How to Spin a Plausible, Silly Political Theory -- and How That Distorts Commentary on Trump
Note to readers: My new Verdict column today takes a break from pessimism and looks (with a small amount of success) for reasons to be optimistic about the U.S. constitutional system. My column here does not build on the Verdict piece, but I encourage you all to read -- and possibly enjoy -- both.
by Neil H. Buchanan
Did you know that the Change Candidate always (at least since 1960) beats the No-Change Candidate in U.S. presidential elections? Other than when formerly Change Candidates run for reelection, this is true -- every time. And even when a president is running for a second term, the basic logic still works -- every time.
Although what I wrote above is mostly true (which I will demonstrate momentarily), the whole exercise is absolutely pointless and nonsensical. Let us put aside for today thoughts of the doom of our constitutional system and work through a demonstration of armchair analysis that would be on the high end of U.S. political punditry, if only I could say it with a straight face.
Moreover, this inanity is not innocuous. Even pundits who do not support Donald Trump end up using this framing in a way that inaccurately and harmfully builds him up into something that he is not.
To be clear, this "change candidate wins" idea is hardly my own. I do not recall where I first came across it, but it pops up in various guises every now and then. Moreover, in the current context it is a theory that I ought to find pleasing, given that I have endorsed Change Candidate Elizabeth Warren this year and defend somewhat-bigger-Change Candidate Bernie Sanders against the ridiculous attacks coming from self-styled centrists. Meanwhile, I have had little good to say about No-Change Candidates Joe Biden, Pete Buttigieg, Amy Klobuchar, or (especially) Michael Bloomberg.
Even so, the more I have thought about the initial appeal of this argument, the more trivial and less believable it has become. (And this is true even without challenging what the term "change candidate" might mean.) Still, let us run backward in time to make the superficial empirical case, first looking only at years in which neither candidate was running for reelection:
-- 2016 - Trump wins, Clinton loses
Duh. This is the leading (and for some, the only even moderately developed) example in most of the "Change Candidate wins” think pieces. (I suppose those ought to be called "think-but-don't-think-very-hard pieces.") Trump is the ultimate disrupter, Clinton the ultimate status quo candidate (representing not just "Obama's third term" but carrying her husband's specter with her as well). She lost, he won. Q.E.D.
-- 2008 - Obama wins, McCain loses
Again, an easy case. Obama shocks everyone by becoming an electable black man in America, beating the former mavericky guy McCain, who at that point was again the loyal insider who had waited his turn and was rewarded for his time spent patiently in the pre-Trump Republican Party's well behaved boys club.
-- 2000 - Bush wins, Gore loses
Simple narrative mirroring the conventional wisdom on 2016. Bush, despite being a president's son, is supposedly new and different. Gore is more of the same after eight years of Bill Clinton.
-- 1988 - Other Bush wins, Dukakis loses
This is the most challenging one to fit into the narrative, and it is arguably a wash if one sees this as an election with no Change Candidate on either side. Dukakis was the safe insider, and Bush was the incumbent Vice President. Still, Bush claimed to be a new sort of kinder, gentler Republican, and he successfully painted his opponent as the candidate of the "Harvard boutique" while Yalie-to-his-DNA Bush ate pork rinds.
-- 1976 - Carter wins, Ford loses
I include this here because Ford was never elected, but in any case, Carter was the fresh-faced outsider, while Ford was the long-serving House insider who had pardoned Nixon.
-- 1968 - Nixon wins, Humphrey loses
Again, Nixon is not really an outsider, but he is the Change Candidate at least in contrast to Humphrey, who bore the brunt of anti-Vietnam anger aimed at his boss.
-- 1964 - Johnson wins, Goldwater loses
Oops. I wrote above that the theory works "every time," but this is nearly impossible to fit into the narrative. So a good pundit would chalk it up as "the exception that proves the rule." (Or maybe "LBJ was really running for reelection, even though he did not win in 1960." But see below re reelection examples.)
-- 1960 - Kennedy wins, Nixon loses
Back to easy cases. Brash young outsider beats continuation-of-the-previous-eight-years insider. Change wins, as it always will.
