Lecture notes

Day 1

Lectures from Chicago, Monday, July 27, 2026.

Chicago

Monday, July 27, 2026

3 lectures

Open the full notes (PDF)

Lecture 1

The Middle-Income Trap

A lecture by Somik Lall

Co-Director, Growth Academy; Director, Strategy and WBG Institute for Economic Development, Development Economics Vice Presidency, World Bank Group.

The bottom line

Middle-income economies rarely stall for lack of ideas; they stall because incumbents preserve the status quo, resources sit trapped in unproductive firms, and countries never switch growth models. Escaping the trap means moving in sequence from investment to infusion to innovation, and rebalancing Schumpeter's forces, by disciplining incumbents, rewarding merit, and capitalizing on crisis.

Two traps, not one

World Bank Group's 2024 World Development Report was titled The Middle-Income Trap, but the sharper frame is creative destruction. The term was coined in 2006 by World Bank economists in Latin America, ground zero for economies that surge toward the frontier and then slow. That slowing is measurable: a growth slowdown is about three times more likely in a middle-income economy than in an advanced one, and slowdowns arrive sooner and last longer where economic freedom is lower.

The pattern is stubborn. Since 1980, middle-income countries' average per-capita income relative to the United States has barely moved, near six percent, with China the only real bright spot. These economies hold about three-quarters of the world's population but only about a third of its output, yet nearly all aspire to join the club of developed nations, not merely to clear the Bank's high-income threshold of about $14,000, a low bar, since the United States sits roughly seven times higher. The transition is hard because there was never a real theory of growth for the middle: Solow–Swan models fit low-income economies and Lucas–Romer models fit advanced ones, but neither explains the switch, and the trap is really two, one near $4,000–$5,000 and another near $8,000–$10,000, where countries stall for failing to change models.

The real bottleneck: misallocation of capital, talent, and energy

Before innovation, the binding constraint is how existing resources are used. On capital the diagnostic is blunt: unproductive firms are not competed out and productive firms do not grow. Where an American firm surviving to age 40 is about eight times its birth size ("up or out") emerging-economy firms too often "flat and stay."

Talent is misallocated just as badly. As economies develop, the number of jobs demanding specialized skills rises many-fold, yet half the labor force is often locked out; when the United States was middle-income in the 1950s, 97 percent of doctors and lawyers were white men, whereas women and minorities now hold 52 percent of those professions, leaving the economy about 50 percent larger than the discriminatory pattern would have allowed. Energy is used no better: middle-income economies run energy intensity about two and a half times, and emissions intensity about three and a half times, that of high-income economies, gaps about using energy better, not switching fuels. Aggregated, if China used energy as efficiently as the United States, a Chinese worker would be about 80 percent as productive as an American, not 20 percent.

Switching models: investment, infusion, innovation

The escape route is a sequence, not a leap. Between investment and frontier innovation sits "infusion" (adapting global ideas to local circumstances) the missing ingredient in most middle-income countries: low-income economies concentrate on investment, lower-middle add infusion, and only upper-middle combine all three. Korea is the exemplar, sending engineers to learn electronics from Japan, extending collateral-free credit to firms with foreign licensing agreements, outcompeting the Japanese in Asian markets, and (once license prices rose) switching deliberately into innovation. Brazil chose the opposite path, subsidizing domestic R&D and patents behind a tariff wall; but its patent office could not tell new ideas from recycled ones, so firms collected subsidies while productivity relative to U.S. workers fell about 40 percent, leapfrogging straight to innovation rarely works.

Underneath both stories is Schumpeter's trio of forces, creation, preservation, and destruction. Most middle-income countries have no shortage of creation; they have a preservation problem, with incumbents (often state-owned enterprises) holding down the new. Schumpeter saw incumbents colluding to block entry, Aghion and Howitt cast entrants as the heroes who push the frontier, and Akcigit and Bill Kerr show that incumbents too can be innovators, so the task is to weaken preservation and rebalance the forces.

Evidence at a glance

Slowdowns are structural. A growth slowdown is about three times more likely in a middle-income economy than in an advanced one, and lasts longer where economic freedom is lower.
The convergence gap is frozen. Middle-income per-capita income relative to the United States has stayed near six percent since 1980; only China has broken out.
Firms are stuck. Unproductive firms survive and productive firms fail to scale ("flat and stay" rather than "up or out") and in Ukraine the productivity–growth link, once U.S.-like, later vanished entirely.

From X-rays to MRIs: the policy agenda

Discipline incumbents, reward merit, capitalize on crisis. Stop diagnosing growth with the X-rays of firm size, market concentration, and income inequality; use the MRIs of value added, socio-economic mobility, and emissions intensity.

