Air Pollution

Paper Session

Sunday, Jan. 8, 2017 3:15 PM – 5:15 PM

Hyatt Regency Chicago, New Orleans
Hosted By: American Economic Association
  • Chair: Nicholas Muller, Middlebury College

Corrective Policy and Goodhart's Law: The Case of Carbon Emissions From Automobiles

Mathias Reynaert
,
Toulouse School of Economics
James M. Sallee
,
University of California-Berkeley

Abstract

Firms sometimes comply with externality-correcting policies by gaming the measure that determines
policy. We show theoretically that such gaming can benefit consumers, even when it
induces them to make mistakes, because gaming leads to lower prices by reducing costs. We use
our insights to quantify the welfare effect of gaming in fuel-consumption ratings for automobiles,
which we show increased sharply following aggressive policy reforms. We estimate a structural
model of the car market and derive empirical analogs of the price effects and choice distortions
identified by theory. We find that price effects outweigh distortions; on net, consumers benefit
from gaming.

Air Pollution and Labor Supply: Evidence From Social Security Data

Felix Holub
,
University of Mannheim
Laura Hospido
,
Bank of Spain
Ulrich Wagner
,
University of Mannheim

Abstract

How does environmental quality impact on human capital? A sizable empirical literature examines the health effects of air pollution, but much less is known about outcomes such as worker productivity and labor supply. A fundamental challenge in estimating the short-run impact of air pollution on labor supply arises from unobserved economic shocks that shift air pollution and labor demand simultaneously and thus induce bias in the estimated relationship between air pollution and labor supply.<br /><br /><br /><br />
<br /><br /><br /><br />
Our paper exploits administrative data from the Spanish social security system to circumvent this problem. Specifically, we estimate the causal impact of air pollution on sick leaves taken by workers with full-time employment contracts. The terms of these contracts are unlikely to respond to short-term economic fluctuations because they are largely determined by a highly rigid collective bargaining process. We combine daily information on sick leaves in a representative panel of 250,000 employees with daily measures of air quality and weather in their respective places of residence. Controlling for worker fixed effects helps us to mitigate sorting bias that might be present if individuals with bad health move to less polluted places.

Our econometric model relates individual sick leaves to air quality in the municipality of residence, controlling for weather and seasonal effects. Using rich information on personal characteristics in the data set, we explore the heterogeneity of the treatment effect with respect to the individual’s gender, age, income, and economic sector of the employer. For some pollutants, our estimator exploits exogenous fluctuations in naturally occurring emissions across time and space. Because social security covers more than 95% of employees in Spain, our estimates are close to the population effect.

Rare Events and Risk Perception: Evidence From the Fukushima Accident

Renaud Coulomb
,
University of Melbourne
Yanos Zylberberg
,
University of Bristol

Abstract

We study changes in nuclear-risk perception following the Fukushima nuclear accident of March 2011. Using an exhaustive registry of individual housing transactions in England and Wales between 2007 and 2014, we implement a difference-in-difference strategy and compare housing prices in at-risk areas to areas further away from nuclear sites before and after Fukushima incident. We find a persistent price malus of about 3.5% in response to the Fukushima accident for properties close to nuclear plants. We show evidence that this price malus can be interpreted as a change in nuclear-risk perception. In addition, the price decrease is much larger for high-value properties within neighborhoods, and deprived zones in at-risk areas are more responsive to the accident. We discuss various theoretical channels that could explain these results.

What Are the Long-Term Effects of Prenatal Air Pollution Exposure? Evidence From the BHPS

Patrick Gourley
,
University of Colorado

Abstract

The detrimental impacts of air pollution on human health are significant. High levels of pollutants can increase mortality in the elderly, reduce worker productivity, and decrease birth weights. While a consensus is beginning to emerge on the contemporaneous effects of air pollution on health and productivity, the long-term effects are still largely unknown. Because of data difficulties and omitted variable bias, it is usually troublesome to determine how air pollution from years or decades earlier will affect current economic outcomes. Restricted data from the British Panel Household Survey (BHPS), however, includes narrowly defined current residence and location of birth. Linking over 11,000 responses from the BHPS to historical pollution data allows me to isolate the impact of prenatal particulate matter exposure on adult outcomes. I find that those with higher levels of exposure are more likely to be disabled and have lower wages. One standard deviation increase in average particulate matter exposure during gestation corresponds to an additional 700,000 members of the labor force that will file for disability across the United Kingdom. There is also some evidence that the second trimester of gestation is when the fetus is the most vulnerable to exposure and that there is a lower threshold that must be met for particulate matter to have an impact. This paper gives some of the first evidence that prenatal air pollution exposure affects adult outcomes decades later.

Flexible Fuel Vehicles, Less Flexible Minded Consumers: Price Salience Experiments at the Pump

Alberto Salvo
,
National University of Singapore

Abstract

This paper addresses a puzzle: why do many of Brazil's energy consumers driving "flexible fuel" gasoline-ethanol vehicles forgo energy savings at the pump, choosing the fuel that yields the lower mileage per dollar of spending, in some cases by a substantial margin? In a large-scale set of randomized experiments with 10,400 consumers at the pump---the first of its kind---I raise the salience of the price difference across both fuels, just as the consumer pulls up. The largest treatment effect I obtain is to shift one-tenth of consumers, who absent the intervention would have chosen expensive gasoline, to instead choose very favorably priced ethanol. While statistically significant, this shift is small compared with the higher likelihood that the favorably priced fuel is chosen among college-educated subjects relative to their less schooled counterparts. I estimate the increase in consumer welfare from mandating higher price salience at the pump to be equivalent to a 1 to 3% (general) reduction in fuel prices, depending on the relative price point.
JEL Classifications
  • Q5 - Environmental Economics