A new paper, presented at a workshop in Kampala, shows potentially very significant Rates of Return for Uganda’s Senior Citizen’s Grant. Depending on the specific assumptions, the SCG, which is an old age pension provided on a universal basis in selected districts, could pay for itself 1.4 times over within a 10 year period, if scaled up across the country – through its effects on education alone.
The paper is the result of a collaboration between researchers from Maastricht University, Manchester University and Makerere University as well as the Government of Uganda. It adds to the evidence on the economic impact of the Senior Citizen’s Grant as Ugandan MPs outline the process of building the political will at an event in London later this month.
The new paper, ‘Social protection investments, human capital, and income growth: Simulating the returns to social cash transfers in Uganda’, assesses the short‐ and mid‐term effects of two cash transfer programmes: the Senior Citizen’s Grant and the Vulnerable Family Grant, if they were scaled up to cover the entire country. The latter programme was a household transfer targeted at vulnerable families, which was discontinued after the pilot phase, while the Senior Citizens’ Grant was scaled up.
The results of the analysis show first of all that the Senior Citizen’s Grant has a much higher Rate of Return than the poverty targeted Vulnerable Family Grant, suggesting that the Government of Uganda made the right choice in scaling up the former and not the latter. This difference is mainly due to the much higher coverage rate of the Senior Citizen’s Grant: 60% of households have at least one member above 65 and would therefore receive the SCG, while the VFG only reached 15% of households. The much lower administrative costs of the universal programme also contributes to its higher rate of return.
The paper uses panel data from the Uganda National Panel Survey (UNPS) to estimate linkages between income, child health and school attainment, and subsequently uses this data to carry out a micro-simulation to predict the income rates of return of the two programmes, through their effects on education. Figure 1 shows the analytical framework linking the Senior Citizen’s Grant with household income, improved education, and then how this leads to increased income through the effect of improved education.
The analysis shows positive, but not very large, effects of the transfers on education. However, the effects increase significantly over time, and they are large enough that the Senior Citizen’s Grant pays for itself within a period of 13 years. As the authors note, it is important to be aware that this assessment only considers the monetary value generated through improved education. There are many other potential channels through which the Senior Citizen’s Grant is likely to generate value, for example through improved adult health and increased agricultural investments, impacts that are not accounted for in the analysis.
In addition, another paper by the same group of researchers, builds on, and qualifies these results, by examining how the rate of return changes as the assumptions behind the analysis are changed. In particular, it notes that a cash transfer will have higher value for low-income households than for high-income households. By applying what the authors call ‘prioritarian welfare weights’ they show that the real rate of return to the programmes is likely to be much higher, when the actual social value of the grant for low-income households is considered. Figure 2 shows the conceptual difference between the ‘utilitarian welfare function’ assumed in the original analysis, and the more realistic ‘prioritarian welfare weights.’
As Figure 3 below shows, the rate of return of the Senior Citizen Grant changes significantly when the new welfare weights are applied. It is now very likely to pay for itself after just 10 years, with benefits worth as much as 1.4 times the costs within the period, depending on the welfare weight applied. Again, this is only through its effects on education, ignoring all the other positive effects of the programme. Note that the figure, from Franziska Gassman’s presentation in Kampala, is based on an assumption of the propensity to consume 100% of income, while the analysis in the original paper assumes 80% is consumed. In effect, using 80% propensity to consume means that the simulation examines the effect of only 80% of the value of the grant. Given that the benefit level of the SCG is quite low, it is arguably more realistic to assume that 100% of the grant is consumed.