Liberals, Conservatives and Independents

A blogger called Scott Alexander laments the forming of political opinions on a tribal basis, citing the ebola quarantine and other issues.

But he provides no solution. Using his rubric, any argument or position can be considered “politicized” or tribally motivated.

He notes that the Rotherham rape scandal, where Pakistani gangs in Rotherham, England took thousands of girls as sex slaves, and authorities actively did nothing for fear of being considered racist, and those who wanted to publicize the issue were sent to diversity training, fit perfectly into conservative narratives. Consequently, “the Left then proceeded to totally ignore it, and the Right proceeded to never shut up about it for like an entire month, and every article about it had to include the “diversity training” aspect, so that if you type “rotherham d…” into Google, your two first options are “Rotherham Daily Mail” and “Rotherham diversity training”.

Alexander then notes a blogger who writes on race and genetics and weighed in on Rotherham, for what Alexander suspects are tribal reasons: “I have no doubt that her outrage is genuine. But I do have to wonder why she is outraged about this and not all of the other outrageous things in the world. And I do have to wonder whether the perfect fit between her own problems – trying to blog about race and genetics but getting flak from politically correct people – and the problems that made Rotherham so disastrous – which include police getting flak from politically correct people – are part of her sudden conversion to political activism.”

But who cares what people’s motives are for being right? Anything people say can have suspicious motives, so there is no way to avoid these accusations of tribalism. By making motives central (you’re only saying that because you’re conservative!), Alexander reinforces tribalism. The only real solution is to evaluate arguments themselves rather than motives behind them.

Alexander does genuinely bash the left for ignoring Rotherham. In doing so, he no doubt thinks he’s proving his independence, and thus his freedom from tribalism. But independent posturing is a motive no more noble than tribalism- it’s about taking positions to reinforce a narrative.

And of course the world is full of people who are actually on one side or the other, but triangulate against “the tribes.” Usually Alexander’s own stance is substantively identical to the progressive side, and only stylistically independent. Stylistic elements include an anti-ideological, data-driven, pragmatic outlook (or pose), but that’s exactly the outlook movement progressives maintain; so even stylistically, there’s no real difference between Alexander and “the left,” in its current form.

For instance, he discusses conservatives’ uniting in favor of quarantines to solve ebola, and liberals’ uniting against it, in an above-the-fray way, exuding superiority to both tribes. But then he says “it would be really hard to switch these two positions around. If the Republicans were to oppose a quarantine, it might raise the general question of whether closing the borders and being scared of foreign threats is always a good idea, and whether maybe sometimes accusations of racism are making a good point.” And here he does not prove his balance by treating the liberal position in a similar way. Because he agrees with the liberal position. Once you take a stand, what makes you any different from the “tribalists” taking that same stance?

If, as is often the case, there is something about the form or rhetoric of the argument itself that is tribal, then we can legitimately make a critique. If not, then not.*

In the case of Rotherham, moreover, it is not only true that what happened is terrible in an isolated sense. The desire to reduce the power of political correctness, far from being a corrupting motive sullying the purity of moral outrage, is actually the solution to the problem. Insofar as outrage at Rotherham leads to a reduction in the social status of political correctness, future Rotherhams will be less likely.

It is, contrary to Alexander, very easy to see why someone might get outraged about Rotherham and not one of the many other outrages in the world. Americans identify with England in a way they don’t with South Sudan and Syria the Congo; that’s tribal, but not in a way that’s really relevant to Alexander’s analysis. And most of us don’t understand South Sudan or Syria or the Congo, the politics and motives of these places, or what can be done to solve their problems, but we understand political correctness very, very well.w

***
Nor is tribalism necessarily all that important. In most cases, the Parties structure themselves so that the median voter prevails, so that no actual policy issues are at stake in elections and the people nevertheless get what they want, but the only thing left to fight on is trivial matters. On the rare occasion that an actual major policy change was clearly at stake in an election, and the public didn’t want it, the people of Massachusetts went against tribe voted very solidly for Scott Brown to try to block Obamacare.

Like everyone else in the progressive tribe, Alexander assumes that conservative tribalism is the barrier to action on global climate change. But there’s no analysis that I know of showing that this is the case.

