The front-of-package labelling debate: A contrarian perspective based on Gresham-Akerlof’s Law

Posted by Brandon Schaufele in Economics, Food, Health Claim, Regulation on Mar 3rd, 2010 | 1 comment

Front-of-Packaging labelling on processed foods has garnered considerable attention recently.  The key individuals in the discussion are Marion Nestle, of Food Politics (and NYU), and David Ludwig.  Together, they wrote an editorial in Journal of the American Medical Association (JAMA) on food labelling in the US. 

This is an important article as it summarizes what I feel is a consensus position on food and health claims.  (Here’s the link to the JAMA piece.  You can find the original blog post here and some LA Times commentary here.)  I’m going to sketch out this perspective and then present my thoughts on the topic.

Nestle and Ludwig start by outlining several arguments with respect to front-of-package labelling and health claims.  Here are their key points:

1. Few health claims can be verified.
2. Claims based on particular nutritional factors are misleading because food products may also contain unhealthy ingredients (see this story for an example).
3. Information on front-of-package labels is often poorly understood by consumers.
4. “Healthier”, as a labelling term, processed foods are not necessarily healthy.
5. Front-of-package claims can produce a conflict of interest between the regulatory authority and food companies.

Based on these arguments, Nestle and Ludwig come to a conclusion.  They recommend that “consideration should be given to an outright ban on all front-of-package claims.”  The view that health claims should be prohibited is common among many who think about food and health issues (see here, for example).  

I have a different perspective.  It takes several steps to explain, so please bear with me.

The pro- versus anti-health claim argument pivots on one key point.  Food companies support front-of-package labelling because it is an effective form of advertising, increasing product sales.  Anti-health claim advocates dislike health claims because these labels insinuate an implicit approval of the underlying statement by government regulators.  This gives consumers the false belief that the health claims are true.

It is important to state the key issue precisely.  The problem is that consumers trust food and health claims because the government is prominent in the market. 

Let’s be clear on something.  I do not think that health claims are anything more than advertising.  In fact, I agree with most of Nestle and Ludwig’s arguments.  Simply, I do not think front-of-package labelling bans are a realistic solution.  I do however think that there is a counter-intuitive remedy available. 

This is my take on the problem: Health claims are advertising.  Advertising is a method for conveying information to consumers.  This information has some value.  The government is involved in the labelling process for health and food (some claims are allowed; others aren’t, etc.).  Because government agencies are involved in the process, front-of-package claims are falsely viewed as more legitimate than other forms of advertising.  In other words, the government inflates the value of the information conveyed in the health claim.  The solution to this problem is to devalue health claim information.  The best method to devaluing health claim information is by allowing more health claims rather than fewer. 

This conclusion may seem strange.  Let me explain my logic.

The argument is based on something known as Gresham-Akerlof’s Law.  Gresham-Akerlof’s Law is a statement about the value of diverse information. 

Gresham’s Law is a historical dictum about commodity money.  Usually phrased as “bad money drives out good”, the basic idea was generalized by George Akerlof in his famous Market for Lemons paper.  It is this generalized version that I apply.

Akerlof looks at the used car market to illustrate his point.  Here’s how his argument works.  In the used car market, there are high-quality used cars and low-quality used cars (“lemons”).  Sellers know the quality of their cars, but all “claim” that their car is high-quality.  The problem is that defects are unobservable at point of purchase, so buyers cannot tell the high-quality car from the lemon.  For any given used car, the buyer’s best estimate of its level of quality is the average quality.  More importantly, the buyer will only pay for an average quality used car.  This is where the difficulties begin.  Sellers of high-quality cars will not accept an average quality payment for their vehicles.  Instead they will withdraw their car from the market.  This of course lowers the average quality of remaining vehicles, which in turn lowers the buyer’s willingness to pay for a vehicle.  This process continues until the used car market contains nothing but lemons or no transactions are completed.  In other words, the “bad has driven out the good”.

Discussion of used cars may seem off-topic.  How do I get from here to health claims and front-of-package labelling?

When consumers purchase any food that has labelling, they are really buying two things.  The first is the underlying food.  All foods have health attributes and these health attributes exist independent of any claims, labelling or regulations.  The second feature is information.  Information is not limited to health claims.  Logos, brand names, ingredient lists, etc. are crammed onto labels as well.  Generally, this information adds very little value to the product.  A problem arises when the government becomes involved however.

Let’s walk through an example.  Consider two food companies both which make some claim on their products.  Let Company A’s health claim be true or “good”.  Let Company B’s health claim be unsubstantiated or “bad”.  Next, say consumers are willing to pay a premium for a product with a good claim and place no value on the bad health claim.  Consumers cannot differentiate between good and bad health claims.

What does Gresham-Akerlof’s Law tell us?  Bad health claims drive out good health claims.  In other words, because consumers are unable to distinguish between good and bad health claims, they treat all health claims as average.   Moreover, as more bad health claims enter the market, consumers will eventually treat all claims as bad (average quality declines).  This means that even companies that make good health claims cannot elicit premiums for their products.

How does the government fit into this scheme?  Labelling regulations are very complex in both the US and Canada (see here).  Some claims are permitted.  Others aren’t.  In Canada, nutrition information and ingredient lists are required, but it is not obvious that consumers grasp the basic constituents of their diets.  When the government plays a significant role in the food and health market, consumers shift the burden of quality assessment to the government.  By forbidding particular front-of-package statements, regulators add credibility to all remaining front-of-package claims, good and bad.  In other words, by banning some particularly bad front-of-package labels, the government increases the average perceived quality of the remaining claims above their true value.  

If you recall (I know this is a long post), the underlying issue was that consumers trust food and health advertising because the government is prominent in the industry.  I believe that it would be much more successful to devalue all front-of-package claims, making them equivalent to generic advertising, than it would be wade through the regulatory mess of food and nutrition labelling.  This means the government should throw up its hands and permit all health claims.  This would cause consumers to believe that most health claims are nothing more than advertising, thus rendering front-of-package labelling impotent.

Attempts to ban front-of-package labelling via regulation would likely produce a protracted and expensive battle between food companies and regulators.  Loop-holes in new regulations will be found and labelling will become even more complex.  In the end the consumer will not be better off.  I could be wrong on this, but I doubt it. 

It is interesting to note that Gresham-Akerlof’s Law is usually seen as a failure of markets.  However, when it comes to front-of-package labelling, it may highlight a success of information asymmetry. 

I have one final somewhat ironical point.  The proof of my argument (that bad claims drive out good and that this may be a positive thing) comes from the growing number of voices calling for health claim bans.  As this topic garners more attention, consumer confidence in all (good and bad) front-of-package claims is eroded, thus reducing health claims to standard advertising.  See here for evidence.

Update: The FDA recently took action on this issue.  It looks as though they’re limiting health claims and slapping wrists.  See this Food Politics post for the details.