Calorie posting at chain restaurants

Apologies for thin posting.  It's just been too busy for much posting given this blog is pretty low on my priority list these days.  I do have a number of on-topic things I'd like to blog about and hope to have a bit more time in the weeks ahead.  Today was my last lecture for a long time, and that helps a lot.  Just finals, a PhD prelim exam, research and a ridiculous stack of referee reports due....

Anyway. Here's some new interesting research by Bryan Bollinger, Phillip Leslie and Alan Sorensen on mandatory calorie posting in restaurants.  Here's the abstract:

We study the impact of mandatory calorie posting on consumers’ purchase decisions, using detailed data from Starbucks. We find that average calories per transaction falls by 6%. The effect is almost entirely related to changes in consumers’ food choices—there is almost no change in purchases of beverage calories. There is no impact on Starbucks profit on average, and for the subset of stores located close to their competitor Dunkin Donuts, the effect of calorie posting is actually to increase Starbucks revenue. Survey evidence and analysis of commuters suggest the mechanism for the effect is a combination of learning and salience.

This is a pretty big deal because, while a law forced calorie posting in New York awhile back, the new health care law is going to force all chain restaurants to do this.  And it looks like it will have a big effect.  Six percent may not sound like much.  But six percent of 2500 calories a day (a conservative estimate of average consumption) equals over 15 lbs per year.  Of course, we only eat about half our food away from home and while a large portion of that come from fast food, a lot comes from other non-chain restaurants as well.  Also, some may eat more in other places even if they eat less at fast food joints.  But  I think other evidence suggests chain restaurants seem to cause more problems with health than other kinds of restaurants or food eaten at home.

Pretty interesting stuff.  I have a hard time seeing the downside to this kind of mandatory labeling.  While food labels seem to have had no effect on food consumption at grocery stores this looks like an entirely different story.

Oh, evidence in the paper suggests the effect sticks around over time, too.  At least for ten months (the length of the study).

Update: Amid a slew of spam comments (much like the one below which I will delete) I did receive one substantive comment that I inadvertently deleted with the others.  The anonymous  commenter asked what the market failure was that would justify regulatory intervention.

In this case it's not very deep:  almost all market failures come down to information.  Information is a public good:  it's about as non-excludable and nonrival as goods get.  Asymmetric information poses a whole other set of potential problems.

Market failure doesn't necessarily justify market intervention; the benefits of intervention should exceed the costs.  We don't have numbers here, but, for chain restaurants anyway (which is what we are talking about), the cost of calorie posting must be trivial, since it only has to be done once.  It's hard to quantify the benefits in monetary terms. But this study shows the information affects decisions a lot, which is a tell-tale sign of big benefits.  So this seems like a no brainer from an economics perspective.

The benefit-cost analysis of calorie posting at non-chain restaurants with frequently changing menues would be far more ambiguous since the costs would be far greater and the benefits could be smaller.

Leonhardt says now is a good time to buy rather than rent a home

Way back in the midst of the housing bubble's peak, David Leonhardt wrote a nice (and very influential) article about it being better to rent than buy.  His colleagues at the New York Times also put together a very nice interactive calculator that allows people to easily make their own assumptions and calculations.

Now that the bubble has deflated, Leonhardt has changed his tune. Now he says it's better to buy than rent.  At least in many parts of the country.

I am humbled to say that I was actually critical of Leonhardt when he wrote his 2005 article.  I wrote a letter to the editor of the Times, and the Times even got back to me about publishing it, but I was slow to reply since I was on vacation at the time.  My basic point was that real interest rates were low and that it seemed risk premiums were low.  Under these assumptions, basic financial calculations suggested housing prices made a fair amount of sense.  I didn't see prices rising much further, but I certainly did not see an imminent crash, as Leonhardt seemed to.

I still think my calculations made sense.  But my faith in perfectly functioning markets was way too strong.  My conception of risk was way too low.

But now that price-to-rent ratios have fallen and interest rates remain very low, it's really hard to see how buying doesn't make a lot of sense in a lot of places.  Raleigh, however, isn't one of them.  The price-to-rent ratios have increased here while they've decreased almost everywhere else.

