Monday, May 12, 2008

A Respectable Debate

Andrew Sullivan says that Noam Scheiber says that Obama shouldn't have unmoderated debates with McCain because:
I don't see the upside for Obama.
That's true, there is very little upside for Obama. McCain has very little going for him in this election, and any additional exposure he gets will (most likely) increase his support. It is very easy to suggest things like public financing of campaigns and these unmoderated debates--which would be conducted Lincoln-Douglas-style in front of audiences around the country--when you are the underdog. It is much harder when you are the frontrunner, when there is a cost to doing so.

But I would be disappointed in Obama if he did not try to set a precedent in this race, even if it comes at a cost to his campaign. I'd love to see a series of joint public appearances with McCain. Public financing is a bit dicier, because it is hard to argue that Obama is unduly influenced by money when none of his funding comes from PACs or lobbies and most of it is in small donations. It's hard to know where to draw the line, but I hope that this is an election that shifts the structure of campaigns to a new, better equilibrium.

Friday, November 24, 2006

Reverse Psychology

Everyone knows why stores sell things for $1.99 instead of $2. It's because we tend to give more weight to the first digits in a price rather than the last. So even though we know that $1.99 is functionally equivalent to $2, we'll give the first a bit more consideration.

However, it seems to me that the reverse happens when you are cutting a price. Dell is offering me a deal in which a $799 laptop is cut down to $719. That didn't seem like much of a deal. But I think that if they had cut it from $800 to $719 it would have seemed a bit sweeter. Something about all those 9's obscures the discount.

Wednesday, October 04, 2006

Overstatement of the Year

Quote from the NYT:
Phil Burress, president of the Cincinnati group Citizens for Community Values, said that among Christian conservative voters, the Foley scandal “just sows more contempt for Congress as it becomes nothing more than a playground for sexual perverts.” (emphasis mine)

Tuesday, August 22, 2006

Yglesias gets it wrong

Matt Yglesias has a sort of insane post on marginal tax rates and revenues.

If you look at, say, this chart you'll see that overall federal tax revenue hasn't changed all that much since 1962 -- bouncing up and down between about 16 percent of GDP and about 20 percent of GDP. The top marginal income tax rate, by contrast, has changed a great deal, veering from 91 percent to all the way down to 28 percent. In particular, there was a large structural shift during the Reagan years when the top rate started at 70 percent and has never since gone above 39.6 percent.

The implication being that one can infer a supply-side effect from the combination of a) revenues constant as a share of GDP and b) declining top marginal income tax rate. While this is a tempting idea, it is also rather wrong on two levels.

First, on logical grounds. If supply-siders are correct, then a declining top rate would keep tax revenues constant (or increasing) in dollar terms, but they would be decreasing as a share of GDP. The point of supply-siders is that you cut taxes, the economy grows, and so you can collect the same amount of revenue but from a larger economy. This would in most cases imply that revenue falls as a share of GDP (though, admittedly, this is not a necessity).

Second, on factual grounds. Revenue has stayed constant as a share of GDP not because of a supply-side effect but because the base has broadened: that is, more kinds of income are now taxable and not because of increases in income. In particular, the 1986 tax reform (for the right-wing take on it, see here).

Wednesday, August 02, 2006

Gay Marriage

A brilliant post by Julian Sanchez that deconstructs the argument for a ban on gay marriage.

Thursday, July 06, 2006

I really like this

Greg Mankiw has a reformulated take on the ages-old efficiency vs. equity problem:
The study of economics leaves a person with two strong impulses.

The Libertarian Impulse: Mutually advantageous acts between consenting adults should, absent externalities, be permitted. The ability to engage in such trades is how people in free-market economies achieve prosperity. When the government impedes voluntary exchange, it prevents the invisible hand of the market from working its magic.

The Egalitarian Impulse: The market economy rewards people according to supply and demand, not inherent worth. Markets often fail to provide people the ability to adequately insure themselves against the vicissitudes of life and accidents of birth. We should, therefore, look for ways to help those who end up at the bottom of the economic ladder.

