Piketty has written a data-driven book on economic history, which is accessible to a broad audience. The book is full of graphs (even more so than this review) and involves calculations and equations in its arguments. In my opinion, this makes it a better work of scholarship than most other books about history.

As with most (all?) books on economics, there is a political bias. What bias? Well, this is a book with the word “Capital” written in big red letters across the cover.^[1]A clear reference to Das Kapital by Karl Marx (1867). It has been politically influential: Piketty convinced Elizabeth Warren to advocate for a wealth tax.

The power of capital, relative to income, fell dramatically during the World Wars and Great Depression. Since then, capital has been making a comeback. Piketty expects this trend to continue. The growth of capital is inherently faster than the growth of income in a capitalist society, except during major crises like World Wars. To counteract this, Piketty proposes a global tax on wealth.

Since reading Capital in the Twenty-First Century, I have become aware of the existence Poverty and Progress by Henry George (1879). I would be interested in hearing a Georgist take on the same data. Is there a similar book from a Georgist perspective: a Poverty and Progress in the Twenty-First Century?

---

Prerequisites: Some basic economics knowledge would probably be helpful, but not required. Thinking with equations and graphs is necessary.

Originally Written: August 2020.

Confidence Level: This is mostly not my ideas. I have no expertise in economics, but I know how to read data and technical arguments.

Most of this is a summary of the book. My comments / questions are in italics.

---

---

Before I begin this book review, here is a question to think about:

Consider all of the wealth in your country: all of the stocks & bonds, all of the value of factories & machinery, the price of all of the housing & agricultural land. Now think of the total income of everyone in your country, including both wages and earnings from investments.

How many years of income would it take to earn all of the wealth of your country?

Guess a number and write it down.

Piketty calls this number the “capital to income ratio” and it plays an important role in his book.

---

Since the word “Capital” is in the title, it is important to understand how Piketty uses this word. Capital means wealth – of any kind. Capitalism means that the people with the most wealth dominate – not people with the highest wages (these might be the same people, but often in history, the wealthiest thought it beneath them to work for their income), not people with the broadest support, not people who win games of social propriety, not the most moral or religious people, or anything else. Wealth is what matters. In America, the most common meaning of “capitalism” is laissez-faire or free market economic policy. That is not what Piketty means. To Piketty, state capitalism is not a contradiction – it means that the government’s main goal is to increase its own wealth.

---

Piketty’s greatest strength is his ability to gather historical data from many sources – from income tax returns to Jane Austen novels. There is an online technical appendix, if you want to go through the details of how this works. More books should have an online technical appendix.

His historical sources allow him to estimate economic statistics far earlier than other authors (since the 1700s, not just since the 1950s), so he can paint a much bigger picture.

---

The book is divided into four parts:

• The first part provides some basic background on economic growth.
• The second part looks at the history of the capital to income ratio for developed countries.
• The third part, the longest, looks at inequality.
• The fourth part contains policy proposals.

---

---

Part I: Introduction

Part I addresses basic information and history of economic growth.

---

We’ve already defined capital and the capital to income ratio, which Piketty calls $\beta$.

Income has two sources: income from capital and income from labor.

The fraction of total income which comes from capital is called $\alpha$.

Piketty’s First Fundamental Law of Capitalism is $$ \alpha = r \beta \ , $$ where $r$ is the rate of return on capital.

This is true simply by the definitions of the quantities involved: $$ \frac{\mbox{capital income}}{\mbox{total income}} = \frac{\mbox{capital income}}{\mbox{capital}} \ \frac{\mbox{capital}}{\mbox{total income}} \ . $$

---

Exponential growth is incredible. A 1.5% growth rate means that the economy is 35% bigger after 30 years and 2.7 times as big after 100 years. A 2.5% growth rate means that the economy more than doubles in 30 years and is ten times as large after 100 years.

---

Piketty claims that all economies since the industrial revolution at the technological frontier have a per capita growth rate of 1.5%. There are two ways of seeing higher growth rates:

• Population growth adds to the total growth of the economy. Birth rates are falling throughout the world, and are low in the developed world, so this will contribute less in the future.
• After a war or gross mismanagement, catching up to the technological frontier can give larger growth rates. The growth rate slows significantly once all of the known technology is being utilized.

My initial reaction is that this is implausible. Wouldn’t the growth depend on things like science funding, venture capital availability, the fraction of individuals who embrace risk, and how readily the culture adopts new things? There is little room for individual agency, cultural influences, or policy impact.