Note that I do think one should take the Times' price-to-rent ratios with a heavy dose of salt.  These ratios are hard to measure accurately because the typical rental differs a lot from the typical home.  Making the appropriate adjustments is not trivial.  To invest well you really need to know the local market.  The transactions costs of doing that is what helps to create buying opportunities for small-time investors.

A couple things I think Leonhardt should have emphasized:  The "right" price-to-rent ratio for buying depends a lot on the interest rate and how much one expects rents to grow in the future.  If interest rates go up to 7 or 8 percent, I'd need a much lower price-to-rent ratio.  If interest rates stay low and you expects growth or inflation to increase, then even very high price-to-rent ratios make sense.  Instead of emphasizing these fundamentals, he emphasizes expectations about price increases or decreases. But that sort of gets it all backwards.

Should the government impose limits merchant credit card fees?

Albert Foer, President of the American Antitrust Institute, thinks the U.S. government should impose a limit of 0.5% on the fees charged by banks to retailers for use of credit cards.  Currently Visa and Mastercard charge about 2% of all transactions. 

Is this a good idea?

Interestingly, it seems Visa grew its market share buy raising their fee, which gave banks an incentive to push Visa on retailers and offer nice deals to consumers who used the card in the form of cash back bonuses, frequent flyer miles, etc.

This does seem like a strange market.  Clearly there are market power issues stemming from network externalities and economies of scale. 

When Australia capped the fee is didn't seem to cause many problems, or so Foer claims.  Is that true?  It seems like a simple enough thing to examine empirically.  What share of retail sales in Australia are credit card transactions?  How does this compare to the U.S.?  How did shares change in both countries before and after the change in Australia's policy? It might also be interesting to see how retail prices and the share of credit card transactions changed in the U.S. as the credit card companies upped their fees.

I'm not sure if regulation makes sense or not, or if it does make sense whether it would optimally take the form Foer is suggesting.  But it sure seems like an interesting thing to study.  While I know of some work (e.g. Ausubel) that focuses on consumer behavior and high interest rates, the issue here may be a larger one.  Two percent of ALL transactions is a lot of money!

Life Insurance for Profit: A Story about Vinny and Vito

Update: I wonder if maybe this post is too obtuse.  In case it isn't clear to readers, I was trying to make up a playful parable that mirrors, at least in some ways, the situation with Mr. Paulson (Vito), Goldman Sachs (Vinny) and the SEC (FBI). The implicit fictional characterization of Mr. Paulson (at least) isn't quite fair.  But it did make the story a little more fun to tell.


Vito, an underworld Chicago mob boss, is becoming increasingly concerned about his extended family of criminals.  Tensions with a rival gang have heightened and one of his boys recently got wacked.  Things are bound to get worse in the coming months.

Vito is searching for an upside to this quandary.   So he seeks out his friend Vinny who works in the  insurance business.  Vito tells Vinny about his situation and that he’d like take out life insurance for all the members of his extended family.  A LOT of life insurance.  You see, he’s pretty darn sure at least some of his boys will get wacked.  This way, no matter how the gang war works out, Vito wins.

The key to this scam is to be sure that the insurance company doesn’t know that it is Vito buying the insurance.  So Vinny goes back to the brokerage he works for and tells them about all the new clients he has found.  He doesn’t tell them it's Vito who's buying.  He just describes Vito’s boys--the ones being insured--in generic terms that seem fairly innocuous (male, 45, smokes 2 packs a day, works in security detail, etc.).  The brokerage doesn’t see the scam, brokers the deal, and gives Vinny a fat bonus for selling all that insurance.

To sweeten the deal further, Vinny takes out life insurance on Vito’s boys too.

Vito is elated.  He just can’t lose in his war with the rival gang. In fact, he wins by losing. So war breaks out between the gangs, most of his boys get killed, Vito collects his insurance indemnitities, and retires to the Caribbean, a happy man.  Meanwhile, the city of Chicago burns in the afterglow of the gang war.

