Price responsiveness to unhealthy food taxes and healthy food subsidies

I recently ran across this article published in Current Obesity Reports calling for more rigorous methods to assess the effects of soda and fat taxes.  The article makes a number of good points, but it also misses many more complications that should be included for a rigorous evaluation - for example, how do firms respond when subjected to new food policies?

Another example comes to us via this paper in the Journal of Marketing by Debabrata Talukdar  and Charles Lindsey.  If you want to tax unhealthy foods and subsidize healthy ones, the authors suggest regulators may face an up hill battle because consumers are less sensitive to rising prices for unhealthy foods and falling prices for healthy foods than the reverse.

A portion of the abstract:

“The results from multiple studies confirm that consumers exhibit undesirable asymmetric patterns of demand sensitivity to price changes for healthy and unhealthy food. For healthy food, demand sensitivity is greater for a price increase than for a price decrease. For unhealthy food, the opposite holds true. The research further shows that the undesirable patterns are attenuated or magnified for key policy-relevant factors that have been shown to decrease or increase impulsive purchase behavior, respectively. As the rising obesity trend brings American consumers’ food consumption behavior under increased scrutiny, the focal findings hold significant implications for both public policy makers and food marketers. ”

They conclude:

“For public policy makers, our findings imply that the efficacy of economic policy interventions in inducing healthier food consumption behavior will be much more limited than what is expected under the conventional premise of symmetric patterns of demand response. This may warrant that public policy officials lower their level of expectations regarding the effectiveness of so-called sin tax initiatives in curbing demand for unhealthy food.”

Impacts of Dietary Recommendations

Following the government's dietary recommendations may lead to . . . climate change?

New research suggests the following:

“if Americans adopted the recommendations in USDA’s “Dietary Guidelines for Americans, 2010,” while keeping caloric intake constant, diet-related greenhouse gas emissions would increase 12 percent.”

Rather than trying to anticipate the unintended consequences of such recommendations, the study authors want to add another layer on top of the nutritional recommendations

“The take-home message is that health and environmental agendas are not aligned in the current dietary recommendations,” Heller said.

The paper’s findings are especially relevant now because the USDA Dietary Guidelines Advisory Committee is for the first time considering food sustainability within the context of dietary recommendations, he said.”

As I've pointed out before, trying to integrate nutritional and environmental objectives into recommendations involves value judgement that go beyond scientific evidence. Moreover, focusing just on C02 emissions or nutritional composition (as if that's easy to characterize) ignores many other factors.  On a per-acre basis, which crops are the biggest users of pesticides or water? You might be surprised to find out that it is not corn, soybeans, or wheat but rather many fruits and veggies like lemons, strawberries, etc. 

Rather than trying to add layer upon layer to the dietary recommendations, why not respect people's choices?  The price of food reflects the resources used and the demands on those resources.  If the problem is that prices don't fully reflect water use or C02 emissions, then the idea is to think about assigning property rights in a way that that information-aggregating markets help allocate those resources.  But, I suppose it's less fun to let markets allocate resources.  That would take away our power to tell others what to eat.  

Reducing Food Insecurity

Last week, I mentioned the new USDA report on food insecurity.  I also mentioned a WSJ editorial by James Bovard arguing that food stamps don't reduce food insecurity because the number of people enrolled in food stamps has risen dramatically while food insecurity remains essentially unchanged.

I noted that we don't know the counterfactual (i.e., how much food insecurity would have changed had enrollment in food stamps not increased).  And, I also noted that there is some good academic research on the relationship on food insecurity and food stamp participation.

One of the big problems with trying to tease out the link between these two is that they are jointly determined.  That is, I may enroll in food stamps precisely because I'm food insecure. This sort of selection effect will make it look like being on food stamps is associated with food insecurity, but clearly this is just correlation, not causation.

Here is a careful paper published in the Journal of the American Statistical Association that tries to get at the issue:

“Under the weakest restrictions, there is substantial ambiguity; we cannot rule out the possibility that SNAP increases or decreases poor health. Under stronger but plausible assumptions used to address the selection and classification error problems, we find that commonly cited relationships between SNAP and poor health outcomes provide a misleading picture about the true impacts of the program. Our tightest bounds identify favorable impacts of SNAP on child health.”

One of the measures of "child health" is food insecurity, and this research seems to suggest null to positive effects of food stamp participation and child food insecurity.  

A lot of the discussion on the web related to the USDA report seems to be wrapped up in ideological baggage associated with beliefs about the desirability of cutting or expanding the food stamp program (or, for example, utilizing work requirements).  Those who would like to reduce the size and scope of the food stamp program often try to argue that food stamps do not reduce food insecurity and may actually increase it.  My view is that the best analysis doesn't support such an argument.  There may be other good reasons for reducing the size of the food stamp program, but the food security argument isn't one of them.  

Another argument I made in my previous post was that technological development that leads to lower food prices seems a comparatively good strategy for reducing food insecurity.  

As such, I was intrigued to see this white paper by Graig Gundersen at the University of Illinois on food insecurity.  One of the five drivers he discusses to reduce food insecurity is to focus on the importance of low food prices.  

He also writes, when discussing, what food groups can do (or perhaps what they shouldn't do):

“Third, they can view proposals encouraging organic foods and local foods with skepticism. While proposals to encourage, say, local food procurement by supermarkets can have ancillary benefits, these benefits do not generally extend to low-income households because they cannot afford these items. Instead, the benefits are more likely to extend to upper-income households that can afford these items. Moreover, by devoting scarce resources to encouraging the entrance of these into the food supply chain, this diverts resources away from factors that would help low-income households. ”

Calorie Labeling at Harvard

“There have been some suggestions that we need to post calories at schools and certainly in college dorms … it’s not really a good solution, and Harvard recently removed this information for fear of contributing to eating disorders, because students were going in and selecting their foods based on the calories and not necessarily the nutritional value.”

That's from Anding, Roberta A. Professor, Baylor College of Medicine and Texas Children’s Hospital. “Lecture 20: Obesity—Public Health Enemy Number One.” Nutrition Made Clear from the Teaching Company (HT Bailey Norwood; one transcript is here)

Food insecurity remains essentially unchanged

Yesterday the USDA Economic Research Service released a report on the prevalence of food insecurity in the U.S.  Over 14% of US households (that's 17.5 million households or 49 million people) remain food insecure, a number that hasn't much budgeted since the recession began in '07-08.

As pointed out in a Wall Street Journal editorial today, the USDA's measure of food insecurity isn't a direct measure of hunger.  Rather, the measure is derived from responses to a set of survey questions.  Respondents are shown 10 questions (18 if they have children), and if they respond "yes" more than three times to questions like, "In the last 12 months, did you or other adults in the household ever cut the size of your meals or skip meals because there wasn’t enough money for food?" then the household is classified as food insecure.  

Over at the US Food Policy blog, Parke Wilde remarks:

“In previous years, the United States solemnly adopted targets for reducing the prevalence of food insecurity from 12% (the level observed in the mid-1990s) to 6%. As my chart (based on USDA data) shows, this effort to improve U.S. food security has failed. Yet, neither Democrats nor Republicans talk much any more about any substantial realistic strategy for poverty reduction — with serious objectives, quantitative targets, and implementation steps. Though food assistance is of course important, poverty reduction is the most promising approach to improving household food security in the United States.”

James Bovard in the WSJ notes that the food insecurity results are surprising given the rise in food stamp participation over this period

“In 2013 the USDA reported that federal food programs—most notably food stamps provided by the Supplemental Nutrition Assistance Program (SNAP)—“increase food security by providing low-income households access to food, a healthful diet, and nutrition education.” But food insecurity was more widespread in 2013 (14.3%) than in 2007 (11.1%), while food-stamp recipients rose to 47 million from 26 million.”

Bovard makes an interesting and relevant observation.  However, I'm not sure that I fully agree with his characterization of the literature on the (lack of) causal relationship between food stamp participation, food insecurity, and hunger.  It could be that we would have had even higher rates of food insecurity had enrollment in food stamps not swelled.  As econ-speak: we didn't observe the counter-factual.  

A few comments I've read point out that food insecurity would likely have been lower in 2013 had we not experienced higher food prices over the last several years (particularly for protein over the past year).  That's almost certainly is true.  Just last year, two USDA-ERS researchers (two of the same people who authored the recent report on food insecurity) published a paper (which I previously discussed here) in the journal Applied Economic Perspectives and Policy showing that food stamp participants who live in areas with higher food prices are more likely to be food insecure than food stamp participants who live in areas with lower food prices.  They write:

“We find that the average effect of food prices on the probability of food insecurity is positive and significant: a one-standard deviation increase in food prices is associated with increases of 2.7, 2.6, and 3.1 percentage points in household, adult, and child food insecurity, respectively. These marginal effects amount to 5.0%, 5.1%, and 12.4% increases in the prevalence of food insecurity for SNAP households, adults, and children, respectively.”

If we want less food insecurity, one way to achieve that outcome is to have lower food prices.  How to we get lower food prices? Rain would help (but not too much rain).  The primary systematic way to achieve long-term reduction in food prices is through scientific development and technological innovation that increases agricultural productivity.