This episode is the tenth in a miniseries of weekly short episodes featuring young scholars entering the academic job market who discuss their latest research. In this episode, Shruti talks with Dr. Archana Dang about her paper, “Role of Time Preferences in Explaining the Burden of Malnutrition: Evidence from Urban India.” They discuss India’s double burden of over- and undernutrition, why financial savings might be a good predictor of obesity, the effects of COVID on India’s obesity levels and much more. Dang is a postdoctoral fellow at the Institute of Economic Growth. Her research interests include the economics of health, specifically issues of overweight and obesity in India. Her work has been published in the journal Economics and Human Biology.
SHRUTI RAJAGOPALAN: Welcome to Ideas of India, a podcast where we examine academic ideas that can propel India forward. My name is Shruti Rajagopalan, and this is the 2021 job market series where I speak with young scholars entering the academic job market about their latest research on India. I spoke with Dr. Archana Dang, postdoctoral fellow at the Institute of Economic Growth. She received her Ph.D. and master’s in economics from the Delhi School of Economics.
We talked about her paper titled “Role of Time Preferences in Explaining the Burden of Malnutrition: Evidence from Urban India.” We talked about obesity in urban India, how impatience and lack of self-control leads to overnutrition, its impact on health outcomes especially during the COVID pandemic, and much more.
For a full transcript of this conversation, including helpful links of all the references mentioned, click the link in the show notes or visit mercatus.org/podcasts.
Hi, Archana. Thank you so much for coming on the show.
ARCHANA DANG: Thank you for having me.
Causes of Obesity in Urban India
RAJAGOPALAN: It’s a pleasure. You’ve done this wonderful field study in Rohini in Delhi, and you’re trying to understand the behavioral causes for obesity and overweight individuals in India. You’re using savings, or rather savings as a proxy, for people who have low levels of patience or low levels of self-control. You find that that explains much of the overweight and obesity patterns, at least in urban areas like Rohini in Delhi. Can you talk a little bit about this connection between impatience or low patience, self-control and obesity in urban settings?
DANG: Factors like fall in the price of calorie-dense food and rise in sedentary lifestyle at work and at home have received a lot of strong theoretical and empirical support as major drivers of an increase in the cases of overweight and obesity in both developed and developing countries. Studies on India by Popkin, Meenakshi and Luhar et al., they suggest that people are moving away from traditional food towards the highly processed foods that are usually very high in fat, sugar and calories.
These factors, they explain rising average body mass index [BMI], which is used as an indicator to classify people into different weight categories, but these factors do not explain why some people gain higher increase in BMI than others. Our data, as well as other past studies as you mentioned, in a very uniform context where, for example, in a relatively urban setup where mostly people are employed in sedentary jobs and they have easy access to high-calorie food, there’s a lot of variation in food intake, which could explain this heterogeneity or variation in BMI.
Since it takes time for food habits to culminate into health outcomes like obesity, I think it’s natural to us that people who discount the future more will be more likely to ignore the future health impact of present food choices that they make in the current period. And that is what I tried to capture in this paper. In this paper, I use a field experiment where I examined and quantified the relationship between behavioral attributes of patience and self-control, and the problem of obesity—or more generally, problem of overnutrition and undernutrition.
I built a simple motivating model which shows that lower patience and lower self-control can result in both over- and undernutrition, which India is currently facing. Data from the primary survey conducted in Delhi provides strong support to my model predictions, and I found that individuals with lower patience and poorer self-control, they make unhealthy food choices, which in turn results in higher BMI.
My sample actually had a very low proportion of underweight adults. Therefore, I could not check the issue of undernutrition using my primary survey. To check that, I use a secondary dataset, which does not have direct measures of self-control and patience. Following the earlier literature on time preferences on behavioral attributes, I use savings as a proxy for patience and self-control, where lower savings indicate low patience and vice versa. I found that these behavioral attributes also explained the problem of undernutrition.
