A high share of wages in Added Value in France in 2025
Productivity
France
Europe
Automatic translation made using Deepl
According to Cotis (2009), the share of wages in value added has remained stable since 1980. Recent assessments by Cette, Koehl and Philippon (2019) and Gendre and Thommen (2025) confirm this diagnosis. We disagree, as illustrated by Figure 1, applying a different calculation method that we believe to be more justified. We find that:
the share of wages in value added has been growing in France, at least since 1995 (since 1989 to be exact, see OFCE working paper 2025-23, appendix 5)
the share of wages in France has been higher than in the five other major eurozone countries since the 2008 financial crisis.
Value added sharing summarizes the state of the productive fabric in a single piece of information. It indicates the potential for wage growth and future investment opportunities. Despite nearly 20 years of supply-side policies—from Nicolas Sarkozy’s fiscal shock to Emmanuel Macron’s supply-side policy, via François Hollande’s responsibility pact—France seems to be particularly constrained. In the OFCE working paper 2025-23 (section on evaporation), we point out that, in addition to the usual explanations for the share of wages in value added, the evaporation of value added through tax optimization within the European Union could explain this high level of wages in net value added. Tørsløv, Wier and Zucman (2022), using a different approach, document this channel.
The method we used to construct the Figure 1 indicator differs from previous approaches in that:
Wages (including bonuses and profit-sharing) are divided by net value added and not gross value added as in most other analyses. Net value added is calculated by treating fixed capital consumption (FCC, i.e., the wear and tear of productive capital during the production process) as intermediate consumption. This allows for better comparability between countries and over time. On the one hand, national accountants can treat certain expenditures as intermediate consumption or investment—this is the case for software produced for own account. On the other hand, different or changing production structures over time can result in aggregate productive capital of a different nature over time or space. Taking the CCF into account makes it possible to correct specific conventions or structural differences, at least as long as it is correctly assessed. The objective is to analyze the (possible) allocation conflict, and not removing fixed capital consumption from value added would significantly bias the analysis by artificially reducing the share of wages. See the OFCE working paper 2025-03 (section on net value added) for more details.
Wages (including bonuses and profit-sharing) are increased by an equivalent amount attributed to non-salaried workers. Non-salaried workers (also known as self-employed workers or individual entrepreneurs) are workers like any others, whose contractual relationship with their clients is different in nature from that of salaried workers. We increase the payroll by their remuneration because the number of non-employees varies greatly over time and space. While other analyses generally attribute to non-employees a salary equivalent to that of employees (often in the same industry, at more or less detailed levels of disaggregation), we use mixed income data (net of CCF), i.e., the income of sole proprietors as compiled by national accountants. In fact, in some countries, imputing equivalent wages leads to imputing a payroll for non-salaried workers that is higher than their mixed income. Furthermore, in France at least, mixed income data is available by industry. We use a hybrid method (proportional to the number of non-salaried workers in each industry) for other countries. This method makes it possible to measure the impact of platforms (uberization) on the income of self-employed workers. See the OFCE working paper 2025-03 (section on non-salaried workers) for more details.
We limit ourselves to the field of market sectors excluding real estate services. This field avoids having to include in the calculation of the indicator, on the one hand, the sharing of value added in non-market sectors, where it is largely conventional, since value added is assessed by wage and capital costs. On the other hand, it excludes real estate services activities that include residential services to households, which result more from the formation of rents (land rent, owners’ bargaining power) and an agreement for the valuation of the housing service self-produced by owner-occupiers for themselves (the method involves imputing a rent equivalent to that prevailing on the rental market, with the difficulties that one can imagine and significant variations from one country to another). Implicit rents represent 18% of market value added in France; since the objective is to analyze the distribution conflict between capitalists and workers, it would be problematic to pollute it with rents. See the OFCE working paper 2025-03 (Appendix 3) for more details on the impact of the chosen scope on the evaluation of the indicator.
These three main differences in the calculation of the indicator explain the difference with previous assessments, and other methodological aspects are discussed in detail in OFCE Working Paper 2025-03.
References
Cette G., Koehl L., Philippon T. (2019). “La part du travail sur le long terme : un déclin ?” Economie et Statistique / Economics and Statistics, n° 510‑511‑512, pp. 35–51.
Cotis J.-P. (2009). “Partage de la valeur ajoutée, partage des profits et écarts de rémunérations en france,” Rapport pour la Présidence de la République Française.
Gendre C., Thommen Y. (2025). “Le partage de la richesse produite en France entre le travail et le capital,” Trésor-Éco, n° 363.
Tørsløv T., Wier L., Zucman G. (2022). “The Missing Profits of Nations,” The Review of Economic Studies, 90, n° 3, pp. 1499–1534.