The Atlantic Home

A day in which two big Obama polls have been released (NYT poll here and Pew poll here) seems like a decent enough excuse to revisit the question of the value of political polls. Mark Blumenthal, Ed Kilgore and John Sides (twice!) have all been kind enough to write detailed responses to my original thoughts on the subject. And I'm pretty sure they are winning the argument. Nonetheless,  I don't think we're at Waterloo just yet, and I want to respond to some of their points.

This is dredging up an old controversy, so here is a quick recap: I wrote a short piece for the Atlantic a few weeks ago ("Get Rid of Polls") that piece made three basic points: (1) we should want our democratic institutions to operate according to pre-established mechanisms, not random and constant quasi-referenda; (2) Lots of polls are wrong or misleading or insufficient snapshots of "actual" preferences; and (3) Present opinion polls can affect future opinion polls, due to information cascades. (Which offends my romantic sense of how and why opinions should change.) In a follow-up I added a fourth point, which was mostly a desperate attempt at burden-shifting: Even if you disagree with the three points above, what's the affirmative case for polling?

Well, Mark and John and Ed responded to my three points and answered my question and more. So, in no particular order, here's a few thoughts on the issues they raise [looking back over it after writing, I think the debate probably boils down to #'s 4, 6 and 7]:

1. Let me make one big concession at the start. I wouldn't vote for a law than banned polls, and were I dictator of the universe I wouldn't want to outlaw them. (I'm a fan of free speech; I'd get that first amendment tattoo in a heartbeat; etc.) The position that I would feel more comfortable defending is something like "polls are on balance a bad thing." Or, even more milquetoasty: "polls are not the best use of newspaper resources."

2. I hope that goes along way towards addressing the main thrust of Ed Kilgore's argument -- namely, that I am weirdly in favor of reducing the total amount of information that exists in the public sphere. I think it's fair to say, with Ed, that there should be a strong presumption in favor of putting additional information in the public sphere. But I think that's very different from saying the current level of information is optimal or desirable.

3. I think there is an easy extension of points #1 and #2: Producing information isn't free. A dollar that the New York Times spends on a poll is a dollar that it isn't spending on a Baghdad bureau or a congressional beat reporter. Well, that's not really how tradeoffs work.  But the general point -- you have to measure something against its opportunity cost -- is one I agree with and one that is relevant here. So one question to consider might be this: Could the resources that the Times and the Post spend on polls be better spent elsewhere?

4. I am happy concede John Sides' empirical point that there's little evidence that politicians are "buffeted to and fro by the winds of the capricious public," as expressed in polls. I am also willing to concede that politicians do in fact use polls "to find out how best to sell their ideas." But I'm not sure what this proves. If politicians become more adept at selling their ideas to the public, does that benefit the public? That sounds like a question with a complicated answer. On the other hand, I feel reasonably confident saying that it is not the purpose of the media, at least as traditionally understood, to provide a consulting service by which politicians can hone their rhetorical skills.  

5. John Sides also has a lot of helpful stuff to say about information cascades, but I'm not sure what to make of his conclusion -- namely, that "the role of polls [in influencing future polls and outcomes] is unclear." I agree that it would be difficult to prove how a poll in the present affects a poll in the future. (Or at least I'm not sure how you'd ever tease causation away from correlation.) But it would surprise me if they did not. We all have preferences for all sorts of things that are popular -- neighborhoods, fashions, music and so on -- which will in turn make them more popular. These preferences are endogenous or interdependent or whatever you want to call them. I think it would be strange to discover that political preferences are any different.

(A sidenote: One quality of many situations in which choices are interdependent -- in which my choice depends on your choice and your choice depends on Michael's choice so forth -- is that they have multiple equilibria. The mere fact that we all end up living in neighborhood x or drinking at bar y or supporting candidtate z is not evidence that we are best off doing so.)

6. On the affirmative case for polls: Sides and Blumenthal both point to the value of accountability. in Sides' words, "we should know whether policymakers and policy are acting in accord with public opinion." Ack. I feel a bit dense saying this, but I really don't understand why we should want to know this at all. I am intensely interested in whether or not policy is acting in accord with my preferences. And so, presumably, is everyone else. But I can't figure out why any individual should care about how well policies line up with aggregate preferences. (John or Mark [or someone!]: please be patient and explain this to me once more.)

