Knowledge that Transforms

John A. Brittain, A Regression Model for Estimation of the Seasonal Component in Unemployment and Other Volatile Time Series, The Review of Economics and Statistics, Vol. 44, No. 1 (Feb., 1962), pp. 24-36

T HE present paper represents an attempt to add a further empirical example to the little that we know about how producers of primary products in less-developed countries respond to changes in the relative prices of the commodities which they produce. While knowledge of the manner and magnitude of such behavior is of basic importance in dealing with questions of economic analysis and economic policy in underdeveloped economies,' it is only recently that efforts have been made to document this behavior empirically.2 example to be presented here deals with the acreage adjustment made by Indian cultivators of jute and competing crops in response to changes in the relative price of jute.3 cultivation of jute for commercial purposes has long been a mainstay of what in undivided India was the province of Bengal, and of part of present-day East Pakistan. While jute was also grown in the provinces of Bihar and Orissa, Bengal, with about go per cent of Indian and world production, was the dominant producing region.4 Although jute is an important export commodity, jute cultivation is small relative to that of rice. Tables i and 2 below show that rice is the main staple crop grown in the Bengal districts. However, the agronomical similarity of jute and rice, as well as the considerable experience of producers in their planting and harvesting are indications that the two crops are competitive in production.5 crops chosen for study are typically grown by peasants with small land holdings similar in organization to those in many of the other heavily populated, underdeveloped regions of Southeast Asia. In the census of I92I, for example, it was estimated that an ordinary in Bengal worked approximately 3.I acres, which was substantially lower than in a number of the other Indian provinces at that time.6 First, we will examine the area and price responses of jute growers in East Pakistan following Partition, from I949/50 to I959/60; an attempt will then be made, using simple regression analysis, to determine the magnitude of the elasticity of this response in the undivided Indian provinces of Bengal, Bihar, and Orissa for *I am indebted to James O'Connor, David Schwartzman, and Elliot Zupnick for helpful suggestions, to Michael Dakolias for computational assistance, and to the Columbia University Council for Research in the Social Sciences for financial aid in connection with the research on which this paper is based. 1 relationship of production to price is crucial in discussions of the stabilization of prices and incomes of primary producers. existence of such a (positive) relation was, for example, one of the key assumptions employed by the late Ragnar Nurkse in the stabilization proposal contained in his Fluctuations and Buffer Policies of LowIncome Countries, which formed the basis of the symposium, The Quest for a Stabilization Policy in Primary Producing Countries, Kyklos, XI (I958), Fasc. 2. 2See for example my contribution, The Price Responsiveness of Egyptian Cotton Producers, Symposium Stabilization and Development of Primary Producing Countries, Kyklos, XII (I959), Fasc. 3; and P. T. Bauer and B. S. Yamey, Case Study of Response to Price in an Underdeveloped Economy, Journal, LXIX (December I959), 800-805. In addition to the foregoing studies which focus upon cotton and cocoa, discussion of the response to price of producers of numerous other primary products can be found in William 0. Jones, Economic Man in Africa, Food Research Institute Studies, I (May I960), esp. II5-I26; P. T. Bauer, Rubber Industry (Cambridge, I948), esp. 28-30 and 360-36I, and West African Trade (Cambridge, I954), 425-427; H. Kitamura and S. Yang, Domestic Stability and Development: A Critique of Nurkse's Scheme, Symposium II, Kyklos, XII (I959), Fasc. 3, 3I7-3I8; and Brij Raj Chauhan, Rise and Decline of a Cash Crop in an Indian Village, Journal of Farm Economics, XLII (August I960). 'A focus similar in many respects to the one to be presented can be found in FAO, Jute, Commodity SeriesBulletin No. 28 (Rome, I957), esp. 20-24. 'Jute acreage and production data can be found for the provinces of British India for the years, I890/9I-I943/44, and for the other very minor producing countries -Formosa, French Indo-China, Japan, and Iran -from I908/09 onwards in the mimeographed study by Horace G. Porter and Maurice R. Cooper, Statistics on Jute and Jute Manufactures, With a Brief Survey of the Industry, U.S. Department of Agriculture (Washington, June I945), 29-32. See also FAO, Jute, 63-64. 51bid., 20-2I. Shown in Table 2 is the total area devoted to ten other crops, which are also conceivable substitutes, and will consequently be considered in the production relationship to be studied. 'See the data reproduced in Albert Howard and Gabrielle L. C. Howard, Development of Indian Agriculture (London, I927), 3. number of acres per cultivator in 195I in East Pakistan was 2.2. This estimate, computed from the Pakistan Statistical Yearbook of 1955, is, however, too low because it is based upon the acreage of the principal, rather than of all, crops planted in East Pakistan. It indicates nonetheless the typically small holdings of peasants in this region.

in the determination of consumption patterns is rejected. The calculated marginal propensities to consume vary greatly for both races in different sections of the country (Table i) and this fact gives additional reason for rejecting the theory. The inference is clear that factors other than race are at work in the determination of consumption patterns. Although the data examined here do not rule race out as a factor in the determination of consumption patterns, they cast such serious reflections on the hypothesis as to make it, with all charity, useless. There are too many uncontrolled variables to warrant the conclusion that race alone is responsible for the difference in consumption patterns. It has become clear in this study that the concept of race as a factor in the statistical analyses of group economic behavior, such as in the assessment of so-called Negro-White savings and consumption patterns, has no more validity than left-handedness, eye pigmentation, or height. Before any realistic comparison is made on the amount spent on consumption by Negroes and Whites, adjustments should be made for the difference in their income expectation, age, dependents, education, location, financial reserves, and similar factors. Once these adjustments are made, differences attributed to race may find their origin in other factors. For example,

T HE employment model described in this article was fitted to semi-annual data covering the years 1947 through 1959. The model is believed of interest for the following reasons: i. It allows a good fit for the employment function over the sample period and it produces a rather successful forecast for the first half of I960. 2. Because the employment function contains two interdependent variables, the method of two-stage least squares was applied in its estimation. The advantage of this method from the standpoint of forecasting is demonstrated. 3. Finally, in the construction of the model all but one of the predetermined variables is lagged. Consequently, the reduced form of the model is highly suited for forecasting purposes.'

to study policies affecting money and credit separately from fiscal policy, because the underlying cause-effect relations are so very different. In terms of a summary characterization of the new budget, what I have tried to convey is this: Let us not say that it is designed to show the impact of economic policies; let us rather say more modestly that it shows the impact of fiscal policies; and let us make clear that a detailed and comprehensive analysis of monetary and credit policies cannot be conducted within the framework of the NIP accounts.