Marc A. Thiessen writes:
With three polls showing her in the lead, Sen. Elizabeth Warren (D-Mass.) may soon eclipse former vice president Joe Biden as the front-runner for the Democratic presidential nomination. That’s great news for Republicans, because Warren has a problem: The central message of her campaign is that the economy is working for the very wealthy but it is not working for ordinary Americans. Unfortunately for her, ordinary Americans disagree.
That is great news for Republicans. But that doesn’t mean it’s great news for people who support Republicans because they don’t like left-wing economic policies. In fact it’s bad news for them.
And the reason has to do with the median voter theorem. Here’s what William F. Shughart II writes in “Public Choice,” in David R. Henderson, ed., The Concise Encyclopedia of Economics:
Studying collective decision-making by committees, Duncan Black deduced what has since been called the median-voter theorem. If voters are fully informed, if their preferred outcomes can be arrayed along one dimension (e.g., left to right), if each voter has a single most-preferred outcome, and if decisions are made by simple majority rule, then the median voter will be decisive. Any proposal to the left or right of that point will be defeated by one that is closer to the median voter’s preferred outcome. Because extreme proposals lose to centrist proposals, candidates and parties in a two-party system will move to the center, and, as a result, their platforms and campaign promises will differ only slightly. Reversing 1964 presidential hopeful Barry Goldwater’s catchphrase, majority-rule elections will present voters with an echo, not a choice. If the foregoing assumptions hold, the median voter’s preferences also will determine the results of popular referenda. As a matter of fact, anticipating that immoderate proposals will be defeated, the designers of ballot initiatives will strive to adopt centrist language, in theory moving policy outcomes closer to the median voter’s ideal point than might be expected if decisions are instead made by politically self-interested representatives.
Those 4 assumptions are quite strong and unlikely to hold up in the real world. But that doesn’t mean the median voter theorem is useless. It still has power in explaining two things: why candidates’ platforms in a general election, when 2 parties contend, will often be similar. Think of Nixon and Humphrey in 1968: that’s the first presidential election I followed somewhat closely. I remember pundits referring to them as Tweedledum and Tweedledee. (Some wags at the time referred to them as Tweedledum and Tweedledumber.) There wasn’t a huge difference between the policy positions they articulated (with one important exception: Nixon promised to end the draft.)
Or think about what happened in the Goldwater/Johnson election. Vietnam was a big issue, as the war was just getting going. Goldwater took an extreme position in favor of the war. Did Johnson go all dove? No. He was pro-war also, but less extreme. That led to the old joke that might have actually been based on an actual voter’s comment: “I was told that if I voted for Goldwater, we’d bomb North Vietnam. So I voted for Goldwater and sure enough we bombed North Vietnam.” And LBJ won with about 60% of the vote.
If a politician wants mainly to win, and if his opponent takes positions that are far from the median voter’s view, his best strategy is not to take the same views but to move towards that view. He wants to motivate his base to show up at the polls and so he’s unwise to go all the way to the opponent’s views, but he will go some of the way.
So in this case, with Elizabeth Warren being so far left, if she wins the nomination Donald Trump’s best strategy is to move further left than he would have if he had run against a more-moderate Democratic candidate. So it wouldn’t be surprising to see Trump advocate, say, a $10 or $12 an hour minimum wage, to take just one example.
The only exception would be if the politician is highly principled and if his main goal is to implement particular policies rather than to win. Does that sound like Donald Trump to you?
By the way, I posted about this 2 years ago.
READER COMMENTS
Jon Murphy
Oct 19 2019 at 11:11am
Why would Trump move toward Warren and not Warren move toward Trump?
This is a hypothetical question. I think there is fairly little of substance between them already.
David Henderson
Oct 19 2019 at 11:20am
Good question, but, by the way, I strongly disagree with your statement that there’s little of substance that separates them. On trade and protectionism, they seem equally bad. But think minimum wage, Medicare for All, and wealth tax, just to name 3.
To your question, both will happen. However, her stances in her campaign for the Democratic nomination will somewhat limit her movement away from the left. The bottom line is that we will get both candidates being more left-wing than if, say, Biden were the nominee.
Jon Murphy
Oct 19 2019 at 11:49am
In support of your comment here, I originally had a stronger phrasing (“no substantial difference”) but between the type I wrote it and the time I posted, I had moderated. So, I may moderate even further before too long.
Yes, I agree with that.
Jon Murphy
Oct 19 2019 at 11:12am
I gotta admit: the Median Voter Theorem is something I’ve wrestled with (not conceptually but from a point of view of how useful a theory it is). I’m hoping for good conversation here.
Matthias Görgens
Oct 19 2019 at 2:33pm
The dynamic you describe would ld be a problem if Trump wins and implements promises he made in the campaign.