(One possible exception to my above confusion is this: I might have an individual preference for policies that line up with aggregate preferences. But, you know, I don't. And I don't think most people do, either.)

7. A related point: I am happy to concede, as per #4, that politicians might be influenced by polls. But even if you could prove that polls helped bring political behavior more in line with public opinion, I wouldn't totally know if that were a good or a bad thing. The wooly-headed paternalist in me thinks that the public doesn't always know what's best for itself. And the hard-headed-fan-of-The-Myth-of-the-Rational-Voter in me thinks that the incentive structure in democracy leaves something to be desired. It seems to me that the best case for democracy is a moral one ("people deserve a system that takes their preferences seriously") and not an instrumental one ("democracy will produce the best policies"). So while I'm happy to celebrate elections as a check on political power and a moment of collective expression, I'm less eager to celebrate polls as a tool for relentlessly enforcing the public will. (Of course, the fact that the two sides of the previous sentence are in obvious tension makes me realize that I need to think about this more.)

8. I think the killer point that John and Ed and Mark all make is that there's really no reason for me to believe the world would be better off -- less misleading, more responsive to "actual" preferences, whatever -- if we got rid of polls. There's lots of other disingenuous noise floating around out there. Which is probably true. But I guess I reserve the right to complain about that stuff, too.

Gaak, this was a lot longer than I thought it would be when I started writing.

There were two interesting pieces this morning on the question of whether or not we should tax employer-sponsored insurance (ESI) benefits -- they are currently exempt from income and payroll taxes -- to help pay for universal health care. The first is a great New York Times column from David Leonhardt; the second is a piece in the Washington Post that offers some reasons to be skeptical of the claim that taxing health benefits will lower spending. I've written before about why I think scrapping or capping the tax exemption is a wonderful idea (I made the longest version of the case here) and don't want to belabor those points too much. But I do want to wonder aloud about these paragraphs from the Post piece:

But even as consensus grows, others warn that the effect of taxing health benefits is being greatly overstated. They concede there is a case for limiting the tax exemption -- to raise money for universal coverage, as well as to equalize the status between employer benefits and individual coverage bought with after-tax dollars -- but say that move is not some kind of golden key for bringing spending under control.

[...] Steven Kreisberg, director of collective bargaining for the American Federation of State, County and Municipal Employees, was even more derisive. "Right. I have great chemo coverage so I think I'm going to go get cancer," he said. "This all starts with the economists -- they just apply traditional economic theory that if you give something preferable treatment in the tax code you'll get more of it. But health care is not a traditional good or service. It's a vital service."

Hmmm. This makes me think that it might be worth distinguishing between two claims that someone who supports changing the tax treatment might make. The first is that the preferential tax treatment is one of several things that causes people to overconsume health services and makes health care more expensive. A second claim might be that the tax treatment is the only cause of higher health spending, and that closing the loophole will be the silver bullet, "the golden key," the Pick Your Metahphor.

The Post piece starts with the assumption that the second claim is on the table. But look, no one is actually making the second claim. The second claim is crazy. The first claim, on the other hand, is obviously correct. If you get rid of a giant tax subsidy for health care, people will consume less in the way of health care. We can debate the degree to which limiting or eliminating that subsidy might bring spending under control, but there really isn't any debate that it will do something. Or at least there shouldn't be.

(A final note: The Post piece has some discussion of other factors that might shape higher spending -- pressure from producers, lack of information among consumers, etc. I think it's worth observing that these factors are also influenced by the tax treatment, which encourages consumers to be less information savvy and encourages consumers to develop high cost treatments that they know will be consumed with tax free dollars.)    

This is the third and final installment of my interview with Kenneth Arrow. Part one is here and part two is here. In the section below we discuss climate change: Dr. Arrow was a lead author on several of the Intergovernmental Panel on Climate Change reports, which are considered one of the most authoritative sources for estimating climate impacts. My questions are in bold.
Read More

[UPDATE: part three is here.]

Part one is here. There is one more part to come, which I'll post tomorrow. Part two, which is pasted below, is mostly about health economics. Arrow wrote the definite modern health economics paper -- "Uncertainty and the Welfare Economics of Medical Care"; I believe the PDF is here -- which is required reading for every student of the subject. My questions are in bold. Read More

[UPDATE: part two is here and part three is here.]

