April 2023

This month, we feature a conversation with Rebecca Diamond. Rebecca is a Professor of Economics at the Stanford Graduate School of Business. We discuss her research from a series of papers she wrote about food deserts and nutritional inequality.

Leo Feler: Welcome to the UCLA Anderson podcast Series. I'm Leo Feler. I'm a Senior Economist with the UCLA Anderson Forecast. Today our guest is Rebecca Diamond. Rebecca is a Professor of Economics at Stanford's Graduate School of Business. Rebecca thanks so much for joining us today. I'd like to focus on your paper looking at food deserts and nutritional inequality. To start, what is a food desert?

Rebecca Diamond: A food desert is a place or a neighborhood where there are very few or maybe even no full-service grocery stores within one mile or half a mile or five miles… you can have different measures of how intense the food desert is.

Leo Feler: What's the popular view of why food deserts might lead to nutritional inequality?

Rebecca Diamond: There's a theory that one reason low-income households consume less nutritious foods is due to lack of access. If you look in low-income neighborhoods on average, the types of food products on the shelf tend to be less healthy than the average food products on the shelf in higher income neighborhoods. There tends to be fewer full-service grocery stores in low-income neighborhoods, so one theory about why low-income people consume less nutritious food could be the stuff in the stores around them is less nutritious and their access to nutritious foods is restricted.

Leo Feler: In the data you look at and in the work you do, do you actually observe food deserts?

Rebecca Diamond: It's absolutely true that if you look in low-income neighborhoods there are fewer full-service grocery stores, the density of them is lower, and the stuff on the shelves in the average store in those neighborhood including convenience stores tends to be less healthy products. If you define a food desert as no full-service food store within half a mile, there are lots of places that are food deserts. If you define a food desert as no full-service food store within one mile, there are definitely a good chunk of places. But if you go all the way out to five miles, within metropolitan areas of the US, there are not that many places where food deserts exist… they exist, but they’re few and far between.

Leo Feler: What do you observe about low-income residents in these “food desert” areas? Do you see that their consumption is limited to only their immediate surroundings or do you see them traveling further to go to full-service grocery stores?

Rebecca Diamond: We have data that tracks the food purchases of tens of thousands of households across the US. We can see at a granular level exactly what people bought and where they bought it, and a striking fact is that even in low-income neighborhoods or even low-income food desert neighborhoods, the vast majority of groceries are still purchased at full-service grocery stores. That being said, low-income households in food deserts are going to have to travel further to full-service grocery stores, so it may be unpleasant and costly to get there but they do find a way to get there, suggesting that the food that they see when they go shopping is more reflective of a typical full-service food store than just the convenience store that might have been down the block from them.

Leo Feler: These individuals might have to pay a greater time cost or transportation cost because they're traveling further to go to these full-service supermarkets.

Rebecca Diamond: Yes, and there are definitely neighborhoods that are low income that are not food deserts and then there are neighborhoods that are low income and are food deserts. So it's not just if you're in a low-income neighborhood, you're necessarily in a food desert, but food deserts are more likely to occur in low-income neighborhoods.

Leo Feler: You examine data to analyze not only the prevalence of food deserts but also whether the existence of food deserts causes nutritional inequality. Can you talk about some of the data and techniques you use to assess whether food deserts cause nutritional Inequality?

Rebecca Diamond: This is a classic case of supply versus demand. You see in low-income neighborhoods there are fewer full-service grocery stores and you see low-income households in those neighborhoods consuming less healthy food. So one theory could be the lack of supply of healthy food leads to less healthy food choices or it could be the lack of demand or desire to purchase healthy foods means stores in these neighborhoods stock healthy foods at a lower frequency. Just observing this correlation of low-income households consuming less healthy food and those neighborhoods stocking less healthy food doesn't tell you whether the lack of access is causing fewer purchases or whether less demand for healthy food is causing lower supply of healthy food.

To disentangle this issue, we put together a detailed dataset of full-service grocery store openings, when a new store opened in a given area, and we knew exactly the location of where that store opened and when it opened. And then in our consumer purchase data where we can see the granular food purchases and shopping choices, we were able to track the same household over time, before versus after the store opening. If the lack of access to a full-service grocery store were a key cause of less healthy food purchases, then you would expect to see food purchases getting healthier after the full-service grocery store opened in a neighborhood. Now it may not happen overnight; you might not notice the store right away and you might have to change your routine and start shopping there more often, so potentially there could be very long run effects that we can't look at. Right now, we can look at about three to five years after a full-service grocery store opens in a neighborhood.