Most economists feel both of these impulses to some degree. The difference between right-leaning and left-leaning economists is how strongly they feel each of them. Right-leaning economists have a stronger libertarian impulse, whereas left-leaning economists have a stronger egalitarian impulse.

[...]

Here is a conjecture: Whenever a policy appeals to both the libertarian impulse and the egalitarian impulse, economists will offer a relatively united view, as they do on the topic of immigration.


I personally have a strong libertarian impulse, but am also very much persuaded by egalitarian arguments regarding accidents of birth and alleviating market-based risk. Most things that happen are random, or at least unpredictable. In that way, the market is a lot like nature. When natural disasters happen, we don't blame the people hurt by them for their losses. In the same way, if free trade displaces textile workers, it is not really accurate to blame the textile workers for not having foreknowledge of what industries would be profitable in the future. It's random.

At the same time, of course, we do blame people for taking on too much risk. Like those who keep rebuilding Floridian beach houses. And there are clear problems of moral hazard with insuring unemployment. There's a very fine line that must be walked...

Sunday, July 02, 2006

Farm Subsidies Transfers

Greg Mankiw points out this Washington Post exposé on farm subsidies. But the particular program that the piece covers is not a subsidy at all. A subsidy is a negative tax on a good. If I pay you an extra $10 for every acre of corn that you grow, I've subsidized your corn-growing activities. This distorts your decision to grow corn because you'll grow more than you otherwise would. You may not even grow corn at all if I weren't subsidizing you. Since growing corn provides no positive externality and markets for agricultural goods are reasonably close to perfectly competitive, too much corn will be grown. What does that mean, exactly? "Too much corn" means that we could potentially improve everyone's well-being by moving some resources from corn-growing to other productive activities. You, the subsidized corn farmer, won't respond to price signals that tell you to abandon corn-growing for a profession that creates greater value. Moreover, the subsidy will squeeze out unsubsidized farmers, like those in third-world countries.

That's a farm subsidy. What the Post's story is really about is a program that was included as farm subsidy methadone in the Republicans' ultimately unsuccessful 1996 attempt to rid the U.S. of farm subsidies once and for all. Cutting farm subsidies is not a Pareto-improving action. Farmers are hurt (and so is Archer Daniels Midland.) They most likely have to find new work and a new home. They have to be trained to do a new job. They may never reach the same level of income that they once had because they've invested so much in farming-related human capital. But removing farm subsidies is a potential Pareto-improvement. It creates so much new wealth that the losers can (theoretically) be compensated for their lost income by the winners (everyone who benefits from lower produce prices, for example.) This program is an attempt to do that. (Though Republicans were not really interested in compensating anyone, they included the program simply to win Democratic votes.) It is not, as the Post makes it out to be, one of the truly pernicious farm subsidies that we should be most worried about.

However, this program is indeed a very bad way of compensating farmers. The program in a nutshell: if you own land that used to be farmed you get paid a per-acre transfer as long as that land is not developed and regardless of whether that land is still being farmed.

To see why this is a bad plan, consider what a good plan would look like. A non-distortionary compensation program would give everyone who was farming at the time subsidies were cut (or even better, a few months before subsidies were cut) a single payment or series of payments equal to the approximate amount lost through the removal of subsidies. This doesn't alter anyone's decision to farm (from the non-transfer, non-subsidy equilibrium) because it doesn't affect the cost or benefit of farming an additional acre of corn. If farming 500 acres of corn after subsidies were cut is profitable, farming 500 acres of corn would be no more or less as profitable after the transfer as it was before the transfer.

The plan as it was implemented has a number of problems. The biggest problem is that the transfer doesn't follow the people that were hurt, it follows the land. Why do we have an interest in whoever happens to own land that was farmed once upon a time? We don't; we have an interest in the people who were harmed by cutting farm subsidies. The transfer should be issued directly to the farmers, not to the land. This problem goes even deeper, however: it distorts the allocation of land. Suddenly, land that had been farmed had a fixed payment attached to it. Each acre of this land basically comes with an annuity and the price of this land is accordingly higher. Since farmers' property is now artificially worth more, this is in fact an incentive to sell land and stop farming. The plan actually creates an inefficiently low amount of farming. A related problem is the provision that land cannot be developed if you collect payments on it. This is simply silly, why prevent land from being used for more valuable purposes?

