More Highways, More Pollution?

Air pollution decreases even as we drive more

According to a new report by the Public Interest Research Group (PIRG) “building new highways will do little to alleviate traffic congestion in the long run and likely will exacerbate already severe air pollution problems in metropolitan areas across the country.” PIRG asserts the only solution is to restrict how much people can drive and build more public transit.

The report, “More Highways, More Pollution,” is the latest effort by anti-automobile activists to torture the data until it gives the desired confession. In fact, despite a tripling of urban driving during the last 30 years, air pollution has dropped dramatically, while areas that have more freeway lane-miles per capita have lower traffic densities.

Driving and Air Pollution

PIRG claims that building more highway capacity will increase driving and therefore air pollution. But if building highways inevitably increases air pollution, we should have seen a huge increase in air pollution during the last few decades. In fact, just the opposite occurred. While 60 percent of the nation’s ozone monitors violated the 1-hour ozone standard in the late 1970s, only 10 percent do so today. Areas with the worst pollution achieved the greatest gains. For example, the San Bernardino area exceeded the 1-hour standard more than 150 times per year in the late 1970s, but only 20 to 30 times per year today.

The nation has likewise made great progress on airborne particulate matter (PM). Only a few percent of the nation’s monitoring locations still violate the PM10 standard. About 20 percent violate EPA’s new and much more stringent PM2.5 standard, but PM2.5 levels have steadily declined, dropping by about 40 percent during the last 25 years. Recent monitoring data show these declines are continuing.

Although at least 90 percent of urban carbon monoxide (CO) emissions come from motor vehicles, more than 99 percent of the U.S. has met the federal carbon monoxide (CO) standard since the early 1990s. The remaining few locations recently came into attainment.

Despite most nitrogen oxides (NOx) coming from motor vehicles, the entire U.S. has been in attainment of the nitrogen dioxide standard for more than a decade. Air toxics have also declined. Although benzene comes mainly from motor vehicles, ambient benzene levels dropped more than 70 percent around the U.S. between 1989 and 1999.

Highways and Air Pollution

PIRG reports that total miles driven in urban areas tripled between 1970 and 2002. Thus, America achieved extraordinary air pollution reductions despite vast increases in driving. These facts are lost on PIRG, which ignores past pollution trends and claims pollution is “severe” and getting worse. In PIRG’s imaginary world “The experience of the last 30 years has shown that limits on tailpipe emissions — while necessary — are not enough to resolve the problem of vehicular air pollution. Any strategy to reduce health threats from air pollution must include a strategy to curtail the growth of vehicle travel.”

Ironically, even PIRG’s own numbers don’t support its conclusions. For example, PIRG notes that “Vehicles are 80 to 99 percent cleaner per mile than vehicles produced in the late 1960s” (emphasis in original). Let’s assume then that per-mile emissions of the average vehicle declined 90 percent since the late 1960s — the middle of PIRG’s range. If driving hadn’t increased, this would have led to a 90 percent decrease in total emissions. Or put another way, if emissions equaled 1.0 in 1969, they would equal 0.1 today. Now let’s add a factor of three increase in driving: 0.1 * 3 = 0.3. In other words, even with a tripling of driving, technological improvements in vehicle emissions control reduced total emissions by 70 percent.

On-road measurements show vehicle emissions are dropping about 10 percent per year due to fleet turnover. Driving is increasing only about one to 3 percent per year, for a net decline in emissions of 7 to 9 percent per year. Because more-recent vehicle models continue to start out and stay cleaner than earlier models, fleet turnover will continue to clean the air. New SUVs have been as clean as new cars for the last several years, so the popularity of larger vehicles won’t affect future air quality. EPA regulations that phase in during the next few years require an additional 70 to 90 percent reduction in new-vehicle emissions from both cars and diesel trucks, ensuring that most remaining motor-vehicle pollution will be eliminated during the next 20 years or so. PIRG’s fanciful claims not withstanding, technology will continue to win the battle against air pollution without the need to restrict people’s travel choices.

