Is Hillary Due For A Comeback?

There has been some scoffing at Clinton pollster Mark Penn's memo issued yesterday arguing that Hillary Clinton can still win more delegates than Barack Obama. The memo contains a certain amount of campaign boilerplate:

Hillary is the only candidate who can deliver the economic change voters want--the only candidate with a real plan and a record of fighting for health care, housing, job creation and protecting Social Security.

But, hey, he's paid (and very well) to say things like this. And there's independent polling data that seem to support his argument.

Start with Pennsylvania, which votes April 22. Quinnipiac today released a poll showing Clinton leading Obama there 52 to 36 percent. Whites back Clinton 58 to 31; blacks back Obama 71 to 10. Since Pennsylvania's population is only 10 percent black, that accounts for Clinton's big lead.

Then look at Ohio, which votes March 4. Here Quinnipiac shows Clinton ahead 55 to 34 percent. Whites back Clinton 64 to 28; blacks back Obama 64 to 17. Ohio's population is 11 percent black. Quinnipiac's Peter Brown (whom veterans of the campaign trail will remember as a first-rate reporter) explains why Clinton seems to be doing so well in Ohio (and, by implication, demographically similar Pennsylvania) after losing eight straight contests:

Ohio is as good a demographic fit for Sen. Clinton as she will find. It is blue-collar America, with a smaller percentage of both Democrats with college educations and African-Americans than in many other states where Sen. Obama has carried the day. If Clinton can't win the primary there, it is very difficult to see how she stops Obama.

Quinnipiac's result is similar to two other recent Ohio polls. Rasmussen has Clinton ahead 51 to 37 percent; SurveyUSA has her ahead 56 to 39 percent. The only Ohio poll taken in January, by the Columbus Dispatch, showed Clinton ahead of Obama 42 to 19 percent. Obama has apparently made gains since then. But so has Clinton.

In the other big state that votes March 4, Texas, it seems that there has been no public poll since last April(!). Texas's population is 12 percent black and 32 percent Hispanic, so we can expect the Democratic primary electorate there to be about 20 percent black and perhaps 15 to 20 percent Hispanic.

One primary Penn did not stress in his memo was Wisconsin. The Clinton campaign line has been that the post-Super Tuesday February contests are all dismal ground for their candidates. But the Wisconsin polling data tell a different story. Scott Rasmussen shows Obama leading Clinton by only 47 to 43 percent. This is similar to Strategic Vision's Wisconsin survey, which shows Obama ahead 45 to 41 percent. Wisconsin's population is 6 percent black and 3 percent Hispanic.

How can Clinton be doing so much better here than she did in Maryland and Virginia? One reason is that there are smaller percentages of black voters in these states. Another, probably more important, reason is that the white Democratic primary voters are different. In Maryland and Virginia, they tended to be quite upscale and on the young side, especially in the big suburban counties outside Washington, D.C. In Wisconsin, Ohio, and Pennsylvania, they're much more downscale. At a time when Clinton and Obama are essentially tied in national polls, it stands to reason that if Obama is ahead in states like Maryland and Virginia, Clinton will be ahead in Wisconsin, Ohio, and Pennsylvania.

Texas is another, interesting story. Texas doesn't have party registration, and, historically, huge numbers of white voters participated in the state's Democratic presidential primary--1.3 million in 1980, 1.8 million in 1988, 1.5 million in 1992. That number plunged downward to 786,000 in 2000 and 839,000 in 2004, even though the state's population grew from 14 million in 1980 to 2 million in 2004. The obvious conclusion: An awful lot of white Texans began voting in the Republican primary again. This year's Texas Democratic primary could turn out to be largely a battle of minorities, with blacks voting heavily for Obama and Latinos, as in most other states so far, heavily for Clinton. In this battle Obama will undoubtedly have an organizational advantage, both because his campaign-- unlike hers-- has done organizational work in the post-Super Tuesday states and because of the strength of pre-existing black turnout organizations. As for white Democratic primary voters, upscale Texans still tend to be heavily Republican, though a little less so than 15 or 20 years ago--very much contrary to the pattern in Northern Virginia and Montgomery County, Md. White downscale voters in southern states have generally gone for Clinton, but not by overwhelming margins. Of the four states we've looked at here, Texas appears the most problematic for Clinton, though she's on far stronger ground there than in the already concluded post-Super Tuesday contests.

Next: A look at the prospects for delegate counts in these upcoming primaries

By Michael Barone