Click on the image to download the spreadsheet.
I think I’m losing my mind (and yeah, I know; some of you will present evidence that this happened a LONG time ago).
Let me apologize in advance if I wrote this post before. I thought I had, but I can’t seem to find it. So here goes, perhaps again…
About a week after the election, Cindi Scoppe wrote about the terrible won-loss record of the candidates that The State had endorsed in 2016:
Two-thirds of the candidates our editorial board endorsed in last week’s election lost. We have never seen numbers like that since I joined the board in 1997 — and as far as I can tell for decades before that. Normally, it’s more like 25 percent….
Of course, all that means was that two candidates lost, as the paper had only endorsed in three races in the general, instead of the usual 10 or 20 that we’d back in the days when we had the staff to do it.
But taken as a percentage (which is a pretty meaningless thing to do with a sample of three), I’m sure it was a bitter pill. I wondered why Cindi hadn’t offered the running total from over the years to show just how much of an anomaly that was. Apparently, she just didn’t have the numbers at hand. But I did, at least through 2008, my last election at the paper. And it just took a few minutes to update the table with results from 2010, 2012, 2014 and 2016.
Why did I have those numbers? Because in 2004, I got fed up. We’d hear from bitter candidates who did not get our nod who claimed that they didn’t want it anyway, because our endorsement was “the kiss of death.” Well, I knew this wasn’t true, not by a long shot. (I also knew by their behavior that these very people were usually quite eager to get our endorsement, until they didn’t. Then it was sour-grapes time.) But I didn’t know how wrong they were. I didn’t have numbers.
Then there was the other problem: Democrats regularly claimed that we only endorsed Republicans, and vice versa. I knew that was untrue, too (any casual, unbiased observer knew better than that). But again, I couldn’t quantify it.
I had resisted keeping track of such things in the past, for a couple of reasons. First, endorsements were arguments as to who should win, not predictions of who would win. A lot of people failed to understand that, and would demonstrate their lack of understanding by saying we got it “wrong” when our endorsee lost. No, we didn’t. We weren’t trying to make a prediction. And why would we have kept track of how many Dems or Repubs we backed, when we didn’t care about party?
But as I said, I was fed up, and I wanted to lay all the lies to rest permanently. So I dove into our musty archives for several hours, and came up with every general-election endorsement we had done starting with the 1994 election. Why that date? Because that was my first election as a member of the editorial board, and since then we’d had 100 percent turnover on the board — so it was ridiculous to hold any of us responsible for editorial decisions made before that date.
And I stuck to general elections, to keep it simple. After all, that’s the only time one is choosing between Democrats and Republicans. And digging up the primary endorsements would have taken more than twice as much time. I’ll acknowledge this freely, though: Our won-loss numbers wouldn’t have been as good if I’d tried to include primaries, because we were staunch centrists, and primary voters tend to have more extreme tastes than we did.
What I found in 2004 was that since 1994, about 75 percent of “our” candidates had won, and we’d endorsed almost exactly as many Democrats are Republicans. I updated the numbers after the 2006 and 2008 elections.
Anyway, after Cindi’s column, I updated my spreadsheet with numbers from the years since I’d left the paper, including 2016, and here’s what I found:
The running percentage of “wins” had dropped slightly since 2008, with 72.26277372 percent of endorsees winning since 1994. Even though the paper had a big year in 2014, with eight out of 10 endorsees winning. (When I had first compiled the numbers in 2004, our batting average was .753.)
The partisan split became more nearly even. As of 2008, we were favoring Democrats slightly with 52.6 percent of endorsements going to them. Now, that’s down to 50.37 percent, about as dead-even as you can get: 68 Democrats, 67 Republicans and one independent since 1994. The paper has favored Republicans 13-8 since I left.
Anyway, since I’d gone to the trouble of running the numbers, I had meant to write a post about it. If I did before now, I can’t find it. So here you go…
Here’s the spreadsheet.