This project analyses whether there is any evidence of income uncertainty and adverse shocks affecting choices relating to the schooling and work of children in rural Colombia. We set up a simple theoretical framework to analyse this, and go on to test the predictions of the model empirically. We measure income uncertainty using original data that elicits from respondents their own perception of the variation in future income draws. Behavioural changes to this type of risk are "ex-ante", in the sense of taking place in response to a spectrum of possibilities which may or may not occur. We measure shocks from the realisation of adverse events such as deaths and illnesses of household members, and crop losses. Any responses to these events may be considered to be "ex-post" ways of dealing with adversities.

Sample: Approximately 11,500 very poor households in rural Colombia