More notes on Piketty's Capital

Right, I have now finished the book, for which I already published two sets of notes on the way.

However, there is so much material for discussion in the book after the point where I stopped my notes that, to make it a bit more readable, I will split the final notes in two more installments. I therefore return to part 3, looking at inequalities, from Chapter 9 onwards, and will stop at the end of Chapter 11.

Piketty being the specialist of inequality, there is little surprise that he gives a very good description of its historical evolution. 

It’s hard to pinpoint one thing in particular, but despite having over the last few years read each of his papers or articles that came my way, I kept finding things I did not know, that will probably alter my views of economics and politics in the future, even if I don’t necessarily know in what way yet. For instance, I did not realise that Germany’s top 1% had a markedly higher share of revenue than in any other continental Europe country. It is also interesting to look at the very different evolutions of minimum wages in France and USA (although that had very little impact on inequality at the very top, which is what took off of late).

There are also surprises, at least for me, in some of his comments. For instance, while Piketty mentions among the possible causes of the explosion of inequalities that there has been a divergence in the negotiating powers of Capital and Labour, he quickly adds that this is not his favoured theory. Well, I find that rather extraordinary when one looks at what happened over the same period (essentially the last 35 years). We had the conservative revolution, the coming into force of several agreements allowing full mobility of capital (of course, some countries like China don’t give full mobility –but they don’t see the phenomena that Piketty describes), the fall of the Eastern bloc (which reined in unfettered Capitalism somewhat), the weakening of unions in pretty much every Western country, myriads of ways for companies to avoid tax and regulation… All of that greatly contributed to the rise of the share of profits in added value, and a great weakening of the position of all employed but the very top executives. I have since then noticed that Piketty acknowledged in interviews that this is an important factor that is somewhat missing in the book. I will return to this later -what is clear is that, while there may still be some discussions on the relative importance of various causes, the effect is indisputable. And since Piketty does suggest a way to address the effect, that is not a vain conclusion.

Piketty mentions education as a force going against inequalities. Although he will quickly add that this is never going to be enough, he also seems to say that this effect is not even controversial. However, I find the data he provides extremely unconvincing. As far as I can see, he mentions the case of France, which spent a lot on education for all without seeing inequalities shrinking because “everyone went up a rank” (in education) –adding that without the spending, inequalities would have increased greatly; and that of USA, who reduced spending around 1980 and saw inequalities shoot up (but they did so many things favourinf inequality at that time that it’s hard to point to the education spending.

I really don’t want to pick up a fight with education, which I see as desirable by itself and not just for economic prospects (Piketty says as much), and there may well be many known studies that I should know making this effect entirely uncontroversial, but I must say that what the book presents does not seem sufficient to conclude. In any case, Piketty is under no illusion that education would remove inequalities, as we shall see.

Further, Piketty takes some time to explain why in the situation of the very top managers makes it impossible for their wages to be set at the marginal level of productivity. I don’t dispute that this is the case, and understand Paul Krugman’s comment that one should not slam Piketty for not discarding traditional economics (even if that does not mean he endorses every one of its conclusions), rather that it demonstrates that even within that framework, inequalities can be described and explained.

However, when it comes to the idea of salaries being set at the marginal level of productivity, this goes a bit beyond what Krugman said. The thing is, it’s not just that there are some issues in some circumstances, it’s that the theory makes no sense whatsoever. Its logic would require full employment and symmetrical flexibility (both a laughable proposition in today’s world –and the latter in any world. But even if those conditions were met, the simple truth is that this is not remotely like the way compensations are decided –and I write as someone who has sat on salary boards (not for top multinational executives obviously –but my point is that the hypothesis makes no sense for most salaries, not just for superstars). Looking into power relations and expectations management would get you much closer to the truth -this is linked to the underplaying of the loss of union power.

Also, top salaries tend to be paid for who you know (whether among potential clients or people in positions of power) more than what you know. In a world where the top 0.1% essentially isolate themselves, this should be a very strong positive feedback loop for inequality. Piketty does mention some positive feedback loop, but mostly through channels such as increasing tuition fees. That is a strong factor between, say, labour and middle class and upper middle class. But university degrees cannot explain the massive divergence at the very, very top, often happening to people who were not necessarily near the top of the academic pile. Certainly many of them did not top ethics, either.

Piketty does mention that evidence points to an impact of the top marginal tax rates, in that they seem to lead to an explosion in top pay (of course, there is less incentive to run silly risks for more pay when 91% of it will be taxed). He does not explicitly say so, but I would say that the reduction came in strong correlation with a change in social norms (which Piketty points towards, rightly I feel, as a major determinant of very top pay) and a proliferation of pseudo-justifications of extremely high revenues. There seems to have been a lot of positive feedback.

But what he does not mention is that, at the same time, the safety net has collapsed in the very same (English-speaking) countries. This must have helped top executives share a bigger portion of the pie, since lower-paid workers were facing a rather grim situation, where were they to lose their salary (or salaries, some people having to work more than one job), they would be destitute. However, one could have expected the late 90s to be better in that respect (since unemployment was very low). But Piketty’s data series indicate that they were a time where wage inequality was going up very fast (faster than just before and after), as measured by the proportion of the top 10% and 1%. So maybe that weakening of the situation of the less fortunate, while socially destructive, was not the main factor in the rise of the top.

There is a discrepancy between the data and the text at one point, when Piketty describes the extent of the rise of the top 1% in the UK and USA. Because in the text he has to group countries and give approximate value, and with the UK starting at the bottom of a group and finishing at the top, he asserts that the rise of the top 1% has in the UK been half of what it has been in the US. However, the figures are from 6% in 1980 to just under 15% in 2010 (just under 16% in 2007) in the UK, and from 8% in 1980 to just over 17% in 2010 (just over 18% in 2007) in the US. So a similar rise, but from a lower start in the UK, so higher in proportion. Admittedly, this is excluding capital appreciation, which should have a somewhat bigger impact in the US but still, it may be important to remember that the UK actually had just as much of an increase in the share of the top 1%.

However, I must admit that the UK is a country that should always be looked at with suspicion, as it has to a considerable extent a tax haven economy, to an extent that I cannot see in any other major developed country. For instance, I do not know the impact of non-domiciled on various data series, but it could be considerable, as billionaires could report essentially no capital.

The part about developing countries was informative for me –I would not have expected at all to see Argentina so high in inequality over the last few years (of course, I completely expected it during the dictatorship, but the series breaks then). Indeed, it is still rising. I am aware, though, that those series are based on tax returns and thus that they are prone to quite a lot of uncertainty in countries where the fiscal administration is not strong –although that could probably also be said of several developed countries, at the very highest level of revenue at least. In any case, those countries are not the most important for Piketty’s reasoning.

In chapter 10, Piketty moves towards the inequalities in capital. Let’s give it away: they are even stronger than the revenue ones. This should not be surprising to anyone with eyeballs, however it is rather amazing to see how blind common wisdom has been. Piketty talks about the Modigliani curve, that states that people accumulate capital while they work which they later spend, to the point that they have very little when they die. In fact, there is no such trend: the older the population group, the higher share of capital it holds. This is quite consistent with Piketty’s thesis that, beyond a certain level, capital essentially feeds itself and reaches a pattern of divergence.

He draws a very long history of capital and its returns. I find it pretty unconvincing actually. For instance, capital is supposed in this book to be measured by its market value. But for much of history, most of it was agricultural land, for which there was essentially no market –it was inherited or conquered. Also, I keep feeling that the estimated returns do not take a good account of the costs of keeping capital at its level. That is true even of agricultural land in Antiquity: you may argue that it gave you the means to purchase an army. But you probably also needed to have one in order to keep your land –try and cash in on purely financial terms and your pot of gold may go. Actually, this was the base of the feudal society, where landowners had the duty to defend their serfs. Probably they realised very early that it was a “service” that they had better provide in any case. 

Also, clearly a return of 4% in excess of growth during two millennia would quickly swallow everything of course. Piketty mentions that crisis reset that somewhat at regular interval –but then is the given return all that meaningful? We can’t consider the bust as an outlier in a boom and bust cycle. Clearly, there was great social stability, if you were not wealthy you could not become so –but since most of the population was not allowed ownership in any case, it was even more a consequence of the system of economics.