Now, let us look at the years in which an incumbent is running for reelection. In almost all of them, the challenger loses, a result that seems to be in direct opposition to the Change Candidate Wins story. One could, I suppose, merely say that the powers of incumbency are categorically different and thus make the Change theory inoperative in those years, but it is not even necessary to rely on that excuse. Why? Because in every case that the incumbent won reelection, the other party nominated a safe insider non-Change Candidate.
-- 2012 - Obama wins, Romney loses
Whatever one thinks about Romney today, in the wake of his lone stand to impeach Trump (on one count, while offering a ridiculous argument to acquit him on the other), he was the quintessence of non-Change. Obama's hopey-changey thing might have been four years old by then, but Romney managed to make Obama look fresh and new.
-- 2004 - Bush wins, Kerry loses
Kerry was the Democrats' Romney. Safe, boring, patrician and a complete insider. He would have made anyone look more disruptive by comparison.
-- 1996 - Clinton wins, Dole loses
Oh wait. What I just said about Kerry and Romney? Double that for Dole.
-- 1992 - Clinton wins, other Bush loses
This is a big data point (aka anecdote), because it is one of the two times that an elected incumbent loses. So you not only have the clear Change Candidate winning, but you have the super-duper non-Change candidate losing.
-- 1984 - Reagan wins, Mondale loses
OK, this is getting embarrassing. What I said about Romney, Kerry, and Dole? Triple it.
-- 1980 - Reagan wins, Carter loses
The other big one where a challenger beats the incumbent. Reagan is not merely a Change candidate, he is the Sanders of his day, freaking out his party's insiders so much that they actually tried to draft Ford at the convention to head off sure defeat.
-- 1972 - Nixon wins, McGovern loses
Another exception to the rule.
For those readers who are still with me (thank you!), it is already possible to see how this game is played. First, universal "rules" are not universal. Not really a surprise there. More careful pundits will hedge by sprinkling "almost," "virtually every," or "perhaps" throughout such analyses. But second, even with the exceptions, it still looks like a pretty impressive list. In 15 elections, the theory seems to hold up nicely in 13 of them.
Except that it doesn't. The obvious starting point is 2016, where the whole Change beats non-Change idea seems so powerful. In an otherwise interesting op-ed telling Democratic elites to calm down about Sanders, former Treasury Secretary under Clinton Robert Reich makes a familiar claim about Hillary Clinton and Jeb Bush in 2016: "Clinton and Bush started with all the advantages, but neither could credibly convince voters they were not part of the system."
So they both lost because they were not the Change candidate, right? Wrong. Bush's humiliation in the primaries was due to his being an incredibly weak candidate. If Trump had not come along, Bush would not have been called "low energy," but his lower energy levels would still have dragged him down. President Christie, anyone? Paging President Kasich!
And what of Clinton? Did she, per Reich, fail to "credibly convince voters that [she] was not part of the system?" I have no recollection of her trying, so I guess the answer is no. But that is not why she lost. Even putting aside the whole popular vote versus Electoral College thing, data guru Nate Silver's post-election analysis remains essential reading: "The Comey Letter Probably Cost Clinton The Election So why won’t the media admit as much?"
All of the Monday morning quarterbacking in the world -- "She should have campaigned in Wisconsin!" "She focused on attacking Trump too much!" -- does not change the fact that Trump's non-majority win was hugely assisted by Russian intervention (including the oft-forgotten but intense efforts to depress turnout by African Americans) and media pile-ons ("What about her emails?”), as well as the simple fact that Comey's letter (per Silver's analysis) was easily enough to explain the cracking of the Blue Wall in Pennsylvania, Michigan, and Wisconsin. Trump would have -- and should have -- lost badly, even though he was the Change Candidate.
The 2000 election was another non-majority win, with more than a soupcon of Florida recount madness. Even W's win over boring insider Kerry in 2004 was incredibly close, turning on the outcome in Ohio, where the Republican Secretary of State had engaged in massive voter suppression. (And it did not hurt that then-Republican Michael Bloomberg invited the party to use NYC -- and the 9/11 site -- as a campaign staging area, aka the Republican National Convention.)
