Saturday, September 22, 2012

The 2012 presidential election and the problem of counterfactuals (continued)

My immediate impetus for writing this post was an intriguing passage in a recent National Journal piece by Ronald Brownstein with the attention-grabbing title "Heartland Monitor Poll: Obama Leads 50 Percent to 43 Percent".  But before I get to Brownstein's piece, to help explain why I found some parts of his analysis especially interesting, I'll have to provide a little background (and some of it will draw on things I've written before).  So please bear with me.

=> As I have noted in some previous postings (including this one), one of the more important factors shaping the outcome of this year's presidential  election is how voters deal with what analytical philosophers call the problem of "counterfactuals".  That is, in assessing situations, policies, and outcomes and making choices about them, we have to judge them not only in themselves but also, explicitly or implicitly, in comparison with what we think are the realistically available alternatives—including things that didn't happen (i.e., "counterfactual" outcomes) but that plausibly would or could have happened if some alternative course of action had been followed.

(If that sounds abstract, here's a concrete illustration from a non-political context:  Let's say a doctor tells you that, unless you have a painful and costly operation, there is a good chance you will lose a limb, or die, or experience some other unpleasant outcome.  Supposing you have the operation, and you don't die, but in the process you do spend a lot of money, experience a lot of pain, and take a while to recover.  If you want to decide whether or not the operation was worth it, it would be silly to focus exclusively on the empirically observable facts like the expense incurred and the pain of the operation.  Any sensible judgment also has to take into account the likely, but now counterfactual, outcome if you hadn't undergone the operation.  And it would also be silly to dismiss that factor by saying that we don't really know what the outcome would have been, minus the operation, since it didn't happen.  In the real world, a lot depends on whether or not you trust the doctor's diagnosis—which, in turn, depends on how much you trust the medical science the doctor is drawing on and his judgment in applying it to your case.  Can you ever be 100% sure which counterfactual is most accurate and appropriate, or which doctor you should place the most trust in?  Alas, no.  Does that mean you can or should simply ignore potential or retrospective counterfactuals in making decisions or assessing their consequences?  Only if you're determined to act stupidly and/or thoughtlessly.)

Getting back to this year's presidential election, it is generally acknowledged that Obama's chances for getting re-elected have been hurt by the fact that the economic situation is lousy, unemployment remains painfully and unacceptably high, the pace of recovery from the big crash of 2008 has been slow, etc.  So it's natural that a lot of Americans are disgruntled.  But how their unhappiness gets translated into voting depends, at least in part, on how they answer two interrelated questions.  First, who do they think deserves most of the blame for the situation?  And second, whether or not they are entirely happy with how well Obama and his administration have dealt with the country's economic problems, do they think that another (realistically available) president, pursuing different policies, would have produced better results so far and would produce better results in the future?

=>  The significance of that second question (or set of questions) has been highlighted especially sharply by public reaction to the 2009 economic "stimulus" (the American Recovery and Reinvestment Act). Ever since it was passed, against almost monolithic Republican opposition, Republican politicians and the right-wing propaganda machine have been relentlessly pushing the line that the Obama "stimulus" failed, produced no useful results, didn't save or generate any jobs, was just a waste of money, etc.  And public opinion surveys, as well as a lot of other evidence, indicate that a great many people believe them.

That conclusion may seem plausible, given that unemployment has remained stubbornly high, the economic recovery  has been disappointingly slow (due in no small degree to Republican sabotage and obstructionism, but that's a separate issue), and so on. But that conclusion, or impression, happens to be wrong, no matter how many people believe it.

Yes, it does seem clear in retrospect that the 2009 "stimulus" was  too small—especially since the economic crash of 2007-2009 turns out to have been more severe than most analysts recognized at the time—and in various ways it could have been better designed. And yes, the economic situation is still lousy for most people.  But the real question has to be posed a little differently:  What would have been the likely consequences if (counterfactually) no serious stimulus had been passed at all, and passed quickly? The answer, almost certainly, is that the results would have been disastrous.The self-reinforcing downward spiral of the economy would almost certainly have gone into free fall, official unemployment might well be closer to 15% than 9%, and instead of our current Great Recession we would probably have something more like another Great Depression. Instead, within a few months of the passage of the ARRA the economy stabilized and then gradually began to recover.

In this connection we can quote Mark Zandi, who is not a left-liberal neo-Keynesian but a quintessential mainstream economist who was an economic adviser for John McCain in 2008.
[W]e would be in a measurably worse place if not for the stimulus. I don’t think it is any coincidence that the great recession ended [i.e., the economy stopped contracting] at precisely the same time that the stimulus, and in this case when I say stimulus I am talking about the [American Recovery and Reinvestment Act] … was providing its maximum economic benefit.
 This graph captures one key aspect story. It shows the month-by-month record of net job gains and losses in the private sector from January 2008 through November 2011. Notice any difference between the first three months of 2009 (the first three blue lines) and the period since then?

In short, the 2009 economic "stimulus" worked—probably not as well as it could have, but definitely a lot better than the alternative option of not passing it at all, which is the appropriate criterion. In my (possibly fallible) opinion, the only people who try to pretend otherwise are either economic illiterates, or sincere but misguided economic cranks theologically committed to pre-Keynesian economic dogmas (who get showcased or channeled by propaganda outlets like the editorial page of the Wall Street Journal), or cynical partisan propagandists who don't really care about the facts one way or another.  (For some further elaboration, see here.) 

