69

In a town hall meeting on June 2, 2019 with Representative Tim Ryan (D-OH), the candidate stated the following (emphasis mine)

Now, a lot of people don't know, but Youngstown, Ohio is 50 percent African American. I've been working on these issues a long time. We need to address when it comes to African American women what's happening with their health 300 to 400 times - 300 to 400 times higher percentage of death and pregnancy or child birth. Even along education and wealth, income levels, that's an issue we need to address. An African American baby born in Youngstown, Ohio has a higher infant mortality rate than a baby born in Iran. I am deeply committed to these issues around justice when it comes to the African American community, and you can bet that I will surround myself with men and women in my administration that will help us put and excellent plan together to make that happen. Thank you.

So is his claim correct? Does an African-American baby born in Youngstown, Ohio have a higher infant mortality rate than a baby (not of a specific ethnicity) born in Iran?

134

Highly likely. The infant mortality rate in Iran is 13 (per 1000 births); the infant mortality rate for African Americans in Ohio (but not specialized to Youngstown) is 15.6.

There may be difference in definition (in the USA, some neonatal deaths that are included in the IMR are classified as stillbirths in other countries and excluded), but Ohio would not be the only state where black mothers have Third World levels of infant and maternal mortality.

  • 14
    ... though this is a difference of 12%, or 1.12 times. That's clearly not the same as the 300 to 400 times (equivalently, 30000%-40000%) difference quoted in the claim. Nor is it even close to a 300%-400% difference, which I suppose is what Tim Ryan probably meant to say. So while your source supports the simple claim that the AA infant mortality rate is higher, it suggests that the specific numbers given - even if we charitably assume that Ryan really meant to describe a difference a factor of 100 less than what he literally said - are wildly false. – Mark Amery Jun 14 at 12:03
  • 8
    That number is in relation to maternal mortality, which does appear to average about 3:1 nationally. cdc.gov/vitalsigns/maternal-deaths/index.html – richardb Jun 14 at 13:18
  • 21
    Iran is not a third world country, so infant mortality rates higher than those in Iran are not necessarily third-world rates. – Mike Scott Jun 14 at 15:44
  • 10
    @Servaes, the 300% comparison is to white residents of Ohio IMR. – Andrew Lazarus Jun 14 at 18:38
  • 7
    @copper.hat It referred to non-aligned nations during the Cold War, fifty years ago. Nowadays, it means one of the Least-Developed Countries. – Mike Scott Jun 14 at 19:03
9

It's plausible, but unclear. According to official statistics, Iranian women have a significantly lower maternal mortality rate than Ohio blacks. However, who has the higher infant mortality rate is harder to judge, because it has a less clear definition, the stats for the two populations are close, the rates are very similar and year-on-year changes potentially flip the result from one year to the next, and different sources disagree on what the rate is - for both Iran and Ohio - by enough to flip which of them has the higher rate.

The World Bank put the maternal mortality ratio for Iran in 2015 at 25 per 100000 live births. (Source: https://databank.worldbank.org/data/reports.aspx?source=2&series=SH.STA.MMRT.) America's Health Rankings, on the other hand, puts the rate for blacks in Ohio at 46 per 100000 live births in 2018. (Source: https://www.americashealthrankings.org/explore/health-of-women-and-children/measure/maternal_mortality/state/OH.) Thus the comparison to Iran is probably true insofar as it applies to maternal mortality.

The claim about infant mortality, however is harder to be sure of. In part this is because of differences in reporting standards, explored (in the context of US healthcare) in Joseph Ahern's article Do Differences in Reporting of Live Births Affect Comparability of Infant Mortality Rates?. (Spoiler: the answer is "yes".) Per official definition, any death of an infant who is born alive but then dies prior to their first birthday is supposed to count towards infant mortality statistics. In practice, however, the authors note that, for the purposes of statistical reporting, hospitals in some jurisdictions routinely consider premature babies born prior to a certain gestational age not to have been born alive - even if, in reality, they were - despite this practice having no justification in the definitions used by either the (international) WHO or the (American) CDC. For instance, they note that some hospitals in New York City do not count actually-alive births of babies of under 23 weeks' gestational age as live births. The article notes that the average "infant mortality rate" of hospitals in NYC that follow the official definitions strictly is 3.54 per 1000 live births, while those that apply a 23 week cutoff have a rate of 2.68 per 1000 live births. Though not explicitly stated, it seems to be implied that this one-third difference owes entirely to the difference in definitions.

This caveat is worth bearing in mind as we note that:

Depending upon which of these different stats we believe - if any of them - the infant mortality rate for African Americans in Ohio could be higher or lower than the overall rate in Iran. I don't know how to judge which of the stats are true, so I can't tell. Regardless of which is true, it's possible that the nominal "infant mortality rate" is really measuring a different thing in each country due to differences in reporting, potentially rendering the comparison between them meaningless.

Finally, note that America Health Rankings puts the overall infant mortality rate for Ohio at 7 per 1000 live births, suggesting that blacks have an infant mortality rate of only two times that of the general population. On maternal mortality, the stats are 20.3-per-100000 for the whole population vs 46-per-100000 for blacks. So Ryan's suggestion that blacks are 300-400 times more likely to suffer a death in childbirth - even if we charitably assume that he really meant 300%-400% more likely - seems not to be true. The real stat is more like "100% more likely".

  • It was stated that black persons are 50% of the population. That means you shouldn't compare "black" vs "average" but "black" vs. "not black". If there are 15.6 per 1,000 deaths for black people, and half the population is black, then for an average of 7.8 per 1,000 there would have to be no non-black child deaths at all, because the black deaths make that average already. – gnasher729 Jun 16 at 22:26
  • @gnasher729 Blacks are 14.2 of the Ohio population, not 50%. Given an overall rate of 7 and a black rate of 13.5 (not 15.6 - if we're going to directly compare rates, we should probably compare them from the same source wherever possible), that puts the non-black rate at (7 - 13.5 * 0.142) / 0.858 = 5.92 deaths per 1000. Which is still roughly a factor of 2, nowhere near the factor of 4-5 needed to justify "300%-400% more likely" or even the factor of 3 that would be needed to justify a claim of "300% as likely". – Mark Amery Jun 17 at 9:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .