Geary Institute Working Papers

Many countries provide extensive in-kind public transfers for specific needs of particular client groups such as the elderly, the disabled, and children. However, this may crowd out private expenditures on the goods in question and, to some extent, undermine the case for not simply giving cash. If the target group belongs to a larger household the mechanism behind this crowding out could be either altruism or agency. This paper is concerned with three nutrition programmes for children in UK households: free lunch at school for children from poor households; free milk to poor households with pre-school children; and free milk at day-care for pre-school children in attendance regardless of parental income. We exploit a reform that removed eligibility to the first two programs from working poor households. We find significant crowding-out of private food expenditures – a free school lunch reduces food expenditure by around 15% of the purchase price of the lunch, and a free pint of milk reduces milk expenditure by about 80% of the market price. We conclude that this is due to altruism rather than agency problems because milk expenditure crowd-out is similar across milk programs that have different delivery mechanisms.

Governments, over much of the developed world, make significant financial transfers to parents with dependent children. For example, in the US the recently introduced Child Tax Credit (CTC), which goes to almost all children, costs almost $1billion each week, or about 0.4% of GNP. The UK has even more generous transfers and spends about $25 a week on each of about 8 million children – about 1% of GNP. The typical rationale given for these transfers is that they are good for our children and here we investigate the effect of such transfers on household spending patterns. The UK is an excellent laboratory to address this issue because such transfers, known as Child Benefit (CB) have been simple lump sum universal payments for a continuous period of more than 20 years. We do indeed find that CB is spent differently from other income – paradoxically, it appears to be spent disproportionately on adult-assignable goods. In fact we estimate that as much as half of a marginal pound of CB is spent on alcohol. We resolve this puzzle by showing that the effect is confined to unanticipated variation in CB so we infer that parents are sufficiently altruistic towards their children that they completely insure them against shocks.