"Correspondence, Abstraction, and Realism"

Philosopher of social science Daniel Little on "realists" versus "instrumentalists":

Correspondence, abstraction, and realism: Science is generally concerned with two central semantic features of theories: truth of theoretical hypotheses and reliability of observational predictions. ... Truth involves a correspondence between hypothesis and the world; while predictions involve statements about the observable future behavior of a real system. Science is also concerned with epistemic values: warrant and justification. The warrant of a hypothesis is a measure of the degree to which available evidence permits us to conclude that the hypothesis is approximately true. A hypothesis may be true but unwarranted (that is, we may not have adequate evidence available to permit confidence in the truth of the hypothesis). Likewise, however, a hypothesis may be false but warranted (that is, available evidence may make the hypothesis highly credible, while it is in fact false). And every science possesses a set of standards of hypothesis evaluation on the basis of which practitioners assess the credibility of their theories--for example, testability, success in prediction, inter-theoretical support, simplicity, and the like. ...

Whatever position we arrive at concerning the possible truth or falsity of a given economic hypothesis, it is plain that this cannot be understood as literal descriptive truth. Economic hypotheses are not offered as full and detailed representations of the underlying economic reality. For a hypothesis unavoidably involves abstraction, in at least two ways.

First, the hypothesis deliberately ignores some empirical characteristics and causal processes of the underlying economic reality. Just as a Newtonian hypothesis of the ballistics of projectiles ignores air resistance in order to focus on gravitational forces and the initial momentum of the projectile, so an economic hypothesis ignores differences in consumption behavior among members of functional defined income groups. Likewise, a hypothesis may abstract from regional or sectional differences in prices or wage rates within a national economy. Daniel Hausman provides an excellent discussion of the scope and limits of economic theories in The Inexact and Separate Science of Economics.

Another epistemically significant feature of social hypotheses is the difficulty of isolating causal factors in real social or economic systems. Hypotheses are generally subject to ceteris paribus conditions. Predictions and counterfactual assertions are advanced conditioned by the assumption that no other exogenous causal factors intervene... But if there are intervening causal factors, then the overall behavior of the system may be indeterminate. In some cases it is possible to specify particularly salient interfering causal factors (e.g. political instability). But it is often necessary to incorporate open-ended ceteris paribus conditions as well.

Finally, social theories and hypotheses unavoidably make simplifying or idealizing assumptions about the populations, properties, and processes that they describe. Consumers are represented as possessing consistent and complete preference rankings; firms are represented as making optimizing choices of products and technologies; product markets are assumed to function perfectly; and so on.

Given, then, that hypotheses abstract from reality, in what sense does it make sense to ask whether a hypothesis is true? We must distinguish between truth and completeness, to start with. To say that a description of a system is true is not to say that it is a complete description. (A complete description provides a specification of the value of all state variables for the system--that is, all variables that have a causal role in the functioning of the system.) The fact that hypotheses are abstractive demonstrates only that they are incomplete, not that they are false. A description of a hockey puck's trajectory on the ice that assumes a frictionless surface is a true account of some of the causal factors at work: the Newtonian mechanics of the system. The assumption that the surface of the ice is frictionless is false; but in this particular system the overall behavior of the system (with friction) is sufficiently close to the abstract hypothesis (because frictional forces are small relative to other forces affecting the puck). In this case, then, we can say two things: first, the Newtonian hypothesis is exactly true as a description of the forces it directly represents, and second, it is approximately true as a description of the system as a whole (because the forces it ignores are small).

This account takes a strongly realist position on social theory, in that it characterizes truth in terms of correspondence to unobservable entities, processes, or properties. The presumption here is that social systems generally--and economic systems in particular--have objective unobservable characteristics which it is the task of social science theory to identify. The realist position is commonly challenged by some economists, however. Milton Friedman's famous argument for an instrumentalist interpretation of economic theory (Essays in Positive Economics) is highly unconvincing in this context. The instrumentalist position maintains that it is a mistake to understand theories as referring to real unobservable entities. Instead, theories are simply ways of systematizing observable characteristics of the phenomena under study; the only purpose of scientific theory is to serve as an instrument for prediction. Along these lines, Friedman argues that the realism of economic premises is irrelevant to the warrant of an economic theory; all that matters is the overall predictive success of the theory. The instrumentalist approach to the interpretation of economic theory, then, is highly unpersuasive as an interpretation of the epistemic standing of economic hypotheses. Instead, the realist position appears to be inescapable: we are forced to treat general equilibrium theory as a substantive empirical hypothesis about the real workings of competitive market systems, and our confidence in general equilibrium hypotheses is limited by our confidence in the approximate truth of the general equilibrium theory.

