What’s the harm in “what if?”?

October 24, 2011 in counterfactualizing

This morning in the graduate seminar we’ll be discussing Niall Ferugson’s Virtual History, which (per Andrew Gelman here and here) seems to me an agreeably rigorous thought experiment on the nature of causation in history. A cause is that x without which no y; to establish the causality of x one ought to be able to show that without it, no y, which means a comprehension of the counterfactual.

It strikes me as odd that some historians remain unhappy with this concept. Tristram Hunt suggested awhile back that this was because counterfactuals are inherently conservative; I think this is true only if you are wedded to a radical or progressive concept of history in which the end is foreordained. (See responses to Hunt here.) It is of course also possible that counterfactuals discomfit because, past a certain nearby point, there can be no evidence for them. Or perhaps for the same reason Gelman likes them—they are kin to a quantitative modeler’s mindset.

12 comments

October 24, 2011 at 9:06 am

Sandwichman

“Tristram Hunt suggested awhile back that this was because counterfactuals are inherently conservative;”

That’s silly. Here’s a counterfactual that isn’t inherently conservative:

A. C. Pigou succeeded Alfred Marshall in his chair at Cambridge in 1908. Pigou took academic economics in a more abstract theoretical direction, which substitutes “minute adjustments at the margin” for the laissez-faire “system of natural liberty”. The take away is the same: “this is the best of all possible worlds.”

S. J. Chapman, another of Marshall’s “favorite pupils”, was briefly a candidate for Marshall’s chair but withdrew. Chapman’s work was more “realistic” (similar to Institutionalism in the U.S.). Had Chapman become the leading academic economist in the U.K. rather than a career civil servant, a more Keynes-like approach to economic problems may have prevailed long before the Great Depression and may have developed the analytical tools and institutional innovations to eliminate poverty and establish a “co-operative commonwealth.”

October 24, 2011 at 9:29 am

jazzbumpa

I’d like to see a really good defense of counter-factuals. Seriously, I just don’t get it. They seem not inherently conservative, but inherently whatever-the-hell-you-want-them-to-be. You construct a counter-factual based on what wasn’t, and that leads to conclusions (?!?) about what might have been. How can there be rigor in such a process? How can one avoid arriving at self-serving, pre-conceived conclusions? (Or does one even want to?) Most fundamentally, what is the difference between a counterfactual and just making up some line of B.S.?

And note Sandwichman’s tentative conclusions above. How can you ever get to anything more definitive than a “. . . may have . . . ?” What real value does this bring?

And welcome back. I’ve missed the EOTAW crew.

Cheers!
JzB

October 24, 2011 at 11:00 am

Craig

“How can one avoid arriving at self-serving, pre-conceived conclusions? (Or does one even want to?)”

Can’t, don’t. In my view, that’s the principal harm: a logical one. “From a falsehood, everything follows”. Since everything follows, what people will select from that universe of falsehoods is exactly thing they want to select. Using deliberately false premises allows the historian (if historian she be) to engage in the perfect type of special pleading (since only a narrow range of falsehoods – tailored to the conclusion – then get into your analysis).

I’m not saying that an analysis of counterfactuals is fruitless; the consideration of the counterfactual is part and parcel of an analysis of historical causation. So what I mean is, if you construct narrow counterfactuals to try to test the knowledge you are constructing about how history hangs together, I think that’s a very good thing. If you construct broad, sweeping counterfactuals to try to establish new knowledge, you’re always going to run into the problem of its non-falsifiability. Far from a conservative approach, indeed, you have relativism run amok.

Counterfactual history is the opposite approach to that of the skeptic. It is (I think quite deliberately) intended to be provocative, fruitful and challenging. The problem is, the fruit comes from a poisoned tree. Rather than the bloody hard-won (and often thin and tenuous) truth of historical research, the “truth” of counterfactual history slops in great sploshes out of a rolling bucket of endless false (but oh so interesting!) possibilities.

