When interest rates rise, fixed-rate mortgages generate a financial incentive for owners to keep their homes, creating “rate lock”. Does rate lock dampen the negative impact of rising interest rates on house prices? To estimate rate lock’s causal effect on market-level house prices, we instrument for the rate lock incentive in the outstanding local mortgage stock using unexpected family size shocks that induce moves at times with different mortgage rates. We find that when interest rates increased over 2021-23, a one standard deviation increase in the rate lock incentive, corresponding to a 0.3pp lower average outstanding mortgage rate, caused 2.6pp higher nominal house price growth. To understand the mechanism, we compare moves of owners who purchase homes just before and just after sharp mortgage rate increases. A 1pp lower outstanding mortgage rate reduces moves from owning to renting by 33%, a force increasing the price-to-rent ratio, and reduces overall moves by 42%. Using these estimated effects on mobility, we calibrate a dynamic structural model to quantify how much rate lock offsets the negative aggregate price effects of a higher cost of capital. Model simulations indicate that the 2021-23 tightening would have reduced the price-to-rent ratio by 12.3% with adjustable-rate mortgages, and hence no rate lock, versus only 4.4% with fixed-rate mortgages. Rate lock thus dampens, but does not fully offset, negative price effects of higher interest rates.

To study the impacts of debt relief versus cash transfers, I compare saving and consumption responses to student loan forbearance and stimulus checks in the 2020 CARES Act. Borrowers non-optimally use much of the liquidity received from for- bearance to voluntarily prepay 0%-interest student debt instead of high-interest obligations, despite prioritizing high-interest debts when receiving stimulus checks. Consistent with this flypaper effect, the marginal propensity to spend (MPX) out of forbearance liquidity is less than half that of stimulus checks. A calibration exercise estimates that the flypaper effect makes forbearance less effective and more costly as a countercyclical fiscal tool.

We present a new model of competition between digital media platforms with targeted advertising. The model adds new insights around how user heterogeneity and overlap, along with user and advertiser substitution patterns, determine equilibrium ad load. We apply the model to evaluate the proposed separation of Facebook and Instagram. We estimate structural parameters using evidence on diminishing returns to advertising from a new randomized experiment and information on user overlap, diversion ratios, and price elasticity from earlier experiments. In counterfactual simulations, a Facebook-Instagram separation increases ad loads, transferring surplus from platforms and users to advertisers, with limited total surplus effects.

We use parcel-level data to provide new facts on the level and distribution of land available for residential development, focusing on New England housing markets between 2007 and 2021. Most buildable parcels are small, with large buildable parcels scarce in most geographic markets. Large buildable parcels are less available in more populous markets, become more scarce as populations grow, and have become more scarce over time. Markets with fewer large parcels experience higher price growth and lower residential development relative to price growth. We present evidence consistent with developer returns to scale in parcel size, meaning that fragmentation of buildable land across small, disjoint parcels increases house prices by lowering construction productivity and making development less responsive to demand. In counterfactual simulations from a simple calibrated model, we show that recombining small buildable parcels into larger ones, while holding the total amount of buildable land fixed, would increase supply, raise construction productivity, and reduce house price growth.

This paper studies the impact of the emerging market for “buy now, pay later” (BNPL) installment loans on consumption patterns using detailed financial transactions data. For liquidity-constrained consumers, BNPL increases total spending and overdraft fees, but also facilitates greater expenditure smoothing. For all users, BNPL access boosts spending on retail goods, suggesting static substitution towards retail consumption. These inter-temporal and static substitution effects are much larger than a reasonably-calibrated lifecycle model with liquidity constraints can capture. Our findings are more consistent with a “liquidity flypaper effect” whereby the additional retail liquidity from BNPL “sticks where it hits,” increasing short-term spending.

We study how lender responses to consumer biases affect credit card debt. Using account-level credit bureau data, we document a new fact: many consumers make near-minimum credit card repayments while also overpaying lower-interest installment debt. We estimate an empirical model in which lenders optimally set minimum payments given this behavior. Because some liquid consumers make near-minimum payments, lenders lower minimums to slow repayment, increasing high-interest debt balances and interest revenue. Our model implies that this lender response increases revolving debt by $46 billion. This finding illustrates how optimizing lenders can amplify the effects of consumer biases in equilibrium.

This paper uses mutual fund manager data to examine how managers' performance incentives generated speculative demand during the 2020-2022 cryptocurrency boom and bust. We find that managers with strong relative performance incentives began investing in crypto after their competitors began investing in it, consistent with a model of rational performance hedging. In contrast, managers who invest their personal wealth in the funds that they manage, who have strong direct performance incentives, were significantly less responsive to their competitors' investment decisions. Our findings suggest that relative performance incentives can encourage managers to mimic their competitors instead of trading on their beliefs. In equilibrium, this competitive hedging motive can magnify the scope of speculative demand.

State governments use production subsidies to attract companies and facilitate economic development. Such programs benefit state residents by increasing local labor demand but may also encourage firms to pursue suboptimal production strategies. To assess the net welfare impact of these competing effects, I develop a general equilibrium framework with multiregional production, rich firm heterogeneity, and production subsidies that vary across states and between firms. Eliminating subsidies would increase total US welfare by 1.1%, despite costs to peripheral states in the Deep South and Northwest.