The typical politician is usually being pulled in two directions at once. She has public opinion (measured by polls, recent election results, or gut instinct) pulling her toward the ideological "center," and she has activists, interest groups, and sometimes members of her own party pulling her toward the ideological extremes. (Admittedly, this is a pretty simplistic version of the pressures an officeholder actually faces, but it captures the main dynamic.)
Implicit in all of this is the notion that activists and interest groups are always trying to pull officeholders toward a more extreme position and are almost always disappointed.
Which way does she end up leaning? That depends on the circumstances. Is it an issue to which voters are paying close attention, perhaps one that will help them decide their vote in the next election? If so, she might end up going with what her constituents want. Conversely, if the public isn't paying close attention (which is the situation on most issues) and if activists and interests seem to care passionately about it and can make her life hell in the next primary if she disappoints them, she might end up leaning toward the more extreme position. This is part of the reason that officeholders tend to be more ideologically extreme than their districts (PDF).
Implicit in all of this is the notion that activists and interest groups are always trying to pull officeholders toward a more extreme position and are almost always disappointed. Liberals were happy that Obamacare got signed into law but really wanted a public option or a single payer system. Abortion opponents may be gladdened to see Texas enact a law that makes it more difficult to obtain an abortion, but that's not as far as they'd like to see the law go.
Sometimes, however, this model doesn't hold. In a paper (PDF) presented at last week's American Political Science Association conference, Daniel Galvin offers details on the fascinating relationship between the United Auto Workers union and the Michigan Democratic Party. Galvin shows that in one of the strongest remaining bulwarks of union activity in the United States, Democratic officeholders have steadily moved to the center, often voting clearly against issues of importance to their union allies. And they have often done so not in spite of the union but because of it:
[U]nion leaders supported adaptation by Democratic candidates and officeholders. Union officials actively recruited centrist candidates who could “go base-plus” and lent their formidable electoral muscle to support almost anyone who could help the Democratic Party build electoral and legislative majorities. [Emphasis in original.]
Galvin paints a picture of a decidedly strategic union that sought out candidates for governor, Congress, and state legislature who were not in its pocket and gave them leverage to behave more moderately. He quotes one UAW leader as saying, "I can hardly remember having [a candidate] that I wanted. I spent a lot more time in the 'live with' category. And that's what you have to do." Some union leaders would actually advise candidates to run for office while bashing the UAW.
This appears to be part of a broad strategic decision by the union over many years of political activity. Rather than stay to the left and lose relevance and elections, they chose to moderate in an effort to remain competitive. They might not get what they wanted all the time, but if it kept Democrats more competitive in the state, they would still get more than they would if more Republicans were elected. As Galvin suggests, this approach has borne some fruit: "Democrats in Michigan managed to remain more electorally resilient than neighboring Democrats, and part of the explanation is that many moved to the right in their policy positions."
Could this serve as a model for unions or other interest groups in other states? The situation in Michigan is pretty singular—it's hard to conceive of a political relationship quite like that of the UAW and the Michigan Democratic Party. Nonetheless, in an era when the parties are running hard to their extremes, there is an electoral opportunity for the party that allows its candidates to moderate. Yes, it involves giving up on some longstanding policy goals, but not as many goals as you'd give up by always losing.