Blog

Should dietitians endorse specific brands?

There seems to be a bit of a storm over this piece in the New York Times related to the decision of groups affiliated with the the Academy of Nutrition and Dietetics to allow a “Kids Eat Right”  label from the association on Kraft cheese slices (thanks to Kevin Klatt for the pointer).  

I have a few mixed thoughts on this one.   

In general, it seems like a bad idea for an organization that aims to disseminate unbiased information to endorse specific products or companies (or specific policies for that matter), regardless of whether it's Tom's organic asparagus or Kraft cheese slices.  I recognize that, in principle, the label is not an endorsement of the product, but rather an acknowledgement of a contribution to the Kids Eat Right campaign, but that's probably a distinction without a difference to most food shoppers who see the label.

On the other hand, there seems to be a FoodBabe element to the discussions surrounding all this in the sense that there's a lot of hand wringing over an "evil" processed food.  Because the slices are a processed food it must - defacto - be bad, as is any company that makes it.  But, the most evil ingredient listed in the story - "milk protein concentrate" - is about as benign as they come.  I'm not a nutritionist, but seems to me we should be more interested in overall dietary patterns rather than specific foods.

It's little wonder that my recent survey showed little trust in dietary recommendations.  News stories often hype results from eat-that-no-don't-eat-that studies that shouldn't be used to make causal claims.  That changing dietary advice, coupled with a sometimes superior attitude about what people should be eating regardless of cost or taste often turns people off.  Throw all that on top of the large number of booksellers hocking specific diets that claim to cure all our ills, and you've got a recipe for distrust 

Finally, the NYT story mentions the following:

“Over the last few years, the academy been criticized from some of its members and health advocates over what they contend are its overly cozy ties to industry. Companies like PepsiCo, Kellogg and ConAgra regularly attend the organization’s big annual meeting, where they make presentations to dietitians, hold seminars and parties and provide free samples of their products.”

The implication seems to be that dietitians should be free from connections with industry.  But, that's silly.  It's not necessarily a bad idea for food companies to engage with associations like this.  After all, if the members of the association are doing research relevant to the industry and the foods people are actually eating, then that would be reflected in  the industry showing up and contributing at their meetings.  True, one must be cautious of conflicts of interest, but one must also recognize the power of working with the companies actually selling people food to enact dietary change.  

Food Labels - Environmental Edition

I had a recent request on Twitter asking for my thoughts on environmental labels for food.  The question seems to be motivated by recent discussions about the USDA and the Dept of Health and Human Services possibly incorporating environmental considerations into the federal dietary guidelines.  As I've previously noted, this move makes me a bit nervous because it entails non-scientific value judgments about how to integrate disparate issues (health and environment) into a single overall recommendation.  But, even as I said there: 

“It isn’t necessarily a bad idea for the government to convey the scientific evidence on the environmental impacts of producing different foods.”

Of course, that still leaves many unanswered questions.  What do we mean by environment?  Just C02, or is it water quality, or deforestation, or what?  Is the label voluntary or mandatory?  How will food companies respond to the label?  What do consumers understand about the label?  And so on.

In principle, it is possible to imagine something like a nutrition facts panel for environmental issues.  However, the two are not as analogous as might first appear.  First, scientists have a pretty good idea how to measure the fat, carb, and protein contents of food, whereas measuring C02 or deforestation impacts is tricky business with a lot of uncertainty.  Moreover, the nutritional content of a processed food is relatively stable regardless of where the raw ingredients came from, which plant or facility was used to manufacture it, how it got to the store, or how you transported and cooked it.  None of this would be true for an environmental label, which would require more more extensive (and more costly) monitoring and tracing, and if it is at all accurate, one could have two Wheaties boxes that are nutritionally equivalent but with very different environmental impacts.  That may be all the more reason to inform consumers, but the point I'm trying to emphasize here is the much higher cost and greater uncertainty in informing about nutrition vs the environment.  

Finally, an perhaps most importantly, nutritional outcomes are, by and large, what we economists would call "private goods."  The nutrient contents of the foods I eat affect me personally and not others (let's put aside Medicare/Medicare, which is another issue I've touched on here, here, here, and here).  In these cases, the effects of a label on my choices, and ultimate welfare consequences are more straightforward.  Let's compare that to environmental labels, which signal attributes associated with public goods and possible externalities, where we suspect there are likely to be problems with coordination, free riding, etc.  I suspect most economists would tend to look toward getting the property rights or the prices right as the "solutions" in these cases rather than looking toward labels (here's a paper I wrote on that issue).

Finally, I'll note there is a long literature in agricultural economics on food labels - focusing on when and under what conditions labels enhance social welfare.  The results of this literature are hard to summarize (meaning the effects are complex).  Here are a few good places to start if you're interested in the topic.

My answer to the question: should food products contain environmental labels?  I don't know.  There are far too many unanswered questions to say anything more precise than that.

Lost Pleasure

I'm confused by this article by Sharon Begley that appeared in Reuters yesterday.  She writes:

“U.S. health regulators estimate that consumers will suffer up to $5.27 billion in “lost pleasure” over 20 years when calorie counts on restaurant menus discourage people from ordering french fries, brownies and other high-calorie favorites.

The lost-pleasure analysis, which is criticized by some leading economists and public health groups, was tucked into new regulations published last month by the U.S. Food and Drug Administration which require chain restaurants, grocery store chains selling prepared food, large vending machine operators, movie theaters and amusement parks to display calorie counts.