In the end, Piketty concludes that the return on capital being superior to growth has been a historic reality, not a logical conclusion. Clearly, growth has generally been rather low, and a conjunction of elements seem to drive capital returns to a value that is, in any case, much higher than 1% (whereas growth is often lower and rarely much higher). That is probably true, but I feel that studying the said conjunction of elements would be important for the policy recommendations. As it is, Piketty seems to accept r>g as a fact and tries to deal with the consequences, rather than taking it on directly -which is sensible enough.

As an aside, Piketty notes that the fact that growth went from 0.2% yearly until the 17^th century to 0.5% in the 18^th and 1% in the 19^th hardly made a difference as the gap between g and r (at around 5%, indeed slightly higher during the early industrial revolution) was not affected much. True, but this should have made a huge difference to the proportion of capital to revenues, since he earlier explained that they trended towards s/g (see previous notes), whereas according to his series it pretty much stayed constant (admittedly in part due to colonisation –internal capital to revenue dropped a little, but nowhere near in the implied proportion, and in any case some of that revenue came from the external capital, so even that drop would be an overstatement of the internal factors). This should indicate that β=s/g should be used cautiously, and probably s is strongly affected by changes in the economy.

Later Piketty looks at the early 20^th century, and the illusion (or bad faith?) that the repartition of wealth was pretty much a given, that could not change (so it would be futile to try). Amazingly, it would change dramatically but a few years later, with the outbreak of WW1. He also shows the breakdown between the types of capital in early 20^th century France, which is interesting. However, he marvels unconvincingly at how “modern” those portfolios look in their diversification. Well, maybe he has other data, but this table being an aggregate, it does not necessarily follow that individual portfolios were diversified. Still, it looks like balanced in the aggregate at least, nothing like bubbles with an investment fad in a single area (think tulips…). 

Something fascinating in this historic perspective is that when countries tried to introduce taxation, at incredibly weak levels, there was a deluge of justifications for not doing so, fears of destitution and so on –the contrast could not be starker with the actual figures. Inheritance tax was even described as totally unnatural, since you should consider that when a son takes over from his father, it is in fact a continuation as if they were one and the same person. Uh?

In general, I don’t think that his main point should be the equation and actual valuation. What seems to have been the case historically is that capital has been able to maintain a high return even as it grew massively, not, I feel, because of the theory of preference for the present (this has been put forward as explaining the returns of capital, but is just plain silly at levels where capital is worth many years of revenue), but largely because capital is able to chase high returns pretty much anywhere, something that labour largely cannot do, and because the wealthy class has been able to influence the political process in ways that the less fortunate have not. Piketty concedes that eventually r must go down, but also that it can take a very long time, which in the meantime would lead to levels of inequality that clash with a democratic society. Since I reckon we are already there in many countries, it is hard to dispute that conclusion.

Chapter 11 looks at the breakdown between inherited and earned revenue.

It is quite a spectacular evolution. Essentially, inheritance collapsed –just collapsed- between 1920 and 1950, and then stayed in the doldrums until 1980. It seems to have had a very strong effect on people perceptions, leading to an impression of the death of inheritance and the rise of a meritocratic society. This is something that got me to reflect much beyond economics. Could that be the source of the feeling that got me to name my blog “Anachronicles”? Is it at least in part that I was merely channelling values from my parents’ (or even grandparents’ –my parents were born in 1949 and 1951) childhoods and feeling the dissonance with my own time? It may be that the “trente glorieuses” were glorious not just because of the GDP growth, but because they had a purpose and values that were more appealing than what followed (and certainly more appealing than what preceded them –but that’s not exactly a hard bar to clear).

But inheritance came back with a vengeance. News of its demise had been greatly exaggerated. In general, this chapter is excellent in the clarity with which it breaks down the various components driving the evolution of the “inheritance flow”, and presents data on each of them, which helps get a very convincing view of where it’s likely to head absent an effort to act upon it. As throughout the book, anytime Piketty is in descriptive mode, he is simply awesome.