In this plan farmers are compensated in an extremely inefficient way: attaching an annuity to each acre of land inflates property values by the amount of the annuity. So, when farmers eventually sell their land, they are essentially paid the amount of the annuity by whoever purchases the land. And whoever purchases the land has a strong incentive to leave it undeveloped, even if it would be best to develop the land. But I wish that the Washington Post had pointed out that the basic idea is good. The losers from farm subsidy cuts are no different from the losers from free trade. Both should be compensated for their losses, not only because it is equitable to do so, but also because they will be more willing to support policies that increase the size of the economic pie.

Sunday, June 18, 2006

The Usefulness of Fiscal Policy

Brad Delong writes:
These days, except in exceptional circumstances--in a liquidity trap, when interest rates are already so low that the Fed can't or daren't lower them further, or when the fiscal expansion comes as a sudden surprise that the Fed does not have time to immediately offset--fiscal policy has next to no stimulative effect at all because the Federal Reserve takes steps to make sure that it does not. That's the big reason that claims in 1993 that the Clinton tax increases were going to send the economy into recession were wrong.
Suppose this proposition is true. The Fed always adjusts monetary policy so that aggregate demand is unaffected by fiscal policy. Then what does fiscal policy do?

(1) It changes the makeup of aggregate demand. Larger government deficits increase the supply of bonds, pushing up interest rates. Higher interest rates mean that fewer people want to invest. This is the "crowding-out" effect. Private investment is far more productive than government debt, so for this reason we shouldn't like deficits.

(2) The other macro effects of fiscal policy have to do with aggregate supply (not in the sense of changing G - T, but in the sense of changing marginal tax rates, subsidies, etc.) Lowering most marginal tax rates increases aggregate supply, so does funding research and development, decreasing unemployment benefits, etc.

(3) There are also the micro effects of individual government programs. These are the things that government spending and tax policy are ultimately meant to do like build infrastructure, redistribute income, provide defense, etc. This is a mixed bag: If these provide a public good or correct an externality they are good, if they regulate a competitive industry or subsidize farmers, they are bad.

These are the things that fiscal policy can do. The results for total budgetary policy are sort of ambiguous, but what about just a typical stimulus package of unfunded additional government spending? Deficits would increase, which is definitely bad. Aggregate supply would be unaffected. It is likely to decrease micro efficiency. So something like this is pretty unambiguously bad. (Of course it depends somewhat on the construction of the stimilus package, it could be worse or better.)

However, it seems politically impossible to ever argue that fiscal policy is impotent with respect to aggregate demand for this reason. We may all know that it's true, but it's hard to imagine that any Congressperson could use this against a bill. Suppose Senator A argues that deficits are just bad because the Fed will offset the stimulus. Senator B asks Ben Bernanke, who says that, no, in fact he doesn't stabilize output, only inflation. And since, in politics, what people say is more far important than what people actually do, Senator B wins.

Thursday, June 15, 2006

Google ads

One Google ad in my gmail tells me:
Why You Shouldn't Invest?

The only realistic way to return
5536% in 18 months with one stock.
Haahahahahahahaaaaahahahahaa...so I follow the link and find this:
Through a combination of the 3 proven investment strategies, you would find companies like the following before they skyrocketed:

Strategy (#1 , #2 , #3)

*Medifast, Inc. -- 5536% Profit in just 18 Months
*Ask Jeeves, Inc. -- 1395% Profit in just 8 Months
*SCO Group, Inc. -- 1382% Profit in just 11 Months
The three strategies are "growth", "value", and "insider trading". It appears that they've just pulled some of the highest-performing stocks ever and listed them as results of the proven investment strategies.

So many things are running through my head...efficient markets...survivorship bias...what does economic theory tell me? Contradicting things. On one hand, it's painfully obvious that this is total crap. If I had invested in the whole market in the 90's, I too could claim that I was invested in a stock that returned 5536%. Rational people use the same model I do: the market is essentially random, so returns from any investment tips this company gives will be no better than average, thanks to mean reversion, and probably worse, thanks to stupidity. On the other hand, it appears that the company pushing the ad is somewhat successful. Thus (some) people must not have rational expectations.