Comments

[anne]

"Friedman argues that the realism of economic premises is irrelevant to the warrant of an economic theory; all that matters is the overall predictive success of the theory. The instrumentalist approach to the interpretation of economic theory, then, is highly unpersuasive as an interpretation of the epistemic standing of economic hypotheses. Instead, the realist position appears to be inescapable: we are forced to treat general equilibrium theory as a substantive empirical hypothesis about the real workings of competitive market systems, and our confidence in general equilibrium hypotheses is limited by our confidence in the approximate truth of the general equilibrium theory."

Perfect.

[Roger Chittum]

"Economic hypotheses are not offered as full and detailed representations of the underlying economic reality."
If only. In truth, mere hypotheses are offered all the time as full and detailed representations of the underlying economic reality--even often by reputable practitioners constrained by the duration of a sound bite, the need to provide a coherent textbook explanation, and/or by the desire to argue to a policy end rather than to analyze the situation completely. Referring to MT's post yesterday, when professional economists do this, there is a misrepresentation problem. Where's the accountability for it?

[Phillip Huggan]

I still don't see why you can't in theory have nominal interest rates being negative. Pulling the assets out of cirulation seems simple enough. Negative interest rates would cost money just like Keynesian spending does. In theory maybe it is good no one has tried negative interest rates because it is a tool that might lead to depression lots if misapplied.
In practise I like the way things are because there seems to be a chronic underinvestment of public spending in the USA economy and Keynesian spending seems slanted towards addressing these productivity/quality-of-living shortfalls. While an economy is continually being degraded by rich and stupid people lobbying to make themselves ever richer, it is probably best to have the Keynesian emergency exit.
But for instance, were the world's economy to resemble wherever Northern Europe is headed; were externalities like pollution, education, poverty, foreign aid/investment, employment, assured food and water supply, longevity, career mobility, free time...were these to be costed/recognized-legally, you may not want to change the composition of your economy towards public spending in a Keynesian way. Alternatively, you may have some sort of liquidity trap somehow in a planned/communist/heavily-public-mixed economy and not want to have added public sector investment.
In either scenario it would be useful to have at least the intellectual framework to push nominal interest rates negative, remembering in practise rich people might use the tool as an excuse to really get greedy and (more) evil.

If a country has an economy that maximizes quality-of-living, but faces externally caused liquidity trap or something, you'd want to keep things as they are by using negative interest rate monetary policy, not inflate some optimally performing private credit for sub-optimal public spending. Probably rich people would just misapply and ruin this 22nd century economic tool anyway, so maybe it deserves no further study.

[mrrunangun]

Same argument can be used to debunk global warming theory's claims to be science.

[Phillip Huggan]

Typos.
"In theory maybe it is good no one has tried negative interest rates because it is a tool that might lead to depression lots if misapplied."
meant to say negative interest rates would lead to subsequent inflation. Just imagine housing prices 2006 if nominal interest rates had gone to -5% post 9-11.

"...not inflate some optimally performing private credit for sub-optimal public spending"
Assume an "optimal" economy credit and cash allocation with external liquidity trap, it would be cash/loans that gets inflated away by using present Keynesian spending instead of future negative nominal interest rates, not credit.

I look at the macro viewpoint like this if economics ever gets really advanced and adds neocons to the commie dustbin:
1) if you have an optimal societal (100% of population plus future, not just 1% richest now) allocation of cash, use high interest rates and no Keynesian spending.
2) if you have an optimal allocation of credit, use negative interest rates and heavy Keynesian spending.

Right now USA is so underinvested in public spending (including curbing GHGs) and so subject to corporatism and so in debt that #2 minus the negative interest rates is fine. But in the future or in some over-commied country, you might want to try negative interest rates to stimulate private sector lending (probably is similiar to banking bailouts except it doesn't reward incompetant bank management) rather the Keynesian spending.

