What is the practical difference between capital deepening and TFP? They both seem to be about productivity.

Why the distinction matters

Both capital deepening and TFP raise output per worker, but they behave very differently in the long run.

• Capital deepening ($g_{K/L}$) is growth in capital per worker. If you give every factory worker a second machine, output per worker rises. But the second machine is less productive than the first, the third less productive than the second, and so on. This is the law of diminishing returns to a single factor.
• Total factor productivity ($TFP$) is growth in output that is not explained by either labor or capital inputs. It captures technology, organizational improvements, regulatory quality, and how efficiently the existing capital and labor work together. TFP is not subject to diminishing returns in the same way — better ideas keep paying off.

A tale of two countries

Imagine two economies with the same headline labor productivity growth of $2\%$ per year:

Loading diagram...

Cabralia is producing more per worker by piling on capital — an extremely high $34\%$ investment share. Eventually, the marginal product of capital falls so far that additional capital is not worth its cost, and growth grinds toward the rate of TFP alone ($0.2\%$). Norvalia is producing more per worker mostly through better technology and organization, which is not bounded by diminishing returns the same way.

Forecast implication: Cabralia's trend growth is likely to decelerate. Norvalia's can persist or even accelerate if its institutions continue to support innovation.

How analysts measure each

TFP is almost always measured as a residual in growth accounting:

where $\alpha$ is capital's share of national income (commonly $\approx 0.30$ in developed markets) and $g_K$, $g_L$ are growth rates of the capital stock and labor input. Capital deepening per worker is then $g_{K/L} = g_K - g_L$.

Because TFP is a residual, it absorbs measurement error and any factor of production you forgot to measure. That is why mature-market TFP estimates are noisy — and why analysts cross-check with patent counts, R&D intensity, and total factor productivity series from independent sources like the Conference Board.

Common exam trap

Candidates often confuse capital deepening (capital per worker) with growth in the total capital stock. If labor is growing at $1\%$ and capital is growing at $1\%$, capital per worker is flat — there is no capital deepening even though absolute capital is rising.

Try the related practice question on identifying capital deepening vs aggregate capital growth.