Wednesday, August 3, 2016

My Vote Counts, Part 2: Spoilers

I was going to save discussion of spoilers until later, but I probably need to get into it before we get into state-by-state voting patterns so we can come up with some sample scenarios for gauging how likely it is third party spoiler is going to flip a state and bring about the end of life on this planet that will inevitably result from the election of the Greater Evil (whichever one of them that is). All facts and figures come from either 270towin.com or Wikipedia unless noted otherwise. If you're one of those people that rejects Wikipedia as a valid source of information because it sometimes contains erroneous or incorrect information, go learn how to evaluate sources and then come back.

As I mentioned a couple posts ago, the last time a third party candidate won a state was in 1968 when George Wallace took 46 electoral votes. Twenty years before that, Strom Thurmond got 39 electoral votes. In 1912, Robert M. LaFollette of the Progressive party took Wisconsin's 13 electoral votes. The fourth and final time a third party candidate has won electoral votes in the 20th Century was 1912, when Teddy Roosevelt took 88 votes. Unless I'm missing a small state on one of the maps I'm looking at, third parties have only taken states 10 times throughout American history. Given the relative obscurity of both Jill Stein and Gary Johnson, I think it's a safe to say that won't happen this year. Let's stick to more recent spoiler candidates who may have taken just enough votes to flip a state and see if the election would have turned out any differently without them.

Our first spoiler is John Anderson in 1980. If we assume that Anderson stole votes exclusively from Jimmy Carter, his candidacy cost Carter 14 states: Arkansas, Delaware, Kentucky, Maine, Massachusetts, Michigan, Mississippi, New York, North Carolina, Pennsylvania, South Carolina, Tennessee, Vermont, and Wisconsin, These states had a total of 177 electoral votes, so if we take them from Reagan's 489 votes and add them to Carter's 49, Reagan still wins 312 to 226. So we can't blame Anderson for the Reagan Revolution.

The next spoiler is Ross Perot in 1992. If we add his votes to George H.W. Bush's in all the states that Bill Clinton won, Bill only wins 2 states (his home state of Arkansas and New York) and the District of Columbia. If you believe the conventional wisdom about spoilers, Clinton would have done even worse than Jimmy Carter in 1980 (with 42 votes to Carter's 49) if it hadn't been for Ross Perot.

If the idea of Bush I beating Bill Clinton in a Reagan-style landslide sounds a little fishy, you're probably right, and the reason is that most "lesser evil" arguments rest on the false assumption that a third party spoiler's votes come entirely from a single major party. In reality, that's rarely the case. It's likely that many or even most Perot voters were disaffected Republicans, but there were also disaffected Democrats, third party voters, and people who don't normally vote casting ballots for Ross Perot in 1992. Without Perot in the picture, Clinton probably wouldn't have taken so many southern states that usually go to the Republicans, but he likely would have won more than 2 states and D.C. because not all of the people who voted for Perot would have voted for Bush. Some would have voted for Clinton, some would have voted for another third party candidate, and some would have stayed home. Given his positions, I assume that Perot took more votes from Bush than Clinton and suspect he tilted the race to Clinton, but there's no way to know for sure without some indication of what Perot voters would have done if there was no Perot.

Perot ran less successfully in 1996, but let's skip that and move on to He Who Must Not Be Named himself, Ralph Nader in 2000. Before we can even get around to claiming that Nader played the spoiler, there are some unpleasant facts about Al Gore that we have to be willing to ignore:
  • Al Gore lost 7 states that Bill Clinton won in both elections and Jimmy Carter won in 1980, all by more votes than Ralph Nader got in those states: Arkansas, Kentucky, Louisiana, Missouri, Ohio, Tennessee, and West Virginia.  
  • Al Gore also lost 2 states that Bill Clinton won in both elections: Nevada and New Hampshire. He lost Nevada by more votes than Nader got. We'll get to New Hampshire in a minute. 
If Gore would have taken any of these states that the last 3 Democratic winners before him won, he would have won the election. Even if we remove the three swing states (Nevada, New Hampshire, and Ohio) that's still 6 states that Clinton and Carter took but Gore couldn't including Gore's home state of Tennessee. Even Walter Fucking Mondale won his home state (if nothing else). Since I've got my heart set on making Nader a spoiler and those 6 remaining states are in the South, I'm going to give Ross Perot credit for Bill Clinton's wins there. That doesn't explain why Carter won them in 1976, though, so let's write that of to GOP voters still being too disillusioned by Watergate to come to the polls. Nixon's always handy for stuff like that. Now let's look at the two states that Bush won where Nader's numbers were higher than the margin of victory: Florida and New Hampshire.

