为什么有些国家富有，有些国家贫穷？

流行版本

Two dinner-party versions cancel each other out. One: “GDP is all that matters — growth is the only number a government should chase.” The other: “GDP is meaningless — it counts oil spills as progress and ignores everything human.” The first treats a thermometer as the patient’s health; the second throws out the thermometer because it doesn’t measure happiness. Both miss what GDP is for: a measure of market output, which turns out to be tightly correlated with the things the critics say it ignores — clean water, schooling, child survival, the leisure to ask whether money buys happiness.

最强支持论点

The Easterlin paradox — rich countries aren’t reported as happier than middle-income ones — was the empirical spine of the “GDP misleads” case for decades. Then Stevenson and Wolfers (2013) reopened the data with larger samples and found a robust, roughly log-linear relationship between income and life satisfaction that holds at every level — within countries, between countries, and over time. There is no satiation point where more income stops buying reported well-being. If you care about how people say their lives are going, income is one of the strongest predictors you have. The factor-of-100 gap is also a factor-of-many gap in stated life satisfaction.

最强反对论点

Simon Kuznets, who built the national-income accounts that became GDP, warned in 1934 that “the welfare of a nation can scarcely be inferred from a measurement of national income.” The architect told us not to use it as a welfare measure. Marilyn Waring pressed the sharpest version: the accounts systematically erase subsistence agriculture and unpaid care work — the labor that feeds and raises most of humanity — because it never crosses a market. A country can grow its GDP by commercializing what families used to do for free, and the number rises while nothing improves. The measure is not neutral; it sees what is sold.

判断

GDP is an imperfect but indispensable starting point. Track it alongside health, schooling, distribution, and environmental damage, not instead of them — the HDI and its cousins exist precisely to keep the other columns open. But the measurement debate does the opposite of what its loudest users intend: it does not dissolve the puzzle, it confirms it. Every honest revision — informal economy in, environmental costs out, capabilities counted — leaves Norway vastly richer than Niger on every axis a reasonable person would name. The gap is real. Now we have to explain it.

流行版本

Two bumper stickers fight to a draw. “Africa needs more aid — the rich world is stingy.” And: “Aid doesn’t work — it’s money down a hole.” Both treat aid as one undifferentiated thing. It isn’t. A vaccine campaign and a balance-of-payments loan to a kleptocrat are both “aid,” and they do opposite things. The interesting argument is not for or against aid in general; it is about which kind of aid does what, and why the Solow logic that motivates the whole enterprise misfires at the national scale.

最强支持论点

Jeffrey Sachs makes the “big push” case, and it is pure Solow: poor countries are stuck in a low-capital trap; inject enough capital at once — clinics, roads, fertilizer, schools — and you exploit diminishing returns in reverse, kicking the economy onto a self-sustaining growth path. The historical exhibits are real. The Marshall Plan rebuilt a devastated Europe in a few years. South Korea absorbed enormous US aid in the 1950s and 60s before its takeoff. The Millennium Villages Project tried to run the experiment village by village. The logic is airtight inside the model: where capital is scarce, its return is high, so capital should work hardest exactly where it is poured in.

最强反对论点

William Easterly answers that aid mostly funds planners, not searchers — grand top-down schemes that no one is accountable for, captured by elites, generating Dutch disease that strangles exports, and blunting the very pressure to reform that produces growth. Moyo adds that many of the largest per-capita aid recipients saw incomes fall over the decades of heaviest assistance. The Marshall Plan and Korea are the wrong analogy: those were already-industrial societies with skilled workforces and functioning institutions, rebuilding what they had lost — not poor agrarian economies building it for the first time.

判断

Aid works for specific, targeted things — vaccines, deworming, bed nets, oral rehydration salts. These save lives at extraordinary cost-effectiveness, and the case for them is overwhelming. What aid has never done is drive sustained national growth. No country has developed primarily on foreign aid. Korea, China, and Botswana grew through institutional reform, export discipline, and domestic policy — aid was at most a lubricant. The Solow model’s mistake was never algebraic. It was the assumption that capital is the binding constraint, when the real constraint is whatever decides whether capital gets used productively. That “whatever” is the subject of the next three stages.

