Austrian Theory Is Not An Economic Theory

So the old Keynesian-vs-Austrian debate has flared back up in the blogosphere. Economic Policy Journal has the full story, but the short version is that Robert Murphy, a famed Austrian economist, made an incorrect inflation forecast, which Paul Krugman and Brad DeLong have jumped on to argue that the Austrian explanation for what ails the US economy is misguided.

Getting a forecast wrong is, of course, OK. But Krugman and DeLong’s main gripe is that the incorrect predictions coming from the Austrian school in recent years have not forced the school’s practitioners to retool their theory.

Here’s Krugman:

Basically the whole [Austrian/Austerian] movement has been wrong about everything for two and a half years — wrong about interest rates, wrong about the effects of austerity on GDP in Europe. Yet where is the reconsideration?

I thought it might be useful to take a step back and talk briefly about methodology; because, while I think Krugman and DeLong are more or less right, they do seem to be slightly misrepresenting the Austrian view. In particular, Austrian theory doesn’t make specific predictions about economic phenomena; the theory simply serves as a blueprint for thinking about relationships between economic variables.

Major_Freedom, a commenter on Robert Murphy’s blog, has the full story:

Krugman writes:

“On the other hand, the unfortunate Romer-Bernstein prediction of a fairly rapid bounceback from recession reflected judgements about future private spending that had nothing much to do with Keynesian fundamentals, and therefore sheds no light on whether those fundamentals are correct.”


“The fact is that while Keynesians predicting a fast recovery weren’t really relying on their models.”

The same exact principle applies to the Austrian model. The Austrian model DOES NOT “predict” that price inflation will be X% when the money supply increases by Y%. Krugman is being unfair. While he claims that Keynesians are immune from bad predictions, he doesn’t grant that same immunity to Austrian theory. To him, Austrian theory says that price inflation goes up Y% when the money supply goes up X%. But that is not correct. Austrian theory does not contain that argument at all. In fact, it holds the exact opposite.

Rothbard explains:

“Mises agreed with the classical “quantity theory” that an increase in the supply of dollars or gold ounces will lead to a fall in its value or “price” (i.e., a rise in the prices of other goods and services); but he enormously refined this crude approach and integrated it with general economic analysis. For one thing, he showed that this movement is scarcely proportional; an increase in the supply of money will tend to lower its value, but how much it does, or even if it does at all, depends on what happens to the marginal utility of money and hence the demand of the public to keep its money in cash balances.”

Krugman is not correctly describing the Austrian theory. He wants his readers to believe that Austrian theory predicts hyperinflation whenever the Fed prints a lot of money, and that’s that. If it doesn’t occur, then Austrian theory is wrong. Krugman is repeating this lie over and over again, almost as if he doesn’t want his readers to look closely into Austrian theory lest they “convert”. So he keeps “reassuring” his readers that Austrian theory is wrong because we haven’t had hyperinflation yet, so don’t bother with it.

Bob Murphy for his part made a bet with Henderson about price inflation, yes. But this bet was NOT a consequence of any Austrian models. It was Bob Murphy’s model. Austrian theory does not make predictions of the form

Price inflation = a + b * (Money supply growth)

What Bob Murphy did was what Krugman claimed Romer and Bernstein did. They, like Murphy, made a prediction that did not turn out correct, but they were not Keynesian or Austrian models.”

I think this comment is accurate, but I also think that it exposes an even bigger flaw with Austrian theory: namely, that the theory isn’t really an economic theory.

Economics, after all, is a positive science, whose purpose is to make accurate predictions about the world in which we live, so that consumers, businesses, and policymakers can make better choices. This is how John Stuart Mill described the science’s purpose nearly 200 years ago; and how Milton Friedman did the same in the 1950s. In short, if the purpose of Austrian theory isn’t to make predictions about economic phenomena, then the theory is not an economic theory – at least not in the behavioral sense.

Which is obviously fine: normative, non-behavioral endeavors in the social sciences are clearly useful endeavors, which warrant considerable attention.

What’s NOT fine, though, is when normative, non-behavioral social scientists try to influence public policy. Austrians claim to not make explicit predictions about economic phenomena, but the fact is that they’ve been trying to influence policy in meaningful ways – whether by arguing that we need to cut government spending now now now, or by arguing that all unions and regulations need to be abolished, so that the “free market” can operate more efficiently.

The point is, if you’re unwilling to make predictions (using empirical tools like statistics), then you’re not allowed to have a say in economic policy matters. And, to be clear, I’m not arguing that normative matters aren’t important in policy debates; we should obviously always care about things like distributive justice and moral rights. But we don’t call the people arguing for normative goals “economists” – we call them philosophers.

In this context, the whole Krugman/DeLong-Murphy debate is somewhat pointless because you essentially have the aged-old debate between economists and philosophers, and each group is speaking past one another on fundamentally different playing fields.

4 Responses to “Austrian Theory Is Not An Economic Theory”

  1. Shenpen Says:

    It is not that simple. Both models have big problems.

    The problem of the Keynesian model is that it takes the economy as it is, it takes a real-world model, a ” a positive science, whose purpose is to make accurate predictions about the world in which we live”, meaning it takes government interventions as a fact. Because they are a fact. There are plenty of them in every country. Example: Keynes’ famous argument about sticky wages, which are mostly from interventions. But if you include politics – which is inherently unpredictable, they could ban or tax anything tomorrow – you don’t have a model of economics, you have a model of politics and economics, or political economy, and every time politics changes you have to update the model. Thus the Keynesian never knows where is truly eternal economic model ends, and temporary political circumstances begin. And is it even possible to model the huge number of political laws and rules currently in effect? How can you predict this subsidy here is more important than that tax there?

    Austrians take a “pure” model, a market free from interventions, in order to figure out the market logic, the purely market logic, without the distortions. They built a model of pure economics, with eternal rules that would function even in a sci-fi future. Unfortunately, nowhere in this world is the worl purely economical, without political interventions. Thus their theory is about a perfect model world, not about the real world. To bridge it into the real world takes considerable effort, and the problems is that the most popular, Rothbardian branch of AE takes their model world a bit too… religiously to make a serious effort.

    Summary: Keynesianism is a real-world model, but not an economic model, but a mish-mash of eternal economic law and temporay political circumstance. Austrianism is a purely economic model with eternal laws, but not of this world as this world has no pure markets.

  2. JSeydl Says:

    But we have to deal with the world as it is, Shenpen. JSM called the science “political economy” because it really is inseparable from politics. It’s great to have a frictionless, zero-intervention model that gives insights about how the market operates, but if the model has no actual basis in the real world, then what’s the point?

    I never said there aren’t problems with Keynesian models. There certainly are. But at least Keynesian models attempt to model real-world phenomena, even if many of the predictions are wrong. This, in my opinion, takes a lot more courage than outright refusing to predict anything…

  3. L. Reichard White Says:

    Keynesian predictions, even when correct, are pretty much useless — they lag real world, as do their prescriptions, by at least six months. That means that Keynesians don’t know what’s happening real-time. In fact, they are six months late because of data acquisition, etc. delay. Not to mention sampling problems. Likewise, their fixes have at least a six month lag time. That means they are at least a year late.

    ‘Dumb’ economists still can’t predict recession, Ben Bernanke, Jan. 2008, @ Mayflower Hotel

  4. L. Reichard White Says:

    I left a link to a video, but it didn’t take. Let me try HTML: ‘Dumb’ economists still can’t predict recession Ben Bernanke, Jan. 2008, @ Mayflower Hotel

Leave a Reply