[Bruce Wilder]

I see Bruce's Thumbnail Epistemology is desperately needed:
Three kinds of knowledge:
1.) What might be: Theoretical, imaginative, analytical, categorical, conceptual -- the test is logical validity;
2.) What is: Empirical, experiential, positive, synthetic, factual, the particular, generalization, observation, measurement -- the test is pragmatic probability, guilt beyond a reasonable doubt, 95%, objective measurement;
3.) What ought to be: Normative, conative, morality and ethics, wisdom, aesthetics -- the tests are of meaning and satisfaction, taste, subjective evaluation.

Common sense notions of knowledge assume all three kinds of "knowing" combined together. But, getting from one to the other always involves an unsupported leap. The only proper test of a theory, qua theory, is, is it logically valid?

We need theories, because the key relationships in a functioning system -- good ol' "cause and effect" -- are not observable. No one has ever "observed" one thing "causing" another. We have to develop analytic theories -- ways of thinking -- that allow us to build a model of how a particular system is working. The useful "hypothesis" is an observable that allows us to distinguish one possible case from other possible cases.

Theories are useful, because they let us imagine what we can not observe, and because of that logical thing. A good theory eliminates the illogical cases, the cases that cannot possibly be. The only thing we "know" for certain about the world is that the world operates logically, which is another way of saying that it exists. So, we imagine worlds that might exist, worlds that operate logically, and try to sort which world we are actually in . . . which is the case.

"Correspondence" is not really possible between pieces of the world and theory. A theory is a logically valid way of thinking about what is possible, not about what is. A theory is about every conceivable case, (literally "conceivable"). Something "true in theory" is "true" (valid) in every case. Concepts do not have exceptions.

We apply theory to construct a model of a particular instant thing in the world, to understand how a particular system, a particular machine works. If we learn enough about enough instances, we might generalize, but generalizations, in contrast to concepts, just naturally have exceptions. Hypotheses test the correspondence of the model, not the theory, and only to what can be observed. And, the good stuff -- the "causal" relationships -- cannot be observed; that's why we need theory, to figure out logically what the "causal" relations must be. If we could observe causal relations, we wouldn't need theory or models, to figure out how the world works.

The metaphor of the correspondence between a map and some territory has considerable merit, but I think it is wrong to make of theory, mere description. Theory is about what might be -- that's what makes Huggan's negative interest rates kind of fun. Are negative interest rates possible? That's not the same as looking to see if they ever happen. Nor is logical speculation about negative interest rates "descriptive"? It is about finding order, and what kind of ordered relationship between concepts is implied by negative interest rates. It is about finding order in relationships, and description may follow of the ordered relationship, but the prize is not a description per se, but the discovery of an ordered relationship.

And, the "truth" of an hypothesis is of hardly any account, warranted or not. No one cares whether the hypothesis is "true". One tests hypotheses, to determine, which model better fits the case, which may well be partially an aesthetic judgement (knowledge category 3).

How should we make sense of Little saying, "hypotheses abstract from reality"? What the heck does that phrase mean? If anything, hypotheses abstract "to" reality, since an hypothesis is a statement, deduced from a logically constructed model, that can be construed as an assertion or prediction about what will be and can be observed in a particular case (aka reality).

Another Little developed queuing theory. The four Little Theorems don't "describe" standing in line. They identify the concept of a queue, and outline the relationship between rates of arrival, waiting and service. One could apply the theory (1) to analyze what goes on in the waiting room of an emergency room, by building a model, and using hypotheses to test how well the model "fit" the case. (2) And, then there would be the normative problem of developing some strategy of what to do: what's a good policy? (3)

Truth means different things in different categories of knowledge. In theory, "truth" is logical validity or analytic completeness in identifying necessary and sufficient. In interpreting evidence, "truth" is factual "probability". In choosing and in loving, "truth is beauty and beauty, truth" as the poet wrote.

Frankly, fuzziness worries me less, in social science, than overdetermination and the "causality" of intentional control. But, hey, that may be just me.

