Reading Thomas Piketty’s Capital in the Twenty-First Century from front to back was a mistake.
Better to read the last hundred pages first, with their recommendations for the confiscation of wealth and marginal income tax rates nearing 100 per cent, and then read the preceding 470 pages to decide whether the flimsy evidence, conjecture and questionable theories the author offers justify such draconian measures.
Mr Piketty’s thesis rests on a simple insight. If the rate of return r is higher than the growth rate of wages of the economy as a whole g, the incomes of those who own more capital will grow faster than the incomes of those who own less. Indeed, his oft-cited inequality r>g has many virtues. Like E=mc2 it is short enough to fit on a t-shirt, yet seems oddly profound. It is even true. The consequent theory that, absent revolution or government intervention, the rich inevitably keep pulling ever further ahead sounds plausible too. It is, however, on closer inspection, complete nonsense.
Karl
To see why, suppose Karl is a worker who earns £1 a year. His wage grows at a rate of g=0.01, or 1 per cent a year.
By contrast, his friend Friedrich is heir to a fortune worth £100 that earns a return of r=0.04, or 4 per cent a year. Assume as well that both save at a rate of s=0.05, or 5 per cent of their income each year.
In year one, nasty rentier Friedrich enjoys an income four times larger than Karl without having to work. How much income does each receive in year two? Karl gets £1.01 in wages, and collects an additional 2/10th of one penny in interest on his savings. Friedrich’s capital has grown to £100.2 and that pays him the same income of £4 plus an additional 8/10ths of one penny. Carry this forward and it will become apparent that Karl’s income grows faster than Friedrich’s, and eventually overtakes it. It will take 155 years, but the direction of travel is what matters here.
Of course it is still possible for Mr Piketty’s mechanism of inequality to work, if Karl chooses not to save. If people differ in their savings behaviour, perhaps because they discount future consumption with different rates of time preference, the most thrifty or the most patient individual will eventually accumulate everything. Robert A. Becker described this possibility in a paper in 1980, but it is an idea Mr Piketty only introduces rather sheepishly on page 359 (and without attribution). It is easy to understand his reticence – acknowledging this channel for the propagation of inequality sounds too much like blaming the poor for their circumstances. Yet as is often the case, concern for the downtrodden coexists happily with snobbery aimed at those in the middle. Once, only the children of the very rich could expect to inherit assets from predecessors. But the average size of estates has fallen as inheritance has become more commonplace. This democratisation of family inheritance might be welcomed by someone so inimical to inequality, but no, all the machinery of the state must be deployed to prevent the rise of a class of ‘petits rentiers’.
Thrift is the villain here. Thrift makes capital accumulate faster, and it is the rising ratio of capital to income that demonstrates so clearly Mr Piketty’s mechanism of inequality leading us back to the Victorian era. But what if the ratio of capital to income has not been rising as the figures in the book appear to demonstrate (leaving aside the question of whether this would be such a bad thing)? In fact, that ratio, along with the share of output captured by capital have been so remarkably constant over so many decades that economists exploit this property when building their models of the economy. So here lies more of Mr Piketty’s sleight of hand. Mr Piketty’s theory rests on a model that describes the way reproducible capital (machinery, equipment) is used to produce consumable goods and also accumulates over time. Yet the capital that appears in his graphs is not capital as conventionally defined at all, but a much broader measure that represents total assets or wealth (a more accurate title for the book would be Assets in the Twenty-First Century). Mr Piketty acknowledges the dissonance in a one-sentence-long caveat on page 169, but then carries on.
This is no isolated instance, but rather a pattern that is repeated throughout the book, and it is this behaviour that really gets to the heart of the matter. Mr Piketty is no crank. He publishes papers in highly regarded peer-reviewed journals.
Though willing to distort economic reasoning in the service of ideology, he cares too much for his academic reputation to not leaven the polemics and half-truths with little escape clauses, so short and carefully worded they are barely discernible to anyone but fellow economists.
Then there is the straw man tactic. Acknowledging that the distribution of national income between interest payments, rent and profits on one side, and wages and other forms of employee compensation on the other, is remarkably constant would leave much of his theory in tatters. So Piketty turns the argument around, stating on page 22 that: “There is little evidence that labor’s share in national income has increased significantly in a very long time.” Who knew the only defence of capitalism rests on its ability to deliver Marx’s peculiar vision of utopia?
Lenin
Has inequality risen over the last 35 years? Yes, but this has nothing to do with over-accumulation by Lenin’s hated “coupon clippers”. Instead it is the growing disparity in wages themselves that drives rising inequality. Shifting technology plays a role, but Mr Piketty is not the first to wonder whether the structure of modern corporations – where the directors who set the salaries of senior managers are senior managers themselves – is fit for purpose. Yet suppose ‘supermanagers’ are overpaid, enough to justify Mr Piketty’s plan to cap salaries by taxing them at rates approaching 100 per cent. That would not, as Mr Piketty implies, mean that wages would rise for other workers. Instead, as Mr Piketty well knows but cannot bear to acknowledge, the savings would be passed along to the undeserving shareholders, incentivising even more saving and investment unless that is taxed away as well.
So what if we were to implement the full Piketty programme? We would start by confiscating private wealth on a sliding scale ascending to 20 per cent. Thereafter each citizen would have to supply the authorities annually with an updated inventory of all they possess, paying 2 per cent on its value. Income tax rates would reach 80 per cent on incomes of about £300k, so after adding in national insurance payments, and accounting for the remaining tax brackets, £200k would be the upper limit for net income.
Would anyone invest in risky new ventures? Would they save at all, or would they rush to spend everything before the government took it away? Or might they flee somewhere else? In George Orwell’s 1984 the world is divided between three totalitarian jurisdictions, but in Mr Piketty’s world Big Brother’s reach is global – as are his taxes. There is a pattern here too – the Soviets were apparently not the last to realise that scientific socialism could never truly succeed if the laboratory rats were free to escape the experiment.
Michael Ben-Gad is a professor of economics at City University
Key Points
Thomas Piketty’s book recommends the confiscation of wealth, and marginal income tax rates nearing 100 per cent.
Thrift makes capital accumulate faster, and it is the rising ratio of capital to income that demonstrates so clearly Mr Piketty’s mechanism of inequality.
Acknowledging that the distribution of national income between interest payments, rent and profits on one side, and wages and other forms of employee compensation on the other, is remarkably constant would leave much of Mr Piketty’s theory in tatters.
For the more advanced student:
If Friedrich starts out with k units of capital in year one, his income is rxkx(g-rxs) (1+rxs)t-1 in year t, and if Karl’s initial wage is w, his income will be wx(gx(1+g)t-1-rxsx(1+rxs) t-1)/(g-rxs).
As long as g>sxr – as is generally the case – Karl’s income overtakes Friedrich’s in year t=1+ln(rx(gxk+sx(w-kxr))/gxw)/ln((1+g)/(1+rxs)).