At least according to this study, Nader was a spoiler in Florida, but there's a caveat: "While Nader voters in 2000 were somewhat pro-Democrat and Buchanan voters correspondingly pro-Republican, both types of voters were surprisingly close to being partisan centrists. Indeed, we show that at least 40% of Nader voters in the key state of Florida would have voted for Bush, as opposed to Gore, had they turned out in a Nader-less election. The other 60% did indeed spoil the 2000 presidential election for Gore but only because of highly idiosyncratic circumstances, namely, Florida’s extreme
closeness." In addition to ignoring the "highly idiosyncratic circumstances," you also have to ignore the possibility that more Florida Democrats voted for Bush than Nader, whether intentionally or due to confusing ballots. Since I can't seem to find any sourced data on those numbers, we'll give Florida Democrats who voted for Bush a pass. Since I'm holding the Democrats blameless, I'm going to show the same courtesy to the GOP and discount claims of purged voter registrations an other vote tampering in favor of Bush. If you ignore everything but Nader, it's obvious that Nader was the reason Bush won Florida. 

Even with a Bush win in Florida (damn you Ralph Nader!), Gore still could have won the electoral vote if he'd taken New Hampshire. If we go with the conventional wisdom that would send Bill Clinton back to Arkansas with under 50 electoral votes in 1992, then Nader was clearly the spoiler there, too. If, however, we accept that not all Nader voters would have voted for Gore otherwise, it's not quite as clear. If we use the 60/40 split from Florida, Bush still wins New Hampshire in a race without Nader. Even if you assume that 70% of Nader's voters would have voted for Gore, Al only wins by about 1600 votes, and that's if you keep the right-leaning Libertarian and Reform parties in the race to spoil it for Bush. Of course, if you're going to start assuming that the breakdown of Nader voter loyalties varies that much from state to state, you probably also need to consider the possibility that Gore would have lost one or more of the six states (Iowa, Maine, Minnesota, New Mexico, Oregon, and Wisconsin) where he won by fewer votes than Ralph Nader got. 

Spoilers definitely exist, but it's not quite as simple as assuming that any vote for a third party will spoil the race for the lesser evil. For purposes of deciding whether a third party vote is an acceptable risk in a given state, I'm going to go through a few different scenarios for each state:

Worst Case Scenario (Third Party takes 25% of vote from a single party)
For all the Nader hate out there, Ross Perot was actually the most effective spoiler since at least 1980. In 1992, Perot took 18.91% of the popular vote (Nader only took 2.74% in 2000). His best state percentage-wise was Maine, where he took 30.44% of the vote. In 28 states, he took over 20% of the vote, with 20 of those states in the 20-25% range, 7 in the 25-30% range, and Maine at just over 30. I'm going to use the middle of that range as my worst case 3rd party bogeyman.

Average (Third Party Takes 10% of vote from a single party)
If you average the percentage of the popular vote for Anderson, both Perot runs, and Nader, you get a 9% and some change, so I'm going to round it up to 10. Given that all of those candidates had a lot more going for them than the 3rd party candidates running in 2016, this is probably the more realistic worse case.

Polls
Most polls I've seen put Gary Johnson at 5-10% and Jill Stein at 5% or less, so that's probably the highest percentage each candidate will actually get in most states barring a major game changer between now and the election. 

Now we're ready to look at the states (I hope).

Here's this thing. 


No comments: