Cash transfer programmes and conditionality: why not a lump-sum payment trusting the poor?

Most developing countries have introduced large-scale programmes of direct transfers with the intention of reducing poverty and vulnerability. Most of these programmes deliver monetary resources to poor households, provided that they comply with certain conditions associated with improving the human capabilities of their children. The dominant political discourse has been that it is necessary to help the poor in decision-making in order to break the inter-generational transmission of poverty. However, the great popularity that conditional cash transfer programmes have achieved contrasts with the lack of empirical evidence demonstrating the specific contribution of conditionality. It seems that policymakers have preferred to focus on the practical benefits of conditions, rather than worrying about their actual effectiveness. This research seeks to analyse critically conditionality in antipoverty programmes, which I consider important for the future development of improved forms of social assistance in developing countries. More specifically, it compares the effectiveness (in terms of welfare generation, human capital investment and labour supply) of two completely different cash transfer programmes offered in Ecuador: one with the typical conditional structure (Bono de Desarrollo Humano) and a recently implemented programme in which households receive a lump-sum payment without attached conditionalities (Crédito de Desarrollo Humano). For this purpose, I use different quasi-experimental methods together with two specific impact evaluation designs, which represent a significant methodological contribution to an evaluative literature where experimental methods have predominated. The analysis is based on linear, probit and ordered logistic regression models. Theoretical concepts pertaining to social protection and social assistance in developing countries, as well as the different theoretical perspectives about cash transfers, and the arguments for the inclusion and exclusion of conditions, frame the research questions and the interpretation of results. This research draws on the household- and individual-level data from the Ecuadorian Registro Social 2008 and 2014 databases, which compile surveys applied to families in areas with a high incidence of poverty. The empirical analysis shows that the non-conditional programme Crédito de Desarrollo Humano (Human Development Credit) has consistently superior effects on overall welfare, the average educational attainment of the households and the housing conditions, when compared to the conditional programme Bono de Desarrollo Humano (Human Development Bonus). Surprisingly, the conditional programme has no effects (or even small negative effects) on these outcome variables. This points to the low importance of conditionality in practice, which seem to play a secondary (or even harmful) role in generating welfare among the poor, in comparison with other constituent elements of the programmes (such as the amount of transfers and complementary policies). Therefore, the evidence clearly suggests that the presence of conditionalities in social assistance is neither necessary nor sufficient to ensure that families overcome poverty in a sustainable manner reaching higher levels of well-being, increasing their human capital and improving their living conditions. Moreover, neither programme has a significant impact on the probability of school enrolment/attendance of the first-born child. Although some statistically significant effects were estimated, their magnitudes are very small. Therefore, the findings indicate that conditionality is not sufficient to guarantee that parents invest in the human capital of their older children. Probably, conditions should be accompanied by constant awareness campaigns or better control and verification mechanisms. However, conditionality seems to be necessary for this purpose since the unconditional cash transfers had no effect on the parent’s behaviour, which is by itself an interesting and important result, since there is the idea that this programme –which promotes the generation of productive enterprises– could be causing at the same time an increase in child labour. Finally, neither of the two social assistance programmes negatively affects the probability of being unemployed. Both the beneficiaries of the Bono de Desarrollo Humano programme and the beneficiaries of the Crédito de Desarrollo Humano are equally likely to be unemployed than poor individuals who do not receive any kind of benefit. At least in this case, welfare programmes are not generating disincentives to work among those in poverty. Overall, the analysis suggests that cash transfer programmes can achieve better results if they leave conditionalities aside and opt for the productive inclusion of the poor by linking cash transfers to interventions/policies that aim to directly increase household productivity. Through elements such as mutual trust and training, more effective and sustained escapes from poverty might be achieved