Easterbook’s premise is that a “third party spoiler” running a “vanity” campaign in 2012 could cost Barack Obama the election, siphoning off Republican-friendly voters who are turned off by their party’s nominee but reluctant to back the president. Specifically, he points to Michael Bloomberg, and advises Obama to “appoint him to something important ASAP” — before the New York billionaire throws the election to Mitt or Tim or … whoever.
This is completely and thoroughly untrue — and it’s not an even remotely ambiguous matter. Let’s take them one at a time.
Jimmy Carter lost in 1980 because of John Anderson? Well, that was the Carter campaign’s fear throughout the summer and fall of 1980 (and the Reagan campaign’s hope). It’s why Carter refused to participate in any debates in which Anderson — a liberal House Republican who had bolted the GOP after briefly seeking its nomination in order to wage a third party campaign — participated. But it’s not at all how things played out.
When Anderson began his independent bid in the spring of ’80, he reached as high as 24 percent in national surveys. But he had trouble raising money and steadily lost ground in the ensuing months. By September, his support was in the mid-teens, and it declined further after he and Reagan engaged in a nationally televised one-on-one debate. It was around that time that pollsters began to realize that the assumptions about Anderson’s effect were wrong. Sure, it seemed logical to assume that his message would resonate primarily with Carter voters. But at a practical level, he was simply another option for frustrated voters who had already decided not to back Carter for another term. Polls found Anderson voters nearly as likely to list Reagan as their second choice as Carter. The election was fundamentally a referendum on Carter, and with unemployment, inflation, interest rates and gas prices soaring — and with national confidence sagging, thanks to a protracted hostage crisis — the majority of Americans were set to toss the president out. The only question was which of his opponents they’d support. This explains why Anderson’s support dropped so steadily as the election neared: Anti-Carter voters came to fear wasting their votes on Anderson (and maybe even helping Carter score a backdoor victory).
Thus, Anderson finished with just 5.7 million votes on Election Day — or 6.6 percent of the vote. Reagan’s margin over Carter? 8.4 million votes, or 9.7 percent. In other words, even if you awarded every single vote for Anderson to Carter, the president still would have finished with nearly three million fewer popular votes than Reagan. Moreover, we know that many Anderson voters preferred Reagan — not Carter — as their second choice; so without Anderson in the race, Reagan still would have won easily. The notion that Easterbrook is now advancing — that Anderson denied Carter a second term — was actually debunked on the spot 31 years ago. Here’s how Newsweek reported it:
John Anderson’s impact on the race was largely overshadowed by the broad-based Reagan landslide. It was in one sense tempting to view him as a spoiler; Anderson’s vote was actually greater than Reagan’s margin of victory in thirteen states, among them New York, Wisconsin, North-Carolina and Connecticut. But had Anderson not run, Carter would have picked up barely half (49 per cent) of his vote; 37 per cent of Anderson voters said they would have backed Reagan.
Next up: Ross Perot and 1992. Easterbook argues that George H.W. Bush lost because the Texas billionaire “siphoned off conservative votes.” False.
The endurance of this particular myth, regular readers will recognize, is particularly bothersome to me, and I’ve written about it several times. Easterbrook is hardly the only one who’s still pushing it. I can’t tell you how many times I have heard someone matter-of-factly make this claim in the past few years. Generally, it’s from conservatives who like to pretend that Bill Clinton’s ’92 victory was a fluke, but there are liberals who still fall for it too.
Instead of recounting all of the details of the ’92 race here, I’ll simply refer you to one of my previous posts on the subject. If you want the Cliff’s Notes version it goes like this: (1) Economic anxiety was high, causing Bush’s poll numbers to drop to poisonous levels — by the fall of ’92 he was not an incumbent who, on paper, should have won reelection; (2) Not a single public opinion poll from the middle of July (when Perot dropped out the race) through the end of September (when Perot returned) gave Bush a lead over Clinton — not even in the immediate wake of the August ’92 GOP convention. In fact, Clinton’s average lead in this period was double-digits — and the race was not tightening at the time Perot jumped back in; (3) A comprehensive national exit poll found that Perot voters were divided almost evenly on their second choice and that Clinton — in a two-way race — would still have beaten Bush by 5.8 million votes (his actual margin was 5.3 million in initial ’92 tally). Here’s how the Washington Post summarized the exit poll:
Ross Perot’s presence on the 1992 presidential ballot did not change the outcome of the election, according to an analysis of the second choices of Perot supporters.
The analysis, based on exit polls conducted by Voter Research & Surveys (VRS) for the major news organizations, indicated that in Perot’s absence, only Ohio would have have shifted from the Clinton column to the Bush column. This would still have left Clinton with a healthy 349-to-189 majority in the electoral college.
