The Atlantic has a very nice article summarising the discussion of whether the flu vaccine prevents death among older people. The basic problem in evaluating interventions like these is self-selection, which can make “cohort studies” unreliable. Not really news to most economists, but it seemed to have created quite a fire storm among public health and MD researchers. A very worthwhile read.
Whether this season’s swine flu turns out to be deadly or mild, most experts agree that it’s only a matter of time before we’re hit by a truly devastating flu pandemic—one that might kill more people worldwide than have died of the plague and aids combined. In the U.S., the main lines of defense are pharmaceutical—vaccines and antiviral drugs to limit the spread of flu and prevent people from dying from it. Yet now some flu experts are challenging the medical orthodoxy and arguing that for those most in need of protection, flu shots and antiviral drugs may provide little to none. So where does that leave us if a bad pandemic strikes?
I presented my paper “Natural Hazards and Child Health” and a poster on my paper “The Demand for Sex Selective Abortions” at the Population Association of America’s annual meeting (held in Detroit). It was a great conference in a horrible hotel.
I gave a seminar at the CSDE seminar series Friday 17 April. The topic of the talk was “The Demand for Sex Selective Abortions”. The abstract is here (it is the same paper as I presented at University of Michigan – see below).
I gave a seminar at UBC 17 March on my paper “Gone with the Wind? Hurricane Risk, Fertility and Education“. I am in the process of revising the paper based on comments from seminar participants and referee reports I have gotten. An revised version should be ready in the not too distant future.
I gave a seminar at the Population Studies Center at University of Michigan Monday 9 March 2009. The paper, “The Demand for Sex Selective Abortions,” is still work in progress, but the presentation is here. Note that since this is work in progress methods and results might/will change, especially in light of the very useful comments I got during my visit. The abstract is:
One of the major changes that have taken place in India over the last two decades is a significant shift in the sex ratio at birth, as techniques for prenatal sex determination have become more widely available. There has, however, been little analysis of which factors influence the decision to abort female fetuses at the individual level. Furthermore, the sparse literature does not address the relationship between fertility, spacing and the demand for sex selective abortions, which may lead to biased estimates of the determinants of sex selective abortions. Using data from the three rounds of the National Family and Health Survey this paper relies on the observed spacing between births to examine the determinants of the demand for sex selective abortions. By employing a discrete hazard model it is possible to simultaneously control for the fertility and abortion decisions, while taking account of censoring and unobservable characteristics that might affect either.
Many newspapers in Denmark reported on 17 February 2009 that there has been an alarming increase in the number of abortions among girls below 15 years of age. One paper (Berlingske.dk in Danish) reported it as a more than 200 percent increase over a decade. In 1998 there were 33 abortions among girls less than 15 years of age, while in 2007 there were 113 for the same age group. Looking for an explanation one paper (Jyllands-Posten in Danish) interviewed a doctor who said that the increase was due to deteriorating sex education in schools and lack of communication between parents and girls.
Both of those explanations might be true, but how much has the number of abortions really gone up?
The top graph show the development in the absolute numbers which is what the newspapers based their story on. Now that would be perfectly fine if the number of girls stayed the same over this period of time, but clearly that was not the case (and it takes all of 10 minutes to determine that from Denmark’s Statistical Office). Hence, the bottom graph shows the development in the number of abortions per 1000 girls. The blue line is the rate for abortions per 1000 girls using those ages 12, 13 and 14, while the top one is the number of abortions per 1000 girls using those aged 13 or 14. True, the numbers have gone up, but 1998 begins to look like an anomaly and if that is the case the increase has been about 45 percent.
Furthermore, these numbers are so small there is bound to be a fair bit of noise. Unfortunately, I could not find earlier data on the number of abortions and there is also no on-line data on the number of births for girls below 15 years of age. One of the papers has the numbers for the first half of 2008 and if the number is the same for the second half the rate using girls 12-14 as base would fall back down to 1.06.