If we can show that majorities of voters believe in human-caused global warming, that would indicate that the barrier to persuading the median voter is something other than a widespread rejection of “scientific consensus.” The polling seems ambiguous. Pew finds that 67% of adults believe in global warming, but only 44% believe it is primarily human caused; but some of these people may have been hedging. When Gallup didn’t give people the option of global warming “denial,” instead presupposing that temperatures have increased and asking what the cause was, they got a solid 57% saying human activities are the cause. This divergence obviously indicates tribalism (in Pew, the way to prove conservative tribal loyalty was to say global warming doesn’t exist, while the way to prove you’re moderate was to say it exists but has “mostly” natural causes, while in Gallup saying global warming is caused by nature is the only option to prove conservative tribal loyalty, and the centrists apparently went with anthropogenic), but overall I’d take it as indication that the median voter agrees with liberals that humans are causing global warming. At least, the opposite isn’t clearly the case.

But that doesn’t guarantee they support cap and trade. The median voter may not cite Bjorn Lomborg and make the cost-benefit arguments associated with the conservative tribe, but he may also just not care much about global warming, not think much can be done about it, and not be willing to sacrifice anything economically. (Note how Barack Obama mostly appeases the blue tribe on global warming as a hot button issue, while letting the U.S. oil boom happen and bragging about it, thereby appealing to the median voter.)

Of course you could argue that conservatives are better organized and are thereby blocking action, but remember that they needed coal state Democrats to help them block cap-and-trade, which proves the power of organization, but not of conservative tribalist organization by itself.

***
As for ebola quarantines, Alexander treats it as utterly irrational that liberals and conservatives lined up in different positions. “Weird that so many people suddenly develop strong feelings about a complicated epidemiological issue, which can be exactly predicted by their feelings about everything else.” But maybe some general principles explain why people line up on issues. And after all, they don’t do so perfectly by any means, but only to such an extent that we at least have a concept of “a conservative” and “a progressive.” Are these labels merely descriptive of a random combination of issue positions, or is there something at all deeper behind them? Alexander himself suggests the latter in noting progressive and conservative narratives and the way different events and ideas do or do not fit with each. But these narratives (or some narratives- Alexander may or may not be emphasizing the most important ones) are things people believe! That being the case, their political positions may be rational. Liberals may care more about civil liberties for foreigners than the “complicated epidemiological issue,” at least at the relevant margins.

And even to the extent this is an epidemiological issue, people may not see it as a complicated one- they may see it as a matter of “common sense,” something they don’t need specialized knowledge to figure out. How and when we should rely on common sense and when we need specialized knowledge to make decisions is itself a complicated social and epistemological issue.

But probably tribalists have acquired specialized knowledge. Everybody knows that strong tribalists are the best-informed people, exhibit strong confirmation bias, and become more confident in their initial opinions as they become better informed.

* Indeed, I actually despise tribal arguments. My mood affiliation, as Tyler Cowen would say, is entirely on the side of rejecting them. So by tribal affiliation, I should be on the side of Alexander, and other members of the “independent” tribe like Tyler Cowen and Megan McArdle who have been linking to him. By taking a stance against my tribe in this post, I am of course proving my independence from even the independents, which of course proves I must be even more right than they are, which is doubtless my real motive in taking these positions, which then calls them into question, which…

Evaluating a Debate on the Minimum Wage

In this, the third in a series of posts evaluating Intelligence Squared U.S. debates, I consider a debate on the proposition: Abolish the Minimum Wage.

Russ Roberts delivers the 1AC, making the standard theoretical case against the minimum wage, and argues that it applies to our specific circumstances because employers are manufacturing more with fewer workers, meaning there are substitutes for low-wage labor. He hints at an argument for quantifying effects in terms of the magnitude of the impact of losing a job rather than the number of people losing a job. He compares 1.7 million low-wage workers getting a raise to the 3.4 million people looking for work.

Jared Bernstein, a Democratic economist, argued in the first place that abolishing the minimum is outside the mainstream, supported by the likes of Michele Bachman and Herman Cain.

Now, was this pander to the liberal Intelligence Squared audience (in this case, a D.C. audience) really necessary? A theme of this series has been the attempt by some speakers to get the audience to think tribally, in culture war terms. Though we can certainly expect the audience to oppose Bachmann and Cain, we can also expect them to oppose the minimum wage in the first place. In fact, to give them some credit, the audience may oppose Bachmann and Cain because these two are the sort of people who oppose the minimum wage, rather than supporting the minimum wage because Bachmann and Cain oppose it. One might think the opposition (pro-minimum wage) team would have nothing to lose by sticking to substance. Still, perhaps there was danger that the audience might be more open-minded in response to a substantive policy debate, and so it makes sense that Bernstein would want to give the audience a helpful framing device (mainstream vs. right-wing extremist.)

Bernstein then argues that globalization and technological advance have led to economic growth no longer reaching low-wage workers, making the minimum wage necessary.

