Daron Acemoglu on Inequality, Institutions, and Piketty
What this covers
Daron Acemoglu and host Russ Roberts examine whether grand theories of capitalism's inevitable trajectory—particularly those of Marx and Thomas Piketty—hold up under scrutiny. Acemoglu's central claim is that such "general laws" fail because they strip away institutions, which actually determine how technology and markets translate into inequality outcomes. The conversation moves through three domains: why Marx's prediction of perpetual wage stagnation proved wrong; why Piketty's r > g thesis is economically unremarkable and lacks empirical support; and how institutional shifts—not capital dynamics—drove inequality swings in twentieth-century Sweden and apartheid South Africa.
Acemoglu distinguishes between three separable problems: static inequality itself; equality of opportunity and social mobility; and political inequality. He argues the top 1% focus obscures rather than illuminates genuine distributional concerns, since apartheid South Africa saw the elite's share decline even as it engineered one of history's most brutal hierarchies across every institution. On labor markets, he notes that gains from competition flowed to workers only where institutions—franchises, mass education, competitive employers—counterbalanced employer power; where coercion persisted (slavery, indenture), wage rises never materialized regardless of marginal product of labor. The discussion also addresses why Piketty confuses empirical correlation with causation, finding no support when one actually checks whether periods of r > g correlate with rising inequality, and how social mobility weakens the claim that capital returns automatically concentrate wealth across generations.
Acemoglu argues that grand 'general laws of capitalism' (Marx's tendencies, Piketty's r>g) systematically fail because they abstract away from institutions and their endogenous evolution, which actually mediate how technology and markets translate into inequality outcomes.
- Marx's predictions failed because the political/social institutions of his era (feudal power balance, no franchise, no mass education) changed, reshaping wages via human capital and competitive labor markets
- Piketty's r>g is economically expected, doesn't imply rising inequality, omits institutional determinants, and lacks correlational/causal empirical support
- Cases like Sweden and South Africa show inequality swings driven by institutions, not r-g or capital destruction
Institutions, not competition or technology alone, determine how economic gains are distributed between capital and labor.
- Competition only shared gains with labor because of the specific labor-market and political institutions of that era, which were the exception in human history; where labor coercion existed—such as slavery in Jamaica, Cuba, and Haiti—large swings in the marginal product of labor (e.g., rising sugar prices) did not translate into higher wages, demonstrating that the equitable distribution of growth depended on institutions evolving together with quasi-competitive markets, not on competition alone.
“you had sugar prices fluctuate a lot and improve and increase the value of marginal product of labor of slaves in Jamaica and Cuba and Haiti. That did not translate into increases in the wages of labor.”
- Marx's prediction of stagnant wages failed because the political and social context changed: the spread of mass education dramatically increased the supply of skills during the 19th century, and an institutional environment that supported continual technology creation and the scaling-up of competitive employers (rather than protecting incumbents) raised the demand for skilled labor, so that demand for labor and skilled labor together kept pushing wages up.
“the social and the political context determined the two most important factors shaping wages: human capital and technology.”
- Economic outcomes are not determined by technology and preferences alone; institutions—government actions, bargaining, coercion, rent-seeking by powerful groups—put supply and demand together, and crucially they have a life of their own, changing in response to the distribution of political power, technology, and political demand, which makes simplistic laws of capitalism's evolution untenable because they ignore both the impact of institutions and their endogenous evolution.
“Government actions, actions by powerful groups in society through non-market means--bargaining, coercion, rent-seeking--all of those are in general part of this ensemble that put preferences and technology together to generate the outcomes that we care about. And broadly speaking we would refer to these as institutions”
- Whether unions are good or bad for an economy cannot be answered universally because nobody actually wants a competitive labor market—everyone prefers to be a monopolist and form coalitions—so it depends on which side of the balance you are on: where employers are monopsonistic and workers powerless (as in late-18th-century Britain, apartheid South Africa, or with abusive child/safety conditions in early-20th-century America), labor organization is important to balance power; where unions are excessively strong (as in parts of Continental Europe or late-1990s Portugal), they push for wages that curtail innovation and harm other workers and capital.
