Thursday, June 29, 2017

Chris Dillow doesn't get libertarianism


Chris Dillow - why libertarians should read Marx. All good points, but:

There’s astonishingly little in Marx about a centrally planned economy: if you want an argument for central planning, you should read that hero of the right, Ronald Coase instead (pdf). Marx was admiring of capitalism in some respects. It has, he wrote, given “an immense development to commerce” and has “accomplished wonders far surpassing Egyptian pyramids, Roman aqueducts, and Gothic cathedrals.” And I think you’d be surprised by just how much attention Marx paid to the facts: once you get past the first few chapters, there’s massive empirical work in Capital volume I*. And there are many differences between Marx and social democrats – not least of them being that Marx was no statist.

What’s more, many of the ideas associated with Marx were largely elaborations of his predecessors: Paul Samuelson called him a “minor post-Ricardian”. The labour theory of value, the interest in the division of income between classes and the idea of a falling rate of profit are all as Ricardian as Marxian.

The problem is, Chris, that libertarians don't give a shit about resisting state power. If they did, they wouldn't be advocating transvaginal ultrasounds, criminalizing labour bargaining power, and so on.

Libertarians are libertarians because they hate the poors. So if Marx is advocating for the poors, he's the enemy of libertarianism. Full stop.




Wednesday, June 28, 2017

Welcome back to the middle ages


It's not really as bad as the middle ages, but zero productivity growth is something that should really truly concern anyone who believes in investing in equities:

NY Fed - low productivity growth: the capital formation link.

Unfortunately, one problem is that this kind of investigation requires growth accounting, and if it's one thing I learned last semester it's that growth accounting requires such insane assumptions that it's hard to tell if the numbers it yields are worth anything at all. (For example, are you generating the capital stock through an assumed depreciation rate and by assuming investment equals capital? Or, if a brand name is "capital", are you including advertising as investment?)

Basically these exercises require making up a shit-ton of numbers and assuming variables remain constant, and their inability to work across countries (i.e. if you use a method to generate capital stock and depreciation rate for Belgium, that same method will generate nonsensical numbers for Japan and Sweden, and I know cos I did the work). The method always fails dismally out-of-sample, so I don't see how you can write a serious article on the topic.

There's also an ideological impurity in this work, in that it asserts a worker should only see an increase in real income if their productivity improves: that's nice, except if employers have increased market power in the labour market due to say 30 years of anti-labour legislation, they won't be passing much of their improved earnings to workers.

And actually, that reminds me of the one argument for why the Industrial Revolution happened in England. The idea is that with labour being scarce and with cheap coal able to drive machines, the English industrialists invested heavily in capital to substitute away from labour; the result of this was that the scarce labour being employed became more and more productive because it was combined with more and more capital, and a positive feedback loop caused the English economy to spiral up to a high-capital equilibrium.

In that case, the problem with the US is lack of effects from labour scarcity: partially from labour being replaced with overseas labour, and partially from existing labour scarcity not causing a higher labour price (wage inflation) due to anti-labour political policy at home.

Basically, there's a completely different argument for why capital formation in the US is still poor.