Once again, the government claims a victory when they regulated a system that wasn’t broken in order to help a few (in this case, most of whom didn’t need helping) while inconveniencing many more people.  As I will argue, it also promotes an already unhealthy perspective on food consumption that is ruining the eating experience in the United States.

I’m referring to a recent study done on the effects of a New York City law that requires restaurants to post calorie counts for the items they sell.  The Stanford University study looked at Starbucks figures on food and drink purchases between January 2008 and February 2009.  Here’s a Stanford news release with a summary of the findings and a link to the NPR interview that occurred today with one of the study authors.

Basically, the study concluded that the new law caused a 6% reduction in calories per transaction for Starbucks customers.  For transactions over 250 calories, the reduction was 26%.  By perusing Starbucks’ online nutrition data, it’s pretty clear that the difference means that people who ordered a basic coffee or similar drink didn’t change their habits much (6%), while those who ordered a drink plus a food item changed their habits more (26%).  The study authors acknowledge this: they note that if you analyze beverages separately, there is basically no statistically significant change.  Almost all the reduction was due to changes in food purchases.  Keep that in mind; it will be important later.

Now, given this data, the study authors ask — is the effect worth it?  Do the benefits outweigh the costs of regulation?  From an economic standpoint, the cost of putting up menus with calorie counts is relatively small (though many restaurants have claimed this alone is an annoyance).  As for the effect on businesses, Starbucks doesn’t seem to be hurting because of this law.

In fact, the study authors note that for Starbucks restaurants within a reasonable range of a competitor such as Dunkin Donuts, they actually experienced an increase in sales.  (Of course, to confirm what’s going on here, they’d need to correlate this effect with a simultaneous decrease in Dunkin Donuts sales, but they don’t have such data.)  Notably, however, the increase in sales was due to beverages when near a Dunkin Donuts, whereas the loss of revenue that was experienced across the board at all Starbucks was, as already noted, almost entirely due to food.  That seems a giant hole in the presumption that calorie counts are causing Dunkin Donuts customers to go to Starbucks, since the calorie counts in beverages such as plain coffee aren’t that different.  I don’t know why the beverage sales went up more near Dunkin Donuts restaurants, but we shouldn’t assume that it was simply due to a law whose measured effects only changed behavior around food items.

Nevertheless, the authors go on to speculate about whether such regulation is warranted.  They postulate that 25% of calories ingested by consumers come from chain restaurants.  We could argue about such a figure, but let’s assume it to be plausible.  If calories are reduced by 6% across the board for chain restaurants, the average 2000-calorie diet would be reduced by about 30 calories per day.  They think that such a difference is negligible, but for people who are basing entire papers on small variations in calorie counts, they don’t actually consider the larger repercussions.  If you reduce consumption by 30 calories per day, that’s equivalent to 30 * 365 = 10950 calories per year, which is a little over 3 pounds per year worth of excess stored fat calories.  If people actually only changed their behavior in chain restaurants by reducing their intake by 6%, they’d lose an average of 3 pounds per year (or gain 3 pounds less, at least).  Since most obesity is caused by gradual accumulation of fat over many years, such an effect could potentially be quite significant over a few decades.

Of course, such an argument presumes that people don’t make up for the tiny amount of lost calories in some other way.  If they give up 30 calories at Starbucks but eat an extra cookie at home with dinner, the effect is ruined.  Such a scenario seems likely, given the mentality I’ve seen anecdotally in many dieters who say things like, “I didn’t have the muffin at my coffee break, so I can have a cookie with dinner.”  A study would need to be done to assess such factors, but the study authors are probably correct that unless intake is reduced by at least 100 or even 500 calories per day, no significant weight loss will start to occur.  In fact, one could argue that this law might create the reverse effect.  While you’re unlikely to order more than one cookie or muffin at Starbucks when they cost a few dollars each, you might be much more likely to have an extra cookie or muffin at home when you paid only 10-20% (or less) of the cost at Starbucks.  I’m not saying I have evidence to back this up, but it wouldn’t surprise me at all to hear that consumers who buy less at Starbucks end up eating slightly more calories overall when they are less paying less attention to calories.  In the end, the effect would probably be a wash, though.

Now, despite the authors’ focus on these tiny effects, they don’t focus on two major holes in their theories that undermine the results of the entire study.

(1) Why consumers reduced food but not beverage consumption

The study answers this question, but it plays little role in their conclusions.  They did a separate study in which they interviewed people and asked them to estimate the amount of calories in various items at Starbucks.  The results are not at all surprising, given what we already know about how their patterns changed.  The consumers tended to overestimate drink calories while underestimating food calories.  In other words, we have a study that shows that when consumers see a calorie count posted that is higher than expected, they eat less, while if they see a calorie count that is lower than expected, there is no real effect.  Recall that almost all the measurable effect was on food purchases.

Why do consumers overestimate Starbucks drink calories?  I don’t know, but again, anecdotally, it seems most people I know have decided that fancy coffee drinks must have more calories than a Big Mac.  That’s almost never the case.  (In an embarrassing moment in the NPR interview, the author of the study had to correct the host and inform her that no Starbucks beverage has over than 1000 calories, much less the “thousands” of calories the host claimed.  In fact, by perusing the Starbucks nutritional info, it was clear that even the most decadent drinks generally have 300-500 calories, and only a few have just a little over 500.  Still a lot, but no where near “thousands” of calories.)

In any case, it’s very important to note the influence of what consumers already thought about calories in a given item.  The research seems to suggest that consumers would generally only decrease consumption in items where they underestimate calorie content.  This only makes sense if people are rational.  If they want to decrease their calorie intake, they will be most sensitive to places where they didn’t realize that a particular item had high calories.  On the other hand, if they go to a fast food restaurant and already know that sandwiches there have a ridiculously high amount of calories, posting the calories may have little effect.

