Are the poor so present-biased?

Estimates of "present-bias" among the poor may be exaggerated if poor individuals are credit-constrained and expect to have greater liquidity in the future. I conduct an experiment in rural Pakistan which provides causal evidence of this effect. I use windfalls to generate fully exogenous variation in subjects' liquidity constraints. I show that fluctuating liquidity has a significant and sizeable effect on measures of time-inconsistency, which does not operate via cognitive functioning. Importantly, I establish that the causation runs from tighter liquidity constraints to appearing "present-biased" rather than truly present-biased individuals making choices which lead to tighter liquidity constraints.