Professor Jenny Schuetz is one of my favorite colleagues here at USC. She’s in the real estate group, and her work focuses on urban economic geography. She gave a seminar Wednesday for Metrans on work she has done on the commercial real estate effects of transit investment. You can see a draft of her paper, entitled Do Rail Stations Encourage Neighborhood Retail Activity, here. Jenny’s contributions to the literature are careful empirical analyses of urban economic theories of location. Her contributions stand out because she works incredibly hard to get the detailed data that she needs in order to answer questions well. Data on the commercial real estate sector can be hard to get, and planners tend to pin pretty high hopes on transit’s transformative power while not systematically testing their claims.
The question tested here are straightforward: Does new transit investment prompt more retail activity within a buffer of the station area?
But just because it’s a straightforward question does not mean getting a good answer is easy. Why not? Because like just about all aspects of urban research where we’d like to see cause and effect, there are many possible explanations for why we might see an effect if we do, in fact, see an effect. If I do a great job with route planning, for example, it’s likely that I’m targeting places that are already growing. So investment there may simply be a continuation of the existing trend. We need to see some change in the trajectory, and that’s hard. For places in decline, we have to show that, post-investment, there’s some slower decline or turnaround. And that’s hard to show when you are just looking at data at one point in time and noting that land values are higher near rail stations. That’s good, but it’s not the sort of evidence we need to be able to say that the investment was a game changer.
Jenny contends with this problem by looking at the conditions from 1992 to 2009 both case and control areas surrounding train stations in four California cities: San Francisco/San Jose, Los Angeles, Sacramento, and San Diego. Her controls are tracks outside the normally accepted walking area around stations. If trains are really making a difference in retail, you should see higher performance in the tracks within the station capture area than for those outside. Granted, this is an arbitrary boundary, but the research shows that the capture area commonly used around transit stations is pretty sound. There are so many weirdos like me who will walk 2 miles to a station. (What else am I going to do for exercise? A nightclub?)
Anyway, the resulting model includes some information about the station characteristics, locational metrics (station density, proximity to highway, distance to CBD (a combo of variables I very much like)), and some local population statistics we would normally associate retail attraction. (Retailers like areas with moneyed folk in them.) Her dependent variable is retail employees per square mile and she’s got about 500 station areas in the study.
She finds no statistical difference between rail-accessible land compared to the controls in San Francisco and San Diego; in Los Angeles and Sacramento, she does a find a significant difference, but negative: that is, rail-accessible areas lost retail employment compared to controls, save for suburban station areas. Rail development seems to offering suburban locations a chance to get more retail in two of the cities.
I do have some questions, since this is a draft paper. One question just concerns the dependent variable as a measure; retail employment. I’ve been skylarking (skylarking = thinking unencumbered by either theory or data) about changes in the retail sector over that time period, and its possible that there is something going on with the type of retailer that gets attracted to station-areas. Planners have great faith in local small businesses, and I’m sure that, when aggregated, they create a lot of employment. But it’s entirely possible that you might see new small businesses start of up in a station area (whee!) but on the whole, those employ fewer people in the retail sector than a big box located off a highway. Or there might be something technological going on, where employment in the retail sector is systematically declining, and those places without investment have what retail they have and those linger for awhile, but those new opportunities for retail don’t really blossom particularly fast around new commercial land supply around train stations.