---

Inflation was effectively zero during the 1800s. Authors wrote what their characters’ incomes were explicitly, assuming that these numbers would still make sense decades later. Most developed countries went through a period of high inflation, if not hyperinflation, during or immediately after the world wars. Since then, central banks have maintained a low but steady level of inflation.

---

---

Part II: Capital / Income Ratio

This was the best part of the book. It looks at the ratio of capital to income for hundreds of years in developed countries.

---

We now answer the question I posed at the beginning of this review. My initial thought was: maybe we own one generation’s income in wealth – so about 50 years. Then, I remembered economic growth, and that wealth was accumulated using the income of the past, and revised my estimate to maybe 20 years. I was totally wrong. If you guessed that the total wealth owned in developed countries is equivalent to 5-8 years of income, then congratulations ! You have an good intuitive understanding of how much capital we have.

I guess this means that Piketty taught me that we are significantly less capitalist than I thought, which is definitely not his goal.

---

Piketty divides capital into 4 types: agricultural land, which was dominant historically, but is much less so now; housing, which has become increasingly important recently; other domestic capital, which is primarily the country’s industrial and commercial capacity; and net foreign capital, which is the amount of wealth in other countries owned by this country’s citizens minus the amount of wealth in this country owned by other countries’ citizens.

Piketty does not believe that “human capital” is a meaningful concept. It is just another way of measuring income due to labor. Comparing income due to capital vs income due to labor is better than comparing capital to “human capital”.

While capital is measured using stocks and bonds, they do not contribute to the national capital stock themselves. They are an asset to one individual but a liability to another. As long as both individuals are in the same countries, these cancel out. Owning foreign stocks and bonds contributes to the net foreign capital.

---

Without any further exposition, here are the graphs of capital vs time for France & Britain. The horizontal scale is time, with nonuniform spacing. The vertical axis is the capital to income ratio. 700% means that the total wealth is worth 7 years of national income.

France’s & Britain’s patterns are similar, so I will discuss them together.

In 1700, both France & Britain had very capital-intensive economies. Most of this capital was in the form of agricultural land.

Over the next 200 years, the total capital / income ratio stayed the same. This does not mean that the total capital stayed the same. It means that both grew extensively at the same rate.

What the capital was used for changed. The importance of agricultural land declined dramatically and was replaced by other domestic capital – particularly industrial capital in Britain.

The amount of foreign capital also increased during this time, because of colonialism. It is interesting to note that the earlier French & British colonial empires in North America did not contribute significantly to their nation’s capital. Even at the height of colonialism, most of the wealth of the biggest colonial empires was domestic.

Strangely, the French Revolution had little effect on the total amount of capital in France.

Between 1910 and 1950, capital suffered tremendous shocks: two world wars, inflation to allow governments to escape their war debts, and decolonization. The amount of capital fell from about 7 times national income to less than 3 in both countries.

Crises have a disproportionate impact on capital. People living paycheck-to-paycheck have a few bad years, and then are back to where they were before. People living off of interest can lose the wealth collected over generations in a few years.

I’ve heard the exact opposite from other sources. Crises tend to most affect the most vulnerable people in society, while wealthier people can ride them out. Does anyone know why there is this difference? Perhaps it’s because Piketty is looking more at total wealth and other authors look more at human suffering.

Since 1950, capital has been making a comeback. Although capital is not as dominant as it was before WWI, Piketty expects that it will reach that level and surpass it in a few years.

Most of the recent increase in capital has been due to housing. Housing had previously been the portion of capital that changed the least. From 1700-1950, the total value of housing remained about 1 year of national income or slightly less. Now, housing is worth about 3 years of national income.

Capital in agricultural land has continued to fall. Other domestic capital has increased slightly. Net foreign capital is close to zero, and sometimes negative. Other countries own more of France & Britain than France & Britain own of other countries.

---

Next, we’ll look at Germany:

There are only a few things different between Germany and France & Britain.

The data starts at 1870 because Germany did not exist before then. The borders of Germany have changed much more significantly than the borders of France or Britain, so care has to taken to make sure that this is a fair comparison.

Germany had a much less significant colonial empire than Britain or France. Net foreign capital was much less significant immediately before WWI. Now, it is close to zero, but positive, unlike France & Britain. Income grew faster than capital in the decades before WWI.

If we ignore colonialism, this is actually true in France & Britain too.

Capital has not made nearly as much of a comeback as it has in France & Britain. The value of housing has only doubled, instead of tripled, relative to income since 1950. Piketty also mentions that stock prices tend to be lower in Germany because the board of directors has to include representatives from labor groups, environmental groups, and the local government. Since stock doesn’t give full control over a company, it has a correspondingly lower price.

