In village economies, limited commitment models have been proposed as theoretical alternatives to the complete markets hypothesis and have been shown to be empirically relevant. Yet, village institutions might be able to enforce some contracts. We construct a theoretical model which shows how households can insure through both formal and informal contracts when some verifiable production takes place in an environment of incomplete markets. This theoretical setting nests the case of complete markets when all risks can be insured by formal contracts (because all states of nature would be verifiable) and the case where only informal agreements are available (agreements specifying informal transfers that needs to be self-enforceable). We derive two equations of interest, the income equation which is partially determined by the formal contract and an Euler-type equation where consumption growth is affected by lagged consumption instead of being a martingale. Using semi-parametric specifications, we derive testable restrictions of our model. We estimate both equations using data of village economies in Pakistan. Empirical results are consistent with the model. The estimation and allows to show that the incentive constraints due to selfenforcement bind and that formal contracts are used to reduce the probability of binding the constraint.