=> On the other hand, the Democrats, including Obama, have been remarkably ineffective in telling their side of this story.  And even when they have tried to make their case, a large proportion of the public hasn't found it plausible or convincing.

In some remarks that Barney Frank made at a Congressional hearing in July 2009 (reported by the Washington Post), he tried to capture some of the reasons why making this sort of case is difficult:
REP BARNEY FRANK: Not for the first time, as a -- a -- an elected official, I envy economists. Economists have available to them, in an analytical approach, the counterfactual. Economists can explain that a given decision was the best one that could be made, because they can show what would have happened in the counterfactual situation. They can contrast what happened to what would have happened.

No one has ever gotten reelected where the bumper sticker said, "It would have been worse without me." You probably can get tenure with that. But you can't win office. [....]
=>  Maybe, maybe not. There's clearly a lot of wisdom in Barney Frank's remarks. But some intriguing results from a few recent surveys may suggest that his formulation might have been a bit too pessimistic. (All those qualifiers are deliberate.)

Consider, for example, that recent piece by Ronald Brownstein in the National Journal with a title that Republicans should find a bit ominous: "Heartland Monitor Poll: Obama Leads 50 Percent to 43 Percent". (Brownstein is analyzing results from "the latest Allstate/National Journal Heartland Monitor Poll"; but, as he notes, these results "are in line with most other national surveys in recent days showing Obama establishing a measurable lead, including this week’s new Pew Research Center and NBC/Wall Street Journal polls." The whole piece is worth reading, but one passage has an especially direct bearing on the question of how voters are approaching the analysis of counterfactuals:
The survey also shows why it may be difficult for Republicans to center the election on the famous Ronald Reagan question to voters that the party highlighted at its national convention last month: Are you better off than you were four years ago?

That question divides likely voters almost exactly in thirds: in the poll, 31 percent say they are better off than four years ago, while 34 percent say they are worse off and 34 percent say they are about the same. Romney, predictably, wins more than four-fifths of voters who say they are worse off; the president, equally unsurprisingly, attracts almost nine in 10 of those who consider themselves better off.
Let me just interject that these dramatically high correlations are not as obviously predictable or self-explanatory as Brownstein seems to imply. For example, since the early months of Obama's presidency the stock market has rebounded dramatically, corporate profits are rising, and a lot of Wall Street financiers and corporate executives have been making out like bandits (I mean that mostly in a figurative sense, though not entirely). Yet all the available evidence suggests that many of them have turned sharply against Obama.

Actually, I suspect that in many cases the answers to those survey questions were influenced, at least partly, by a reverse logic. That is, if respondents favored Obama, they would be more likely to say they were better off, and vice-versa for people who would like to see Obama defeated. (I wouldn't describe all the people in the second category as Romney supporters, since it's unclear how many people positively favor Romney, as opposed to disliking Obama so much they are willing to vote for any Republican candidate who might defeat him.)  This is all speculative, of course, and it underlines the fact that one always has to approach survey results with caution. But be that as it may ...

... here are the most intriguing results.
Crucially, though, Obama holds a commanding 57 percent to 34 percent advantage among those who say their finances are unchanged. One reason for that critical tilt in his direction: Voters who say their finances are unchanged also say, by a resounding 53 percent to 33 percent margin, that they believe the country has been better off over these past four years because Obama, rather than another candidate, won in 2008.

Overall, 48 percent say they believe the country is better off because Obama won in 2008, while 41 percent say the nation would be in a stronger position today if another candidate had won.

In a related finding, 47 percent of likely voters said they believed Obama’s economic policies helped “avoid an even worse economic crisis and are laying the foundation for our eventual economic recovery.” By contrast, 45 percent said that his agenda has “run up a record federal deficit while failing to end the recession or slow the record pace of job losses.” That’s hardly a ringing endorsement and well within the survey’s margin of error — but it represents only the second time since the Heartland Monitor began asking that question in September 2009 that a plurality has attributed positive effects to Obama’s agenda.
Along with most serious economists, I happen to believe that the conclusion eventually reached by 47% of these respondents—i.e., that Obama's economic policies helped "avoid an even worse economic crisis"—is almost certainly correct. So this result might be interpreted, optimistically, as evidence that you can't fool all the people all of the time. (Only 45% of them.)

At all events, it would appear that at least part of the electorate can be convinced, as least part of the time, by the argument that "It would have been worse without me." (Of course, voters may be more receptive to that argument when a sizable chunk of the electorate has concluded that the actually available alternatives are definitely unacceptable.)  If so, that's good,  because this kind of argument often should be taken seriously in politics ... up to a point.

=> That's one interesting piece of the puzzle, but of course the final outcome remains to be seen. At the moment, the odds that Romney & Ryan will win the election do seem to be dropping like a stone. As Brownstein sums up the latest evidence:
Taken together, all of these small movements toward Obama have produced, at least for now, a tangible advantage for the president over Romney as the race hurtles toward its final weeks.
However, even if we overlook the uncertainties involved in interpreting polling data, there's still more than a month left before the election, and all sorts of things can happen, in the US and abroad, that could alter the situation in unpredictable ways. Also, a huge amount of money will get spent between now and November 6, and in some states the outcome may turn on the success of Republican efforts to discourage, intimidate, or simply disenfranchise potential Democratic voters—a factor which is, again, unpredictable. Stay tuned ...

—Jeff Weintraub