The midterm of his discontent. Photographer: Saul Loeb/AFP/Getty Images
The midterm of his discontent. Photographer: Saul Loeb/AFP/Getty Images

Harry Enten at fivethirtyeight.com today gives us some historical numbers on midterm elections. The numbers support what I said yesterday: There is more downside risk for Democrats going forward in this election cycle than for Republicans, at least when it comes to larger partisan tides.

Using data going back to 1938, Enten compares Gallup polling of a generic Congressional ballot -- anonymous Democrat versus anonymous Republican -- in the first quarters of midterm-election years with results in October of the same years. His conclusion? The position of the president’s party tends to deteriorate over the course of the midterm election year. The average decline is a bit more than five percentage points. In 19 election cycles he studied, the president’s party has only rallied five times, and four of those produced just a one- or two-point change. On the other hand, in four different elections the president’s party has lost more than 10 percentage points over the year. The biggest gain for the president's party (6 percentage points) happened only once. Meanwhile, the president's party lost 6 percentage points or more nine different times.

Why? Enten doesn’t analyze it, but the most likely big movers of public opinion over that kind of medium-length time frame are negative -- recessions, scandals and the like.

Is there any good news for Democrats here?

Not much. First, most of the large poll movements in Enten’s chart appear to be regression to the mean. Three of the four instances in which polls changed more than 10 points -- all negatively for the president's party -- came after the president’s party held a 20-point lead or more in the generic ballot early in the presidential year. So perhaps a major decline for Democrats is unlikely. Unfortunately for Democrats, however, the one significant movement in favor of the president’s party was probably a result of exactly the same phenomenon. In 1958, Republicans improved from an awful 20-point deficit to a less awful 14-point gap. They received a shellacking in November, anyway. With the generic ballot measure essentially even now, there is no precedent for a sizable rally by the president's party under similar circumstances.

Beyond that, what Democrats can hope is that President Barack Obama's job approval rating has bottomed out after he shed his re-election honeymoon bounce over the first nine months of 2013, and that the marginal improvement that began last fall will continue throughout 2014. Enten’s data suggest that is a plausible scenario. It's just not one that will dramatically improve Democratic prospects. This just isn't a good national environment for Democrats.

To contact the writer of this article: Jonathan Bernstein at Jbernstein62@bloomberg.net.

To contact the editor responsible for this article: Francis Wilkinson at fwilkinson1@bloomberg.net.