“The Bad Target”: The Rise of Consumer Redlining

Danny Caine
In addition to being your trusty podcast host, I’m also a poet. One of my poetry books features a poem called In the Bad Walmarts Parking Lot. I’ll spare you the whole thing, but the poem starts with the line, every town with two Walmarts has a good one and a bad one. I wrote this from my lived experience. At the time, I was living in Lawrence, Kan., a city of two Walmarts. Indeed, there was pretty broad agreement about which one was more pleasant to go to. As a poet, I use this as a springboard into writing a silly little sonnet about being sad in that oceanic parking lot. But as a sociologist and law professor, Andrea Chandra Shaker took the question of the bad Walmart in a totally different direction. As it turns out, every town with two Walmarts has a good one and a bad one is not only a poem, but a statistically proven fact. And as is always the case with this season of building local power, race plays a part. A professor of law at UC Riverside, Chandra Shekhar is the author of the recent paper, The Good Starbucks, Consumer Redlining in Large American Chain Stores. In it, Chandra Shekhar not only proves that towns with two locations of the same chain have a good one and a bad one, but also that the bad one is almost always in a poor or Black or Latinx neighborhood. Her findings tell all kinds of compelling stories about racism, corporate power, and the American retail experience. We’re so glad to have her here to tell those stories. Andrea Chandrasekhar, thank you so much for being on Building Local Power. I can’t wait to talk to you about your work and your research. I think my first question for you, just tell us a bit about your academic history and your journey and what led you to study consumer redlining.

Andrea Chandrasekher
Sure. And first all, thank you so much for having me. So I am a professor of law, but I’m also an economist and a statistician by training. When I was in graduate school, I completed a law degree as well as a PhD in economics and a master’s in statistics. And what this means is that in my research, I almost always ask questions that are empirical in nature, questions that one can answer by analyzing large data sets with statistics and econometric modeling. In terms of what led me to study consumer redlining, that’s a longer story. the idea actually came to me from my own personal experiences as a shopper. My family and I regularly shop at two different Target stores. There’s the one in the Black neighborhood where we go to get the Black hair care products that are typically not sold at all Target stores. And then there’s the one in the White neighborhood where we go to get everything else. Now you might ask, why do this? It takes more of my time to shop at two different stores. It takes more gas money. Why not just get everything that we need from the Black Target store? The reason is that the White Target that we go to offers a much higher quality of consumer experience. The floors are cleaner. The shopping carts are cleaner. There isn’t sticky spilled stuff on them. The shelves are stocked and organized. The lines are shorter because they have more employees working there, which also means that there’s always someone available to help me when I can’t find something. And most importantly for me, the laundry detergent isn’t locked up, which is my biggest pet peeve. All of these things together create a much more pleasant shopping experience for me. And that’s something that really matters to me, how pleasant or unpleasant the shopping experience is. And then about a year or so ago, I started to think more and more about the ridiculousness of going to two different Target stores. And that’s when I began to put my researcher hat on. That’s when I began to notice a more general trend where stores in Black and Latinx neighborhoods seem to offer a lower quality of consumer experience than stores in white neighborhoods. And what really surprised me is that this seemed to happen within the same chain of stores where presumably the consumer experience is supposed to be broadly consistent across different locations. So like within Starbucks or within Home Depots, the Black-located and the Latinx-located stores seem to always offer a lower quality of consumer experience. As I looked more and more into this, I discovered that this practice, the practice whereby chain store businesses offered a lower quality of consumer experience in poor, more BIPOC neighborhoods was called consumer redlining. And even though the practice had a name in the academic literature, interestingly, no academic had ever rigorously studied it before with large data sets. And so that’s what I set out to do.

Danny Caine
It’s really interesting. when I first encountered your research, it really struck me because like, I live in Cleveland right now, everybody knows what the worst Walmart in Cleveland is. And I had just like, accepted that without really thinking about like, this is a chain, should all be the same. what are the kind of the socio economic things that determine why this Walmart is the worst, in addition to just the fact that it’s not very pleasant to shop there. So you say that this already had a name in the literature, consumer redlining. I had also heard the term retail redlining before. And is there a difference between those two things and can you explain it?

Andrea Chandrasekher
Yeah, great question. So as I mentioned, consumer redlining is where a chain store business offers a lower quality of consumer experience at their stores in poorer, more BIPOC neighborhoods than at their stores in Whiter more affluent neighborhoods. Retail redlining, on the other hand, occurs when businesses just avoid opening stores at all in BIPOC consumer redlining, chain businesses do in fact open stores in BIPOC neighborhoods. They just offer a lower quality of consumer experience there.

Danny Caine
Okay, so we’ve got consumer redlining kind of defined in the literature as a concept, but not studied terribly much and you want to go about doing this. Tell us what you did. Like, how do you go about studying and redlining?

Andrea Chandrasekher
Yeah, absolutely. So it starts with looking at patterns, or in statistical terms, correlations. I started out by looking to see if the neighborhoods that had larger than average Black and Latinx populations were also more likely to have chain stores that had lower than average Yelp ratings. And I found out that, that was true. But then I said to myself, well, that could be for a lot of different reasons, right? It could be that the lower quality of consumer experience at the Black-located chain store, for example, is due to the fact that the neighborhood itself has lower market potential or generates lower revenue. If either one of those two things is true, right, lower market potential or lower revenue, then perhaps it’s rational for a profit-maximizing chain business to give the Black location a lower labor budget and thus fewer employees. And then that would explain the longer lines and the dirtier facilities. Similarly, it could be that the lower quality of consumer experience at the Black-located chain store is due to the neighborhood perhaps having higher than average crime. If so, then perhaps it’s rational for a profit-maximizing firm to give the Black location less stock because they’re worried about theft. So in my analysis, I wanted to be able to control for demographic factors like neighborhood market potential, neighborhood crime, neighborhood income, education, poverty level, basically anything that is both correlated with race and could also have an independent impact on the quality of the consumer experience at the store. Statistically speaking, in other words, I wanted to hold all those other factors constant and then determine if there was still a difference in chain store Yelp ratings between BIPOC locations and White locations. Multiple regression analysis, I’m gonna say that again, multiple regression analysis is the statistical technique that I used that allowed me to do that. It’s basically the equivalent of taking two neighborhoods that are else equal. So same average crime level, same average market potential, et cetera, except that one is in a Black neighborhood and one is in a White then asking within chain, how do the ratings of the Black located chain store compare with the ratings of the White located chain store? And so that’s what I did.

Danny Caine
So what did you find?

Andrea Chandrasekher
So, in total, I analyzed over 400 large American chain stores. I found that chain stores in Black and Latinx neighborhoods have on average lower Yelp ratings, about one quarter star lower ratings than those chain stores in White neighborhoods, even holding constant neighborhood demographic factors like market potential, crime, income level, et cetera. Now it’s important to remember that this is an average, right? This one quarter star lower rating. It’s an average across all industries of chain stores. So in some industries, the difference in Yelp star ratings was larger and in others it was zero. The difference was largest in big box and discount stores about one half star. coffee and snack shops, there was a one quarter to one third star difference. And in fast food shops, there was a one third star difference. So in other industries, there was no difference. So for example, for automotive supply, grocery stores, gas stations and cell phone stores, there was no measured difference in the average Yelp ratings between BIPOC and White-located chain stores.

Danny Caine
I feel like I remember reading through your research, there was something, it was like not statistically perfect, but there was something with a four and a half star difference. Was it maybe dollar stores?

Andrea Chandrasekher
Sometimes what can be interesting is taking your larger data set and making subsets and smaller subsets and smaller subsets to ask questions that are interesting to you. However, as you do that, your statistical power, in other words, the number of observations, the number of data points that you have goes down. do that analysis, because it’s interesting and then you have to kind of caveat the results and say well look my data set is smaller when I ask this very specific question but you know here’s here’s evidence that might be interesting to you but is more suggestive rather than conclusive and I believe I was looking at a particular chain of dollar stores I don’t remember what it was or maybe it was all dollar stores together and so the sample size was much smaller but the average star difference was huge. It was like on the order of four stars between the Black-located chains and the White-located chains. But in order to really believe that, I would need much more data.

