I recently saw this headline claiming that abortion rates increased after Texas defunded the abortion provider Planned Parenthood in 2011. After claiming that the new policy resulted in the closure of some “women’s health clinics,” The Huffington Post article stated, “the abortion rate among teenagers in the state rose 3 percent over what it would have been had the clinics remained open.”
That sentence set off alarm bells in my head. Note what it didn’t say; that the abortion rate actually increased. The “would have been” language is usually code talk for statistical gymnastics in order to justify a position or theory that can’t be proven by the facts.
That is certainly the case here. In this research, the obvious point being made by the researcher is that by cutting funding for Planned Parenthood, Texas conservatives actually achieved the opposite results they intended: an increase in teen abortions. That would be news; except it isn’t true. Though that didn’t stop the HuffPost’s headline writer from writing that is was.
Here is the graph from the paper that shows the real data about teen abortion rates in Texas:
In 2012, the first year after cutting off funds to Planned Parenthood, abortion rates dropped. While there was a slight increase in 2013, a sharp drop in 2014 left the rate far below where it started out in 2011. In other words, after the defunding of Planned Parenthood in 2011, abortion rates for teens continued their overall downward decline. No headlines there; although it is sure that a number of people who would otherwise be dead will be pleased with this news as soon as they are old enough to understand it in a few years.
To make her political point about the reckless actions of Texas conservatives, the researcher stated “these data cannot produce convincing causal estimates” and used that to justify her construction of what she called “a Synthetic Texas,” in which she constructed her estimates of what would have happened if the clinics had remained open. Then she compared her Synthetic Texas with the Real Texas to see if she could come up with a better conclusion that the real data presented:
Of course, there were several problems with the data when constructing Synthetic Texas. So she had to “intuitively” utilize data, “estimate” a model, make an “assumption” about whether the data would do what she wanted it to do, etc. But the really big problem, as you can see above, was, “Unfortunately, as of now, abortion data are only available up to 2013.” So the researcher simply had to, unfortunately, cut off her comparison of Real Texas v. Synthetic Texas at 2013, which just happened to be the year that rates increased slightly–right before the sharp drop of 2014.
So even though teen abortion rates in Texas sharply decreased in the three years following the defunding of Planned Parenthood, the researcher uses the artificially created gap between Real Texas and Synthetic Texas to claim “the funding cuts increased abortion rates by 4.9 percent 1-2 years after the funding cuts and 3.1 percent over three years.”
The researcher pulled the same stunt when dealing with the teen birth rate. The researcher acknowledges, as seen in the table below, that “teen birth rates fell after 2011” in Texas. But, of course, this doesn’t fit the scenario of being able to show the harmful effects of defunding Planned Parenthood. So the researcher compares the teen birth rates in Texas with those in other U.S. counties in order to make an “analysis [that] can be viewed as estimating to what extent teen births rates could have declined further in the absence of family planning funding cuts.” The result of this analysis is that the researcher concludes “that decreasing funding for family planning in Texas … led to an increase in the teen birth rate by 3.4 percent.” Even though the actual teen birth rate decreased by almost 19 percent.
This fake research led to fake news in more than just the Huffington Post. The Austin Chronicle, the Daily Kos, and several other outlets picked up the story. Seems as if they all got the latest talking points memo.