Editor’s note: this is a guest post by Alan Abramowitz, who is Alben W. Barkley Professor of Political Science at Emory University, and a member of the TDS Advisory Board. It was originally published on May 20, 2009.
Is support for abortion rights hurting Democratic candidates at the polls and, if so, would abandoning the Party’s traditional pro-choice position help Democrats win over pro-life voters? These questions are being raised with increasing urgency following the release of new Gallup and Pew polls that supposedly show a substantial decrease in support for the pro-choice position among the American public.
The findings of the Gallup and Pew polls are rather surprising given the stability of public attitudes on the abortion issue over several decades. Moreover, a number of other polls conducted before and after the 2008 election found no dramatic change in public opinion on this issue. For example, the two most respected academic surveys of the American public, the General Social Survey and the National Election Study, found no decline in support for abortion rights between 2004 and 2008. More importantly, the evidence from the 2008 National Election Study indicates that Barack Obama’s support for abortion rights was a net plus for his candidacy and that attempts by Democrats to win over pro-life voters by abandoning the Party’s support for abortion rights would probably do more harm than good.
Every four years since 1980, the American National Election Study has asked a sample of eligible voters to choose one of four positions on the issue of abortion: abortion should “never be permitted,” abortion should be permitted “only in case of rape, incest, or when the woman’s life is in danger,” abortion should be permitted “for reasons other than rape, incest, or danger to the woman’s life, but only after the need for the abortion has been clearly established,” or “a woman should always be able to obtain an abortion as a matter of choice.”
The results in 2008 were very similar to those in other recent election years: 13 percent of voters supported a total ban on abortion, 26 percent supported allowing abortion only under highly restrictive conditions (rape, incest or danger to the woman’s life), 19 percent supported allowing abortion under less restrictive conditions but only if a there was a clearly established need, and 42 percent supported allowing abortion as a matter of choice. For analytical purposes, I combined the first two options, banning abortion completely and allowing it only under highly restrictive conditions, into a single pro-life category. I left the rather vague third option, allowing abortion only if a need had been clearly established, as a middle category, and I used the fourth option, allowing abortion as a matter of choice, as the pro-choice category. This resulted in 39 percent of voters being classified as pro-life, 19 percent being classified in the middle position, and 42 percent being classified as pro-choice.
There was a strong relationship between abortion position and presidential vote in 2008. Pro-life voters supported John McCain over Barack Obama by decisive 62 percent to 38 percent margin. But pro-choice voters supported Obama over McCain by an even more decisive margin of 73 percent to 27 percent. Those in the relatively small moderate group favored McCain over Obama by a fairly narrow 55 percent to 45 percent margin.
According to the NES data, pro-choice voters supporting Obama made up 30 percent of the electorate while pro-life voters supporting McCain made up only 24 percent of the electorate. These results suggest that Barack Obama’s support for abortion rights helped him more than it hurt him in 2008. Before accepting this conclusion, however, we need to control for the influence of partisanship because opinions on abortion are strongly correlated with party identification and 90 percent of Democratic and Republican identifiers voted for their own party’s presidential candidate in 2008.
In order to evaluate the impact of the abortion issue on the performance of the presidential candidates, we need to know whether partisan defection rates were affected by opinions on abortion.

Table 1 displays the relationship between abortion opinion and partisan defection among Democratic and Republican identifiers (including leaning independents) who had a clear abortion position. About one-fourth of each party’s voters were cross-pressured on the issue of abortion: 27 percent of Democratic voters took the pro-life position and 25 percent of Republican voters took the pro-choice position. However, the results in Table 1 show that pro-choice Republicans were more than twice as likely to defect as pro-life Democrats. Pro-life Democrats were only slightly more likely to defect than pro-choice Democrats but pro-choice Republicans were much more likely to defect than pro-life Republicans. In fact, the relationship between abortion opinion and defection was statistically significant only for Republicans.

Based on these results, the abortion issue appears to have produced a small net gain in support for Obama in 2008. Pro-choice Republicans who voted for Obama made up 2.6 percent of the electorate while pro-life Democrats who voted for McCain made up 1.4 percent of the electorate, resulting in a net gain of 1.2 percent of the vote for Obama.
Some critics of the Democratic Party’s current position on abortion have suggested that the Party could make substantial inroads among pro-life voters who now support the GOP by abandoning its support for abortion rights. The evidence from the 2008 NES displayed in Table 2 raises serious doubts about the viability of such a strategy, however. Republican voters who were pro-life on abortion tended to take conservative positions on many other issues: the overwhelming majority described themselves as conservative, opposed marriage or civil unions for same sex couples, opposed a larger government role in providing health insurance, supported the war in Iraq and approved of President Bush’s job performance.

Based on their ideological identification and other issue positions, there appears to be little likelihood that pro-life Republicans would respond positively to appeals from Democratic candidates on the issue of abortion. Moreover, in addition to alienating the pro-choice majority of Democrats, such a shift would also alienate the pro-choice minority of Republicans who appear to be much more open to appeals from Democratic candidates on a wide range of issues. These findings suggest that rather than abandoning the Democratic Party’s traditional support for abortion rights in a futile pursuit of pro-life Republican voters, Democratic candidates would be better off focusing their efforts on appealing to Republicans who support the Democratic Party’s traditional position on abortion.