Occasional blogging, mostly of the long-form variety.

Sunday, August 17, 2014

Artificially Equalizing Unequal Views on Inequality

The mainstream political press has a horrible tendency to rely on false equivalencies and to conclude, far too simplistically and without nuance or detail, that 'both sides are equally to blame.' It would take much more effort, and seem like taking sides, to call things accurately to the best of their ability, or discuss policy positions in terms of facts and likely outcomes, or contextualize negotiations and other matters for their audience. Even important issues can wind up being poorly covered. Some conclusions, no matter how well supported, are deemed unacceptable; others are essentially predetermined, regardless of the pesky facts. A Washington Post piece from July provides a good example.

The Article

"A think tank wanted to study inequality. It couldn’t get conservatives on board" by Jim Tankersley covers an interesting tidbit. (Its subtitle, "If most of the research on a big economic question comes from one political point of view, can Americans trust it?" seems like an awfully loaded and misleading question, however. To be fair, editors often choose the headlines, not the authors – but Tankersley is listed as the editor for this and other "Storyline" pieces.)

The setup—a new Beltway think tank (Washington Center for Equitable Growth) wants to study an important issue:

What she wanted, [Heather] Boushey said, was to fund an intellectually honest investigation of arguably the hottest issue in American economics right now: the widening gap between the richest Americans and everyone else. She wanted to learn more how and why that gap might hurt the nation’s overall economic performance.

The think tank has some members with a liberal background, but apparently it aims to be inclusive and is officially "non-partisan." Accordingly, it extends a wide invitation:

Boushey was hoping to sign up an ideologically diverse cross-section of thinkers to her effort. . . .

The grant recipients, she explained, would examine questions of inequality and growth from every possible angle, liberal and conservative. They’d report what was true and what wasn’t. Her think tank would live with the findings, whatever they might be.

The think tank's trying to be inclusive and strives to be empirical and data-driven. Sounds good so far. Next, proposals are submitted and grants are given:

There are no identifiably conservative economists among the grantees, however.

Uh-oh! That doesn't sound inclusive! What gives?

No conservative economists applied for grants, among the 70 submissions the center received, Boushey said.

...Oh. Well, it's rather hard to offer a grant to someone who can't even be bothered to apply. Not even one of the 70 applicants was a conservative? That seems remarkable. Why was this? Tankersley theorizes:

Perhaps that’s because the center has, in its early months of existence, engaged in some high-profile blog-and-social-media fights with conservatives over questions of data and policy when it comes to inequality.

Alas, no links are provided for this. This sounds a bit whiny, that the new think tank was mean, so conservatives didn't want to play. The think tank's twitter feed seems fairly anodyne and not combative. Back in May, Boushey did appear on PBS' NewsHour to debate Kevin Hassett of the conservative American Enterprise Institute (AEI) about Thomas Piketty’s recent book, Capital in the 21st Century, but it's a pretty polite exchange of disagreements. The most charitable possibility (for conservatives) would be that they judged that their contributions would be unwelcome. (It's a paid gig, though. And wouldn't getting paid to skewer a liberal position be tempting?) But Tankersley's speculating anyway, so let's move on. He floats another idea:

Perhaps it’s because it’s relatively rare in Washington for liberal and conservative thinkers to team up to pursue big questions – especially ones that, depending on the answer, could shake party platforms or basic ideological beliefs.

This is a much more convincing hypothesis, especially if one's familiar with the usual conservative stances on inequality. Tankersley continues:

It’s also true, in this case, that a lot of conservative economists have already decided that inequality – as opposed to mobility or the even broader “economic opportunity” – isn’t a problem.

Aha! Yes, this is much stronger, more accurate stuff – conservatives have often stated that inequality is overstated, or that it isn't a problem, or that efforts to ameliorate it cause more harm than good, or all of the above. Tankersley's link is to conservative AEI and its latest 117-page collection of essays, which argue all of these positions. This is good context for the reader.

This is the real challenge posed faced by Boushey, and her center, and the work that will now begin thanks to their grants: If most of the research into an economic issue is being conducted from one side of a political debate, consumers of that research could – and should – look skeptically on its conclusions, until persuaded otherwise. The burden is on the researchers, on their data and the stories they tell from it.

It’s a heavy burden. It would help to have a more diverse group of thinkers shouldering it.