“I cannot answer the question as to whether unions are good for the functioning of the economy or not. It really depends on what side of this tradeoff we are”
- Whether reducing inequality requires limiting or increasing government power depends on the specific historical context: in cases of direct exploitation like apartheid, slavery, or feudalism, government power is the main method by which the wealthy extract from the rest, so limiting it helps; but in the Gilded Age of robber barons, trusts were so economically powerful and able to use the legal system that crushing opposition required a stronger government to perform antitrust regulation—which the Progressives achieved—and is why Marx's prediction of increasing concentration did not pan out.
“if you are thinking of the Gilded Age... then we have a system there where these trusts are so powerful economically and are so able to use the legal system that they crush any opposition... what you need to do is, you need to do standard antitrust regulation. But to be able to do the antitrust regulation, you need the government to become stronger.”
- Marx held a materialistic conception in which the economic base determined everything else, so although political factors mattered they were shaped by economic currents; from this framework he derived strong tendencies—the general law of capitalist accumulation (wages stagnant or labor's share declining), the general law of declining profits, and the general law of decreasing competition toward monopoly capitalism.
“he had a conception in which the base, the economic relations, determined everything else.”
Piketty's r-g framework lacks empirical support and misses institutional determinants of inequality that actually drive outcomes.
- Piketty's own data from the US and France clearly show that the increase in inequality over the last 50 years is not a capital phenomenon but a labor-income phenomenon, yet he argues that despite this, in the future it will be capital that dominates—a claim for which there is no evidence, constituting a sleight of hand between his stated emphasis on capital and what his data actually demonstrate.
“his data from both the United States and France quite clearly show that the increase in inequality is not a capital phenomenon. It's a labor income phenomenon.”
- Piketty (like Marx) does discuss institutions such as minimum wages, wealth taxes, and social norms regulating super-manager pay, but these are not an integral part of his core theory; his major prediction—that the gap between the interest rate r and growth rate g determines the capital share and inequality—has institutions making no appearance, and where they are mentioned they are an outside construct rather than something that is itself changing and needs to be explained.
“the major force of his approach, which is that it's going to be the gap between the interest rate and the growth rate that is going to determine both the capital share and inequality in society... I think that's where institutions make no appearance.”
- Piketty's 'central contradiction of capitalism' is misnamed because r>g is implied by economic efficiency: in growth theory, economies where r is less than or equal to g are called dynamically inefficient because savers are not properly compensated and the economy generates too much capital, so that everyone could be made better off by consuming some of it; thus r>g is exactly what a market economy should and generally does generate, not a newly discovered fault line of capitalism.
“we call economies that do not satisfy r > g 'dynamically inefficient' economies... if r were ever less or equal to g... then you could actually make everybody better off in the economy without making anybody else worse.”
- Piketty does a fantastic job unearthing informative tax-return data on inequality, but he violates the empirical literature's standard practice of testing claims through correlations and regressions and then worrying about causality; instead he lays out striking patterns and asserts they are explained by r-g or the World Wars without ever doing cross-country, over-time, or cross-household correlations—and when one does a simple correlational check, periods where r exceeds g do not show increasing inequality.
“he then takes the next step, and he says, Well, here are the patterns and these are explained by r - g being positive... But how do we know that? ... He never takes that step.”
- Focusing on r-g leaves out far more important institutional determinants of inequality, as illustrated by Sweden and South Africa: both saw huge 20th-century swings in inequality, none of which is related to r-g or to wars and the destruction of capital, but all of which is deeply institutional—the formation of the Swedish social welfare state and the apartheid regime in South Africa that held black wages down while redistributing to whites, then its collapse.
“there are huge changes in inequality in both of the countries over the 20th century. But none of it is related to r - g. None of it is related to wars or destruction of capital. So, they, are all deeply institutional.”
Inequality's real policy significance lies in opportunity, mobility, and political voice rather than top-one-percent share alone.
- Concentrating solely on the top 1% misses crucial dimensions of inequality; in apartheid South Africa—one of the most unequal societies—inequality was about whites versus blacks and coloreds across every institution from marriage to education to political representation, all designed to keep blacks poor, yet the top 1% share actually declined during apartheid, so looking only at the top 1% would misleadingly suggest the elite was weakening.
“top 1% share declines throughout the period of the apartheid. So, if you just looked at the top 1% you'd say, well, this economy is doing well; the elite must be getting weaker over time.”