He says he has evidence that the minimum wage causes little increase in unemployment, and hopes “we’ll have lots of time to get into it.” He vaguely claims that most studies show the benefits outweigh the costs, but does not attempt to quantify either benefits or costs, or give a suggestion for how to compare them.

So through two speeches, it is unclear whether we have disagreement about whether and how much the minimum wage reduces employment. We have potential epistemilogical clash between Russ Roberts’s Austrian, deductive paradigm and Bernstein’s empiricism. I fault Bernstein, as the first negative, for failing to provide this clash. He ought to have given us some analysis on how to weigh his [unspecified] studies, results and methodology against or in light of what Roberts shows we can expect deductively. But again, it isn’t even absolutely clear that they disagree on results on can expect; they may simply disagree on how to weigh higher unemployment against higher wages.

Bernstein provided, as I mentioned, no analysis on how to weigh these things, while Roberts’s arguments on this point were rather weak. Roberts spiked the argument on small employment effects, saying “When you lose your job or can’t find one, the effect isn’t small, it’s 100 percent.” He needed a much more explicit and systematic case for quantifying by magnitude of impact rather than number of people impacted. His point that only 1.7 million workers earn more as a result of the minimum wage, or would earn more if we increased the minimum wage, or something while there are 3.4 million out of work who will thus have a harder time finding work, is dubious because we don’t know how many of those 3.4 million would find work without the minimum wage.

The state in which Bernstein’s speech left the debate allowed great scope for a killer 2AC, but people always bring prepared speeches to Intelligence Squared U.S. rather than debating from the flow. The 2AC and 2NC simply gave high school novice style speeches (my evidence says x. Yeah, well my evidence says not x.)

Now, to the Q&A. Russ Roberts says, look, there’s studies showing one thing and another, and if either were thoroughly convincing, economists would all be convinced. Because the world is a complicated place, nobody can do a thoroughly persuasive study, so everyone just accepts the studies that fit their ideological convictions. So we have to debate these convictions themselves. Bernstein did not address the epistemological debate, but at this point decided to be honest about the empirical research yielding a variety of results, and that overall he would predict that a 10 percent increase in the minimum wage leads to a 1 percent decline in the employment of teenage workers.

Pressed later by the moderator to address the affirmative team’s deductive case, he said these deductions were a theory and had to be tested empirically. He did not say why this is the case (why shouldn’t we test the propositions deductively? Why and how are they subject to empirical testing?) Roberts restated his epistemological argument, which Bernstein never really addressed.

Having acknowledged, as I noted above, the likely decline in employment resulting from the minimum wage, Bernstein finally attempts to address the question of weighing costs and benefits. “99 percent of affected workers get a pay raise. Now, our opponents have consistently said, if 1 percent loses a job…yet somehow 99 percent get a raise, it’s a bad deal. That’s just, I think, economically a very — a very misleading –” (here he may have been cut off.)

But “Economically” has nothing to do with it. Economics can’t tell us whether it is good or bad to sacrifice one person’s job to give 99 people a raise; this is an ethical question. But why would anybody favor such a thing- sacrificing some people’s jobs to give a larger number of people a raise? Is there any reason for holding a notion of distributive justice to take one person’s job away and spread his wage among 100 poor people? And if it isn’t a matter of distributive justice, is there some other reason for doing this?

Later in the Q&A, Jim Dorn makes a moral argument that I flowed across from Bernstein’s dubious moral theory: “The question is whether individuals should be free to choose on their own terms…If the government says the employer has to pay you $9 an hour and you’re only producing $5 a hour, you’re going to lose your job.” Bernstein responded with a study showing that 90% of low-wage workers favor raising the minimum wage, and claimed that this proved the minimum wage didn’t interefere with workers’ freedom, suggesting that he views freedom in terms of participating collective political decision-making rather than individual decision-making. Dorn tries bringing the focus back to the individual: “Well, ask this lady that lost her job in New Hampshire.” Bernstein responds: “If you’re going to find one case of one person who loses their job and say, “We should abolish the minimum wage,” you’re nuts.

But that only returns us to the question of why this is nuts- why should we consider 99 people getting a raise more important than one person losing a job? What criterion is Bernstein advocating for judging policies, and why that criterion? What is wrong with Dorn’s criterion of judging policies on individual freedom?

Dorn’s own argument could be more precise. Probably, though, his position is based on the primacy of the individual (he doesn’t believe in freedom of contract for its own sake), so two things become important: the magnitude of the impact on an individual, and the individual’s moral claims. The first is obvious: if we take the individual to be of central importance, then we care most about those on whom the policy has the strongest impact. Bernstein doesn’t actually dispute that job loss is, all in all, the highest magnitude impact in the round. He just thinks it is insane to care about this.