Public health advocates alerted Reuters to the inclusion of the analysis, which they say makes such regulations more vulnerable to challenges by industry because it narrows the gap between the government’s projections of a regulation’s benefits and costs. ”

Here's the problem - I can't find any evidence the FDA did such a thing in relation to calorie counts on menu labels.  Here is the FDA's final rule, which puts total costs at about $1.7 billion, which is far less than the $5.27 billion from "lost pleasure" cited in the article.  Indeed, the net benefits cited in the final rule are only $510 million - far less than the supposed "lost pleasure". Maybe there is another cost-benefit analysis not mentioned in the federal register?

Also, the article makes reference to a paper by Jason Abaluck, but the only paper of his I can find that seems to have any relation to this topic is this one, and it includes no such "lost pleasure" calculations that I see in my quick look at it.  Again, the paper may very well be out there and I've overlooked it. 

What I can find, after a bit of internet searching, is much concern about including "lost pleasure" in relation to tobacco labeling policies from back this summer, including the concern mentioned by some economists about using "lost pleasure" in this context.  

It's clear why many public health advocates don't like the idea of "lost pleasure" in a cost benefit analysis.  However, a good cost benefit analysis needs to include ALL the costs and ALL the benefits; and it must also consider the longer-term second order effects.  Moreover, many of the articles seems to suggest that "good economists" would never include "lost pleasure" in a cost benefit analysis - an implication that is wholly false.   

A lot of the discussion in these articles, including the one in Reuters, seems to suggest that "consumer surplus" is an invalid way to measure the benefits/costs of a policy.  That's baloney.  And, indeed I think you'd have a hard time finding many economists who wouldn't say that any policy analysis of food (or tobacco) taxes or bans SHOULD use "consumer surplus" which captures "lost pleasure" from being unable to consumer the same consumption bundle as was the case pre-policy.    

At issue seems to be the question of whether the same holds true for a mandatory labeling policy.  The argument is that information does not change prices or available options per se, and as such more information can only make consumers better off (more consumer surplus).  On the surface that's true - and I suspect that is the point the cited economists are making in objecting to using "lost pleasure" in this particular context.  However, it is not unreasonable to include "lost pleasure" in a cost benefit analysis of a label or information policy if:

  • food/tobacco companies respond to the new law by removing some options;
  • food/tobacco companies respond to the new law by changing prices; or
  • consumers continue making the same purchases after the policy, but only do so now with more "guilt" (there can only be a value of information if people change behavior; if you just reduce the pleasure they get from buying a product without changing behavior, then there is indeed "lost pleasure").

When is Mayonnaise Mayonnaise?

A lot has been written recently about the lawsuit Unilever has filed against a new upstart Just Mayo made by the company Hampton Creek.

At issue is the fact that the FDA defines mayonnaise as containing egg yolk, and Just Mayo is a vegan alternative that uses yellow pea instead of egg. Unilever claims that Hampton Creek is falsely advertising their product as mayonnaise (even though it only uses the term "mayo"), as the word is defined by the FDA. By the way, you now have a hint as to why Miracle Whip isn't labeled as mayonnaise.

Most of what I've read on the web has been harshly critical of Unilever. This piece in Mother Jones is one example, framing the issue as a David vs. Goliath (or "Big Mayo"), and an NPR piece is titled "Bit Mayo vs. Little Mayo".

I've been interviewed by a couple reporters about the suit (e.g., see this CBC story), and I have to admit to having mixed feelings about the case. I have found it a bit surprising that many of the folks I normally think about as being in favor of consumer protection have been sided with Hampton Creek, I suspect because it is more appealing to take the position against any food giant.

What are the issues?

Going back to some of the concerns over 100 years ago that led to the passage of the Pure Food Act, there have been attempts to regulate the usage of certain terms and labels. The idea is that if you see a food labeled "mayonnaise", for example, that you can be sure it hasn't been adulterated - that food processors haven't watered down the product or used sawdust or some such nefarious substance to reduce cost. These sorts of rules exist for all kinds of food products (here, for example, is a list of approved labeling terms for meat from the USDA).

From Unilever's perspective, they're just following the FDA labeling rules that have ostensibly been created to protect the consumer against fraudulent labeling. Thus, I can see their point of view that Just Mayo is "unfairly" competing by implying they're making a product that does not conform with the FDA guidelines. While this may seem like a silly case, the implications could be much wider; when should a firm be able to use labels with the term "organic" or "free range" or "milk" or "bread"? Should certain standards be met to use these terms so that consumers aren't mislead?

On the other side, these sorts of rules could very well prohibit innovation and competition. I haven't tasted Just Mayo, but supposing it is comparable to "regular" mayonnaise, they've created a process that removes an animal ingredient, which is appealing to a segment of consumers. Moreover, what if a company can create a less expensive, non-egg based mayonnaise that consumers find tasty? Wouldn't the upstart have a hard time entering the marketplace and competing if they are unable to describe their product using word that best fits it? A lot of the labeling rules seem to be just plain bureaucratic overreach, and they seem to do more to line the pockets of lawyers and regulators than protect the consumer.

Ultimately, I have a hard time seeing how these kinds of name or labeling rules make much sense for products like mayonnaise. These are branded products. If a consumer buys Just Mayo and doesn't like the taste as well as conventional mayonnaise, they can go back to their old brand. In econ-speak, these are experienced goods, and firms have reputations that allow consumers to easily search and identify eating satisfaction with different brands. There seems little reason the market couldn't work this out.

Ultimately, I think there are good arguments on both sides of this case, and it isn't obvious what would be the consequences of the unraveling of these sorts of "food purity" laws. Sometimes it's hard to know when consumer protection becomes protectionism.