2008? The Great Recession hit just as the election approached. 1992? Clinton almost certainly would have lost badly had Ross Perot not disrupted things (and Perot lost, by the way). In fact, Clinton was only the nominee because the party's heavy hitters sat out the election when the first Gulf War sent Bush's approval ratings skyward. 1980? Carter was frankly a sh_tty candidate, plus the economy was in terrible shape (stagflation), plus .. what was it? oh right .. the Iran Hostage Crisis.
Add in the fact that these "difficult cases" along with other non-exceptions (1976, 1968, 1960) were incredibly close, and what is left of the theory? Well, this is where pundits would earn their keep. It is always possible to say that the candidate who won the close election "put himself in a position to win" by being the Change Candidate. Except that this means that the other candidate put her/himself in a position to win by being the non-Change Candidate.
The reason that this theory bothers me so much, I think, is not merely that it reduces multivariate analyses to a single, simplistic explanation. That is bad, but it happens. It is worse because it treats Trump's rise as either inevitable -- after all, if the Change Candidate always wins, no one was more of a change than Trump, even Bernie Sanders, so Trump had to win -- or somehow sensible.
Because Trump won, even his detractors start to talk about him as if he represents a majority of voters who "wanted change." He is then turned into a supposedly fearsome opponent, with pundit after pundit saying that the Democrats need to be careful to find a candidate who can go "toe-to-toe with Donald Trump on a debate stage." (Seriously, how many times have near-identical variations on that phrase been typed by lazy media hacks?)
As I wrote last summer:
But the fact is that he lucked his way into the presidency, and anything close to a fair election in 2020 would see the man with comically low approval ratings lose decisively.
There probably will not be anything close to a fair election this year, but Trump is still likely to lose to any Democrat, especially if recent trends continue. But no matter the background factors, any analysis that builds him up as the all-but-unstoppable voice that rose up to represent Change not only relies on faulty empirics but creates a pro-Trump narrative that is both vacuous and harmful.
by Neil H. Buchanan
Did you know that the Change Candidate always (at least since 1960) beats the No-Change Candidate in U.S. presidential elections? Other than when formerly Change Candidates run for reelection, this is true -- every time. And even when a president is running for a second term, the basic logic still works -- every time.
Although what I wrote above is mostly true (which I will demonstrate momentarily), the whole exercise is absolutely pointless and nonsensical. Let us put aside for today thoughts of the doom of our constitutional system and work through a demonstration of armchair analysis that would be on the high end of U.S. political punditry, if only I could say it with a straight face.
Moreover, this inanity is not innocuous. Even pundits who do not support Donald Trump end up using this framing in a way that inaccurately and harmfully builds him up into something that he is not.
To be clear, this "change candidate wins" idea is hardly my own. I do not recall where I first came across it, but it pops up in various guises every now and then. Moreover, in the current context it is a theory that I ought to find pleasing, given that I have endorsed Change Candidate Elizabeth Warren this year and defend somewhat-bigger-Change Candidate Bernie Sanders against the ridiculous attacks coming from self-styled centrists. Meanwhile, I have had little good to say about No-Change Candidates Joe Biden, Pete Buttigieg, Amy Klobuchar, or (especially) Michael Bloomberg.
Even so, the more I have thought about the initial appeal of this argument, the more trivial and less believable it has become. (And this is true even without challenging what the term "change candidate" might mean.) Still, let us run backward in time to make the superficial empirical case, first looking only at years in which neither candidate was running for reelection:
-- 2016 - Trump wins, Clinton loses
Duh. This is the leading (and for some, the only even moderately developed) example in most of the "Change Candidate wins” think pieces. (I suppose those ought to be called "think-but-don't-think-very-hard pieces.") Trump is the ultimate disrupter, Clinton the ultimate status quo candidate (representing not just "Obama's third term" but carrying her husband's specter with her as well). She lost, he won. Q.E.D.
-- 2008 - Obama wins, McCain loses
Again, an easy case. Obama shocks everyone by becoming an electable black man in America, beating the former mavericky guy McCain, who at that point was again the loyal insider who had waited his turn and was rewarded for his time spent patiently in the pre-Trump Republican Party's well behaved boys club.
