Friday, October 28, 2011

Hockey and development???

From a new paper by Alberto Chong and Pascal Restrepo:
We provide evidence of the Peltzman effect by tracking the professional path of each
hockey player that ended up in the National Hockey League from 2001 to 2006. We
take advantage of the fact that visor use has not always been compulsory throughout
a player’s career, which allows us to compare the change in behavior of users and
non-users of visors when they are forced to use them. We find that whereas the
average penalty minutes per game is 0.8, visors cause a substantial increase of 0.2
penalty minutes per game. Players become more aggressive when forced to wear a
visor, partially offsetting its protective effect and creating potential spillover effects to
other players.
When I arrived to work at the IADB, one of the first stories I heard was how Alberto had an RA the whole summer looking for pictures of hockey players and if their were using visors... now I understand!

If interested in link the Peltzman effect to development related issues, here there is a nice blog post by Dean Karlan.

Thursday, October 13, 2011

Is Jeffrey Sachs in a highway to hell?

Here again is Jeffrey Sachs trying to make a public defense of the evaluation of the Millennium Villages Project (MVP).

He claims that "MVP is based on rigorous measurement, detailed comparison of the villages with other sites, and peer-reviewed science." He claims  list of the scientific publications{the list was deleted!} publications using MVP data, but (with all respect) non of the journals that were in the list before the page was deleted  is specialized in development economics or impact evaluation. Even more, I think he ought to directly address the "scientific counterevidence", like the one from the CGD guys (more about this here).

If the road to hell is paved with good intentions, maybe Sachs is in a highway to hell.... on the positive side, I at least must say that I admire that he is not comfortable with staying at his office to give advice to the world, but he tries to implement some real policies.

Wednesday, October 12, 2011

The essence of treatment heterogeneity

If the effect of a treatment is heterogeneous and the decision to take-up is based on individual characteristics unobservable to the econometrician, then we are in the presence of "essential heterogeneity".  This concept was coined by Heckman and he has been working in the issue for more than 2 decades, producing various papers with co-authors (including one of my professors from Chile), that are generally obscure and not easy to follow for applied researchers.

Martin Ravallion has a new paper that can be a good starting point to understand the problem. He shows that under  essential heterogeneity  the use of a randomized treatment assignment as an IV for take-up can be worse than the naive OLS specification in estimating the causal effect of a program. As a nice extension, he allows the unobserved heterogeneous characteristics to also affect the covariates. 

Tuesday, October 4, 2011

Open data for all

The Center  for Global Development has launched  an initiative for data transparency, in order to allow researchers to replicate their results. While doing this, they took the opportunity to criticize Jeff Sachs for not doing the same in the evaluation of his millennium villages project.