Umm… what? Where did this come from? How does this follow? And a "diverse group" needs to include people who explicitly don't want to study the issue? I suppose "until persuaded otherwise" and the mention of "data" soften this a bit, but this is quite the sharp turn. It also ignores that there's a great deal of solid research on inequality already. For example, Thomas Piketty's aforementioned book, Capital in the 21st Century, is a new, major work in the field, and it's earned praise for its extensive research and synthesis of previous studies. (It's also received plenty of attacks from conservative outlets.) This think tank story, and discussions about inequality in general, do not take place in a vacuum. Let's recap Tankersley's account and conclusion:

1. A think tank wants to study inequality. It invites conservatives to participate.

2. Conservatives completely decline to participate (zero of seventy applicants).

3. It's mentioned but isn't emphasized in the piece that the think tank is sponsoring data-driven work. (A chart of the grantees and their paper topics is included. The findings of previous major studies on inequality are not mentioned.)

4. The conservative position is that inequality doesn't matter. (It's not really explored that justifications for this tend to be ideological versus data-driven. More on this in a bit.)

5. Conservatives have their own think tanks upholding their position.

6. Somehow… the new think tank's conclusions will be suspect and the burden is on its members, despite conservatives rejecting their invitation and rejecting their entire endeavor (not to mention rejecting the findings of previous studies).

Tankersley sure seems to arguing that if conservatives don’t agree with something, it can't be valid (or at the very least, that it must be viewed with great suspicion). There are several glaring problems here.

It's completely reasonable and fair to let different political factions have their say. It's utterly ridiculous, however, to automatically judge that their positions are equally valid and sound – especially when the facts say otherwise. Serious policy discussions require some level of qualitative analysis and judgment. Tankersley offers a variation on the usual predetermined "both sides are equally to blame" conclusion – in this case, he essentially lets conservatives veto the propositions that inequality matters and that it should be studied. He doesn’t provide much context for the reader, but essentially, he's falsely equating an empirical approach (and extensive data from it) by one "side" in a debate with a conservative ideological position against it. (For instance, conservatives will indeed argue that efforts to lessen inequality are ineffective or counterproductive, but their central, much more common argument is that such efforts are morally wrong.)

But Tankersley's conclusion doesn't even follow his own account. He links AEI's collection of papers arguing that inequality doesn't matter (and is overstated, etcetera). How is this point of view not represented, then? Why does the Washington Center for Equitable Growth specifically need to represent it, too? (Especially when they tried to, and offered an invitation that was rejected?) How the hell is this their fault?

Put another way, Tankersley's stated concern is about "research into an economic issue . . . being conducted from one side of a political debate." Based on his own account, either the other side has refused to engage in research, or the other side is already engaged. Yet somehow, the conservatives are still blameless and the fault lies with the more liberal – in this case, the intentionally centrist – faction.

Realistically, the news business is based on sales, ads, and page hits – content is churned out, and some of its quality will suffer. A short, quickly written piece won't give a full account of an issue. However, it should give a roughly accurate account of an issue and some basic context, and Tankersley's piece is striking for its abrupt and ill-fitting conclusion. It reads as if he was writing up an interesting tidbit, then realized he had to end by blaming "both sides," so he took a sharp swerve. He could have tried a familiar Beltway gambit – holding this incident up as an example of political intransigence and then bemoaning the lack of bipartisanship in Washington – and he kind of does this briefly (the less context on the issue one provides, the easier it is to pull off this move). But that route is slightly problematic here, because one "side" is investigating the issue and the other is refusing to do so – so fault must be invented in the name of "balance." Basically, the piece offers a socially driven conclusion versus an empirical or logical one.

The Bigger Picture

Tankersley's piece exemplifies some persistent problems in coverage of policy and political disputes. Let's return to that loaded and misleading subtitle: "If most of the research on a big economic question comes from one political point of view, can Americans trust it?"

This asks for a simple "yes" or "no" in a polarizing fashion, not for deeper reflection. (As Neil Postman put it, poor questions "insinuate that a position must be taken; they do not ask that thought be given.") Why does any political point of view automatically possess virtue, insight and accuracy? The stated attitude is also a major problem in "debates" about climate change and evolution, among other things. What if one "side" rejects empiricism and facts? What if one side consistently lies? What if one side had obvious reasons for arguing in bad faith? In the context of the paucity of conservatives in academia, PZ Myers has tackled this issue (emphasis mine):

Of course if [Jonathan Haidt] believes there’s no real difference between left and right, it’s a problem that the right is poorly represented in academia.

