Agricultural Investment Behaviour and Contingency : Experimental Evidence from Uganda

Underinvestment in agriculture -- a major cause of rural poverty -- may be due to difficulties to detect 'contingency', defined as the influence one may exert on the distribution of returns on investment. Recently experienced contingency may create a mismatch between perceived and actual contingency, leading to sub-optimal investment behaviour. To test this, we use an experiment with poor farmers in Uganda used to low levels of contingency, as many factors (e.g., the weather, pests, price fluctuations) obscure the link between farm investment and outcomes. We find that in situations in which some contingency is present, investment levels respond positively to recently experienced contingency. In situations in which no contingency is present (`non-contingency'), time spent exploring a new investment opportunity responds negatively to recently experienced non-contingency. The findings that perceived contingency influences investment behaviour, and perceived contingency can be readily changed, may inform new behavioural policies to promote agricultural investment