To top it off, _Virtual History_ (like a lot of Ferguson’s own work) is a particularly bad book, because while much of the content is sweeping and simple, it doesn’t admit of its own limitations in that way. It’s a denial of the infernal and enraging complexity of reality.

October 24, 2011 at 12:26 pm

eric

Oh, that seems to me unfair, Craig. When you say, “the consideration of the counterfactual is part and parcel of an analysis of historical causation. So what I mean is, if you construct narrow counterfactuals to try to test the knowledge you are constructing about how history hangs together, I think that’s a very good thing,” you’re saying pretty much what Ferguson says in his 90-page introduction to the book: he sets forth the argument about causation, and he sets forth fairly constricted rules about how to construct a legitimate counterfactual. In fact when you say, “the “truth” of counterfactual history slops in great sploshes out of a rolling bucket of endless false (but oh so interesting!) possibilities,” you’re saying exactly what Ferguson argues against—he demands that alternative choices be documented in the actual historical record, and he asks an assessment of probabilities as against mere possibilities.

Now, if you wanted to argue that the actual counterfactual chapters do not obey Ferguson’s introductory strictures, that would be another thing.

And as for admitting of its own limitations, well. As Richard Hofstadter said, “if a new or heterodox idea is worth anything at all it is worth a forceful overstatement … this is one of the conditions of its being taken seriously.”

October 24, 2011 at 2:13 pm

Craig

Oh Eric, don’t get me wrong; I am not saying that Ferguson himself is guilty of the crimes, merely that the book is (and it clearly is in my view, although it’s been several years since I read it). At this point, I don’t even remember what work was his and what wasn’t.

But as for admitting of its own limitations? Look, if a method can’t be used to assert anything new, merely to test existing knowledge (and I very much believe that counterfactuals are only useful in that sense), and even then only in a limited way, then I think we’re fools if we don’t acknowledge the limitations of the method.

Finally, as to “an assessment of probabilities as against mere possibilities”, there’s Ferguson with his thumb on the scales again. No, I won’t have it. It’s not really history, although it may well be entertaining (_Virtual History_ mostly isn’t, it’s mostly dull as ditchwater), diverting and thought-provoking. But is it good for historians and future historians to evaluate such things? Most definitely!

October 24, 2011 at 3:56 pm

eric

It’s not really history…. But is it good for historians and future historians to evaluate such things? Most definitely!

I think we agree on this—as, I suspect, does Ferguson.

October 24, 2011 at 9:05 pm

jazzbumpa

I think I’m even more confused now.

he demands that alternative choices be documented in the actual historical record, and he asks an assessment of probabilities as against mere possibilities.

How can there be a historical record of things that never happened?

How does one assign probabilities? Where do history and statistical rigor interesect?

Cheers!
JzB

October 25, 2011 at 5:46 am

eric

How can there be a historical record of things that never happened?

You have to be able to document that alternative choices were considered at the time.

Jazzbumpa, your reading for this week is Ferguson’s introduction to Virtual History.

October 25, 2011 at 6:34 am

david

It’s the historical profession’s answer to “what would Courtney Love do?” I’m all for it.

October 25, 2011 at 7:25 am

JazzBumpa

Eric – I’ll take that on as soon as I can.

Meanwhile this strikes me a good use of counterfactuals:

Specifically, if demand was strong but businesses were concerned about future regulations, they would increase the hours of the workers they already employ rather than hiring additional workers. We have seen no evidence of this in the data: the average work week for private employees has been roughly flat for the past year. Similarly, if demand were strong, firms could easily expand using existing capacity without taking on the cost and risk of added capacity. However, the share of total potential industrial output in use remains 3 percent below its long-run average. Low capacity utilization is inconsistent with concerns about future regulatory risk, but aligns with weak demand holding back current production.

October 25, 2011 at 7:33 pm

andrew