Some Stata resources

Stata is a wonderful program. Regrettably, its functions have odd names and in general working with it is rather excruciating. That’s why I was happy to discover the following resource site from Princeton (particularly the page on time series analysis as I’m not sure if we’re going to cover it). Also, see this Stata/R comparison.

Freakonomics was likely wrong (on some topics)

I just randomly stumbled upon something rather peculiar related to our earlier readings. In Freakonomics one of the chaptersdeals with the effects of legalizing abortions on future crime rates. Authors Levitt and Dubner claim that the drop in crime which appeared out of the sudden in early 1990s can be attributed to the fact that the unwanted children (who are more likely to become criminals) were not born as a result of Roe v. Wade.

While I enjoyed the book, I was a bit skeptical of its methods (especially the excessive temptation to control for a large number of variables), chains of causations, and conclusions. It turns out I was not the only one. Five years ago Christopher Foote and Christopher Goetz, two economists from Boston Fed, found a mistake in Levitt’s original paper; it turned that Levitt and his original collaborator, John Donohue of Yale, forgot to run one of the tests! Running it significantly decreases the effects of abortion on crime rates. The Economist article on Foote’s and Goetz’s research also brings up the problem of controlling (I finally feel vindicated!):

Ted Joyce, a professor at Baruch College (part of the City University of New York), who has had his own methodological disagreements with Messrs Donohue and Levitt, also thinks the debate is stretching the data too far. He points out that if you add controls for 50 states and 12 years—as Messrs Foote and Goetz do, and as Messrs Donohue and Levitt meant to do—you are, in effect, holding another 600 things constant. This robs the data of most of their variety, and of much of their ability to explain anything.

When doing the assigned reading we completely disregarded the possibility of it being incorrect. I can’t recall — have we dealt with selection bias already?

Sources of data

DataMarket is a startup from Iceland which deals with aggregation of various data sets. The company, founded in 2008 and first appearing live in May 2010, currently offers the ability to search more than 13,000 datasets that contain more than 600 million facts; the data providers range from OECD to FIFA to Icelandic Directorate of Fisheries. According to company’s blog most datasets can be exported and therefore used in our beloved Stata.

The second large source of data I recently discovered is Windows Azure. Backed by Microsoft, the services introduces itself as

DataMarket is a service that provides a single consistent marketplace and delivery channel for high quality information as cloud services.

I’m yet to further explore its possibilities, however knowing about these two sources may prove very valuable me and my classmates.

First good article on leadership in a long time

Blogpost for my econ 350 class.

William Deresiewicz writes in The American Scholar that great schools train great hoop jumpers and why it’s wrong.

At Gettysburg College it’s impossible to escape “leaderification.” It seems as though every element of life here is about leadership. Each club, regardless of size or activity, has a Fortune 500-like leadership board; every job for a student workers develop leadership; there is even a new center for leadership with an inspirational mission statement — “There are no bystanders at Gettysburg College. We believe anyone can be a leader.”

There is no definition of the word leader. Generally its meaning is obfuscated in management-speak and tautologies. Leader is the one who inspires and leads others.

The sad fact is that not everyone is meant to be a leader. As a matter of fact, a vast majority of people are not suitable to be leaders. Unfortunately, the status quo feeds everyone’s heads the exact opposite. In my opinion this has two major consequences that dramatically hinder collaboration: everyone wants to have a say in everything; and everyone thinks their idea is better. Non-working free riders is suddenly desirable as one finally doesn’t have to take into consideration their thoughtful input.

Working in groups on important projects has recently been impossible. Merit or knowledge are not taken into consideration as field is artificially leveled through leadership programs. Criticism is absolutely inappropriate (how could one be wrong when they’re constantly told they’re great leaders?), entitlement goes through the roof.

It would have been vastly preferable if the leadership fad died and the capable would learn the art themselves (Google’s Larry Page read books on business as a “young man”), if leadership was about creativity and not being trapped in dogmas, the results of other people’s thinking (as Steve Jobs noted during a Stanford commencement speech in 2005). Such reversal seems unlikely. In the meantime I will continue to dread group assignments.