A great Ron Brownstein column on the demographic questions hanging over the presidential race ends with a lame sixth question: "Can anyone win a mandate to govern?"
The answer is "no." A mandate is not a real thing, so there's nothing a candidate can do—up to and including a Democrat carrying North Carolina and Indiana—to win one. Probably the best way to think of a mandate is as a historical artifact of the poorly sorted congressional politics of yore. Politicians in that framework were cross-pressured between partisan and ideological loyalties. A president with a "mandate"—think Ronald Reagan in 1981 or Lyndon Johnson in 1965—could unify the ideological factions within his own party while fracturing the other side's coalition. Modern politics just doesn't work like that. Olympia Snowe, Scott Brown, Susan Collins, and Mark Kirk all vote a more reliably conservative line than Ben Nelson, Joe Manchin, or Claire McCaskill. Different politicians disagree with each other about different things, but the parties are basically coherent ideological teams. And although there may be compromises to be forged, there are no mandates to be won.I think Yglesias is basically correct here, but he's leaving out part of the answer. Modern opinion polling (and its ever-increasing use in political positioning) plays a role, as well. National political candidates simply don't campaign on viewpoints that are held by a small minority of the electorate. Even when politicians actually hold unpopular views, they rarely, if ever, are a public point of contention. As a result, it's hard to say that an election result ever represents a definitive rejection of a particular set of views (and consequently a definitive embrace of another set of views).
Absent any means of predicting the public's response to particular policies or positions, you'd expect candidates to generally clash on topics where their views were the furthest apart. But because polling lets candidates know in advance who will win those arguments, you find that candidates usually refuse to engage on issues where they disagree the most, and often end up instead stepping on each others' toes trying to win a slim majority on subjects where both sides generate strong support.
(One irony is that, on issues where a whole spectrum of views are possible, two positions with roughly even amounts of support are likely to be quite similar in substance. Of course, "I'm slightly better than the other guy" isn't very persuasive, so parties are left trying to exaggerate their differences on these issues, while ignoring the issues on which their differences are more pronounced. This is probably largely responsible for the strange phenomenon of low-information voters believing "both parties are the same." Come to think of it, this may also explain much of the attraction of culture war issues for campaigns: more binary viewpoints mean less opportunity for the losing side to triangulate, and less confusion over the parties' distinctions.)
Another way of saying the same thing: a "mandate" represents the discovery that voters widely support (or reject) a particular point of view. When everyone knows what the voters think far in advance of an election, any political party worth its salt will embrace (or dodge) those views to whatever extent it can. As a result, policy outcomes are less election-driven and more poll-driven; policies with widespread support are less likely to be propelled into existence by major electoral victories and more likely to emerge, in more-or-less the same form, no matter who wins an election.