World Bank Reprint Series: Number Sixty Montek Ahluwalia Rural Poverty and Agricultural Performance in India Reprinted from The Journal of Development Stuidies (1977) The most recent editions of Catalog of Publications, describing the full range of World Bank publications, and World Bank Research Program, describing each of the continuing research programs of the Bank, are available without charge from: The World Bank, Publications Unit, 1818 H Street, N.W., Washington, D.C. 20433 U.S.A. WORLD BANK BOOKS ABOUT DEVELOPMENT (FOR SALE BY THE PUBLISHER INDICATED) Research Publications International Comparisons of Real Product and Puirchiasinig Powerby Irving B. Kravis, Alan Heston, and Robert Summers, published by The Johns Hopkins University Press, 1978 Experiments in Family Planniinlg: Lessons from the Developing World by Roberto Cuca and Catherine S. Pierce, published by The Johns Hopkins University Press, 1978 Income Distribution Policy in the Developing Countries: A Case Study of Korea by Irma Adelman and Sherman Robinson, published by Stan- ford University Press (in the Commonwealth, Oxford University Press), 1978 Interdependence in Planning: Mtultilevel Programminlg Studies of the Ivory Coast by Louis M. Goreux, published by The Johns Hopkins Univer- sity Press, 1977 The Mininzg Industry and the Developin1g Countl.ries by Rex Bosson and Bension Varon, published by Oxford University Press, 1977 Patterns in Household Demand and Saving by Constantino Lluch, Alan Powell, and Ross Williams, published by Oxford University Press, 1977 Unskilled Labor for Development: Its Economic Cost by Orville McDiarmid, published by The Johns Hopkins University Press, 1977 Electricity Economics: Essays and Case Studies by Ralph Turvey and Dennis Anderson, published by The Johns Hopkins University Press, 1977 Housing for Low-Income Urban Families: Economics and Policy in the De- veloping World by Orville F. Grimes, Jr., published by The Johns Hopkins University Press, 1976 Village Water Supply: Economics and Policy in the Developing World by Robert Saunders and Jeremvy Warford, published by The Johns Hopkins University Press, 1976 Economic A nalysis of Projects by Lyn Squire and Herman G. van der Tak, published by The Johns Hopkinis Univereity Press, 1975 The Design of Ruiral Developmient: Lessons from Africa by Uma Lele, published by The Johns Hopkins University Press, 1975 Economy-Wide Models and Development Planning edited by Charles R. Blitzer, Peter B. Clark, and Lance Taylor, published by Oxford Uni- versity Press, 1975 Pattern1s of Develovpnent, 1950-1970 by Hollis Chenery and Moises Syrquin with Hazel Elkington, published by Oxford University Press, 1975 A Systemn of International Comparisons of Gross Prodtuct and Pulrchlasing Power by Irving B. Kravis, Zoltan Kenessey, Alan Heston, and Robert Summers, published by The Johns Hopkins University Press, 1975 (contiriued on inside back cover) Rural Poverty and Agricultural Performance in India hbi Montek S. Ahluw-aliO* Thlis paper exunhilnes time series evidence on ruratlpovertl over the past twvo dl'dL'.s. The time sries showrs that the incidence of poverty, fluctuates in re.pon.s' to variations. in real agriicultural olutput per headi, but there is nzo significant time trend. Thlere is a statisticallyt signifi cant in verse ,ela i ioai.'iip between rural poverty* and agricultural per fan nan ce Jbr India as a whole, suggesting that a-griculltral growvth by itself tends to reduce the incidence of poverty. The analivsis for individual states presents a somewhat different picture. The in verse relationisliip between output per head and ruralpovertv is observed in several states but thlere is also evidence that there are processes at wvork which te;id to inierease the incdidene of poverty, independenithi o 'variations in agricultural output per head. 1. INTRODUCTION Recent years have seen the developnment of an extensive and disqluieting literature on trends in rural poverty in India and their relationship to agricultural growth. A recurring theme in much of this literature is that agricultural growth has been accompanied by a steady deterioration in distributional terms, involving not onlv an increase in relative inequality but also an increase in absolute impoxelrishmenit. Indeed, it is argued that these trends are the natural consequence of the type of agricultural growth which can be expected ;within the existing institutional structure in Indian agriculture.' This latter proposition has important