Conditional Cash Transfers in Uruguay
Introduction
Bergolo and Cruces (2021) compute the MVPF of a marginal increase in transfers through the Family Allowance Assistance program (AFAM), a conditional cash transfer program run by the Federal Government of Uruguay that targeted low-income households with children and/or pregnant women. Covering 42 percent of children under the age of 18 and costing 0.35 percent of GDP, AFAM is Uruguay’s largest social assistance program.
To receive transfers, households must apply and pass two tests: an income test and a proxy means test. The income test verifies if household per capita income falls in the bottom 20 percent of income distribution. The proxy means test verifies if the household’s poverty score — computed from a large set of socioeconomic variables — falls below a given threshold. Once enrolled, beneficiaries must to comply with the program’s health and educational requirements to continue receiving transfers.
The paper uses administrative data and exploits the program’s poverty score threshold to implement a difference-in-discontinuity design and identify AFAM’s causal impact on the labor supply of beneficiaries. The paper finds that the program reduces labor force participation by about 6 percentage points (a 13% drop) among all beneficiaries, and by about 8.7 percentage points (a 19% drop) for single mothers. The paper uses these results to estimate the MVPF of expanding the cash transfer program for single mothers.
MVPF = 0.6
Net Cost
The paper includes three components in the computation of the AFAM’s net costs to the government: the direct cost and two fiscal externalities related to the impact of AFAM on registered employment.
First, the direct cost of the transfer is given by the average monthly benefit of UYU 1,086.
Second, AFAM causes registered employment earnings to fall, which reduces tax revenue. The paper assumes that AFAM’s target population is only subject to the payroll tax, which is about 32% of earnings. AFAM beneficiaries’ income is exempt from income taxation up to the threshold amount for eligibility, and the paper assumes lost revenues through other tax basis (such as consumption and capital taxes) can be ignored. To compute the lost tax revenue due do reduced payroll taxes, the paper takes average gross earnings of UYU 9,597 of the target population (single mothers) and multiplies that by the payroll tax rate (32%) and by the change in employment caused by the program among single mothers (19.3%), for a total of UYU 9,597 x 0.32 x 0.193 = UYU 592.7.
Third, an expansion in AFAM transfers would attract more participants into the program, which would increase the total cost of the program. To estimate the lost revenue through this channel, the paper first estimates that 55.3% of single mothers induced not to work by AFAM would have earned incomes above the program eligibility threshold. Therefore, the extra government expenditure due to higher induced take-up is given by the average transfer multiplied by the fraction of the population who would not have received transfers but for the increased generosity of the program: UYU 1,086 x 0.193 x 0.553 = UYU 115.9, where 0.193 is the estimated fraction of single mothers who lose formal employment due to the program and UYU 1,086 are the direct transfer costs.
The total net cost to the government is therefore URU 1,086 + URU 592.7 + URU 115.9 = UYU 1,794.6. The paper notes that several potential sources of government costs have been excluded because they were unobserved, such as changes in take-up of other social programs and AFAM’s administrative costs.
Willingness to Pay
The paper assumes that enrollees value transfers through AFAM in a 1-to-1 fashion and that other aspects of the program – such as complying with conditionalities – do not affect the willingness to pay. Therefore, the willingness to pay is given by the average monthly transfer amount received by the relevant population (in this case, single mothers): UYU 1,086.
MVPF Summary
Dividing the willingness to pay of UYU 1,086 by the net cost to the government of UYU 1,794.6 the authors get an MVPF of 0.61.
Using alternative assumptions about the propensity for formal employment yields MVPF estimates 0.37, 0.67, and infinity.
References
Bergolo, Marcelo, and Guillermo Cruces. “The Anatomy of Behavioral Responses to Social Assistance When Informal Employment is High.” Journal of Public Economics 193 (2021): 104313. https://doi.org/10.1016/j.jpubeco.2020.104313
Policy Features
- Category
- Social Insurance
- Sub-Category
- Cash Transfers
- Beneficiary Type(s)
- Adults, Mothers and pregnant women, Parents
- Average Income
- 9597
- Country of Implementation
- Uruguay
- Year of Implementation
- 2009
- Empirical Method
- Regression Discontinuity
- Research Type
- Primary
- Peer Reviewed
- Yes
- MVPF Publication Link
- doi.org/10.1016/j.jpubeco.2020.104313