And even in Ohio, the hypothetical Bush “margin” without Perot in the race was so small that given the normal margin of error in polls, the state still might have stuck with Clinton absent the Texas billionaire.
In most states, the second choices of Perot voters only reinforced the actual outcome. For example, California, New York, Illinois and Oregon went to Clinton by large margins, and Perot voters in those states strongly preferred Clinton to Bush.
Repeat after me: Ross Perot did not “cost” George H.W. Bush the 1992 election. If you see or hear a commentator using this claim as supporting evidence, immediately discount whatever argument that commentator is advancing. The poor economy doomed George H.W. Bush in 1992 — not a short billionaire from Texas.
And then there’s 2000: Gore, Bush, Florida, Nader — you remember it all. Easterbrook has a point on this one, I suppose (even if I would argue that more Nader voters would probably have backed Bush as their second choice than we generally realize). But really, the ’00 example serves to illustrate the very narrow and specific set of circumstances required for a third party candidate to swing the election. The race came down to Florida, where Bush’s official margin of victory was 537 votes — and where Nader received nearly 100,000.
The bottom line is that a serious third party candidacy next year — one on the level of Perot ’92 — will simply serve as an indication of much deeper trouble for Obama. Perot took off in ’92 because swing voters will deeply uneasy over the economy and, thus, dismayed with Bush’s perceived job performance. With or without Perot in ’92, Bush would have lost — just as Carter (running under similar condition) would have lost with or without Anderson in 1980. The same goes for Obama next year: A double-dip recession will probably doom him, whether he has one general election opponent or two. Conversely, if the economy continues to improve, there simply won’t be room for a major third party candidate; and if a self-funding third party candidate like Bloomberg were to force the issue, it wouldn’t be a threat to Obama, since swing voters will happily reelect a president they like if they perceive the economy to be on the rebound. This is why there was no room for an independent candidate when Reagan ran for reelection in 1984, and why Perot was unable to gain traction when he ran again in 1996.
Update: In the comments section, “The Voice of Reason” provides a useful demonstration of the logic that people can employ to keep these myths alive:
You suggest that the “Anderson spoiler effect” was debunked, yet the quote from the Newsweek is revealing: “Anderson’s vote was actually greater than Reagan’s margin of victory in thirteen states.” I’ll remind the author that we do not elect a president on popular vote, but on the electoral college so using popular vote numbers as statistics in this case is misleading. The spoiler effect is state to state and can come down to tipping only one state’s electoral college vote, i.e. Florida in 2000.
For the record, the 13 states in which Anderson’s total was greater than Reagan’s margin were: Maine, Vermont, Massachusetts, Connecticut, New York, Delaware, North Carolina, Michigan, Wisconsin, Arkansas, Kentucky, Maryland and Mississippi. The total number of electoral votes in the these states (as of 1980): 150. Reagan’s electoral vote margin over Carter: 489 to 49. So even if you took every vote that John Anderson received in every one of these 13 states and added it to Carter’s tally, Reagan still would have won a decisive electoral college victory: 339 to 199.
But that’s beside the point, because — as noted above — we have known for 31 years that not that every Anderson voter would have gone to Carter as his/her second choice; in fact, Anderson voters were only slightly more likely to favor Carter as their second choice than Reagan. In other words, Anderson’s absence wouldn’t actually have flipped many — or maybe any — of these 13 states to Carter.
Update II (starting to sound like Greenwald here…): Combing through the rest of the comments, I notice several people make the same basic argument for the ’92 mythology: Forget the exit polls — Perot’s real impact was in undermining Bush and softening him up for Clinton. For instance, “Amity” writes:
The 1992 election was not simply a question of who was whose second choice on election day. Early in the campaign, Perot had managed to wedge himself into what was at that point essentially a three-way dead heat. Steve Kornacki knows perfectly well that Perot did not spend the entire campaign as some interesting oddity — he set the agenda early and masterfully in his public appearances and left the other candidates flat footed.
Kornacki would have us believe that the 1992 election was a foregone conclusion and Perot literally may as well not have existed. But the history of the late 20th century has shown that there is no Republican economic crisis so great that Democrats can’t find some way to flub it. There was nothing about the situation in 1992 that made it automatically impossible for Bush to take the situation at hand, co-opt Perot’s populist economic agenda, and wipe the floor with it.
Yet it was Clinton who did that instead, and Bush dithered. You could argue that Bush dithering had been his problem all along, long before Perot was in the picture, but that’s just begging the question. Clinton’s popularity through 1992 waxed in exact proportion to the extent to which Perot’s waned.
Of course by November Clinton had become most Perot supporters’ second choice. That doesn’t mean that Perot’s campaign started out (implausibly) as a Clinton fifth column — rather it serves as an admirable, though by no means singular, demonstration of the reason why Clinton’s campaign is 20 years later still studied by everyone serious about electoral politics in America.