[a]

"Along these lines, Friedman argues that the realism of economic premises is irrelevant to the warrant of an economic theory; all that matters is the overall predictive success of the theory. The instrumentalist approach to the interpretation of economic theory, then, is highly unpersuasive as an interpretation of the epistemic standing of economic hypotheses. "

Yes - but wouldn't Friedman grant this? He's not trying to persuade us about the epistemic standing of the hypotheses; he's trying to say something about the epistemic standing of the theory's conclusions. They are true if they have overall predictive success, false otherwise.

[a]

This is slightly OT, but here's an honest question. Krugman has a model up in his latest post,

http://krugman.blogs.nytimes.com/2008/12/29/optimal-fiscal-policy-in-a-liquidity-trap-ultra-wonkish/

"What the model says: when monetary policy is up against the zero bound, the optimal fiscal policy is to expand government purchases enough to maintain full employment."

Does the model make a distinction between the *type* of government spending, i.e. between spending it on useful projects (e.g. building bridges that improve future productivity) to spending it on digging ditches and filling them in?

Thx.

[Greg Byshenk]

This is the sort of discussion that tends to waken my inner pragmatist.

Little writes that "This account takes a strongly realist position on social theory, in that it characterizes truth in terms of correspondence to unobservable entities, processes, or properties." But let us suppose that we choose to characterize truth in this way; what does this actually buy us?

Some would like to think that it gets us a "real" truth, corresponsding to something "real". But in fact, or in practice -- as opposed to the realm of pure supposition -- we find that such is not actually the case. For our "correspondence" is characterized as "correspondence to" something that is -- by definition! -- unobservable. Which means that we have -- in principle! -- no way of determining whether any such supposed "correspondence" actually holds in any case. Indeed, the only thing we can do is "systematize[] observable characteristics of the phenomena" and test our theories as "instrument[s] for prediction" -- which is precisely equivalent to the position of the instrumentalist.

In short, this sort of "realism" is ultimately empty. Yes, if we wish we may suppose that our theories "correspond to" something, but that matters not at all to the actual practice of theory creation or testing, because the supposed "correspondence" is one that cannot (by definition) be determined or even tested.

[wjd123]

To know the cause of something isn't to know how something functions in society. They are two seperate questions. I would suggest that before economist opine on the proper economic course for the future that they be able to answer both questions.

[Beezer]

I see Bruce's Thumbnail Epistemology is desperately needed:
Three kinds of knowledge:
1.) What might be: Theoretical, imaginative, analytical, categorical, conceptual -- the test is logical validity;
2.) What is: Empirical, experiential, positive, synthetic, factual, the particular, generalization, observation, measurement -- the test is pragmatic probability, guilt beyond a reasonable doubt, 95%, objective measurement;
3.) What ought to be: Normative, cognative, morality and ethics, wisdom, aesthetics -- the tests are of meaning and satisfaction, taste, subjective evaluation.

Boy, I'd like to see some studies estimating the cost of accomplishing #3. I don't believe there'd be much dissension about the desired results of an ethical, moral, wise society that's aesthetic to boot.

Sign me up.

[robertdfeinman]

I posted something along similar lines yesterday.

My thesis is that Malthus' ideas leaked into Darwin metaphorically which then ended up in Herbert Spencer and so on until we have the present situation where greed is not only natural, but good.

The full argument:

The Theory of Everything - Blame Darwin

I conclude that one can make as strong a case for altruism as a driving force in human nature and as a way to construct a society as one can the pseudo-Darwinian "dog-eat-dog" ideology.

[Roger Chittum]

I think I get it now. Still I’m concerned about the misrepresentation problem. Perhaps that would be resolved if social scientists preceded everything they said or wrote with a disclaimer along these lines.

WARNING! The following is a discussion of a virtual world that differs in very material ways from the real, physical world in which we live. Although I take this stuff very seriously and am trying to improve my skills—like players of Grand Theft Auto or Dungeons and Dragons—many of the things I say (I won’t tell you which ones) are not meant to describe or make predictions in the real world. Therefore, nothing I say should not be construed as advice for surviving and prospering in the real world or for establishing norms for the ordering of the real world. Basically, if you’re not a gamer like me, you should just ignore me.

