Q&A: Elizabeth Delmelle, Associate Professor of City & Regional Planning

This summer, geographer and geographic information scientist Elizabeth Delmelle joined the Weitzman faculty and became the director of the School’s Master of Urban Spatial Analytics (MUSA) program, which equips students to use geographic information system mapping (GIS), spatial analysis, and data science to address the public policy and city planning conundrums of our day. Prior to Penn, Delmelle was an associate professor at the University of North Carolina at Charlotte, where she became an authority on the changing dynamics of the rapidly growing city of Charlotte. 

What attracted you to Weitzman’s Urban Spatial Analytics program?

A lot of things. I did my PhD in geography and urban regional analysis. We learned different methods, mainly quantitative, for addressing urban issues. When I was doing my dissertation, I really got into looking into other fields, like computer science, for methods that can be used for visualizing and analyzing neighborhood change, which is a multidimensional concept. A neighborhood is not just one thing. There are people and there's income. There's crime and there's housing. I didn't feel at the time that there were great methods for analyzing these changes across multiple characteristics over time and across space. So, I became kind of obsessed with this quest for methods that could better enable us to visualize these changes and analyze them.

Not a lot of people in geography were doing computational and quantitative urban geography. In geography, there's a big rift between the qualitative folk and the quantifiers. So, I was a bit on my own in the small group of people who were doing this very quantitative and computational work. I found a home teaching in a more traditional GIS [geographic information system] program, where I applied the methods to urban applications—neighborhoods and transportation. The MUSA program is very aligned with what I had been doing and followed the similar trajectory from GIS to more computational methods.

What sets MUSA apart from other GIS programs is that there was a decision made several years ago to not just teach ordinary GIS, using a kind of point-and-click mapping, but to ask students to learn programming, machine learning, and statistics. These are the notoriously most disliked skills and courses in traditional GIS programs.

Credit goes to the previous director of MUSA for making this change and maintaining its popularity. There was this foresight that this would be the way of the future and the program evolved to match the demands of what the job market would be. Now everybody wants these skills, and now everybody is trying to figure out how to change the mindset and get students to take more coding classes. There are so many GIS programs across the country, but a program that really teaches these intense computational methods, applied within the urban context, is a rare thing and in high demand.

One area of your research challenges the assumptions people tend to make about how things work in cities and what directions planning should take. For example, you have done a lot of research on the Lynx Blue Line, a light rail line completed over the last 20 years in Charlotte, and how that has impacted the city.

Yes, we like to make people mad [laughing]. My former colleague, Dr. Isabelle Nilsson and I led a number of studies on the impact of the new light-rail system on neighborhoods, basically answering the question, does investing in new transit cause gentrification and displacement? In a city like Charlotte, there was just a single rail line proposed and some people were saying, "No, don't build it. It’ll totally change the neighborhood.”

Dr. Nilsson and I found all these anecdotal studies in different cities where people were protesting light rail projects based on this assumption. There was one project in Nashville that people were successful in blocking by saying that it would cause gentrification. And we found this interesting Op-Ed that said that all of the gentrification in a Los Angeles neighborhood was because of a train line. That was causing this backlash against investments.

We decided to see what the data said. We looked at the topic from several different perspectives, asking "Does transit cause lower income individuals to disproportionately leave after you've built a light rail station? Can we measure displacement?" We looked at that across the entire country and we found no evidence that that was the case.

Then we followed up with other different outcome variables. Does it cause neighborhood income diversity changes? Does it change eviction rates? Does it change housing values? Does it change who's moving in? And, the large majority of the time, we could not find any statistically significant evidence that it was the train causing these results.

In the popular press, you read one thing about gentrification due to new light rail systems, but when you read our papers, you read another version. The reason for this is “the means versus the extremes.” Quantitative data is really good at looking at averages. And when you have an average, you have some observations that are above the mean and some observations that are below.

When you read a story in the press, a journalist is going to find the observations that are above the mean, because that's the interesting story. They're going to find the one or two cases where something dramatic has happened and focus on that. Whereas we're much more restrained. We say, "Well, no, on average, everywhere, it's not going to cause systematic change." In a few cases, where you have a neighborhood that has a lot of other things going for it–maybe it also happens to be close to a downtown and have some cool restaurants–it's probably going to gentrify anyway. If you add a train station there, well, that makes it even more attractive. But in a neighborhood that doesn't have a lot of things going for it–maybe it has a lot of crime, poor school quality, no desirable housing–if you just add a train, that train on its own is not going to cause a whole bunch of people to move to the neighborhood simply for the train. It can never be one thing that causes neighborhoods to change. It's the whole context. We kind of spoil the party when we say, “on average,” and just look at these boring means.

In other research you've done, you found that light rail, in a city like Charlotte, might not be the best investment for helping people get to their jobs.

That's right. Charlotte is an auto-dominated city. It was constructed around the automobile and does not have a lot of walkable neighborhoods. The few neighborhoods that are walkable–what they call Uptown or Downtown–work because transit works well in walkable, dense urban environments where there's a mixture of land uses. In the few neighborhoods in Charlotte where that's true, there are few remaining low-income people because the price of land is too high. Lower-income and largely minority residents are moving to older suburban houses, constructed in the 1970s and '80s. They are cookie-cutter suburbs with a lot of cul-de-sacs. So, we have low-income people spread out in older suburban neighborhoods and the jobs that they might want to get to are also dispersed. So, there's no mass transit solution currently that will effectively get that population to the jobs they might be interested in at this moment.

I tend to be controversial and say, "You can invest in light rail, and it's good to think about the future, but that's not really going to help the low-income population. It's just going to give more resources to people who already have some. What those neighborhoods need are all of the supporting infrastructure. They need sidewalks. They need to be walkable. They need mixed land use. We need to do all of the ‘boring’ stuff before we can have the cool train.”

How are you liking Philadelphia?

On my way to Philadelphia, we took a small detour. After living in Charlotte for a long time, last year we were in the Netherlands. I was on sabbatical, living in The Hague and living a car-free lifestyle. That detour has made the transition from Charlotte to Philadelphia easier. I'm able to ride my bike to campus. We are living in West Philly, Cedar Park, and learning that part of the city, and having this urban environment and atmosphere feels good.