More Fun with Statistics

I didn't really pay much attention to the typical 24-hour partisan finger-pointing flurry accompanying the stats showing an uptick in US infant mortality rates.  However, looking back on it, it is a good lesson about how statistics are routinely misused in this country (via Captain's Quarters).  Critics of the administration and the health care system used the statistics to try to show something is wrong with the US. 

Two weeks ago, Nicholas Kristof wrote a column on the first increase in the American infant-mortality rate in decades, taking the opportunity to excoriate Americans and the Bush administration as uncaring and unresponsive to the deaths of children. He compared the US unfavorably with Cuba and China

Unfortunately, this conclusion was flawed:

babies that would die in the womb or at stillbirth elsewhere are born alive in the US. Many of these survive completely, but because of their precarious state, they tend to die at higher percentages than normal births. That's why the numbers rose slightly for 2002. The CDC doesn't expect to see another increase like it.

The solution might be to look at the survival rate at year one as a percentage of total pregnancies, not just births (though you would have to exclude abortions).