An FBI investigation into the incident uncovers the whole scam.  They couldn’t pin any crimes on Vito who ran a tight operation and made sure no crimes could be traced back to him.  But the FBI was able to show quite clearly that Vinny withheld crucial information from the insurance company when he brokered the insurance deal.  If the insurance company knew it was Vito who was buying, there's no way they would have gone for it.  So Vinny was prosecuted and thrown in jail.

The question is: Would the gang war have happened and Chicago burned if Vito was prevented from taking out life insurance on the boys?  Who's really guilty here, Vito or Vinny?

Warmest 12 months in recorded history, and then some

A picture stolen from Brad Delong, who writes:

"We have just experienced the hottest twelve-month period in at least the past thousand years."

So much for emailgate, which, as it turns out, only shows how bad scientists are at handling the politics in which they are immersed.

Anyway.  It's clear the planet is warming and that it is almost certainly due to human caused CO2 emissions.  The recent volcano eruption may cool us off an iddy bit for an iddy while. 

Much harder questions persist about the costs and benefits of curbing emissions and induced warming.  But as time goes on, it seems to me the costs of action decline and the cost of inaction increase.  Would it really be such a crime to develop a global system of property rights to manage this problem?  Maybe while we're at it we can work on development of property rights for the other big global commons, the oceans.

Perhaps the largest issue here is not the price level we should put carbon but rather the development of the institutions to deal with truly global commons.

Interesting times...

We need more evidence on impacts of genetically modified crops

Here's the story at the New York Times.

One key sentence: 

"The improvement in water quality could prove to be the largest benefit of the [genetically modified] crops, the report said, though it added that efforts should be made to measure any such effect"

It amazes and disappoints me tremendously how little we know about the links between agriculture and water quality.  We know agriculture, broadly speaking, is perhaps a key source of all kinds of water pollution.  But which parts of agriculture are most important, and how much policies and structural changes ranging from conservation tillage, buffer strips, GM crops, shifts in cropping patters, and growth in confined livestock operations, really matter is, I think, a complete unknown.  There are some black box models that some have coaxed into spitting out numbers, but I really don't believe them.  I expect few do.

What I'd love to see is hard data linking changes in agriculture to changes in water quality outcomes. This hasn't been done because it is very hard to do.  But it is possible.  And if one could manage to do a good job for one particular policy or structural change, then there would be a litany of follow pieces to show impacts for other policies and structural changes.  Any dedicated number crunchers want to write a dissertation?

Krugman on Environmental Economics

This is a really nice article by Krugman on the economics of climate change. There's not much new here for wonks who follow this stuff closely. But it's the nicest summary of issues that I know of written in a way that anyone can understand. I'll be making this required reading for my principles of economics class (ARE 201).

What I would liked to have seen in the article but didn't:

(1) Some discussion of carbon emissions besides burning of fossil fuels. Land use change, mainly deforestation, comprises about 20% of emissions worldwide. It's a lot harder to price sequestrations from trees, but that form of emission reductions is cheap. Without putting these emissions under the cap we'll have a large loss in efficiency, and possibly a big problem. For example, biofuel production could accelerate deforestation.

(2) Some discussion of how unequal the likely costs of inaction will be. The DICE model and nearly all others obscure distributional issues which is where a lot of the action is likely to be.

(3) A fleshed out argument for how Weitzman's doomsday scenario might actually play out. It is hard to believe that climate change could cause direct catastrophic consequences to the developed world under any scenario. If there is a possibility of truly catastrophic consequences, it will be for certain poorer regions in the world. A permanent dust bowl in the midwestern U.S. coupled with bad agricultural outcomes in other parts of the world, especially Africa, could cause commodity prices to rise a lot. We'll hardly notice higher commodity prices. But the developing world--the parts that remain largely agrarian and will get too hot and the urban poor who rely on cheap grains--could get hammered. This is the doomsday scenario that is most plausible and no model I know of even tries to consider it. Not even Weitzman. It's all representative agent modeling on a global scale. Of course, doomsday for part of the world could draw in a lot of the rest of the world via civil conflict.

Here [after jump] is a condensed version of the article (with an eye toward undergraduates):