I spoke with Kenneth Arrow, Stanford Professor and Nobel Prize-winning economist, for about an hour last week. Dr. Arrow is perhaps most famous for the eponymous Arrow's Impossibility Theorem, one of the cooler ideas in public choice economics. But his work is wide ranging and we talked mostly about other topics. (I also couldn't think of anything to ask about the Impossibility Theorem.)

I have divided this interview into three parts. The first, published below, is mostly about the state of modern economics, the concept of general equilibrium analysis, and behavioral economics. The second part, which I will publish tomorrow morning, is about health-care economics. (In 1963 Dr. Arrow wrote what is probably the seminal paper in modern health economics, and very relevant to the current debate.) The third part, which I will publish as soon as I finish the editing, is about climate change. (Dr. Arrow was also a lead author on several of the Intergovernmental Panel on Climate Change reports.) My questions are in bold.
Read More

I should have mentioned this a few days ago, but I forgot. Anyway, as the headline of this post suggests, I'm guestblogging for Andrew Sullivan this week. I have a couple more interviews to grind through (I'm talking to Kenneth Arrow later this afternoon) and I'll post those here, but all of my other blogging will be on the Dish.

But I will note that the Dish doesn't have a comment section, so if I write anything stupid over there please feel free to leave any and all mean comments here.

[Update: Part two of the interview is now posted here]

Over the weekend I spoke with Thomas Schelling, who won the 2005 Nobel Prize in Economics for his work on game theory and collective bargaining. Schelling's early work was about war and arms control, but I wanted to speak with him about a different collective bargaining problem that's been in the news: Global climate change.

Climate change presents the world with an incredibly complicated bargaining arrangement. The big costs will be born by future generations, and born more heavily by countries other than the United States. The responsiblity for those costs is also distributed unevenly around the wrold, and there is no enforcement mechanism for a global agreement. And underscoring all of this is a great deal of uncertainty about what will happen 50, 100, or 500 years in the future.

I was hoping Schelling could help me make sense of this. Part one of our conversation is below. We discussed the financial crisis first, then moved on to discuss Waxman-Markey and the upcoming Climate Change Conference in Copenahgen. I will publish part two, in which we discuss international collective bargaining, tomorrow morning. My questions are in bold.
Read More

Will Wilkinson, Megan McArdle, Dave Weigel and I had a laconic little debate on Twitter yesterday about the USA Today article on the politicis of stimulus spending. Painful though it is to admit, 140 characters proved limiting! So we transferred over to the old-fashioned blog world, where Will commened on my last post:  

I think you're right that the article does not really establish what it implies, which would require comparison to the Republican counterfactual, but what it implies is almost certainly true. If it isn't, and Obama failed to get more for districts with high concentrations of poor voters than these districts would have gotten under McCain -- even with the blank check of a big stimulus during a perceived economic crisis -- then he may be the most ineffectual Democratic president ever, and progressives ought to be pissed.

That's fair, but I think the claim that what the article implies "is almost certainly true" depends on what you think the article is implying. The controversial implication -- picked up by a good number of big conservative blogs -- is that Obama is handing out stimulus dollars as a reward for voting for him. That strikes me as intuitively false, but pretty much impossible to prove or falsify. (After all, it's a claim about Obama's intentions: how on earth would you know one way or the other?)

But I agree with Will: What is almost certainly true -- or just true -- is that counties that voted for Obama get more money than counties that voted for McCain. But my problem with the article is that I find this fact totally uncontroversial and unsurprising. You "almost certainly" could have figured this out back in February, when the stimulus bill was passed and signed. The reason: If you have a piece of legislation that distributes dollars through Medicaid or TANF or whatever, those dollars will go to low income-voters who support candidates who, in turn, support redistributive social services. After I wrote the post yesterday a couple of people referred me to some old work by Andrew Gelman, and I think this chart sums up the point ("less wealthy voters across all categories support Democrats") quite nicely:

This isn't a coincidence, much less a conspiracy! In all likelihood, this is just "democracy": voters support candidates that will advance their interests. It is not always the case that the interests of the voters are well served. But there's no need to get alarmed when they are.
 

