This is a guest post by Emory University political scientist Alan Abramowitz.

As we marked the 41st anniversary of the historic Roe v. Wade decision this week, the issue of abortion continues to divide Americans. While recent polls show that most Americans want to keep Roe as the law of the land, large percentages and sometimes majorities of Americans also support limitations on when and under what circumstances abortion should be permitted. Many clearly have ambivalent feelings about this highly emotional issue. Nevertheless, Americans remain sharply divided on the fundamental question of whether abortion should remain a woman’s choice in the large majority of cases. And when it comes to abortion, it is this question of abortion as a woman’s choice that is the most significant politically.

The public division on abortion is nothing new. It has existed almost since the Supreme Court issued its ruling in 1973. What has changed, however, is that opinions on abortion have become increasingly divided by party as the positions of party leaders have diverged on the issue. The graph below displays the trend in the average position of Democrats and Republicans, including leaning independents, on a four-point abortion scale that has been included in American National Election Study surveys since 1980. The choices on this scale include completely banning abortion; permitting abortion only in cases of rape, incest or danger to the life of the mother; allowing abortion under other circumstances; and allowing abortion as a matter of choice.

Although the average position of all voters changed very little between 1980 and 2008, the distance between Democrats and Republicans increased dramatically over that period. The gap continued to grow after 2008, as well.  In 2012, Democrats and Republicans were more polarized than ever. A record 63 percent of Democrats took the most pro-choice position on this question compared with only 29 percent of Republicans. In contrast, 54 percent of Republicans favored banning abortion entirely or allowing it only in cases of rape, incest or danger to the mother’s life, compared with only 23 percent of Democrats.

Another question in the 2012 ANES survey also shows this polarization. It asked respondents to place themselves on a nine-point scale indicating how strongly they favored or opposed permitting abortion as a matter of choice. Overall, voters divided almost evenly — with 45 percent in favor and 44 percent opposed. But most striking was the polarization: 59 percent of voters placed themselves at the two most extreme positions on the scale, and 77 percent placed themselves within one unit of either pole. In contrast, only 11 percent placed themselves in the middle of the scale. By the standards of public opinion on policy issues, this is an extraordinary degree of polarization.

That gap is also driven by partisanship.  The graph below shows that the overwhelming majority of Democrats were on the pro-choice side of the scale while the overwhelming majority of Republicans were on the pro-life side of the scale.  In fact, more than 50 percent of Democrats placed themselves at either 1 or 2, and almost 60 percent of Republicans placed themselves at either 8 or 9 on the scale.

Nevertheless, a substantial minority of voters in each party held views on abortion that were in conflict with their party’s position. In fact, 23 percent of Democrats came out as strongly pro-life (8 or 9 on the scale), and 21 percent Republicans came out as strongly pro-choice (1 or 2 on the scale).

These results suggest that abortion could have been a major wedge issue in 2012 — potentially prompting defections among those whose views differed from those of their party’s candidates. But despite efforts by candidates in both parties to exploit divisions among the opposing party’s supporters, such defections were limited. Only a small minority of voters whose opinions on abortion conflicted with their own party actually voted for the opposing party’s presidential candidate. Moreover, these defections essentially canceled each other out. According to the data from the 2012 ANES, 17 percent of strongly pro-life Democrats voted for Mitt Romney, and an identical 17 percent of strongly pro-choice Republicans voted for Barack Obama. Neither presidential candidate gained a clear advantage from voter defections on the issue of abortion in 2012.

This is likely to be true in the 2014 midterm elections, as well. A large majority of Americans now hold opinions on abortion that are consistent with the position of their party. Moreover, roughly equal proportions of Democrats and Republicans hold opinions on abortion that conflict with the position of their party and these voters still cast their ballots overwhelmingly for their own party’s candidates. This may indicate these voters are more concerned about other issues or don’t know the positions of the parties on abortion.

The abortion issue is likely to matter only to a few competitive races in which one candidate takes a position that is clearly out of line with the views of most voters. While this could conceivably affect either Democratic or Republican candidates, based on recent election results the most likely scenario would involve a Republican candidate who opposes abortion even in the case of pregnancies caused by rape. In 2012, according to data from the American National Election Study, 75 percent of all voters and 64 percent of Republican voters favored allowing abortion for pregnancies caused by rape.

In 2012, two Republican Senate candidates, Todd Akin in Missouri and Richard Mourdock in Indiana, lost what appeared to be very winnable races because of defections by Republican voters. According to state exit polls, Republican defectors outnumbered Democratic defectors by 21 percent to 4 percent in Missouri and by 20 percent to 7 percent in Indiana.  It seems likely that Akin’s and Mourdoch’s outspoken opposition to abortion even in the case of pregnancies caused by rape contributed to these extraordinarily high defection rates among Republican voters and to their defeats.

Democratic strategists can only hope that Republican primary voters will nominate more candidates like Todd Akin and Richard Mourdock in 2014.