no way to guarantee winning a majority coalition – or holding a swing vote – in parliament. The size of the population is unclear and a cause of great political tension. An effort to issue new biometric national identity cards to prevent fraud has been delayed by a controversy over the degree to which ethnicity should be recorded. Moreover, the current presidential electoral system undermines political parties and incentivises leading candidates to run as individuals rather than representing political party members. Changing the system of governance without significant electoral reform is a gamble with unpredictable odds.

The current voting system – the Single Non-Transferrable Vote (SNTV) – works on the basis one single vote per person, for one candidate within a multi-member district. The number of candidates per electoral district (a province) depends on its rough population size (Nimroz has two seats, Balkh has 11, Nangarhar has 15, Kabul has 33 and so on). Any number of candidates may stand for election – and in each election hundreds have done so in most provinces. What this means, however, is that a great many votes are cast but few of these end up being for winning candidates. Those who win do so with relatively few votes and the margins of victory are exceedingly slim. Among the current 33 members of parliament from Kabul Province, for example, the most popular member received 16,500 votes, with a majority of members receiving fewer than 4,000 votes. Overall, the total parliamentary delegation in Kabul received approximately 100,000 out of 480,000 total votes cast – a ‘waste’ of more than 75% of constituents’ votes. Voters’ ability to hold their elected representatives to account is thereby greatly reduced.

 

The SNTV system also reduces the strength of political parties because it is difficult to apportion votes to different candidates from the same party within a multi-member district. If a party fields only one popular candidate in a constituency, it may receive a large number of votes but win only one seat. But if a party fields several popular candidates and the vote is split among them, it could win no seats because popular candidates running as independents could take all the top spots. Overall, independent candidates who get support from targeted vote blocs within a province, including warlords who attract voters either through intimidation or corrupt patronage, tend to do better. As elected officials, they are not beholden to a party.

 

Almost everyone agrees that eliminating SNTV is key to improving both elections and the function of parliament – except for sitting members already elected under SNTV. But there are strong disagreements over what to replace SNTV with. Like with the debate over presidential versus parliamentary democracy, this has ethnic dimensions that has led to political gridlock and the perpetuation of a highly undesirable status quo. The Special Electoral Reform Commission (SERC) in December 2015 recommended a change to the electoral system under its mandate from the NUG agreement. During the commission’s deliberations, some members favoured a proportional representation system that would have voters choose a political party on the ballot, which would include a slate of candidates running under that party’s banner. Then the number of candidates who actually won a seat would be determined in proportion to the number of votes cast for each party. This system would greatly enhance the influence of political parties on the electoral system and elevate the importance of party platforms.

 

Other members favoured a change from the current multi-member districts to single-member districts, in which the seat goes to whoever gets most votes, whether they choose to affiliate with a party or not in a ‘first past the post’ arrangement. This system was used to elect members of parliament under the 1964 constitution. It would have the important benefit of ameliorating a problem in large or ethnically divided provinces where security and access to the ballot determines who gets elected more than the strength of a candidate’s campaign. In the ethnically diverse Ghazni province, for example, in 2010 all 11 members of the province’s parliamentary delegation were Hazara because security was greater in Hazara areas and turnout was low in others. If single member districts are drawn carefully, representation can be more evenly distributed because even insecure districts would be guaranteed a seat whether one person or 100,000 people come out to vote. If district boundaries are unfairly gerrymandered, however, a change to single member districts could embed polarising tensions into the electoral system, with destabilising results. With trust at such low levels in Afghan politics, it is difficult to see how a nationwide process of drawing electoral district boundaries can be conducted quickly or quietly.

 

The debate over a change of electoral systems ended with no result. The SERC ultimately recommended a hybrid system in which some seats in multi-member districts would be reserved for political parties, and others for independent candidates. The parliament ultimately failed to pass new legislation to move away from SNTV. President Ghani passed a decree empowering the Independent Election Commission (IEC) to study the issue further, whereby the IEC recommended reducing the size of the constituencies. To date, the cabinet has failed to act on this. In March 2018 a coalition of 20 political parties, including major parties with different ethnic compositions, sent an open letter to the IEC calling for a switch before the next elections to a hybrid representation system as recommended by the SERC. This fraught debate demonstrates that even though almost everyone agrees that the current electoral system is undesirable, the political stakes are too high to agree on fundamental reforms.