But what if there is a real difference, a difference of substance, in how left and right approach the evidence and how they respect the methods of science? I think the field of social psychology is suffering because they haven’t hired enough serial killers for their tenure line positions. They’d certainly do a good job of making psychologists question their assumptions about the importance of health, happiness, and security in human welfare, and also, someone would be around who could finally intimidate those prissy-pants on the Human Subjects Review Board. Should we complain about the deficiency, or should we recognize that some behaviors are antithetical to the cooperative and responsible pursuit of knowledge?

The problem isn’t that academia excludes conservatives. It’s that it is a rare conservative who doesn’t prioritize the moral foundations (to use Haidt’s own terms) of respect for authority and loyalty to the ingroup above breaking through conventions and assumptions to test the truth. Also, it’s the rare conservative who will accept a job with high admission requirements that also pays a pittance.

(As commenters in a Lawyers, Guns & Money thread on Tankersley's piece point out, conservative think tanks likewise probably pay much better than the new think tank's grants.)

Returning to the subject of inequality, it's not a secret that conservatives generally defend it and oppose efforts to lessen it; this is a core ideological principle. As conservative David Frum explained to Bill Moyers in 2009:

The Republicans are not the party of equality. They're the party of liberty and they're the party of efficiency. . . .

Liberty leads to inequality just as attempts to reduce equality lead to adoption of liberty.

The data don't always cooperate with the conservative position, however. Paul Krugman's written a great deal about inequality over the years, and a recent op-ed, "Inequality Is a Drag," summarizes some recent research:

It’s true that market economies need a certain amount of inequality to function. But American inequality has become so extreme that it’s inflicting a lot of economic damage. And this, in turn, implies that redistribution — that is, taxing the rich and helping the poor — may well raise, not lower, the economy’s growth rate.

You might be tempted to dismiss this notion as wishful thinking, a sort of liberal equivalent of the right-wing fantasy that cutting taxes on the rich actually increases revenue. In fact, however, there is solid evidence, coming from places like the International Monetary Fund, that high inequality is a drag on growth, and that redistribution can be good for the economy.

Earlier this week, the new view about inequality and growth got a boost from Standard & Poor’s, the rating agency, which put out a report supporting the view that high inequality is a drag on growth. The agency was summarizing other people’s work, not doing research of its own, and you don’t need to take its judgment as gospel (remember its ludicrous downgrade of United States debt). What S.& P.’s imprimatur shows, however, is just how mainstream the new view of inequality has become. There is, at this point, no reason to believe that comforting the comfortable and afflicting the afflicted is good for growth, and good reason to believe the opposite.

Specifically, if you look systematically at the international evidence on inequality, redistribution, and growth — which is what researchers at the I.M.F. did — you find that lower levels of inequality are associated with faster, not slower, growth. Furthermore, income redistribution at the levels typical of advanced countries (with the United States doing much less than average) is “robustly associated with higher and more durable growth.” That is, there’s no evidence that making the rich richer enriches the nation as a whole, but there’s strong evidence of benefits from making the poor less poor.

(Hullabaloo has covered inequality quite a bit. My most extensive post on the subject is this 2010 one.)

The conservative American Enterprise Institute's aforementioned work on the subject is a collection of essays entitled, Opportunity for All: How to Think about Inequality. I'm not going to go through all 117 pages in detail (you can read it for yourself), but I did want to note a few key arguments. In AEI President Arthur C. Brooks' introduction, he states:

The conventional wisdom on inequality is built on three assumptions: (1) Income inequality is inherently unjust; (2) it is bad for the economy; and (3) government redistribution is the best way to remedy it. According to this narrative, narrowing the gap between what wealthy and working-class Americans earn should be our top political priority, and policies such as raising taxes or increasing the minimum wage are the answer.

It's not encouraging when the opener offers such a huge straw man argument (and tinier ones). "Income inequality is inherently unjust"? As Krugman and others have noted, some inequality is necessary and useful, but beyond a certain amount it is indeed "bad for the economy." Direct government intervention can work, as can indirect intervention, and whether it's the "best" option or not, sometimes it's the only real option (see the New Deal). That's the practical argument. The moral counterargument to Brooks is that, if the political will existed, it would be relatively easy to have a system that still featured inequality but also much more "equality of opportunity" and a higher standard of living for everybody. Compared to other industrialized nations, the U.S. does relatively little social spending. Income and wealth inequality have increased over the past few decades, and the rich are being taxed at low levels by historical standards; U.S. taxes could become far more progressive and the richest 1% would still be left immensely privileged, if merely filthy rich versus obscenely wealthy.

