Thursday, March 24, 2011

This Bears Repeating

Tyler Cowen over at Marginal Revolution cites a tidbit that is worth repeating. Over and over. Folks should IMHO be pounded over the head with this until the lightbulb goes on:
…what separated those with modest but significant predictive ability from the utterly hopeless was their style of thinking. Experts who had one big idea they were certain would reveal what was to come were handily beaten by those who used diverse information and analytical models, were comfortable with complexity and uncertainty and kept their confidence in check.
Cowen cites that from Tetlock and Gardner's Why Most Predictions Are So Bad, at Forbes. This shouldn't be a revelation to many folks, but apparently it is. It's a multivariate world, whether you we like it or not. Base your predictions on one variable, and you'll usually get it wrong.

Tuesday, March 15, 2011

Use Rail to Solve Problems That Trains Are Good Solutions For

A modest proposal to guide new rail projects:
I am a fan of train projects when those projects start with a problem that might be solved by a train, and then work forward to the train. The problem is that in America, those routes are difficult to build, because they're places where there's already a lot of stuff. Rights of way are expensive and time-consuming to obtain, and the project is bound to be blocked by well-organized NIMBYs.

And so the idea seems to have become to build trains where it's possible to build trains, and hope that development follows.  But trains succeed where they are better than some alternative form of transportation.  In the case of Tampa to Orlando, they're worse than a car, and there isn't even any air travel to replace; in the case of Fresno-to-Bakersfield, it may be better than a car for a few passengers, but there are too few passengers to make the trains better than cars for the environment.

Meanwhile, projects that do make economic sense, like an actual high-speed Acela, or Southeastern High-Speed Rail Corridor, are going nowhere.  They might have a better chance of success if rail advocates hadn't abandoned them in favor of building whizzy demonstration projects with dubious economic appeal. 

Sound thoughts via Megan McArdle here, for sure. The Acela, which goes right behind my house, could travel the 222 miles from Boston to NY with a handful of stops in, what, 90 minutes at 150 mph? Or in 2 hours at 120 mph. So, make that happen? Oh, no! Keep it slow, take 4 hours or more, and charge almost as much as a plane ticket. Spend money, accomplish almost nothing.

See, this gets right at Obama's alleged goal of investing in ideas that will really make us better and keep us ahead. Let's get rid of "if we build it they will come" and replace it with"if we really need it we should build it." File this under not brain surgery.

Tuesday, March 8, 2011

So, Do We Do Acid Tests Or Not?

Does evidence matter? If something that seems like a great idea has been shown not to really pan out, are we capable of moving on and trying something else instead. Or is it the case that once an imagined virtue has been embraced by the heart, it must forever endure? Are we impervious to evidence?
Head Start -- the romance of a government program that would provide care, nutrition, education, and skills to impoverished preschoolers in order to erase, to the degree possible, the handicaps poverty imposes. That was the idea in 1965, when Head Start was founded. Lyndon Johnson declared, upon signing the enabling bill that "Today, we reach out to five and half million children held behind their more fortunate schoolmates by the dragging anchor of poverty." Head Start, he promised, would be their "passport" out.

By 1987, even the program's founder, Yale psychologist Edward F. Zigler, declined to claim educational benefits for the program. But as the Thernstroms concluded, "Everyone could agree that poverty was hard on blameless children, so any federal effort purporting to help them was difficult to attack without seeming mean-spirited."....

A just-released study by the Department of Health and Human Services delivers incredibly harsh news about Head Start. A large, nationwide survey of 4,600 preschoolers who were randomly assigned to either the Head Start (experimental group) or no program (control group) were studied on 114 different measures ranging from academic skills to social-emotional development, to health status. The study found no statistically relevant effects from the Head Start program by the end of first grade.

So, what's the deal? What say you? How much evidence is enough for us to be willing to conclude that something that seemed like a no-brainer just isn't helping? Isn't it true that if one has occasion to cite studies in support of their views, then one must be willing to adjust their views when a host of studies indicates something contrary to what you'd hoped?

And what's the best case that can now reasonably be claimed as the virtue of Head Start?