P. N. Rosenstein-Rodan, “Alternative Numerical Models of the Third Five Year Plan of India” (1964)

This brief single-authored technical note in development economics tests the realism of India’s Third Five Year Plan by recalculating its fiscal and growth magnitudes under alternative assumptions. Its scope is deliberately narrow and numerical: from a base of 12,500 crores national income, 10,000 crores total investment, and 15,700 crores government expenditure, Rosenstein-Rodan compares three models distinguished by the capital-output ratio, marginal savings rate, and foreign aid requirement.

We shall also consider the ways of financing Indian development if the marginal rate of savings in India were not 27 per cent (and 38 per cent if the capital-output ratio were to be 3:1) as implicitly assumed in the Pant and Little memoranda, but considerably lower, within the range of 20 to 25 per cent (say 23 per cent).

The note’s central thesis is that the Pant-Little assumptions are too optimistic if the true capital-output ratio is nearer 3:1 than 2·2:1. Rosenstein-Rodan’s key conceptual move is sensitivity analysis: he shows that a single technical coefficient alters growth, tax yields, savings demands, borrowing feasibility, consumption growth, and the required scale of external assistance.

The extent of the changes between models $a$ and $b$ (and $c$) show how crucial the capital-output ratio is, i.e. how highly other variables depend on its assumed value.

Model $a$, the Pant-Little case, assumes a 2·2:1 capital-output ratio and yields 6 per cent annual national-income growth. Even here, Rosenstein-Rodan notes that the combined burden of new taxation, government enterprise surplus, borrowing, and rising savings may be excessive. The deeper objection, however, is empirical: India’s agriculture, investment composition, and gestation lags make the low ratio doubtful.

The assumption of a capital-output ratio 2·2:1 seems, however, to be dangerously optimistic.

He therefore shifts attention to the more realistic 3:1 ratio. This is not merely a pessimistic adjustment; it reflects a planning philosophy attentive to delays between investment and output and to the limited reliability of “record” performance across sectors.

While only a more detailed analysis of possible alternative compositions of investment and of the lags inherent in them could give a more reliable estimate, it seems far more realistic to assume a high capital-output ratio of 3:1.

Model $b$ keeps foreign aid at 1,100 crores while adopting the 3:1 ratio. Growth falls to 4·8 per cent, existing taxes yield less, taxes would have to rise to 16 per cent of national income, and the implied marginal propensity to tax reaches 39 per cent. The savings implication is still more severe.

Total Savings in model $b$ are assumed to rise from 8·4 per cent of National Income in the first year to 15 per cent in the fifth—implying a marginal rate of savings of 38 per cent—which seems indeed to be far too heroic a target.

The intervening discussion of savings is the analytical core of the note. Corporate savings may have high marginal rates, but the corporate sector is too small; private savings form the bulk but cannot be assumed to accelerate indefinitely; government savings can be raised by policy effort, but not through a cumulative yearly process. Hence the feasible marginal savings rate is closer to 23 per cent.

Taking the three groups of savings into account, a combined marginal rate may at best be estimated at somewhere between 20 and 25 per cent—say 23 per cent, as considered in model c—but it is not very probable that it should reach 27 per cent, not to mention the very high rate of 38 per cent in a nontotalitarian economy without too drastic austerity measures.

Model $c$ is Rosenstein-Rodan’s preferred reconstruction: capital-output ratio 3:1, marginal savings 23 per cent, and foreign aid of 3,000 crores, of which 2,500 crores is available for net investment and 500 crores for refinancing short- and medium-term external debt. Aid fills not only the foreign-exchange gap but also a domestic resources gap.

Accordingly, a much higher amount of foreign aid, which alone can secure success in achieving a rate of growth of 4·8 per cent per annum, is assumed here.

The note’s relevance lies in its disciplined refusal to separate growth targets from institutional feasibility. Rosenstein-Rodan turns development planning from aspirational arithmetic into a consistency test among output ratios, savings behaviour, taxation, borrowing, consumption, and aid. Its conclusion is that India’s plan can be made plausible only by accepting a higher capital-output ratio, a moderate savings target, and substantially larger foreign assistance.