The Roe Effect & Abortion
On January 17, 2003, James Taranto opined that one of the long-term effects of abortion was to make the country more conservative.
It's almost a truism that women who have abortions are more pro-choice than those who carry their pregnancies to term, and it stands to reason that they generally have more-liberal attitudes about sex and religion. It also seems reasonable to assume that parents have some influence on their children, so that if liberal women are having abortions, the next generation will be more conservative than it otherwise would be.
His conclusions may or may not be true (e.g., an alternative explanation is that that baby boomer culture, as a whole, is liberal. Thus, the younger generations more conservative beliefs could simply be an outgrowth of normal rebellion), but this idea is interesting enough it deserves more study. On December 9, 2003, Taranto started calling this idea the Roe Effect.
Why should it surprise anyone that those people lucky enough to have been born since 1973 would be more conservative than their elders, especially on abortion?
Up to this point, I believe Taranto had made a compelling, albeit circumstantial, case for the Roe Effect. Today he confused me.

Now, do you notice anything about these lists of states? Here's a hint:

High abortion rates Low abortion rates

District of Columbia (Gore by 76.2%)
New Jersey (Gore by 15.8%)
New York (Gore by 25.0%)
Maryland (Gore by 16.2%)
California (Gore by 11.7%)
Nevada (Bush by 3.5%)

Utah (Bush by 40.5%)
South Dakota (Bush by 22.7%)
Kentucky (Bush by 15.1%)
North Dakota (Bush by 27.6%)

This certainly looks to us like confirmation of the Roe effect. We'll have more analysis of the Guttmacher numbers in a future column.

Why is this confirmation of the Roe effect? If anything, it suggests that Roe effect either does not exist or it works completely opposite to Taranto's prediction. Taranto is an experienced columnist, he may have a perspective I have obviously missed.

For my part, I do not think this data, by itself, can be used as evidence to support or falsify the proposed Roe effect. If you could control for migration, immigration, and state culture (a difficult proposition indeed), one would have to need a "liberal" measure for each state going back at least 18 years. If Taranto's Roe effect exists, then there should be a significant difference between the change in liberalism in New Jersey than in Utah (with Utah showing a greater amount of increased liberalism over time than New York, since the Roe effect predicts more liberal children would be aborted in New York).

Assuming the Roe effect exists, I think it's impact will be felt in two ways. One, the conservative states (where abortion rates are low) will continue to grow faster than the liberal states (where abortion rates are high) when you control for immigration and migration, thus they will have more political representation over time. Two, when enough children born after 1973 are voters (and enough liberal boomers grow old and die), abortion will again be illegal.

As for the abortion numbers Taranto quoted. It merely shows the unsurprising link between abortion and the Democratic party. The parts of the US where abortion is more common naturally votes for the party that supports their peculiar institution.

James Taranto responded. He was not drawing inferences about how the Roe effect may have impacted each state (which didn't make sense). He was using the data to test one of the assumptions behind the Roe effect.
I should have been more precise. The data seem to confirm a premise of the Roe effect, namely that liberal women have more abortions.
That makes perfect sense to me. A big thanks to Taranto for clarifying.

Send this Post
Email this entry to:

Your email address:

Message (optional):