Remittances—the money or goods that migrants send home to support their families and friends—have become increasingly important in developing countries. In nations like Haiti, Honduras, and Tajikistan, remittances account for more than 20% of their respective GDPs. Interestingly, many of these top recipient countries also rank among the most corrupt in the world, at least according to Transparency International’s Corruption Perceptions Index. While that correlation does not by itself establish causation, it does invite the question of whether large flows of remittances have any meaningful impact on corruption within the recipient country.

Surprisingly, this question has been “virtually ignored” in academic literature. Only a few studies investigate the connection between remittances and corruption, and the handful of papers on this topic come to quite different conclusions.

The more common view in this literature is that larger flows of remittances into a given country are, all else equal, likely to increase corruption within that country. The principal argument for why this would be so is that when remittances increase, there’s more money available for corrupt officials to capture. For example, the individuals who receive remittances will often want to use this money to secure goods and services that can be obtained only by bribing the official gatekeepers to these resources. Essentially, as the supply of potential bribes increases, so too does the demand of bribe-solicitors.

A second argument as to why remittances may be associated with greater corruption is that when individuals receive money from their relatives or friends abroad, those recipients are less reliant on their domestic governments. They are therefore less directly impacted by corruption, and so less likely to take action to demand cleaner government; this makes it less costly for government officials to engage in corruption. Put another way, remittances make the recipients more likely to tolerate corruption. (It’s worth noting that this argument is in some tension with the first argument: The first argument presumes the remittance recipients are more likely to be adversely affected by corrupt officials—through increased solicitations of bribes—while the second argument presumes that recipients’ increased income makes them less likely to be affected by domestic corruption.)

Although the idea that remittances are likely to increase corruption may seem intuitive, it has been challenged. Indeed, there is some evidence that high remittance flows may, after controlling for other factors, be associated with lower levels of corruption. Why might this be? Those arguing that remittances tend to reduce corruption take the two arguments sketched above and turn them on their heads:

First, while it may be true that remittance recipients, simply by virtue of having more money, become more attractive targets for corrupt low-level officials, it could be that the recipients’ increased economic standing empowers them to resist paying these bribes. Additionally, some of the low-level government officials who might be inclined to solicit bribes are themselves remittance recipients, and so may be less reliant on bribes as an additional source of income. (This latter point is a variant on the familiar argument that raising bureaucrats’ salaries can reduce petty bribery.)

Second, while those who believe remittances increase corruption argue that recipients’ lower dependence on the state make them more tolerant of corruption, those taking the opposite position argue that this lower dependence puts pressure on officials to reform corrupt institutions in order to regain lost political and monetary support from the populace. Additionally, even if recipients’ outside source of income marginally decreases their incentive to pressure the state to improve institutions, this may be more than offset by the fact that their higher incomes substantially increase their capacity to push for greater accountability, for example through increased political activity and support for government opposition.

Those taking these contrary positions have each found support in certain, admittedly limited, empirical studies. What explains the divergence in these findings? While it is hard to say anything definitive given the small number of studies and the difficulties of isolating and assessing the proposed causal mechanisms, a closer examination of the existing research suggests two important factors that may account for the differing results.

• First, the impact of remittances on corruption in the receiving country may depend on whether that country’s government is democratic. Less democratic regimes are in general less accountable to the public and so will be less responsive to increased pressure for reforms from remittance recipients. Furthermore, insofar as remittances free up state resources that would otherwise be spent on providing goods and services to the public, less democratic regimes are likely to use those resources to finance activities that deter political competition (particularly military spending), a result which tends to increase corruption.
• Second, the institutions of the sending country (that is, the country where the migrants are working) may play a role in determining the behavior of remittance recipients back home. Migrants who live in countries with more democratic institutions may transmit certain values—sometimes called social remittances—along with their money, and these values inform how recipients spend (or don’t spend) those funds. Indeed, some studies suggest that social remittances may explain why recipients are less tolerant of bribe solicitations. Recipients of money sent from more democratic countries are also more likely to promote democratic values in their home countries through political participation and support of anticorruption agendas.

If these conjectures are accurate, then in a country like Tajikistan, an autocracy whose remittances come mainly from migrant workers in Russia, remittances are likely to worsen the country’s corruption problem. But in a country like Haiti, which struggles between autocracy and democracy and has a substantial number of migrants who live and work in the U.S., remittances might have a more positive (though admittedly marginal) effect.

Again, though, right now we simply don’t have enough information to confidently say how remittance levels affect corruption in the recipient countries. Future research on this topic should investigate more specifically the relative importance of the sending and receiving countries’ institutions, and should consider how various migration routes influence corruption in the origin country. As the research progresses, migrant-receiving countries may learn that they can further their anticorruption agenda abroad by offering support and economic opportunities to migrants at home.