Rema Hanna and Benjamin Olken (2018) explore the trade-offs associated with universal basic income and targeted income transfers as a means for developing countries to fight poverty. Their findings suggest that although universal basic income helps to overcome targeting errors, targeted programmes in developing countries allow for a bigger transfer on a per-beneficiary basis and hence are more effective at improving social welfare.
Background
In the fight against poverty, many governments around the world resort to income transfer programmes that aim to target poor populations. However, a key challenge with this mechanism is that, typically, in developing countries, a large proportion of the poor work in the informal sector, which prevents governments from observing their income and identifying those who are eligible for the transfer. An alternative solution is to use proxy income measures, however, these lead to targeting errors such as giving the transfer to those who are not poor (Inclusion error) or failing to provide the transfer to poor individuals (Exclusion error).
A way to overcome targeting errors is to apply a universal basic income programme, where everyone is given equally a transfer. Since the government does not need to verify income, these programmes are relatively easier to implement and involve lower administrative costs. On the other hand, for most developing countries, these programmes need to be financed through domestic taxation, which again is challenging due to the high level of informality and a large number of people outside the tax net. Hence, to provide a transfer, the government might need to increase marginal tax rates substantially for those few inside the tax net.
Case Studies
The authors evaluate the social welfare implications of implementing these types of programmes by using nation-wide household data and simulating different eligibility thresholds in Indonesia and Peru. Both countries run various transfer programmes and also have a high percentage of their population working in the informal sector.
When examining the welfare effects of each programme, they assign different weights to exclusion and inclusion errors so that they reflect how important it is that the poor have access to the transfer and how much benefit there is if non-eligible individuals also get to enjoy the transfer. They also consider that by allowing more people into the programme, if the total budget for the programme is fixed, then the transfer given per person falls. Finally, the comparison also considers the savings in administrative costs that arise from not having to target individuals in a universal basic income scenario.
Overall, when comparing utility levels between the two programmes, they find out that narrowly targeted programmes, where the government distributes large transfers per person to the poorest of the poor, are optimal from a social welfare perspective.
Conclusion
While the paper concludes that despite targeting errors, targeted programmes might be a better option, it also argues that the results do not consider dynamic changes in household income. In fact, if survey data is not collected frequently, then identifying poor households becomes a more challenging task and therefore the size of the relative inclusion and exclusion errors might increase, which might lead to a different optimal threshold.
Finally, the authors also argue that the choice of programme depends principally on the policy context. A programme might receive better political support and funding if everyone gets to benefit directly from it. Equally, a targeted programme might be harder to implement if the proxy income measure for eligibility lacks transparency or if a country has high levels of corruption. Thus, governments should consider other factors, besides targeting errors, when choosing between the programmes.
Comments