Thursday, 24 September 2015

The Historical Piketty

The lasting significance of Thomas Piketty was recently proven when Morgan Stanley issued a research paper, co-authored by the economist Charles Goodhart, with the media-friendly quote: "Is Piketty history? We think so". This was predictably broadcast to an expectant world by the usual suspects, including the BBC where Duncan Weldon appears to have been fully assimilated by the neoliberal Borg [update: he's a bit more sceptical here]. If 2014 saw a steady progression through the stages of the Kubler-Ross model - denial, anger, bargaining, depression and acceptance - we have now arrived at the start of a new cycle in which Piketty is dismissed with the airy claim "It doesn't matter anyway, coz". Weldon sets the scene: "The last three decades have seen huge changes in the global economy. Three trends have dominated: falling interest rates, weak wage growth and rising inequality. Documenting the last of these trends made the French economist Thomas Piketty an unlikely best-seller. But might these trends be about to reverse? Is Piketty wrong to worry about rising inequality?"

The Morgan Stanley theory (outlined here) is that demography - specifically a rising dependency ratio: the number of young and old relative to the working population - will lead to rising interest rates, strong wage growth and thus falling inequality. There are two parts to this. First, an ageing population is expected to lead to an unwinding of the global "savings glut" that developed on the back of the rapid expansion of the labour supply after 1990. This will serve to increase consumption relative to saving and thus drive up interest rates. Second, the fall in the working population relative to the total will bid up wages. This will further fuel inflation (because those wages will be spent), leading to further upward pressure on interest rates. The authors consider some mitigations in developed economies - that more old people will continue working, that immigration will expand the labour supply, and that fertility rates may pick up - but dismiss them as marginal at best.

This is a welcome message for many. Not only does it suggest we needn't worry about inequality, because it will right itself over time, but it also promises a return to healthy interest rates that will benefit both savers and bankers. Given that OAPs are assumed to be running down their savings in this new model, the term "savers" can be taken as a proxy for "the rich" - those with surplus income. As Morgan Stanley put it: "Piketty is history, not the ineluctable future. If these global demographic trends, as we argue below, drove inequality higher, then their reversal could lower inequality too. Labour had lost much of its power to command higher wages between 1980 and 2010. Now labour will become increasingly scarce. The labour share of income, having trended down in most DM [developed market] economies since 1970, is now likely to rebound." You'll notice that no distinction is made within the category of labour, such as between the top 1% of earners and the rest, which even some sympathetic observers have noted remains an issue, and one that the Frenchman explicitly addressed.

The pun is clearly deliberate. What is significant is not merely the suggestion that the Piketty storm has passed, but that history can be safely ignored - it is not a reliable guide to the future. This echoes the "marxisant" criticism of Piketty's claims last year, namely that rising inequality is no more inevitable than the tendential fall in the rate of profit proposed by Marx. Of course, this original critique and its current iteration both ignore the Frenchman's specific point that history does not show an inexorable, linear progression but an oscillation in the mid-twentieth century, which contradicted the assumption of Simon Kuznets. In Piketty's view - which was backed by solid historical data - rising inequality was a reversion to a historical norm that was potentially incompatible with democracy (this political dimension continues to be sidestepped by the right). The focus on the capital/labour share of income (i.e. wages) ignores Piketty's more fundamental point about the accumulation of capital over time due to the rate of return being higher than growth (r > g). Claiming that this is no longer relevant massively misses Piketty's point: patrimonial capital is a legacy of history that affects society today. We don't reset the wealth counters each morning.

I haven't read the Morgan Stanley paper (a full version doesn't appear to be online [update: the authors provide a summary here] and I'm not going to buy it), so my understanding is based on selective quotations and second-hand synopses, but there is clearly a problem with their claim that an ageing population will lead to higher interest rates. Conventional wisdom suggests that too much saving (which increases with age) is deflationary, but a recent BIS (Bank of International Settlements) statistical study (quoted by the Morgan Stanley paper) suggests the opposite: "a larger share of dependents (ie young and old) is correlated with higher inflation, while a larger share of working age cohorts is correlated with lower inflation." The study authors suggest a possible explanation: "those cohorts which consume more goods and services than they produce (ie the dependents) could exert an inflationary pressure through excess demand while those who produce more than what they consume (ie the working age cohorts) could exert a disinflationary pressure through excess supply".

While this turns the standard theory on its head, the idea is not unreasonable; though it's worth noting in passing that it is at odds with the traditional claim that wage demands - particularly by young workers - drove inflation in the 60s and 70s. If the BIS study authors' supposition is true, this would suggest that high levels of underlying inflation (i.e. ignoring shocks like the 1973 oil crisis) were the product of the family, not of social factors such as trade unions, which would be an amusing historical irony. That said, my issue is not with the BIS study, which has plenty of caveats and calls for further research, but the way that it has been interpreted in the Morgan Stanley paper. Specifically, they have taken a study of 22 advanced economies (16 of them in Western Europe) and assumed that the observed correlation will hold good for developing nations as well. This ignores cultural and social differences between the West and the rest, and in particular the role of the welfare state.

It is reasonable to believe that the growth of state spending in advanced economies over the twentieth century redistributed money from workers to dependents - via health and social care, education and pensions - and that this in turn reduced precautionary saving and fuelled inflation (because more money was being spent), which combined to drive up interest rates. It is also reasonable to assume that globalisation (the increase in labour supply) put downward pressure on median wages at a time when political pressure to reduce state spending was also deflationary, thus leading to lower inflation and falling interest rates in the early-90s. The problem is that China is not the UK. Without similar mechanisms, we cannot necessarily expect the same dynamics. For example, without the creation of a Chinese NHS and the provision of decent state pensions (rather than reliance on family and limited private provision), an increase in the dependency ratio may encourage further precautionary saving as a percentage of GDP, not less.

A second issue with the Morgan Stanley paper is that while it makes great play of the anticipated fall in the working-age population as a percentage of the total, it has little to say about its compositional calibre. It is a mistake to think that labour is fungible - i.e. that anyone can do anyone else's job - and that a falling supply of brain surgeons will therefore drive up the wages of bin-men. The expansion of education in the developed world - which has run for 150 years and has been accelerating over the last 50, with little sign of letting up (e.g. we're now mandating education to 18 in the UK) - is being repeated in the developing world. This will rapidly increase the cohort of labour capable of cognitive work (graduate under-employment shows that this is already a reality in the West), so the downward pressure on wages may still persist even if robots don't make inroads into whitecollar roles as quickly as expected. Meanwhile, we can be confident that automation will continue to reduce the demand for manual labour (or force down wages as an alternative to capital/labour substitution), because that is a process that has been observable for decades now, despite the increase in the global labour supply.

The Morgan Stanley analysis is inadequate and popular for the same reason: it explains contemporary dilemmas away by appeal to a single factor, demography, which now appears to be heading in a benign direction (benign, that is, for bankers if not the NHS). This is the invisible hand at work: inequality will decline as the market reaches equilibrium; no state intervention is needed. The thesis is also attractive to those for whom generational conflict is a more palatable explanation for rising inequality than class conflict, and those reluctant to acknowledge that global imbalances in savings and investment reflect geography (and thus international politics) more than demography. The most striking feature of the analysis and the admiring media reception is the complete absence of any reference to the welfare state. It's not just that an "ageing population" is now being recast as the solution to some of our problems, but that the phrase - with all its negative connotations - has become divorced from the NHS for the first time in decades. History has been wiped clean.

1 comment:

  1. Hey, thanks for this! I was just reading about Japanese demography, and I stumbled onto mentions of the MS paper. Seemed pretty simplistic to me compared to TP.