Old Age Pensions: SOUTH AFRICA Flashcards
Author of the paper?
Duflo (2013).
Evidence of larger improvements in child health when women receive income or assets:
Progresa program in MEXICO. Payments were made to women and it showed to have significant effects on children’s health, nutrition and education.
The experiment:
SOUTH AFRICA (1993). Universal and noncontributory (UCT). All women older than 60 and all men older than 65 were eligible for benefits, subject to a means test. Households were randomly surveyed and the height and weight of children (ages 6-60 months) were measured.
How many African children <5 years old live with a grandparent?
More than 25%.
Effect of pension on weight for height (short-run nutrition and health) on GIRLS:
Having a woman eligible increases W4H by 0.6 std devs.
Having a man eligible increases W4H by an insignificant 0.056 std devs.
Effect of pension on weight for height (short-run nutrition and health) on BOYS:
Having a woman eligible increases W4H by an insignificant 0.28 std devs.
Having a man eligible has a negative and insignificant effect.
Why use 2SLS estimates?
To reflect the fact that men claim the pension less often than women.
2SLS estimates of W4H for GIRLS:
Women’s pension increases W4H by 1.19 std devs.
Men’s pensions have a small, negative and insignificant effect.
2SLS estimates of W4H for BOYS:
Women’s pension increases W4H by an insignificant 0.58 std devs.
Men’s pensions have an insignificant and negative effect.
Effect of pension on height for age (long-term determinant of health and nutrition) on (young) GIRLS:
Having an eligible woman in the household increases H4A by 0.71 std devs.
Having an eligible man has a small and negative effect.
Effect of pension on height for age (long-term determinant of health and nutrition) on BOYS:
Having an eligible woman increases H4A by an insignificant 0.18 std devs.
Having an eligible man has an insignificant and negative effect on H4A.
2SLS estimates of H4A for (young) GIRLS:
Pensions received by women increase H4A by at least 1.16 std devs.
Pensions received by men had no effect.
2SLS estimates of H4A for BOYS:
Pensions received by women increase H4A by 0.28 std devs.
Pensions received by men hd no effect.
What implications does this have for public policy?
Maybe target public transfers to women rather than to men.
Why might grandmothers have a stronger incentive to invest in the children?
They will benefit from them longer. Female recipients are younger and live longer.
Why do we look at young girls when measuring H4A?
The effect on H4A for older girls is less because height is a long term determinant of health and older girls will have been exposed to the program for a shorter amount of time.
