Earlier this week, I penned a comment piece for City AM on Thomas Piketty’s book Capital in the Twenty-First Century. One of the things that I outlined there was that I believe for some countries Thomas Piketty’s reported analysis of why the capital-income ratio has increased since the 1970s/1980s doesn’t hold.
A few people have subsequently asked what I meant by that. Well, Piketty’s emphasis is that the increased capital-income ratio has arisen mainly because per capita growth tends to be lower in periods when the economy is not playing catch up and there is slowing demographic growth. These are the conditions in which Western economies find themselves now. Therefore an increased capital-income ratio relative to the post-war period is inevitable given the dynamics of capitalism and the structure of society. To show this, Piketty invokes a ‘law’ of capitalism.
In the long-run (itself a time period hard to define), we would expect the capital-income ratio (Piketty calls this β) to be equal to the savings rate (s) divided by the growth rate (g) of the economy (see Piketty's appendix online for more details). Thus, to give the example Piketty uses, if β = s/g and if s = 12% and g = 2% then the capital-income ratio should be 600% - i.e. the stock of capital would be six years of national income.
Piketty claims that this law 'allows us to give a good account of the historical evolution of the capital-income ratio' – in his opinion: to explain the U shape of the capital-income ratio for many western countries, which was high from the 18th century to mid-war period, and stayed low until the 1970s/80s, after which it increased again. Certain caveats apply of course. This is an 'asymptotic law', i.e. it is only valid in the long-run and does not hold in any given year. It also applies only to capital humans can accumulate – and not therefore to pure natural resources.
However, there is a third crucial assumption which the law requires in order to work. That is, it is ‘valid only if asset prices evolve on average in the same way as consumer prices.’ In other words, again in Piketty’s words, ‘if the price of real estate or stocks rises faster than other prices, then the ratio β…can again be quite high without the addition of new savings.’
Fortunately, Piketty gives us a nice summary table of the growth rates and savings rates in rich countries between 1970 and 2010 – providing us with an opportunity to test how accurate the law has been over that 40-year period. In the table below I use his table to calculate the implied β given the average annual growth and savings rates over the forty-year period and compare them to the capital-income ratios Piketty graphically shows in 2010.
Growth rates and saving rates in rich countries
|Country||Growth rate (%)||Private saving rate (%)||Implied β||Rough β as shown graphically in book|
|United States||2.8||7.7||2.8||circa 4|
Of course, given this is an asymptotic law we would not expect these values to all match well with the 2010 values. Nevertheless, for some countries the implied β does seem close to the 2010 value. What is striking though is that for the UK, the implied β given our fundamentals is much, much lower than the observed 2010 value. In other words, our savings rate and growth rate suggest the capital-income ratio for the UK should be much lower than what is actually observed.
Why might this be the case? Well, the key seems to be Piketty’s third assumption. In Britain, house prices have most definitely not moved on average in the same way as consumer prices. As the chart below shows, house prices over this period have increased by roughly 40 times since 1970, whilst the RPI index has increased by less than 15 times.
Indeed, Piketty’s own chart shows the increasing importance of housing wealth since the 1970s in Britain in terms of raising the capital-income ratio.
New academic evidence suggests that housing wealth is the key driver across a range of countries. Yet if housing wealth is the key driver of the increased capital-income ratio in the UK then we have to look at why house prices have gone up so much since 1970. And here, we come back again to the effect of planning laws, and in particular green belt legislation, becoming a binding constraint. Due to the relatively inelastic supply of housing, investment into it doesn’t tend to reduce the rate of return in a way a responsive market would. Prices rise to reflect instantly the higher returns from the asset. Whilst there are structural reasons why we would expect house prices in central London to be high, we would not expect the rise to continue forever, and a significant liberalisation of planning laws could see the situation reverse.
The key implication therefore is this: Piketty’s whole thesis is that the wealth-to-income ratio is returning to levels last seen in the 19th century. His own figures suggest the UK’s capital-income ratio is much higher than implied by its fundamentals - s and g. If this is a concern and is being driven by house prices, uplifted by both planning regulations and artificial government stimulus, then the implications for policy would be very, very different from Piketty’s calls for extraordinarily punitive taxes on all wealth and high incomes. In fact, Piketty has inadvertently made the case for significant planning liberalisation and shown that state-imposed constraints not only push up house prices but also wealth inequality – making it all the more surprising that those on the political left don’t focus on planning liberalisation as a policy campaign.