流行版本

Two popular readings collide. “China proves authoritarianism works — strong leaders deliver growth that messy democracies can’t.” And: “China succeeded despite authoritarianism, not because of it.” Both leap straight to the political form and skip the thing that actually changed. The question worth asking is narrower and harder: what did China do, and is the doing separable from the doer?

最强支持论点

The strong version of the authoritarian-advantage case points to East Asia’s record — Singapore, South Korea under Park, Taiwan, China — and to the weak cross-country correlation between democracy and growth once you control for institutions and education. Daniel Bell pushes it furthest with “meritocratic governance”: a competent, insulated technocracy can plan for decades, override NIMBY veto points, and execute infrastructure at a scale electoral cycles punish. On this reading, the binding variable is state capacity — the ability to set a goal and actually deliver it — and authoritarian states sometimes have more of it.

最强反对论点

The same Chinese system that produced Deng’s reforms produced Mao’s Great Leap Forward — 30 to 45 million dead in a famine engineered by the absence of any mechanism to tell the leader he was wrong. Authoritarian systems have no error-correction: when they get it right the upside is fast, and when they get it wrong there is nothing to stop them. And the growth itself traces to specific institutional reforms, not to the one-party form — household decollectivization that gave farmers the residual, Special Economic Zones that created secure-enough property rights for foreign capital, WTO accession that locked in openness, and a quiet extension of property rights to private entrepreneurs. Those are institutional moves that happened to occur under authoritarian rule.

判断

China’s growth came from institutional content that happened to be delivered under authoritarian rule — not from authoritarianism as such. The replicable part is the content: secure property rights, market incentives, export discipline, openness to foreign technology. The political form is not the lesson; it is the packaging. This is why the ideas stage bridges into the institutions stage. Ideas are non-rival and available, but a society adopts them only when its rules make adoption pay. China rewrote those rules. That is what catching up actually looks like.

流行版本

Two extremes shout past each other. “Colonialism explains everything — poor countries are poor because the West looted them.” And: “Colonialism is an excuse — it ended sixty years ago; blame present governments.” The first collapses a complex, well-identified mechanism into a slogan; the second waves away an effect that the cleanest natural experiments in development economics keep confirming. The real question is how much, through what channel, and what it does and doesn’t excuse.

最强支持论点

The empirical case is unusually tight. AJR’s settler-mortality instrument isolates institutional quality from reverse causation. Melissa Dell’s study of Peru’s colonial mita — the forced-mining-labor draft — finds that districts subjected to it are measurably poorer today, with the boundary visible in household consumption centuries later. Nathan Nunn shows that the regions of Africa most ravaged by the slave trade are the poorest now, with the damage running through shattered trust and broken state-formation. Three independent designs, three different natural experiments, one consistent answer: the institutional scars of extraction persist for centuries. The deepest of those scars, the Atlantic slave trade, is the empirical anchor under Nunn’s result.

The narrower question — did the West get rich because of slavery? — is its own walkthrough: Did the West get rich because of slavery? Here the slave trade enters as one channel by which extractive institutions were imposed and persisted in the colonized world; that walkthrough takes the West-because-of-slavery argument as its central frame.

最强反对论点

Jared Diamond answers that geography came first — the head start in domesticable crops and animals, the disease environments, the accidents of latitude that shaped who could develop and who couldn’t, long before any colonizer arrived. David Albouy challenged the AJR settler-mortality data directly, arguing the mortality figures are noisier and more selectively coded than the headline result admits. And the counterexamples bite: Ethiopia was essentially never colonized and is poor; Botswana was colonized and became one of Africa’s richest and best-governed states. If colonial institutions were destiny, neither should exist.

判断

Colonial institutions left real scars — the AJR, Dell, and Nunn evidence is too consistent across too many independent designs to dismiss. But the story is more complex than a single cause. Colonialism is a major, well-documented source of institutional dysfunction across much of the developing world; it is one cause among several, interacting with geography and contingency rather than overriding them; and it does not relieve post-colonial governments of responsibility for what they built on the rubble. Botswana inherited extraction and chose differently. The scar is real. It is not a sentence.