Vote buying—the practice of providing or promising cash, gifts, jobs, or other things of value to voters to induce them to support a candidate in an election—is illegal in 163 countries, yet it is a widespread and seemingly intractable problem in many parts of the developing world. In Ghana, for example, incumbents distribute outboard motors to fishermen and food to the rural electorate. In the Philippines, politicians distribute cash and plum short-term jobs. In 2015, Nigerian incumbents delivered bags of rice with images of the president ahead of the election. And Werner Herzog’s 2010 documentary film Happy People shows a politician cheerfully delivering dried goods along with musical entertainment to an utterly isolated village of trappers in Siberia (49 minutes into the film). Thus, recent instances of vote buying are more varied than the simple cash for vote exchange; they include awarding patronage jobs and purposefully targeting social spending as a reward for political support.
Vote buying not only distorts the outcomes of elections, but it also hurts the (usually poor) communities where this practice is rampant. It might be tempting to say that at least those who sell their votes receive something from their government, but in fact, once these citizens are bought off, their broader interests are left out of the government’s decision-making process, as the incentive to provide public goods to that group disappears. A study in the Philippines, for example, found that vote buying correlates with lower public investments in health and higher rates of malnourishment in children.
While some commentators occasionally (and condescendingly) suggest that vote buying is a product of non-Western political norms and expectations, this could not be further from the truth. Although wealthy democracies like the United States today experience very little crude vote buying, vote buying in the U.S. was once just as severe as anything we see today in the developing world. In fact, during George Washington’s first campaign for public office in 1758, he spent his entire campaign budget on alcohol in an effort to woo voters to the polls. By the 19th century, cash and food occasionally supplemented the booze, particularly in times of depression. Even as late as 1948, a future president won his senate campaign through vote buying and outright fraud.
Yet while U.S. politics today is certainly not corruption-free (see here, here, and here), it has managed to (mostly) solve the particular problem of vote buying. Does the relative success of certain U.S. efforts hold any lessons for younger democracies? One must always be cautious in drawing lessons from the historical experience of countries like the U.S. for modern postcolonial states, both because the contexts are quite different and because suggesting that other countries can learn from the U.S. experience can sometimes come off as patronizing. Nevertheless, certain aspects of the United States’ historical strategy to combat vote buying might be relevant to those countries struggling with the problem today. Let me highlight a few of them:
- A truly secret ballot: A formally secret ballot (the “Australian” ballot) was a key reform in the United States and is now ubiquitous across the world. (Different U.S. states adopted the secret ballot at different times over the course of the 19th century. In 1891, Kentucky became the last state to eliminate the oral ballot.) Yet while a secret ballot is a necessary first step to combat vote buying, a secret ballot alone is not sufficient so long as other, informal methods of voter monitoring are available. For example, the secret ballot is of little use in small communities where a local official can look the voter in the eye before and after they cast their “secret” ballot. The secret ballot must therefore be augmented by accompanying reforms that make it difficult to informally monitor voters. In the U.S., vote buying continued for several decades after the introduction of the secret ballot, until the political machines were dismantled and it became impractical for party officials to monitor how the radically enlarged electorate voted. Supplemental reforms to the secret ballot which might cut down on informal voter monitoring include banning electioneering near polling places (as many U.S. states do), assigning non-local people at random to staff election booths, and drawing larger voting districts.
- Systematizing social spending: In the 19th and 20th centuries, political machines in U.S. cities provided the poor electorate with targeted support (cash, jobs, etc.) in exchange for their votes. These machines, with their vast networks of local agents, used specific kickbacks to reward the faithful and to penalize the unfaithful. As Professor Susan Stokes documents, progressive era reforms in the United States made social spending more formalized, rule-bound, and programmatic; as a result, political parties could no longer penalize or reward voters according to their choice on Election Day. And while social programs change with the party in power, they change in systematic ways. The general principle seems to be: insulating decision-making about individual beneficiaries of social spending from partisan politics helps to combat vote buying. While this strategy is of little help in combating more direct forms of vote buying (such as direct transfers of cash or gifts from politicians to voters), limiting the ability of political machines to use the public purse to buy votes is an important step in the right direction.
- Civil Service Reforms: The role of civil service reforms in combating vote buying is quite similar to that of social spending reforms. Jobs are probably the most valuable handouts that a winning party can give to its supporters. The power to make countless political appointments feeds a clientelist system and kept U.S. political machines running for generations. Blowback came in the late 19th century, when a disgruntled potential appointee assassinated President James Garfield. His successor, President Chester Arthur, signed into law several measures to control the patronage appointee system. For example, the 1893 Pendleton Act required entrance exams for new bureaucrats. Raising the bar for entrance to the civil service and whittling down the number of political government appointees are good first steps. But separating the civil service from successive monopolization by political parties is a long process and in the U.S. it took decades. Almost a half-century later, the Hatch Act was enacted as a direct response to politicians coercing voters with contracts and jobs in the Works Progress Administration (the largest New Deal agency). The Hatch Act explicitly forbids bribing voters, but it takes the further step of severely limiting the campaign activities federal employees may engage in. While formalizing civil service and social spending are essential reforms for combating vote buying, if the U.S. is any indication, these efforts take time.
