Productivity Debate: Measurement

 

There is a debate going on right now (some of it behind paywalls, see below) about how to compare productivity across countries. The debating question is whether Europe (EU) has fallen behind the US because of lagging productivity and whether the comparison can even be made given how Technology and Productivity are measured. The issues involve: Measurement, Economic Theory, Economic Models, Forecasting, Growth Mania, the Steady State Economy, Sustainability, Prosperity, Human Development, etc. etc. In this post, I will deal with Measurement.

Notes

References

Bertuzzi, L. (2026) EU launches major tech push to break US and China dependence

European Commission (2025) The Draghi report on EU Competitiveness

Reserve Bank of Australia (2026) Productivity

Nature Editorial  (2025) End GDP Mania: how the World should really measure Prosperity.

Kalman, R. E. (1980) System-Theoretic Critique of Dynamic Economic Models," International Journal of Policy Analysis and Information Systems, 4 (March, 1980), 3-22.

Stone, R (1980) Review: Whittling Away at the Residual: Some Thoughts on Denison's Growth Accounting: A Review Article  Journal of Economic Literature, Vol. 18, No. 4 (Dec., 1980), pp. 1539-1543

Articles

• Challenging the Narrative of European Decline: Revised, Free Post May 21, 2026, Paul Krugman

• Why Isn’t Europe Poorer Than the US? Jun 1, 2026 Dalia Marin

• The Mismeasurement of Europe’s Productivity May 29, 2026 Philippe Aghion

• China’s Long March to Technological Supremacy May 27, 2026 Johan Rockström

•  Europe's Competitiveness Bogeyman Jun 4, 2026 Daniel Gros 
  
• The Draghi Report on EU Competitiveness Sep 9, 2024 Mario Draghi
• Measurement Without Theory Aug 1947, Tjalling C. Koopmans

Links

• A Steady State Economy for Germany the Deutsch Empire (DE) has been in a steady state since 2010.
• 
  

Measurement 

Measurement Models

Theory

There has been another long running debate in economics (Koopmans, 1947, see the article above) about the link between measurement and theory. When reading the Productivity debate, it might not be entirely clear how much implicit theory exists behind the qualitative analysis. A lot of Neoclassical Economic ideas are assumed in the arguments, and have to be exposed for general readers.

To avoid requiring an advanced degree in macros economics, causality in Economics can be understood using Direct Graphs and the Kaya Identity (above, the same identity used by the IPCC to explain Climate Change): more population (N)  leads to more employment (L) leads to more production (Q) leads to more Capital Accumulation which leads to more energy use (E). Ot is a simple linear model that is true by definition but incomplete.

There should be a dynamic feedback loop between Capital (K) and Production (Q) the accounts for Multifactor (MFP, see above) or Total Factor Productivity (TFP).

If TFP is not measured, the the productivity coefficient (q, called an Intensive Variable by the IPCC) is (at equilibrium and using loop reduction theorems) more complex, q/(1-ik). What we measure by computing Q/L from raw data (as is done in the Productivity debate) is confounded with dynamic Capital Investment reducing the true measure of Labor Productivity.

See the About page and the Boiler Plate for more discussion.

Confounding may not be a major issue if the i and k coefficients are small, but there are many other dynamic feedback loops that surround aggregate production that also go unmeasured in Labor Productivity calculations (wages, education, Learning-by-Doing, etc., etc.). Somehow, we have to control for dynamic effects and qualitative analysis (without models) does not do that.

Another theoretical measurement problem is "what do we mean by Q?" Are we counting goods actually being produced (toasters, automobiles, etc.), are we counting money (in native currency), etc. This is another theoretical issue involving markets.

Statistical Agencies actually do no count toasters and automobiles, let alone services, when calculating Q; they count the value of these goods and services in local currencies, that is Prices (P). The question is, which prices? One side of the debate (Krugman) argues for using Purchasing Power Parity (PPP) when comparing countries. Another side of the debate (Aghion) argues for using the GDP Deflator. The goal is to calculate Real GDP or Q. I always use the GDP deflator to compute Real GDP because it is more general. Computing PPP involves all sorts of assumptions about goods and services comparability across countries. And, all the Principal Components (PCA) measurement models are based on correlation coefficients which are dimensionless and comparable across countries. And, again simply dividing Q/P does not control dynamic price effects (the dynamic price effects can still be estimated in systems models, still using Real GDP).

Another old controversy surrounding the Solow-Swan model when used for Growth Accounting (see Stone, 1980 above) is accounting for the Residual after dynamic Labor and Capital input have been estimated. The Directed Graph above "whittles" away the residual by including other factors that come up in the Productivity debate: Technical Change (TECH), Market Fragmentation (MARKET), Education (ED), Regulation (REG), Innovation (INNOV) and Learning-by-Doing. For all the dynamic reasons mentioned above, the simple division of (Q/L) becomes quite confounded.