Gerhard Tintner, The Theory of Production Under Nonstatic Conditions (1942)

Gerhard Tintner’s journal article develops a nonmathematical account of production planning when future prices, interest rates, and productive conditions cannot be anticipated with certainty. Its central move is to replace the maximization of a single anticipated profit with the evaluation of a probability distribution of possible profits. Production decisions depend both on what the firm anticipates and on how it values the resulting prospects. The article first distinguishes four cases—subjective risk, subjective uncertainty, technological risk, and technological uncertainty—and then illustrates their logic through a two-period production model and a sequence of graphs.

Flexibility and adaptability provide the intertemporal foundation of the argument. Present production does not merely generate present output: it alters the conditions under which later production takes place.

The amounts of products produced and factors used in the processes of production at an earlier point in time influence the conditions of production at a later point.

Consequently, a firm may organize its initial activities to preserve advantageous responses to future contingencies. Tintner contrasts this problem with Hicks’s treatment of “single-valued” anticipations. When all future prices and productive conditions are known, dated inputs and outputs can be treated as distinct commodities, and the firm maximizes anticipated discounted net profit through familiar marginal conditions. Time alone does not require a fundamentally different theory of choice. The distinctive problem arises when alternative futures must be evaluated before present commitments are made.

Under subjective risk, technical and technological conditions remain perfectly known, but anticipated prices and interest rates have a joint probability distribution. This induces a distribution of total discounted net profit. Tintner introduces a “risk preference functional” to describe how the firm ranks such prospects, refusing to identify rational planning exclusively with maximizing expected profit.

The individual in question may easily take other features of the probability distribution of anticipated net profits into account—for instance, its dispersion, its skewness, kurtosis, etc.

Drawing on Menger and Marschak, Tintner makes the valuation of the distribution a separate component of production theory. Expected-profit maximization is a special case rather than a universal behavioral rule. Once a preference functional is specified, the firm chooses planned inputs and outputs across the relevant periods to maximize it. These choices constitute a plan at the initial date, not an irrevocable commitment: subsequent information or changed anticipations may justify revision.

Subjective uncertainty adds another probabilistic layer. Under risk, the firm confidently anticipates a particular distribution; under uncertainty, it assigns probabilities to different possible distributions. Their means, standard deviations, correlations, or other defining characteristics may themselves be uncertain. Tintner thus represents uncertainty through prior probabilities over the forms of distributions, allowing still higher orders when those priors are not confidently known.

Using the joint probability distribution of anticipated future prices and interest rates, together with the a priori probability distribution, we can derive again a probability distribution of anticipated total discounted net profits.

This preserves the earlier decision procedure: derive profit prospects, evaluate them through a preference functional, and select the production plan. Tintner’s uncertainty is therefore not an absence of probabilistic representation. It is uncertainty about the probability law itself, brought within a generalized framework of probabilistic choice.

The technological sections shift attention from market conditions to productive possibilities. Tintner distinguishes technical conditions within a given state of the arts from technological change, including inventions that alter production functions. Both can generate uncertain coefficients or parameters determining the relationship between inputs and outputs.

Many of the situations of risk and uncertainty which arise in economic life come not so much from anticipated price and interest contingencies as from anticipations of changes in technical and technological conditions.

Obsolescence makes this extension economically important. Agricultural yield fluctuations exemplify technical variability without a change in the state of the arts; yields may also be correlated with product prices. Technological change, by contrast, may affect both product and factor prices. The argument therefore accommodates relationships between productive and market contingencies rather than treating their distributions as necessarily independent. Technological risk assumes a known distribution of production parameters; technological uncertainty assigns prior probabilities to alternative forms of that distribution.

The graphical examples make the planning procedure concrete. A firm uses one variable factor to produce one commodity in two successive periods. Greater first-period input reduces the second-period input needed for a given output, a carryover effect illustrated by applying lime to agricultural fields. With certain prices, ordinary profit maximization determines both periods’ activity. With economic risk, the second-period factor price instead follows a triangular distribution. For each initial input choice, Tintner derives the best subsequent response and then the resulting distribution of total profit. An illustrative preference functional rewards a higher mean, penalizes standard deviation, and rewards positive skewness.

The uncertainty example makes the range of that price distribution itself uncertain. The technical-risk example instead assigns a uniform distribution to a parameter governing second-period productivity. Each repeats the same conceptual sequence: initial commitments shape later opportunities; optimal later responses generate profit distributions; preferences over those distributions determine the initial plan. Technological uncertainty is left as an analogous extension rather than illustrated separately. The examples are constructed demonstrations, not empirical estimates or universally prescribed attitudes toward risk. Their significance lies in showing how production theory can integrate adaptable planning, uncertain technology, and heterogeneous preferences without reducing every decision to expected profit alone.