Working Paper: NBER ID: w26645
Authors: Bronson Argyle; Taylor D. Nadauld; Christopher Palmer
Abstract: We establish two underappreciated facts about costly search. First, unless demand is perfectly inelastic, search frictions can result in significant deadweight loss by decreasing consumption. Second, whenever cross-price elasticities are non-zero, costly search in one market also affects quantities in other markets. As predicted by our model of search for credit under elastic demand, we show that search frictions in credit markets contribute to price dispersion, affect loan sizes, and decrease final-goods consumption. Using microdata from millions of auto-loan applications and originations not intermediated by car dealers, we isolate plausibly exogenous variation in interest rates due to institution-specific pricing rules that price risk with step functions. These within-lender discontinuities lead to substantial variation in the benefits of search across lenders and distort extensive- and intensive-margin loan and car choices differentially in high- versus low-search-cost areas. Our results demonstrate real effects of the costliness of shopping for credit and the continued importance of local bank branches for borrower outcomes even in the mobile-banking era. More broadly, we conclude that costly search affects consumption in both primary and complementary markets.
Keywords: Consumer Credit; Search Frictions; Price Dispersion; Durable Consumption
JEL Codes: D12; D83; E43; G21; L11
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| search frictions in consumer credit markets (G51) | significant deadweight loss (D61) |
| search frictions (F12) | decrease in consumption (E21) |
| costly search in one market affects quantities in other markets (D43) | price dispersion (L11) |
| search frictions (F12) | affect loan sizes (G51) |
| search frictions (F12) | decrease final goods consumption (E20) |
| borrowers assigned high interest rates due to search costs (G21) | purchase older cars (R48) |
| borrowers facing high markups (G51) | lower loan take-up rates (G51) |
| higher search costs (D83) | lower loan sizes (G51) |
| higher search costs (D83) | diminished durable consumption (E21) |