In spite of Mali’s international recognition as a beacon of good governance in the region, decades of free and fair elections and extensive civil liberties failed to engender public support for democracy following the March 2012 coup d’état. Demonstrations against the military junta were half as large as demonstrations in support of it. What looked to be strong democratic institutions from the outside were often hollow shells that privileged the elite class and marginalized everyone else – not a system ultimately worth fighting for. This essay discusses some of the constraints to democratic accountability in Mali and what can be done to mitigate them. In light of the July election date currently being touted by influential Western powers, it is critical that the same mistakes not be remade, pushing Mali back into a democracy in name only.

During its two decades of democratic rule, Mali was praised for the fact that power changed hands peacefully from one political party to another, that most elections were judged free and fair by election observers, and that civic and political associations abounded. I was therefore surprised on visits to the country over the past decade by the general apathy toward and isolation from the democratic process expressed primarily by rural villagers who make up 80 percent of Mali’s population.

For democracy to work, for it to engender a modicum of accountability to a majority of its citizenry, requires a minimum level of civic engagement on the part of that population. At the most basic level, voters should place some value on the ability to freely select leaders. Further, voters should be able to make an informed choice when selecting those leaders. Even compared to other African countries, Malian voters do not appear to be in a position to achieve these goals. When it comes to voters expressing a preference for democracy over other forms of government such as military or single-party rule, Mali ranks in the 72nd percentile of African countries.1 Mali is also the third worst African country in terms of adult literacy (UNDP 2012) -- a critical skill necessary for engaging in civic and political life.

Such constraints have consequences for government accountability and representativeness. Corruption and nepotism pervade Mali’s political elite class who are seldom punished electorally or otherwise. And with the world record for low voter turnout in the 1990’s, Mali’s voter participation has improved little in recent elections (IDEA 2013). To be more than a democracy in name only in the future requires a deeper understanding of democracy’s past failures, some of which I discuss below. These and other shortcomings have not disappeared since the coup, so rushing Mali into another round of internationally sanctioned elections risks erecting another democratic facade that fools everyone but its own people – at least until the walls come crumbling down.

To better understand the roots of Mali’s low levels of accountability and lack of public support for democracy, my research attempts to identify and measure constraints to the democratic process. Here, I discuss three key findings: severe information asymmetries between voters and politicians reduce voter control over elected leaders, the persistence of inegalitarian social norms undermines civic participation, and institutional arrangements dampen political competition and facilitate illicit collusion. While not an exhaustive list, this discussion sheds light on some of the reasons Mali’s democracy failed in the past, and suggests avenues for future policy.

This research project began with a hunch – that Malian voters, particularly rural ones, had inadequate information about what their local governments could do for them. From early conversations in rural villages, I learned that few people knew the kind of budget (albeit small) that town councils had at their disposal to build public goods such as schools and clinics and wells. I also learned that the idea of a reciprocal contract – paying taxes in exchange for public goods – had not permeated much of the rural periphery. I surmised that if Malians expected little of their local governments, then there was nothing to stop elected leaders from living down to very low standards. To test this idea more systematically, I designed and implemented a civics course providing information about the capacity and responsibility of local governments as well as how the democratic process works (or should work) in Mali.

This intervention was randomly assigned to some communes and not others so I would have a proper counterfactual to evaluate its effects. The course, taught in 370 villages by local instructors over three weeks, had an average of 30 participants per village. Men and women attended at almost equal rates and participation was voluntary following a village-wide assembly advertising the course. Comparing survey responses among a representative sample of households in “treated” and “control” villages, I found the intervention improves sanctioning of poor-performing politicians, in particular by increasing the value of the performance dimension relative to other dimensions voters care about such as kinship and patronage. I also found that villagers from treated communities are more likely to challenge leaders at town hall meetings.

That’s the good news: providing information to villagers about what local governments can and should do raises expectations and changes behavior at the ballot box. The bad news is that, in the aggregate, the civics course did not improve other forms of civic participation such as attending meetings or contacting local leaders. Among the same representative households, I compared detailed information on participation in civic life (type of civic activity, reason for activity, date, number of participants, etc.) across treated and control villages. Upon closer examination, I found that while the course had no net effect, it had a significantly positive impact on men and a negative impact on women. In other words, the course prompted higher rates of civic activity among men whereas women participated less after the course came to their village.

Returning to these villages to understand why the course had a perverse consequence on women unearthed another important constraint to democracy in Mali. Pre-existing norms hinder engagement in the democratic process. From qualitative interviews with men and women, participants and nonparticipants, emerged a very strong gender norm against female participation in the public sphere. In the words of one man, “Women have nothing to do with civic or political activities; they should only busy themselves with taking care of their home -- that’s their place.”