Evidence in the literature suggests that these behavioral measures do not change over time. In my sample, I find that these attributes are not correlated with age. Therefore, the major implication of my results is that these psychometric or behavioral measures, such as impatience and self-control, they are stable, and they are powerful predictors of dietary and lifestyle choices and therefore BMI. Thus, these psychometric measures, they could help us in early identification of individuals who are at a potential risk of becoming overweight or obese later as adults.
Teasing Out the Contributing Factors
RAJAGOPALAN: I want to zoom out a little bit and try and understand the mechanism at work here. You rightly pointed that there could be two, three different factors for these high BMI category individuals. One is, of course, the price mechanism that is at play, which is that very fatty, high-calorie foods are now lower in price. They’re more easily available in urban settings—processed food. That could be one mechanism at play.
The second is, of course, genetics. A very large predictor of some of these health conditions is genetic, which I know is much harder for you to find or control for in your study. Then the third is these behavioral tendencies of lack of self-control and so on. Given the overall problem of obesity, especially in urban areas, how much do you think each of these different explanations contributes?
DANG: I would say that in this paper, I could not capture all these three factors that you mentioned, but my study shows that bias for the present or low self-control leads people to make unhealthy food choices, which in turn increases their BMI. This is evidenced by increasing reliance on convenience food as households are moving away from cooked meals and towards packaged food and restaurant meals. In this paper and in the survey, I collected data on individuals’ food intake and used it to test whether food intake is a mechanism that’s driving the relationship between behavioral attributes and body weight outcome.
I found that indeed, food intake is the channel that explains the relationship between these behavioral attributes and body weight outcomes. Unfortunately, I can’t really say how prices and genetics play a role here because, in this paper, I could not test for that. Of course, in India, there is no such study which looks at these factors. I think these are great questions for future research.
RAJAGOPALAN: You’re essentially trying to measure impatience and lack of self-control, and it’s not quite that easy to do. Now, one would imagine that impatience has to be very specific to the individual. Each individual has different thresholds. Now, when we start digging into it further, you have different levels of self-control for, say, savings versus food consumption versus alcohol or some addictive substance consumption versus Netflix binge-watching.
There are all different levels of self-control. One might be very good at some and very poor at others. You use savings as a proxy. To what extent does that work in an Indian milieu? For instance, in India, a lot of people may not save as much because there is a big social safety net that is provided by extended families or joint families and kin. Some of them might be public-sector employees, in which case there is a fair bit of social security or a welfare net that is provided. How does one think about some kind of baseline savings measure against which you judge self-control or impatience?
DANG: The idea behind doing a survey in Delhi was to get direct measures of self-control and patience. We do that using a field experiment where we ask respondents whether they would prefer a smaller amount today or a larger payment which is available on a later date. If they choose former option more often, they’re more likely to be impatient, but if they choose latter option more often, they are more likely to be patient.
Precisely because of these disadvantages that you mentioned, that savings might be affected by a lot of other factors—for example, health shock aids and a bunch of other factors—for that reason we wanted to do a field experiment where we directly estimate these behavioral attributes. And, moreover, to capture their two preferences, we provided them actual monitoring incentives. And the method that I use in the paper has been used by Meier and Sprenger and Bradford et al.
It has been used a lot, in fact, in the literature for precisely the same reason to estimate these behavioral attributes. Studies have suggested that this method that I use is easier to understand for respondents, as compared to other methods that are there in the literature. These methods are also better in predicting real-world behavior outcome, for example, BMI in our case.
RAJAGOPALAN: Now I’m going to ask you a slightly strange question which—this is a very economist question, actually, so bear with me. We use BMI as a standard. We put people in categories. Now, I don’t want to go into the medical literature of whether BMI is a good measure or not. I know that’s quite controversial, and there’s a lot to say about it.
From a point of view of an economist, could it be that some people are just optimally overweight in the sense that there are certain lifestyles and foods that are worth being fat for? It’s in the sense that—is this a good way of thinking about—are these people optimally lazy or optimally overweight? They seem to be pretty happy with wherever they are. They don’t wish to lose weight. In effect, it may not really be a problem of self-control. They are exactly where they wish to be.
