Gerhard Tintner and Yves Aureille, Les problèmes de la mesure du niveau d’activité de la main-d’œuvre en France (1968)

Tintner and Aureille’s journal article combines a theory of labour-input measurement with an analysis of French demographic and employment prospects through 1970. Its central claim is that population growth, employment growth, and growth in the quantity of labour supplied are distinct processes. Changes in participation, sectoral composition, working time, and labour quality intervene between them. The article proceeds in two parts: a general decomposition method and a proposed measure of labour input, followed by a historical and prospective application to France.

The methodological argument begins by challenging the apparent neutrality of aggregate worker counts. Even counting hours leaves differences in the effectiveness of work unexamined:

Ce faisant, on se fondait sur l'hypothèse implicite que le travail de ces différentes catégories de personnes actives est homogène en qualité et en quantité.

English translation: In doing so, one relied on the implicit assumption that the work of these different categories of economically active persons is homogeneous in quality and quantity.

For the authors, structural change is therefore constitutive of labour-input measurement, not a secondary correction. Their general method expresses an aggregate as a sum of products of explanatory factors, then decomposes its absolute change into additive contributions. Successive substitutions combine initial and final factor values, producing an exact accounting identity. These contributions depend on the sequence of substitution; they distinguish sources of measured change without independently establishing causal effects.

The proposed labour measure classifies people jointly by sex, age, occupation or economic sector, and qualification. Each employed class contributes its headcount multiplied by weekly hours, annual working weeks, and a coefficient representing labour quality. Summing these products yields a common unit:

Ce niveau d'activité de la main-d'œuvre mesure alors le montant total du travail effectué dans une entité économique donnée pendant une année, exprimé en nombre d'heures de travail « équivalentes » à l'heure de travail d'une classe de travailleurs de référence choisie arbitrairement.

English translation: This level of labour-force activity thus measures the total amount of work performed in a given economic entity during one year, expressed as a number of working hours “equivalent” to the working hour of an arbitrarily chosen reference class of workers.

The conceptual move is from heterogeneous workers to comparable labour services. Employer hourly costs, including taxes and social contributions, are suggested as a practical quality weight proportional to marginal productivity. A footnote restricts the strict validity of that assumption to perfect competition. The full model separates seven contributions: population size, age-sex composition, participation rates, sectoral and qualification distribution, weekly hours, annual working weeks, and labour quality.

Statistical limitations substantially narrow the empirical application. The simplified model drops changes in labour quality, combines demographic and participation effects into changes in workforce size and structure, and assumes working time depends only on sector. The French calculations consequently measure total hours rather than implementing the full quality-adjusted ideal. Their four components are employment size, sectoral composition, weekly hours, and annual working weeks.

The application places postwar demographic recovery against France’s earlier population stagnation. Renewed births, immigration, and repatriation after the Algerian War increased population, but the simultaneous expansion of younger and older age groups reduced the proportion aged 20–64. The projection assumes declining mortality, stable age-specific fertility, and net immigration of 150,000 annually. Participation and schooling rates, known from the 1954 and 1962 censuses, are extended by linear interpolation and extrapolation. Longer schooling and earlier retirement offset much of the labour-force increase that population growth alone would imply.

The authors distinguish this estimated available workforce from the actual labour force, reconstructed through civilian employment, conscripts, career military personnel, and unemployment. The Algerian War matters because military expansion temporarily reduced civilian labour availability. Meanwhile, agriculture lost workers through mechanisation and the relatively weak responsiveness of food demand to rising incomes. Since agricultural working hours were longer, movement into industry, commerce, services, and administration reduced aggregate hours even where employment increased.

For 1949–1964, total labour input rises only from approximately 42.32 to 43.07 billion hours. Employment growth contributes 2.186 billion hours, but sectoral redistribution subtracts 1.055 billion and fewer annual working weeks subtract 1.042 billion; increased weekly hours add 687 million. The prospective result reverses the expectation that postwar demographic recovery must expand labour input:

Contrairement à l'opinion généralement admise, le niveau d'activité de la main-d'œuvre dans l'ensemble de l'économie française devrait diminuer sensiblement au cours de la période 1965-70.

English translation: Contrary to the generally accepted view, the level of labour-force activity in the French economy as a whole should decrease appreciably during the period 1965–70.

For that period, the decomposition attributes a positive contribution of 966 million hours to employment growth, outweighed by reductions associated with sectoral redistribution, weekly hours, and annual working weeks. These are conditional projections, not observed outcomes. Numerical inconsistencies in the supplied text—including differing projected labour-force totals and a mismatch between the stated 1965 hours and the table—counsel caution about exact endpoints.

The article’s lasting relevance lies in its separation of demographic potential from realised labour input, and of employment expansion from hours supplied. Its conclusion expects decline through 1969 before demographic rejuvenation begins to increase activity:

Si, à long terme, les variations démographiques conditionnent bien la quantité de travail disponible à l'économie, cette relation de cause à effet n'est pas absolument stricte et reste sujette à une inertie considérable.

English translation: Although, in the long run, demographic changes do condition the quantity of labour available to the economy, this cause-and-effect relationship is not absolutely strict and remains subject to considerable inertia.

Demography sets long-term possibilities; institutions, workforce composition, and working-time practices govern how slowly and unevenly those possibilities become labour services.