Economics is personal, whether we like it or not

I love to read a bit of name-dropping. A fine example thuds on to the page at the start of the latest piece by Raghuram Rajan on the public spat between Paul Krugman and Reinhart and Rogoff:

For those who know these two superb international macroeconomists, as I do…

He’s trying to argue that the metaphorical punch-up, which was about whether or not too much public debt causes economies to slow down, should have between conducted entirely calmly and free of vitriol. Rajan believes that nobody can be so sure of their position that they should dig their heels in quite to the extent that Krugman, Reinhart or Rogoff did. “Economics is an inexact science, with exceptions to almost every pattern of behavior that economists take for granted”.

But Rajan’s shameless self-propagandising-by-association somewhat undermines his argument that economic debate shouldn’t be personalised.

Rajan makes the curious point that: “Perhaps respectful debate in economics is possible only in academia.” Last time I checked, most academia was so infested with infighting that it made the UK House of Commons look like the picture of restraint.

All social science starts from a subjective, unstated position. It just isn’t the neutral, value-free discipline that its physics-envious mainstream practitioners pretend. Economic behaviour is unpredictable and complicated, so those who study it are hardly likely to achieve a consensus. This is a point which Rajan himself recognises:

economic behavior is complex and can vary among individuals, over time, between goods, and across cultures. Physicists do not need to know the behavior of every molecule to predict how a gas will behave under pressure. Economists cannot be so sanguine.

…what seem like obvious, commonsense policy solutions all too often have unintended consequences, because a policy’s targets are not passive objects, as in physics, but active agents who react in unpredictable ways.

As Deirdre McCloskey said long ago, economics is rhetoric. Like all social sciences it’s a process of trying to convince others of the validity of your position using all the techniques at your disposal. So a statistical regression doesn’t “prove” a theory to be true or false — it’s just evidence (albeit maybe very good evidence). A clever model doesn’t provide an “answer”. It simply posits that, under certain circumstances, a particular relationship might hold. All sorts of hidden or clearly stated assumptions underlie any model.

The inevitably rhetorical and inexact nature of economics, not to mention its dependence on values, means that it is always likely to be personal. It’s a human belief-system whose outcomes are destined to be all-too-human. It’s almost impossible for economists or any other social scientists to progress without developing a set of beliefs about how the world works, and these beliefs are likely to be based on personal experience and political persuasion. That’s why Krugman and Reinhart and Rogoff shout at each other so loudly. Rajan is so keen to tell readers that he is mates with a Nobelist and two academic superstars because he wants to persuade us that he’s a bigshot who’s worth listening to.

Nothing wrong with that — we’re all trying to convince. Power always enters into the equation. But it would be a good idea to admit the rhetorical nature of economics as well as its non-neutrality. Economists should probably state their assumptions and try to make clear what biases they bring to their research.

I actually find it slightly creepy that economists would try to portray themselves as white-coated scientists tinkering behind the scenes, especially when such monumental political decisions are at stake. It’s no bad thing that some debates are conducted in public, even if a bit of snarkiness creeps in. And in the end who doesn’t like watching a bit of a brawl?

If only economists were remotely close to following Rajan’s advice:

All of this implies that economic policymakers require an enormous dose of humility, openness to various alternatives (including the possibility that they might be wrong), and a willingness to experiment.