---

Finally, we’ll look at the US, which is totally different:

The first thing to notice is how steady the capital to income ratio has been. It has been between 4 and 5 years of income for almost the entire history of the country. I don’t think this is because there have been fewer changes. It’s because there have been multiple diverging processes that approximately canceled each other out.

It also means that the US is less capitalist than France or Britain – at least using Piketty’s meaning of capitalism.

The value of agricultural land has always been less than in France or Britain. Although the US has significantly more land, the existence of the frontier kept land prices extremely low. The lower price more than compensated for the larger land area.

Slavery accounted for a significant fraction of the total wealth of the country – about a quarter of the total wealth of the country and almost half of the wealth of the South.^[2]This is the only way to really measure “human capital” – the amount a person could be sold for in a slave market. This “wealth” stopped being counted as wealth between 1850 and 1880 when people stopped being property.

At the same time, other domestic capital was increasing dramatically as the US industrialized. This increase (0.5 → 3 years from 1770 to 1910) was even more dramatic than in Britain (1.5 → 3 years) or France (1.5 → 2 years).

There were shocks between 1910 and 1930, more due to the Depression than to war, but they were much smaller than in Europe.

Capital has not grown relative to income nearly as much since 1950. Piketty explains this because income (and labor income inequality) has grown more in the US than in Europe.

Lower housing prices almost certainly also play a role. Notice how the expansion of housing capital has been much slower.

Net foreign capital has been small throughout the history of the US and is currently slightly negative.

---

Along with the individual histories, Piketty has a bit more theory: what he calls the Second Fundamental Law of Capitalism.

In the long run, the capital to income ratio depends only on the economy’s growth rate $g$ and savings rate $s$: $$ \beta = s / g \ . $$

The savings rate describes how income is converted into capital. A higher savings rate means more wealth in the long term. If there is a high growth rate, wealth accumulated in the past, when the economy was smaller, becomes less important. So a higher growth rate means that there is less wealth relative to income (not less wealth in total) in the long term.

This equation comes from a simple differential equation which Piketty discusses in the online technical appendix, but not in the book. I spent an afternoon messing around with variations of the differential equation on my whiteboard.

The Second Law of Capitalism is less exact than the First – it’s only true in equilibrium, not at every time. It does help to explain the differences between developed countries. Japan has a low growth rate and high savings rate, so it has more capital. The US has a high growth rate and low savings rate, so it has less capital. Europe is in between.

Marx followed similar logic (without the differential equation to make it precise), but assumed that economies don’t grow over the long term. If you set $g=0$, then that implies $\beta \rightarrow \infty$. Capital will accumulate forever and labor will become increasingly undervalued, inevitably leading to a proletariat revolution.

---

---

Part III: Inequality

Before we look at the dynamics of inequality, it’s important to understand why inequality is a problem. Piketty does not do this at all, which limits the audience who will appreciate the book.

---

I can think of 3 reasons why inequality might be a problem:

• Inequality is bad because poor people are too poor.

Most economic statistics are reported for the average or median person. So if the average person is doing well, we declare that all is well. However, the people who could probably use help the most are not average. They are the very poorest. Focusing entirely on average measures can mask the suffering of the very poor. Including inequality gives us more information about the very poor.

We should care for the poor. But there is a more direct way of measuring this. Instead of measuring the average, then looking at inequality, we could just look at how the bottom 10% are doing / have done directly. Caring about inequality is only a substitute for when that data is unavailable.

• Inequality is bad because rich people are too rich.

Envy drives human behavior more than a desire for comfort. People are happier, not when they reach a certain standard of living, but when they have more than other people. High inequality is bad because it means that there will be more envy in society.

There might be evidence indicating that this is the case. Even if there were, I do not find this a morally compelling argument. We should not combat inequality because we want to satiate the envy of the masses.

• Inequality is bad because it leads to injustice.

Although economics is not a zero sum game, there are some interactions that are important to us which are.^[3]A zero sum game is any interaction in which no one can gain anything except by taking it from someone else. A positive sum game is any interaction that can leave both sides better off – or the … Continue reading Inequality in wealth often leads to inequalities in other areas. For example, the wealthy can gain more political influence then a typical person, which makes a democratic government more aristocratic.^[4]It can also make a monarchy more aristocratic and lead to political systems with broader support, as seen repeatedly in British history. This undermines our notions of justice and equality before the law.

This certainly can be a problem. It’s not immediately clear to me when the better solution is to reduce inequality and when the better solution is to make the other areas of our lives more resistant to the influence from wealth.

