Why is TFP called a "residual"? It feels like we are giving up by lumping everything into one term.

Short answer: TFP is called a residual because we compute it INDIRECTLY. We measure GDP growth, capital growth, and labor growth — then assign whatever is LEFT OVER to "productivity." We do not measure productivity directly. The label "residual" is accurate but does not mean we are giving up — it means we have a disciplined accounting framework that isolates the unexplained portion.

Reading the symbols: $g_Y$ = output growth, $g_K$ = capital growth, $g_L$ = labor growth, $\alpha$ = capital share of income (~0.30 in developed economies), $g_A$ = TFP growth.

The mechanical computation

Starting from the Cobb-Douglas production function $Y = A K^{\alpha} L^{1-\alpha}$, taking logs, and differentiating gives the growth-accounting equation:

$$g_A = g_Y - \alpha g_K - (1-\alpha) g_L$$

We DIRECTLY observe (or estimate) the right-hand side terms. TFP growth $g_A$ is the BACKED-OUT remainder.

Loading diagram...

Why this is a feature, not a bug

The residual approach has three powerful properties:

1. It isolates everything we did NOT measure. Better technology, better management, better institutions, smarter logistics — all show up as TFP because they affect output without increasing capital or labor inputs. The framework lets us QUANTIFY the unexplained portion even if we cannot break it apart.

2. It provides a falsifiable benchmark. If you claim "the new manufacturing process boosts productivity 2%," you can verify by measuring whether actual output growth exceeds the predicted contribution from measured capital + labor by ~2%. TFP gives you a check.

3. It rules in or out competing growth narratives. When commentators say "China is catching up via technology transfer," growth accounting checks: is China TFP growth high, or is China growth coming from capital deepening (more machines per worker)? The decomposition reveals which.

Example: TFP as detective work

Suppose two countries both grow GDP at 5% per year for a decade:

Country	$g_Y$	$g_K$	$g_L$	$\alpha$	TFP $g_A$
Finland	5%	2%	0.5%	0.30	$5 - 0.6 - 0.35 = 4.05\%$
Vietnam	5%	8%	1.5%	0.30	$5 - 2.4 - 1.05 = 1.55\%$

Both grow at 5%. But Finland TFP growth is 4% — driven by genuine productivity gains. Vietnam TFP growth is 1.5% — most of its growth comes from capital deepening.

Loading diagram...

This is the analytical power of treating TFP as a residual — it forces a clean accounting and exposes what the headline GDP growth number hides.

The "measure of our ignorance" framing

Economists call TFP "the measure of our ignorance" half-jokingly. The serious version: TFP captures EVERYTHING that affects output but cannot be measured as a quantifiable input. As measurement improves (better data on human capital quality, intangible investment, etc.), some of what we currently call TFP gets re-classified as input. Over time, the TFP residual shrinks — but never to zero, because some growth genuinely comes from intangible technological and institutional improvements that we will never measure perfectly.

For the full framework see our TFP article.