What Can Save the Democrats?

During a stop in Madison, Wis., this week, I was grabbed by a phrase I had never heard before -- "sifting and winnowing." The friend who used it later explained that the term was from an 1894 report on higher education and is immortalized on a sign outside Bascom Hall on the University of Wisconsin campus. The plaque reads, "Whatever may be the limitations which trammel inquiry elsewhere, we believe that the great state University of Wisconsin should ever encourage that continual and fearless sifting and winnowing by which alone the truth can be found."

For those of us who analyze politics for a living, that is a great description of the process we undertake. Whether studying historical patterns, election results, or polling data, or when just asking questions and listening, we sift through all kinds of information. Then we winnow it all down. We attempt to discern what seems most relevant, and we then try to reach conclusions and develop a narrative that explains what has happened, may be happening, or will happen -- and why.

The catch is that different analysts see things differently, interpret them differently, and often reach different conclusions. Until the votes are counted on Election Night, we don't know who was doing the best job of sifting, winnowing, and drawing conclusions.

As the 2010 campaign begins to unfold, many smart and talented people are making varying predictions about the outcome of the midterm elections. Most analysts agree that Democrats will suffer losses, at least giving up the 16 House seats that the president's party generally loses in first-term, midterm elections. That's where the agreement tends to end.

In my view, Democrats have been in a free fall since summer, and unless something significant changes, they are headed toward the losses of the magnitude we saw in the midterm elections of 1958, 1966, 1974, 1994, and 2006. One difference between this year and 1994 and 2006 is that the party in power started developing serious problems more than a year ahead of the election.

Although no two cycles are exactly alike, history suggests that the indicators we're now seeing mean that the Democratic majority in the House is in grave danger and that Senate Democrats could easily see their ranks shrink to 52 or 53 seats. Today's signs are much like those that led me to predict in August 2006 that "unless something dramatic happens before Election Day, Democrats will take control of the House. And the chances that they'll seize the Senate are rising toward 50-50."

What could change the current trajectory, preventing the Republicans from gaining more than the 40 seats they need to take control of the House and from winning more than six or seven seats in the Senate? Some observers argue that if the Democrats pass some kind of health care reform bill, scaled down or not, they would appear less ineffectual and would change the current thinking that they have wasted the better part of the past year and come up empty-handed. That sounds plausible, but only if the public's perception of the Democrats' health care plan changes significantly. Democrats have not exactly been winning many perception battles lately. And in the end, would they really help themselves by enacting something that most voters say they don't like and don't want?

So what if pollsters can pick apart the Democrats' health care reform legislation and find considerable public support for many of its pieces. If people don't like the total package, those exercises are purely academic. In short, it's hard to see how passing health care is the way for Democrats to get out of this mess.

What about unemployment? The February Blue Chip Economic Indicators survey of 52 top economists estimates that unemployment for this year will average 10 percent, the same level that President Obama's Council of Economic Advisers forecast earlier this month. Some political scientists have said that there isn't a strong a correlation between high unemployment rates and midterm election losses by the president's party. In the post-World War II era, however, unemployment has never been over 8 percent during an election year except when the two parties shared control in Washington. The only midterm election held when unemployment topped 8 percent was in 1982, when Democrats controlled the House and Republicans had the Senate and the White House. Even in that year, unemployment crossed into double digits only two months before the election.

In short, unemployment isn't likely to fall enough to throw Democrats a lifeline in 2010. What could improve Democrats' prospects? All my sifting and winnowing has yet to give me a plausible answer.