Sunday, 20 April 2014

Further notes on Piketty's Capital

I published a first set of notes on the book a little while ago. I still have not finished it, but am now much further. It's still an excellent read that I would recommend to anyone.
Some of the slightly odd trends are maintained, but also the further developments answer several of the questions that came out of the first chapters.


Chapter 5 - The Capital/Revenue ratio in the long run
Some quite revealing observations there -and it is of a great benefit to have timed data, rather than just a snapshot of the current situation, as we can compare dynamics between countries.

Piketty discusses Germany, and states in passing that the German trade surplus is due to “the alea of German competitiveness” and that we should thus expect it to go down over the near future. This will come as a complete contradiction to the analyses that we tend to read on European Tribune (in particular Migeru’s very detailed description of the imbalances). I must say that I find those analyses (and mine) more convincing that this mere statement that it’s about competitiveness –unless of course that term is used to convey something else than I understand. Certainly, it is not competitiveness due to productivity gains. If by competitiveness we mean the effect of actively depressing domestic demand in a fixed exchange-rates system, then I guess we can reconcile the two views. However, I very much doubt that this is what the word will bring in mind to most readers. In particular, it is clear that it is very much a consequence of political choices, and this is not, I felt, what seemed to come out of Piketty’s brief explanation.
Things may change in the rest of the book (although I know that some reviewers have mentioned it as a general tendency), but I keep getting the impression that structural causes are overplayed while the consequences of political choices are not highlighted as much as their impact would suggest.

In any case, much of what he presents is quite revealing, and does give indications of the relations between the capital to revenue ratio and policy choices:

-Private capital shot up in the UK by a level that savings could not explain, apparently mostly because privatisations happened at very low prices (same with Russia in the 90s). This would be, to a lesser extent, a factor in France and Italy as well. Hard to believe that it was not by design. This, remember, is central to Naomi Klein's Shock Doctrine. Something to keep in mind for forming suggestions to address the diagnosis (something that will be covered later in the book -although I already know from reading reviews what the main ones will be).

-Italy has had a much higher private capital to revenue ratio than comparable countries since the early 90s, mostly because the State (Berlusconi?) chose to borrow from the rich rather than tax them.

-And an extremely interesting line of study (from graph 5.6): Piketty mentions that the price of assets (he seems to mean company assets, and thus exclude financial and real estate ones, which clearly moved a lot) went up, explaining about a third of the rise in capital to revenue ratio. He states that the ratio of market value to book value went up in all 8 main developed countries since its value "in 1970-80".

Actually, though this is true and a very important observation, I find the description somewhat strange. In all countries, the ratio was higher in 1980 than 1970. In all countries but France and Canada, 1980 or 1981 is a low point (Canada starts slowly rising in 1976, and takes off in 1990. France keeps going down until 1984, returns to 1970 value in 1986 with a sharp upward trend, drops again from 1988 to 1995 and takes off again in 1995 -following political changes), after which it takes off.
Note: for France, 1984 is just after the government froze public salaries and dropped the policy of regular raises of the real minimum salary.

Now, the narrative of a major change in trend right at the time of the Thatcher-Reagan revolution is very different from a return to long-term valuation (note: he does not actually says it returns to long-term valuation, although that seems to be the impression left, and his data series starts in 1970) following a vague starting point in the "1970-80" period. I know Piketty has little sympathy for the Thatcher-Reagan model, so what's going on? Is he afraid that he would lose much readership if he made it too explicit?

Chapter 6 - The Capital/Work share in the XXIst Century 
This is a very important chapter.


I had mentioned that thus far, there had been no mention of the Capital/Labour breakdown (or variations in the return on Capital). Well, not only does Piketty bring it up, he also states why he concentrated on the capital value to revenue ratio: the historical (by which he means up to the end of the XIXth century) data is more reliable for the value of capital than for the income it generated.
Well, I can accept that reason, though it clarifies that it is not because it is a more meaningful analysis (as I had been under the impression that he was saying and had difficulties understanding), rather than it makes it more likely to draw a long-term historical perspective -an extremely useful task, of course.

He introduces a difference between average and pure return of Capital, to take into account the time and work spent managing it. I am slightly surprised by the scale of the difference if that is the only thing differentiating them, and indeed Piketty suggests that the cost of managing may be overestimated (and certainly has strong economies of scales).
However, if we must go for “pure” return, we should (theoretically –I don’t know if we have the data to do so at the moment, though if we do, it would probably be thanks to Piketty’s work) add the additional revenue that people get from having high capital. Not only will people be more likely to get a high degree when born in a wealthy family, but they are far more likely to get high-paying roles, if only because many organisations like having someone from a wealthy background on their books, in the hope that they may look kindly on the venture. Never forget that George Bush junior was given a directorship because, and I quote, “his name was Bush”.
As a note: in his introduction to Part 3, which looks into inequalities, Piketty hints that the correlation between high revenue from capital and high revenue from labour will need to be looked at. I have not yet reached the point where he does, although he notes in comments that (as I suspected) it has gone from negative in the early XIXth century to positive now. So he does take that into account, and when it flows from capital to labour, it is a form of capital revenue.

Piketty once again summons the great novelists of the XIXth century to describe the then well-known fact that Capital had a return of around 5% (sometimes, more rarely, 4% appears to be used in their descriptions). I agree that it makes it clear that this is what “everyone knew”, however, having seen so many glaring mistakes in the economic valuations of ownership, particularly when it comes to the decisions of buying or renting, I am rather careful of equating what everyone knows with an actual economic relationship. Though Piketty may have more of a point with that period, since distorting factors like inflation and taxation were much weaker. The apparent return of government bonds would have been rather close to the actual one, although not entirely devoid of risk, as a war was likely to trigger the inflation whose lack made the calculation easier. Even then, bonds would have then been paper documents that you needed to protect from theft at a cost, and certainly real estate needed maintenance. I doubt that novelists would have been spot on, and probably erred in underestimating costs.

Closer to our times, he has and shows robust series. He takes the time to explain why we would expect (and, to a clear extent, we do see it) that return on capital would go down with higher levels of capital. Still, returns have held up rather well. Piketty seems to suggest that this is due to the diversification of types of capital. No doubt it played a part, but I would expect more emphasis on deliberate choices to help it. Companies receiving subsidies would be a start. But maybe the clearer example is that copyrights and patents laws have become more restrictive even as technological advances would have pointed towards more sharing, anti-trust laws have been watered down, and in general more policing is done to protect owners and corporations than the general public.


Piketty notes the distinction between purely financial (and thus, mostly, not inflation-protected) and other assets -which he both includes in his definition of Capital. And it turns out that the higher rungs of capital owners have a much smaller proportion of purely financial assets these days (whereas many rentiers were getting rents from government bonds in the XIXth century, and non-rentiers had nothing anyway). This will be important when the time to discuss solutions will come, as inflation could hit the wrong targets.
Although at this point (maybe later), Piketty does not attempt to distinguish between the various ways of generating inflation. A proposal, like Steve Keene's, of simply crediting everyone with an equal amount (a negative poll tax?) until reaching full employment (or ideas such as the living allowance) would indeed have an inflationary impact, but it would be of a secondary order to the first order effect of boosting the relative position of the lower rungs, and thus hit the right target after all.

In this chapter, Piketty takes the time to refute some old theories (while being quite understanding with how they may have been formulated when data series were much more limited), such as the idea of a fixed capital/labour share of income.
However I fear that he may give slightly too much importance to the β=s/g relation, where β is the capital/revenue ratio, s the savings rate and g the rate of growth. It is, indeed, the only stable state for any set of s and g.

I have learnt to be very mindful of using accounting identities in causal analyses. Apart from obvious objections of noisy series (growth rates and savings rates are not constant, so how could they be the basis of an infinite limit?), which Piketty does highlight, I see further problems:
-This law, first written by Harrod and Domar, related to capital as a production factor, not as a market value, and it's not quite the same interpretation in either case
-Piketty talks rather a lot of what happens to β when g changes, "for a fixed savings rate". Well, not only is it unlikely that there would be total independence between economic circumstances and the decision to save, s in Piketty's work must be understood as net of all depreciation (ie not the person's decision to save, but what's left of savings after you have paid to keep up capital). In that case, the more capital you have, the more depreciation, and thus you'd need more and more savings to keep up. So s should go down when g is reduced.
-Which way does it go? Piketty seems to feel that β shooting up is a consequence of g going down (and indeed, as ideas of secular stagnation hint at, it's very likely that g is going down). But it has a major contracting effect on g: if you don't have enough customers, you don't have demand and, after a while, you produce less. Yes, that's what's been happening for a while, or at least so I'd argue.

Using the identity too liberally would give the illusion of mostly structural factors at play and underplay the impact of policy choices. That is the recurrent feeling from reading that Piketty is showing what's absolutely undisputable but very mindful of pointing out to political causes, even as he reminds us that they are a crucial component. I would guess that he is trying to make sure that he cannot be dismissed. After all, his book does not have to be an end point, and he does give you plenty of material do derive your own thoughts and conclusions.

Anyway, this part two provided a great description of the evolutions of Capital, something that is surprisingly rarely studied for societies living in a system called "Capitalism". The reasons for this silence may well lie in the next part, about inequalities.

Part Three - The Structure of Inequalities
I will not have very much to add or comment on that part, where Piketty is clearly in his element. Actually, since much of my understanding has been derived from his work, it is unsurprising that I would not change much from the descriptive parts.
But I must mention that it is very clearly written, and that, while little may be news to those who have read every Piketty (and Saez) paper that they could find over the past few years, it is illuminating.

In his Chapter 8 - the Two Worlds, Piketty makes the important point (one that pundits like Friedman and Brooks love to ignore) that the top decile really is a mix of very different realities, and even the top centile still is the top 1/1000 and the rest. Actually, the point is so strong that I feel he could have concentrated on the top 1/1000 even more (certainly for English-speaking countries), but I glanced at the rest of the book and see him doing so.

Piketty contrasts France (the longest data series, and a somewhat trending egalitarian country, if not to the 80s Scandinavian level) with USA (which, although the others are a few years behind, you could use for English-speaking countries in general), which is in an entirely new experiment with inequalities for a developed country.
Much has been made by reviewers of the point that the new experience in inequality in the US is a labour revenue phenomenon. However, it appears to be a combination: the first centile of wages received just over 10% of wages in the US in 2010; the first centile of revenues (thus including capital) received 20% of all revenues. For France, the corresponding numbers are around 7% and 9%. It's hard to refute Piketty's description of a general rise in inequality.

There is little to add to his excellent work -and the stagnation of capital revenues (except for housing rental) in France that he showed in chapter 5 were a hint of the gap between the two countries. Piketty also evokes the fact that tax havens have greatly expanded, so that the reality of capital inequalities is probably highly understated by the data (true for both countries, but believing recent articles probably much more so of USA).

Still, one thing I have not seen mentioned but which seems important is the difference in pension systems. As far as I understand it, a pension through capitalisation system would pretty much force everyone to hold capital (even if in the form of Social Security assets in USA), whereas one by repartition such as in France would not. This should have a strong effect of reducing capital inequalities for countries with a capitalised system. That France should nevertheless have significantly lower inequalities thus speaks volumes.

I'll stop this set of notes at the end of chapter 8


No comments:

Post a Comment