DANG: That’s correct. In the paper, I build up a very simple model which basically tells us that these behavioral attributes that capture impatience and self-control—individuals, they choose their food consumption optimally, based on these behavioral attributes after resolving the tradeoff between their present food intake and its impact on future health outcomes, which in our case is BMI. In the model, I find that some individuals are optimally overeating, relative to moderate eating, which provides an ideal change in BMI.
I find that these people with lower patience or self-control deviate more from the moderate eating, which is the ideal way of eating, as compared to people who are more patient and self-controlled. What I’m trying to say here is that these people are optimally choosing their food intake, and these behavioral attributes affect their food intake, but the level of the deviation from the ideal food intake depends whether they are, say, more impatient or less impatient.
RAJAGOPALAN: I think what I was trying to get at was that as economists, we fundamentally engage in subjective choice, whereas BMI is this objective standard against which everything else is being measured. I guess that’s the question I’m getting at. Is BMI a good way of thinking about these questions of self-control and impatience?
DANG: As you mentioned, there is debate in the literature as to which measure to use for obesity per se, but we actually don’t have any option apart from BMI because weight and height, they’re very easy to use. Given the field conditions, it’s not easy to capture other variables or, say, other way of measuring obesity. That’s the reason that I, in my survey, was able to capture weight and height, although I knew that it’s not a great measure to use obesity.
RAJAGOPALAN: Let me ask you the question a little bit differently. Do you think it’s possible that there’s a slightly different mechanism at play when we think of, say, something like savings related to obesity? The mechanism that you’re using is that people who have high time preference are the ones who are also more likely to over-consume calories, and therefore they have higher BMI. Is it possible to think of this a slightly different way, that the people who are overweight or obese or don’t care too much about their health expect to live not as long, and therefore they save less?
DANG: I think that’s a very good question, but I’m not sure if there is any relation, as I did not test for this directly. Having spoken to many people in the field, my sense is that people don’t think that obesity will shorten their life expectancy and therefore would make them more myopic. Moreover, if this would have been true, we would have expected people to become more impatient with increases in their BMI over time. In the literature, we find that these behavioral attributes tend to remain very stable over time. As I mentioned in the beginning, in my sample these attributes are not correlated with age. I feel that this would not work otherwise.
RAJAGOPALAN: A lot of Indian food habits are very, very tightly linked to caste, at least traditional food habits. I understand that some of these food habits that are so tightly linked to caste are breaking down in urban areas. How much do you think caste has a role to play in this increase in calorie consumption in urban settings?
DANG: In an urban setup where I did my surveys, I actually felt that in terms of different type of food intake, in terms of caste—because I did not actually collect information on caste in the surveys, I actually do not know what was the caste of that particular individual. What I found was that, as I mentioned, there was heterogeneity in food intakes, but I can’t really comment on how this variation can be explained by a variable like caste.
RAJAGOPALAN: What are some of the normative implications of the results that you have found? I want to understand this from two, three different points of view. One, India is one of the lowest countries in the world in terms of the demographic. The second is that India is also one of the fastest-urbanizing nations in the world. Now, these two things put together would suggest that there are a lot of young people who are potentially going to be in the setting or milieu that you’re studying right now. What are some of the implications for long-term health of Indians that you find from a study like this in Rohini? What can be generalized?
DANG: In my survey, I actually do collect information on whether they have any health condition like, say, Type 2 diabetes or cardiovascular diseases. Although they self-reported this information, I could not really take samples to test that. In another paper, where I use a similar survey, I do find that people who are overweight and obese, they’re more likely to have these health conditions. I think this corroborates the medical literature as well.
Economic research in India has traditionally focused on undernutrition, because India is home to largest number of undernourished people in the world. And even currently, undernutrition among children and women is still very high. Overweight and obesity, which was generally considered a problem of a richer country, has now become a major health issue in India, too. Therefore, India is currently battling with coexistence of both undernutrition and overnutrition, which is termed as double burden of malnutrition.