Not sure the latter is at much of a risk of happening.
David Henderson
Oct 19 2019 at 5:53pm
Good point, but for better or for worse, he seems to have been better at keeping promises than previous presidents.
Mark Z
Oct 19 2019 at 7:06pm
I think this overlooks the importance of the decision potential voters make other than who to vote for: whether to vote at all. I’d expect the closer one candidate moves to the position of another candidate, the less likely people a significant distance away from the candidates are to vote, because who wins seems to make less of a difference. And depending on where on the spectrum voters are concentrated, it may behoove a candidate to move in the opposite direction of the other candidate’s movements. E.g., if two candidates, A and B, start at -1 and 1, respectively (where the median voter is at 0) but the modes of the voter distribution are at -5 and 5, then A moves to -5, B’s best strategic option may be to move to 5 (or at least move in that direction) because the gains from increasing turnout among the ‘positive’ voters.
In practice, if the vast majority of potential voters lean solidly to one side or the other (which I think is true), then the main focus of a campaign should probably not be to win over the few genuine undecideds or possible converts among moderates on the other side, but to drive up turn out on one’s own side. So a candidate who observes his/her opponent becoming more extreme may reason (perhaps correctly), ‘I can now afford to become more extreme myself and give my base more of what they want, ginning up turnout, without losing too many voters near the center.’
zeke5123
Oct 21 2019 at 5:20pm
I think in your example you would likely move to 0. There is still a meaningful difference between you and -5 (i.e., 5 units). So, voters +1 through +5 would still be motivated to vote for you. Plus, you’d pick up -1 and presumably -2 voters.
But I think your point remains; centrist don’t necessarily drive the base.
Chris
Oct 22 2019 at 1:05pm
While I generally agree with your view and that was the first thing I thought of as well, I think there are a few counterpoints: By moving farther away from a centrist position, you are increasing the likelihood of your base voting for you, but also of your opponent’s base voting against you. For instance, the democrats could send a houseplant to the general election and I would vote for it over Trump. Trump’s extremism is driving my voting much more than any of the positions a democrat brings. I feel like this is the general view of most Republicans as well; they are voting against Democrats, no necessarily for their candidate. It’s why conservative attacks seem to be less ideology based and more fear mongering based; demonizing liberals.
This likely gets further complicated by the electoral college. As Democrats tend to be packed into fewer districts, the more optimal goal of a democratic candidate may be to limit republican turnout in swing districts. One method of reducing that turnout would be to go centrist so centrist republicans are more likely to stay home.
Mark Brady
Oct 19 2019 at 9:29pm
“So in this case, with Elizabeth Warren being so far left, if she wins the nomination Donald Trump’s best strategy is to move further left than he would have if he had run against a more-moderate Democratic candidate.”
Are we to assume that “left” means more federal spending for civilian ends? I for one don’t wish to concede that usage. And I invite my classical liberal friends to eschew that practice.
David Henderson
Oct 19 2019 at 10:22pm
Yes, I think “left” means for federal spending for civilian ends as well as higher taxation and more regulation.
I could have used the word “statist,” but Trump is highly statist also, just in a different way, for example, immigration.
Mark Brady
Oct 20 2019 at 2:27am
“Yes, I think “left” means for federal spending for civilian ends as well as higher taxation and more regulation.”
Okay, but I don’t favor that more extended usage either. Not least because I don’t want to label lower spending and taxation and less regulation as “right.”
“I could have used the word “statist,” but Trump is highly statist also, just in a different way, for example, immigration.”
In fact he’s pretty statist in most ways! That said, I suggest we should always discuss the median voter theorem with reference to a particular dimension of government policy, e.g., higher vs. lower taxation, or more vs. fewer immigration controls.
Mark Z
Oct 20 2019 at 9:35pm
Hasn’t increased government spending on non-defense pretty much always been regarded (both on the left and right) as a left of center position, and reduced non-defense spending a right of center position? Warren herself, and most people on the left, certainly seem to believe so, indeed many people think this is a key divider of left and right. I’m not sure what efforts to redefine the terms away from the common usage really accomplish. Perhaps the terms are often useless, but then it’s easier to just not use them. They seem like a clear example socially constructed concepts, that only mean anything when they mean what they mean in common usage.
Jason S.
Oct 19 2019 at 9:45pm
David, I don’t think this is right, unless you think that the ideological position of the median voter is somehow endogenous to the ideological position of the Democratic nominee.
The standard assumption is that the position of the median voter is fixed. If Democrats nominate a further-left candidate, that makes a further-right Republican candidate more viable. With perfect information about the ideological location of the median voter, there is absolutely no incentive for Trump to move left if Warren is nominated, unless Warren moves right.
