Inequality: The Gap That Obsesses, the Progress We Forget

What the Inequality Debate Fails to Measure

In France, wealth is viewed with suspicion. This is a cultural fact before it is an economic one. For the past twenty years, the wealth tax (ISF, then IFI), inheritance taxation, “tax justice,” and “ultra-rich” have structured public debate as a series of totems — with one constant: the conviction that the gap between the top and the bottom of the distribution is, in itself, the problem to be solved. The work of Thomas Piketty and Gabriel Zucman did not create this conviction. They gave it an academic framework, long-run data series, and a formalized structure. They made it unassailable — or at least difficult to challenge without being seen as a defender of privilege.

It is precisely this difficulty that makes the exercise necessary.

Let us begin with what their critics rarely do. Piketty and Zucman are right on several empirical points. Wealth inequality has increased in most OECD countries since 1980. Legal tax avoidance mechanisms — stepped-up basis in the United States, split ownership structures and Dutreil pacts in France, foundations everywhere — allow a fraction of ultra-wealth to structurally escape inheritance taxation. Any critique that denies these facts loses credibility before it even begins.

But between these findings and the policy conclusions drawn from them lies a chain of choices — of measurement, framing, and values — that is rarely made explicit. It is this chain that this op-ed seeks to dismantle.

What the Debate Confuses

The inequality debate constantly blends three distinct registers. Empirical facts: wealth concentration has increased. Methodological conventions: how do we define the income of the ultra-rich? Values: which indicator of social progress should be prioritized — reducing relative gaps, or raising the living standards of the least well-off?

Piketty and Zucman measure with remarkable precision the gap between the first and the last runner in the race. The problem is that they look almost exclusively at that gap, without ever checking whether all runners have moved several kilometers forward. The result: normative choices presented as scientific findings. This slippage is not unique to their camp — liberals often do the same in reverse. This op-ed will try not to fall into the same trap.

The r > g Framework: Solid, but Conditional

The relationship r > g — the return on capital structurally exceeding growth — is the cornerstone of Piketty’s argument. Per Krusell (Stockholm) and Anthony Smith (Yale) showed in the Journal of Political Economy (2015) that this dynamic depends on a crucial assumption: an elasticity of substitution between capital and labor greater than 1. If this elasticity is below 1 — as most empirical estimates suggest — concentration converges toward a finite limit. The Piketty dynamic does not disappear, but it ceases to be inevitable.

This is not a minor caveat. It is the condition for the validity of the entire framework. A result conditional on contested assumptions is not a law of nature. It is a thesis — and presenting it otherwise is excessive.

Zucman: A Real Problem, a Misleading Measure

One of Gabriel Zucman’s most widely cited findings is that the 400 wealthiest Americans pay, proportionally, less tax than the middle class. This figure relies on including unrealized capital gains in income. If your assets rise by ten billion this year, Zucman counts ten billion as your income — even if you have sold nothing, received nothing, and that amount could collapse tomorrow.

Zucman’s argument is serious: these gains represent real economic wealth that can durably escape taxation, through mechanisms such as stepped-up basis in the United States or split ownership in France. But this argument concerns a long-term structural advantage, not the current effective tax rate. The Congressional Budget Office, which excludes such gains, calculates effective tax rates for the top percentile at around 30%. The two methods answer two legitimate but different questions. Confusing them produces striking headlines and flawed diagnoses.

Relative Inequality: Effects That Cannot Be Ignored

Piketty does not merely measure gaps; he argues that relative inequality has independent causal effects on health, social cohesion, and democratic quality. The literature is substantial: Wilkinson and Pickett (The Spirit Level, 2009) documented robust correlations between inequality and public health indicators; Marmot established social gradients in health that persist at equal income levels; Gilens and Page suggest political capture of institutions by the very wealthy.

These works are not without weaknesses — sensitivity to samples, causality still debated: are unequal societies unhealthy because they are unequal, or are inequality and social pathologies both symptoms of a deeper institutional deficit? Yet they are too serious to be dismissed outright. This op-ed does not claim that relative inequality has no effects. It argues that these effects, real but debated in magnitude, are not sufficient on their own to justify the systemic fiscal overhaul proposed by Piketty and Zucman.

Three Trajectories, One Question

Consider Bangladesh. Between 1990 and 2020, the country reduced infant mortality by 85%, increased girls’ school enrollment to 98%, and multiplied its per capita income fivefold. Its relative inequality increased slightly. Applied to this case, Piketty’s indicator would signal deterioration. No serious observer would accept that conclusion.

The objection is valid: Bangladesh is a developing country. Take South Korea. Between 1980 and 2020, GDP per capita increased tenfold and life expectancy rose by fifteen years. The Gini coefficient moved from 0.33 to 0.35. Judged by inequality, Korea in 2020 is less just than in 1980. Judged by lived reality, it is incomparably more so.

Even in France, between 1990 and 2022, the living standards of the bottom 10% increased by more than 30% in real terms (INSEE), while the Gini coefficient rose slightly. The gap widened; the floor rose.

Looking only at the gap without looking at the floor is a mutilation of the diagnosis. But looking only at the floor without looking at the gap is another. The choice between these perspectives is a value judgment — not a scientific result.

What This Op-Ed Assumes

To assert that the absolute condition of the most vulnerable should take precedence over relative gaps is a moral choice, not an empirical discovery. Liberal critics of Piketty often commit the very sin they denounce: presenting their framing as neutral rather than as a value system just as partial. This op-ed acknowledges its choices. It does not claim they are objective.

It also acknowledges this: the tax avoidance mechanisms documented by Zucman — stepped-up basis, split ownership, instrumentally used Dutreil pacts — are real problems. Recognizing that Piketty’s overall diagnosis is overstated does not imply defending the fiscal status quo. That would be another form of intellectual dishonesty — and perhaps the most costly one, because it hands Piketty exactly the opponent he needs to be right.

Three Proposals

Correct inheritance-related tax asymmetries. Transferring wealth without taxing accumulated capital gains is an unjustifiable advantage. In France, the combination of split ownership, the €100,000 allowance renewable every fifteen years, and Dutreil pacts allows significant business wealth to be transferred at very low effective rates. This is not a wealth tax. It is the correction of an asymmetry that even Piketty’s critics should acknowledge.

Complement relative indicators with measures of absolute poverty. The European poverty threshold, set at 60% of median income, has a perverse property: if median income falls during a recession, the threshold falls as well, and people can become materially poorer without the indicator capturing it. An absolute poverty threshold, indexed to a basket of goods, would provide a more honest picture of lived conditions.

Demand methodological transparency. Any study on effective tax rates should specify whether it includes unrealized capital gains and present both figures. Any study on inequality should present both relative and absolute indicators. If these requirements were applied, half of the headlines about taxing the ultra-rich would need to be rewritten — and half of complacent liberal op-eds as well.

Conclusion

The problem is not primarily that some people are very rich. It is that too many people are still too poor. Confusing these two diagnoses means designing policies to address the wrong problem — with real costs for those they claim to protect.

The inequality debate lacks neither data nor indignation. It lacks clarity about what is being measured, in the name of which values, and toward which objectives. This clarity is not a technical matter. It is a democratic requirement. It applies to all sides — including the one writing these lines.