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.