Cleaning out my office on Friday, I came across an interesting piece by Gerald F. Seib. He’s a Wall Street Journal reporter who covers Washington politics. The piece is about ending gridlock in Washington, and Seib argues that that will happen when voters punish politicians who “eschew compromise.” He then gives data showing that there has been a shift in voters’ attitudes on this. Seib writes:

Precisely 50% said they would pick the candidate who favored compromises to gain consensus, surpassing the 42% who said they would prefer the candidate who sticks to his or her positions.

That may not seem like much of a shout-out for compromise and consensus, except that it represents a stark reversal from just four years ago. Then, a clear majority, 57%, said they preferred candidates who stuck to positions, while just 34% opted for those who would compromise. Do the math: That’s a 23-point advantage for opposing compromise four years ago, compared with an eight-point edge in favor of it now. A 31-point swing is a significant movement.

So far, so good, in the sense that he makes his point. But then Seib digs beneath the data, writing this:

A lot of the pro-compromise sentiment shows up among Democrats, who think Republicans have stiffed their president, Barack Obama, and want that to change. But they aren’t alone. Even among Republicans, the share saying they favor a candidate who will compromise has risen 13 percentage points since 2010, the year of the midterm in which the GOP took back control of the House.

Notice that first sentence: Democrats want Republicans to compromise. Really? Who’d a thunk it? I know hardly anyone who doesn’t want his opponent to compromise. Seib was presumably writing about voters being willing to punish their own side for not compromising. This is not evidence.