According to the nationally representative dataset called the National Family Health Survey, one in every five adults is either overweight or obese, and it’s no more an urban phenomenon. It has percolated to rural areas as well. In fact, recent study seems to suggest that the problem of overweight and overnutrition has percolated in lower socioeconomic groups as well.
Therefore, overnutrition has not received as much attention in the literature as undernutrition has. This motivated me to work on the overweight and obesity epidemic in India. Also, the motivation to research on obesity is also very personal to me, because I grew up as an obese child in a Punjabi family, and I was body-shamed throughout my childhood and my teenage years.
I was close to 90 kilos when I joined college. I wasn’t the only obese person in my family—so my mother, my elder sister, my cousins, so many of us were overweight. Then obviously, as I grew up and realized my lifestyle wasn’t healthy, and I had to do something about it. So not only did I adopt a healthier lifestyle, I decided to study this issue and ended up getting a Ph.D. in it.
Surprising Field Results
RAJAGOPALAN: I want to ask you a couple of questions about what you found in the field. One of the things I was genuinely surprised by was that in your sample, about 34% of adults are overweight; about 51% are obese. That genuinely surprised me. The proportion of adults who are underweight was almost negligible; it’s less than 1%. Is there something peculiar going on in Rohini or in Delhi? This genuinely surprised me. What do you find when you enter the field and you start studying these questions?
DANG: I was expecting that I would at least find undernutrition in slums. Basically, I surveyed diverse adults ranging from people living in penthouses to people living in slums. Surprisingly, I find that even in slums, the proportion of overweight and obesity was quite high, but this is in line with a lot of new literature that looks at the prevalence of overweight and obesity in terms of socioeconomic group.
For example, a paper by Siddiqui and Donato, which is a very recent paper, they highlight that states like Kerala, UTs like Delhi, Punjab, et cetera, they are comparable to developed countries. In these states, the problem has percolated to poorer people as well. I was surprised a bit, initially, but then when I found out that there were other papers that found similar results, I thought that this corroborates my findings in the field.
RAJAGOPALAN: It’s quite interesting that you talk about these different socioeconomic groups, because one would imagine that if it’s the price mechanism at play, then you can understand why even people in lower-income groups are now sadly facing this problem of overconsumption of calories.
On the other hand, you see that it’s the same with people living in penthouses in Rohini, where you would imagine that there are gym facilities, and they have access to very, very high-quality, healthy, nutritious food, maybe even home-cooked food because a lot of them may have domestic help. But you find a similar pattern there, so it’s no longer the price alone that can explain this. Is that a good way of thinking about it?
DANG: Yes. In my survey, I actually do not find significant differences across these socioeconomic groups. I want to mention that the paper that I was talking about—the recent paper—in fact, it also finds that the prevalence of overweight and obesity has gone down in higher socioeconomic groups, which is similar to what we are seeing in developed countries where poor people are more obese than richer people.
RAJAGOPALAN: Do people talk about wanting to lose weight, or is it a very—you know, the way you think in Delhi like khate peete ghar ke log [someone from a prosperous and well-nourished family]. I mean, being overweight is in one sign of prosperity or good health almost.
DANG: I actually also collect information on whether they exercise or not, and I do find that about 40% of adults in my sample, they do go out and do some form of physical activity. I find this culture more prevalent in higher socioeconomic status and not in lower socioeconomic.
RAJAGOPALAN: Across groups, do people wish to lose weight? Do they wish to be a different BMI or body type than they are?
DANG: Yes. When I was measuring their weight and height, they would say that their weight should have been 10 kilos lower—“I should exercise more, I should eat less.” Yes, they are wanting to lose weight, but I think they are maybe not able to do that.
RAJAGOPALAN: For all the reasons that you said?
DANG: Yes, exactly.