Monday, May 29, 2006

Why the Borough of Swarthmore sucks

I'd argue it's stuck in a bad equilibrium. There is no reason that Swarthmore, PA should not be a place full of pleasant restaurants and coffee shops. After all, Williamstown, MA has about the same number of students and, if anything, fewer professionals not associated with the school. Yet it manages, apparently, to sustain many nice restaurants and such.

What's Swarthmore's problem? It's stuck in the suburbs. That is, in Williamstown people and their money are stuck in Williamstown and so businesses open and cater to them. In Swarthmore, however, people can go elsewhere: Philadelphia, or other suburbs. So the money of the residents of the Swarthmore is not stuck in Swarthmore proper. Rather, it gets distributed among other suburbs and the city.

Why won't this change? Well, if you are a restaurant person thinking of opening a nice restaurant, where are you going to go? Given that people already leave Swarthmore to find a restaurant -- no one would consider staying in Swarthmore on the off chance that there might be a nice place in town -- your best bet is to locate where the people go: some other suburb or the city. Even if some people would stay, a good portion of people would by habit go off elsewhere for nice food. And you certainly aren't going to get random people from surrounding 'burbs visitng Swarthmore for food. This is very different from Williamstown, where the people, and the money, are de facto trapped.

How did this equilibrium come to be? I don't actually know, but I'd speculate that is has to do with a) the Borough of Swarthmore not allowing the sale of alcohol in its boundaries and b) the downtown being smaller than other downtowns.

I'd further claim that this has something interesting to say about international trade or urban studies: somehow placing a town in suburbia versus the middle of nowhere can lead it produce much lower quality services because of the opportunity for people to go elsewhere.

Why the bus comes just when you are about to give up on it

Explanation 1: The wait time for the bus follows a normal distribution with a mean of t and some variance. You have some sense of the mean, and perhaps the variance. Once you have waited that mean amount of time, you start to get impatient and want to give up on the bus arriving. However, the probability of the bus arriving at t+1 is much greater than the probability of it arriving at time t+2 and etc. So as you wait past time t, the probability of the bus not showing up very soon becomes vanishingly unlikely.

Redoing this with, say, an exponential process would yield very different results.

Explanation 2: Once you are about to give up, either the bus comes and the bus has arrived just when you are about to give up, or else you do give up and you walk or take a taxi or something. Point being, that it is definitionally true that the bus arrives just when you are about to give up, because otherwise you give up and you never find out how long it takes the bus to arrive.

Why your lane is always the slowest

A drive from Philly to DC makes me realize that your lane is always the slow one because lanes vary in speed randomly, but you're only aware of your relative speed when your lane is the slowest. At the moment that your lane is the fastest, there is a failure of empathy and you don't look around and think, wow, I'm going faster than those poor shlumps. You only look around when you think, poor me, my lane is the slowest. So you only register relative speeds when your lane is the slowest, not when your lane is the fastest. Hence, your lane is always the slow lane.

Wednesday, May 24, 2006

Did you know?

At present, the federal fiscal year commences on October 1st of the preceeding calendar year. And until fiscal year 1976 the fiscal year commenced on July 1st of the preceeding calendar year. This means that between fiscal year 1976 and 1977 there was a transition quarter which does not belong to any year. Oh the fascinating things one learns in Washington.

Monday, May 15, 2006

Alpha Divisions

Brad Delong objects to this Washington Post piece on the grounds that an hour of studying at 3 am the night before an exam is apt to do little to help you. If this is an intro. stat., that isn't actually so implausible. I'd venture that the other objectionable thing about the article is that "the Indian tutor, who said his name was Mike, spent an hour walking Del Monte through such esoteric concepts as confidence intervals and alpha divisions." Now, I know some statistics and confidence intervals are not esoteric, they are rather basic. Yet "alpha divisions"? Alpha divisions I've never heard of, and nor has Google in a statistical context. See alpha is used to talk about a confidence interval, but the concept of alpha division does not exist (so far as I know).

