The Three Broken Laws of Development
For half a century, three propositions felt like physics: richer countries emit more carbon, have fewer children, and grow more unequal. Each is now true in fewer places than you'd think.
There are propositions you grow up assuming are true because every dataset you've ever seen seems to confirm them. They feel less like findings and more like physics — patterns so reliable they need no defending. In the social sciences, three such propositions sat near the top of the list for most of the twentieth century. Richer countries emit more carbon. Richer countries have fewer children. Faster-growing countries grow more unequal.
Each of these still describes most of the world today. But each has stopped describing some of it — in ways that, taken together, suggest something larger is happening. This essay walks through the three, one at a time, using the same underlying dataset: 335,022 observations across 194 nations, mostly from the World Bank and Our World in Data, all openly licensed. You can follow along in the live lab linked from each figure.
The Law of Carbon
In 1990, every country that got richer also burned more fuel. Not anymore.
Until quite recently, the relationship between national wealth and national emissions looked like physics. You could plot every country's GDP per capita against its CO₂ per capita for any year in the twentieth century and the cloud of dots would slope upward in a stubborn, near-perfect diagonal. Get richer, burn more. The pattern was so reliable it was used as the foundation for emissions forecasts well into the 2000s.
Then it broke. Quietly, in the United Kingdom, first. In 1990, the average Briton was responsible for 10.2 tonnes of CO₂ each year — squarely average for a wealthy European. By 2022, that figure had fallen to 4.8 tonnes. Meanwhile, UK GDP per capita had grown from $33,000 to $53,000 in constant prices, a 61% rise. A Briton today is substantially richer than in 1990, and emits less than half the carbon. The same pattern, with slightly smaller margins, has appeared across most of Western Europe and parts of North America. Climate economists have a flat, antiseptic word for it: decoupling.
| Country | CO₂/cap 1990 | CO₂/cap 2022 | Δ CO₂ | Δ GDP/cap |
|---|---|---|---|---|
| United Kingdom | 10.2 t | 4.8 t | −53% | +61% |
| Denmark | 10.2 t | 4.8 t | −53% | +50% |
| Sweden | 6.7 t | 3.6 t | −46% | +56% |
| Germany | 12.7 t | 7.9 t | −38% | +51% |
| France | 6.6 t | 4.5 t | −32% | +39% |
| United States | 20.0 t | 14.4 t | −28% | +64% |
| China | 2.1 t | 8.9 t | +317% | +1,190% |
| India | 0.7 t | 2.0 t | +184% | +290% |
What did these economies do, between 1990 and now? Three things, roughly in this order. They de-carbonised their electricity — coal was replaced first by gas, then by wind, solar, and (where it survived) nuclear. The UK closed its last coal plant in 2024; Spain crossed 50% renewable electricity in 2023. They terciarised their economies — a pound of British GDP in 1990 came largely from steel, automobiles and chemicals; a pound today comes from software, finance, design and care. They priced carbon, slowly — the EU's emissions trading system, dormant for a decade, became materially expensive from 2018 onward.
One thing should be said clearly. The decoupling story is a story about a minority. Of the 194 nations tracked here, only about thirty have reduced their per-capita emissions while continuing to grow. The rest — most of Asia, all of Africa, much of Latin America — still sit on the old diagonal. China's emissions per capita have more than tripled since 1990. India's have nearly doubled. The global average has barely moved.
The law of carbon is not abolished. It is conditional. It now applies to most of humanity but not to the wealthy core.
That is a real shift — large enough that the IPCC, the IEA, and the World Bank have all reorganised their forecasting frameworks around it in the last decade. But it is not yet a victory. It is a proof that the law was never physics. It was a habit. And habits, it turns out, are easier to break than the equations once suggested.
The Law of Fertility
Wealth was supposed to lower the birth rate predictably. The 2020s have produced a much weirder map.
The demographic transition is one of the most reliable findings in twentieth-century social science. As countries get richer, women get more education, infant mortality falls, and the average number of children per woman drops from six or seven to around two — replacement level. The curve was so predictable that demographers used it to forecast world population for sixty years.
The world has now produced two large embarrassments to that curve. The first lives at the rich end. South Korea, with a GDP per capita of $53,200, has a total fertility rate of 0.78 — the lowest ever recorded in peacetime. Singapore is at 1.04 despite a per-capita income of $133,600. Spain, Italy and Japan all sit between 1.16 and 1.30. None of these countries was supposed to reach below replacement. They were supposed to settle at replacement. The convergence floor has fallen out from under the transition curve.
The second embarrassment lives in the middle. Israel, with a GDP per capita of $48,100 — richer than France or the UK — has a fertility rate of 2.89. That is higher than any other OECD country, and higher than many middle-income ones. France and the United States, at 1.78 and 1.66 respectively, are also stubbornly higher than their European peers despite similar wealth. There is no economic explanation that fits the whole map; the determinants are cultural, religious, and policy-driven in ways the original curve never captured.
| Country | GDP/cap 2022 | Fertility 2022 | What it tells us |
|---|---|---|---|
| South Korea | $53,200 | 0.78 | The new floor |
| Singapore | $133,600 | 1.04 | Wealth ≠ births |
| Spain | $46,700 | 1.16 | Southern Europe collapse |
| Italy | $52,300 | 1.24 | — |
| United Kingdom | $53,100 | 1.56 | — |
| United States | $72,700 | 1.66 | Below replacement |
| France | $53,700 | 1.78 | Stubbornly high |
| Israel | $48,100 | 2.89 | The exception |
| Niger | $1,700 | 6.13 | The other extreme |
Meanwhile, at the other end of the income distribution, the transition is also misbehaving. Niger's fertility rate is 6.13 children per woman. Chad's is 6.21. Mali's is 5.69. These are not anomalies — they are the central tendency for the Sahel, and they have been falling much more slowly than the curve predicted. Niger's GDP per capita has crept from $1,300 to $1,700 over twenty years, and its fertility rate has dropped by less than one child. The transition is happening, but on a much longer timeline than the model expected, and from a higher starting point.
A single global "transition curve" is no longer a reliable description of the world. There are two ends pulling apart in opposite directions.
What you see in 2022, if you let the lab settle on the GDP-versus-fertility view, is no longer a single curve. It is two clusters: a low-income, high-fertility African cluster around 5-6 children, and a high-income, sub-replacement Eurasian cluster between 1 and 2. The middle is hollowing out faster than anyone predicted. Demographers now talk about a bifurcating transition, and the consequences — labour shortages in one half of the world, demographic dividends in the other — are about to reshape geopolitics in ways the old curve never anticipated.
The Law of Inequality
Growth was supposed to raise the Gini. In half of Latin America, it has done the opposite.
Simon Kuznets's 1955 hypothesis was that as economies industrialise, inequality first rises, then falls. The data of the postwar era seemed to confirm it, and through the 1980s and 1990s the dominant narrative in development economics ran in one direction: growth lifted everyone, but it lifted some faster than others, and the Gini coefficient tended to drift upward in the rich world. By the 2010s, Thomas Piketty's Capital in the Twenty-First Century had codified the finding that, since roughly 1980, inequality has been rising almost everywhere in the OECD.
Look at the table below. The OECD pattern is real. Sweden's Gini coefficient has climbed from 24.3 to 29.3 since the 1970s — a 5-point rise. The United Kingdom is up 5.6 points since 1968. The United States is up 5.1 points since 1963. Germany is up 4.4 points since reunification. There is no serious dispute that the welfare-state economies of Western Europe and North America have become measurably more unequal across the past two generations.
| Country | Earliest Gini | Latest Gini | Change |
|---|---|---|---|
| Chile | 56.2 (1987) | 43.0 (2024) | −13.2 |
| Brazil | 57.9 (1981) | 50.3 (2024) | −7.6 |
| Mexico | 48.5 (1984) | 42.6 (2024) | −5.9 |
| France | 37.1 (1970) | 31.8 (2023) | −5.3 |
| Ecuador | 50.5 (1987) | 45.9 (2025) | −4.6 |
| United Kingdom | 26.8 (1968) | 32.4 (2021) | +5.6 |
| United States | 36.7 (1963) | 41.8 (2024) | +5.1 |
| Sweden | 24.3 (1975) | 29.3 (2023) | +5.0 |
| Germany | 29.3 (1991) | 33.7 (2022) | +4.4 |
But look at the top of the table. Chile's Gini has fallen by 13.2 points since 1987 — the single largest reduction in inequality recorded by any country with continuous data. Brazil's has fallen by 7.6 points. Mexico's by 5.9. Ecuador's by 4.6. Argentina's path has been more turbulent but ends roughly where it started. These are not statistical artifacts. They reflect real, sustained policy choices — Brazil's Bolsa Família conditional cash transfers, expanded after 2003 and now reaching 50 million people; sharp real-terms rises in minimum wages across the Southern Cone; the commodity boom of 2003-2014 that disproportionately raised low-wage incomes; and Chile's post-2019 social pact, accelerated by the constitutional process of the early 2020s.
The OECD has drifted upward; Latin America has drifted downward; the two movements have, by 2024, brought parts of Latin America closer to OECD inequality levels than to their own historical norms. Chile at 43.0 is now less unequal than the United States at 41.8 is far from. Brazil at 50.3 has met Mexico halfway, and Mexico has met Argentina.
Inequality is not a function of growth. It is a function of policy.
That sentence sounds polemical, but the data is unforgiving. If the Kuznets curve were a law of nature, Chile and Sweden could not be moving in opposite directions. They are. The single most powerful predictor of where a country's Gini will be in 2030 is not its growth rate; it is whether it spends the political capital to compress its income distribution. That is a finding the development field has been slow to absorb, and it is the most important of the three this essay has tried to show.
What the three laws share
The three findings here are easy to misread as good news. They are not. The carbon decoupling story applies to perhaps 30 countries out of 194. The fertility map has produced collapsing populations as often as flourishing ones. The Latin American inequality decline has slowed since 2014 and remains far above European levels. None of these laws has been repealed. They have only been shown to be conditional.
But that distinction is the important one. When something looks like physics, it implies fatalism — emissions must rise with wealth, families must shrink as countries develop, inequality must follow growth. When something is shown to be conditional, it implies politics: there are conditions under which it holds and conditions under which it does not, and a country can move between those conditions deliberately. Britain decarbonised on purpose. Korea hit 0.78 children per woman because nobody designed a working family policy. Chile reduced its Gini because a generation of governments prioritised it.
The map of development in 2026 is no longer dominated by inevitabilities. It is dominated by decisions — visible, dated, attributable, country by country. That is the harder picture, and the more honest one. The lab linked from this essay is, in the end, an attempt to make those decisions visible at a glance, so that the next generation of them can be argued for on the basis of what has actually happened, not what was once believed must.