Monday, September 24, 2012

When the Positive to Normative Two-Step Goes Wrong: Scientists, Engineers, and Ideologues

. Monday, September 24, 2012

Noah Smith has a nice post running down the history of modern macroeconomics through the lens of Greg Mankiw's "scientists vs. engineers" taxonomy. This is another way of framing the Saltwater vs. Freshwater debate, but instead of calling it a "debate" Mankiw wants to make it a division of labor. Smith, however, is having none of it:
Note that if the microfoundations are misspecified, it doesn't matter whether a model satisfies the Lucas Critique or not. Even if "tastes and technology" really are policy-invariant things, if tastes and technology don't really work the way the model says they work, the model will not give you useful advice about policy
Now, terrible microfoundations might not lead to a terrible model. But if by some lucky happenstance, crappy microfoundations produce a model that matches the macro facts, then it might as well be one of those "aggregate-only" models that Greg Mankiw labels "engineering". In this case, RBC has no theoretical advantage over a New Keyneisan model.
Emphasis added. On another day I might have wanted to get involved in this discussion on its own terms, but for now I'm interested in something else: the shift from positive argument to normative argument. It's a subtle two-step, but it happens all the time in the social sciences. Here's how it generally works. Some social scientist will say: "Given that the world works according to mechanism X (as I've just demonstrated), we should do Y." Often the "we should do Y" is masked by caveats, hedges, and other devices which allow one to walk back the initial claim; nevertheless it is almost always there in some form in almost every piece of social science research. Or if not in every particular research article, in the paradigm within which the article operates.

The problem arises when when the positive argument -- the part that goes "Given that the world works according to mechanism X" -- isn't true. And by "not true" I don't mean "not strictly true but true enough as an approximation of reality that the model works out alright". By "not true" I mean sufficiently false that the model doesn't actually work out alright. In that situation, the positive argument goes away and the normative argument is all that remains. When people advance normative arguments as if they had a valid positive argument underlying them, but they do not actually have such a positive argument, they are not doing anything like science, as Smith rightly notes:
So basically, I charge that the New Classical/RBC/freshwater macroeconomics paradigm is not really science, and not really like science...not yet, anyway. In science, evidence rules all; if a model doesn't fit the evidence you toss it out.
Once again, I'm less interested in the particulars of Smith's case than about the phenomenon he's highlighting. To go back to Mankiw's taxonomy, if the "scientists" aren't doing science and they're not doing engineering either, then what are they doing? Restated, once the foundation of the positive to normative two-step crumbles, what intellectual project remains? It is ideology, and its practitioners are ideologues.

Take an example closer to (my) home. Stephen Walt writes a blog with the grandiloquent subtitle "A Realist in an Ideological Age". This is meant to suggest that Walt recognizes the world as it actually is whereas his opponents are in denial or a state of intellectual confusion. It is a claim that the positive argument is on his side -- the world works according to mechanism X, which in this case means that (all) states pursue their national interest above all else, especially security as measured by relative power differentials -- therefore his normative prescriptions (and perhaps his frequent proscriptions as well) -- that therefore the United States should do Y, which is to also maximize its national interest, especially security as measured by relative power differentials -- are sound wisdom. This attitude is made clear by the William Arthur Ward quote at the top of his page:
The pessimist complains about the wind; the optimist expects it to change; the realist adjusts the sails.
Walt is clear on which of these he believes himself to be. The realist cannot change the wind. The realist can only adjust to it. And in fact the realist only should adjust to it, since complaint is wasted effort and optimism is folly. Yet in almost every one of his blog posts Walt writes something like this:
Instead of harping on our "global responsibilities," Americans ought to focus instead on their national interests. The litmus test of any foreign policy commitment is not what it will do for others, but rather what it will do for us.
Walt has spilled quite a lot of ink describing how and why the United States has lost its way, has forgotten its national interest, and how this has caused a series of catastrophic mistakes in both foreign and domestic policy. Can you tell the problem?

The problem, simply, is that if the United States is not pursuing its national interest, as defined by Walt, then there is no reason to believe everyone else is either. And if everyone else is not, then the "wind" is not blowing in the direction that the realist believes it is: the positive argument that supposedly justifies the normative claim is false. Perhaps some conception of "national interest" other than maximizing one's relative power is in operation. Perhaps states pursue a range of interests at different times and in different places. Perhaps the entire notion of a singular "national interest" is misguided. Perhaps different groups or individuals use the power of the state to pursue group or individual interest. Whatever. Something about Walt's foundational positive argument is not correct. As a result his normative advice is a non sequitur.

