Parking increases citywide car use, SSTI researcher finds

By Chris McCahill

Automobile use has been on the rise in cities for nearly a century and so has the supply of parking. Because driving often seems unavoidable, policymakers, developers and the public push endlessly for more parking to meet demand. That push, however, might only be making matters worse.

My research team, which includes Norman Garrick, Carol Atkinson-Palombo and Adam Polinski from the University of Connecticut, has produced a new study on the subject, which strongly suggests that abundant parking in cities causes people to drive more, shedding important light on the question of cause and effect. We presented the research at this year’s annual Transportation Research Board meeting in D.C.

Admittedly, this isn’t a new idea. For the past several decades, a handful of researchers have studied the purveyance of free parking in the U.S., its hidden costs, and its effects on travel behavior. By examining individual streets, buildings and neighborhoods, study after study has shown that people respond to the price and availability of parking by adjusting their driving habits. That’s especially true in cities, where there are usually more options for getting around.

Yet as cities grow, more often than not, so do their parking supplies. City records that we uncovered in an earlier study show policymakers demanding more parking year after year, even as they worry about the impacts it could have on the built environment and car use. And so we’ve been left to wonder whether cities are simply adapting to the automobile in order to grow, or perpetuating urban automobile use.

In trying to answer this question, we borrowed a classic approach from the field of epidemiology. In the same way that health experts showed how heavy smoking increases the chances of getting lung cancer, we wondered if we could evaluate whether decades of parking growth in cities caused people to drive more and more over time. That’s not easy to do. Cities are complex and parking data is scarce—often nonexistent.

We used the only data sources available to us—historical aerial photographs and census data—to track changes in parking and automobile use as best as we could in a carefully chosen batch of older, medium-sized cities. The data take us back as far as 1960. We then applied a simple framework known as the Bradford Hill Criteria to infer causality.

Most of the nine criteria are fairly straightforward. We asked questions like:

  • Is there a strong relationship between the two?
  • Is the effect plausible?
  • Is the effect large enough to matter?
  • Do other studies show similar findings?

Some questions aren’t fully answered, but by and large the answers are ‘yes’. One of the more interesting findings we uncovered, related to the question of which came first, is that parking growth prior to 1980 is a powerful predictor of automobile use in the following two decades. For a city that added 40 parking spaces per 100 residents before 1980, we expect eight percent of additional residents to then drive to work, as shown in the figure below. For a city that added only 10 to 15 spaces, we expect no increase in driving. (Cities in this lower range also grew more quickly, on average.) This suggests that policymakers were either exceptional at predicting demand, which most evidence disputes, or that parking growth set the pace for changes in driving habits.


The same relationship isn’t nearly as strong when we try to predict parking growth from prior increases in driving.


Overall, we found that every 10 parking spaces per 100 residents and employees in a city are linked to 7.7 percent of commuters driving. So the difference between having 20 and 50 spaces per 100 people is that 60 versus 83 percent of commuters drive, respectively. That covers the full range of values we observed in 2000, but the relationship also held over the entire 40-year study period.

Other factors certainly helped push driving upward as well. With data for all of those factors and a much larger study, we might be able to tease out their various effects and understand how they work together. For now, the Bradford Hill Criteria have let us single out parking as one likely, and important, cause.

Cities throughout the U.S. are finding that they can grow without adding huge amounts of parking, avoiding its hidden costs and slowing future traffic growth. San Francisco and Seattle have started by pricing on-street parking to manage demand. Chicago reduced its parking requirements near transit stations, while other places (including downtown Nashville and recently Fayetteville, Arkansas) have eliminated them completely. Historical urban centers like Cambridge, Massachusetts, and even traditionally car-oriented places like Tysons Corner, Virginia, impose parking maximums and incorporate other transportation demand management strategies. More cities and towns should take these lessons to heart.

Chris McCahill is a Senior Associate at SSTI.