A fundamental issue behind the “realignment or reaction” debate over the meaning of the Obama victory is the relative importance of demographic trends as opposed to the persuasion and mobilization activities of parties and candidates (along with real-life events like recessions and wars).
Positions on this issue do tend to become somewhat polarized. Some “demography as destiny” apostles are almost certainly reacting to the tendency of the chattering classes to treat electoral campaigns as sui generis events in which the entire electorate is up for grabs, and the ebb and flow of campaign dynamics and external events control everything. And on the other side of the divide, those who resist “demography as destiny” explanations rightly object that eight years of the Bush administration, culminating in a financial panic and a deepening recession, not to mention the widely varying strategies and resources of the two parties and the two presidential candidates, must have had some significant effect on the results.
The ever-resourceful Chris Bowers of OpenLeft offers a useful demonstration today of both the strengths and weaknesses of the “demography as destiny” approach. Staring at the exit polls and the changing composition of the electorate over time, Chris provocatively suggests that Barack Obama was probably no better a candidate than the much-derided Mike Dukakis. What really changed between 1988 and 2008, he argues, was the relative size of pro-Democratic and pro-Republican elements of the voting-age population. By looking at Obama’s and McCain’s performance in various groups this year, and reconfiguring them to represent their size in 1988, Chris concludes that a candidate as “good” as Obama would have lost decisively that year, while a candidate as “bad” as Dukakis might have swept to victory this year.
Reflecting Chris’s longstanding views on the decline of White Christian America, the optic he chose for this analysis was religious categories of white voters, and unsurprisingly, he concludes that “The Rise of Non-Christian America” and “The Rise of a Non-White America” were two of the three most important moving parts in the slow change from defeat in 1988 to victory in 2008. The third, “Internet Rising,” isn’t directly about demographics, but plausibly represents a “cultural trend” that crucially affected media, fundraising, and organizational opportunities that Obama exploited but did not create:

Internet Rising: Over the past twenty years, the rise of the network neutral Internet has been, by a long way, the biggest change to the national media landscape. Fully 55% Americans are not just online, but have high-speed connections, and the Internet is now the #2 source of news in the country, trailing only television. Twenty years ago, virtually no one was online, and the world wide web wasn’t even created.
The Rise of Non-Christian America. The number of self-identified non-Christians has slightly more than doubled since 1990 and now totals 21.6% of the population.
The Rise of a Non-White America: While the Census does not measure ethnic / racial statistics the same way every ten years, thus making long-term trends somewhat difficult to pin down, twenty years ago non-whites formed 15% of the electorate, much lower than the 26% in the 2008 election. That is an increase of about one-half percent every year for the past twenty years.

Now the immediate problem with Chris’ analysis is that the Democratic-Republican numbers on racial and religious categories bounced around a fair amount prior to and after 1988. Perhaps both the Dukakis and Obama performances can somehow be stipulated as normative, but there’s nothing obvious about that conclusion. The bigger problem emerges when Chris turns to the question of what, exactly, Obama or Democrats can do to build on the demographic trends he views as crucial:

[N]ow that Democrats are the governing party in D.C., we need to actually go one step further and pass legislation that will itself help continue these trends. While there isn’t much we can, or should, do on the religion front (not the sort of business the government should be involved with), this does mean comprehensive, progressive immigration and media reform.

Since the “internet rising” development isn’t really a demographic trend, and since it’s not a slam dunk that nontraditional media will continue to provide a progressive or Democratic advantage, it’s unclear to me that why “media reform” should be a higher priority than, say, a new foreign policy or universal health care. Immigration reform is a good thing in itself, and will in the long run probably contribute to progressive politics, but the short-term costs in terms of a backlash are tangible as well, which is one reason it didn’t happen in the last couple of Congresses despite significant Republican support, which may now vanish.
Truth is, if you really think demography is the main determining factor in politics, with most elements of governing and campaigning being marginal, then there’s not a whole lot to do between election days other than waiting for the next incremental change in the composition of the electorate. I guess you could make the case that an Obama administration would in that circumstance be free to govern as progressively as possible, since voters have predetermined positions that won’t much be affected one way or another. But in the end, I doubt that Chris Bowers, or that many other advocates of “demography as destiny,” really think events in the real world of politics and governing matter so very little.