Think about the famous Romer & Rosenthal article in _Public Choice_ about how zero-based budgeting in school elections makes budgets bigger. When presented with an extreme choice (shut down the school) and a slightly less extreme alternative (increase budgets and taxes significantly), the median voter is more likely to select the latter than they would if the alternative were simply a default budget.
Matthias Görgens
Oct 19 2019 at 11:28pm
The median voter theorem says that under its assumptions it’s optimal to be just off the other candidate, if you care about maximising the number to votes. (Or alternatively, you introduce some uncertainty and randomness, and then have candidates optimise the probability of winning.)
You are right that the other guy moving away from the median voter makes your own moving away in the other direction more viable; just like in any other game your opponent making less than optimal moves lets you get away with a few mistakes yourself.
But a strategy becoming more viable, doesn’t mean it’s becoming closer to optimal.
P Burgos
Oct 20 2019 at 12:59am
I keep reading libertarians say that Warren is too far to the left to be electable, yet I never hear a detailed discussion of whether her “leftist” policies are actually to the left of what the majority of the electorate supports. My recollection is that raising the minimum wage is supported by more than 50% of Republicans, as is the idea of raising taxes on the wealthy (which Trump promised to do in his 2016 primary campaign, not coincidentally). I haven’t followed the details of all the Dem candidates policy proposals, but Medicare is really popular with voters, so in theory a campaign that promises “Medicare for all” could be very enticing to voters, depending on the details.
Regardless, my main point is that generic analysis of voters and candidates as being “left” or “right” is too simplistic, and that it matters whether the policies that politicians campaign on are popular or not, whether or not those ideas are “left or right”.
Mark Brady
Oct 20 2019 at 2:28pm
Well said.
And readers may wish to compare President Nixon’s plan in 1974 with current proposals. https://khn.org/news/nixon-proposal/
Mark Z
Oct 20 2019 at 9:59pm
I’m wary of using opinion polls as a gauge of public sentiment; as critics of “Medicare for all” are fond of pointing out, support for it varies wildly depending on the framing.
That said, I don’t think an opinion that is fairly popular cannot still be described as far left or far right. Policies may be judged in terms of how much of a deviation they are from the policy status quo. An uncomfortable truth is, there are probably a decent number of policies that would be unprecedented in the US, would clearly run afoul of the bill if rights and our most sacred norms and morals, that would be supported by a third of people in opinion polls.
P Burgos
Oct 21 2019 at 9:22am
I think that there are some positions which have been consistently left or consistently right, yet Trunp’s candidacy and presidency make me a little bit wary of labeling anything as left or right anymore. I think we are in a period of time in which there is a great deal of dissension within the two parties as what constitutes their orthodoxy, and I am not sure where they are going to end up.
I agree with you about the electoral wisdom of running on “Medicare for all”. If polls don’t show a consistent preference for it, why go there at all? Past attempts to reform health insurance in the US have all been electoral failures. But there are other things that are a bit more clear cut, like raising the minimum wage or raising taxes on the wealthy or on corporations. That isn’t to say that those are good ideas, just that they consistently poll well, and that finding is robust to various ways of posing the question.
Todd Ramsey
Oct 20 2019 at 11:00am
IMO this discussion ignores the likelihood that some voters choose for reasons other than strictly rational analyses of policy. Rather, some voters choose on an emotional assessment of, perhaps, “likeability”, then attribute their choice to some policy statement of the candidate.
Without comment on whether Warren would be a good President, she seems a particularly uninspiring candidate on an emotional level. She seems a less likeable, less charismatic version of Hillary Clinton, if such a thing is possible.
I know that snarky line adds nothing to the discussion, and I apologize if I offended you. It struck me funny so I wanted to put it in cyberspace.
Alan Goldhammer
Oct 20 2019 at 11:03am
Whoever wins the Democratic Party nomination is guaranteed 40% of those who vote in November. Trump, assuming he is on the ballot, is probably guaranteed somewhere in the low 30s. The rest of the voters will decide who wins in 2020.
I’m not a Warren fan by any means and think that there are alternate approaches towards providing healthcare ‘for all’ and figuring out a better approach to the tax code. She should not be underestimated as a candidate as she is a very good speaker and debater. I’m not saying, but I’ve already contributed to someone running for the Democratic nomination and it’s not one of the front runners.
IVV
Oct 22 2019 at 1:11pm
Right, but which candidate results in the best outcome for upper middle class workers? Oh, right, never any one of them.
At this point, I don’t know if it’s the class or being Generation X that’s talking.
Todd Kreider
Oct 22 2019 at 5:34pm
Just one week later and the four most recent polls show Biden with a strong lead:
Biden….31
Warren…21
Sanders..20
Comments are closed.