Total Factor Productivity and the Sources of Growth

Core Idea

Growth accounting decomposes output growth into contributions from capital accumulation, labor growth, and total factor productivity (TFP)—the residual measuring technological progress and efficiency improvements. The Solow residual shows that most long-run growth comes from TFP, not capital deepening, highlighting the importance of innovation, education, and institutions. TFP growth varies across countries and over time, explaining divergent growth rates.

Explainer

Think about why some countries grow rich and others stagnate. The Solow model you studied showed that capital deepening — adding more machines per worker — generates growth, but at a diminishing rate. Eventually an economy reaches its steady state where capital per worker stops rising. Yet we observe sustained long-run growth in wealthy economies for over a century. Where does it come from? The answer is total factor productivity (TFP): the ability to squeeze more output from the same inputs. TFP captures everything that isn't capital or labor — technology, organizational efficiency, institutions, infrastructure quality, and the knowledge embedded in the workforce.

Growth accounting makes this precise using the Cobb-Douglas production function. If output Y depends on capital K, labor L, and technology A as Y = AK^α·L^(1-α), then output growth can be decomposed: ΔY/Y ≈ α(ΔK/K) + (1-α)(ΔL/L) + ΔA/A. The first two terms are contributions from capital and labor growth, weighted by their income shares (α and 1-α). The residual — what's left after accounting for factor accumulation — is the Solow residual, which equals TFP growth. This is why TFP is called a residual: we cannot measure it directly, only infer it from what factors alone cannot explain. It is growth that comes from doing things smarter, not just from adding more inputs.

The empirical finding that reshaped macroeconomics: the Solow residual is large. In most developed economies, TFP growth accounts for roughly half to two-thirds of long-run output growth per capita. Capital deepening matters, but it runs into diminishing returns — doubling capital does not double output. TFP has no such constraint. A better algorithm, a new molecule, a reformed legal system, improved management practices — these shift the production function itself, permanently raising the output achievable from any given stock of inputs. Economists call this "working smarter, not just harder."

This explains divergent growth rates across countries in a way pure capital accumulation cannot. Two countries with identical savings rates and labor force growth will converge to the same steady-state income level in the basic Solow model — yet we observe massive, persistent divergence. The missing piece is TFP. Countries differ in their rates of innovation, technology adoption, institutional quality, and human capital formation. Endogenous growth theory pushes further: rather than treating TFP growth as manna from heaven, it models the specific investments — in R&D, education, and infrastructure — that generate it. The policy implication is direct: if most long-run growth comes from TFP, and TFP comes from institutions, innovation, and education, then those are the levers that matter most for development strategy.