By highlighting the importance of civic participation and treating men and women as equals, the course effectively increased the salience of this gender norm. As a result, women met with both implicit and explicit sanctions for engaging (or attempting to engage) in civic activity following the course. One woman reports: “We are scared of imposing ourselves in civic affairs for fear that our husbands will think we have surpassed our limits because we took a course on democracy.” A village chief substantiates this fear: “Women’s participation in these activities should have its limits…an educated woman will forget that she is inferior to men and could even fail to obey her husband. Our custom requires total submission of women to men.” In other villages where women mobilized to lobby the mayor’s office for a development-related project, they were shut down by their husbands or the village chief.

While women’s knowledge and attitudes toward local government improved with the civics course in much the same way as men’s, these intellectual changes do not translate into changes in behavior. This confirms a finding echoed elsewhere in the literature that the business of changing norms is a long-term one. Brief injections of information or even the imposition of democratic institutions cannot undo long-held beliefs and practices.

To complement my research on voting and civic behavior, I also study politician behavior. In a competitive multiparty democracy, we should expect opposing parties to expose the misbehavior of incumbent politicians, rendering an information intervention like mine superfluous. However, Mali’s electoral system of proportional representation at the local level creates opportunities for mutually-assured corruption. When all viable political parties are represented on a local council, political elites can more easily agree to illicit collusion where they jointly enrich themselves at the expense of voters. Part of such a bargain is that every party on the council agrees not to expose the misbehavior of other parties. Using local election results and data on local public goods provision, I show it is only when one party loses a seat on a town council that that party acts as the opposition, threatening to expose bad behavior among parties on the council.

Observations of elite collusion at the local level resonate with Malian politics at the national level. Recently deposed president, Amadou Toumani Touré, won in 2002 and 2007 as an independent without a party. Rather than reject partisan politics, he used a strategy of “reciprocal assimilation of elites,” in which parties chose to be co-opted to maintain access to resources rather than remain in the opposition and lose the ability to support their patronage networks (Baudaise and Chazal 2006). A group of Mali experts agrees that “Touré’s ‘rule by consensus’ had become a mere euphemism for absolute rule with checks and balances existing only on paper” (Lecocq et al 2013).

In sum, I find that information deficits, in particular information about what voters can expect of their governments, imposes a constraint on democratic accountability in Mali. Externally providing civic and political information to voters can change behavior in an anonymous and relatively costless act such as voting, but cannot overcome existing social barriers that inhibit more visible and costly acts of civic participation. Second, information is unlikely to be provided by opposition parties when they have institutional incentives to collude rather than compete as they often do in Mali’s system of proportional representation at both the local and national levels.

These findings suggest several policy implications for domestic and foreign-sponsored governance programs. Information interventions should include details about what citizens can expect of their governments such as what level of government is responsible for providing which services and the types of resources it has at its disposal. Civic education programs, on which the US government alone spends roughly 30 million USD per year, should be wary when promoting democratic ideals inconsistent with local norms. One low-cost remedy is to deliver democracy promotion interventions to women and men separately; this was advocated by men in treated villages who said custom and religion are intolerant of the mixing of genders in the public sphere. Another short-term fix is establishing recourse for women who experience repression. Leading by example has demonstrated more lasting effects on gender norms. In India, for instance, political reservations for women have taught parties that women can be good leaders and have helped make policy more representative of women’s interests (Bhavnani 2009; Chattopadhyay and Duflo 2004).

In the run-up to elections in Mali for which Western governments are now pushing so hard, the findings suggest ways to improve turnout and stimulate opposition. First, clear and transparent information should be provided to voters, not only about candidates and their performance records, but about the major policy issues at stake and resources the government has to address them. Because so much of Mali’s electorate is illiterate, information should be provided orally, over the radio or through visits to rural areas. In parliamentary elections, some proportion of seats should be reserved for women. Researchers are testing an intervention in Sierra Leone that hosted and filmed debates between rival candidates and then broadcast them at community screenings across the country (J-PAL 2013). Such an approach could similarly encourage true political opposition in Mali.

While outside intervention can support a healthy democracy, it can generate perverse consequences if not done carefully, learning lessons from the past. Perhaps it took so long for Mali’s democracy to crumble precisely because it was being propped up by external actors blind to the reality on the ground. The last thing Mali needs is another president who is legitimate in the eyes of the international community but illegitimate in the eyes of his own people. Yes, elections are critical if Mali is to have a chance at a government accountable to its people, but elections that offer no real choice to voters and that do not engage the majority of the population are worse than no elections at all.