How responsive are private transfers to income? Evidence from a laissez-faire economy

In recent years there has been burgeoning interest in how private transfers respond to household income, or "transfer derivatives". Much of it is fueled by the specter of "crowding out": the more responsive are private transfers, the more likely they could be supplanted by expansions in public transfers, weakening or destroying the latter's distributive impact. Yet there is also an emerging consensus that actual transfer derivatives are trivial, at least for the United States, suggesting that crowding out might loom larger in the minds of some economists than in the data. But perhaps crowding out is a fait accompli in the United States. Accordingly, we focus on the Philippines, which has a much smaller public sector. We also pursue a novel theoretical angle, which generates a sharp, non-linear relationship between private transfer receipts and income, in the form of a spline. Our empirical work applies recently developed estimation techniques that treat the spline's threshold as an unknown parameter, and they uncover transfer derivatives much larger than conventional methods reveal. Notably, private transfers patterns are consistent with altruistic preferences, effective risk sharing, or both. In accord with theoretical predictions, however, strong transfer derivatives prevail only among low-income households. Our findings suggest that attempts to aid the poor could be thwarted by private responses, which leak benefits to richer households in the form of lighter burdens of support for less fortunate kin. So the problems that operative private transfers create for public redistribution policy, first pointed out by Becker and Barro over 25-years-ago, do indeed matter empirically. (C) 2003 Elsevier B.V. All rights reserved.