Danny Caine
That number kind of jumped out at me. I limitations of the smaller data set, but I just wanted to ask you about that. so for many industries and many chains, there is a difference between ratings for stores in the White neighborhoods and stores in the Black neighborhoods. And you’ve done all this statistical work to make sure that you’re not actually seeing market value or theft or anything else, that this is an actual difference in quality of service between chains in White neighborhoods and chains in Black neighborhoods. If it’s not theft, if it’s not market value, what do you think is the explanation? Is this evidence chains are actually providing worse service in Black and brown neighborhoods?

Andrea Chandrasekher
Yeah, so it’s a great question. And I want to be really careful in stating my results, right? What I’ve technically shown in my paper is that within chain, the Yelp ratings for Black and Latinx located stores are lower than the Yelp ratings for White located stores. And that those lower ratings are not due to other potentially confounding factors like market potential or crime. then the question is, right, what is it then that’s causing folks to rate these BIPOC located businesses lower? Is it that the customer experience really is worse there? So I investigated this question in many ways and from many different angles. I’ll just mention a couple here. The first thing I did was look at the text of the actual reviews the multi-regression analysis that I described earlier. I was only looking at the numerical Yelp score ratings, but as a supplementary analysis, I wanted to look at the actual text of these ratings to see what folks were actually complaining about, right? Because if they were complaining about, the traffic around the store being worse for the BIPOC located chain store than for the White located chain store. That’s not something that the chain business is really responsible for. So I was looking at the texts of these Yelp reviews and in my text-based analysis, I found that the reviews for BIPOC located chain stores were statistically more likely to use words like dirty or disorganized, slow, wait, line, and other words with a negative valence. And the White-located stores were more likely, text for those stores were more likely to have words that had more of a positive valence. It’s not slam dunk evidence that things are actually different between the two locations, but it’s suggestive. And then the other thing I did was a hybrid analysis where I used the text review data and the numerical ratings data. Specifically, I used the review text to limit the scope of my analysis to reviews that were very likely to be about the store’s physical appearance or about the service there. So for example, reviews that were mentioned the word bathroom or store, service, employee, manager, cashier, those types of reviews. And what I found was that amongst those reviews that were very likely to be about the store’s physical appearance or the service there, the Black and brown located chain stores had lower average Yelp numerical ratings than the White located stores. Again, suggestive evidence, but admittedly, this is all more suggestive evidence rather than conclusive. It’s from reviews rather than being from evidence about the stores. It doesn’t actually show that the consumer shopping environment is worse at BIPOC located stores nor does it say anything about the ways in which it’s worse if it is indeed worse. Are the floors actually dirtier? If so, how much? Are the wait times in line actually longer? If so, by how much? So that’s the question that I’m turning to now in my research. What is really happening in the actual stores that’s tangibly different in terms of the physical environment and the service environment? So it’s a great question that you’re posing. I just don’t have the answer to that exact question yet. That’s what I’m working on.

Danny Caine
I am very curious to hear more about your current work and where you’re taking this. Let’s just stay with this study for a couple more questions. this just popped into my head. simple question. It seems like you have a huge data set and you’ve done many, many things with this. How long did this take you from start to finish to do all this work?

Andrea Chandrasekher
Yeah, so it took about a year. I used my sabbatical year to assemble all these different data sources and analyze the data. Yeah, a really thorough empirical project takes about a year. Yeah.

Danny Caine 
Okay, wow. Well, amazing work and it’s very So I wonder, my last question about this current do you have an idea of what maybe the underlying causes are behind this phenomenon? I know we don’t have like, you know, 100 % perfect evidence that it’s happening. We have the kind of indirect evidence you were talking about. But do you think this is like part of the business strategy? Do think it’s kind of essential to these large chains gaining so much power and market share to have this kind of inequality and service between neighborhoods and classes of people?