The focus on the individual’s moral claims also follows from individualism. If we are to judge costs and benefits on individualistic grounds, we naturally will be leery of sacrificing one person to bring benefits to another. But speaking of sacrifice presupposes that someone has a claim on something in the first place. Voluntary exchange or contract is a basis for such a claim, allowing one person to gain benefits without hurting others. Of course, high magnitude impacts can trump such claims and be consistent with an individualist criterion.

Bernstein’s implicit criterion is number of people affected. Presumably, we could take from 10% of people and give to the other 90%, or take from 40% and give to the other 60%, and this would be a good thing. This is different from utilitarianism, which seeks to quantify total benefits, and therefore considers the magnitude of effects as well as the number of people effected. It is different from Rawlsianism and other forms of egalitarianism: it can be any 10% or 40% of people we are taking from, not just the rich. You just have to take money from one person, however poor, and spread it among any number of other people, the more the better, and you are doing right, as far as I understand his position.

However, the only reason anybody cares about the minimum wage in the first place is the magnitude of its impact on a small number of struggling people. The opposition itself argued that much of the cost is spread throughout society, so clearly they think it’s desirable to greatly help a small number of people in dire need at the expense of small costs to larger numbers of people spread throughout society.
So the Bernstein theory of distributive justice is convoluted and nonsensical, and certainly not intuitive. Yet he views it as straightforward and common sense, and he knows the audience will agree with him, so much so that he doesn’t feel the need to debate it, even though he is at a debate.

The reason the audience thinks he has common sense on his side doesn’t really have to do with distributive justice. Nobody’s idea of distributive justice is “let’s take everything away from a few poor people, and spread it out among a larger number of poor people.” Rather, the audience favors the minimum wage because it wants to reward work. People don’t like welfare or distributive justice; they want to give money to those they see as deserving. Thus, they like the minimum wage and dislike things that make much more sense from the point of view of distributive justice.* Thus Bernstein, prompted by an audience questioner, makes one of the most popular arguments for the minimum wage: “[I]f you abolish the minimum wage…you collect fewer payroll taxes. You will have to pay more in food stamps. You will have to pay more in unemployment insurance. [???] You have to pay more in welfare benefits. And it is very much a transfer of the burden of poverty to the taxpayer in ways that minimum wages and living wages help to mitigate.” If we cared only about distributive justice, and not deserts, what could be more rational than transferring the burden from employers to taxpayers?

One questioner suggested making the EITC generous enough to replace the minimum wage. The EITC is of course specifically designed to give welfare only to the deserving, to workers, and yet Karen Kornbluh’s response was that the minimum wage is more consistent with the dignity of labor: “I do think there are a lot of people who just want to earn a living…they would like to work for a salary…And so I don’t think we should underestimate how much people want to get their living from their job.” Your employer giving you the money, even if they are forced to, just feels more like you have earned the money than if government gives it to you.
***
At the beginning of the debate, Bernstein sounded like your typical Democrat talking about the economy: low-wage people are getting screwed.

“As globalization, and technological change, and a lot of other stuff that we can talk about throughout the evening have evolved, economic growth no longer reaches working families the way it used to. And the further you go down the pay scale, the less growth you’re likely to see.” However, these arguments are of course by no means friendly to the pro-minimum wage position. Russ Roberts used them to argue that minimum wage workers are replaceable:

“But my question for you is, since you reject the logic that we made, and you say it’s only an empirical question, I’d like to know why you think employers over the last 30 years have relentlessly punished not deliberately, but through their actions, have relentlessly punished low-skilled workers because they are relatively unattractive compared to workers outside the United States and automation. Do you really think that work — that employers don’t respond to economic incentives?”

Bernstein then found himself forced to argue like an out-of-touch, right-wing, supply-side economist, clinging to the late 90s to show that low-income people haven’t had it uniformly bad:

“Employers have not relentlessly punished low-wage workers. In fact, in the 1990s, a period when the minimum wage was increased in 1996, the unemployment rate among low-wage workers fell to the lowest it had been in 30 years. The national unemployment rate fell below 4 percent. That was the time of welfare reform. And hundreds of thousands of low-wage, poorly-educated single moms came off welfare, got into the labor market. And with the complimentary boost of the minimum wage and a higher EITC, poverty rates fell very steeply.”

* Here I don’t mean to really endorse distributive justice myself.