Economic Growth

Robert Solow's 1956 paper A Contribution to the Theory of Economic Growth set up the workhorse model of growth economics for fifty years and won Solow a 1987 Nobel: output is a function of capital and labour, capital accumulates through saving, labour grows exogenously, output per worker rises only if technology improves. A second paper a year later, Technical Change and the Aggregate Production Function (1957), took the model to the data: capital accumulation explained only one-eighth of US productivity growth 1909–1949 — the remaining seven-eighths was a residual he labelled technical change. The Solow residual turned out to be most of growth, and seven decades later the discipline still does not fully agree on what fills it.

What fills the residual is, first of all, a statement about what capital cannot do. Add machines to a worker and output rises, but each machine adds less than the one before; keep saving and the economy approaches a steady state where investment merely replaces what wears out, and output per head stops climbing. Growth that persists therefore cannot be growth by accumulation — something outside the machinery has to keep shifting the whole relationship. Bringing that something inside the model meant noticing how ideas differ from machines. A lathe serves one worker at a time; a method serves everyone at once. Because ideas are non-rival, the return to finding one scales with the size of the market that adopts it, so the diminishing returns that stall capital need not stall knowledge. Growth can then sustain itself — but only if whoever searches can capture enough of an idea's value to justify the search, which is why patents, research funding, and sheer market size enter growth theory as first-order variables rather than institutional detail. The harder question is why sustained growth began where and when it did, and there the candidate explanations are better read as claims about which constraint bound first than as rivals. Cheap and then cheaper energy. A culture that made useful knowledge cumulative, published, and testable. Institutions secure enough that an inventor could expect to keep some of the gain. Geography, soil, and disease. The capital and coerced labour drawn out of colonised economies. Each is well evidenced; none is sufficient alone. That is less a failure of the field than a property of the question. The transition happened once, on a planet that cannot be rerun with a single variable changed — so the dispute is about relative weights, and it stays open for reasons no additional data from the past will settle.