The Recovery Act The New York Times says that doubts are rising about Obama's stimulus plan, and raps the states on the knuckles for spending a lot of that money on the rural areas rather than cities. James Pethokoukis and the The National Review (!) say enough funds aren't being used for infrastructure and job creation projects. And USA Today says the money that is being used for infrastructure projects isn't being used properly.

I find these complaints pretty strange, since (1) we all knew the original bill contained a relatively small amount of infrastructure spending; and (2) all of the state allocation decisions are made by the states themselves. (It's fun to blame Obama for some stuff, sure, but not decisions over which he had no control!) More importantly, one of the points made by the Government Accountability Office's stimulus report (which came out this week, and which most of the above critics cite) is that state highway spending -- the biggest infrastructure chunk -- is ahead of schedule.

The GAO looked at 16 states, and all of them were ahead of schedule (click for a full version of this chart):
 

Let me leap to the defense of my corporate slavedrivers at the Atlantic Media Company and say that I find this piece by Slate's Jack Shafer pretty thoroughly unconvincing. I'm biased, of course. But can Shafer really believe that "Every new off-the-record venue drives a measurable quantity of political discourse out of the public sphere and into the private"? If that's really a problem, then he has a lot more to complain about than the Atlantic Media Salons!

Fortunately, the quantity of political discourse isn't fixed or zero-sum. The potential public benefits of off-the-record discussions are many (as Shafer must know) and it is obviously not that case (as Shafer suggests) that each additional minute spent off the record leads to a one-for-one drop in the amount of time spent on the record.

That said, I do think there is a different issue here. My problem with the original with the Washington Post salons was that the business side was promising that the editorial side would operate under some constraints. The Atlantic Salons, to the best of my knowledge, don't have that problem. (I think the critical sentence from Atlantic owner David Bradley's memo on the subject is this one: "There is no constraint placed on either the moderator or our guests as to the questions raised or the opinions expressed.") But a good question, acknowledged by David, is whether this can be achieved in practice. And I think Slate Group editor Jacob Weisberg was right to bop me on the head for missing this point: "invited to a nice dinner at the publisher's house, journalists don't need to be told to behave."

(Also: I agree with what Megan McArdle writes here. I wrote this post and then read hers, only to realize she covered a lot of the same in-defense-of-the-corporate-masters territory.)


Photo: I can't find any pictures of the redesigned Atlantic on Wikimedia!

Via Catherine Rampell, I see the new economics foundation (hip lowercase in the original) has released its second Happy Planet Index. The index attempts to rank the world on the basis of how efficiently countries produce happy lives -- that is, lifespan and level of satisfaction divided by rate of resource consumption. Costa Rica, the Dominican Republic and Jamaica come out on top.

Whether this is a meaningful measure is an open question. Measuring happiness raises philosophical and practical questions that are perhaps best handled with a dorm room and a bong, not a think tank. Nonetheless, I sympathize with the NEF"s -- excuse me, the nef's -- impulse. Gross Domestic Product is in many ways an unsatisfying measure of national wellbeing, and it seems intuitively true to me that a very high level of consumption is neither a necessary nor a sufficient condition for happiness (whatever it is).

On the other hand, what the Happy Planet Index makes clear is that some basic level of development is necessary to have a satifying life. Here, for example, is there map of the world based on the level of self-reported happiness. (The self-reporting is done on a 1-10 scale, in response to the question, "All things considered, how satisfied are you with your life as a whole these days?")

I think it's prretty impressive how well this would line up with most basic human development indices, at least at the lower end of the scale. It might be academically interesting that the United States falls into the same happiness category as Brazil or Saudi Arabia, but it's sad and not at all surprising that Sudan, Chad and Zimbabwe are just incredibly miserable and unhappy places to spend a life. It turns out that being dirt poor does not make people happy.


Photo: The name of this is "Free Sad Dirty Abandoned Child Creative Commons." Really!

I have been meaning to write something about this ever since Jim Manzi linked to Indur Goklany's Cato paper climate change a couple weeks ago and wrote that "before anybody gets on a high horse about how CO2-laden economic development is such a threat to the poor of the developing world, he really ought to have a response to this analysis." So, in the interest of scrambling back atop the high horse, let me say a couple of things about Goklany's analysis (PDF).
Read More

Jeez, the number of theories grows by the hour! Who dares pick now? If there really is an indictment in the offing, I fear that breathless speculation won't nudge it along any faster.

