Many people took issue with my post about Milton Friedman's death. Actually I don't think they were taking issue with my post; most of them were taking issue with Milton Friedman's existence. Whatever. Everyone has their own heroes. While I remain puzzled by the vast distaste directed at someone whose fundamental assertion was that you should have the ability to make the important decisions in your life, it is not likely that the debate is resolvable.
However, several people took issue with economics as a science -- and organized system of theories and facts capable of making verifiable predictions. I fall on the side of saying that it is, but I understand why people question it. It just seems fuzzy, not to mention it seems like any facts you could gather are heavily tainted by politics. On the other hand, it is not like several other fields that we classify as science aren't a little fuzzy too.
This is the argument that the SciAm blog made in their post on the subject:
So is economics a science? A science should make testable predictions about the world. Well, Friedman's most famous prediction was a pretty good one: he foresaw the possibility that high unemployment could accompany high inflation, a phenomenon better known as stagflation. That foretelling earned him the Nobel Memorial Prize, although Friedman's monetary theory is currently out of favor. Economists are performing all kinds of controlled experiments these days to test their assumptions about human behavior and the validity of poverty relief programs. And even if economics hasn't quite achieved testabillity yet, you could say the same for string theory. (Whether that does any favors to economics is your call.)
Economic predictions are necessarily imprecise, given that people behave in ways that are unpredictable in detail. But climate models and reconstructions are uncertain, not to mention fraught with political baggage, as is economics, and yet we're prepared to grant climate research a measure of legitimacy. Part of the uncertainty of economic reasoning stems from the field's reliance on an abstract and often critiqued view of human decision-making: people like money, to oversimplify. Such abstraction in itself isn't bad. Scientists make progress where they can. Physicists are no strangers to making simplified assumptions if those simplifications tell them something fundamental about the subject they're studying. There's a joke about a dairy farmer who has asked a theoretical physicist for help getting his cows to produce more milk. After thinking about it, the physicist begins her answer, "ok first assume a spherical cow." If economists assume a rational utility-seeking agent, it's so they can say something concrete, relevant, and just maybe predictive.
I agree. I also like the comparison with clilmate science. Both economics and climate science involve large mathematical models of controversial validity. Both make prospective predictions, but predictions that by their nature can only be probabilistic. Both abstract from a system too complicated to not be simplified.
- Log in to post comments
I like spherical cows.
I remember my physics teacher in high school always had us catapulting spherical cows into some large and solid -- for mechanics problems. There is nothing like the impact of a spherical cow onto the Empire State Building to make you love inertia.
I find economics much more interesting than physical science precisely because it's hard to run experiments.
Besides, what's "testable"? I provide a testable prediction of a particular economic theory every time I buy stock to diversify my investment portfolio. It's just not testable to the same number of significant figures as physical science. But it's as applicable and relevant to the real world.
Economics is not a science (and it has little to do with stock market predictions); there are few if any reports of testable and reproducible works in this discipline. Your comparison to climate science is off the mark; that is a much younger field, and one that is producing testable hypotheses. Economics is really more similar to theology; economics practitioners have their beliefs and those beliefs are immune to facts or analysis. Economics lacks the ability to `move on` in the way that science does.
"Besides, what's "testable"? I provide a testable prediction of a particular economic theory every time I buy stock to diversify my investment portfolio. It's just not testable to the same number of significant figures as physical science. But it's as applicable and relevant to the real world."
Posted by: JW Tan
No; economics makes probabilistic predictions, about large numbers of events. For example, if you were going the direction predicted by a theory, but almost everybody else was going in the other direction, that theory would have been discredited.
I don't think social scientists should be expected to have "natural sciences envy" any more than organismal biologists should have "physics envy". The limits of testability in social sciences are always going to be fuzzier than they are in fields like physics or chemistry. You simply can't exercise as much experimental control over humans or their institutions as you can over mice, magnets, or molecules. That doesn't mean that no quantitative research can be done in social sciences; it simply means that the methods and the interpretations of results have to take into account the lower level of experimental control.
Not being a student of econ, I can't comment on Friedman's methods. I'm personally in the left-libertarian quadrant of the political spectrum, and agree that "you should have the ability to make the important decisions in your life". However (and again, I haven't read Friedman, so I can't agree or disagree with him on specifics), I don't buy that all participants in the economic system come to the table with the same degree of power or the same access to information. For example, anyone who has ever tried to buy TV airtime to express controversial views knows that the same amount of money can't always buy the same thing for two different people.
The 2 commenters on my post miss the point.
When I buy stock for the purposes of diversification, I am making a testable prediction of part of Markowitz's mean-variance theory. I am predicting that the variance of my portfolio returns will fall with the addition of this stock. I can test this with returns data over a period (in fact it works with time periods as short as a year), using whatever statistical tests I want.
