Casey Stengel said, “Predictions are dangerous, especially about the future.” (Or, according to some sources, “I never make predictions – at least not about the future.” But what’s the point of being a famous wit if the folk process isn’t allowed to edit your wisecracks for you?) On the other hand, Eugene Volokh cites Grant Nelson for this reply:

“If you can’t predict the future, what can you predict?”

Still, writing down in advance what you think will happen is a good way to discipline your optimism.

I was very optimistic at the end of last week. The Gallup and NYT polls rocked me back, showing a big swing toward the Republicans in the final weekend, but Ipsos-Reed, Pew and ABC show no no such movement, and Ipsos-Reed notes that a weekend poll is always more Republican than a weekday poll, which accounts for some of the apparent swing. Zogby, the only outfit that got the 2000 Presidential race right, makes it 51-49 for the Democrats in the generic Congressional ballot. Ipsos-Reed has a 2-point edge for Republicans among the likely voters, but it also has 10% undecided, and their big tilt (57-32) toward “wrong track” rather than “right direction” makes them sound likely to break Democratic. Some 14% of the expected electorate has voted early or absentee, breaking Democratic 51-41. So now I’m a little optimistic.

The big question mark is the turnout models, which in the past have understated the minority vote, and the balance of who’s mad enough to turn out. That Ipsos-Reed gives the Democrats a 2-point edge among “hardcore voters” — people who have been paying a lot of attention to the election — may or may not be a hint here. I’d be surprised if “issues” such as the New Jersey Senate switch and the Mondale memorial service really have much traction among people marginal enough in their interest in politics to be on the edge between voting and not voting, but that wouldn’t be the first time I’d been surprised.

Tom Edsall told me six months ago that last year’s gerrymandering had been so efficient that there simply weren’t enough seats actually in play to make a Democratic takeover of the House plausible, and nothing since has changed that. I gather from people who have been following it closely that the best guess is roughly no change in the House.

On the other hand, the Democrats figure to do very well in the gubernatorial sweepstakes, picking up Pennsylvania, Illinois, and Michigan virtually for sure, plus probably half a dozen smaller states (New Mexico, Arizona, Wisconsin, Kansas, Maine, Tennessee, Massachusetts, Arkansas, Minnesota, Oklahoma, and Wyoming are all candidates, roughly in decreasing order of likelihood, with Massachusetts being about even money.) The Democrats figure to lose New Hampshire and Alaska, probably South Carolina and Hawaii, and maybe Maryland. (Georgia and Oregon now look safe. It’s barely concievable that California is in play, but I’m pretty sure Simon has managed to blow it.) Two of the big prizes, New York and Texas, are clearly out of reach, and the South Florida voting problems probably put Florida out of reach as well, even if the late polls showing a swing toward Jeb are wrong.

But the big action is in the Senate. The most honest and accurate thing anyone can say about the Senate is that no one knows what’s going to happen. In some ways, any prediction obscures that basic fact. Still, here goes:

Democrats have hopes for pickups (again, in descending order) in Arkansas, New Hampshire, Colorado, Texas, South Carolina, North Carolina, and Tennessee. The first three of those look better than even to me, with Arkansas nearly a lock. If that Texas poll showing Kirk down by only 1 point is right, and if it’s based on a turnout model that understates the likely impact on minority voting of a Latino-black ticket, then I’d have to bet on Texas as well, but those are two big ifs. The other three are longshots.

On the minus side, Georgia, Missouri, South Dakota, Minnesota, and Louisiana are all in play. (I think the Republicans have written off New Jersey by now.) Of those, I would rate Georgia and Missouri worse than even money, South Dakota a tossup, and the other two leaning Democratic. Of those twelve, I give the Democrats the edge in five

So there are twelve races in play, and the Democrats only need to win five to hang on to control of the Senate.

Of those I rate one (Ark.) likely Democratic, four (Colo., NH, Minn., LA) leaning Democratic, two (SD and TX) real tossups, four (GA, MO, NC, SC) leaning Republican and one (TN) likely Republican. That suggests a net Democratic pickup of one seat. To retake control (ignoring for the moment the Lincoln Chafee question), the Republicans would have to take everything I’ve put on their side of the ledger, both of the tossups, and one of the Democratic “leans.” That doesn’t seem very likely to me, but it doesn’t seem unlikely enough to give me a good night’s sleep, either.

[For those of you who, like me, have been taking some comfort in the Iowa market prices showing the Republicans with only a 30% chance of taking the Senate, here’s the bad news: the prospectus says that the Republican Senate contracts pay off only if the Republicans have 51 seats, which is one more than they need to organize the Senate. So the analysis above is somewhat more optimistic than the market prices would indicate.]

Author: Mark Kleiman

Professor of Public Policy at the NYU Marron Institute for Urban Management and editor of the Journal of Drug Policy Analysis. Teaches about the methods of policy analysis about drug abuse control and crime control policy, working out the implications of two principles: that swift and certain sanctions don't have to be severe to be effective, and that well-designed threats usually don't have to be carried out. Books: Drugs and Drug Policy: What Everyone Needs to Know (with Jonathan Caulkins and Angela Hawken) When Brute Force Fails: How to Have Less Crime and Less Punishment (Princeton, 2009; named one of the "books of the year" by The Economist Against Excess: Drug Policy for Results (Basic, 1993) Marijuana: Costs of Abuse, Costs of Control (Greenwood, 1989) UCLA Homepage Curriculum Vitae Contact: