For decades, international economists have grappled with a paradox: why does so much global capital flow toward wealthy countries when poorer countries, where capital is relatively scarce, should theoretically offer higher returns?
A new study by Dan Su, Assistant Professor of Finance at Cheung Kong Graduate School of Business, offers a fresh explanation in his latest paper published in the International Economic Review. His paper, The Capital Matthew Effect: Directed Technical Change and International Capital Flows, argues that the amount of capital a country has can shape the direction of technological progress—and that this can create a self-reinforcing cycle in which capital continues to flow toward countries that already have plenty of it.
The study’s central finding challenges a common assumption behind the long-standing “Lucas paradox.” Rather than assuming that poor countries offer inherently higher potential returns but are held back by weak institutions or financial frictions, Su finds that since the 1970s, marginal returns to capital have, on average, been higher in richer countries than in poorer ones.
That changes the question. If returns are already higher in capital-rich economies, the real puzzle is why those economies can maintain higher returns despite already possessing large amounts of capital.
Su’s answer centers on the direction of technological change.
Countries with abundant capital have stronger incentives to develop technologies that make capital—and complementary resources such as highly skilled labor—more productive. As these technologies improve, additional capital does not necessarily experience rapidly diminishing returns. Instead, technological progress can increase the productivity and value of capital, helping sustain high returns and attract further investment.
Capital-scarce economies face different incentives. Where capital is expensive and labor relatively abundant, firms may be more likely to develop technologies that economize on capital, substitute for capital, or improve the productivity of lower-skilled workers. Such innovation can still raise overall productivity and lower the reliance on capital, but it does not necessarily increase returns to capital.
Su calls this the “Capital Matthew Effect”: an initial advantage in capital can reinforce itself through technological specialization.
The mechanism forms a self-reinforcing cycle: capital abundance encourages capital-enhancing technological development; those technologies raise returns to capital; higher returns attract additional investment; and the larger capital base further strengthens incentives to continue along the same technological path.
If the divergence created by directed technological change outweighs the traditional force of diminishing marginal returns, capital can continue flowing from capital-scarce countries toward capital-rich ones.
The paper tests this argument using both country-level and firm-level data.
Drawing on the Penn World Table and the World Bank’s wealth database, the study compares marginal returns to capital across richer and poorer countries. Before the 1970s, the relationship between income levels and returns to capital was not consistently significant. Since then, however, the relationship has become persistently positive: richer countries have, on average, recorded higher marginal returns to capital than poorer countries.
The finding remains broadly intact under different country classifications and alternative measures of capital returns, as well as after excluding small economies and oil exporters.
The study also finds that a country’s initial capital abundance significantly predicts its subsequent long-term capital flows. Globally, capital has tended to move toward countries that were already relatively capital-rich.
The direction of productivity growth also matters. Technological progress that complements capital and other complementary factors is more likely to raise returns and attract investment than productivity growth driven primarily by capital-substituting technologies.
Firm-level evidence points in the same direction. Using Global Compustat’s data covering listed companies in more than 80 countries, the study finds faster capital-biased technological progress among firms in countries that were initially more capital-rich and has increasingly open financial systems. The paper does not rule out other explanations, including institutional differences and financial frictions, but suggests that directed technological change may be one important force shaping long-term international capital flows.
The findings raise difficult questions for policymakers.
If international capital imbalances are mainly caused by institutional weaknesses or financial barriers, reducing those obstacles should help capital flow toward countries where it is scarcer and gradually equalize returns.
Su’s framework suggests the opposite. If capital-rich countries have already developed technologies that generate higher returns on capital, more efficient and integrated financial markets may make it easier for global capital to flow toward them.
The paper does not argue against financial openness. Instead, it emphasizes the need to consider its longer-term economic consequences. Economies that consistently receive capital may face risks of financial instability, while economies experiencing persistent capital outflows may need to consider the possible links between capital outflows and premature deindustrialization.
1. The direction of technological progress may matter more than its speed
We often measure technological progress through R&D spending, patents, or productivity growth, but rarely ask what technology is actually enhancing or replacing.
Some technologies complement capital and highly skilled workers; others reduce the need for capital and raise the productivity of less-skilled workers. Both can drive growth, but they have very different effects on returns to capital and the allocation of resources.
For capital, the key question is therefore not just how fast technology advances, but in which direction. Technologies that amplify the value of capital can generate higher returns even when technological progress itself is modest.
2. A true first-mover advantage may be self-reinforcing
Having more capital, talent, or data today does not necessarily create a lasting advantage.
What matters is whether a company can develop technologies and organizational capabilities that reinforce those resources and generate higher returns over time.
When resource accumulation, technology choices, and rising returns reinforce one another, later entrants face not only a resource gap but also a self-reinforcing cycle.
3. Does greater openness narrow the gap between leaders and followers?
When firms have access to similar technologies, we might expect technology to level the playing field.
But if leading firms have already developed mechanisms that amplify the value of their capital and talent, greater openness may instead accelerate the concentration of resources in those firms.
Ultimately, the study offers a new way to understand the relationship between capital, technology and global development gaps: when capital abundance shapes technological development, and technological development in turn raises returns to capital, initial differences between countries may become self-reinforcing.
