Tim interprets Thomas Piketty's theory (pdf) that wealth concentration is rising because returns on capital are high as evidence against Marxism, as this predicts a tendency for the rate of profit to fall.

I fear he's being a little hasty here.

For one thing, Marx only saw a tendency for profits to fall, and cited numerous "counteracting factors" which could reverse this tendency. And for another, his idea of a falling profit rate was really just a rehash of the idea of the "stationary state" which pretty much all the classical economists had - that a time would come when diminishing returns overcame technical progress and caused growth to cease.

Let's though ignore that. There's a question here: if returns on capital are so high, how come companies are buying back shares rather than investing, and that so many are talking about secular stagnation?

My chart highlights the issue. I've measured the US profit rate as non-financial pre-tax profits expressed as a percentage of the previous quarter's non-financial corporate assets, taken from the Fed's financial accounts. This shows that the profit rate trended down from the 50s to the early 80s, but has risen since, albeit punctuated by the tech crash and financial crisis*.

This is consistent with the trend in the income share of the top 1%, which also fell from the 50s to late 70s and has risen since**.

However, the turnaround in profit rates hasn't been accompanied by a rise in capital spending. The share of non-residential investment in GDP is lower now than it was in most of the late 70s and early 80s, when profit rates were lower. Higher profits, then, aren't providing much motive to invest.

One solution to this puzzle is that there's a discontinuity between past investments and future ones, or - if you prefer - between existing assets and growth options. Whereas past investments have been profitable, future ones are not expected to be so. The cliche's true - past returns really are no guide to future ones.

This is possible because investments are inherently lumpy and heterogenous. The fact that investments in iPads or Coca-Cola bottling plants have been profitable tells us nothing about the likely profitability of new future ventures.

This explanation strikes me as wholly reasonable. But it poses a serious problem for conventional neoclassical thinking. It means capital can't be aggregated and that the idea of a smooth, easily differentiable marginal product of capital is hooey - as some Cambridge economists pointed out (for different reasons) years ago***.

And this is why I say Tim is being hasty. If returns on capital are high, it's a problem not just for Marxism, but for neoclassical economics too.

* I'm using US data as there's more comparable longer-term data than there is for the UK. I suspect the trends in the latter would be pretty similar.

** Exactly what the link is is another matter. It could be that better management has raised profit rates and shareholders have rewarded top bosses for doing so. Or it could be that socio-economic change has raised profit rates and bosses have seized the proceeds of this for themselves. I shall leave the reader to guess which explanation I favour.

*** I know - if capital can't be aggregated the notion of an aggregate profit rate, as shown in my chart, becomes, ahem, problematic. I don't think this undermines my general point.