[Bruce Wilder]

a, to answer your honest question, no, it doesn't matter. The gov't could build pyramids or break rocks. The problem is that the circular flow has spiraled down to a level, where resources are not fully employed, and there's no practical equilibrating mechanism to bring it back up; the objective of policy is to spiral the circular flow back up to full employment. Anything else is a bonus.

knzn considered this question in a blogpost titled Are Useful Projects Really Better than Useless Ones? If you prefer your Keynesian economics 80 proof and neat, and I do, knzn is your man.

[Bruce Wilder]

Beezer, "conative" is a good word. The noun form is "conation".

[stick]

Bruce has, again, hit the nail on the head. However, I would argue that all three "kinds of knowledge" are implicit in any social theory. Theorists are not blank slates, and the social theories they construct bear the mark of their normative values that, in turn, determine what empirical evidence is or is not relevant to their inquiries.

Where social science gets into trouble is in its embrace of positivist notions of science. The "laws of the market" are not of the same order as the law of gravity. Social theories are logically valid assumptions of social reality that guide human inquiry, and their validity is to be established in their ability to explain empirical reality and solve human problems. When social theories become ontology they are no longer theories... they are dogmatic assertions.

[a]

Bruce: my question was directed at the model, not reality. If the model doesn't make the difference (about the type of government spending, whether it's productive or not), then I would say so much the worse for the model, because in reality it surely (IMHO) makes a difference, at least in the medium-term. So it would seem that Krugman's is the case of a model which would fail the instrumentalist test.

[Phillip Huggan]

Yes the actual work done matters. Paying people to build wind turbines is better than paying the to cut themselves or rip apart garbage bags in the streets.
An employer hires someone when they can make a profit and the person's salary then spills over as he invests and consumes it unto the bank, baker, landlord, utility, etc.
You could just give an unemployed person cash to get the spillover and keep the rest of the gang employed, but you might as well earn the "public profit" while you are at it and get the spillover plus useful work. You are handicapped by not having market employer price signals, but there are enough market failures and public underinvestments like paying people to retrofit homes and build transit and rail lines and provide daycare...killing three birds with one stone by giving people money to keep gang employed and addressing underinvestment/market-failures and doing profitable/almost-profitable work as opposed to just keeping gang employed by paying someone to loiter.

[Lee A. Arnold]

I had a different line of thought on this. I'm not convinced that the "realism/instrumentalism" distinction is particularly apt in economics. Historically the instrumentalist position in the philosophy of science was suggested by physics in the fast-changing realm of the 20th century. But the physical sciences had other, static entities: materials and movements which remained constitutive of the subject, and of course are found repeatedly. Global warming is a case in point: to disprove it, you would have to disprove radiation physics and heat physics -- or else prove that the extra energy flux is immediately dissipated to space, and with no other effects upon Earth. But in economics there are no such hard, constitutive principles like "energy" to predominate in the contents of the discussion. The 19th-century hope was that "utility" might do the trick, but it's mostly ordinal and it is not unchanging. At this moment, all that remains of constitutive fundamentals in economics is something like the lesson: "incentives matter," --and of course they do, but again, incentives change all the time. They appear to be accountable by money, because money is accountable in an arithmetic sense, but otherwise it doesn't act like matter-energy: it can be created and duplicated and destroyed. Therefore money might count utility and incentives, but only so far in our experience as we find money is exchanged. (And some of those experiences are disputed on moral grounds.) No doubt economics has uncovered an invaluable amount of knowledge using accounting identities to find out aspects of why things have changed and moved around. Aspects but not the whole story. It still must find the fundamentals with which to map its explanations. Just so, "instrumentalism" as a buzzy idea can serve economics as a premise to create hypotheses. Indeed monetarism as proof of Miltonian "instrumentalism" came a Poppered balloon. But for instrumentalism to really be claimed in the philosophy of science of economics, you would first need precise, quantitative determinism that makes winning predictions over an aggregate of harder fundamentals.

Perhaps economics won't be strictly quantitative because it takes off into boolean realms. All relata are dyadic in the last reduction, and that basic split is triangulated by the purposive observer. Then the observer becomes a center, by drawing the relata into a useful array. Ideas, innovations, individuals and institutions are all centers making arrays, are all the same basic type of object: a one-to-many, many-to-one step of organization. They are continually involved in purposive behavior and are always doing some kind of economizing. (see: New Chart, for Descartes.)