I am starting to read up on the topic. I see epistemology as the attempt to articulate what criteria we are using to evaluate the usefulness of economic analysis. Various initial thoughts:
1. I emailed Pete Boettke for advice on what to read to get me started on the topic, and not surprisingly he had useful suggestions. I told him that I of course know about Milton Friedman’s classic position.
2. One can argue that there is no need for epistemology. You could just assume that good economics is what leading economists do. Without having to articulate what they do, just try to do things similarly. However, for someone like myself, who is inclined toward heterodoxy and to doubt that leading economists are doing useful economic analysis, that is not the right answer.
3. One of the articles that Boettke suggested was by Dan Hausman, who wrote
Instead of attempting to discover what methodology neoclassical economists actually practice and to think seriously about how that methodology might be justified, … critics … have usually relied on indefensible philosophical theories of science to support broad condemnations. … Philosophers have, however, little to offer by way of informative well-supported systematic theories of the scientific enterprise and that little does not lend itself to mechanical application.
In other words, if you have a problem with how economists are doing things, that is your problem, not the economists’. Those who can, do, and those who can’t, do epistemology.
To put it another way, I read Hausman as saying that the task of the epistemologist is to figure out what economists do and then justify it. Again, from where I sit as a heterodox economist, this is hardly satisfactory.
What I contend in my forthcoming book is that economic theories are interpretive frameworks. These cannot be tested decisively. They are not falsifiable in the Popperian sense. Think of AS-AD. There is no combination of output and price movements that can falsify it. If they move together, you call it a demand shock. If they move in opposite directions, you call it a supply shock.
What I do not address in the book is the question of how best to evaluate competing interpretive frameworks. I believe that it is an important question. I am not convinced that the best way to answer to the question is to study how frameworks become popular in the profession. I think that factors such as path dependence (model X got published in a good journal, so something that pertains to model X can also get published in a good journal) and ideological preference play outsized roles in such evaluations in practice.
Somewhat related: Noah Smith on objectivity and economics. Pointer from Mark Thoma.
You might find the philosopher of economics Alex Rosenberg’s work interesting, especially his book Economics: Mathematical Politics or Science of Diminishing Returns?
The article linked below (and several responses to it) was the most memorable article we read in my research methods course in grad school. Me and one other guy in the class took the author’s side while the professor and the rest of the class took the critics’ view.
I think the author does a great job critiquing Friedman and arguing for his alternative.
http://econpapers.repec.org/paper/agsumaesp/13880.htm
The problem with classical epistemology and its modern descendants is the search for silver-bullet answers to very hard problems.
In some ways, the answer is as simple as this: we must attend to the particular theory, argument, or framework, and evaluate it on the merits. “The merits” sounds vague, but really just means attending to the applicable questions, even just to raise them when it seems no one else has thought to. Is this unreasonable precise for the subject matter? Do we have evidence from history, studies, experiments, testimony, or the existence of consensus among specialists on the topic? Does the reasoning hold together or fall apart under scrutiny? No one of these concerns is a silver bullet, and ultimately it takes a leap of faith to come to a conclusion. But we do so all the time, with varying success.
The attempt to divine a method or procedure that will guarantee truth has failed time and again and is simply ill-advised. There are no shortcuts in scholarship and science.
Sounds a bit like a revisit to the “approach” taken by Mises.
Since they purport to describe reality, the key and minimal requirement is that distinct interpretive frameworks must be able to compete with each other by means of some substantial reality test, such that less useful or accurate frameworks gradually get driven out of the market not thorough popularity or prestige but by performance. It it both a sufficient and necessary condition for the possibility of meaningful epistemology. If you don’t have that, there is no point in even trying. If you do have that, what more do you need except to try and continually refine and improve the tests?
I agree, but the problem is leading economists and leading economic frameworks have really lousy predictive track records. Nobody’s model works very well “out of sample”.
It’s like medicine before the mid-19th century. Credentials, rituals, institutional bases, prestige, resources, and major interventions into organic processes, but to no obviously favorable result.
You can’t expect the guild to declare itself irrelevant via empirical analysis of itself.
Economist at their best are historians + math and at their worst are mystics+math, but would like the epistemic status of physicists or modern physicians. Your requirement will let the rest of us judge when they are achieving the latter.
“What I contend in my forthcoming book is that economic theories are interpretive frameworks. These cannot be tested decisively. They are not falsifiable in the Popperian sense. Think of AS-AD. There is no combination of output and price movements that can falsify it. If they move together, you call it a demand shock. If they move in opposite directions, you call it a supply shock.”
This is actually very Popperian. Popper called such theories ‘metaphysical research programmes’. ‘Metaphysical’, because they weren’t falsifiable in and of themselves, but rather offered a theoretical framework for constructing other falsifiable theories. Whenever such theories were falsified, the metaphysical research programme wouldn’t be held responsible, but always some other assumption. Because of this, MRPs have a tendency to become dogmas, so Popper was concerned with other ways to criticise an evaluate them.