In the same vein as Brooks, AEI author Aparna Mathur writes:

Like most Americans, I believe in equality of opportunity and not equality of outcome, which is clearly neither desirable nor attainable.

This is both a false dichotomy and a straw man. (Throw in appeal to the majority, too. What efficiency!) Of course, liberals (and the think tank centrists from earlier) don't seek "equality of outcome"; this is a wearily common conservative straw man tactic, essentially attacking communism and pretending that it's what liberals or European social democrats (or centrists) want. The goals are to reduce inequalities of both opportunities and outcomes, not to eliminate them altogether. In liberalism, governing is a balancing act, and theory and practice inform each other. Outcomes aren't the only concern, but they can't be ignored, either. The entire point of empiricism is to collect good data, which, among other things, entails – duh – looking at outcomes, measuring them and trying to determine causes with greater acuity. If the data show great inequality of opportunity, and also show great inequality of outcome – and the perennially preferred conservative policies to address the former have been shown to be insufficient or simply ineffective – perhaps something else is needed, hmm? If the data show that greater government spending does help, but conservatives oppose that, shouldn't they have to give a good reason and not just their position? If slightly higher taxes on the rich would help fund that spending, why not do it? (And do we really have to pretend that conservative arguments against even slight increases in taxes are always in good faith and with sound reason versus self-interest? Some conservatives are at least forthright about it.) It makes sense that most national political coverage avoids discussion of motives and bad faith, but it's not as if it sticks to the data instead – on inequality, climate change and many other subjects, it's been proven acceptable (for conservatives, at least) to simply say the data don't matter. Only rhetoric (which it would be rude to fact-check) can be allowed to prevail.

"Debates" about inequality do not take place in a vacuum – a long history and context exists. As Krugman observed in a July op-ed, "On Inequality Denial":

Not only do the usual suspects continue to deny the obvious, but they keep rolling out the same discredited arguments: Inequality isn’t really rising; O.K., it’s rising, but it doesn’t matter because we have so much social mobility; anyway, it’s a good thing, and anyone who suggests that it’s a problem is a Marxist.

What may surprise you is the year in which I published that article: 1992.

(Krugman's far from the only one writing about this stuff, but he's awfully good at it.)

That's a fine summing up of the arguments from AEI and other conservative outlets. In the same piece, Krugman also covers criticisms of Piketty's work:

At the risk of giving too much information, here’s the issue. We have two sources of evidence on both income and wealth: surveys, in which people are asked about their finances, and tax data. Survey data, while useful for tracking the poor and the middle class, notoriously understate top incomes and wealth — loosely speaking, because it’s hard to interview enough billionaires. So studies of the 1 percent, the 0.1 percent, and so on rely mainly on tax data. The Financial Times critique, however, compared older estimates of wealth concentration based on tax data with more recent estimates based on surveys; this produced an automatic bias against finding an upward trend.

In short, this latest attempt to debunk the notion that we’ve become a vastly more unequal society has itself been debunked. And you should have expected that. There are so many independent indicators pointing to sharply rising inequality, from the soaring prices of high-end real estate to the booming markets for luxury goods, that any claim that inequality isn’t rising almost has to be based on faulty data analysis.

Yet inequality denial persists, for pretty much the same reasons that climate change denial persists: there are powerful groups with a strong interest in rejecting the facts, or at least creating a fog of doubt. Indeed, you can be sure that the claim “The Piketty numbers are all wrong” will be endlessly repeated even though that claim quickly collapsed under scrutiny. . . .

This picture makes some people uncomfortable, because it plays into populist demands for higher taxes on the rich. But good ideas don’t need to be sold on false pretenses. If the argument against populism rests on bogus claims about inequality, you should consider the possibility that the populists are right.

Exactly. But then both sides aren't equally to blame, and there's no good reason to oppose policies helpful for the middle class (and poor) that powerful groups do in fact oppose. That's awkward for media outlets that traditionally toe the establishment line.

Moral arguments have their place (and can have great value), but the honest use of data is extremely helpful for cutting through biases and grounding those arguments. 'My ideology states that I don't need to care' is a position, not a true argument, and shouldn't cut it in serious discussions. Validating it amounts to empowering the fanatic's veto – it isn't even-handed; it's irresponsible. Having an open mind entails giving someone a fair hearing, not ignoring all evidence, judging all opinions equally sound or turning off one's bullshit detector. False equivalencies do not serve the audience (but unintentionally or not, they do serve other parties).

(Cross-posted at Hullabaloo.)

1 comment:


Great work!