Swing voters

In today’s (20-VIII) NY Times, T. Edsall makes an important point {Obama campaign, are you listening?} about the 2012 election. The constant media drumbeat about “swing voters,” now defined as about 3-5% of the electorate, and further refined to include only those in the swing states, like Ohio or Florida, misses a key lesson of the election cycle 2008-2010: if Obama and the Democrats are to win, they must convince their Democratic base, and especially the new constituencies of that base, to show up.

In 2010, looking at a sample of lost Democratic House seats, in most cases candidates lost because of substantially lower Democratic turnout, especially in heavily Democratic towns or counties.  In a smaller number of cases [among the 20 or so I checked], Republicans did win by attracting what appeared to be swing voters to win their seats. [Gov Christie, in NJ in 2009, also did so.]

Let me give the example of elections in my home state of Virginia. The Republican candidates in the two most recent non-Presidential year state-wide elections, George Allen (Senate, 2006) and R. McDonnell (Governor, 2009) got virtually identical totals:  1.166 million and 1.163 million votes – Allen lost, McDonnell won.  Why?  In 2006, Jim Webb got 1.175 million votes, to become a Senator; running for Governor in 2009, C. Deeds got 0.82 million votes, to get crushed.  That’s a loss of over 350,000 votes, roughly 30% of the Democrats’ total.  Where did Deeds, a conservative Democrat, lose votes?  Let’s take what Allen, in 2006, called the “Communist part of Virginia,” the heavily Democratic towns of Arlington and Alexandria.  In 2006, Webb got 86k votes, while Allen got 32k; in 2009, McDonnell got those same 32k votes, but Deeds got only 59k, a decline of 31%.

Presidential elections bring out a lot more voters, especially in 2008.  Consider that Bush, in 2004, and McCain, in 2008, got exactly the same number of votes in Virginia (1.72 million), the same pattern we saw above with Allen and McDonnell.  Kerry lost; Obama won.  Why?  Obama convinced an additional 500,000 Democratic voters to show up, so he polled 1.96 million votes to Kerry’s 1.45 million.

In the 2009 election, McDonnell got 67% of the Republican voters of 2008 to show up;  Deeds got only 42% [sic] of Obama’s voters to turn out.  In the two towns mentioned above, Obama got 130k votes, Deeds only 59k.  African-American voters, in places like Alexandria, Richmond, and Newport News, overwhelmingly abandoned Deeds.  Young people in my Virginia extended family simply did not vote at all in 2009.  They form a nice sample: of the 5 under-30s eligible to vote in 2004, none did; in 2008, it was 6 for 6 (all Obama); in 2009, I believe it was 0 for 6.

Swing voters matter, but Obama’s most important task in states like Virginia is to make sure that the roughly half a million additional Democratic voters who turned out in 2008, show up again in 2012.  Obama got an amazing 35% more votes than Al Gore received in 2000.  McCain got only 3% fewer votes than George Bush received in 2004 (59.9 m v. 62m, and got 18% more than Bush received in 2000).

Nationwide, we can count on Romney getting that same 60 million or so votes in 2012.  Obama got 17.6% more votes than Kerry (69.4m v 59m).  The difference in the election was far less voters swinging from Bush to Obama, than new Democratic voters swamping the Republicans in states like Virginia (which went for Bob Dole in 1996, quite apart from going to Bush in 2000 and 2004).

If Obama is going to win, he needs to focus on his Democratic base, particularly on his base of new voters from 2008.  He needs to re-energize the young voters who proved so critical in places like Virginia and North Carolina.  He needs to re-energize the African-American community, who turned out in much larger numbers than usual in 2008, and voted 95% for Obama.  He needs to solidify his appeal to Latino voters (67%; Kerry got roughly 60%).  Those three groups, not some ill-defined swing voters, will decide the 2012 Presidential election.  Consider that African-Americans and Latinos constituted 16.9% of the electorate in 2000 and 17% in 2004, but 19.5% in 2008.  African-American turnout, for the first time, matched White turnout, in percentage terms (roughly 2/3rds of eligible voters).  Rising Latino and African-American participation translated into 4.4 million additional voters (2008 v 2004): 88% of the new voters of 2008 came from these two groups.  Obama got roughly 2.75 million votes more than Kerry got from those two groups of new voters alone.

Republicans know this; they have responded with a grotesque campaign designed to eliminate as many African-American and Latino voters as possible, through new voter registration laws and other anti-democratic (and anti-Democratic) activities.  The worst abuses take place in states like Ohio, likely to be critical in the Electoral College.  Nationwide, consider that the White population grew by 5.7%, while the Black population increased by 12.3% and the Latino population by 43%: Latino population growth (15.2 million) exceeded that of all other groups combined (US Census Bureau, Overview of Race and Hispanic Origin: 2010, table 1).  For simple demographic reasons, Latino and Black voters will be even more important in 2012 than in 2008.

