Policies need not only to be well designed to effectively address market failures, but their parameters also need to be part of agents’ information sets. This is illustrated by government student loans in the Netherlands which are intended to alleviate liquidity constraints. Despite generous loan conditions, take-up rates on these loans are low. Some have argued that this is due to limited knowledge about these conditions. We examine the importance of information constraints through a randomized experiment. Half of the students who responded to an Internet questionnaire were given factual information on loan conditions, whereas the other half did not receive such information. Six months later, students who received information have better knowledge about the loan conditions. While OLS regressions reveal a large and significantly positive association between knowledge about loan conditions and borrowing, our instrumental variable estimates suggest that this is not a causal effect which would rule out that the low take-up rate is caused by information constraints. Classification-JEL: I22; I28; D83 Keywords: Field experiment; Student debt; Student loans; Loan conditions File-Url: https://papers.tinbergen.nl/08039.pdf File-Format: application/pdf File-Size: 205290 bytes Handle: RePEc:tin:wpaper:20080039