TARP, the 2008 bank bailout, is an unlovely thing, defended by its supporters less on the merits than on the grounds that it prevented things from being much worse than they were. Ben Bernanke, for example, has been quoted as saying TARP helped prevent a second Great Depression. Those irate citizens who oppose the re-election of anyone who voted for the TARP bill must have a different idea from Bernanke’s about what would have happened if it hadn’t passed. But it’s hard to find any sort of account of just what alternative reality they envision.
Notice that the real debate here is not about what happened in the wake of the TARP bill—everybody agrees that the economy is still quite crummy—but about what would have happened without TARP. In other words, it is an argument about “what ifs”—or, as social scientists call them, “counterfactuals.” Of course, you never get to observe a counterfactual, essentially by definition: It’s what didn’t happen. The best you can do is to approximate that alternative reality, generally by trying to identify cases that are similar in relevant ways—as in pointing out that in the 1930s, Hoover let the banks fail, and look what happened. Then the argument shifts from being about TARP as such, to being about whether that case really is relevant, and whether it was in fact the preventable bank failures that caused the Great Depression. The economists are just warming up, and everybody else is looking for the exits.
This thing about counterfactuals is part of what drives many people nuts about social scientists. We don’t just stick to reality; we’re always insisting on talking about what ifs, and using theory and models to decide which examples and analogies are really relevant. Just stick to the facts, people want to say.
And yet those who criticize TARP are also playing “what if”: “If we hadn’t passed TARP things could hardly be worse than they are now—just look at how high unemployment is”. But there is that word “if”—if we hadn’t passed TARP, maybe unemployment would now be 15 percent or maybe in the view of some it would now be 5 percent. We can’t ever know for sure, and that’s a reality many people hate.
Now we know this blog is supposed to be about higher education, not finance (fortunately for us!). But this problem about counterfactuals is everywhere in policy discussions about higher education. What if public tuition rates hadn’t gone up so fast? What if a lot more low-income people had entered college over the last decade? What if a full-time bachelors’ program only lasted three years?
You can’t answer those questions just by looking at what really happened. The only way to approach them is by putting forward theories and models, and looking as carefully as possible for cases that look relevant, that give you some evidence about what would have happened. Even a “gold standard” random assignment experiment, like a drug trial, can only tell you that the drug worked for people who are a lot like you—but of course none of those people are exactly like you.
Analogical thinking, closely related to counterfactual analysis, of course abounds in higher education. Is borrowing to pay for college more like borrowing to start a business or more like borrowing to buy a speedboat? Are grants to students to help pay their living expenses more like welfare or more like hiring someone to do a piece of work? Is getting more people through college now like getting more people through high school was for the U.S. in the early 20th century?
Arguments from analogy and about counterfactuals go best when people on both sides spell out their theories and models and say explicitly what makes their comparative cases good—that is, relevant—ones. Too often, though, we find that one side of a disagreement about counterfactuals spells out their theories and models, which of course are always vulnerable to doubt, while the other side is content to just let the facts “speak for themselves,” which of course they never do.
You can’t responsibly say that the President’s stimulus package failed unless you are willing to stick your neck out and explain what you think would have happened without it, and why we should accept your judgment. Just saying it couldn’t be worse than it is doesn’t cut it—obviously things aren’t as bad as they could possibly be.
So, in our world of higher education:
- If you think people borrow too much for college, you need to say what would happen if they borrowed less, and defend your prediction;
-If you think too many people are going to college, you need to tell us what would happen to those people if they didn’t go—and back your claim up with evidence;
- If you think colleges cost too much, you need to tell us what they should do more cheaply, and what would happen if they did that; and
- if you want to spend more money on student aid, you need to tell us where the money will come from, and how it will change the lives of those who get it and those who lose out—and you need to give us reasons to believe you.
If people on both sides of controversial questions spent more time spelling out their counterfactuals and evidence, and less time simply expressing their opinions, we certainly wouldn’t see an end to disagreements but surely we could have more productive discussions.
3 Responses to Contrary to Fact
pncourant - September 3, 2010 at 8:06 am
Your argument, of course, is exactly what we should be teaching in college. I would add only one small amendment to this excellent piece. Sometimes people are not simply expressing their opinions without evidence, counterfactuals and models. They are, rather, deliberately misapplyng the the evidence in support of their own financial or political interests. Much of the argument in opposition to the efficacy of the stimulus package would seem to be of exaclty this form.
dank48 - September 3, 2010 at 9:05 am
Not to forget the Nirvana Approach, “term used by American economist Reuben Kessel for the logical error, common in everyday economic debate, of comparing an imperfect economic policy in practice with a perfect policy in hypothesis to the detriment of the former, i.e. comparing a known practice with an unknown ideal. In a sense this error is unavoidable, for, when any new policy is proposed in order to cure the faults discovered in an existing economic technique, it cannot be anticipated what faults the alternative policy my in turn develop. But the term serves as a warning not to conclude that a new policy whose faults are not known is necessarily superior to a policy whose faults are known.” (Seldon and Pennance, Everyman’s Dictionary of Economics)I don’t know that “misapplying the evidence” out of economic or political self-interest is the only possible, or even most likely, alternative to unsubstantiated opinion. It’s actually possible for people to disagree with, say, me, for example, without my having to resort to selfishness as an explanation for their failure to comprehend the brilliance of my position.
eslombard - September 5, 2010 at 2:42 am
I doubt that I am saying anything unknown to our readers, but….The student faculty ratio in 1940 at U of Pennsylvania was 1:7; today it is 1:6. In 1940, the tuition there was $400 a year; at UCLA in 1941, it was $29. a semester; at USC in 1942, it was $300. a year. Incidentally, we used to speak at Pennsylvania then of a “gentleman’s D”, but I thought then and still think that it was an absolutely marvelous education. Yes, inflation, but…There are all the jusifications for labs, etc. I would like to believe that big business is meeting all those additional expenses to the university and society, but probably not. It must fall into the category of what we have endowed banks and big business with, a form of corporate socialism.One in seven high school graduates went to college in 1940. The high school graduation percentages were less than we have today, but my reckless estimate would be that a high school diploma then signified more than the average diploma does today. A factor that should not be ignored among others is that too many students and their families have perceived a social distinction conferred by college attendance, but alas, their students’ mediocrity must somehow be served. Parents want to postpone marriage as long as decent and hope for the most promising social climbing for their offspring. The unions are only too happy to keep young people out of the job market as long as possible. Employers are only too happy to have the universities give future employees a bit of polish that the employers should furnish themselves. The increased numbers of graduates whet the appetites of employers who are frequently ready to select candidates with the most college; sometimes, it’s a matter of which PhD to hire for a position that requires nothing of the kind. We’ve conspired to create an inflation in available manpower with degrees of doubtful value in the workplace. It’s curious that the majority of undergraduates never work in their major fields once out of college. I have come to esteem former students who did not go on to or finish college (like Bill Gates)but found their own unique paths. I’m left with an impression from 1870 letters written by Americans and others who did not go to school past the age of 14, but who wrote marvelously subtle and thoughtful letters,something that I have rarely found among my own students at whatever level. It is easy to say, but somehow something is missing be it from our families or the schools and maybe changes in our culture. Finally, I do note what appears to be increasingly top heavy college administrations. It’s difficult to determine if they are the cause or the effect of higher education as big business. Surely someone must have written a book or dissertation on the subject.