Anand Mathew

Job Market Paper

Credit Where Credit Is Withdrawn: Weather Shocks and the Household Coping Portfolio in India

with Sankalp Mathur (CAFRAL)

In September 2018 the default of Infrastructure Leasing and Financial Services, an infrastructure financier rated AAA weeks earlier, cut the wholesale funding of India's non-bank lenders. The squeeze was hardest in the districts that relied on them. We use it to ask what a contraction in formal credit does to the way households absorb a dry shock. We follow 210,560 households from 2014 to 2019 and set them beside a novel dataset of 483 million loan records from the credit bureau, aggregated to the same districts. A dry shock takes the same share of income whether exposure to the contraction was higher or lower. Total spending barely moves in either case, but the composition of the basket differs. Where exposure was lower, the cuts begin with recreation and restaurant meals and stop before health, education and a staple such as milk. Where it was higher, the same shock reaches all three. Fewer households hold a non-bank loan, and among those who lose formal access, about as many stop borrowing as keep or find an informal lender. Few households borrow from a non-bank directly, so the effect runs through the district's credit market. Credit does not stop a dry shock from being felt. It determines how far down the coping portfolio a household has to go.

Earlier versions presented at the World Inequality Conference, Paris School of Economics (June 2026) and the WATE-Florida Workshop (October 2025). To be presented at the 96th Annual Meeting of the Southern Economic Association, Houston (November 2026).

Working Papers

Cannabis Access and College Persistence: Evidence from Washington State's Dispensary Lottery

Single-authored

Abstract

This paper examines how proximity to recreational marijuana dispensaries affects postsecondary educational outcomes in Washington State. Leveraging the quasi-random placement of dispensaries via a state-run license lottery and a staggered difference-in-differences framework, I estimate the causal impact of dispensary access on college enrollment and retention rates. Using campus-level administrative data from the Integrated Postsecondary Education Data System (IPEDS) and detailed dispensary location data, I find that institutions located within a 10-minute driving radius of a dispensary experienced a statistically significant 7.9 percentage point decline in student retention rates. In contrast, there is no evidence that dispensary proximity affects college enrollment. Robustness checks across alternative distance thresholds confirm that the retention effect is stable and significant for thresholds between 10 and 15 minutes, while enrollment results remain null. These findings suggest that while local cannabis availability may not deter students from enrolling, it can negatively affect persistence among those already enrolled.

Presented at the 95th Annual Meeting of the Southern Economic Association, Tampa (November 2025).