The Asian population, too, grew by 43%: the US added as many Asians as Blacks between 2000 and 2010.  “Asian” in the Census Bureau’s definition means East, Southeast, and South Asia only; that grouping includes the Philippines (Filipinos are the second largest Asian population in the US, after Chinese; those two groups account for just under half of all “Asians”).  A report by the Asian American Legal Defense and Education Fund (AALDEF) on Asian voting in 2008, which relies on exit-poll interviews with more than 16,000 voters, found that 31% of them were voting for the first time – 81% of the first-time voters chose Obama.  Their respondents claimed to have voted 76% for Obama (other sources suggest he got as little as 63% of the “Asian” vote).  One statistic stands out: their sample includes only 5% Filipinos, who were the second most Republican Asian constituency (Vietnamese are heavily Republican), and who are thus substantially underrepresented in the sample.  “Asian” voters give particularly high importance to the health care issue: across all nationalities, 70% of “Asians” in the exit polls of 2008 supported a universal health care system.

Among voters 18-29, Obama made even greater advances.  Bush and Gore split that vote in 2000, while Kerry won it, 54-46, about the same margin as Clinton in 1996.  Obama got an amazing 66% of this vote in 2008.  Obama got 69% of first-time voters to select him.  Among young Asians, Obama got 88% of the vote.

Unsurprisingly, given the Republicans’ misogynist agenda, Obama did best with young women, who gave him 69% of their votes.  {And his candidacy particularly energized young Black women, who had the highest percentage turnout of any large constituency.} Pew post-election surveys found that 28% of voters under 30 had attended a campaign event, which must be some sort of record.  Certainly here in Northern Virginia, young Obama volunteers were everywhere in 2008.  Nationwide, the youth vote went up only 1 percentage point, but in some key states – such as North Carolina, Virginia, and Indiana – it went up 4 or 5 percentage points .  The combination of higher turnout, and greater margin, among African-Americans and young voters swung those states to Obama in 2008.  Aside from Jimmy Carter taking North Carolina in 1976, it was the only time since 1968 that a Democrat carried any of those three states.   Together, they have 39 Electoral votes.

Indiana may be a traditional Red state, but its African-American population increased nearly 17% between the 2000 and 2010 censuses, and the Latino population went up 82%; in contrast, the White population went up only 2.8%.  Obama barely won Indiana in 2008; to win in 2012, he needs to concentrate not only on swing voters, but on getting out the vote in Latino and Black communities, and on energizing young people again.

Florida has little to do with the youth vote, at least among Whites.  In Florida, the White population has gone up substantially, by 13%, but the Black population increased by more than double that rate (28%) and the Latino population by 57%.  No wonder the Republicans are doing everything in their power to prevent Blacks and Latinos from voting, as they did successfully with Blacks in 2000, when their efforts stole the election by eliminating enough Black votes for Bush to be declared the winner in Florida, and overall.  In Florida, the new White voters are often over 60, and we can see already the heavy focus on Medicare in campaign discussions since Romney named Ryan as his VP choice.  Given that Obama did not do well with voters 65+ (-7% in 2008, as against Kerry’s -1%), and that recent polls suggest Romney has a 10-point lead among this demographic, we can expect to see even more about Medicare.  This group votes in higher percentages than any other, so getting back to a 7-point deficit would probably assure an Obama win.  Dropping back down to the levels of Democratic support in 2010 (-21) would spell disaster, just as it did then.

As for the gender gap, the contrast between the two parties could not be starker: Obama crushed McCain among unmarried women (2:1), but it’s worth noting that McCain actually defeated Obama among married women, 54-46 (Gallup Poll election analysis).  The astounding gaffe by the Republican Senatorial candidate in Missouri, Rep. Todd Akin, can only draw the comparison in sharper lines.

So, quick memo to Barack Obama and company:  think state-by-state.  In Florida, go for the Medicare crowd, but don’t forget that 43% of Florida’s population is African-American, Latino, or Asian.  In Indiana, Virginia, and North Carolina, focus on turnout among young people, African-Americans, and Latinos.  Above all, get away from the mindless fascination with “swing voters” and think more carefully about what has really decided recent elections.

Statistical data are drawn from official state election returns, available on state government websites, and from the many fine studies done by the Pew Research Center, available online at: http://www.people-press.org

Leave a Reply

Your email address will not be published. Required fields are marked *