Air Pollution

“As the Wind Blows: The Effects of Long-Term Exposure to Air Pollution on Mortality,” by Michael L. Anderson. September 2015. NBER #21578.

What are the health consequences of exposure to conventional fossil fuel emissions? The U.S. Environmental Protection Agency sets ambient air quality standards using research on mortality rates in different metropolitan areas that have different ambient pollution levels. Such observational studies infer causality between air quality and mortality through statistical controls, and thus are subject to all the doubts that accompany statistical inference.

Just as medicine increasingly has relied on clinical trials rather than observational studies to understand the effects of drugs, diet, and behavior on human health, economists have increasingly published the results of natural experiments in which people are exposed to pollutants in a manner that is plausibly random and health effects are observed. In my Working Papers column in the Winter 2012–2013 issue, I described a paper that examined the summer cap-and-trade system for nitrogen oxide (NOx) emissions in the Eastern and Midwestern United States and its effects on ozone levels and asthma drug expenditures.

This paper compares the mortality rates of people who live upwind and downwind of interstate highways in the Los Angeles metropolitan area where the winds often blow in the same direction. Many studies have measured air quality near major highways. Ultra Fine Particles (UFP) and NOx plumes don’t decay to background levels until they get 600 meters downwind, while upwind levels decay to background levels within 100 meters. During “carmegeddon” in Los Angeles in July 2011, when Interstate 405 was shut down over a weekend and car emissions fell accordingly, UFP levels were 83 percent lower downwind but there was no change upwind.

Even with the smaller estimated health effect, Anderson finds that reducing the downwind extra exposure to NOx and ultra-fine particles would be cost-effective.

Given the persistent wind direction, wouldn’t people avoid downwind houses and thus make exposure non-random? The answer appears to be no. One reason for this is that people cannot perceive UFP and NOx readily with their senses. People can perceive conventional particulate matter and ozone with their senses, but those emissions decay to background levels within 100 meters downwind of highways. Anderson examines housing values and shows no difference in values upwind and downwind of highways, so there is no obvious evidence of non-random sorting of people into upwind and downwind houses. In the data in this paper, the median individual over 75 has lived at his current location for 25 years and 78 percent of them have lived at their current location for over 10 years.

Anderson finds that moving from the upwind to the downwind side of the road (within 600 meters) increases average NOx levels by 43 percent and mortality rates for those above age 75 years by 3.6 to 6.8 percent. This estimate is 60 percent smaller than one of the studies the EPA currently uses as a basis for its standard setting.

But even with the smaller estimated health effect, reducing the downwind extra exposure to zero in dense urban areas like Los Angeles would be cost-effective, according to Anderson. This would add 0.24 years to life expectancy at birth, or 372,000 total life-years in a sample of 1.55 million individuals. That equates to $37.2 billion in benefits if each life-year is valued at $100,000. There are 2.9 million cars in the Los Angeles basin. Using federal tax credits to spur replacement of all those cars with zero-emission electric vehicles would cost only $21.8 billion, passing the cost-benefit test.

CAFE Standards

“Do Consumers Recognize the Value of Fuel Economy? Evidence from Used Car Prices and Gasoline Price Fluctuations,” by James M. Sallee, Sarah West, and Wei Fan. July 2015. NBER #21441.

Do consumers appropriately evaluate future energy costs in their decisions to purchase capital goods? That is, are they willing to pay up to $1 today for an investment that saves $1 (present value) in future energy expenses?

Advocates of the Corporate Average Fuel Economy (CAFE) standard, which mandates reduced energy use by automobiles, argue that consumers undervalue future energy savings in their decisions to purchase cars and trucks. Exactly 10 years ago in these pages, Molly Espey argued to the contrary: consumers seemed to value fuel economy appropriately when purchasing 2001 model-year cars (“Do Consumers Value Fuel Economy?” Winter 2005–2006).

This paper examines a large sample of vehicles sold from July 1993 through June 2008. The authors compared the sale prices of identical cars with different odometer readings and thus different total future operating costs. They tracked those prices month-by-month, along with changes in fuel prices, which altered the cars’ future operating costs. The researchers compared the car prices to determine if consumers changed their valuation of otherwise identical cars because of the change in future fuel costs.

Consumers appeared to conduct full valuation of gasoline costs over the remaining lifetime of the vehicle. A $1 increase in the present discounted value of the fuel cost over the remaining life of the vehicle resulted in a $1 decrease in the price paid for the vehicle. The policy implication is that no CAFE policy is necessary because consumers are willing to pay the cost of improved fuel economy on their own.

Cigarette Taxes

“Cigarette Taxes and Youth Smoking: Updated Estimates Using YRBS Data,” by Benjamin Hansen, Joseph J. Sabia, and Daniel I. Rees. June 2015. NBER #21311.

In these pages last winter, Kevin Callison and Robert Kaestner presented evidence that cigarette taxes have become so high and smokers are now such a small percentage of the population that current consumption is price inelastic (“Cigarette Taxes and Smoking,” Winter 2014–2015). That is, remaining smokers have such a strong preference for smoking that further cigarette tax increases will not reduce use further and cannot be justified by health benefits.

This paper asks whether the same is true for youth smoking. The standard answer, using 1991–2005 data from the Youth Risk Behavior Survey, was that a $1 increase in the per-pack tax reduced smoking participation among high school students by 3–6 percentage points and frequent smoking (smoking in 20 of the last 30 days) by 2–4 percentage points. However, data from 2007–2013, when several states increased their cigarette tax, did not find any relationship between higher taxes and further declines in youth smoking or smoking frequency. Also, adding state-specific time trends to the 1991–2013 data eliminated the effect of tax increases on youth smoking. The authors speculate, analogous to the authors of last winter’s article, that the only young smokers left are price inelastic, like their elders.