The thing about using grocery store openings is that it’s clearly a change in supply or access and not a discrete, sudden change in demand preferences, so that allows us to disentangle the question of “is it the lack of demand for healthy food” or “is it the lack of supply or access to healthy food” that’s influencing household purchase behaviors.

The first thing we see is that indeed households notice this new grocery store. We can look at the share of their purchases that occur at the chain that's opening in their area. We see that share tick up substantially, so people clearly notice the store -- it's not as if it opens and these people just don't shop there; they definitely shop at the new store.

But the really telling fact is that the share of their groceries that comes from full-service grocery stores barely changes. It goes up a tiny bit. Basically what's happening is a convenience of transportation. So, these households used to drive a long way to the good grocery store, and now it's very convenient and they drive a short way or walk to the nearby grocery store -- so that's definitely welfare-enhancing: these people prefer to have a nearby grocery store. Perhaps it’s not surprising that people prefer to have a grocery store nearby. When we look at the types of food people choose to buy, perhaps it’s also not so surprising that they buy very similar foods to what they bought before at the further away full-service grocery store.

So that indicates that it's not so much a lack of supply of healthy foods that's determining these food choices, it's the preference and demand differences across high versus low-income households that leads to different food purchases.

Leo Feler: What is the time period you’re looking at when you’re doing this analysis?

Rebecca Diamond: It's mostly in the early to mid-2000s into the 2010s.

Leo Feler: Do you think that access has become more liberalized in that we now have UberEats and DoorDash and Instacart and meal kits and different ways of accessing healthy foods through web platforms in a way that didn't exist in the early 2000s? Do you think that would have an effect on the access that people are experiencing to healthier foods?

Rebecca Diamond: I haven't studied any data on that. So, for that to have an impact on narrowing nutritional inequality between high and low-income households, you would first have to argue that those platforms are especially providing healthy food as opposed to less healthy food. I don't think it's obvious that DoorDash and Instacart disproportionately provide access to healthy food. I think they just provide access to food. We often order a lot of unhealthy food on those platforms. And then number two, those are expensive ways to access food and are much more targeted at the higher income end of the market. My guess would be that if you looked at the low-income neighborhood share of people ordering food on UberEats or DoorDash or Instacart, it’s a small share, and if they are using it, it's infrequent. So, I think it might shift what higher income people are eating, potentially. Maybe it shifts high income people to eat less healthy foods and that narrows nutritional inequality.

I think the lower end of the market is much more driven by Walmart. But again, even if Walmart is stocking healthier foods, that doesn’t mean lower income households will be buying healthier foods. We find that demand is the main focus, not lack of access or supply.

Leo Feler: What does it mean that it's a demand-side issue for why lower-income households are consuming less healthy foods? And how do we incentivize or do we want to incentivize low-income households to eat healthier foods?

Rebecca Diamond: I think that's a hard question. It's hard to think about policies that change preferences. So, there are two ways you can think about this. You can take the economist view, which is we just take preferences as given -- what you like is what you like -- and then we can influence prices potentially. So you could subsidize healthy food to the lower-income end of the market, maybe through food stamps, and try to incentivize people to purchase healthier food. We look at that in the paper, and it's expensive, because there's a strong intrinsic demand difference that you have to overcome. You have to make healthy food really cheap.

Or you could try to think about what shapes preferences for food, which is beyond the scope of our study, but there are other works suggesting that where you grow up as a child has a strong influence on which brands of food you prefer and likely the sort of cuisine and type of food you prefer. So, thinking about childhood as a place where people develop their food preferences, that's a very long run view. That could be a place to intervene as opposed to just subsidizing the heck out of healthy food.

Leo Feler: This also opens the potential for education through school lunch programs. If a lot of these lower income households aren't getting healthy food at home, can we incentivize healthy eating by influencing what's available in school cafeterias?

Rebecca Diamond: I think making school lunches healthier but also increasing awareness about healthy foods when you're a kid – that school lunches are healthy and what's healthy about them and why it's good for your body and for your long run health -- just teaching that in school can potentially influence behavior.

Leo Feler: Excellent, Rebecca thanks so much for joining us and for this conversation. Insightful to know that it's really a matter of changing demand preferences, which is a much harder task than changing supply constraints.