The 2004 constitution calls for direct elections for the lower house of the Afghan parliament (Wolesi Jirga), provincial councils, district councils, village councils and mayors of the major municipalities. The Wolesi Jirga has 249 members who are directly elected from multi-member provincial constituencies. The 102-member upper house (Meshrano Jirga) is composed of one representative per province elected from among the provincial councils, and one per province from the province’s district councils, with a third appointed directly by the president. So far, however, district council elections, village council elections and mayoral elections have not been held. The last parliamentary election was in 2010, meaning that the current parliament has overstayed its five-year constitutional term by four years and counting.

 

There is often a debate in democratic transitions about whether it is best to have local or national elections first. Not much time was spent on this question in Afghanistan, where the international community wanted to maintain a light footprint and establish Afghan sovereignty as quickly as possible. The Bonn Agreement called for presidential and parliamentary elections within two and a half years, whereby the transitional government would become fully sovereign. As it happened, parliamentary elections were delayed by a year for logistical reasons, and presidential elections were prioritised and held in 2004. Focusing on the presidency made sense in a country where most infrastructure was undeveloped or destroyed, massive amounts of reconstruction assistance needed to be managed from a central location, and international relations were a critical part of stability and development.

 

De-prioritising local elections increased the risk of alienation of rural constituencies from the government, however, particularly in a system with few checks on presidential appointments to local positions. Holding local elections could have increased the legitimacy of local leaders and introduced some local accountability if government officials were ineffective or corrupt and would have increased understanding of the democratic and electoral process among citizens who had had little experience with it over the past decades. A 2015 impact evaluation of the Afghanistan National Solidarity Programme, which facilitated local development projects based on locally elected Community Development Councils (CDCs), found that where CDC elections were held, voter participation in the 2010 parliamentary elections increased. In fact, local communities had traditions of quasi-democratic self-rule in the form of local (usually all-male) councils (shuras) that would discuss and decide certain justice and governance issues.

 

Those who favour a strong central government with nationally focused elections note that warlords or corrupt actors would have an advantage in local elections and would pervert the system. This is likely true in the short term, although less so in safe, pluralistic communities, including many urban ones. On the other hand, holding large and confusing national elections has enabled wholesale fraud in several of the past ballots, which favours corrupt and powerful actors with less accountability to constituents.

 

Tellingly, once the Taliban insurgency reached a critical mass in 2008, many plans for stabilisation called for the establishment of community councils to address local grievances as a key to reducing violence. These tended to be ad hoc, however, and actually establishing the constitutionally elected village and district councils was never seriously considered. As discussion turns to ways that the Taliban can be fragmented and reconciled with the government, local elections emerge as a potential way to enable local political diversity without changing the overall reform direction of the country. Creating or re-empowering local governance bodies could be a useful first inroad for the Taliban into power, without forcing out a body of incumbents.

In mandating the Wolesi Jirga, provincial council, mayoral, district council and village council elections, the 2004 Constitution states that the polls must be ‘free, secret, universal, and direct’ (emphasis added). The word ‘direct’ is standard parlance for election systems and was used for local and parliamentary elections in the 1964 Constitution as well. It means that individual citizens must cast their vote for the candidate or party that is running for a designated seat and not for a representative or elector who will ultimately decide on who fills the seat.

 

The US system for electing the president is indirect because it relies on an Electoral College to cast the final votes for the president. US voters in fact are choosing ‘electors’ at the ballot box who have indicated their candidate preference. The number of electors each state has is proportional to national population. It is the electors who actually cast the deciding votes for the president weeks after the national vote is held. This system was designed to give states of differing sizes a proportionate say in who becomes president. Electors also have the ability to vote against a candidate they believe represents a grave mistake for the country (depending on state law). An indirect election can therefore be used to ensure balance of regional influence and moderate extreme influences in an electorate.

 

Afghanistan has a long tradition of indirect democracy that has facilitated both national and local governance. Shuras and jirgas provide a form of local indirect democracy, with leaders chosen on the basis of seniority and judgment to represent the views of a community and make decisions on their behalf. Since 2001, there is anecdotal evidence that indirect elections for positions outside the constitutional requirement of direct voting have been more inclusive and less controversial. Most prominently, delegates to the Emergency Loya Jirga and the Constitutional Loya Jirga in 2002 and 2003 respectively, used indirect election methods. In the Emergency Loya Jirga, members of designated districts selected 20–30 representatives who then gathered to cast secret ballots for the specific delegates who would attend Loya Jirga itself. In the Constitutional Loya Jirga selection process, a caucus system was used whereby designated constituencies gathered and chose delegates based on a more traditional consensus-based process than strict voting procedures.

