## Tuesday, December 3, 2013

### Presidential Party and Real GDP

I see Matt Yglesias has a post about whether the economy does better under republican or democratic presidents. Here's the graph he should have included:
I started at Eisenhower in 1953 because the data series itself starts abruptly in the middle of Truman's term. More over, as inaguration starts a thrid of the way into first quarter, I have adopted the convention of counting first quarters under the previous president--this is certainly the most reasonable option, since it takes time for presidents to start enacting their own policies. Also, I used Real GDP instead of Real GDP per Capita because the latter preferred series only starts in 1960. This substitution is valid if (as I suspect is true) population growth is uncorrelated with presidential parties. What the figure shows is that every democrat has outperformed every republican with two exceptions:
1. Reagan actually managed to do a little better than Carter, and
2. Obama, dealing with the recovery from the worst recession by far in this series, has done worse than everyone.
It is easy to make excuses here. For example, republicans will want to blame Obama for the slow recovery, while Democrats will want to excuse Obama on the basis that Bush II caused the recession in the first place. Similarly, republicans will want to excuse Nixon and Ford, who had large oil shocks to deal with, while Democrat's certainly have plenty of policy grounds to criticize (price controls? really Nixon? and how about Ford's comically dim-witted "Whip Inflation Now" strategy?). The sample size really isn't large enough to rule any of these hypotheses out.

But, just remember that looking at this data and claiming that Republicans are actually better for the economy is a pretty extraordinary claim that requires extraordinary evidence. There were 6 republican presidents and 5 democrats. 5 out 6 republicans beat only 1 out of 5 democrats. And that 6th republican hardly did any better, beating only 2 out of 5 democrats.

## Monday, December 2, 2013

### Subsidizing private health insurance

Terence Chai Cheng has a new article making it's way to the pages of the Journal of Health Economics (there is an older ungated version here) using evidence from Australia to show that in countries with both a universal government-funded public health insurance and a private health insurance market, reducing subsidies (ie tax-exemption) for employer-sponsored private plans results in a net reduction in government healthcare spending. Now it may seem obvious that eliminating subsidies reduces government spending, but this is actually non-obvious because we must remember that these countries also have a government-funded universal healthcare system to fall back on--we might have hypothesized that the reduction in subsidies would be offset by increases in the cost of public health insurance.

That is more or less the line that some other papers studying "dual" systems have argued. From Chetty and Saez (2010), for example:
"Taking crowd-out into account lowers the estimate of G(b) by a factor of more than 100. An analyst who ignored crowd-out and applied existing formulas (e.g., Chetty 2006a) would infer that a \$1 million expansion in public health insurance programs would generate \$280,000 in surplus net of the required tax increase needed to finance the expansion. This analyst would mistakenly conclude that an expansion in the overall level of public health insurance would yield substantial welfare gains. Taking crowd-out into account implies that we are near the optimum in terms of aggregate public health insurance levels, as a \$1 million across-the-board expansion would generate only \$1,700 in net social surplus."
Now, Chetty and Saez are talking about the US, not Australia, and they are talking about measures social wellbeing, not government expenditures. Still I think their findings add some relevant perspective.

Cheng notes that there are two other countries--the UK and Spain--that have the same dual system with subsidized private insurance alongside universal public health insurance. But in a somewhat older paper, Amy Finklestein reminds us that Quebec was once also a member of the dual health systems club, with both a universal Medicare public health insurance, and a subsidy for employer-sponsored private health insurance. Now, the Canadian and Australian health systems are very different, so it isn't at all surprising that the elasticity of private insurance with respect to the public subsidies is wildly different in the two countries. What is somewhat surprising is that the elasticity was actually much a much bigger magnitude in Canada, where a 1% reduction in subsidies lead to a 0.5% reduction in private insurance coverage, than in Australia where it results in just 0.13% reduction in private insurance coverage. Finklestein does not do the public expenditure analysis that Cheng did (and I'm not sure "public expenditure" is necessarily the right metric to care about), but comparing these elasticities it certainly seems immanently plausible that the Canadian government saved money by reducing subsidies for private coverage.

## Sunday, December 1, 2013

### Sunday Night Music

Allow me to add some rock & roll (and some twentieth century, for that matter) to the blog's music selection. In honor of Mr. Lou Reed, who we lost a few weeks back, I've selected what might be the Velvet Underground's single most energetic number: What Goes On, off their eponymous third album.

The driving guitars are matched in fury only by perhaps "I Heard Her Call My Name" from White Light/White Heat, the preceding album. However I find this track much more melodic, which is reflective of the album of a whole. The Velvet Underground, a personal and dear favorite of mine, can be seen as a dramatic shift away from the wild experimentation the Velvets had become noteworthy for. Here we note the absence of John Cale, who represented the most avant garde taste in the group, and the curtailing of Andy Warhol's role in production. What we're left with is an utterly listenable collection of ten songs that ranges from feisty (this track, "Beginning to See the Light"), gorgeous ("Pale Blue Eyes", "I'm Set Free"), and cute ("Story of my Life", "After Hours") to mildly subversive ("Jesus"). The transexual tale "Candy Says" seriously bolsters their already solid counterculture cred, while "The Murder Mystery", the most experimental but probably weakest song on the record, features overlapping vocals from each band member. The result is somewhat cacaphonous, but never short of intriguing.

# Mortgage Calculator

Payment:\$Compound Frequency:(monthly) Annual Property Taxes:\$
Nominal APR:%
Term Length:Years
Down Payment:\$House Price:\$

## Friday, November 29, 2013

### Is there any sane reason for Black Friday?

Happy Thanksgiving!

Actually, yesterday was thanksgiving, but I didn't do any blogging yesterday. That makes today Black Friday, a shoppers' holiday on which all the stores are packed with people. There are usually many injuries.

Something that I have always wondered about is whether there is any sane reason to shop on black friday. In principle, there are supposed to be all kinds of crazy good deals everywhere. But as an economist, I'm skeptical.

Is there any economic reason to suppose that there are tons of great deals to be had on black friday? I don't see it. For one thing, no producer anywhere has any incentive to lower any of their prices below what they paid for it--it's better not to sell than sell at a loss! Store markups range from 20 percent to 35 percent over wholesale costs, and that has to cover both profits and store operation costs. Thus, if you are looking at a "discount" of more than 30 percent, there's something fishy going on. I can think of only exceptions to this rule:
1. trying to unload undesirable or damaged products
2. fishing for customers using a few good deals as bait
These are the only reasons I can think of for a store to sell a product below cost. The first point is the only sane reason for discounts, in my humble opinion. If you can't sell a product, you lower the price. If that discount happens to bring the total price down to below your willingness-to-pay, then it makes sense for you to buy it, and it makes sense that the store would be willing to take a loss because it turns out not to be worth as much as the retailer had predicted it would be worth.

The problem with the first reason is that there is no inherent reason to wait until Black Friday to offer such a discount--you can sell a product at a loss any day of the year. What does change on Black Friday is that there is a huge increase in demand, which means that your probability of selling a product at a given price is much much higher on Black Friday than most other days of the year. Hence, you can probably sell an undesirable or defective product at a higher price on Black Friday than any other day of the year. This is, I think, a satisfying answer to the Black Friday puzzle because it would be individually rational for the retailer to offer such a discount specifically on Black Friday, since the discount is acutally less than what they would normally have to offer, and also rational for shoppers to shop for such discounts on Black Friday, since they know that retailers won't offer it on other days of the year. However, the welfare implications of this are somewhat disturbing: ultimately, this means that the existence of Black Friday makes consumers in the aggregate worse off, since prices are actually higher than they otherwise would be.

The second reason is probably also at least somewhat true. Stores often do offer money-loosing bargains that are designed to generate publicity and bring customers into their store as opposed to their competitors. There's nothing wrong with that, per se, and if you happen to be one of the few who get the bargain while supplies last, good for you. But, let me offer a bit of a warning to bargain hunters: stores would only do this if they expected it to cause people to spend more than they normally would have in their store--be on the lookout not just for discounts, but also price hikes. Essentially, this scheme is one of cross-subsidization where they tease you with a bargain or two, but really aim to overcharge on all the impulse buys. Moreover, remember that stores can and do raise their prices in order to offer discounts.That's not to say that it isn't rational to hunt for good deals on Black Friday, but you should realize that they are counting on behavioral economics leading unsophisticated shoppers to overspend. The final price, not the percentage discount, is the only thing that matters. Ever. Period.

I often hear about a third explanation for Black Friday discounts: price discrimination and customer loyalty. I think most people have an intuitive sense of how offering discounts is a way for retailers to essentially charge one price for people with low price elasticities (ie, inattentive loyal customers) and a lower price for customers with higher price elasticities (bargain-hunting defectors). The formal model I think about here is this one published in the AER. In this model, period sales give rise to implicit price discrimination as retailers are essentially able to charge a regular high price to inattentive people who don't like to shop around, without totally loosing their sales to bargain hunters who will still show up whenever they hold a sale. However, this can't explain Black Friday, because in this model holding a discount is profitable if and only if relatively few retailers are holding discounts. Black Friday cannot arise in this model, because holding a discount while everyone else is having a discount would be a money-loosing venture on the part of retailers, who would therefore decline to participate.