A colleague and I recently put together an analysis of abortion attitude data, a version of which was published today on MercatorNet. The piece quantifies the degree to which the climate surrounding this issue has shifted over the last 15 or so years.
We combined over 30,000 survey interviews from Missouri, spanning 1992-2006, and looked at changes in pro-life/pro-choice self-identification (national Gallup Poll numbers are similar). We find there have been dramatic shifts in the pro-life direction: in 1992, the electorate was 30% pro-life and 43% pro-choice. The two labels reached a rough parity in 1997, and the pro-life label has since grown to a 41% to 30% advantage. In other words, the turnaround has been nearly complete.
Some of the demographic subgroup changes are especially interesting. For example, young women had been the most strongly pro-choice group in 1992; now they are the most strongly pro-life. Other dramatic shifts have occurred among voters who rarely or never attend church services, and those with post-graduate degrees; they had been very pro-choice in 1992, but have abandoned that label in droves.
The most likely cause of the attitude changes, we speculate, is the silencing of confrontational clinic protests---coupled with the ascendancy of partial-birth abortion as the new frame for the issue. The public has clearly changed its mind as to who the "abortion extremists" are.
There have also been implications for the party coalitions. The Democratic coalition is now much more divided on this issue than the Republican coalition is, which may explain why Democrats in most states no longer make such a big deal of their pro-choice stance. The pro-life label no longer carries the stigma it once did, and pro-life candidates should not shrink from identifying themselves as such.
A slightly different version of the article, which includes graphical displays of the trend and subgroup data, is available for download from my consulting website. These graphs didn't really fit MercatorNet's format or style, but please do download this document and take a look at them: they show at a glance just how dramatic the over-time changes have been. They also give much more detail about how the data break out; the MercatorNet version had to truncate some of those details for journalistic purposes.