 

At the local level, CDCs are elected as part of a World Bank-sponsored national development programme. The rules call for secret and direct elections, in which women are required to participate. A mid-term evaluation of the National Solidarity Programme in 2006 found that the resulting councils were largely seen as representative, but that it was doubtful the direct and secret balloting procedure was used in many cases. Also, elections for temporary district councils undertaken as part of stabilsation programmes used indirect methods. The District Development Assemblies (DDAs) were chosen in a two-tier process by which selected community leaders from around a district gathered to vote secretly for a DDA, which helped to assign priorities and make decisions on the distribution of development programmes within the district.

 

There are definite downsides to indirect elections. They dilute citizens’ voting rights and are subject to elite capture. Without fair administration, minority groups can be excluded. But indirect elections can be a way to mitigate the distortion of regular voting rights through insecurity, disenfranchisement and corruption of the voting system itself. In the cases cited above, indirect election processes were employed as pragmatic shortcuts to get around either logistical difficulties of full voter registration and voter education or alternative problems of exclusive elite capture that would not fairly represent the interests of a larger polity. They arguably enabled balancing of political interests in ways that were quicker and more efficient than a more thorough process of civic education about election systems and the development of credible institutions to operate checks and balances within the legal and political system.

 

While citizens’ understanding of elections and democratic governance principles has increased since these indirect election systems were developed, security conditions have deteriorated dramatically and trust in electoral authorities has declined. It may therefore be reasonable to continue to consider indirect electoral mechanisms to get avoid results that are perceived as exclusive or inequitable, particularly at local levels. If, like in the Emergency Loya Jirga, citizens give their proxies to genuinely respected community leaders, then it may be easier to select more representative leadership than has been the case in opaque and corrupt direct election processes. In the event of an eventual Taliban deal, indirect elections may more easily allow for reconciled Taliban to join local politics.

 

At the national level, use of an Electoral College system in Afghanistan would help to ensure that in presidential elections each province would have a guaranteed share of influence based on its population. This would address the current disparities in voter access across the country, where voters in insecure areas cannot get to the polls and in conservative areas where women face greater barriers to voting. The practical effect would be to increase the influence of provinces like Helmand and Zabul, which have historically low turnouts that leave its voters with little say in the outcome of Presidential elections. It would also reduce pressure to stuff ballots in insecure areas to compensate for perceived disenfranchisement, thereby increasing overall perceptions of electoral integrity. Such a system depends on having an accurate and accepted census, however, which has been an impossible task since the Bonn Agreement.

Once the first Afghan Parliament was inaugurated in November 2005, the UN role in sharing administrative duties ended and the Afghan government became officially fully sovereign. Since then, the international community and the Afghan government have had a complex relationship whereby international donors provide essential assistance and policy advice to pursue their own interests while also trying to respect Afghan sovereignty and allow for Afghans, who know their country best, to develop effective solutions to security, political, and development problems. The challenges of this convoluted relationship have led to the mantra that Afghan elections and the peace process must be ‘Afghan-owned and Afghan-led’.

 

This ambition holds true from a moral and a legal perspective. But the frequent use of the phrase masks the fact that the international community, including neighbours outside the Western donor community, has enormous influence over political arrangements in Afghanistan. International assistance is vital to funding and maintaining the state. And, tellingly, at times of deep political or security crisis, the international community, led by the US, NATO and the UN, has intervened to mediate a solution to a crisis. The forming of the NUG in the wake of the controversial 2014 elections is just the latest example.

 

The international community is therefore both a safety net to avoid political and security disasters, but also bears some responsibility for tipping the scales in one direction or another to resolve crises in ways that serve Western security interests but may destabilise Afghan politics. At the same time, the international community’s ability to use its leverage is constrained by the fact that if substantial international assistance is withdrawn from Afghanistan, the state is likely to collapse and the core goal of the international community to prevent safe havens for transnational terrorists will not be met. Afghan political leaders, and the Taliban, know this and a fragile balance of power is somewhat maintained but with a steep cost in violence and instability. Afghanistan’s neighbours are also wary of the chaos that could escape Afghanistan’s borders in the event of state collapse, and Pakistan in particular fears having a government in Kabul that would act too favourably toward India. Therefore, countries in the region have also intervened significantly in Afghanistan’s internal politics by supporting proxies that serve foreign interests but keep the situation unstable.

