Distinguishing beliefs from preferences in food choice

That's the title of a paper I co-authored with Glynn Tonsor and Ted Schroeder, which is forthcoming in the European Review of Agricultural Economics.  The abstract:

In the past two decades, there has been an explosion of studies eliciting consumer willingness-to-pay for food attributes; however, this work has largely refrained from drawing a distinction between preferences for health, safety and quality on the one hand and consumers' subjective beliefs that the products studied possess these attributes, on the other. Using data from three experimental studies, along with structural economic models, we show that controlling for subjective beliefs can substantively alter the interpretation of results and the ultimate implications derived from a study. The results suggest the need to measure subjective beliefs in studies of consumer choice and to utilise the measures when making policy and marketing recommendations.

We show applications related to tenderness, added growth hormones in beef, and country of origin labeling.  Here are a couple excerpts:

The reason why the conventional ‘reduced form’ model yields a potentially misleading result is that it does not take into account the fact that most people believe that the generic steak is safe. The reason the premium for natural over generic was so low in the ‘reduced form’ model was not because people did not care about safety but rather because they, on average, believed the health risks from growth hormones and antibiotics in the generic steak were low.

and

the results reveal that, at the mean beliefs, consumers are WTP a premium of only about USD 1.68 . . . for a US origin steak relative to the ‘weighted average origin’ steak. The reason why the value is so low is that most people believe the unlabelled steak is highly likely to come from US origin.

The estimates allow us to make interesting calculations like:

of the total WTP premium for guaranteed tender steak, 46 per cent is due to perceived value of added tenderness; the remaining 54 per cent is due to other factors. A similar computation reveals that of the total WTP premium for natural steak over the generic steak, only 38 per cent is due to perceived added healthiness or no hormone use; the remaining 62 per cent is due to other factors.

and

The implication is that when a product has a mixed-origin label, people are
apparently pessimistic, believing the joint-labelled product to have a much
higher likelihood of coming from the less-preferred origin.

Food Conspiracies

This past weekend, I was a guest on a radio show that is broadcast through a network to about 100 stations across the US.  I was talking about my book, The Food Police.  Having done dozens of these kinds of shows over the past six or seven months since the book release, I figured that I've heard just about every question there was to ask.  I was wrong.  

After some standard back-and-forth questions with the host, the line was opened to callers.  Here are a few of the claims I heard - each from a different caller: 

• Adding fluoride to water doesn't prevent cavities and causes joint pain, teeth browning, cancer, and Alzheimers   
• Canola oil is an "unnatural" newly created synthetic product that causes cardiac problems and high blood pressure
• Organic farmers are small farmers; small farmers treat their soil better than large farmers
• With GMOs the genes they inject into DNA.  They are unnatural and become free floating in the soil; 70% of babies have the Bt toxin in their blood as a result of GMOs; the implication is that GMOs are very dangerous
• A new wave of cancer patients are successful fighting cancer by moving to a diet of organic produce 

Some of these are more grounded in reality than others but overall I think I lost a little bit of faith in my fellow man.  I don't mean that in a belittling way.  But it makes me wonder what it is in human nature or what incentives exist in media/internet that would take a little grain of truth and turn it into some of these beliefs that are so at odds with the evidence.  

How do you order at restaurants?

Yesterday, Alexandra Sifferlin at TIME.com wrote about some research I've worked on with Brenna Ellison at the University of Illinois.   Brenna collected data on what people order at restaurants and correlated that with what other people at the same table also order.  Here is a snippet from the TIME blog:

The researchers then created a model to assess how customers felt about their choices. Based on the popularity of menu items, the researchers determined the probability that individuals were satisfied with their choice. Interestingly, this gauge of satisfaction was influenced by expected factors such as price and calories, but also by fellow diners’ menu choices as well. They found that even if a customer initially felt less satisfied about their choice of say, a salad, they felt better about it if their friends ordered an item within the same menu category.

“The big takeaway from this research is that people were happier if they were making similar choices to those sitting around them,” study author Brenna Ellison, an economist at the U of I, said in a statement. “If my peers are ordering higher-calorie items or spending more money, then I am also happier, or at least less unhappy, if I order higher-calorie foods and spend more money.” So if you’re hoping to eat better, try dining with friends who do too.

You can read the actual research paper here.

What Explains the Difference in the Way Americans and French (and Brits) Eat?

I ran across this fascinating paper entitled "Do Prices and Attributes Explain International Differences in Food Purchases" by Pierre Dubois, Rachel Griffith, and Aviv Nevo that is forthcoming in the American Economic Review (an earlier version of the paper is here; a gated forthcoming version can be found by searching here).

According to the paper, French consumers eat about 1777 calories every day.  Americans, by contrast, eat 2103 calories (UK falls in the middle at 1929).  The differences don't end there.  49% of our calories come from carbs; but for the French its only 38%.  A much larger share of French calories comes from fat than those of us in the US (46% vs. 37%).  When one digs a little deeper - it becomes clear why: The French eat more dairy and oils than Americans.

Now, here is the key question which Dubois and colleagues ask.  Why do people in the US, UK, and France eat so differently?   

The authors consider three possible explanations: 

1. differences in prices across countries,
2. differences in the food options available (and nutrient content of foods) across countries, or 
3. differences in what people like to eat across countries (i.e., differences in preferences).