-- 2000 - Bush wins, Gore loses
Simple narrative mirroring the conventional wisdom on 2016. Bush, despite being a president's son, is supposedly new and different. Gore is more of the same after eight years of Bill Clinton.
-- 1988 - Other Bush wins, Dukakis loses
This is the most challenging one to fit into the narrative, and it is arguably a wash if one sees this as an election with no Change Candidate on either side. Dukakis was the safe insider, and Bush was the incumbent Vice President. Still, Bush claimed to be a new sort of kinder, gentler Republican, and he successfully painted his opponent as the candidate of the "Harvard boutique" while Yalie-to-his-DNA Bush ate pork rinds.
-- 1976 - Carter wins, Ford loses
I include this here because Ford was never elected, but in any case, Carter was the fresh-faced outsider, while Ford was the long-serving House insider who had pardoned Nixon.
-- 1968 - Nixon wins, Humphrey loses
Again, Nixon is not really an outsider, but he is the Change Candidate at least in contrast to Humphrey, who bore the brunt of anti-Vietnam anger aimed at his boss.
-- 1964 - Johnson wins, Goldwater loses
Oops. I wrote above that the theory works "every time," but this is nearly impossible to fit into the narrative. So a good pundit would chalk it up as "the exception that proves the rule." (Or maybe "LBJ was really running for reelection, even though he did not win in 1960." But see below re reelection examples.)
-- 1960 - Kennedy wins, Nixon loses
Back to easy cases. Brash young outsider beats continuation-of-the-previous-eight-years insider. Change wins, as it always will.
Now, let us look at the years in which an incumbent is running for reelection. In almost all of them, the challenger loses, a result that seems to be in direct opposition to the Change Candidate Wins story. One could, I suppose, merely say that the powers of incumbency are categorically different and thus make the Change theory inoperative in those years, but it is not even necessary to rely on that excuse. Why? Because in every case that the incumbent won reelection, the other party nominated a safe insider non-Change Candidate.
-- 2012 - Obama wins, Romney loses
Whatever one thinks about Romney today, in the wake of his lone stand to impeach Trump (on one count, while offering a ridiculous argument to acquit him on the other), he was the quintessence of non-Change. Obama's hopey-changey thing might have been four years old by then, but Romney managed to make Obama look fresh and new.
-- 2004 - Bush wins, Kerry loses
Kerry was the Democrats' Romney. Safe, boring, patrician and a complete insider. He would have made anyone look more disruptive by comparison.
-- 1996 - Clinton wins, Dole loses
Oh wait. What I just said about Kerry and Romney? Double that for Dole.
-- 1992 - Clinton wins, other Bush loses
This is a big data point (aka anecdote), because it is one of the two times that an elected incumbent loses. So you not only have the clear Change Candidate winning, but you have the super-duper non-Change candidate losing.
-- 1984 - Reagan wins, Mondale loses
OK, this is getting embarrassing. What I said about Romney, Kerry, and Dole? Triple it.
-- 1980 - Reagan wins, Carter loses
The other big one where a challenger beats the incumbent. Reagan is not merely a Change candidate, he is the Sanders of his day, freaking out his party's insiders so much that they actually tried to draft Ford at the convention to head off sure defeat.
-- 1972 - Nixon wins, McGovern loses
Another exception to the rule.
For those readers who are still with me (thank you!), it is already possible to see how this game is played. First, universal "rules" are not universal. Not really a surprise there. More careful pundits will hedge by sprinkling "almost," "virtually every," or "perhaps" throughout such analyses. But second, even with the exceptions, it still looks like a pretty impressive list. In 15 elections, the theory seems to hold up nicely in 13 of them.
Except that it doesn't. The obvious starting point is 2016, where the whole Change beats non-Change idea seems so powerful. In an otherwise interesting op-ed telling Democratic elites to calm down about Sanders, former Treasury Secretary under Clinton Robert Reich makes a familiar claim about Hillary Clinton and Jeb Bush in 2016: "Clinton and Bush started with all the advantages, but neither could credibly convince voters they were not part of the system."