Nevertheless, one point in the chapter felt unconvincing. Piketty describes the “reset” at the end of the war (due to policy decisions more than actual physical destruction of capital, so for France at least we get a better understanding of the drivers of the drop in capital that was mentioned in the first part). He then uses that to explain the Modigliani triangle in 1947 –the only time actual data somewhat approached one, namely the only time when the age groups just before or just after retirement really had more capital than the others (a genuine Modigliani triangle would have capital drop down to zero for the older groups, that was not the case even then). He reckons that it was because most capital had been wiped out, and that elder generations did not have the chance to “re-accumulate”, using the power of compound interest in the process.

This would seem to explain why as soon as 1960 we were back to an “ageing capital”, which has kept ageing since, but for 1947 it is unconvincing. A general drop in value of capitals should have affected every generation in a similar fashion, and for people who died in 1947 there would not have been enough time for younger generations to make it up from later savings. So something else must have happened also. I would reckon that rentiers did not really reduce their spending during the war, even though a diminished capital no longer provided them with the returns to fund them, whereas people who had wages would have adjusted in proportion to the inevitable drop. If that is true, it would give us an important insight as to how various classes react to a crisis.

Moving further, Piketty discusses (to refute it) the idea that this rent on capital would be due to a market imperfection, and argue that, to the contrary, it is a natural market outcome that is reinforced by perfect markets, since perfect markets would mean that each marginal unit of capital would go to its most productive use (or, at least, to an efficient one, where efficient is defined as an indifference curve of risk and returns, plus any other factor that one may have, such as ethics).

There are several implicit ideas in that statement and I will take them in turn as I believe that there is a fundamental inaccuracy in them which is likely to drive us to the wrong kinds of root causes for the indisputably real situation that he has so painstakingly described.

First, what he claims is the definition of “perfect markets” is actually perfect market for investing. But the fact that there is an opportunity for a certain level of return on investment may well be due to a monopoly situation –in fact, it is almost always, to an extent, linked to that. Big business success stories invariably have a level of monopoly, usually as the main ingredient, but if not, at least as a seasoning. So more competition in the business markets may well drive down the rate of return, something James Galbraith, Dean Baker, Robert Reich, Mark Thoma and others have argued.

But the crucial point is when he explains that, however uncomfortable one may be with the idea, it is only natural that, as long as capital is a factor of production, it should earn a return.

Yes, sure. But that is not the same as earning a positive revenue in Piketty’s definition. Remember, throughout the book, he has defined revenue as net of any depreciation –that is, even depreciation in market price. Yet nothing in a perfect market ensures that a factor of production will get a return over and above what is needed to maintain its own value. It is plainly not the case of the major input factor, labour: plenty of people work yet do not earn enough to maintain themselves in good health. And that is, of course, ignoring the 100% depreciation that we must all, eventually, encounter.

The only minimum level of return, as far as I can see, that market forces seem to ensure, is the rate of added depreciation. That is, if producing will accelerate depreciation of the capital, then at any lower level of earning the capital would be better left idle. But this is a far cry from a constant return of 5% over and above any depreciation.

This would have a major impact on the prospects of capital in a world of secular stagnation (as Keynes foresaw, calling it “the euthanasia of the rentier”) –and as I have written before, it should be clear by now that we have long reached a situation in which our capital and technology enables us to produce far more than we can meaningfully consume, whether from lack of demand, dearth of inputs or environmental constraints. In a world awash with capital and facing stagnation or even recession, it is not obvious why capital should command a clearly positive net return. Could it be that Piketty is facing the irony of another major treatise (Malthus, Pareto spring to mind) published just before the key relationship it highlights ceases to hold?

Well, I don’t expect it to happen just yet, because rentiers put a hell of a fight. This is where the r>g, or even more precisely the kind of stable r is more of an empiric observation than an economic one. Throughout most of history, the wealthy have managed to influence things so that they got a certain level of return on capital. As such, governments that were created along great hopes of enlightenment have shown a distinct preference for protecting the right of property rather than any of the other professed ones. Recently, it is hard to believe that the waves of deregulations were never influenced by those who had the most to gain –safe in the expectation that, if things were to go pear-shaped, governments would make them whole.

So I would differ on the causal analysis with Piketty. However, I would not yet be sanguine enough to differ on his projection over the next few decades. He describes that as being “absent new policies to mitigate”. In my view, it’s rather “absent a break in the trend of adding new policies to maintain the return on capital”. But it feels likely enough at the moment.