In particular, Karl Rove studied Carville’s strategy and took the lessons learned to heart — which you might think Gore could have done too.
The basic problem with assessments like this is that they place far too much weight on campaign messaging, tactics and events — and not nearly enough on the basic structural factors that dictate the outcomes of most presidential elections.
When it comes to analyzing ’92, by far the most important thing to remember is that the race featured an incumbent president (whose party had controlled the White House for an unusually long 12 years), rising unemployment, and sagging public confidence. After spiking to an artificially stratospheric 91 percent after the winter ’91 Gulf War, Bush’s popularity steadily eroding as the bad economic news mounted. By Labor Day, his approval rating had fallen under 60 percent. By Christmas, it was under 50 percent. By March of 1992, it was nearing 40 percent — a dangerously low level that generally signals looming defeat for an incumbent. In fact, by early March, polls found Bush losing to both of the Democrats’ top two prospects — Bill Clinton and Paul Tsongas.
All of this happened without Perot and his big ears and charts in the picture. It wasn’t until February 20 — two days after the New Hampshire primary, in which Bush was humiliated by Pat Buchanan’s surprisingly string showing — that Perot, then largely unknown to most Americans, announced on “Larry King Live” that he’d be willing to run if volunteers put him on the ballot on all 50 states. The Perot movement quickly gained steam and climbed to impressive heights in national polling. By late May, he was leading the three-way horse race, slightly ahead of Bush and far ahead of Clinton.
It’s not a mystery what was going on: Voters had already given up on Bush. The economy explains why and the steady, months-long decline in his popularity is the proof. But as the Democratic nominating process wrapped up and Clinton emerged, voters were eager for another non-Bush option. Clinton had been absolutely battered by scandal — Gennifer Flowers, the Vietnam draft, and so on. It is hard to appreciate now, but the conventional wisdom in Democratic circles at the time was that Clinton, in the words of then-Sen. Bob Kerrey, would be cracked open “like a soft peanut” by the GOP attack machine in the fall. Clinton’s personal popularity was horribly low; voters found him profoundly slippery and untrustworthy. But Democrats didn’t have much of a choice: Bush’s artificial post-Gulf War popularity had scared off all of their A-list prospects, and Tsongas — with his pro-Wall Street message and charisma-challenged style — was unacceptable to too many key elements of the party coalition.
Thus, there was a vacuum: Voters did not want to reelect Bush, but they also didn’t want to vote for Clinton. Perot surged — and Clinton struggled to break 20 percent. This was not a case of Perot’s message waking up voters to the unpleasant realities of Bush’s record and convincing them to turn on him. They already had done so (for absolutely predictable reasons). Perot was merely capitalizing on this predictable discontent. And it’s not as if his message was even original: Tsongas had been doing the anti-deficit outsider thing for a year before Perot arrived on the scene — and ended up garnering nearly 5 million votes in the Democratic primaries.
The Perot phenomenon soon crested. His behavior grew erratic as the media ratcheted up its scrutiny. Meanwhile, anti-Bush voters began growing more comfortable with Clinton as they were exposed to him more. With Clinton emerging as a viable option, there was less an less room for Perot. By July, Perot’s support had slipped back to the 20s, with Clinton’s rising. Perot then dropped out during the Democratic convention in mid-July, after which Clinton surged to leads of 20-25 points in national polls. Here’s the important part: He maintained wide leads for the next 2 1/2 months, during which it was a two-way race. Even Bush’s post-GOP convention bounce in mid-August wasn’t enough to put him ahead of Clinton, and polls throughout September weren’t even close. In a two-way race, Clinton was dominant — and there was no sign that Bush and his vaunted “attack machine” could do anything about it. Of course they couldn’t: Not with unemployment soaring, confidence sagging and Bush’s approval rating under 40 percent. The race was fundamentally about the incumbent.
When Perot returned to the race at the start of October, little changed. Sure, Perot performed ably in debates, rehabilitated his own image, which had been damaged by the bizarre circumstances of his July exit (he had blamed Republican “dirty tricksters” and suggested they had tried to sabotage his daughter’s wedding), and managed to snag nearly 20 percent of the popular vote. But his presence never altered the basic dynamic of the race: Voters wanted Bush out and Clinton cleared the (very low) hurdle of being an acceptable enough alternative.
The idea that Perot came along and mucked up a fundamentally winnable election for Bush just doesn’t mesh with reality. Without knowing anything about his opponents, any election forecasting model would have predicted a Bush defeat based on the economic conditions that prevailed in the run-up to the ’92 election. Perot was not the cause of Bush’s political misfortune that year — he was a symptom of it.