As with all critical reviews, I am likely to spend more words talking about weaknesses in the argument than about the strength of the book, so let me say upfront that it is a very important work, where not only Piketty but several other researchers working with him spent many years collecting data sets bridging very different time periods across several countries. This alone is invaluable, and the foundation for much (though admittedly not all, as Galbraith pointed out) of the current studies of inequality. To that he has added a considerable effort to present it with clarity, very much including notes about how a graph should be read. People accustomed to doing their own analyses may find some of the explanations long-winded (although even experts will make silly mistakes more often than they'd like), but the upside is that it is readable by pretty much anyone willing to spend the time -and considering the subject, I feel that this is an important thing. Also, Piketty keeps sending the reader to an online database if he wants the details, while acknowledging that he is presenting a somewhat simplified description. That is very good practice. All in all, I would very much praise his work (certainly not limited to this book), and recommend reading this book.

General remarks

Piketty writes a rather long introduction which helps understand what he has been trying to do. In it he mentions that it is as much a history book as an economics one -and I feel that he is right. He is on the right track with his remarks, very early in his career, that too many economists (especially English speaking ones) are far too infatuated with complex numbers and forget that economics comes from social sciences. Indeed, he repeatedly uses novels as evidence of what societies looked like, especially when we were short of data. I have some sympathy for that, and do not begrudge drawing from works of fiction (which would also help many readers who are more acquainted with Austen or Balzac than Smith, Ricardo, Marx or Keynes), though I feel he overdoes it somewhat at times, such as when he brings up Cameron's Titanic. Titanic, whilst a commercial success that many of his readers (not I) would have seen, is not a contemporary work and thus may say as much about our fantasies of the period as its reality.

As often when the reviews are so extreme, I feel that the reality may be somewhat less so. So, while I understand Galbraith's criticism, I feel that it comes out as much more negative than deserved. Others have even criticised him for not being revolutionary enough, considering the needs of the times. Well, he is writing an academic study about one issue only (for instance, while he mentions environmental constraints, that is not the crux of his work, though arguably the strongest reason why the current system needs to be overhauled), and not a political pamphlet.

Conversely, while this has been deemed the most important economics book of the decade, such a statement may lead to unrealistic expectations. Certainly, it is a stronger history and economics book than purely an economics one. Some of the confusions that inevitably derive from methodological choices (some of which may have been unavoidable) are not addressed as much as you'd expect with such a tag. I feel that this was what had Galbraith annoyed, as well as his not having been mentioned as much as he would have liked (and maybe should have been).

Introduction

The introduction is a great read, and whoever feels the book is too long should at least read it. It also describes methodological choices that will then be used throughout the book. One of them is highly important as I feel that it will have some of the strongest impacts on the conclusions.

In essence, throughout the book, Piketty will call Capital everything that can be owned AND value it at market value. To be fair to him, he does mention that it is not ideal as markets can be moody. Yet, he says -and I have some sympathy for that- what else can you do, how else can you add hectares, bonds, patents and plants? That is true, provided you want to give a value to the sum of capital, which may not always be needed. But also, I feel that, while he repeatedly notes that market effects explain some of the variations he shows, he probably does not quite take enough account of that when deriving his conclusions.

On the revenue side, Piketty uses the net national revenue, which is a modified GNP, to take account of the fact that some of the GNP must be used to fight off capital depreciation, and thus is not a revenue at all. He tries to be consistent with that in his analysis, although you do find several instances in the book where he seems to omit depreciation and cost of ownership in his estimation of the return of capital. It is important to note that there is in that no distinction between revenues from work or from capital.

Much of the quantitative analysis (there is, to be fair, a lot of qualitative one too) that follows tracks the ratio of the value of capital to net revenues. Piketty states that this allows much better analyses than what had been often done in the past, which was to look at the breakdown between capital and labour in the sources of revenue, and says that this will be made clear. One third into the book, I still have not seen a clear explanation of why this gives better insights (it might be in the technical appendix, of course). Certainly, this view suffers more from the problems associated with using market values, and with the lack of distinction between the sources of revenue (although it has its advantages, as it gives an idea of how much capital accumulation there has been).

Looking at long-term evolutions, Piketty mentions forces of convergence and divergence among revenues. He does mention that some of the forces of convergence are theoretical -but does not highlight quite how weak their cases are. In particular, he mentions the view that production modes would require increasing competencies (call centre operators may debate the point, as would many skilled craftsmen whose work has been replaced by robots), which should increase the proportion of revenues from labour.

But that is not how compensation systems work: a durable unemployment and weak unemployment benefits will force very highly skilled people to work for very little pay, provided that all they have is their skills. Compensations are far more the result of power plays than of the importance of the contribution of any given factor. Of course Piketty knows that, and does mention that evidence does not back the case for convergence. Still, I feel that even the theoretical case is very weak.

Piketty briefly talks, during his introduction, of the notion of fair inequalities, and provides a decent definition. In that I understand that he is echoing Rawls, although not by name. This is fine, and I would find myself in much agreement, but I feel it is worth pointing out that the level of optimum fair inequality is not necessarily different from zero, at least for some measures of inequalities.
I would reckon that it is not quite the case, but I also notice that many of the arguments in favour of higher inequalities tend to be very weak and based on very questionable assumptions.

However, zero inequality is by essence impossible, if only because it cannot be achieved for all measures at the same time. For instance, would you look at income or disposable income? Do you adjust for specific needs (like being responsible for children) and if yes how? Do you adjust for local cost of living (which is not always a choice. Some people have careers which force them to live in major cities)? Do you take an instant view, or rather look at a lifelong progression? Those different perspectives will give a different result, and even full equality in one of them will mean inequalities in others. Piketty is of course right to suggest that what makes sense is not a straight refusal of inequalities, but that they be based on common utility.

Before closing the introduction, Piketty mentions where he comes from, as a child of the early 70s. He tells us that, having come of age on the year that the Berlin wall fell, he is immunised to all the simplistic rejections of Capitalism. That is fine, as long as the argument cuts both ways, and as long as elaborate criticism is acceptable (to be fair to Piketty, I am pretty sure that he agrees to that). Yelling "Russia" is itself a simplistic rejection of Communism (the case would be stronger for rejecting an immovable system), and even more so of a role for the state in the economy. Yet I see that far more often in the media. It is telling that, despite intellectual shortcuts being much more frequent on the other side, he should feel the need to point this out. I look forward to seeing academics starting their books by explaining that, having come of age during the Reagan-Thatcher years (or, in Piketty's case, just before the Shock Doctrine descended upon Russia), they are immunised against simplistic support for Capitalism.

Chapter One - Revenue and Production

A historical overview of returns on capital in chapter one suggests quite a few examples gravitating around 5%. Among them, Piketty shows the cost of renting an apartment in Paris, relative to its selling value. I find this rather objectionable, as ownership of an apartment (or house) involves significant costs, akin to the depreciation of any equipment. Everything I've read and experienced points to a cost of around 1% of the value of the property where land itself is an important part of the price (such as Paris), and higher elsewhere. This would be an important adjustment to the reported value of housing capital, and I believe that underestimation of those costs is one of the reasons leading to many overestimating the comparative advantages of ownership.

Similarly, later in the book, Piketty states that a 5% return on slaves (yes, slaves were capital at some point) would give them an average price of 20 years of average wages, while they were cheaper. But you had to house and feed the slaves you owned, plus they would not have lived forever (OK, if they were to have kids then I guess you owned them so the calculation would become actuarial). So such a price would probably not have been consistent with a 5% rate, another example where the stated aim of using net revenue may not have been followed.

Chapter Two - Growth: Illusions and reality

In his chapter 2, Piketty states that strong population growth will be a strong equalising factor, everything else being equal. It does not seem to me to follow necessarily.

Strong population growth can either be internal, or from immigration. The case is probably stronger for internal growth, as estates would have more inheritors. However, while this may mean that most people would need to rely on something else than inheritance, this would not necessarily reduce inequalities in any way. For instance, if the first born inherited the whole estate, the effect would rather be to increase inequalities.

As for immigration, I fail to see why that would reduce the importance of capital necessarily. Most immigrants probably have little wealth, but their arrival, all things being equal, would tend to depress wages while increasing the demand for the use of capital (most obviously housing), thus increasing its market value, which after all is what Piketty tracks. A similar case can be made for the impact of economic growth (as a factor watering down inequalities due to capital), which would indeed help workers accumulate savings more easily, however this would also translate in pushing up the market value of existing capital, especially where it cannot be meaningfully increased: landowners would not own a lesser proportion of the land, however high the growth may have been.