- A rapidly rising top 1% share is worth worrying about not for its own sake but because it likely signals how dominant the economic elite is becoming politically; if the top 1% share doubles (e.g., 15% to 30%) and that money flows into politics, allowing figures like Sheldon Adelson to finance candidates with specific economic agendas, society can be increasingly dominated by a small fraction—so the real concern is the political and institutional resilience and fault lines, not just the tax-return numbers.
“if the top 1% is becoming very, very rich at the same time as somehow... money becomes more and more dominant in politics. That, you might actually worry about it”
- If the real goal is equality of opportunity and social mobility, then top-1% inequality is largely irrelevant: taking money from Bill Gates and Warren Buffett and giving it to millionaires so the top 1% share declines does nothing for disadvantaged blacks in inner cities, Hispanics, or the least-educated poor, so the right focus is poverty and the opportunities available to people at the bottom of the income distribution.
“Are we going to make equality of opportunity for blacks in our inner cities... if we take some money from Bill Gates and Warren Buffett and give it to millionaires, so the top 1% declines? Of course that's totally irrelevant.”
- One should distinguish three separable judgments about inequality: first, the static welfare judgment about whether a more unequal allocation is bad holding all else equal (most people are somewhat against it, though increasing inequality also changes incentives); second and more important, equality of opportunity—whether being born into a rich family guarantees the top position regardless of talent or effort, which greater inequality can worsen via credit-market constraints that block the poor from schooling; and third, political inequality—whether a handful of people dominate political voice while the rest are 'like sheep.'
“it's very useful to distinguish three roles of inequality. Or, three sorts of judgments that one can, and should, have about inequality.”
Labor market vulnerability arises from weak bargaining position and coercive intermediaries, not from competitive wage-setting.
- Even in a labor market that looks competitive, vulnerable workers (e.g., non-English-speaking cleaners brought by a head contractor) may be in a weak bargaining position because they fear illegality or retaliation, so an intermediary can capture much of the payment—charging the client $100 while paying each worker $6—and better safety laws or protections could move their wages closer to their true marginal product, producing a more efficient and liquid market.
“the $100 that you are spending actually gets captured by the intermediary. And in fact, if there was better safety laws, perhaps these ladies would be compensated for doing some of the difficult job”
- A positive r-g does not imply increasing inequality because not everyone is a capital owner all the time—people save for life-cycle reasons or to pass wealth to lower-earning children—and social mobility (both intergenerational and within-lifetime, such as being a high earner in your 40s then living off savings in your 60s) breaks the link between the return on capital and where capital owners end up, so the connection between r-g and inexorably rising wealth inequality is tenuous at best.
“social mobility breaks this link between return to capital and where the capital owners are going to be.”
- The common belief that the US standard of living exceeds Mexico's because of the minimum wage is very hard to sustain empirically; comparing very different economies like Mexico or Peru to the US involves much more important first-order factors (institutional and technological parity), and minimum wage only becomes a good explanation for wage differences between economies with substantial institutional parity, such as France versus the US.
“before you can get there you need to have a much greater institution parity... when you are comparing Mexico to the United States or Peru... There are much more important first-order factors.”
- Even granting that capital income is very unequally held—with the top 1% holding perhaps 40% of capital stock—the key fact is that membership in the top 1% changes over time, since a large fraction of today's top 1% were not in it 30 years ago, which undermines the claim that returns on capital relentlessly compound the wealth of a fixed group of capital owners.
“Who is in that top 1%? If you look at who is in that top 1%, that changes over time. The people who are in the top 1% today, many, a large fraction of them were not in this top 1% 30 years ago.”
Piketty's book is important and influential but conflates academic empirical claims with persuasive rhetoric about fairness.
- Piketty's book is undeniably important given its enormous impact, selling millions of copies and transforming and opening new debates about inequality.
“How could anybody say that it's not an important book when you look at the impact it has had?”
- Piketty's book is genuinely important and tapped a deep public angst while building on decades of inequality research, but it suffers from a tension because he tried to do in one volume both the academic work of advancing empirical comparative-statics claims and the popular work of appealing to readers' sense that something is deeply unfair—so it becomes hard to parse which statements are scientific claims about the effect of x on y and which are persuasive appeals made even when the data is not there.
“Thomas tried to do both in the same volume. And I think there is a tension in there; and I think it's not clear what part of it is a scientific, academic statement... and what part of it is an appeal to his readers”