Discipline incumbency. Do not vilify incumbents; harness their talent, capital, and market power to advance development, not to preserve it.
Reward merit. Connect young, small firms to markets, mentors, and finance so the best can scale.
Capitalize on crisis, and free enterprise. Shocks weaken preservation, as when Korea let go of protected conglomerates in the Asian financial crisis; and economic freedom can shift fast, as Vietnam showed after 2017.

Lecture 2

The Economics of Creative Destruction

A lecture by Philippe Aghion

Kurt Bjorklund Professor in Innovation and Growth, INSEAD, London School of Economics and Collège de France.

The bottom line

Long-run growth is driven by creative destruction, and at its core lies a contradiction: innovation needs the prospect of rents, yet yesterday's innovators use those rents to block tomorrow's. Managing that contradiction (through competition, education, and flexicurity) is what lets economies avoid the middle-income trap and become both more innovative and more inclusive, including in the coming age of AI.

A paradigm built on a contradiction

The neoclassical growth theory of the 1950s, in Solow's elegant version, explained growth through capital accumulation but ran into decreasing returns; it needed technical progress for the long run yet never said where progress came from. The answer is creative destruction, Schumpeter's idea that new innovations make old technologies obsolete. Beginning in 1987 with Peter Howitt, a new Schumpeterian paradigm was built on three ideas: growth is cumulative, each innovator standing on earlier shoulders; innovations come from entrepreneurs chasing temporary monopoly rents; and each new innovation displaces the old.

The paradigm contains a built-in tension. Innovators need rents to justify the effort, but once they hold those rents they are tempted to block new entry so they are not themselves creatively destroyed; regulating a market economy is largely about managing this contradiction. In the basic Aghion–Howitt model, the equilibrium growth rate rises with the productivity and size of innovations and falls with the interest rate, so growth is positively correlated with the rate of creative destruction. That is why the United States, with more R&D and more breakthrough innovation, grows faster than a Europe specialized in incremental improvement.

Two extensions: competition and firm dynamics

The basic model implies that only outsiders innovate and that more competition is bad for growth, counterfactual results that forced the theory to evolve. Adding step-by-step innovation, where a laggard must climb each rung before overtaking a leader, introduces an escape-competition effect: leaders innovate to stay ahead while firms far behind are discouraged. Because advanced countries have more firms near the frontier, competition becomes increasingly growth-enhancing as a country develops, and empirically the relationship between competition and innovation is an inverted U.

A second extension models firms as collections of product lines that grow through creative destruction on new lines and shrink when rivals displace them. It predicts a highly skewed size distribution (many small firms, few large ones) and a positive relationship between firm age and size. That gradient is far steeper in the United States than in Mexico or India, because good U.S. innovators can reach venture capital and institutional investors and scale, so surviving long in the U.S. is itself a signal of quality.

Secular stagnation and the middle-income trap

The same tools illuminate two enigmas. In the United States, TFP growth surged between 1996 and 2006 and then declined. The surge coincided with superstar firms (Google, Microsoft, Walmart, Amazon) that harnessed the IT revolution; but weak competition policy let them expand unboundedly through mergers and new establishments, and entry has fallen since 2000. These leaders also stopped sharing knowledge, widening the gap to followers, discouraging laggard innovation, and pushing markups up, so concentration rose and growth fell together.

The middle-income trap has the same structure across countries. Growth can come from catching up with the frontier (the advantage of backwardness) or from frontier innovation, and the two require different policies. Catch-up rewards education, trade openness, technology transfer, and better management, as in China after Deng Xiaoping; frontier innovation demands graduate schools, basic research, the freedom to fail, patient funding, venture capital, and competition. Economies that fail to switch fall into the trap: Korea's chaebols and Japan's keiretsu blocked the move to frontier institutions, and Europe, after catching up by the late 1980s, could not harness the IT revolution, lacking venture capital, patient funding, and a DARPA-style bridge between competition and industrial policy.

Evidence at a glance

Superstars, then stagnation. U.S. TFP growth surged 1996–2006 alongside rising concentration, then declined as entry fell after 2000 and leader–follower technology gaps widened.
AI's potential is large but conditional. Automating tasks in goods and ideas could add about 0.68 percentage point to annual growth over a decade, comparable to IT's 0.8, provided competition is preserved.

Toward growth that is both innovative and inclusive

You need not choose between innovation and protection. Flexicurity, education, and competition make an economy more innovative and more inclusive at the same time.

Flexicurity. Denmark's model (income support and active retraining after job loss) makes creative destruction work better and protects workers' health, vital as AI accelerates churn.
Education that levels the field. The probability of inventing rises steeply with parental income; broad-based, high-quality schooling reduces "lost Einsteins" and raises both innovation and mobility.
Competition, made dynamic. Judge mergers by whether they stifle entry and innovation, insist on data sharing, and pair competition with DARPA-style industrial policy and deep financial ecosystems.
Harness AI deliberately. Preserve competition and open source to capture its growth potential, and pair strong basic education with active labor-market policies to capture its jobs potential.