---

Piketty seems to care most about (2). That is not to say that Piketty comes across as a particularly envious person. He seems genuinely concerned for the envy of humanity.

This is not to say that Piketty is completely unaware of (1) and (3). But they get less than 5 pages each in a book that’s over 600 pages long.

There are multiple chapters discussing the income and wealth of the top 10%, 1%, and 0.1% in various European countries, the US, and Japan for the last few hundred years. He makes a point of using percentiles because no other measure can be used consistently across such varying societies.

In the middle of this, there are 3 pages on the minimum wages of France and the US. He even neglects to mention that minimum wage is not relevant for the same percentile of workers in the two countries or in different years.

I was hoping that some of this would be corrected in Ch. 12, Global Inequality of Wealth in the Twenty-First Century. However, by “rest of the world”, Piketty means the Forbes list of billionaires and the sovereign wealth funds of the oil states and China.

I understand that developed countries and wealthier individuals have much better data and so more can be said about them. But the level of neglect extends far beyond a lack of data. The poorer half of humanity has no place in Piketty’s analysis.

---

What do I think should be in the book that isn’t here?

Piketty looks in detail at how much the wealth and income of the top 10% have changed over the last 200 years. I would be interested in seeing how the wealth (likely negative) and income of the bottom 10% have changed over the last 200 years.

This analysis should be done both within developed countries and between countries. How much more or less wealthy is the average person in Nigeria today than 200 years ago? India? Indonesia? China? Brazil? This is what I was expecting from Ch. 12, not a look at the global elite.

---

---

Part III Again: $r > g$

If this book review is to follow Piketty’s book, I should not spend Part III discussing why we might care about inequality.

---

Part II compared wealth and income. But having a society with lots of wealth doesn’t immediately imply that the society is extremely unequal.

In every society we’ve seen, the distribution of wealth is more concentrated than the distribution of income.

To give the reader some intuition for this, he takes Scandinavia in 1970, Europe today, the US today, and Europe in 1910 as examples of a relatively equal society, a moderately unequal society, a highly unequal society, and an extremely unequal society. He compares the income and wealth of the bottom 50%, the next 40%, the top 10%, and the top 1% for each society.

For some reason, he decides to use the same average wealth for each society, which is completely not historical. I think there is an implicit assumption that the same growth rate (1.5% + population growth) would have occurred, regardless of how equal the society was.

---

Before WWI, in Europe, wealth was almost entirely inherited. Only the very highest paying jobs could provide an income comparable to the poorer end of the elite. This caused a rigid social structure – unless you were born with wealth, you had almost no chance of gaining a significant amount.

In contrast, in the developed countries with the largest income inequality today (the US and Britain), most of the income inequality results from extremely high paying jobs. These jobs didn’t exist 50 years ago because the top tax rate was 80-90%, which would have effectively confiscated it.

Even though we have lots of capital and inequality, it has not created a class of people living off of their ancestors’ investments. Piketty thinks that is just because the wealth hasn’t been accumulated for long enough yet. Once the next generation inherits the wealth of today’s entrepreneurs and CEOs, they will form a class that lives entirely on interest.

Much of the wealth is also held by organizations, such as university endowments and sovereign wealth funds, instead of by individuals. Instead of having people who inherited their own wealth, we now have people whose positions allow them to control tremendous institutional wealth.

---

Piketty thinks that increasing wealth and inequality will continue.

This is a consequence of the Fundamental Inequality of Capitalism: $$ r > g \ , $$ where $g$ is the growth rate again and $r$ is the rate of return on capital.

The rate of return you get for wealth is always greater than the economic growth rate, so the size of large fortunes grows faster than the economy as a whole. Piketty estimates the typical return on investment to be about 5% — and larger for the largest fortunes. The growth rate for an industrialized country at the technological frontier is 1.5% + population growth. With population growth being small and $5% > 1.5%$, we expect wealth to grow faster than income.

Piketty provides historical data for $r$ and $g$, extending all the way back to Antiquity. The only time when the economy grew faster than wealth was during the crises between 1910 and 1950.

I do not trust the earlier data for this at all. I suspect that $r < g$ during other crises as well – for example, during the fall of the Western Roman Empire.

There are several caveats.

One is that, in order to have your wealth grow at $r$ instead of something less than $r$, you have to reinvest almost all of your interest and dividends. If you instead spent most of your interest and dividends, your wealth would not grow at all. Piketty thinks that, while this is important for smaller fortunes, the amount of money involved with the largest fortunes is too large to be spent.

