Macroeconomics › Section G · Economic growth › Topic G.4
Economic convergence
If poorer countries grow faster than rich ones, income gaps narrow over time. Economists disagree about how likely that is, and even optimists agree that catching up takes generations.
Will the poor catch up?
The topic "The recent arrival of economic growth" showed how two centuries of unequal growth opened huge gaps between countries. The natural next question is whether those gaps will close. Economic convergence is the pattern in which poorer economies grow faster per person than richer ones, so that their incomes move closer together. It is one of the most debated questions in economics, because the answer shapes the lives of billions of people. This topic sets out the arguments for and against convergence, looks at the evidence, and works through why even successful convergence is slow.
Two kinds of convergence
Economists distinguish two versions. Absolute convergence is the claim that poorer countries grow faster than richer ones regardless of their other features. Conditional convergence is the narrower claim that an economy grows faster the further it is below its own long-run potential, which depends on its saving, schooling, institutions and policies. Two countries with very different institutions need not converge to the same income at all. The evidence fits the conditional version better. Convergence is clear among broadly similar economies, such as US states over the twentieth century or the countries of Western Europe after 1950, but far less clear across the whole world.
Argument for: diminishing returns
The first argument rests on diminishing returns to capital. If technology stays the same, each extra dollar of capital per worker adds less output than the dollar before it. A worker with no tractor gains enormously from the first one; a worker who already has three gains little from a fourth. Poor countries have little capital per worker, so each new investment should raise output a lot; rich countries already have plenty, so their gains are smaller. In the Solow model, developed by Robert Solow in 1956, this pushes poorer economies to grow faster as they build up capital, until they approach a steady state where capital deepening alone stops raising output per worker. A higher saving rate raises the steady-state level of income, not its long-run growth rate.
Worked example: diminishing returns
Suppose output per worker in a hypothetical economy is \(y = 200\sqrt{k}\) dollars, where \(k\) is capital per worker in dollars and technology is fixed. 1. At \(k = 10{,}000\), \(y = 200 \times 100 = 20{,}000\). 2. At \(k = 20{,}000\), \(y \approx 28{,}280\), a gain of about 8,280 for the extra 10,000 of capital. 3. At \(k = 40{,}000\), \(y = 40{,}000\); at \(k = 50{,}000\), \(y \approx 44{,}720\), a gain of only about 4,720 for the same extra 10,000. 4. Interpret: the same investment raises output almost twice as much in the capital-poor economy. That is the diminishing-returns case for convergence.
Argument for: borrowing technology
The second argument is that it is cheaper to copy than to invent. Rich economies at the technological frontier must discover new methods to grow, which is slow and expensive. Poorer economies can adopt methods already proven elsewhere: mobile phone networks, high-yield seeds, modern logistics. The economic historian Alexander Gerschenkron called this the advantage of backwardness, meaning not that being poor is good but that a lagging country has a stock of known techniques waiting to be used. A third argument is that late starters can learn from the policies of countries that grew before them, and that rising incomes build public support for the institutions that sustain growth.
Argument against: technology keeps rich countries growing
Diminishing returns apply when technology is held fixed. But technology does not stand still. Each improvement shifts the whole production function upward, so a rich economy can keep deepening its capital without its growth fading. Over the two centuries since the Industrial Revolution, new ideas have not obviously yielded smaller and smaller gains. One reason is that ideas are non-rival: once found, a method can be used by any number of workers at almost no extra cost. Endogenous growth theory, associated with Paul Romer, who shared the 2018 Nobel Prize in economics, explains how research and the spread of ideas can sustain growth in rich countries indefinitely.
Argument against: copying is not automatic
The advantage of backwardness is only a possibility. Adopting a technology needs engineers and managers who understand it, reliable power and transport, finance for the equipment, and confidence that profits will not be seized. Where schooling is weak, institutions are extractive or conflict is frequent, the opportunity to copy goes unused. This is why the low-income economies discussed in the earlier topic stagnated for decades despite enormous room to catch up. In the conditional-convergence view, such countries are already close to their own low potential, and they will converge only if the conditions that set that potential change.
What the recent evidence shows
World Bank data, grouping countries by their current income level, show real GDP per person growing at about 3.8% a year from 1990 to 2024 in middle-income countries, against about 1.6% in high-income countries. That is convergence, driven largely by China, India and other Asian economies. But low-income countries grew only about 0.5% a year per person, so they fell further behind. Growth of total GDP can mislead here, because populations in low-income countries are growing fast: growth per person is roughly total growth minus population growth. Some recent studies argue that convergence has become more widespread since about 2000, but the poorest economies remain the large exception.
Worked example: how slow catching up is
A hypothetical rich country has real GDP per person of $60,000 and grows at 1.5% a year; a poorer one has $6,000 and grows at 5.5%. 1. After 30 years: rich, \(60{,}000 \times 1.015^{30} \approx 93{,}800\); poor, \(6{,}000 \times 1.055^{30} \approx 29{,}900\). 2. The ratio has fallen from 10 to about 3.1. 3. To close the gap fully, solve \(\left(\frac{1.055}{1.015}\right)^t = 10\):
4. Interpret: even very fast growth, held for 30 years, leaves the poorer country with under a third of the rich country's income. Its growth may also slow as easy catch-up gains are used up.
Common mistakes to avoid
First, judging convergence by total GDP growth instead of growth per person: fast population growth can hide stagnant incomes. Second, assuming diminishing returns guarantee convergence, when they hold only with technology fixed. Third, reading the advantage of backwardness as automatic; copying requires skills, infrastructure and secure institutions. Fourth, confusing a falling ratio with a falling gap in dollars: in the worked example the ratio fell from 10 to 3.1, yet the dollar gap widened from $54,000 to about $63,900. Fifth, treating absolute and conditional convergence as the same claim.
Key takeaways
Convergence means poorer economies growing faster per person than richer ones. Absolute convergence ignores other differences; conditional convergence is catching up towards each economy's own potential, set by its fundamentals, and fits the evidence better. Arguments for convergence: diminishing returns to capital, the advantage of backwardness, and learning from others. Arguments against: technology keeps rich economies growing, and copying needs supportive institutions. Since 1990, middle-income countries have converged on high-income ones while low-income countries have not. Even rapid catch-up takes generations. The topic "Policies to improve living standards" asks what governments can do to speed it up.
Flashcards
Flashcard
Poorer economies growing faster per person than richer ones
Test your understanding
Question 1 of 5
Countries A and B have similar institutions and saving rates, but A has much less capital per worker. What does conditional convergence predict?
Question 2 of 5
With output per worker y = 200√k and fixed technology, how much does output per worker rise when k goes from 40,000 to 90,000?
Question 3 of 5
Which argument most directly explains why rich countries' growth need not fade as their capital per worker rises?
Question 4 of 5
A low-income country's total real GDP grows 3% a year while its population grows 2.7% a year. A high-income country's real GDP per person grows 1.5% a year. Is the low-income country converging?
Question 5 of 5
A country with $5,000 per person grows at 5% a year while one with $50,000 grows at 1.5% a year. After 20 years, which statement is true?