Highways and Congestion

PIRG believes that building highways “induces” more demand for travel, eliminating any gains from extra road capacity. If people had an infinite demand for automobile travel this might be the case. But in the real world, cities with more highway lane-miles per capita actually have lower traffic densities, as shown in Figure 1.

Figure 1. Relationship between Driving and Highway Miles

Notes: Analysis by Wendell Cox, based on 1999 data from the Texas Transportation Institute for urbanized areas with more than 1 million population. Each point represents an urbanized area. Slope is statistically significant at the 99.9% level.

As transportation researcher Randal O’Toole puts it, “travel that is ‘induced’ by added capacity is actually travel that had been repressed or shifted by capacity shortages.”[1] Between 1980 to 2000 the number of cars in the U.S. increased 50 percent, while total miles driven increased 75 percent. Yet the number of road miles increased by only about 5 percent. Recent increases in congestion aren’t due to building more freeways, but are “due to more driving without a similar increase in freeways.”[2]

Most Americans consider the unparalleled convenience and flexibility of the automobile to be a great benefit. Even with increasing congestion due to lagging road investments, most Americans still prefer driving to other modes of transportation. PIRG wishes it were otherwise, but getting people to use transit is no easy task. Metropolitan planning organizations, the regional agencies that draft metropolitan transportation plans, predict that even spending thousands of dollars per capita on new urban transit services — hundreds of billions on a nationwide basis — would at most reduce single-occupant-vehicle trips by a few percent below “business as usual.” And the urban densities required to make transit viable will increase congestion, because per-capita driving decreases only modestly with increasing density.

Instead of providing infrastructure for the types of transportation Americans most desire, PIRG instead aims to end highway building and increase transit. Where the vast majority of Americans see great net benefits to automobile travel, PIRG sees only costs. Rather than pursue the “public interest,” PIRG seeks to override Americans’ preferences.

False Premises

PIRG’s staff carefully chose and structured the data to tell the story they wanted. But even here they didn’t do a very good job. PIRG used EPA’s official inventory of VOC and NOx emissions for its estimates. However, these inventories have never passed a real-world validation test. For example, PIRG claims, based on EPA estimates, that 29 percent of volatile organic compound (VOC) emissions come from on-road vehicles. But real-world studies in several cities show the actual contribution ranges from about 50 to 85 percent. PIRG could have bolstered its anti-automobile case if its staff were familiar with basic air pollution science.

Indeed, PIRG didn’t even bother to check its central claim that per-capita pollution emissions correlate with ambient pollution levels. They don’t. As shown in Figure 2, there is no correlation between PIRG’s estimate of emissions per capita and actual measured ozone levels. Several factors could explain this. First, PIRG naively used EPA’s incorrect emissions inventory. Second, pollution levels vary based on differences in weather from place to place. Third, urban form affects pollution levels. Higher population density means more emissions per unit area. Suburbanization might increase driving, but it reduces the density of emissions much more. Ambient pollution levels are what matter for health. If higher per-capita emissions don’t translate into higher ambient pollution levels, then PIRG’s entire study is based on a false premise.

Figure 2. PIRG’s Estimate of Per-Capita Motor Vehicle Emissions vs. Ozone Levels

Notes: Each point represents a single metropolitan statistical area (MSA). “NOx + VOC per capita” is for 1999 and is from the PIRG report. All MSA’s with more than 1 million population are included. “Ozone exceedances” is the number of 8-hour ozone exceedances in 1999 at the worst location in each MSA. Ozone data were downloaded from EPA here.

Most people prefer auto-based lifestyles, and technology and fleet turnover are mitigating air pollution. Instead of trying to make motorists miserable, scarce transportation dollars should instead be allocated to reflect the actual relative demand for transit and auto travel. That would allow cities to provide the additional automobile infrastructure necessary to keep up with travel demand, reduce road congestion, and provide a transportation system that meets people’s needs.

PIRG’s report should provide grist for courses in critical thinking, but shouldn’t be taken seriously as a guide for policymakers.

Joel Schwartz is an adjunct fellow at Reason Foundation and visiting scholar at American Enterprise Institute.


[1] R. O’Toole, The Vanishing Automobile, and Other Urban Myths (Bandon, OR: Thoreau Institute, 2001).

[2] Ibid.