So they both lost because they were not the Change candidate, right? Wrong. Bush's humiliation in the primaries was due to his being an incredibly weak candidate. If Trump had not come along, Bush would not have been called "low energy," but his lower energy levels would still have dragged him down. President Christie, anyone? Paging President Kasich!
And what of Clinton? Did she, per Reich, fail to "credibly convince voters that [she] was not part of the system?" I have no recollection of her trying, so I guess the answer is no. But that is not why she lost. Even putting aside the whole popular vote versus Electoral College thing, data guru Nate Silver's post-election analysis remains essential reading: "The Comey Letter Probably Cost Clinton The Election So why won’t the media admit as much?"
All of the Monday morning quarterbacking in the world -- "She should have campaigned in Wisconsin!" "She focused on attacking Trump too much!" -- does not change the fact that Trump's non-majority win was hugely assisted by Russian intervention (including the oft-forgotten but intense efforts to depress turnout by African Americans) and media pile-ons ("What about her emails?”), as well as the simple fact that Comey's letter (per Silver's analysis) was easily enough to explain the cracking of the Blue Wall in Pennsylvania, Michigan, and Wisconsin. Trump would have -- and should have -- lost badly, even though he was the Change Candidate.
The 2000 election was another non-majority win, with more than a soupcon of Florida recount madness. Even W's win over boring insider Kerry in 2004 was incredibly close, turning on the outcome in Ohio, where the Republican Secretary of State had engaged in massive voter suppression. (And it did not hurt that then-Republican Michael Bloomberg invited the party to use NYC -- and the 9/11 site -- as a campaign staging area, aka the Republican National Convention.)
2008? The Great Recession hit just as the election approached. 1992? Clinton almost certainly would have lost badly had Ross Perot not disrupted things (and Perot lost, by the way). In fact, Clinton was only the nominee because the party's heavy hitters sat out the election when the first Gulf War sent Bush's approval ratings skyward. 1980? Carter was frankly a sh_tty candidate, plus the economy was in terrible shape (stagflation), plus .. what was it? oh right .. the Iran Hostage Crisis.
Add in the fact that these "difficult cases" along with other non-exceptions (1976, 1968, 1960) were incredibly close, and what is left of the theory? Well, this is where pundits would earn their keep. It is always possible to say that the candidate who won the close election "put himself in a position to win" by being the Change Candidate. Except that this means that the other candidate put her/himself in a position to win by being the non-Change Candidate.
The reason that this theory bothers me so much, I think, is not merely that it reduces multivariate analyses to a single, simplistic explanation. That is bad, but it happens. It is worse because it treats Trump's rise as either inevitable -- after all, if the Change Candidate always wins, no one was more of a change than Trump, even Bernie Sanders, so Trump had to win -- or somehow sensible.
Because Trump won, even his detractors start to talk about him as if he represents a majority of voters who "wanted change." He is then turned into a supposedly fearsome opponent, with pundit after pundit saying that the Democrats need to be careful to find a candidate who can go "toe-to-toe with Donald Trump on a debate stage." (Seriously, how many times have near-identical variations on that phrase been typed by lazy media hacks?)
As I wrote last summer:
"Honestly, if anyone reading this column were advising a candidate who was running against Trump next year, where would 'we might not stand up to Trump's awesome debating skills' land on your list of concerns? There are all kinds of ways in which things might go badly next year -- most prominently a return of the Russians' successful intervention in the U.S. electoral system, possibly this time including successful hacking of voting machines -- but losing a debate to Trump is about as likely as my being Trump's running mate."The point is that it is not merely casual empiricism that is the problem here. What looks like a universal or near-universal empirical regularity is not merely fatuous, but this particular theory elevates Trump into some kind of awesome figure who was (and maybe still is, because he continues to be all about Change) nearly unbeatable.
But the fact is that he lucked his way into the presidency, and anything close to a fair election in 2020 would see the man with comically low approval ratings lose decisively.
There probably will not be anything close to a fair election this year, but Trump is still likely to lose to any Democrat, especially if recent trends continue. But no matter the background factors, any analysis that builds him up as the all-but-unstoppable voice that rose up to represent Change not only relies on faulty empirics but creates a pro-Trump narrative that is both vacuous and harmful.