Uppsala, Sweden

Last week I went to a course in regression discountinuity (RD) designs in Uppsala, Sweden. RD is about using shifts in policy to identify causal effects on various outcomes.

A classic example is found in the “effects of schooling on earnings”-litterature, where the schooling decision is endogenous because it partly depends on unobservables such as ability. RD-designs can be used to fix this problem such as in a paper by Matsudaira (Journal of Econometrics, 2007) that exploits a mandatory summer school program for students who score less than some cutoff level on a test. The idea is that students very close to the cutoff point are unlikely to have different ability and hence we can compare students who were just bad enough to get into the schooling-program with students who were just good enough to stay out of the program, and thereby hope to indentify the effect of schooling on wages.

Besides that Uppsala (pop. 167,000) is a very beautiful city located in eastern Sweden about one hour by train from Stockholm. They say that the university is the oldest in Europe after the one in Bologna (which dates back to the 11th century). The city is also now as hometown for Carl von Linné who is world famous for his plant-taxonomy (see Wikipedia’sarticle about him). As an aside: While I was in Uppsala they were celebrating his 300th birthday so the Swedish king and the Japanese emperor was in town and the shopping street was temporarily decorated with a very, very, long roll-on lawn and lots of flowers.

To get the idea of the beauty of the city, you can see a pseudo-live webcast from Uppsala University here.

And one more thing. I always thought that the price of beer in Sweden was much higher than in Denmark, and that that was the reason for all the drunken Swedes in Copenhagen. But a beer in a restaurant in Uppsala costs the same as a beer in a restaurant in Denmark (around 50 SEK ~ 40 DKK) and the purchasing power is roughly the same (a Big Mac is $4.84 in Denmark and $4.59 in Sweden – see the Big Mac Index). So what explains all the drunken Swedes in Copenhagen?

How to write boring papers

Professor Kaj Sand-Jensen has written an artikel in Oikos called “How to write consistently boring scientific literature”. Here are his 10 recommendations:

1. Avoid focus
2. Avoid originality and personality
3. Write l o n g contributions
4. Remove most implications and every speculation
5. Leave out illustrations, particularly good ones
6. Omit necessary steps of reasoning
7. Use many abbreviations and technical terms
8. Suppress humor and flowery language
9. Degrade species and biology to statistical elements
10. Quote numerous papers for self-evident statements

On mobile phones and fish

This week the Economist has a nice story of the value of information. The story illustrates how important market failures, such as imperfect information, are for the market outcome. When you have read the story below then think a moment about the informational value of the Internet…

From the Economist.com:

“YOU are a fisherman off the coast of northern Kerala, a region in the south of India. Visiting your usual fishing ground, you bring in an unusually good catch of sardines. That means other fishermen in the area will probably have done well too, so there will be plenty of supply at the local beach market: prices will be low, and you may not even be able to sell your catch. Should you head for the usual market anyway, or should you go down the coast in the hope that fishermen in that area will not have done so well and your fish will fetch a better price? If you make the wrong choice you cannot visit another market because fuel is costly and each market is open for only a couple of hours before dawn—and it takes that long for your boat to putter from one to the next. Since fish are perishable, any that cannot be sold will have to be dumped into the sea.
Continue reading