Andrea Chandrasekher 
It’s a really interesting question. I actually haven’t studied the market dynamics enough to know for sure, but my guess is that this is financially motivated. Now, whether or not it’s essential or like a necessary condition to the success of these large chains, I don’t know. But let me say a little bit more here about what I would guess are these firms’ financial motivations. And this is where I think consumer redlining intersects with the retail redlining, which is what we talked about at the beginning of the hour. If these chain businesses can operate some stores, let’s say the ones in BIPOC neighborhoods at lower costs because they hire fewer employees for those stores or they’re sending those stores fewer cleaning supplies, all the while knowing that they won’t lose any, or let’s say many customers because they know there are very few stores in the BIPOC neighborhood that they have to compete with, right? Because of retail redlining, then just speaking from an economics point of view, that’s a very profitable business model, right? These chain store businesses can minimize costs while maintaining the same level of demand. Now, I don’t know for sure if that’s what’s happening, but if it is, ultimately, I think that’s a very short-sighted and suboptimal approach to doing business because I think, and some of my other research shows, there is actually a lot of money to be made in Black and brown communities, but when the physical and the service environment in the stores is so unpleasant, people will eventually, not right away, but people will eventually go elsewhere to shop or they will shop online.

Danny Caine
Yeah, well that intersects really interestingly with some work we’re doing at the Institute for Local Self-Reliance that we’re studying food deserts. And when these chains will swoop in and force the small independent neighborhood focused stores out of business, they don’t have to replace them because there’s no competition anymore. And so all of a sudden, a neighborhood that had six grocery stores in 1960, five of which were independent, now just has a single Dollar General. And one that’s not even a good dollar general at that, if that’s even possible. So we’ll put the links to that research in the show notes so you can have a look, but it definitely kind of ties into what we’re doing at ILSR right now.

Danny Caine
My last question I think, I wanna hear more about where you’re going with this. You said it’s an understudied phenomenon and you mentioned that you’re doing some more work with it. So tell us what’s next.

Andrea Chandrasekher
Yeah, definitely. So I’m working on two other projects eventually all of this will hopefully become a book. as I mentioned earlier, I’m going to be investigating kind of the mechanisms through which consumer redlining might be operating. I’m attempting to determine, for instance, whether BIPOC located chain stores are disproportionately allocated like fewer employees, fewer cleaning supplies, or less employee training, and whether these disparities then lead to the lower ratings that I’ve found in my current paper, one of the two that I’ll be working on next. And the second one is about the law. So I’m investigating what legal remedies, if any, exist to address consumer redlining. The answer here is not straightforward because retail establishments don’t count as places of public accommodations. The protections of Title II of the Civil Rights Act of 1964 don’t apply. what I’ll be doing is exploring the extent to which other laws, so federal and state consumer protection laws, can be used to address consumer redlining. So lots to do and I’m super excited to do yeah, that’s what I’ll be working on next.

Danny Caine
And I actually, have a feeling that ILSR might have some resources for you. I’m realizing that this work intersects much more with what we do than I even thought when I booked the interview. So that’s great. It’s really exciting to hear about. And when the book comes out, we’d love to have you back.

Andrea Chandrasekher
Great. Terrific! Thank you so much. Thank you so much for having me. I appreciate it.

Danny Caine
Yeah, thanks. It’s a great conversation and I can’t wait to read more research.

Andrea Chandrasekher
Thank you.

Danny Caine
See the show notes for a link to Power Play, our report that finds that monopolies exploit systemic racism to build and maintain their power. You can also find resources about why groceries are so expensive, why chain dollar stores are bad for cities, and more. Plus, we’ve got a link to Shaker’s full paper, The Good Starbucks, Consumer Redlining, and Large American Chain Stores. If you’re interested in more stories about race, corporate power, and healthy American towns, please subscribe to Building Local Power wherever you get your podcasts. We take pride in sharing a compelling story about life under monopoly power every other week. We’d love to have you on board as a regular listener. And of course, we always appreciate a share on social media if you like what you’re hearing. This episode of Building Local Power was produced by me, Danny Kane, with help from Reggie Rucker. It was edited by me and Téa Noelle, who also composed the music. Thank you so much for listening.

• Andrea Cann Chandrasekher’s paper, “The ‘Good’ Starbucks: Consumer Redlining in Large American Chain Stores.”
• ILSR’s “Power Play” report