Mis-Measuring Infant Mortality

September is National Infant Mortality Awareness Month. The main evidence I have observed is that it has given critics an excuse to repeat claims that international comparisons of infant mortality and life expectancy prove the need for more government control of our health care system.

However, infant mortality and life expectancy comparisons are misleading health-care efficiency indicators. They not only ignore important differences in what countries count as infant deaths and the extent of heroic lifesaving efforts, but also many factors unrelated to health-care quality that dramatically alter international comparisons.

Nonviable babies who die quickly after birth are recorded as live births in the U.S., but are more likely to be classified as stillbirths in other countries, particularly if they die before birth is legally registered. That biases our infant mortality rate substantially upward compared with others. One study in Philadelphia concluded that the overstatement was 40%.

American doctors also go to greater lengths to resuscitate very premature babies who are not breathing when delivered. Consequently, babies at very high risk of death are counted as live births here, but not in many other countries, biasing our infant mortality rate sharply upward.

The U.S. also has the highest proportion of pre-term and low-birth-weight babies of any developed country, a category that comprises a large fraction of infant deaths.

To illustrate the impact of such factors, in 2010, our 6.1 death rate per thousand live births (down to 5.68 in 2020) fell to just 4.2, when births at less than 24 gestational weeks were excluded.

Infant mortality also reflects many other factors, including mother’s age, obesity, drug use and other lifestyle factors, all of which further worsen American results compared to other developed countries.

For example, teenage mothers (far more common in the U.S. than other developed countries) are far more likely to have low-birth-weight babies. A study found that if the U.S. birthweight distribution had been the same as for Canada, that would by itself lower American infant mortality below Canada’s.

Beyond overstated infant mortality measures, U.S. life expectancy numbers are reduced by higher rates of death from violence and accidents in the U.S., which does not reflect health care failings.

Larger, more diverse countries also tend to have worse life expectancy results than smaller, more homogeneous countries, where communication problems, cultural differences, variance in population characteristics, etc., are far smaller.

Further, what works for small, compact populations may not scale to far larger countries. Critics often repeat that the U.S. ranks down the list in life expectancy, but what they never mention is that no country more populous than the U.S. ranks higher. In fact, you could add up the populations of half of the countries in the top 10 of life expectancy in some years (e.g., Iceland, Monaco, and Andorra) without totaling California’s population.

Perhaps most importantly, as Scott Ehrlich reported in 2017, while our “official” life expectancies lag many other countries, “on average, there is nowhere you will live longer in the world as someone of Asian, Hispanic, or African descent, than in the United States.”

Critics constantly use infant mortality and life expectancy comparisons as weapons to attack the U.S. as offering inferior health care and to push for ever-more government control. However, the infant mortality and life expectancy differences they say proves their case involve overly simplistic comparisons of inconsistent measures, which omit many important determinants. It is just as convincing as concluding that an above-average death rate in an elite hospital serving the riskiest patients proves it is a low-quality hospital.

Gary M. Galles is a professor of economics at Pepperdine University.