Nonprofit, nonpartisan journalism. Supported by readers.


Why the Democrats have a decent shot at regaining Senate control (or close to it)

Over recent cycles, Democrats are twice as likely as Republicans to make Senate gains during presidential election years; Republicans are twice as likely to make gains during midterm elections. 

If we look at Cook’s rankings for the 24 currently-Republican-held seats that are up this year, he rates just 11 of the incumbents as “safe.”
Senate Photo Studio

Our political system has many quirks, one might even say many bugs, that don’t resemble much else in the democratic world. And one of them is among the reasons that the Democrats have a decent shot at taking over the Senate, or at least cutting deeply into the Republicans current 54-46 majority.

In five words: It’s a presidential election year.

You noticed that of course. But over recent cycles, Democrats are twice as likely as Republicans to make gains in the Senate during presidential election years, and the Republicans are twice as likely as the Democrats to make gains during midterm elections. Why?

In four words: It’s the turnout, stupid. Pardon that outburst. I didn’t mean to call you stupid. Especially not you, MinnPost readers. It just seemed like the best way to end that phrase, which is borrowed by Bill Clinton’s internal mantra from the 1992 presidential campaign: “It’s the economy, stupid.”

Bouncing turnouts

Anyway, just to finish the point, because of some of the vagaries of our strange (and, in my view, sub-optimal) system of politics and government, the voter participation rate always goes up by 15-20 percentage points in presidential election years (usually, in recent cycles, to within shouting range of 60 percent) and then back down in the midterm election two years later to a percentage in the low 40s or even upper 30s. (Even our “high” turnouts for presidential elections are quite lousy by international comparison.  Our low turnouts for midterms — when, bear in mind, the entire U.S. House and a third of the Senate is on the ballot — really, really stinks.)

Article continues after advertisement

But, back to the U.S.-only situation: The groups that tend to vote most reliably – whites, older voters, the affluent, more educated voters – are groups that are relatively rich with Republicans compared to the groups that vote less reliably – nonwhite, younger, poorer and less-educated voters, all of which tend to, on balance, vote more for Democrats, when they do vote. So, obviously, when those less-reliable voters drop out of the electorate because there is not the extra drawing power of a presidential race, you get a more Republican-heavy electorate. And then back to a more Democrat-heavy electorate for the next presidential race when turnout bumps back by about 20 percentage points.

Because of other vagaries too numerous or above-my-pay-scale to understand, this doesn’t always lead to the election of Democratic presidents.

Who gains when

But it does lead to a somewhat reliable fluctuation between presidential elections and midterms. I just toted up the net change in the U.S. Senate over the last 14 elections, including seven midterms and seven presidential election years. In two of those elections (one a midterm and one a presidential) neither party gained in the Senate. But in the six presidential elections in which one party gained seats, the Dems gained in four compared to two for the Repubs. In the six midterms in which the partisan makeup of the Senate changed, the Repubs gained in four and the Dems in two.

Perhaps that’s enough to reinforce the generalization. Now, of course, these elections are not all the same and may vary greatly in conditions of the country and also in the popularity of the incumbent president.

So, if you were to turn to one of the pundits who rates all races, to see how things are looking this year you would find that according to Charlie Cook of the Cook Political report, first of all, of the 34 Senate seats that are up this year, 24 are held by Republicans and just 10 by Democrats. If you do the math, you’ll note that those senators were last elected in 2010, which was a midterm, and the Republicans who are up for re-election were elected in 2010, a very good year for Republicans.

But if you look at Cook’s ratings, of the 10 seats currently occupied by Democrats, he rates eight of them as “safe,” meaning heavily favored to win another term. He rates just one of them (Sen. Michael Bennett of Colorado), as merely “favored.” Colorado is, of course, a purplish state and many — but not all — of the “safe” Democrats are in solid blue states. So that’s obviously a factor as well.

But only one Senate seat currently held by a Democrat is rated as a “toss up.” That’s the Nevada race, where long-time Democratic Sen. Harry Reid is retiring, so it’s an open seat, which is often one reason a Senate seat comes into play.

Just 11 incumbents rated as ‘safe’

If we look at Cook’s rankings for the 24 currently-Republican-held seats that are up this year, he rates just 11 of the incumbents as “safe.” He rates three more as “likely” to be re-elected, three more (including 30-year Senate veteran John McCain of Arizona) in which the state is only “leaning” toward the incumbent.

And then Cook rates an impressive seven races involving seats currently held by Republicans that Cook rates as “toss-ups.”

Article continues after advertisement

All seven are in purplish states, so that’s a big part of why they in “toss-up” territory. In all seven, by the way, the incumbent is running for re-election, so that’s not the reason there are so many. But, perhaps more impressively, all six of the Republicans on Cook’s toss-up list are just completing their first term. That means they all were elected for the first and only time in 2010, a midterm election featuring a smaller, more Republican-rich electorate. None of them has yet demonstrated that they can carry their purplish state in a presidential election year featuring a larger, Democratic-heavy electorate.

So yeah, you can, and should, look at each state, each race, the strength of the challengers and so forth, and all of those are big factors. You can, and should, look at the relative popularity of the major party nominees and whether their relative popularity will help or hurt their party’s nominees for the House and Senate. All of those are valid factors and many more.

But one of the fundamental reasons so many Republican Senate seats are “in play” this year is the one I’ve been yammering on about thus far. Presidential election years = elevated turnout = generic benefit for Dems. Midterm = lower turnout = benefit for Repubs.

2018 will favor Republicans

By the way, if you look ahead to 2018, everything I’ve said is true in reverse. It’s a midterm, featuring a predictably lower turnout, and there are 25 currently-Dem-held Senate seats up for election compared to just eight currently Repub-held seats. But then, in 2020, a presidential election year, 22 currently-Republican Senate seats will be on the ballot, compared to 11 Dems.

(By the way, there are two current senators who were elected as Independents but caucus with the Democrats. For purposes of this analysis, I’ve counted them as Democrats. Neither of them is up this year anyway.)

Is this all part of the Framers’ brilliant plan? Not even slightly. The Framers didn’t know anything about turnout. They didn’t even intend for senators or presidents to be directly elected. I can’t imagine anyone who was designing a system who would think it was a good idea to have voter participation rates regularly bounce up and down the way they do. To me, it’s not a feature, it’s a bug.