Sunday, May 14, 2006

The White Man's Burden

Henry got me The White Man's Burden by William Easterly on the vague condition that I blog about it. It turns out to be a rather excellent book. Better, I'd say, than The Elusive Quest for Growth, even though that book made me fully appreciate the richness of "thinking like an economist; whereas I'm not sure how much I actually learned from The White Man's Burden.

The argument of the book is that there are two types of approaches to development, planning and searching, and because of the inherent complexity of a society, being a flexible and innovative searcher is rather more desirable than being a blue-print type planner. Characteristic of searching is solutions which arise from a demonstrable need at the local level and respond to the local particularities. This is opposed to the one size fits all approach of the planners. Searchers only try to make concrete improvements; planners are enamored of grand utopian visions of totally transforming society. Searchers are accountable to local needs, and thus have to deliver; planners are only accountable to the grandiosity of their visions and thus never really fail because they still have their visions.

This all goes to argue that Jeffrey Sachs is not to be trusted. While a given intervention may sound reasonable, having a grand plan is fundamentally misguided. This is not only because you cannot have one set of interventions that will transform society but also because it is hard to evaluate the effectiveness of any one piece of a grand plan in that you can never say that this one piece failed because there were so many other pieces and perhaps the whole thing was underfunded anyway. And thus it is very hard to hold the institution accountable. Easterly advocates having development organizations focus on narrow geographical areas, and also focus on narrow goals, such that they gain local knowledge, and can be held accountable.

He acknowledges that development organizations have been effective when they can be held accountable: so there has been great progress in certain kinds of health and education things (child mortality is way down, for example). He also acknoweldges that holding people accountable through measuring outcomes creates a problem in that people work on the measurable thing, when it may be only a proxy for something more important. For example, primary school enrolment is a proxy for actually educating students. And if you only measure primary school enrolment, then there is no guarantee that that enrolment means that the students are learning anything.

Yet for all these broad strokes, the much more interesting aspects of the book are his discussions of history, democracy and free markets. He argues recent development is just a continuation of colonial efforts, both in terms of the foolishness of planning and lack of attention to local specificty, and in how colonialism also mixed good intentions with self-interest. This is a link that very few modern development economists explicitely make. He has a long section about how free markets and democracies do not just happen. There is no magic market dust, or special democracy potion that can instantly, or even forcefully, transform a society. They both rely on norms and customs that take time to develop and have to be based on knowledge of local history. He gives examples of land reforms that worked because they based the assignment of property rights on past practice, and land reforms that failed because they were following some blueprint of a planner in Washington (or wherever). And democracy, well, his point is that democracy isn't just about elections. This all goes against both "shock therapy" -- which he claims he supported when he was at the World Bank -- and the last 50 years of American foreign adventures.

These sections, especially about the functioning of free markets, are really rather surprising to see coming from a neo-classical economist. As Henry says about some economists, Easterly has all the "correct" economic opinions: that is, his normative views are largely those that economic reasoning would imply. And yet he realizes that well-functioning free markets and market institutions aren't the natural outcome of all human interaction, but require substantial nurturing and cultivating and rely on extra-market things that most economists neglect.

Amartya Sen's review of the book criticizes it on the grounds that Easterly spends too much time attacking development orthodoxy and not enough time praising its successes. But that misreads the point of Easterly's book (and his present career): his goal is not really to be an even-handed critic, but to be a gadfly, pointing out the uncomfortable truths about development that everyone acknowledges on some level, and people even write about, but that no mainstream development economist pulls together in a comprehensive critique of the enterprise. The book is both fun to read, and also interesting for the ways in which Easterly reaches beyond economics to try to explain how the development orthodoxy fails.

Tuesday, May 09, 2006

So word-of-the-day is useful sometimes...

Because it turns out that "rapine" is a word with the following definition:
rapine \RAP-in\, noun:
The act of plundering; the seizing and carrying away of another's property by force.
Which means that the phrase "raping and pillaging" should probably be "rapine and pillaging" instead.