Tag: highway fatalities

Transit Industry Claims That Correlation Proves Causation

A new report from the American Public Transportation Association (APTA) comes out firmly in support of the belief that correlation proves causation. The report observes that traffic fatality rates are lower in urban areas with high rates of transit ridership, and claims that this proves “that modest increases in public transit mode share can provide disproportionally larger traffic safety benefits.”


Here is one of the charts that APTA claims proves that modest increases in transit ridership will reduce traffic fatalities. Note that, in urban areas with fewer than 25 annual transit trips per capita – which is the vast majority of them – the relationship between transit and traffic fatalities is virtually nil. You can click the image for a larger view or go to APTA’s document from which this chart was taken.

In fact, APTA’s data show no such thing. New York has the nation’s highest per capita transit ridership and a low traffic fatality rate. But there are urban areas with very low ridership rates that had even lower fatality rates in 2012, while there are other urban areas with fairly high ridership rates that also had high fatality rates. APTA claims the correlation between transit and traffic fatalities is a high 0.71 (where 1.0 is a perfect correlation), but that’s only when you include New York and a few other large urban areas: among urban areas of 2 million people or less, APTA admits the correlation is a low 0.28.

The United States has two kinds of urban areas: New York and everything else. Including New York in any analysis of urban areas will always bias any statistical correlations in ways that have no application to other urban areas.

In most urban areas outside of New York, transit ridership is so low that it has no real impact on urban travel. Among major urban areas other than New York, APTA’s data show 2012 ridership ranging from 55 trips per person per year in Los Angeles to 105 in Washington DC to 133 in San Francisco-Oakland. From the 2012 National Transit Database, transit passenger miles per capita ranged from 287 in Los Angeles to 544 in Washington to 817 in San Francisco.

Since these urban areas typically see around 14,000 passenger miles of per capita travel on highways and streets per year, the 530-mile difference in transit usage between Los Angeles and San Francisco is pretty much irrelevant. Thus, even if there is a weak correlation between transit ridership and traffic fatalities, transit isn’t the cause of that correlation.

San Francisco and Washington actually saw slightly more per capita driving than Los Angeles in 2012, yet APTA says they had significantly lower fatality rates (3.7 fatalities per 100,000 residents in San Francisco and 3.6 in Washington vs. 6.4 in Los Angeles). Clearly, some other factor must be influencing both transit ridership and traffic fatalities.

With transit ridership declining almost everywhere, this is just a desperate attempt by APTA to make transit appear more relevant than it really is. In reality, contrary to APTA’s unsupported conclusion, modest rates in transit ridership will have zero measurable effect on traffic fatality rates.

Topics:

A Ban On “Walking While Wired”?

New York state senator Carl Kruger (D-Brooklyn) is crusading to ban pedestrians’ use of cellphones and other mobile devices while crossing the street. It’s for your own good, you must understand:

“When people are doing things that are detrimental to their own well being, then government should step in.”

The Daily Caller asked me to write an opinion piece about this proposal so I just did. Excerpt:

Phone use on the street has become near-ubiquitous in recent years, yet over nearly all that time — nationally as in Gotham — pedestrian death rates were falling steadily, just as highway fatalities fell steadily over the years in which “distracted driving” became a big concern.

In the first half of 2010, the national statistics showed a tiny upward blip (0.4 percent), occasioned by a relative handful of fatalities in a few states. Even a spokesman for the Governor’s Highway Safety Association, Jonathan Adkins, seems to agree it’s premature to jump to conclusions: “You don’t want to overreact to six months of data,” he told columnist Steve Chapman.

Like others who seek quasi-parental control over adults, Sen. Kruger tends to infantilize his charges. He told the Times: “We’re taught from knee-high to look in both directions, wait, listen and then cross. You can perform none of those functions if you are engaged in some kind of wired activity.”

This drew proper scorn from columnist Chapman: “Actually, you can perform all those functions and dance an Irish jig, even with text messages or rock music bombarding you.” That some ear bud devotees don’t take due caution is no reason to pretend they can’t.

C.S. Lewis, Lily Tomlin and Transportation Secretary Ray LaHood all get walk-on parts as well.

Friday Podcast: ‘Drinking Ages and Highway Fatalities’

Does the policy of setting a national drinking age reduce highway fatalities?

In Friday’s Cato Daily Podcast, Jeffrey Miron, senior lecturer in economics at Harvard University, talks about the research he and student Elina Tetelbaum (now a Yale Law student) carried out on that question:

What we find is that the only area where there is any evidence for efficacy of the law are states that adopted a higher drinking on their own without any compulsion. For the states that the feds forced … to raise [their] drinking age, there is no evidence of a beneficial reduction in traffic fatalities… We conclude quite strongly that it’s only when a state chooses a higher drinking age on its own, it’s only when it decides its going to devote enforcement resources and when there’s public sentiment to support that, that you see those sorts of beneficial effects.

Miron and Tetelbaum offer a more detailed look at their findings in the Spring issue of Cato’s magazine Regulation, which will be released March 26.

Topics: