Now at Streets.MN: Do Streetcars Promote Economic Development? :
“It is often said that “Absence of Evidence is not Evidence of Absence”, but this is wrong. This is especially wrong in a context were we have motivated people searching for evidence. Consider the example of Bigfoot. Bigfoot is a supposedly big humanoid / primate living in very small numbers (and presumably hiding from humans). There was a fad in my youth for blurry pictures of Bigfoot to appear in weekly tabloids sold at supermarket checkout stands. However, as XKCD points out, cameras are everywhere now, on billions of phones, and yet we have no more evidence of Bigfoot than before.”
We have evidence in many historical cases of the co-evolution of transport and land use (e.g. London, New York, Twin Cities). In those cases the transport mode developed was faster than alternatives, and therefore increased accessibility, making land more valuable.
We have no evidence that streetcars, of themselves, promote economic development in the context of present-day US cities. That is, there is no case where modern streetcars were built, nothing else was done by the public sector (no road reconstruction, no public subsidies for development, no change in development regulations), and the level of private sector economic development changed measurably, and more than in an otherwise comparable control case.
We have hypotheses as to why there should be no effect, and that is the maturity of the system (streetcars are not connecting places that are presently unconnected) and the lack of positive changes to accessibility (or even negative changes to accessibility) that comes with adding a slow mode to a network which is already faster.
It is often said that “Absence of Evidence is not Evidence of Absence”, but this is wrong. This is especially wrong in a context were we have motivated people searching for evidence. Consider the example of Bigfoot. Bigfoot is a supposedly big humanoid / primate living in very small numbers (and presumably hiding from humans). There was a fad in my youth for blurry pictures of Bigfoot to appear in weekly tabloids sold at supermarket checkout stands. However, as XKCD points out, cameras are everywhere now, on billions of phones, and yet we have no more evidence of Bigfoot than before.
Does this constitute “proof” of the non-existence of Bigfoot? No, because one can never prove a negative (See Popper on Falsifiability). It does however cause the rational among us to become increasing skeptical about our hirsute friend’s likelihood of being real.
The very logical Less Wrong blog discusses this issue: Absence of Evidence Is Evidence of Absence :
“But in probability theory, absence of evidence is always evidence of absence. If E is a binary event and P(H|E) > P(H), “seeing E increases the probability of H”; then P(H|~E) [Probability of H given Not E] is less than P(H), “failure to observe E decreases the probability of H”. P(H) is a weighted mix of P(H|E) and P(H|~E), and necessarily lies between the two.”
We need to think in “but for” way when evaluating economic development claims.
Would the development not occur “but for” this particular investment? Would it occur with streetcars and BRT, with only streetcars, with only BRT, with neither?
We can’t fully know this without running 4 experiments in 4 parallel universes. We can estimate this statistically by looking carefully at multiple cases that have already opened, under multiple conditions, and get likelihoods that effects are as estimated by the model. There are some examples of modern transit lines increasing property values, there are some examples of no effect, and there are even some examples of transit lines destroying wealth (sometimes all found in the same study, e.g. Rail Transit Investments, Real Estate Values, and Land Use Change: A Comparative Analysis of Five California Rail Transit Systems by Landis et al.). Once lots of these studies are done, we can do a meta-analysis, and try to put the complexity of findings into some order. The meta-analysis is much more robust, in that it essentially has the combined sample size of all of the studies it takes in. For instance, see The Impact of Railway Stations on Residential and Commercial Property Value: A Meta-analysis . These studies unfortunately do not generally apply to modern streetcars, which have very different characteristics than high capacity services.
There have been a few attempts to summarize the results of the Streetcar and economic development debate:
- TCRP Synthesis 86: Relationships Between Streetcars and the Built Environmentbasically finds an absence of evidence.
- The Great Streetcar Conspiracy by Cato’s Randal O’Toole critiques many of the claims of streetcar advocates, especially the claims about economic development in Portland. [Here's some recent news about Portland's streetcar revenue.]
- Bob Dunphy, retired from the Urban Land Institute, points me to a set of reports critical of the proposed Arlington Streetcars, one of which he authored.
None of the reports about economic development effects have survived a rigorous peer review process. So in the end, we equivocate. If the new transportation infrastructure notably increases the relative accessibility of a place (compared to other places), it might attract some development that would otherwise go elsewhere. If it signals to developers to coordinate actions, and develop here, rather than there, it might also concentrate development. If we concentrate development, and create more accessibility, we might have some economies of agglomeration further driving growth. If, If, If.
The evidence we do have is that employment in the core cities of Minneapolis and St. Paul is very stable, independent of most of the vagaries of the economy, shifts from low rise to skyscrapers, construction of freeways and skyways, the expansion of the University, the rise of the dual worker household, and so on.
The table below shows employment in Minneapolis and St. Paul from 1970 – 2010.
Do we really believe a small investment in infrastructure serving a thousand additional people per day will move this? That would be the equivalent of urban homeopathy.
|Line Length(miles)||2030 Avg Weekday Project Boardings||2030 Avg Weekday New Riders||Capital Cost (million $)||Operating Cost (million $)||Capital Cost / Weekday Rider|
|Streetcar starter line||3.4||7,200||1,200||200||10.6||$27,778|
|Met Transit aBRT||21.8||34,700||6,700||110.7||9.06||$3,190|
And if you prefer your denominator to be riders per mile of construction rather than riders:
|Cost / Rider / Mile|
Upzoning and development would be factored into all alternatives, especially since a number are already being developed without a streetcar or other improvement. These ridership projections also include the AA’s assumption that streetcar riders perceive a 25-minute travel time savings (implicitly making their travel time negative in some cases), and thus more choose streetcar.
We need to consider that the Met Council would likely ask Minneapolis to cover most if not all operating costs. That means Minneapolis alone is responsible for $10 – 20 million per year, without a legislature or whole large transit apparatus to back the City up. And that’s only one line.
Advocates will advocate, that is their nature. If no one believes their small estimates of economic impact, the estimates will just grow and grow, just so they can get attention. But don’t confuse their advocacy with scientific knowledge, about which we have very little.
Are they the best amenity? Are they the best transportation service possible? Or do they drain resources that would otherwise be spent on something else, like maintaining and improving existing transit systems or serving many more passengers with Arterial BRT?