The Metric Is Not The Goal

A side eddy that came up in the below-mentioned free trade debate is this notion of ‘GDP-maximization’ as a symbol for what is (purportedly) wrong with liberal economics. Foseti says that there are/should be other values besides simply maximizing ‘GDP’ (here serving as a stand-in for wealth). Aretae says economic growth rate is the god-metric.

Once again – and I’m sorry I know this is getting really annoying – but they’re both right. If you focus on ‘GDP’ (or any similar, actually-existing metric) and construct policies around trying to maximize it, or that you think would maximize it in your thought-experiment, seems to me you are indeed likely to end up with some warped outcomes, that antiliberal skeptics can and will point to, as a weak link in your free-market armor, as evidence that liberal economics ignores (what Foseti terms) ‘higher values’.

Yet I still agree with Aretae that economic growth should be the goal! How can this be?

Growth should indeed be the goal. It’s just that we can’t measure it worth shit. We can’t measure anything worth shit, because these aggregate metrics and statistics are all the result of large calculations by nice-comfy-job’ed well-enough-meaning Smart People, and all large calculations are wrong. What we can do is use reason and economic common sense, to understand which policies are likely to point in the right direction. And then make convincing arguments for those polices, based on reason and economic common sense (NOT statistics like ‘GDP’).

And the right direction is, undeniably, economic liberalism, which is to say, the free market, which is to say, freedom. This is probably why it remains true, as Bryan Caplan wrote, that the more people understand and have knowledge about economics (as opposed to ‘GDP’ or something), the more economically liberal they are.

But if you think ‘GDP’ is a synonym for wealth and ‘GDP growth’ is a synonym for economic growth, and/or if you think that economic liberals actually believe in ‘GDP’ or otherwise only look at part of the picture, then Foseti’s appeal to these or those ‘higher concerns’ left unsatisfied by liberal economics has undeniable appeal. But of course ‘GDP’ is just a made-up metric; surely it’s designed to capture something interesting and meaningful, and does an ok job in the main, but even if measured perfectly it’s ultimately – necessarily – an imperfect approximation to the thing it’s meant to signify. (Interesting discussion of some conceptual problems with the GDP construct, which I have no doubt only scratches the surface, is here.) And we can’t and don’t measure it perfectly anyway. But this is sort of a red herring. GDP is just a metric.

So let’s address one of Foseti’s ‘higher concerns’, the externality caused by an Oregon teenager being unemployed when he could be pumping gas due to their full-service-gas-station requirements. The idea being, a ‘you can’t pump your own gas law’, in creating a job for that teenager, could be better than the nominally ‘GDP-increasing’/efficient solution of not having such a law. Do I agree with Foseti that an unemployed young person is a problem? Of course. Do I think the ‘GDP-maximizing’ mentality could lead one to render the teen unemployed with no (or, insufficient) associated benefit? Sure, perhaps. But do I agree with Foseti that such laws therefore make sense? I absolutely do not.

If that teen being unemployed causes an externality, that externality, by definition, is destroying wealth one way or another (this is what externality means). Foseti would probably say that the sort of ‘wealth’ it’s destroying won’t necessarily show up in ‘GDP’, but that’s merely a statement of the incompleteness of ‘GDP’ as a measure of wealth – not a proof that increasing wealth isn’t a good goal. And if it’s destroying so much wealth that it’d be better just to have a self-service ‘tax’ (in the form of that law), then having the self-service law would be the wealth-maximizing solution. It may or may not be the ‘GDP-maximizing’ solution, but if it’s not, that’s (in this hypothetical) a knock against GDP as a metric of what it’s meant to measure. Basically, Foseti in citing an ‘externality’ is saying, every bit as much as Aretae did, that the best policy is the one that maximizes wealth; it’s just that it may not maximize GDP (in which case GDP is failing to capture something important about wealth).

But it would not be so if ‘GDP’ were a perfect metric – if it really were the god-metric. Using the god-metric, the best policy would always show up on top. It’s just that we can’t build, let alone measure, the god-metric.

This all sounds pretty tautological, which it is, which is why once I figured this out I stopped getting into these sorts of discussions.

The ultimate problem for Foseti’s case is that he needs to actually establish his claim – namely, that the externality of the unemployed Oregon teen outweighs the economic damage caused by the inefficient no-self-service law. Unless I missed it, he hasn’t. This would require actual economic argument. Economic liberals have the argument that, if the teen pumping gas were the economically-preferred solution, then people would have been paying up for full service already. Foseti instead just substitutes a call to ‘higher concerns’. And he’s clearly correct that that’s what people do, when they oppose liberal economics as cold and not responsive to other human values. They say that an obsessive focus on ‘GDP’ leads to the neglect of other important values. Is that true? Of course it is. ‘GDP’ sucks, it’s just some human-constructed number that to be measured and calculated, and all large calculations are wrong.

But that doesn’t mean that economic growth shouldn’t be the goal, and it’s not an excuse for substituting your ‘higher concerns’ for those of other people – especially the ones that show up in market prices. Essentially, if you think ‘GDP’ is an imperfect metric, I’m with you, but if you want me to support policy X as economically preferable, you still need to convince me – with actual economic argument – that it increases efficiency and wealth (broadly defined – and I’m willing to entertain quite broad notions of ‘wealth’). But this is hard to do if you make no reference to – or, worse, ignore and wave away as meaningless – revealed preferences whatsoever.

‘GDP’ certainly is, and market prices may be too, an imperfect metric for the thing they represent. I am nothing if not skeptical – to the extent that I can pretty fairly be accused of being an anti-intellectual know-nothingist – of human-created metrics. But at the end of the day, even I think – especially – market prices are probably better than your waving of hands. In short, if you really think it’s so beneficial for the teen to pump your gas, then feel free to express that value – that ‘higher concern’ – by reaching into your pocket and paying full service.