Keli Goff Disagreed When I Called Her Rant “Statistical Malpractice”
As I flipped channels the other day, I came across the Dylan Ratigan show just as his “daily rant” segment was starting. I listened as Keli Goff pieced together an argument using statistics that I found quite amazingly specious, linking together what I assumed were cherry-picked statistics to make a point with which I agree (that contraception should be freely available), but in a way I found undefensible. I waited a couple of days and then went back to get the link to the clip, and put together a tweet:
@i8dc: Statistical malpractice: rant by Keli Goff on Ratigan. http://www.msnbc.msn.com/id/37560195/#46373571. Women will suddenly become illiterate without contraception?
Quite quickly came the reply:
@keligoff: Hmm. Not what I said & speaking of literacy, you might want to try READING the column and stats yourself: http://bit.ly/xRAmVk
Touched a nerve, I guess I did. At least one of us is out of line here. Was I? Tweets are fun things; 140 characters to get a whole thought out. When I wrote mine, I had space left over, and picked a nice juicy follow-up line to try to pique the interest of my 10-ish followers, and anybody who happened across it. Provocative? Yes. Unfair? Maybe; I’ll come back to that.
Goff made two assertions in her reply. First, my tweet did not accurately reflect what Goff said. Second, with a dig at my ability to read, that I should have consulted data she then provided in a link. Following her link, I found a column she wrote which appears to have been the source material for her MSNBC rant, which was likely a heavily edited version.
Goff was critical of me for not being familiar with material that she didn’t reference in her MSNBC rant.
So she was out of line with the second part of her tweet. What about the first? She indicated that she didn’t say that women would become illiterate without contraception. I agree. But did she say something equivalent? From the full transcript of the MSNBC rant, here’s the key passage:
[I]f barriers to contraception access are increased in America, you will see far fewer female voices heard [sic] in the media and in the halls of Congress. How do we know? Because history and a look at other countries where contraception access is limited tell us so.
Goff then lists a few data points, one of which is that “countries with low contraception usage have the lowest number of women who can read.”
So she didn’t say that women would suddenly would suddenly become illiterate. She said that women would be heard less on TV and in Congress, just as fewer women can read in countries with low contraception usage.
Hey, I agree it’s a lousy argument. But it is the argument Goff made.
The problem here is one of causality — Goff clearly argued, using an if-then construction, that there would be a causal relationship between an increase in restrictions on contraception and a decrease in women’s prominence or participation in the media and elective office. She then presented examples that simply do not show any such causal relationship. I think a point-by-point discussion is in order.
“Number one: In countries with the highest fertility rates, countries like Sierra Leone and Chad, where women give birth to six or more children, women have the shortest life expectancies. In Sierra Leone, one in women die in childbirth. During the 17th and 18th centuries, when American women gave birth to between five and eight children, the same number of women died in childbirth here as well — approximately one in eight.”
There’s a big problem here: there are women in America who have many kids, who would make a far better comparison. Unsurprisingly, such women do not die from causes related to pregnancy or childbirth the anything close to the rates in sub-Saharan Africa or colonial America. Also unsurprisingly, men in the countries she cites also have startling low life expectancies. Sierra Leone ranks #198 in female life expectancy, also #202 in male life expectancy. Chad ranks #221 and #223.
Clearly, there are other factors at work — widespread crushing extreme poverty has something to do with this. Also clearly, Goff’s number one supporting fact doesn’t, in fact, support her argument.
“Number 2: Countries with low contraception usage have the lowest number of women who can read. In Afghanistan, which continues to have one of the highest fertility rates in the world, and where contraception knowledge and access remains limited, 87% of women cannot read. In Sierra Leone, the number is 71%.”
Again, Goff tries to compare women in the world’s least developed countries without context. Afghanistan, which suffers from extreme poverty, decades of oppressive theocratic rule, and a decade of war, ranks last (207th) in female literacy. It also just happens to rank 203rd in male literacy. Sierra Leone actually ranks lower in male literacy (202nd) than it does in female literacy. Note, the gap between men and women at the bottom of these rankings is enormous – twice as many men can read as women. But the point is that if there was a causal relationship that had anything to do with contraception, the men would not also fare so poorly when compared to the rest of the world. Again, what we are looking at is not causation, but correlation.
“Number three: Before the 1965 Griswold ruling made contraception widely available to American women, men greatly outnumbered American women in colleges. Today, women outnumber men. Number four: Before Griswold, there were no female CEOs of Fortune 500 companies. In 2011, there were 18. And number five: Before Griswold, there were 20 women in the House of Representatives. Today there are 76 women in the House, and 17 women in the Senate.”
I’m not a constitutional scholar. But I think Griswold is historically important because it furthered the right to privacy that would be built upon in Loving, Eisenstadt, and perhaps most notably in Roe v. Wade. It was not important because it “made contraception widely available to American women.” It didn’t; Margaret Sanger had been providing birth control through physicians for more than 40 years since opening a New York City clinic in 1923. The birth control pill was approved by the federal government in the late 1950s. Griswold applied only to marital privacy, and applied only to Connecticut, which was the only state in the union limiting contraception availability to married women. That’s right, Griswold applied only to married women; Eisenstadt (1972) extended the right to unmarried women.
And again, Goff implies a causal link that doesn’t exist. Griswold was a step–no more–in a decades-long expansion of personal rights, notably civil rights of women and minorities, which have led to the expanded opportunities available today.
Here’s how Goff finished up:
“Today there are 76 women in the House, and 17 women in the Senate. All of whom, I presume, do not want their male colleagues speaking for them or other women on an issue that affects us.”
Really? Male legislators shouldn’t speak for their female constituents on an issue that affects them? Really? How else shall we limit our elected officials? Should those who never served in the military recuse themselves from matters related to the armed services? There’s no such thing as a representative who can have first-hand knowledge of all issues important to constituents; to argue that these legislators are incapable of representing these types of constituent interests is at least silly, at worst quite insulting.
All of these problems with the MSNBC rant could have resulted from a sloppy edit of the source column. But, mostly, they didn’t; Goff’s column adds other facts which provide no support for the argument in the rant. But there was additional support in her lead-in to the statistics:
“Before contraception was widely available, there were far fewer women able to do just that because of the physical, emotional and financial demands that giving birth to and raising sometimes more than a dozen children (something my great-grandmother did) required.”
Now THIS is a great point. But it didn’t make the cut for the rant.