Working Paper 243 - Selling crops early to pay for school: A large-scale natural experiment in Malawi

In this paper we use a natural experiment from Malawi to test the hypothesis that short - term expenditure needs force poor households to sell crops early, when output prices are well below their peak. The experiment comes from a change in the timing of the primary school calendar. In 2010, the school year began in September, three months earlier than in 2009. Although there is no primary school tuition in Malawi, households still incur substantial out -of-pocket school costs. We use difference – in - difference and triple difference specifications to show that the cumulative value of household- level crop sales made before September was significantly higher in 2010 than in 2009. This effect is limited to households with school-aged children, is increasing in the number of school-aged children, and is only present for households in poverty for whom the opportunity cost of liquidity is higher. Because crop prices rise substantially over the last quarter of the calendar year, back - of-the-envelope estimates indicate that poor households paid a per - child penalty of 366 - 1221 Malawi kwacha (2.5-8.5 USD) to finance school expenses three months earlier in 2010 than in 2009. More broadly, these findings highlight the high cost of liquidity for poor households, and provide empirical support for the concern that intra - annual price volatility exacerbates the negative impacts of liquidity constraints.