I find this implausible. Wealth can be spent in remarkably large quantities – whether to hire dozens of servants or to build the mansions of the Gilded Era in the US. Many of the wealthiest individuals also decide to spend their fortune on wildly expensive projects not related to their own consumption – like eradicating malaria or building a Mars colony. Even if it is possible to grow your fortune indefinitely, I doubt that many people would.

Another caveat is that crises destroy wealth, as we saw between 1910-1950. Piketty points out that the only time we have seen capital declining in the past is because of these terrible events.

I’m not entirely convinced that the historical data support this either. We did have r < g between 1950 and 2012, which Piketty explains as a consequence of higher tax rates. But the graph makes r > g look much less inevitable if you cover up the suspect data before 1700 and the projections after 2012. This data shows that for the hundred years prior to this book, the inequality has not been satisfied.

There’s also the question of why the capital to income ratio fell for Germany between 1870 and 1910 and why it would have also fallen in France & Britain during this time without colonialism. If capital were in relative decline during this time too, then I’m not sure when Piketty thinks that $r > g$.^[5]I’m sure that there are many academic articles looking at $r$ vs $g$ which I am totally unaware of, so perhaps these issues have been resolved somewhere. But Piketty has not shown it here.

If $r$ is not inherently greater than $g$, then the main point of the book, that large fortunes will accumulate indefinitely without government intervention, does not have a sound theoretical basis.

---

---

Part IV: Policy

Given that we have a fundamental disagreement on legitimate reasons why someone might want to reduce inequality, it should come as no surprise that I don’t agree with all of his policy recommendations.

---

Piketty, unsurprisingly, advocates for much higher income and inheritance taxes.

---

He has mostly negative views on government debt. Historically, when the government paid decent interest rates, it was a way of transferring more wealth to the (predominantly) rich people who owned government debt. Now that it often pays a lower interest rate than inflation, Piketty doesn’t think it’s as bad. But even now, a panicky bond market can abruptly insist on higher interest rates.

---

Piketty’s main policy recommendation is for a global tax on capital. If this is impossible, then at least the US and EU should establish wealth taxes and work harder to find and eliminate tax havens.

I completely disagree with his argument.

To make this clear, I will contrast his argument with Elizabeth Warren’s arguments for a wealth tax. Even if you disagree with all of her details, at least she has the right structure for the argument:

• Here are some things that the government can spend money on that would improve the lives of the poor.
• The best way for the government to raise this money is to tax wealth.
• Therefore, we should have a wealth tax.

In contrast, Piketty’s argument is structured as:

• Capital has a tendency to accumulate.
• Wealth should be democratically controlled.
• Therefore, we should have a wealth tax.

Piketty’s emphasis is more on limiting the wealth of the rich than on what else could be accomplished with that wealth.

---

This is one of those arguments which only sounds convincing to people who already agree with you.^[6]This type of argument is extremely common, but I think it’s terrible. Arguments written for a broad audience should try to be persuasive even to people who disagree with you.

If you already believe Piketty’s second point, then you probably already support substantial economic redistribution. This book might convince you to support a wealth tax in addition to a higher income tax or nationalizing industries.

If you don’t already believe it, then Piketty’s arguments are entirely unconvincing. And so, this book is one that gets incredible reviews among people already on the Left, but doesn’t actually convert anyone.

---

If I were to take Piketty’s premise that wealth inequality is a major problem and the impressive data sets that Piketty shows, here is what my policy recommendations would be:

• Large inheritance tax.

Piketty supports this, but mostly says that it’s not enough. Large fortunes would still exist because many large fortunes are created by lucky entrepreneurs who then reinvest their wealth. If you (unlike Piketty) distinguish between earned and inherited (or stolen) large fortunes, then one is clearly less just than the other. A large inheritance tax would make sure that we do not recreate what existed in Europe before WWI: a class who lived lavish lifestyles, despite never working a day of their life.

• More affordable housing.

Although Piketty doesn’t discuss this in depth, his data show that the main contributor to the modern growth capital is housing. If we want to reduce the power of capital, we have to reduce the cost of housing. Although this appears to be a bigger problem in Europe, it also has become significant in many American cities.

This is also an area where there could be significant bipartisan cooperation. Adam Smith, in The Wealth of Nations, also criticized high rents:

As soon as the land of any country has all become private property, the landlords, like all other men, love to reap where they never sowed, and demand a rent even for its natural produce.

---

---

Despite disagreeing with much of the book, I am glad that I read it. It is the kind of book that you can productively disagree with. Piketty works more like a scientist than most historians or economists writing to a large audience. He relies heavily on the data and then tries to draw general conclusions from them. Although the author can have a biased interpretation, emphasizing the data helps others to see what’s happening and draw their own conclusions.