Their data reveals a number of interesting findings.  For example, even though Americans eat more calories than the French, we spend less money doing so ($426/quarter vs. $466/quarter).  Part of the explanation is that food prices are generally higher in the France than the US, but interestingly, it isn't across the board in the ways one might expect.  Fruit and Veggie prices are similar in the US and France.  But, the prices for dairy, meats, oils, and prepared foods are 31%, 76%,  16%, and 18% lower in the US than France.  Interestingly, sweeteners and drinks are priced 39% and 43% higher in the US than in France.  So, one thing becomes apparent: the French are eating more dairy and oils than we are in spite spite of the higher prices.  They must either really like to eat those foods or there must be more of those kinds of foods in France to choose from (they also eat about the same amount of meat as we do - as a share of calories - despite meats being 76% more expensive it France).   

Ultimately,  Dubois and colleagues find that all the above factors matter.  The author's models predict that Americans consume an average of 2212 calories each day (slightly more than the "raw mean").  Then, the authors make some interesting projections.  They calculate that Americans would:

• eat 2158 calories if we were exposed to the same food options (or product attributes) as the French
• eat 1890 calories if we faced the same food prices as the French
• eat 1841 calories if we faced the same food options and prices as the French

The authors conclude:

The estimates allow us to simulate counterfactual quantities purchased by households with preferences from one country but facing prices and product attributes from another country. We use the simulations to learn about the relative importance of preferences versus the economic environment. We find that, the average US household when faced with French prices and product attributes, will purchase substantially fewer calories, bringing the level close to that of the average French household when faced with the same environment. However, the composition of these calories would differ. The simulated change is mostly due to price differences. In contrast, when we simulate the average US household’s food basket with UK product attributes this has a substantial impact on reducing calories, whereas changing relative prices in fact increases calories. From these findings we conclude that the economic environment makes a substantial difference on the consumption basket. However, in general, it is the interaction of preference, prices and attributes that explains the cross country differences.

I find these results interesting because there are many Americans who seems to subscribe to a view that the French have some kind of moral superiority when it comes to food and weight.  I read these results to say that the French are, in large part, just responding to the economic incentives they face.  And while they consume fewer calories than we do, it isn't all that clear they're better off given that they must pay more for many of the foods they desire than do Americans. 

I'm in Italy for the next two weeks. I wonder what I'll eat differently due to differences in food prices and availability?

Most overpriced items in the grocery store

Yesterday I received a phone call from a producer for a major cable news station asking if I'd be willing to come on a show and talk about this story that appeared in Business Insider entitled: "5 Of The Most Overpriced Items In The Grocery Store".

After reading the story, I gave the following response to the producer (slightly edited here for the blog).  Although it would have been nice to have a little air time, I'm happy to report that they decided not to run with the story, at least as it was originally premised.

The story equates “overpriced” with the “percent markup”, which is pretty shaky.  There are a lot of good reasons why the percent mark-up may vary across products that has little to do with being “overpriced”.  For example, differences in demand for convenience and other characteristics, differences in costs of packaging, storage, transportation, etc. will cause differences in the percent markup.  

Nonetheless, let’s play along.

1) Bottled water.  On the surface, it does seem crazy that there is a 4000% mark-up for bottled water.  But, part of the reason for the high percent is that the price of water is REALLY cheap to begin with (so the percent will look very high though the actual dollar mark-up in absolute terms is small).  More importantly, how valuable is convenience to you?  A lot of people are willing to pay an extra buck to have more convenient water and not have to fiddle with refilling and refrigerating a re-usable water bottle.  Who am I to say that an extra $0.50 or $1 isn’t worth it to the person whose paying for it?  If it were really the case that bottle water sellers were ripping us off, why doesn’t some entrepreneur enter the market and start selling cheaper bottled water and corner the market?  The fact is that most of the cost is in the packaging, transportation, etc.  When you buy bottled water, you’re paying for packaging and convenience.

The same arguments apply even more forcefully for pre-cut produce.  Who cares if pre-cut carrots and onions are marked up 40%?  I’m not having to do the work!  That’s an extra $1-$2 I’m definitely willing to pay.  And if someone else can figure out a way to do it for less than 40%, you can bet they’d have my business.  Competition – in the long run- will eventually drive down prices to their approximate costs. 

2)   In general, I would characterize something as “overpriced” if people have mis-perceptions; if they believe they’re getting something from a product that they’re not actually receiving.  Two of the examples in the story potentially fit that criteria: name-brand spices and brand-name cereal.  One way to know whether you’re being fooled by marketing is to do a blind taste test.  It is often the case that our brain is more powerful in influencing how we think something tastes than our tongues.  So, with a neutral friend, try it out: can you REALLY taste the difference?  If not, you may be over-paying.

3)  In this light, there are a number of products that many people have “incorrect” beliefs relative to what scientific studies say – thus, they may be paying a premium for characteristics that they’re not actually recieving.  One example is food with a "natural" claim. A “natural” label is pretty vacuous, and I've previously touched on those issues here and here.  Another example is organic food.  People believe a lot of things about organic foods that just aren't true: that they’re pesticide free, that they support small farms, that they are more nutritious, etc.  I’m not saying there are NO benefits to organic, only fewer benefits than most believe.  A lot of the same arguments apply to local foods.  Chapters 5 and 9 in The Food Police have all the details and citations.