Obesity and COVID
RAJAGOPALAN: I want to shift gears a little bit to this other paper that you’ve been working on. This is of course your work with Indrani Gupta. You have a paper on obesity and its interaction with what happened during COVID and, of course, the global pandemic. We know that individuals with comorbidities have more severe cases of the COVID infection and also a higher fatality rate. What do you find in terms of overweight and obese individuals in India with respect to the COVID waves that we’ve had so far?
DANG: As you mentioned, there is global evidence that seems to suggest that among younger adults, obesity appears to be an independent factor of severe COVID outcomes. For all the reasons that you mentioned, that people with higher body mass index are more likely to have abnormal blood glucose, cholesterol and blood pressure, which leads to health conditions like diabetes, cardiovascular diseases and cancer, et cetera.
This dysregulation due to excess body weight can impair the immune system, which increases the risk of infection and respiratory illnesses per se. Moreover, literature from the earlier pandemic tells us that people who are obese—it negatively impacts their immunity. For example, it was found that individuals who are obese, they were identified to have severe influenza mortality and morbidity outcome in 2009 H1N1 pandemic.
Based on all this evidence, in this paper, we use COVID-19 data for India to examine the relationship between overnutrition and COVID-19 prevalence and case fatality rate due to COVID-19. What we found was that the relationship is strong and significantly positively associated with overnutrition.
In a way, this is in line with what the other studies are finding. I think this positive link between excess weight and mortality and morbidity due to COVID-19 has added another dimension to the impact of prevalence of overnutrition. And therefore I think this needs to be conveyed to the policymakers and to the public.
RAJAGOPALAN: What was initially considered an individual health problem has now also converted into a public health crisis, in the sense that during a pandemic, you want to keep the stress on the healthcare infrastructure and hospitals relatively low. And now what was previously—you have hypertension or heart conditions at an individual-level problem, you could think of whether people are optimally fat or happy and overweight versus not, is now a slightly different thing.
As a follow-up, given that you look at COVID case fatality rates, there’s a fair bit of research going on, and scholars are finding that the fatality in India is actually almost 8 to 12 times higher than what the official numbers seem to suggest. Given that, do you think obesity now is less salient, and it was just a case of a very virulent variant like the delta variant and also very poor healthcare capacity, or are you not concerned about the fact that the official numbers are so low?
DANG: I would say that I’m not an expert on these things, because the idea behind this paper was just to look at a very simple correlation between COVID-19 and obesity, just to see whether in India such a relationship exists or not. In fact, we do find that it exists. That was the whole idea behind this paper. I think you mentioned that a lot of researchers are saying that these numbers are underestimated, but we use these official numbers. Using these official numbers, we are finding such a strong relationship.
Had it been the case that we had all the apt numbers or the correct numbers, we would have found an even stronger relationship. I think what I want to say here is that it’s more important to take into account overnutrition because the lockdown has restricted our movements, including exercising. Newspaper articles seem to suggest that during lockdown, a lot of people are consuming highly processed food. The consumption was significantly higher. Therefore, this elevates the risk of being overweight and obese. This needs to be conveyed to the policymakers and the public. That was a good idea.
RAJAGOPALAN: That sounds really important. Given that this pandemic is not going away any time soon, I think what you’re talking about is an important part of the public health problem that policymakers are dealing with. On a personal note, what have you been up to during the pandemic?
DANG: I watch a lot of programs on OTT platforms. I like watching comedy series. I love “Schitt’s Creek,” and I love watching “Modern Family” as well. I can watch these series again and again and never get bored of them. Have you watched both?
RAJAGOPALAN: No, I have watched neither.
DANG: Then you should watch these series.
RAJAGOPALAN: I think watching comedies in general during the pandemic [crosstalk] is overall a good idea.
DANG: Exactly. That’s what used to keep me sane.
RAJAGOPALAN: This was such a pleasure, Archana. Thank you so much for talking to us and sharing your research with us.
DANG: Thank you so much for the wonderful opportunity, Shruti.