When this happens the positive to normative two-step is reduced to just one step: make a normative claim, advance it whenever you can, and defend it at all costs. This is not science and it is not engineering; it is ideology. For the purpose of this post I do not claim that ideology is inferior to science or engineering, but I do claim that ideology which masquerades as science or engineering should be resisted in order to preserve the latter two categories as distinct and worthwhile.

This is something that social scientists struggle with, because it involves admitting that we could be wrong. About everything. Nobody likes being wrong but intellectuals hate it because our status and income is predicated on our not being wrong. So, sometimes, we transition from scientist or engineer to ideologue -- probably without realizing it -- and once we've gotten there it's very hard to find our way back. Eventually we end up advancing arguments that are contradictory in order to defend our ideology, rather than casting aside the ideology in favor of science or engineering, all the while continuing to claim the mantle of science as our own. This is a grave mistake.


Phil Arena said...

Excellent post.

I would say, "For that reason, I never say anything about what we *should* do at the end of my papers", but I probably have. It's awful hard to avoid.

What bothered me about the Smith post is that he seems to think that the mere fact that Homo Economicus is a simplification means that no one should even be interested in models that analyze the behavior of that mythical creature. You are more careful to say that the problem is not analyzing models that rest upon false assumptions (when Smith implied that true scientists don't tolerate false assumptions, my head nearly exploded) but offering policy advice that would only be sound if those assumptions did hold.

Good stuff.

LFC said...

"Something about Walt's foundational positive argument is not correct. As a result his normative advice is a non sequitur."

Yes, but his normative advice in a particular instance or instances could still be very sensible on its own terms.

The real problem, as you suggest though don't make completely explicit, is that Walt seems to think there is an objectively discoverable, singular 'national interest'. He would be much better off if he dropped that pose and the "a realist in an ideological age" slogan (I hate that too) and simply said: 'Look, here's what my view of the US national interest tells me would be wise policy in this instance.' In fact, this is very often, I think, what he does in any given post (though I don't read him every day by any means). The problem, again, is when he insists on claiming that he is the 'genuine' realist article, etc. In fact there are varieties of realism, varieties of conceptions of the natl interest, and Walt should just acknowledge that more straightforwardly than he sometimes does.

Although I don't fully understand the way some IR theorists sometimes use the word "reflexive," I think it often is intended to mean something like "self-reflective" or "self-aware", i.e. aware of the implications of one's own commitments and values and that they do influence not only (and obviously) one's normative views but also, to at least some extent, one's 'positive' analysis. So I think I might say Walt is insufficiently reflexive, though you would not use that language.

LFC said...

P.s. also I wd note that the word "ideology" need not have the negative connotation you give it in this post: i.e., the opposite of science, or opinion-masquerading-as-scientifically-based-when it-isn't.

"Ideology" can also have a more neutral meaning as "worldview", set of overarching or guiding beliefs etc. In this sense everyone has an ideology, including those political scientists who consider themselves non-ideological.

JR said...

I admit to not being well-versed in the current schisms in macro/micro economics, but don't freshwater types also trot out that argument against keynes/neo-keynesians? I saw a lot of that going around after the Sargent/Sims Nobel award last year. Not to mention Vernon Smith being a Cato fellow...

Kindred Winecoff said...

Phil, Yes I chose my language carefully, in part because I knew you'd be reading it and would come after me if I generalized too much.

LFC, you wrote "Yes, but his normative advice in a particular instance or instances could still be very sensible on its own terms."

Indeed, but then it needs to be justified on its own terms, rather than as an implication of a (false) positive claim. That's why I said in the post that I wasn't claiming that ideology was bad per se; merely that it's bad when it tries to hide the fact that it is ideology masquerading as "science".

I think the "national interest" idea is a problem -- I don't think there is such a thing -- but I don't think it's the only problem. That's because I don't think this problem is limited to realists, or to international relations, or to social sciences at all. The real problem is much broader.

JR, as I said in the post I'm less concerned about the specific internecine conflicts among macroeconomists. I'm sure some freshwater types accuse some saltwater types of this, and vice versa. I'm sure it is often merited. As I (tried) to argue, this is a very common thing. I'm sure I'm guilty of it myself sometimes. But if its true that "everyone does it" then that means it's an even bigger problem.

When the Positive to Normative Two-Step Goes Wrong: Scientists, Engineers, and Ideologues
There was an error in this gadget




Add to Technorati Favorites