In various posts about the increasing acrimony over the tight Democratic campaign for the presidential nomination, people have speculated over whether the nomination of Sen. Hillary Clinton, particularly if achieved by superdelegates or the disputed Florida and Michigan delegations, could depress Democratic turnout, particularly among black voters, sufficiently to affect the outcome of a general election.

I have generally assumed that the black vote is largest in states that strongly trend Blue or Red and that an anemic black voter turnout would be important in only a few battleground states.  However, we all know the danger of assuming things.  At PJM, Andrew Walden writes:

The 90% solid black vote tips the balance in favor of Democrats in Missouri, Michigan, Illinois, Ohio, New Jersey, New York, Pennsylvania, Connecticut, Delaware, and Maryland. Without these votes, Democrats would simply cease to function as a national party. The Clinton Foundation’s multi-million-dollar foreign slush funds are what Democrats would be sacrificing to keep that percentage at 90% and stay in power if the Clintons cannot learn to go gracefully.

Unfortunately, the link in the above quote lists the black population by state, which is not the same as the black vote by state, let alone the Democratic black vote by state.

Fortunately, those numbers from the 2000 and 2004 elections have been compiled for a number of these states from reputable sources by the left-liberal Joint Center for Political and Economic Studies.  Voter survey numbers from past cycles are also available through CNN.  Accordingly, what follows is a closer look at these states, supplemented where possible by recent head-to-head polling.  Where available, I have also added exit poll data regarding voter satisfaction with a Clinton or Obama nomination, though I caution against extrapolating from a primary poll to the general public.

In Missouri, the black vote was 8%, down from 12% in 2000.  Kerry got 90% of that vote, but Gore got only 84% of the higher percentage in 2000.  So the Democratic black vote has been between 7.2-10% of the Democratic vote in Missouri.  Bush won Missouri by 7% in 2004, so the Democratic black vote did not tip the balance there in 2004.  To the contrary, the black vote would have to increase substantially to provide the margin of victory for the Democrats.  As noted recently, in Missouri, McCain led Clinton 50%-44% and Obama 51%-40% (a statistically significant lead) in a SUSA poll.  A more recent Rasmussen poll has McCain only a point or two ahead of either Clinton or Obama.  FWIW, the Missouri Democratic primary exit poll this year showed that 21% would be satisfied only if Clinton wins, while 28% would be satisfied only if Obama wins.  A depressed black vote obviously gives McCain an advantage against Hillary, while Obama might have to boost black turnout, perhaps significantly, to gain an edge there.

In Michigan, blacks accounted for 13% of the overall vote, with 90% going to the Dems on average, for an 11.7% Democratic black vote.  In 2004, Kerry beat Bush by a mere 3%, so the black vote was determinative in this case.  This year, the resolution of disputes over the Michigan delegation and the role of superdelegates could have an impact on turnout.

Illinois had a 10% black turnout in 2004, down from 14% in 2000, with 90% going to the Dems on average.  Kerry beat Bush there by 9%, so the black turnout could be determinative in this case.  FWIW, the Illinois Democratic primary exit poll this year showed that 17% would be satisfied only if Clinton wins, while 35% would be satisfied only if Obama wins — a 2:1 gap.

In Ohio, black turnout has been about 9-10% overall.  Gore carried 89% of that vote, while Kerry carried only 84% (perhaps because of that gay marriage referendum).  Thus, the Democratic black vote has been about 8-9%.  Bush won Ohio by 2% in 2004, so the black vote was not determinative, though a large increase in black turnout could be determinative, all other things held equal (e.g., no offsetting anti-black vote).  In Ohio, SUSA had McCain leading Clinton by an insignificant 48%-46%, but ahead of Obama 50%-43% (still within the MoE, but close to an actual lead).  The most recent Quinnipiac poll has McCain ahead by only a point or two against either Clinton or Obama.  This is similar to Missouri, though Obama might not have to boost black turnout above the jump it already took, if the gay marriage vote accounted for the decline in the Democratic share in 2004.

In New Jersey, overall black turnout was 14% in 2004, but Kerry only took 82% of that vote because Bush improved his standing with that block over 2000.  Thus, the Democratic black vote was roughly 11.5%  Kerry’s margin of victory was 7%, so the black vote was a factor in this case.  FWIW, the NJ Democratic primary exit poll this year showed that 28% would be satisfied only if Clinton wins, while 21% would be satisfied only if Obama wins.

In New York, blacks averaged 12% of the overall vote; Gore and Kerry both took 90%, for a Democratic black vote of about 10.8%.  Kerry won New York by 19%, so the black vote was not determinative in 2004.  FWIW, the NY Democratic primary exit poll this year showed that 29% would be satisfied only if Clinton wins, while 17% would be satisfied only if Obama wins.

In Pennsylvania, black turnout was 13% in 2004 — a significant jump from 7% in 2000.  However the Democratic share declined to 83% for Kerry from 90% for Gore.  Thus Gore had a Democratic black vote of 6.3%, while Kerry got about 10.8% — perhaps because the state is his second home.  Kerry’s margin of victory was only 2%, so the black vote clearly made a difference.  The most recent Quinnipiac poll has Clinton +6 and Obama +1 over McCain.  The new Rasmussen poll now has McCain +2 over Clinton, but Obama +10 over McCain.  (Before anyone cheers or panics, note that the Rasmussen poll was a single-day poll taken amid a string of Obama victories and has — like most of these polls – a MoE of about 4.5%)  A depressed black vote here would have implications for Clinton.  Obama looks good here at the moment, though the prior polling had him down by eight points.

In Connecticut, there was only 6% black turnout; it was not sufficient for the distribution to be recorded in the exit poll numbers.  Assuming a 90% Democratic share, the final number would be about 5.4% — far less than Kerry’s 10% margin of victory in 2004.  Indeed, the black share of the Connecticut population is 9%, so the Democratic black vote would have had to substantially over-performed for it to make the differnce in Connecticut.  FWIW, the CT Democratic primary exit poll this year showed that 22% would be satisfied only if Clinton wins, while 24% would be satisfied only if Obama wins.

In Delaware, black turnout was 20% of the total in 2004, but Kerry took only 82%, for a final 16.4% Democratic black vote share.  Kerry’s margin of victory here was only 7%, so the black vote was a definite factor.  FWIW, the DE Democratic primary exit poll this year showed that 25% would be satisfied only if Clinton wins, while 24% would be satisfied only if Obama wins.

In Maryland, black turnout has averaged 23-24% of the total, with the Democratic share averaging a little over 90%, resulting in a 21.5% Democratic black vote share.  Kerry won that state by 13%, so the black vote was a factor, even with this large margin.  FWIW, the MD Democratic primary exit poll this year showed that 16% would be satisfied only if Clinton wins, while 26% would be satisfied only if Obama wins.

Finally, while on this topic, note that Obama guarantees that black turnout will increase by a minimum of 30% if he is the nominee, though the WaPo’s number crunching suggested that was unlikely with respect to Obama’s example of Mississippi — and that Obama would still lose the state if that happened.

Obviously, it is still far too early in the process to draw any sort of firm conlcusion from these numbers.  White’s claim about the import of the black vote turns out to be correct, except with respect to Connecticut and New York.  But the decline in the black vote would have to be significant in most of those states to tip them in favor of McCain, based on the current data.  If the exit poll data was representative of the degree of apathy that might result in a general election, the acrimonious Democratic campaign might be hurting the nominee’s ultimate performance — but that is a highly risky extrapolation.