Jati K. Sengupta and Gerhard Tintner, On Some Economic Models of Development Planning (1963)

Sengupta and Tintner’s journal article compares Dutch long-term planning with India’s Mahalanobis model to develop a more flexible framework for investment allocation. Its central claim is that maximizing growth within a fixed horizon does not suffice for operational planning: saving must support investment, production techniques and coefficients must permit adjustment, and policy objectives must account for consumption and uncertainty. The argument proceeds from a comparison of model structures through successive extensions of the Indian model to a numerical programming exercise. An appendix supplies capital-output estimates; multilingual summaries conclude the article.

The opening distinguishes the problems facing economies with different resource endowments. In less developed economies, planning seeks to raise the growth trend through efficient allocation of scarce capital; in developed economies, intertemporal balance and stability receive greater emphasis. The Dutch N-model seeks an optimal growth path consistent with supplies of labor, capital, and imports, checking aggregate coefficients against detailed intersectoral accounts. Its instruments include saving, emigration, relative export prices, and import substitution. The Indian M-model instead treats capital as the decisive constraint. It first allocates investment between investment-goods and consumption-goods production to maximize national income, then distributes investment among consumption subsectors to meet employment and income targets.

This apparently practical simplification carries strong restrictions:

The triangular coefficient matrix in (4) implies that the dependence of the consumption goods sector on the investment goods sector is unilateral.

Consumption-goods production cannot expand investment-goods supply through exports and imports within this structure. More generally, both models impose complementarity between labor and capital. In the Dutch case, fixed coefficients can prevent simultaneous capital and labor balance. The authors propose varying labor-demand elasticity, either instead of or alongside the saving ratio, so that adjustment does not depend entirely on saving.

For India, they derive the investment-income ratio implied by the model and examine the consequences of saving failing to follow it:

If the intended saving-income ratio does not grow in step with the investment-income ratio given in (8), there would be an imbalance in the system either in terms of unutilized capacity or commodity-cum-factor gaps.

Their calculations of consumption and income losses make balanced growth an operational requirement rather than merely a theoretical benchmark. A production path is not feasible simply because its sectoral investment proportions have been optimized. When available savings constrain investment, the resulting underutilization and foregone output must enter the planning calculation.

The extensions begin by questioning the welfare criterion itself:

From the standpoint of economic welfare, a more realistic criterion in a less developed economy may be the maximization of total consumption rather than total output.

Changing the objective alters the preferred allocation between sectors and its relationship to the planning horizon. The authors thereby expose the conditional character of an “optimal” commitment to investment-goods production: it depends on what is maximized, how far ahead the planner looks, and the assumed one-way dependence between sectors.

A second extension replaces investment growth mechanically determined by previous investment with additional investment responding to previous total income. A reaction coefficient and a shift factor allow this response to change. The conceptual gain is explicit:

Our result is much more general in that it explicitly introduces the possibility of reinvestment out of an increase in total income rather than investment goods.

Public investment policy must consequently consider its effects on saving and investment behavior as well as productive capacity. Long-run growth depends on the output-investment ratio and the strength of the income-induced investment response, not solely on the allocation to capital-goods production.

The authors next introduce diminishing returns through quadratic sectoral production relations. Optimal allocation then involves equalizing marginal investment productivities, and the original linear model becomes an approximation to a nonlinear system. They also explore substitution between labor and capital. With more general production functions, allocation cannot be decided independently of technique: choosing sectoral investment shares requires choosing capital-labor ratios.

The final section translates these concerns into programming under uncertainty. Using Indian Third Five-Year Plan resource constraints and estimated capital coefficients, the authors first maximize income alone. This allocates approximately 22.4 percent of capital and 11.4 percent of labor to investment-goods production. But that allocation does not minimize uncertainty in the investment component. Their revised objective rewards national income while penalizing the standard deviation of investment, with positive weights assigned by the policymaker.

The discrete numerical exercise illustrates how incorporating risk can favor allocations different from the income-only optimum. Its significance is not a universally preferred sectoral ratio, but a procedure for comparing allocations under explicit priorities and uncertain coefficients. The authors envisage extending it to additional sectors, time-varying coefficients, and estimated confidence intervals. Across the article, planning thus becomes a conditional choice among feasible, welfare-sensitive, and risk-bearing alternatives rather than the mechanical execution of a single growth formula.