Monday, February 16, 2015

Kinky budget constraint

This post explains the previous one (I was googling "kinky budget constraint" and found the funny graph depicted in he previous post...)

My experience is that many people misunderstand taxation by brackets, and assume that the higher marginal tax rate associated with a higher bracket (say, for income taxation) must be paid on the whole income, so that earning one more €/$/£ results in owing a lot more than one more €/$/£ in taxes.

Thanks to The Economist, I have found an example of a very badly conceived taxation scheme where the higher tax rate in higher brackets applies to the whole income: stamp duty in the UK:



Observe the nice bunching of transactions (in blue) just below the thresholds...

Oh, and the new rate (which is nicely linear by parts) will decrease the stamp duty for 98% of transactions.

Kinked or kinky?

Economist's humor...


Saturday, February 7, 2015

The suspect quality of undergraduate education in US Universities

From The New York Times: (my emphasis added in bold)


I talked to a half-dozen of Hugh Moren’s fellow students. A highly indebted senior who was terrified of the weak job market described George Washington, where he had invested considerable time getting and doing internships, as “the world’s most expensive trade school.” Another mentioned the abundance of rich students whose parents were giving them a fancy-sounding diploma the way they might a new car. There are serious students here, he acknowledged, but: “You can go to G.W. and essentially buy a degree.”

I went on the university’s website to look for some kind of data or study indicating how much students at George Washington were actually learning. There was none. This is not unusual, it turns out. Colleges and universities rarely, if ever, gather and publish information about how much undergraduates learn during their academic careers.

Flu drugs are not (that) effective ... when assessed properly

The CDC released figures Thursday showing that this season’s influenza vaccine is only about 23 percent effectiveand the federal agency recently sent a letter to doctors, urging them to write more prescriptions for influenza drugs.
“Antiviral flu medicines are underutilized. If you get them early, they could keep you out of the hospital and might even save your life,” CDC Director Tom Frieden told reporters Jan. 9. Flu season lasts, on average, about 13 weeks. We’re about midway through a particularly bad one.
(...)


The discrepancy between the CDC’s assertion that the flu drugs can prevent complications, hospitalizations and perhaps deaths, and the FDA’s insistence that the drugs have only been shown to cut the amount of time that symptoms persist comes down to how they weigh the evidence. The FDA requires randomized clinical trials, the gold standard of evidence in medicine, while the CDC also relies on non-randomized reports. “Observational studies from many countries have consistently found that early oseltamivir [Tamiflu] treatment of influenza patients reduces the duration of hospitalization and risk of severe outcomes such as intensive care unit admission or death,” CDC spokeswoman H. Amy Rowland told FiveThirtyEight in a statement.
But physician Tom Jefferson of the Cochrane Collaboration, an international organization that assesses medical evidence, calls the studies that the CDC is using to tout the drugs poorly designed and unreliable, pointing out that they’re funded by the drugs’ maker and led by researchers tied to the manufacturers. More rigorously designed trials do not back them up. For example, a study published last month that randomly assigned 400 hospitalized patients to receive either the newly approved peramivir (Rapivab) or a placebo was halted early for “futility”when it became clear that patients receiving the drug weren’t benefitting.
(...)
And if the drugs are simply reducing fevers and making people feel better via some other mechanism, then it’s possible they’re nothing more than expensive alternatives to acetaminophen and ibuprofen, which alleviate aches and reduce fevers. (Tamiflu retails for about $130, Relenza goes for about $67, and Rapivab costs about $1,000.) Right now, there’s not enough evidence to draw firm conclusions on the drug’s actual mechanism of action, and Jefferson, his colleagues at Cochrane and editors at the British medical journal BMJ are embroiled in a years-long effort to obtain complete clinical trials data from the maker of Tamiflu.

Friday, February 6, 2015

Recherche

Salle de conférences de l'UQAM ...... j'y ai trouvé ma nouvelle devise (merci à Alice F pour la photo)

Question: qui est A.B.?