What I am not doing is predicting whether a particular stock is a good bet individually (more art than science). I am also making a bet on large numbers of events (of returns, in this case. Not stocks).
Economists know that economics is a science. We try to make it more scientific as practitioners. It's everyone else (usually people who don't practise) who tries to denigrate it.
"Economics is not a science (and it has little to do with stock market predictions); there are few if any reports of testable and reproducible works in this discipline."
Besides aforementioned Markowitz's mean-variance analysis, see Will Vickrey's and Paul Milgrom's work on auctions, the Taylor rule and other theories of monetary policy, Ricardian equivalence, the theory of comparative advantage, Mundell's work on optimal currency areas, just to name a few bits where testable predictions can be made and tested. It's not easy to make work reproducible in all cases, but you can for auction theory, quite easily, for example.
Doing empirical economic work is difficult because economists cannot set up large scale controlled experiments (ex You can't duplicate a county and raise the minimum wage in one version but not in the other), so economists either have to try to find "natural experiments" where a reasonable control case happens to already be present, set up simulated economic situations to determine how people behave and extrapolate from those, or try to use statistical methods to isolate the effects of the variables of interest from historic data. Friedman's major contributions to the economics of monetary policy largely relied on the last type of analysis.
I think economics gets a bad rap because it's very mathematical, so it ends up getting compared to the physical sciences, which it is nowhere close to in its ability to make precise predictions, rather than disciplines that share its difficulties in modeling complex systems and limitations on the ability to peform controlled experiments.
I think economics and ecology are a lot alike---except economics has a stronger mathematical tradition, although this is changing fast in ecology. Both sciences study complex adaptive systems (the damn things are always changing on you!!), replication is near impossible, experiments require so much simplification that it is then hard to undertand how their results should be interpreted in the "real world." Ideology gets in the way of seeing the science clearly (this is actually even more true in Econ than Ecology).
As Ecology grows more mathematical, I've often thought that the field would do much better to have econ-envy than the physics-envy which has gotten us almost nowhere.
The problem with economics, I think, is not the "science or not" question, but rather the fact that as a field it garners far more respect than it has actually earned. Economics has insinuated itself into almost every aspect of public policy decision making but has yet to make an important prediction about any major economic variable. We can't even answer questions like, "will a tax cut help or hurt our economy now?", or "how do illegal immigrants affect the economy?". You can find serious economists on all sides of just about any question. If economic theories were greeted by policy makers with even half the skepticism reserved for climate science I would feel a little heartened. As it stands I would say a comparison to Freudian psychology would be a little more apt. There may be some important ideas in there, but there's a whole load of crackpottery as well.
Now, having said all that I'll add that I'm not counting it out completely. We may get better at it and economic models may finally get interesting. Or we may find that the whole business is just intractable.
if biological anthropology always seemed to end up showing white males are the glory of evolution, would we perhaps wonder as to the 'scientific' validity of this discipline? oh, oops, been there, doen that. one thing to be suspicious about economics is that fundamentally it is a tool by white males. and somehow, it seems that capitalism always rewards the white males with the bounty of maximum utility....
"Economics has insinuated itself into almost every aspect of public policy decision making but has yet to make an important prediction about any major economic variable."
You know, scientists bash creationists for making sweeping statements about what can and cannot be done. Yet the same pro-science people make statements like the above, which are, frankly, silly.
Some counter-examples:
Keynes makes a prediction - you can stimulate aggregate demand by spending a lot of government money (fiscal policy). Government spends a lot of money, and aggregate demand goes up.
Rob Lucas makes a prediction - inflation expectations are rational and affect inflation. He then prescribes a policy - central banks need to establish credibility to stabilise inflation expectations. Central banks establish credibility, inflation expectations stabilise and so does inflation.
I could go on.
The questions that Eric Wallace poses re tax cuts and illegal immigrants are analogous to the question of whether punctuated equilibrium is a major driver of speciation - it's currently under debate. That's not to say that there haven't been a lot of questions answered already.
"if biological anthropology always seemed to end up showing white males are the glory of evolution, would we perhaps wonder as to the 'scientific' validity of this discipline? oh, oops, been there, doen that. one thing to be suspicious about economics is that fundamentally it is a tool by white males. and somehow, it seems that capitalism always rewards the white males with the bounty of maximum utility...."
Speaking as a non-white male from a 3rd world country who thanks my deity of choice everyday for globalisation, capitalism and economics, and my due reward of maximum utility, I would say you need to provide examples.
All this commentary further substantiates one of my pet theories:
Economics is the only subject that everyone thinks he / she understands, but doesn't. And despite not understanding it, people feel free to display their ignorance.
"The 2 commenters on my post miss the point."
(rest snipped)
Posted by: JW Tan
True, but because it was very poorly written. The paragraph after that clarified what hypotheses you were discussing.