Much as I admire both protagaonists, the debate between Simon and Jo on the role of ideology in economics left me cold. I'm not interested in whether economics is mainstream or heterodox.

To see my point, let's take an example from outside macro - the capital asset pricing model. This is an elegant, internally consistent and well micro-founded theory.

And it's wrong. In particular, we've known since the first tests of it that low-risk stocks do better than they should. This is weird: theory tells us there should be a trade-off between risk and return, but the facts tell us otherwise.

Herein, for me, lies the essence of economics - the collision between theories and facts.

There's a parallel here between the CAPM and (some?) RBC models. Both are theoretically elegant. And both fail to survive contact with the real world. Just as the great performance of defensives undermines the CAPM, so the existence of involuntary unemployment rejects RBC theories: the unemployed are significantly less happy than those in work - and in fact unhappier than the divorced or disabled.

But why is the CAPM wrong? There are some nice theories (pdf). But I want to draw your attention to a recent paper by Michael Ungehauer and Martin Weber at the University of Mannheim which I think is a model of good economics.

They show that a key assumption of the CAPM is wrong. The CAPM assumes that investors measure market risk by beta, the covariance of a stock with the market. But Ungehauer and Weber show that people don't measure risk this way. Instead, they assess correlations by using a simple counting heuristic: how often does a stock co-move with the market? This leads to different measures of risk, because it underweights the importance of large price moves.

They established this first by experiments. Then they applied their theory to US data. And they found that if stocks' riskiness is measured by the frequency of co-movement with the market, there is indeed a trade-off between risk and return. It's just that, contrary to the CAPM, risk isn't measured by beta.

In this sense, there's another analogy between the CAPM and RBC theory. The CAPM says that agents' view of risk is determined by beta; RBC theory says their view of the future is determined by rational expectations. For me, both claims must be tested against the facts.

I appreciate that I might just be expressing a personal taste here, but I don't give a toss whether a theory is elegant or not, or mainstream or not. For me, what matters is: does it fit the facts? And: does it work? My response to elegant theories is often like that of Andrew Tyrie to Boris Johnson: "This is all very interesting, Boris. Except none of it is really true, is it?"

Now, you might find this surprising. We Marxists are supposed to be spittle-flecked ideologues, and yet here I am demanding facts and utility.

But of course, there's no paradox at all. As a Marxist, I have no skin in the game of whether the CAPM or efficient theory is right or not: such matters are orthogonal to my concerns qua Marxist. And in fact even if Robert Lucas's main points were right - that business cycles are an optimum response to technology shocks with little welfare cost - a lot of Marxism would survive. Such claims are consistent with the notion that capitalism is exploitative and alienating and leads to unacceptable inequalities of wealth and power.

It's sometimes said that Marxism brings ideology into economics. For me, though, it takes it out.