(2) What segment of the population actually reduced calorie consumption

This is the most important point, perhaps.  Although it is not discussed in detail in the study, the author mentioned in the interview that the greatest reductions in calorie intake occurred among high-income highly-educated populations.  There was only a very small impact on lower income consumers.  We can speculate on the reasons why this may be true, but this is an essential point when it comes to asking the question: is such regulation justified?  Obesity is growing fastest among lower-income families.  High-income families already tend to have a much better track record of eating healthy.  So, in general, this law only seems to impact those who already understand how to eat correctly, while not targeting the population that the law was actually intended to affect.

Indeed, this factor might even offer an alternative explanation for the anomalous Dunkin Donuts statistics.  In restaurants near a Dunkin Donuts, consumption of beverages went up while food consumption went down.  Did the authors even consider the idea that people in such areas might be swapping food calories for beverage calories, rather than simply drawing business away from Dunkin Donuts?  (No, they didn’t.)  Areas with a lot of Dunkin Donuts probably tend to be lower-income areas compared to Starbucks that don’t have such middle-class competitors nearby.  While upper-class consumers might be trying to reduce calories in general — and thus might stop buying food that they think has too many calories — lower-class consumers might be more likely to consider a calorie swap.  If, as studies suggest, they underestimate food calories and overestimate beverage calories, might the effect simply be something like: “Wow — I didn’t realize that muffin had so many calories, but… ooh, I really like lattes, and they actually aren’t quite as bad as I thought, so maybe I’ll get a latte instead of my regular coffee and forget about the muffin.”  In this case, they may actually not be decreasing their overall calorie intake at all (or by a small amount); instead, they are simply swapping one thing for another because the calorie counts give them more to think about.

Again, I don’t have evidence to back this up.  But it seems unlikely to me that in lower-income areas, where more Dunkin Donuts locations tend to be, consumers would suddenly give up their cheap coffee and migrate to Starbucks to buy beverages that cost twice as much or more — all because Dunkin Donuts pastries might be more unhealthy, but then they don’t buy Starbucks food anyway.  (Recall that in the Dunkin Donuts scenarios, there was little impact on food purchases.)  The authors’ conclusions here make no sense.  And once again, there is little evidence that this law is affecting any people other than those wealthy or upper-middle-class populations who are already sensitive to calorie counts.


Why am I against such regulation?  If it’s a small impact, isn’t even that small impact worth it, given that the cost for businesses appears to be negligible?  Won’t this encourage businesses to give lower-calorie options?  This is certainly the argument that the study authors make.

To answer the second question first, here we’re talking about Starbucks.  Not some junky fast-food joint that only carries unhealthy food.  (As other studies preliminary studies have noted the effects aren’t as clear at these fast-food joints; however, other research suggests an impact, particularly on choices made by parents.)  Starbucks is the place where you hear people ordering non-fat decaf lattes (no whip) all the time.  They also tend to offer at least some lower-calorie muffins and such, rather than only donuts and giant pastries.  Even if there is some impact on Starbucks, they already have options for people who want lower-calorie items.  As for a junky fast-food joint, I think we’d need to see a better study to see whether businesses actually need to respond with lower-calorie options.

As for whether the small impact is worth it, well, is that impact what we intended by the law?  We want to fight obesity, but this study seems to indicate that the law actually has little effect on the population sector that is at-risk for obesity.  Instead, we’re just giving high-income folks more to obsess about when it comes to eating, as if worries about natural foods, local organic foods, etc. weren’t enough.

This last point, in the end, is why this law should be abolished.  I don’t want to obsess about calories when I go out to eat.  I’m making the choice to eat out, so I should already be prepared to deal with the consequences.  I oppose calories on menus the same way I support prix fixe menus.  Particularly at more expensive restaurants, I really like the concept that I can just choose a few items, and I don’t need to worry about the cost, which is the same no matter what.  Sure, sometimes that means I don’t get a good deal, but if I wanted a good deal, I’d be cooking dinner at home.  If I’m looking at a menu without prices, I can just feel free to order whatever I want, rather than asking myself whether it’s really worth the extra $5 for some dish.  Now you put calories on my menu, and I’m stuck worrying about both the cost and the calorie content?  No thank you.

One of the many contributing factors to obesity in the United States is that people don’t savor their food.  They don’t sit down, as many Europeans and other cultures around the world do, and spend hours just enjoying a good meal and conversation.  Prices on a menu are already a distraction that makes one worry about economics instead of enjoyment at a good restaurant, but I can understand that some people prefer such a model.  However, I simply can’t endorse the idea that we should give into yet another obsession of calorie-counting to further take away from the enjoyment of a good meal.  If you don’t want to consume calories, don’t order the alfredo sauce or get a giant piece of chocolate cake for dessert.  But if you do, enjoy it — don’t complain to me while you’re eating it that you feel bad about consuming calories.  That detracts from everyone’s enjoyment.

This law takes a cue from those obsessive people who already whine as they consume a rich dish (or immediately afterward) and puts that “guilt” in your face for every dish you might order.  A night out at a restaurant is transformed from a contemplative experience where diners enjoy their food to a balance sheet where we worry about prices and how many calories we consumed today.  We need to educate about better eating, promote moderation, etc.  But giving more fuel to obsessive rich people who are already annoying and can’t enjoy a good meal?  Well, to restaurants who do such things, and to cities who endorse such policies, I’ll increasingly say “no thank you,” and take my business elsewhere.  Or, I’ll eat at home, where I can choose to police my own food if I want to.  (I can cook better than 90% of restaurant food anyway; they’re only getting my business because of the experience.)  But most times, I’ll just follow the credo of moderation, with no need for a Nanny State to watch over me.