Karl Pribram, Probleme der internationalen Arbeitsstatistik (1925)

Karl Pribram’s short monograph, published as the fourteenth Kieler Vortrag, examines the methodological conditions under which labour statistics can support international social policy. Writing as head of the International Labour Office’s statistical department, Pribram makes comparability a problem of institutions and concepts, not merely numerical technique. National figures arise from different administrative systems, legal definitions, and economic structures; apparently identical categories therefore frequently measure different things. The work proceeds from the political responsibilities of statistics through industrial and occupational classification, labour-market measurement, wages and living conditions, and social insurance, before concluding with a qualified programme of international coordination.

Pribram begins with the enlarged demands that regulatory economic and social policy places upon statistics. Governments require not simply records but reliable grounds for intervention and assessments of its likely effects. Political purposes may legitimately determine questions, but must not colour the collection or interpretation of evidence:

Der sicherste — und der einzig sichere — Weg, diese Gefahren zu vermeiden, ist die Ausbildung und Beobachtung von wissenschaftlich einwandfreien Untersuchungsmethoden.

English translation: The surest—and the only sure—way to avoid these dangers is to develop and observe scientifically sound methods of investigation.

This methodological independence is especially necessary where employers and workers interpret results through opposing interests. International comparisons add national competition to these conflicts. Pribram presents the International Labour Office’s statistical conferences of 1923 and 1925 as practical responses: comparative methodological reports enabled national representatives to negotiate recommendations. Their agenda balanced international importance against the prospect of reaching agreement.

Classification supplies the foundational difficulty. Industries can be grouped by material, production process, or product’s purpose, but none of these principles works exclusively. Shipbuilding and shoemaking illustrate how plausible classifications assign the same activity to different groups. Moreover, categories must accommodate changes in production itself:

In jedem industriell fortschreitenden Lande scheint die Änderung in der Struktur der Betriebe ständige Modifikationen in der Klassifizierung der Erwerbszweige zu fordern.

English translation: In every industrially advancing country, changes in the structure of establishments seem to require constant modifications in the classification of branches of economic activity.

Consequently, the conference abandoned a uniform hierarchy of industrial groups in favour of seeking agreement on the lowest-level branches, which could subsequently be combined as required. Even this modest solution encounters differences in the integration and separation of production processes. Occupational comparisons likewise require knowledge of actual functions within establishments: identical job names need not identify equivalent work, while agricultural employment complicates distinctions between proprietor, family member, and wage worker.

Unemployment statistics demonstrate particularly clearly how administrative sources shape the object measured. Trade-union returns cover organised workers and reproduce union classifications; employment exchanges count registered jobseekers; insurance systems count people meeting their eligibility or registration rules. These are not interchangeable populations. Pribram therefore restricts comparison chiefly to relative movements and demands explicit information about representativeness. Labour statistics also extend beyond labour administration: unemployment supplies evidence for business-cycle analysis, while occupational classification helps reveal a country’s economic and social structure.

The discussion of wages distinguishes payment as workers’ income from payment as a production cost. International productivity measurement remains poorly developed, partly because comparable output units are scarce. Wage statistics therefore generally compare contractual rates or actual earnings rather than remuneration per unit produced. Real-wage measurement introduces another distinction: domestic comparisons can hold a consumption budget constant over time, but international comparisons confront different currencies, purchasing powers, consumption habits, and qualities of goods.

Pribram explains the Labour Office’s provisional comparison of wage purchasing power in capital cities through several regional food baskets. Each city’s wage is measured against every basket, and the resulting ratios are averaged. He carefully limits the inference: the exercise compares contractual rates for an eight-hour day, not actual earnings or complete living standards. Measuring labour costs in coal mining presents further problems because countries define wages, shifts, and output differently; agricultural remuneration additionally includes diverse payments in kind.

Nur eingehende Untersuchungen der Struktur der verschiedenen Arbeitsverhältnisse können die Grundlagen für eine vergleichende Lohnstatistik bieten.

English translation: Only detailed investigations of the structure of the various employment relationships can provide the foundations for comparative wage statistics.

This demand for substantive investigation also governs the treatment of housing and social insurance. Rooms, dwellings, and households require definition in relation to national living arrangements. Insurance statistics depend on legislation and administration: even a fatal occupational accident varies with the period within which death is attributed to injury. Pribram advocates more consistent measures of injury consequences and exposure expressed in working hours, while recognising the need for industry-specific methods.

The conclusion rejects wholesale reconstruction of national statistics for international purposes. Partial adaptation and full methodological disclosure offer more realistic advances:

Vielleicht erweisen wir den statistischen Ämtern den größten Dienst, wenn wir sie veranlassen, sich immer wieder auf die Methoden ihrer Arbeit zu besinnen.

English translation: Perhaps we render statistical offices the greatest service when we prompt them repeatedly to reflect on the methods of their work.

International comparison thus becomes a means of improving national statistical practice. The work’s enduring relevance lies in its insistence that comparability must be established through knowledge of definitions, coverage, and institutional origins—not presumed from matching labels or numerical precision.