Friday, 4 April 2014

Cash is the null

Straight cash transfers should always be the null against which in-kind transfers get assessed.

More evidence that the null's pretty decent comes today from Cunha in the latest AEJ: Applied Economics. A randomised control trial of cash versus in-kind food benefits in Mexico's food assistance programme showed that the in-kind transfers didn't beat the null. Here's Cunha [ungated]:
Welfare programs are often implemented in-kind to promote outcomes that might not be realized under cash transfers. This paper tests whether such paternalistically motivated transfers are justified compared to cash, using a randomized controlled trial of Mexico's food assistance program. In relation to total food consumption, the in-kind transfer was infra-marginal and nondistorting. however, the transfer contained ten food items, and there was large variation in the extent to which individual foods were extra-marginal and distorting. Small differences in the nutritional intake of women and children under in-kind transfers did not lead to meaningful differential improvements in health outcomes compared to cash.
While recipients of in-kind food transfers rather than cash did consume more vitamin C, iron, and zinc than did those on the cash transfer, this didn't result in any differential effects on health outcomes for the women and children monitored. Further, recipients of the cash transfers diverted little of the transfer towards alcohol, tobacco or less healthy foods. Since it's much more expensive to provide in-kind transfers than to provide cash transfers (an excess cost of a bit over 17% of the transferred amount), cash transfers win pretty handily.


  1. Interesting timing, just found this paper on Freakonomics today.

    Less Cash, Less Crime: Evidence from the Electronic Benefit Transfer Program


  3. all deals come down to cash presented