Working Papers
Tariffs, Global Value Chains, and the Incidence of Protection: Evidence from US Automobiles (with Luke Heeney and Christopher Knittel) (Link)
In many modern industries, firms compete in differentiated-product markets while relying on complex global value chains for intermediate inputs. In such settings, trade policies such as tariffs on vehicles and parts operate not only through consumer substitution and firm pricing, but also through firms’ cost structures and sourcing decisions. We develop a structural model of the U.S. automobile market that integrates random-coefficients demand, multiproduct firm pricing, and a flexible supply-side framework in which shocks to the cost of imported parts transmit imperfectly into manufacturers’ marginal costs. The model is disciplined by novel model-level data on imported-parts exposure and exploits exchange-rate variation to identify cost pass-through. Our counterfactual analysis quantifies the effects of alternative tariff policies on prices, profits, and welfare. First, tariffs on imported vehicles alone reallocate demand toward domestically assembled products and increase U.S. producer surplus, generating a gain of approximately $1 billion for U.S.-headquartered firms, while reducing consumer surplus by about $14 billion. Second, extending tariffs to imported intermediate inputs fundamentally alters these effects: consumer surplus losses roughly double, and producer surplus for U.S.-headquartered firms declines by about $2.6 billion. These aggregate effects mask substantial heterogeneity: firms with greater exposure to imported parts experience losses, whereas those relying more on domestic inputs are better able to increase profits. Overall, the results show that tariff incidence depends critically on firms’ exposure to global value chains and cannot be inferred from final assembly locations alone. We also find that reinstating EV subsidies generates sizable welfare gains, accompanied by an expansion in EV market share, total vehicle sales, and the number of vehicles assembled in the United States.
Pollution, Population, and Production: A Structural Analysis of Wildfire Smoke and Spatial Sorting (with Christopher Knittel) (Link)
Wildfires in the United States are becoming more frequent and severe, with California bearing the greatest burden. While prior research emphasizes direct damages and health effects, this study examines how wildfire incidents and their smoke influence household migration and firm performance. Exploiting temporal variation in wildfire and smoke exposure from 2011–2023, we combine household-level migration and demographic data with firm-level information on employment, revenue, and survival. Both hazards cause persistent population losses, with smoke having the larger effect: a one–standard-deviation increase in smoke exposure reduces household counts by 0.3–0.4% each year following exposure, whereas comparable wildfire exposure lowers them by about 0.2%. By contrast, smoke exposure has particularly strong effects on older households: those aged 50 and above experience declines exceeding 0.5% in several lags, while younger groups show small, short-run inflows, reflecting differences in preferences toward smoke exposure and likelihood of homeownership. These heterogeneous responses translate into meaningful shifts in neighborhood composition, altering the spatial sorting of households across affected areas. Smoke exposure also leads to sizable and lasting contractions in business activity, reducing firm counts by about one percent per standard deviation, with smaller firms disproportionately affected. To capture the general equilibrium implications of these shifts, we estimate a residential sorting model in which local amenities are endogenously supplied by firms operating within neighborhoods. We find that, on average, households are willing to pay 3.2% of property value to avoid such smoke exposure, rising to 4% among those aged 65 and above. The model highlights that older households are key drivers of local service demand—particularly in health and retail—while education services are least responsive to changes in the older population.
Disentangling Information Frictions and Endogenous Amenities in Residential Sorting Models: Evidence from Changes in Noise Pollution Exposure (Link)
This paper addresses two challenges to estimating the Marginal Willingness To Pay (MWTP) for residential exposure to noise pollution. First, ignoring information frictions among home buyers may bias estimates of the MWTP for an amenity. Second, changes in the amenity of interest may affect other endogenous amenities, further complicating the identification of MWTP. I address both challenges by using a novel instrumental variable based on spatial variation in the salience of noise pollution caused by hourly variation in flight paths due to wind direction. My preliminary results suggest disregarding information friction or endogenous amenities causes nontrivial biases in MWTP estimates. I find the MWTP to avoid 1 dB of noise pollution is approximately $4,742. I find that 80% of households exhibit a negative value of misinformation. The mean value of misinformation is -$118, with a total annualized value of -$2.2 million. Furthermore, households with a head of the family over 65 face significantly higher misinformation costs.
Valuing Noise Pollution in a Residential Sorting Model: Evidence from Changes to Flight Paths (Link)
Noise pollution from airplanes can reduce property values by creating a disamenity for residents. I estimate the effect of noise pollution on residential properties in the Phoenix metropolitan area, using quasi-random changes in commercial flight paths to and from Phoenix Sky Harbor Airport. Then, I estimate the parameters of a residential sorting model with heterogeneous preferences over noise and other amenities. The identifying variation comes from the unexpected implementation of new flight paths and instruments for time-varying housing prices. I find that the average marginal willingness to pay to avoid noise pollution is $3,038 per dB (decibel) and varies from $2,500 to $3,500 per dB, based on household demographics.
Valuing Noise Pollution: Evidence from Flight Path Changes (with Nicolai Kuminoff) (Link) Submitted
On September 18, 2014, the Federal Aviation Administration implemented new arrival and departure paths to Phoenix Sky Harbor International Airport. Overnight, aircraft noise increased in some residential neighborhoods and decreased in others. These changes occurred with no advance warning. We leverage this shock to provide causal evidence on the value of noise pollution. Our hedonic design uses data on houses that were sold both before and after the shock to identify households’ willingness to pay—through housing prices— to reduce noise pollution. We find robust evidence that a one decibel increase in noise caused property values to decline by 1.2%.
Overhead Costs: The Impact of Airplane Noise on Academic Achievement (with Alexander Toy) [Awaiting Permission to Release]
Environmental noise is a pervasive and understudied input in the education production function, with potential implications for academic achievement and equity. This paper examines the impact of airport noise on student achievement by exploiting quasi-random changes in flight paths around Phoenix Sky Harbor Airport. We assemble student-level English-Language Arts (ELA) and math test scores for grades 4–8 from 2012 to 2017 and link them to school-level measures of ambient noise exposure before and after the FAA’s 2015 route realignment. Using a difference-in-differences design we find that a 1 dB increase in average noise reduces both ELA and math performance 1.13% and 1.01% of a standard deviation respectively. These results inform policy debates on school siting, sound-proofing investments, and flight-path management to mitigate academic losses.
Work in Progress
Estimation of Electricity Access and Demand in the Republic of Yemen
The conflict in Yemen in 2019 significantly reduced access to reliable electricity. This study surveyed 1,052 Yemeni consumers to assess their electricity access and preferences for improvements. A stated choice experiment with randomized attributes was employed to estimate household Willingness to Pay (WTP) for various measures of electricity access, including consumption level and daily availability. The survey revealed that 82%, 18%, and 14% of households have access to electricity from solar, private grids, and national grids, respectively. Consumers are willing to pay $1.80 monthly for an additional hour of electricity availability. Moreover, they expressed a willingness to pay $6.50 for the capacity to power small, low-consuming appliances such as televisions, washing machines, and refrigerators and $16.60 for both low-consuming and high-consuming appliances, such as air conditioners or heaters. Further, consumers are willing to pay $13.40 to avoid losing access to electricity apart from the basic lighting and phone charging.
Others
"Specifications and Guidelines for Continuous Emissions Monitoring System (CEMS) for PM Measurement in Emissions Trading Scheme" published by Central Pollution Control Board, India (2013-14). Link
"Risk-Return Analysis of Sectorial Portfolio of Stock", Published at Economics, Management and Financial Markets (Addleton Academic Publishers, New York) (2012). Link