 

Amid this complex dynamic, international leadership can help navigate a way out of the current stalemated political dynamic if it is applied in a coordinated and strategic way. If recent history is a guide, it will otherwise take a destabilising crisis for the international community to act. International actors should first acknowledge that there are flaws in the current political architecture and give cautious support to political reform processes conducted according to shared principles that Afghan actors agree upon – including ideally the Taliban. Such principles might include: the status quo is divisive and destabilising; meaningful inclusion of all non-violent political and ethnic factions is essential; and changes to the current system must be consensual and in accordance with the law. Then international actors could play a mediating role to facilitate a consensus view on the process by which political reform could be achieved.

 

In many ways, the NUG Agreement provides an initial blueprint for political reform negotiations – notwithstanding the severe challenges this power- sharing arrangement has experienced in practice. Finding a way to diversify the powers of the presidency among different groups is a key demand. Devolving some power to the provinces will reduce central government control but may buy political stability. The fact that negotiations over the removal of Governor Atta centred around the core demands of Chief Executive Officer Abdullah in the NUG Agreement discussions, including a shift toward a parliamentary system of government, indicate the former Northern Alliance’s underlying demand for more effective power-sharing among regional and ethnic groups is not going away. Although in the end the crisis was resolved by negotiating a few presidentially appointed positions, the fundamental instability of the system remains unchanged: without addressing the system anyone with power can stall political progress for months to get patronage concessions.

 

Some important changes can be taken by executive action – although those are most susceptible to change and trust levels in the durability of executive action are low. One opportunity would be to support the formation of a commission to formulate amendments to the constitution that was called for in the NUG Agreement. This need not lead to a Loya Jirga right away but could help to define the terms of more inclusive power-sharing arrangements.

 

To achieve more lasting change there would need to be fundamental revisions of the law and the constitution. However, these are nearly impossible for the Afghan political actors to achieve in the current heightened state of tension.

 

The first and greatest opportunity is to facilitate political accommodation is around the 2019 presidential elections. The international community has been very wary of intervening directly in the electoral reform debate because of the sensitivities around international interventions to resolve crises after both the 2009 and the 2014 presidential elections. The Afghan government has made it clear that electoral reform is solely a national issue. On the other hand, from an international perspective the failure of electoral reform has led to political gridlock that affects international security interests. While different factions within the Afghan government have argued over their roles in decision-making, the Taliban have gained territory and an increasing number of international terrorist groups have found a foothold in Afghanistan. More active international mediation of the political differences that have blocked progress on electoral reform could help to break an important logjam and enable a more credible election process in 2019.

 

Another – likely later – opportunity to advance reforms that would bring about more inclusive governance lies in the conduct of a peace process with the Taliban. No one has wanted to re-do the Bonn process or open the constitution to major reforms because of fear that human rights, women’s rights and democratic principles might be set back. But after a decade of deteriorating status quo, one wonders when the slow, steady decline of stability will slip below the worst-case scenario outcome of major reforms and it will seem like the risk is worth taking. Apart from the presence of international forces on Afghan soil, the Taliban’s biggest grievance appears to be their exclusion from the Bonn Agreement and the 2004 Constitutional Loya Jirga. It is likely that a peace process would force a re-examination of the fundamental structures of government and create space for new deals to emerge.

 

Any significant change to the political system or the constitution must take the negative lessons of political exclusion into account. The more major the reform, the more important it is to attempt to include Taliban representatives – as well as the major non-Taliban ethnic and political factions – in the process. Given the instability of the status quo, there is a need to make progress on reforms without waiting for an uncertain peace process. But even without Taliban participation, reforms should aim to create more space at local level for the Taliban and non-violent opposition groups to have a greater and safer space in the Afghan political process. The 2019 presidential elections, preceded by President Ghani’s peace offer to the Taliban extended during the March 2018 Kabul Conference, creates a fluid situation that can be unstable, but also an opportunity to make progress on greater political inclusion.