The three recommended areas of focus are neither exhaustive nor sufficient mechanisms for combating vote buying, but they offer good general principles for young democracies today. As research from the Abdul Latif Jameel Poverty Action Lab (JPAL) has demonstrated, the modern context also allows for powerful new tools, such as well-timed, targeted radio advertisements and television campaigns. It might just be that the key to moving forward in the fight against vote buying is to draw upon the principles derived from historical experience of older democracies, and apply those principles with the aid of modern methods of outreach.
GAB | The Global Anticorruption Blog 
Thanks for the post, Mike! I find your second and third points most compelling because I am not convinced that the ballots cast in US elections are truly “secret” given that, as you allude to, right outside the polls voters will encounter numerous “exit” pollsters immediately asking for information on the ballot cast. Of course, the fact that the ballot itself was cast in secret allows the voter some privacy and she can always lie or decline to answer outside the polls but the presence of these pollsters might make some feel as though they’re votes aren’t really secret.
Thanks Clara!
I would agree that the ballots cast in US elections are not really secret. It seems like lots of people enjoy taking ballot selfies, despite the fact that doing so violates election law in 18 states. But I don’t think that ballot selfies or people outside of polling stations are “monitoring” voters in a way that leads to political organizations rewarding or punishing specific individuals for their decisions within the ballot box. So while the votes might not be ‘secret’, under the general definition of that word, they are actually insulated from the coercive actions of political machines.
Thanks for the post, Mike! I find your second and third points most compelling because I am not convinced that the ballots cast in US elections are truly “secret” given that, as you allude to, right outside the polls voters will encounter numerous “exit” pollsters immediately asking for information on the ballot cast. Of course, the fact that the ballot itself was cast in secret allows the voter some privacy and she can always lie or decline to answer outside the polls but the presence of these pollsters might make some feel as though their votes aren’t really secret.
Reblogged this on Matthews' Blog and commented:
It remains to be seen, the impact.
While I tend to agree with most of the post, Mike, I am left wondering about your statement in the second paragraph that voters are not necessarily effectively “selling” their vote for some actual service that they would otherwise be without. While you site the study in the Philippines, it seems to just show correlation rather than causation. Given that there does seem to be some connection between access to public services and vote buying, I’m left wondering if completely discounting the efficacy of the “transaction” that occurs when voters sells their vote narrows the view of potential solutions. It seems to me that if their is some value add for voters in selling their vote, even if it has other negative effects, then expansion of civil services (in addition to the civil service reforms you bring up) might be an effective tool at combating vote selling, particularly in environments were vote selling has been normalized due to economic realities.
Agree with Nick. I think one of the reasons why this is such a challenging problem to solve is because it may be individually rational and more obvious benefits for the people receiving the benefits (“I get more benefits/goodies when I vote for this person” ) versus very abstract costs (“this is bad for democracy”).
I too see lots of promise in more systematic social policies like CCTs and ones that are difficult for bureaucrats to influence / make more discretionary. I enjoyed reading the post!
You make good points! And untangling the causation/correlation relationship between civil services and vote buying would be a very tall order. If you simplify reality way down, you could reasonably say that increasing civil services might decrease vote buying/selling. But you could also reasonably say that it would just increase the price of the votes sold.
For example, let’s say that rice is handed out on election day in exchange for votes. Now pretend that the rice is a civil service. The government makes some reforms and now hands out more rice. Suddenly, rice isn’t going to buy you very many votes. But maybe something else will, say, chickens. The price of the votes have gone up, civil services have improved, but vote buying remains.
This analogy doesn’t account for the infinitely more complex terrain of reality, but I think it shows some of the underlying difficulties that you’re both getting at.
Thank you for the post. As you mention, some of the reform, such as meritocratic civil service system, took many decades for the U.S. to implement, and even today it is criticized as many government incapable of firing bad people. As result, it would probably be hard for a developing nation to carry out. I like the end of the bullet point on secret ballots because it gives fairly simple reforms like random assigning non-locals to staff polling locations, banning electioneering near voting locations, and drawing larger voting districts. All of these seem beneficial, low-cost, and relatively simple. Are there any other example of intermediate steps in the other two bullet points?
Thanks Nino, and great question! Unfortunately, I have no answer. These were some of the historical recommendations that seemed easiest to implement, drawn from Susan Stokes’ extensive work.
You make a second point in passing that I think is interesting – that of ‘intermediate’ steps. Looking at US history, the Pendleton Act might be considered an intermediate step. It didn’t really get us to where we wanted to go, and maybe the Hatch Act doesn’t do that either. The question, and I don’t know how you would go about answering it, might be, were we in a better position after the Pendleton Act to take another step in the right direction, or were we not?
As for the idea that the government is incapable of firing bad people, I could be wrong on this but I get the sense that this argument has a certain framing. The idea is: this person is not good enough, they should be fired. The idea is not: this person is a democrat, so they can’t be fired. While there may be problems about getting rid of government employees who don’t do their jobs properly, it seems unrelated to the issue of patronage positions.
Great post!
Even though I agree with the strategies you highlighted, I think that none of them can be really effective if not followed by a robust legislation against vote-buying and equally robust sanctions.
Since vote-buying is a multifarious behavior that certainly has deep roots in some societies, I think that it can only be tackled with civic education. This, in turn, should target not only voters but also candidates and political parties.
Yes, you are absolutely right! This is an important aspect of reforms that I completely disregard in the context of the post. For the population, who provides the civic education- perhaps those who have purchased their government offices? That’s why your point that civic education should be targeted at candidates and parties is key.