“In a recent meeting between a World Bank official and a finance minister from a developing country in which the provision of an income support scheme or safety net was being discussed the minister opposed strongly such scheme. When questioned by the World Bank official about the reason for his opposition, the minister’s reply indicated the worry that such schemes could jeopardize the existence of the support network provided by extended families.” (Attanasio and Ríos-Rull 1999)
When we began working with the government of Tanzania to evaluate a pilot conditional cash transfer program, we heard similar concerns from policymakers: would cash transfers provided by the state erode informal safety nets? After several years of evaluation, the answer appears to be no.
In our study “Cash Transfers, Trust, and Inter-household Transfers: Experimental Evidence from Tanzania,” we compare poor households that received conditional cash transfers (or CCTs) to similarly poor households in communities that didn’t have a CCT program. The communities to receive the pilot CCT program were randomly selected, which boosts our confidence that differences between poor households in the two communities are due to the CCTs.
Nearly three years after households began receiving transfers, they reported higher levels of trust in their community members than households in non-CCT communities. They also reported greater trust in particular community members, like shopkeepers and teachers. This is consistent with earlier research in the same communities, where we showed that households in CCT communities had more trust in their locally elected officials. (Unfortunately, we don’t know how the introduction of CCTs affected the people who didn’t receive them in CCT communities.)
But you might say: Well, it’s easy to say you trust people, but will you help each other out? We asked households whether they knew someone who would provide childcare for them if they needed it urgently, and households in CCT communities were nine percentage points more likely to say that they did. They were also five percentage points more likely to report that someone had recently turned to them for assistance. That said, they didn’t report knowing someone who could lend them 15,000 Tanzanian shillings (or about US$9) urgently. Taken together, these responses suggest that the CCT program led vulnerable households to be more embedded in social networks, not less, even if those social networks have limits. It’s also consistent with the limited evidence we see on this elsewhere: CCTs also boosted cooperation among community members in Colombia.
What about cold, hard cash? We also look at gifts—both cash and in-kind—that poor households receive from other households in both CCT communities and non-CCT communities. While poor households do receive less cash in total gifts from other households after about a year and a half of government cash transfers, those numbers bounce back after almost three years of transfers. While we don’t know exactly why these poor households needed more transfers from members of their community over time, what we do know is that even with a formal, government-implemented cash transfer program in place, households still lent or gave each other cash in times of need.
Whether or not government cash transfers crowd out inter-household transfers varies in other settings: there’s some evidence that they do in Mexico and that they don’t in Colombia. Unsurprisingly, it may depend on how big the government cash transfers are.
Conditional cash transfers are one tool to ease poverty for vulnerable households. Many studies indicate that they reduce poverty rates and increase the use of education and health services. We see evidence for both of these in Tanzania as well. Yet policymakers and others periodically express concerns about these other, sometimes less visible community dynamics. While there’s always more to learn, these results suggest that CCTs may help to more fully integrate the poorest households into their communities.
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.