Ludwig Lachmann, On the Measurement of Capital (1941)

Ludwig Lachmann’s journal article examines how capital can be measured in an economy shaped by changing expectations, uneven utilisation, and failed investment plans. Its five sections move from a methodological introduction through a critique of Colin Clark’s capital statistics to an alternative grounded in capital services and depreciation accounting. Drawing on Hayek’s dynamic capital theory, Lachmann argues that statistical measurement must distinguish changes in productive resources from changes in their valuation. His purpose is not to reject quantitative economics, but to identify a measurable object that corresponds to the productive contribution economists wish to explain.

The opening locates the problem within the growing cooperation between theorists and statisticians. National-income research demonstrates the benefits of this cooperation, but capital presents particular difficulties: static concepts cannot adequately describe resources whose uses and values depend on expectations exposed to unforeseen change. Sections II and III therefore examine Clark’s The Conditions of Economic Progress through immanent criticism. Even granting that a “stock of real capital” could be meaningful, Lachmann argues that Clark’s records cannot isolate its growth or productivity.

Clark’s reliance on original cost, market value, and replacement value obscures their different economic meanings. Their convergence in a theoretical long period cannot be assumed over an actual historical interval, during which fresh disturbances may continually arise. The apparent decline in American manufacturing capital after 1929 particularly exposes the problem. Written-down asset values record disappointed expectations, not necessarily the disappearance of equipment or its productive services.

It is this disappointment of investors' initial expectations which the process of writing down asset values reflects.

Misinvestment adds resources, although less usefully than investors anticipated. Counting it at zero understates that addition; counting it at cost exaggerates it. Clark’s treatment effectively identifies misinvestment with no investment, allowing increased output alongside reduced recorded capital to appear as evidence of increasing returns. His British estimates create a related difficulty: adjusting market valuations by consol yields produces a supposed decline in real capital that may instead reflect incomplete adjustment of asset prices.

The capital-output ratio compounds these valuation problems with utilisation effects. During a depression, idle equipment remains in the numerator while output falls; recovery reverses the movement. The resulting changes cannot straightforwardly indicate changes in capital productivity.

The main defect of Mr. Clark's method is thus seen to lie in the lack of identity between the object of his statistical measurement and the productive agent the contribution of which he wishes to measure.

Lachmann’s objection is not confined to the interwar period. Short intervals allow utilisation fluctuations to overshadow growth; long intervals make historical and replacement costs unreliable measures of equipment whose economic role has changed. His examination of Douglas’s earlier manufacturing statistics reinforces the argument that choosing another period does not remove the conceptual difficulty.

Section IV shifts from criticism to reconstruction. Measurement remains necessary for empirical verification, analysis of income distribution, and concepts such as the rate of profit. The decisive move is to distinguish the heterogeneous stock of resources, the services it supplies, and the output produced jointly with other factors.

Hence the relationship determining the productivity of capital has in the first instance to be established between service stream and output stream.

A stock can generate service streams of different sizes and shapes, so its total value has no determinate relationship to output. Lachmann instead connects services to the depletion of resources through production. With constant production coefficients, this permits a relationship between output and the portion of capital used up during the chosen period. Annual depreciation allowances offer a practical starting point because business accounting already assesses results annually, despite the differing durations of firms’ production plans.

Section V develops this proposal through Solomon Fabricant’s distinction between capital consumption and capital adjustment. Yet Lachmann resists equating income-account charges mechanically with productive consumption and capital-account charges with external change. Accounting practice more closely distinguishes anticipated from unanticipated changes: predictable obsolescence enters depreciation, while inadequate allowances during inflation may eventually require a capital write-down.

The crucial distinction concerns whether an unforeseen event leaves a production plan workable or destroys it. Higher replacement costs can be incorporated into revised allowances while the plan continues. Technical or social changes may instead terminate the plan and require a new pattern of resource use.

Hence, in applying period analysis we have to select our period so as to make it co-extend with the carrying out of a coherent production plan.

Revaluations marking such discontinuities must be excluded from the measurement of capital consumed under that plan. Conversely, accounting errors cannot simply be discarded: their effects must be incorporated through corrected depreciation data. Following Fabricant, Lachmann proposes converting straight-line depreciation to a service-output basis and adjusting historical costs to reproduction costs using equipment ages and price indices.

The conclusion defends this approach against the charge that subjective business estimates undermine objective measurement. Heterogeneous capital must be measured through value relationships, which ultimately depend on subjective assessments; second-hand equipment markets do not necessarily coordinate these better than institutionalised accounting practices. Lachmann acknowledges unresolved problems, especially capital-goods price indices under technical progress and variations between sectors. The article’s contribution is therefore a qualified methodological proposal: measure capital’s productive consumption within coherent plans, rather than treating an aggregate asset valuation as an independently meaningful quantity.