You might be familiar with the controversy surrounding Popper’s claim that evolutionary theory wasn’t scientific. It’s occasionally brought up by creationists against atheists online. Anyway, Popper retracted his claim, but I guess that he might just have been placating his critics, because he was right. The basic idea of evolution is non-falsifiable; instead it provides a theoretical framework which, when applied to real historical events, can be used to construct falsifiable theories. In other words, evolutionary theory, at least in its basic form, is a metaphysical research programme.
Lakatos basically copied this idea and renamed is ‘scientific research programmes’, because marketing. He also added a Kuhnian-style paradigm-incommensurability angle which Popper would have disagreed with. Lakatos’s seemingly dogmatic ‘hardcore’ of a research programme are basically the unfalsifiable assumptions of the explanatory framework. Popper regard these as unfalsifiable and, therefore, unscientific, but not beyond rational debate and mutual criticism.
Interesting points.
I once read that Popper’s view on falsifiabiloty was that both acceptance and rejection of a theory were provisional. The common view we are taught in university courses is that only the acceptance is provisional. Rejection has an air of finality to it.
Do you know when/where/how this divergence started?
The common teaching of Popper is a complete mess. He’s mostly used a pedagogical foil in an apocryphal narrative, and so his actual views are perverted and twisted into a caricature so they can play the required role.
Suffice to say, Popper was a fallibilist through and through. The logic of falsifiability has a certain finality, because when expectations are frustrated, then something is wrong in the state of Denmark, so to speak. However, where fault lies for any apparent falsification is always a matter of human judgement, and even guesswork. There is nothing decisive or final about this, and so both acceptance and rejection of theories is always provisional.
No, the asymmetry makes perfect sense. Consider the hypothesis that an clock keeps good time. You then check the clock against the true time. If the clock is wrong, hypothesis rejected, nothing provisional about it. If the click is right, well, maybe it keeps good time, or maybe it’s stopped and you’ve just measured at the exact moment to produce a coincidence. Or maybe you didn’t measure precisely enough to detect a small error. You can provisionally accept the theory with these caveats, but no more.
That assumes you have sufficient quality in your data or that the data don’t change over time/place/etc.
Here’s Popper:
“In point of fact, no conclusive disproof of a theory can ever be produced; for it is always possible to say that the experimental results are not reliable, or that the discrepancies which are asserted to exist between the experimental results and the theory are only apparent and that they will disappear with the advance of our understanding. (In the struggle against Einstein, both these arguments were often used in support of Newtonian mechanics, and similar arguments abound in the field of social sciences.) If you insist on strict proof (or strict disproof) in the empirical sciences, you will never benefit from experience, and never learn from it how wrong you are.”
Popper “The Logic of Scientific Discovery” 1935 p. 28
It is plausible to understand “knowledge” as that “set of connections” amongst bits of perceived information which has (cognitive) meaning to the observer or intelligence receiving the perceptions.
In that case “knowledge” is formed (by cognition or intelligence) from perceptions of information. Perceptions vary; cognitive meanings vary.
Most knowledge is based on information and the “manner” of its perception, as well as its “meaning” to the observer.
[“Personal Knowledge” – Michael Polanyi ?]
What has been “formed” as knowledge in that way by one intelligence may then be perceived as simply information by another intelligence which observes other connections with other meanings.
“Intelligence” may be understood as a capacity to recognize those connections between bits of information – plus – the facilities for perceiving and their use.
Problems with some forms of “modeling” can stem from “syntheses” rather than actual perceptions of information, and “constructs ” of, rather than perceived, connections – and often, combinations of both deviations from reality.
To set out “looking for” what “must be perceived” in order to attain knowledge is likely to change the uses made of the capacities of perception and of the intelligence to recognize connections.
There may be multiple valid frameworks but they need to be able to formulate testable hypotheses and accommodate their success and failure and these frameworks will amount to different ways of saying the same things. If they can’t offer testable hypotheses or can’t respond to their failure they must be rejected.
Be less concerned with definitive than any. Many predicted Fed actions would lead to rampant inflation or assert there is rampant inflation and show no sign of altering. Some predicted government spending would raise interest rates and crowd out private investment. These should be discarded as ideologies.
The market of academia is strange. It is not so much a Hayekian order. A lot of the product of what acadence do is “what academics do.” As in, your product is patially judged based on how well it conforms to standards and customs and trends. So, between the self-referential products and the designed aspect of the internal market, I’d say studying it and offering insights is well-justified.
You might find “Genesis and Development of a Scientific Fact” by Ludwik Flek an interesting, stimulating read. There are many parallels between medicine and economics. Flek emphasizes that a “medical fact”.
An excerpt from “Ludwik Flek on the social construction of medical knowledge”
“In his epistemological works Fleck developed the notion that scientific knowledge is constructed. For him, alleged scientific ‘facts’ do not exist ‘out there’ in nature waiting to be discovered by objective and interchangeable observers. Rather, they emerge as the final result of a social process: the ‘genesis and development of a
scientific fact’. The observer’s training, his preconceived ideas, and his anticipations play a substantial role. Moreover, for Fleck scientific facts are constructed by distinct ‘thought collectives’, each composed of individuals who share a specific ‘thought style’.”
True in medicine. Even more true in economics.