Lecture 3

The Economics of Dynamism: Firm Productivity, Entrepreneurship, and Industrial Policy

A lecture by Ufuk Akcigit

Co-Director, Growth Academy; Arnold C. Harberger Professor of Economics, University of Chicago (on leave); Deputy Chief Economist and Director of Private Markets, World Bank Group.

The bottom line

Every policy is a choice that moves resources from one use to another, so the question is never simply whether firms grow but whether growth is organic and productivity-driven. Middle-income economies stall because unproductive firms are kept alive, transformative entrepreneurs cannot scale, and blanket subsidies buy job numbers instead of efficiency. Better data lets policy target performance, not size, and reward the few firms that truly transform.

Growth accounting: the source of growth matters

We want economies to perform better, but spending money does not guarantee it. In a general-equilibrium world, subsidizing one firm means taxing another, so a policy that looks wonderful when we study only the firm receiving the dollar can leave the economy worse off, and most interventions lack the evidence to justify the trade-off. Decomposing growth into capital, labor, and efficiency, the efficiency term (the Solow residual) is the part we cannot see, which is why politicians prefer visible roads to invisible laboratories.

That preference is costly. Switzerland spends heavily on invisible R&D and labs, giving matching subsidies to universities so firms and universities advance together, an ecosystem, not a line item. Decompositions make the stakes concrete: in Korea and Poland, capital, labor, and efficiency all contributed positively; in South Africa about half of efficiency growth was negative, so whatever was gained was lost, the signature of a stalled economy. Plotting efficiency against income confirms it: Korea, Poland, and China grew as efficiency rose and stalled as it faded.

Static efficiency: are good firms allowed to scale?

Before creating new technology, an economy must use existing technology well. Aggregate productivity is the sum of each firm's productivity weighted by its market share, so the same technologies yield very different outcomes depending on who operates at what scale. The test is responsiveness: does a firm that receives a positive productivity shock actually gain resources and employment? In planned economies the correlation is absent; comparing West Germany with the United States, the responsiveness coefficient is about twice as high in the U.S., so the same innovation produces more efficiency there simply because input markets are less rigid.

Policy often makes this worse by keeping unproductive firms alive. When East Germany's firms were privatized after 1990, unemployment jumped and social unrest followed; after the head of the privatization agency was assassinated in April 1991, contracts shifted from asking a price to imposing labor commitments. Firms hired to satisfy those commitments, creating about 25 percent excess workers, and firms with binding commitments grew employment by about 30 percent. Thirty years on, the East–West productivity gap still exceeds 20 percent, whereas Poland, with no rich neighbor to rescue it, let restructuring happen and converged better.

Entrepreneurs, finance, and the design of policy

Not all entrepreneurs are alike. In Danish data, the likelihood of becoming an inventor or scientist rises sharply with ability, but the likelihood of starting a business falls with it, because most entrepreneurship is necessity-driven and subsistence. Distinguishing the roughly 10 percent of "transformative" entrepreneurs, identifiable early by whether they hire technical talent, from the 90 percent of subsistence founders matters enormously: transformative firms keep growing for a decade, while subsistence firms stay flat. Transformative entrepreneurs, like inventors, tend to be educated and high-ability, yet parental income still predicts who reaches college even where tuition is free.

This heterogeneity should drive policy sequencing. With very little money, spend it all on equal access to education; with more, add a venture-style program that screens and funds promising entrepreneurs; only with ample resources subsidize incumbent R&D, yet intuition jumps straight to the last. Blanket instruments are captured: large firms claim most R&D tax credits because they have the resources to. France's "gazelle" program instead conditioned support on performance, growth above 15 percent for two consecutive years, selecting firms already trying to grow. Finance ties it together: among small firms, the transformative ones are precisely those that borrow to grow, so a credit shock hits them hardest and leaves lasting scars.

Evidence at a glance

Target performance, not size. Among low-productivity firms, about 23–24 percent claim the R&D tax credit but only about 1 percent the gazelle program; among high-productivity, high-growth firms, about 96 percent claim the gazelle.
Finance is decisive. Small transformative firms depend most on borrowing to grow, so a rise in the interest-rate spread produces a downturn from which the economy never fully recovers.

Designing for dynamism

Fix all four tires. Functional economies are alike, education, allocation, technology creation, and finance all work; dysfunctional ones fail for different reasons, so the task is to find the specific blockage.

Chase organic growth, not induced. Subsidies that buy job numbers evaporate when they end; sustain jobs by letting resources flow to productive firms rather than propping up zombies.
Target transformative firms. Use information (who a founder hires, how a firm performs) to reward the few that scale, conditioning support on performance, not size.
Sequence policy to the budget. Education first, screened startup support next, incumbent R&D subsidies only when resources are ample.
Protect finance for growers in crises. Small transformative firms rely on borrowing, so safeguarding their credit access during macro shocks prevents lasting damage.