Information technologies can only partially ameliorate the low take-up of social benefits among forcibly displaced households in Colombia

Keyword: 
Poverty
Topic: 
Poverty - Inequality - Aid Effectiveness
Fiscal Policy - Public and Welfare Economics

One major puzzle in developing economics is that non-take-up rates of social benefits are usually very large. Examples of this disturbing fact range from farmers who resist using fertilizers that are available at very low prices, to parents who opt out from giving their children free de-worming pills that will substantially improve their health. There are dozens of similar examples both in developing and developed countries.

There are various reasons that may explain the puzzle. If people feel ashamed of being identified as needy or disadvantaged after receiving some benefit this may be enough to discourage them from claiming it. Another possibility is that there is limited information about the existence of benefits, or who are the beneficiaries, or the actual value that using the benefit represents for the beneficiary. It may also be the case that the process of obtaining benefits is too costly. This may happen for instance if there is cumbersome paperwork involved in the claiming process or if the bureaucrats in charge are corrupt or engage in red tape.

Little is known about which of these mechanisms is empirically more salient in explaining the high non-take-up rates. The first hypothesis is unlikely to have a lot of leverage as the literature has noted that non-take-up rates are equally prevalent in means-tested and non-means-tested programs. The latter types of programs should be stigma-free, as they do not need any material shortages of potential beneficiaries to be revealed. However, it is difficult to disentangle the relative importance of the imperfect information and the transaction costs hypotheses.

Randomized controlled trials (RCTs) are a powerful tool in this set up as they allow researchers to manipulate one at a time the factor that, in theory, influence non-take-up rates. For instance, by providing timely information on benefits eligibility, one can assess the extent to which low take-up rates depend on the lack of eligibility awareness of beneficiaries vis-a-vis the costs involved in the process of obtaining the benefits. This is precisely the type of RCT that we recently designed and implemented among Colombian internal refugees. But before turning to the specifics of the design and the results from our experiment, let us provide a little context on the problem that motivated this project.

Colombia has hosted an internal conflict for the last five decades. While there is currently an ongoing peace process with the largest rebel insurgency (FARC), the conflict has had devastating economic and social consequences. See for instance two recent blog posts discussing some of the costs that the conflict has produced in Colombia, here  and here. One such consequence is violence-driven internal migration, which is, according to the UN “the biggest humanitarian crisis in Western Hemisphere". Estimates of the stock of internal refugees vary from about 4 million to almost double that figure.

While internally displaced households are legally entitled to obtain from the government a large set of benefits, by 2009 about two thirds of them had not claimed any. This is puzzling as forcibly displaced households are among the most vulnerable people in Colombia, falling systematically below the poverty line much more than any other vulnerable community.

Our RCT was designed to study the extent to which providing timely information to displaced households about their eligibility could improve benefits take-up rates. Using text messages, we communicated to a random half of the benefit-entitled households that they were eligible. At the end of the implementation we conducted personal surveys with treated and control household heads to find out about their eligibility awareness as well as their actual take-up of benefits. Sending text messages is very cheap, and the relevance of the strategy in this context is indisputable. A nationally representative survey of displaced households, carried out in 2008 by NGO CODHES, suggests that the penetration of cell phones among this community is almost 100%.

This contrasts sharply with the status quo process, in which there is no direct communication from the government and thus gathering information about eligibility is the sole responsibility of internal refugees: Upon arrival to their new destination, forcibly displaced households had to apply for inclusion in the unique registry of displaced population, which constituted the list of officially accepted displaced people giving them access to humanitarian assistance and other longer term benefits. To avoid including individuals who would lie about having been displaced to obtain someone else's benefits, the government investigates the applications. In order to receive updates about the status of the inclusion process, applicants must go to specific government offices in a process that turns out to be very costly for them. Costs include transportation expenses, long waiting lines, and the forgone daily income of the household head. Applicants are often forced to repeat the process several times as they are usually told that inclusion in the registry is still pending approval. Moreover, inclusion is no guarantee the flow of benefits. Take-up is largely a demand-driven process that is characterized by similar transaction costs.

Thus, both informational barriers and bureaucratic obstacles are likely to be important drivers of the low take-up of benefits in this context. In our experiment, households assigned to the treatment group are informed about their inclusion in the registry as soon as the government approves it. They are also encouraged to claim their benefits. In contrast, households assigned to the control group must follow the regular procedure, as described above. Because there are no official records of the actual take-up of benefits by eligible refugees, we conducted a follow-up survey to gather such information. In addition to asking about benefits take-up, we gathered a large battery of demographic characteristics and pre-treatment outcomes, which allow us to check that the two groups are sufficiently similar in a variety of aspects other than having received the text message.

Leaving the technical details of the estimation of the effect of the text message on the take-up of benefits to the paper, let us briefly discuss the main result of our intervention. We find evidence that receiving timely information about benefits eligibility increases the take-up of short-term emergency humanitarian benefits by 12 percentage points. This is mainly driven by a 15 percentage point's increase in the take-up of medical care services. The take-up of other benefits like housing support and food assistance do not seem to be significantly increased, perhaps because of our small sample size, an issue that we discuss at length in the paper.

These results are consistent with previous experimental research that shows that the scope of information is limited when other bottlenecks affect access. That is, the that lack of a transparent and accessible information strategy for the displaced households in Colombia only explains part of the low benefit take-up, and the bureaucratic obstacles involved in claiming these should be addressed to alleviate the precarious situation of the country’s conflict driven refugees.
 

Share this