The United States Census Bureau has evidently released the raw data from its November, 2008 voting and registration survey (I say evidently because I can’t find the actual results on their website – you can see for yourself!), and several scholars (see here) who apparently have access to the data have begun analyzing the results. Because this is a survey based on a random sample comprised of some 50,000 Americans, the findings are particularly useful for detecting broad voting trends. However, they are subject to all my caveats regarding the limits on using survey results to understand the larger population from which the survey sample is drawn. In particular, some questions in this survey, such as whether the respondent voted in 2008, depend on the accuracy of the individual’s recall, and we know from experiments that people habitually overstate the degree to which they did socially desirable things, like voting. Keeping this in mind, what did the survey find about voting in the 2008 election and what do the results portend for the future of the two major political parties?
Most scholars have focused on the turnout figures, which suggest that the percent of the voting age population who voted actually decreased slightly by .2 percentage points from 2004 to 2008, from 63.8% down to 63.6%. This is not inconsistent with the other scholars’ estimates, including those by Curtis Gans at American University and Michael McDonald at George Mason. All found that the actual (or best estimated) turnout in 2008 did not quite reach the projected figures that the pre-election hype suggested, with the increase/decrease ranging from about 1.6% to -.2%. Much of the overstated preelection hype was due to the tremendous increase in the use of early voting, something the Census Bureau survey also measured; it found that a record 29.7% of those voting did so through early balloting, a total consistent with what other sources report. Many media sources mistook the early voting as a sign that overall turnout would reach record levels. As we now know, and as the Census Survey indicates, turnout was about what it was in 2004, or perhaps slightly more, but considerably less than what was projected.
Of more interest, however, are the turnout rates among different voting blocs and what this portends for future elections. Consider the following table compiled by McDonald using the census bureau data. Voting by African-Americans, not surprisingly, increased by 4.9%, to 65.2%, from 2004 levels, while turnout among whites decreased by 1.1%, to 66.1%. That is a growth in the black vote, from 2004 to 2008, from 11% to 13% of the electorate (according to exit polls; see the NY Times exit poll website here). The Hispanic voting rate also increased from 2004 by 2.7 percentage points. That translates to 9% of voters in 2008, up from 8% in 2004.
Why are these figures important? Because (based on National Election Survey data and exit poll data) during the last three decades, nonwhite voters have become an increasingly larger proportion of the Democratic voting coalition in presidential elections, along with liberal white voters. According to an analysis by Alan Abramowitz, the proportion of nonwhite and liberal white voters in presidential elections increased from 25% in 1976 to 44% in 2008. What this suggests is that the proportion of voters “naturally” inclined to support the Democratic presidential candidate is on the rise, while the proportion of moderate white voters – who are the swing voters in most presidential elections – is declining.
There is a second trend that should worry Republicans as well: although the youth vote continues to lag behind all other age groups in terms of voting turnout, an increasing proportion of the under-30 age cohort (54% in 2008) is composed of nonwhites and liberals – precisely the voting bloc that tends to vote Democratic. The most Republican voting demographic age group in 2008, on the other hand, were voters 65 years or older. Again, McDonald provides a useful summary table:
There are two important conclusions to draw from this demographic snapshot. First, long-term demographic trends may favor the Democratic Party during the next several elections, unless Republicans can field a candidate that cuts across these demographic lines (say, a black woman? Condi Rice, anyone?) or reversing the voting trends among some groups (can you say immigration reform?) (I will present an extended post examining these voting blocs, particularly the youth vote, in the future.)
More importantly, perhaps, we see that the electoral coalitions for Republican and Democrats are diverging, with much less overlap in their potential voting support as the proportional size of the white moderate swing vote coalition decreases. To the extent that governing coalitions reflect voting coalitions (and vice versa), this means we should expect more polarization in our elected leaders and their policy proposals, not less.
My point – in case you’ve missed the theme of the last 31 posts – should be obvious: Obama is unlikely to bring change, in the form of a less polarized political agenda, because his voting coalition and a divided Congress, won’t allow it. Instead, all the demographics indicators suggest that rather than moving to the moderate middle, as he indicated during the campaign, Obama will instead be forced to tack Left and rely almost exclusively on the Democratic Party to push through his governing agenda – precisely the strategy he has adopted so far.
My broader point is that the media, and pundits, are frequently guilty of overstating the degree to which individual presidents are capable of charting their own political destiny. While there is room for some political maneuvering by presidents, and there are differences in presidential skill (and temperament), these factors are far less consequential than we think. Instead, the reasons behind the increasing polarization in recent decades has much less to do with Obama or Bush (or Clinton or Reagan) and much more to do with the intersection of a changing policy pool and demographic trends among voters – and the officials they elect to Congress. I will devote a detailed post on the causes of polarization, but for now, consider this image: somewhere in Dallas, a former president – cowboy boots on the table – is contemplating his political legacy. I would not be surprised to find him nodding his head in empathy with Obama’s strategic predicament, if not his policies.