The thing is, Piketty's metric will tell us some things, but not all. You can have a lowish ratio of capital to net yearly revenue, but concentrated in very few hands. In fact, if very few people have a chance to accumulate capital (these days, it would probably require a level of capital control, although that has not always been so), this is likely to be the case as market rates would be low. But that would not mean a less unequal society, in fact probably the opposite. Looking at the ratio of capital to median disposable income, for instance, would probably give you very different insights. Then looking at its concentration would go even further. Maybe he covers that in the third part.

Chapter 3 - The transformations of Capital

It is essentially in passing and far from central to his thesis, but figures that Piketty presents in this chapter really should be highlighted in the public discourse.

You see, he talks about the breakdown between public and private capital. And then takes a look at the net position.

So, first, it should be clear how ridiculous the claims that France is a fundamentally statist society are. Private gross capital is over 4 times bigger than public gross capital (in net terms, the ratio is one to twenty, but I guess the gross figure is more relevant). That despite having a mostly public school and health system. And public housing. And a lot of historical buildings.

The other thing is that, while France has a little bit (not much) more debt than the UK, the other country having long historical records, its net position is positive by around 30% of GDP, and on a growing trend showing little interruption (in parallel with the debt, so the net position has not changed in the past few decades). The UK, on the other hand, currently has a zero net position, and its assets have at best been moving sideways since 1920, or even gone slightly down since 1950. Surely looking at both is a more relevant picture than merely looking at the debt side of things.

Chapter 4 - From Old Europe to the New World

Piketty, after spending a long time on France and the UK (for historical reasons), extends his analysis to other countries, such as Germany, USA and Canada. There I feel that the confusions inbuilt in measuring capital by its market value start to have a rather strong impact on the conclusions.

He touches upon it -noting that the drop in the relative value of capital after the bloody 1914-1945 period (and stagnation for some time after that) far exceeded the physical destructions. He does mention the effect of the drop in market price as being important, although he concludes that volume effects were more important (around 30% and 70% respectively). And he does mention, but maybe too quickly, the political choices at the time which restricted "capital" formation, although he really means in terms of ratio to revenue, as of course capital increased very fast during the boom years in Europe -only revenue did, too. Yet, as I know from reading reviews of the book, there will be little said in the rest of the book of the major weight of political choices in boosting capital relative to revenues, especially relative to earned revenues.

Piketty highlights that, despite the colossal growth of capital in Germany after 1946 (probably greatly helped by the cancellation of public debt, though he makes no mention of that), the level remains lower than in France or the UK, being just over 4 years of net national revenue in Germany, against 5.5 to 6 in France and the UK. He does mention that Germany did not have a housing boom, but claims that most of the difference comes from other factors. I find it reconcile that with his figures: housing capital is shown to be close to 3 years of national revenue in the UK, close to 4 in France where the bubble never popped, and just over 2 in Germany. Moreover, high real estate prices will also mean higher prices for housing used for professional purposes, very much so in heavily service-centric economies such as those of France and the UK.

This is not a trivial point. The intensity of capital ownership is implicitly used as a proxy for inequalities, but among types of capital, housing is probably the most evenly shared (though far from equally, of course), and also one that cannot be easily sold when it is the place where you live. So, to get an impression that Germany is a much more equal and meritocratic society (actually, poverty rates have shot up) because, largely due to its population going down, it has lower housing costs would be highly misleading.

Piketty also notes that, should you use the value of the capital owned by company instead of their market values, then the "paradox" disappears. This would be in addition to the housing price effect -should we thus conclude instead that German has a higher ratio of capital to revenues? We can see that using market value rather than physical capital has major implications for the analysis.

Much the same is the case for USA and Canada, though instead of it being due to a dwindling population, it is because they are so damn big, so that only in the zoned areas will property prices be able to go up much. They too have around 2 years of revenue in housing capital, and just over 4 years total capital. They are still "young" countries, and immigrants typically have next to no capital, so that the capital that they do have is probably more concentrated than in Europe, a point that is yet to be addressed in detail at the point where I am in the book.

That will be it for now - as I said, my main aim is to trigger discussions, so hopefully they will be forthcoming.