But what I do find genuinely intriguing is what the Palin diehards have been saying. You have, for example, the good people of Conservatives 4 Palin comparing the Alaska governor's resignation to Julius Caesar. And you have Bill Kristol speculating that her gamble could turn out to be quite shrewd. This strikes me as sociologically interesting. If her resignation -- which was a really a halting ramblefest -- doesn't shake the faith, what will?

Picture: somewhere, for sure, from WIkimedia Commons. And for more there is always Palin's Twitter feed.

That was one of the questions that came up when my friend Conor Friedersdorf and I did an episode of blogging heads TV yesterday. More generally, the question was: Why should we care when politicians have affairs? And why should journalists spend so much time writing about them?

Because of the common human aversion to hearing one's own voice, I haven't watched the BHTV episode. But the position I remember taking is basically this: There is little reason to care about a politician's sex life per se, except insofar as (1) he might have used campaign funds (in the case of Edwards) or state funds (in the case of Mark Sanford) to carry out the affair; or (2) his or her sex life implies faults that might bleed into areas of legitimate public policy (e.g., exploitation of an employee); or (3) the politician has broken a law to hide a dalliance (e.g. perjury).** Otherwise, I think it is fair to draw a line between public and private lives, even for public figures.

But I think there is one big exception to this: When politicians pretend that a certain kind of family life is a qualification for office. In that case, I think, they are shifting the standard by which they can be judged. (That's why I thought Sarah Palin's request that the media not cover her family rang largely hollow: She repeatedly brought up her family as a reason why you should vote for her!) If a politician runs for office on the basis of his or her private life, they're either pleading for a double standard, or pleading for it not to be private.


**I guess I have some slightly mixed feelings about # 3. Basic ethical conundrum: Do you have a moral obligation to give legitimate answers to illegitimate questions?

Photo: I used this one mostly because it was the easiest thing to find on Wikimedia.

It was fast. Very fast.

At 8.04am, Politico's Mike Allen publishes an article: "For $25,000 to $250,000, The Washington Post is offering lobbyists and association executives off-the-record, nonconfrontational access to "those powerful few" -- Obama administration officials, members of Congress, and the paper's own reporters and editors."

At 10.33am, Washington Post editor Marcus Brauchli sends out an email:

A flyer was distributed this week offering an "underwriting opportunity" for a dinner on health-care reform, in which the news department had been asked to participate.

The language in the flyer and the description of the event preclude our participation.

We will not participate in events where promises are made that in exchange for money The Post will offer access to newsroom personnel or will refrain from confrontational questioning. Our independence from advertisers or sponsors is inviolable.

There is a long tradition of news organizations hosting conferences and events, and we believe The Post, including the newsroom, can do these things in ways that are consistent with our values.

I think it's worth observing that what Brauchli is objecting to is not the idea of having corporate underwriting for an event with reporters or administration officials. (Those sort of events happen all the time; indeed, the Atlantic Media Company does them.) The editorial side can host an event and the business side can sell advertising for it, all while theoretically maintaining the traditional divide between the two. What is (or was, at this point) problematic about the Post situation is that the business side seemed to promising that the editorial side would compromise itself by hosting an event that was closed, off the record and non-confrontational.


Photo: Even a small child can see what's wrong with the Washington Post, or something.

Paul Krugman says he's been getting hatemail calling him the new Thomas Malthus. (Which actually strikes me as pretty thoughtful, high-minded hatemail.) Paul responds by pointing out that Thomas Malthus was actually right for just about all of human history, and reprints a graph from Brad DeLong to prove it.

I have slightly different graph on this point, drawn from Gregory Clark's fantastic A Farewell To Alms. (IMHO the book really is worth a read.) The graph is, quite simply, the economic history of the entire world:

For pretty much all of human history, population growth constrained growth in real standards of living. (That's the "Malthusian Trap" above: as standards of living improved, population increased, which put a strain on resources and drove down standards of living, which in turn drove down population growth, rinse & repeat.) The industrial revolution broke this trap, although it's worth pointing out the fairly obvious fact that this is not true for the entire world -- which is why the graph is labeled the "Great Divergence" and not the "Unmitigated Triumph."

An interesting question about the history of economics is whether (and why) we should continue to assume the kind of rapid growth that has characterized western economies since 1800. It hasn't been around forever.