implications for policy. It raises doubts about the scope for achieving even the fairly minimal welfare objective of alle-iating absolute poverty in the future, at least through the kind of agricultural development that is currently deemed feasible, i.e., growth without radical institutional change. The object of this paper is to evaluate the empirical basis for this assessment of past trends and future prospects. The principal sources of data for our study are the various consumption surveys conducted by the National Sample Survey (NSS) which report the distribution of the population across per capita expenditure classes. These surveys have been used in several existing studies on rural poverty but these studies typically have not made full use of the available information. Some of them, for example, Rajaraman [1975] and Lal[1976], relyuponcoinparisons between two arbitrarily chosen points in time. Bardhan [1971] reports four *The author is Chief of the Income Distribution Division of the World Bank's Develonment Research Center. The opinions expressed in this paper are those of the author andnotof theWorld Bank. Iam indebted to Shail Jamnrorcompuier programming assistance and to P. Bardhan, B. S. Mlinhas, M. Lipton. T. N. Srinivasan, 0. Yenal, S. Kuznets and A. Vaidyanathan for many helpful comments. RURAL POVERTY AND AGRICULTURAL PERFORMlANCE IN INDIA 299 observations for India as a wihle between 1960-1 and 1968-9 but only two for the individual states. An . ;ry study by Minhas [1970] was based on data for seven years, but his time series extends only up to 1967-8.2 In this paper we will expand the data base cover NSS data for 14 different years spanning the period 1956-7 to 1973-4. This expanded data set provides the basis for a systematic time series analysis of trends in ruiral poverty for India as a whole as well as for individual states. Throughout this paper, our concern is principally with the extent of absolute poverty in rural India, defined with respect to a fixed poverty line in terms of real per capita consumption. We have attempted, first, to document changes in incidence of poverty over time, and second, to relate these changes to some measures of agricultural performance, The analysis is based on two alternative measures of the extent or incidence of absolute poverty. The first measure is the percentage of the rural population below the fixed poverty line. The second measure is Sen's Poverty Index, which takes account not only of the percentage of the population in poverty but also of the gap between the poverty line and the mean consumption of the poor, as well as the extent of inequality amongst the poor. While the bulk of the debate has been conducted in terms of the percentage measure, the Sen Index has obvious advantages in measuring the true intensity of the poverty problem.3 The paper is organised as follows. Section II deals briefly with the construction of poverty lines in terms of per capita current expenditure. Section III presents our results on trends in the incidence of rural poverty over the period 1956-7 to 1973-4 for India as a whole as well as for the individual states. Section IV attempts to relate observed changes in poverty to agricultural performance. A summary view of the evidence on changes in rural poverty and factors affecting these changes is presented in Section V. II. THE POVERTY LINE IN CURRENT PRICES The first step in our analysis is the definition of an appropriate poverty line for measuring absolute poverty, A fundamental limitation of this approach is that any such line is necessarily arbitrary. In this paper we finesse the problem by choosing our poverty line primarily to conform to past practice, without attempting to justify it as measuring some objectively defeiisible minimum standard. Accordingly, the poverty line used throughout this paper is a consumer expenditure level of Rs. 15 per person for 30 days at 1960-1 rural prices. This line has a well-established pedigree in the Indian literature, As shown by Bardhan [1971], an expenditure level of Rs. 15 in 1960-1 at rural prices roughly corresponds to Rs. 20 per person at all-India 1960-1 prices, which is the minim um level originally adopted by the Pianning Commission in 1962.4 Dandekar and Rath [1971] also adopted this line on the grounds that it corresponded to the expenditure level at which food consumption (on average) provided the 'norm' of 2250 calories per day. However, it is important to emphasise that attempts to interpret this line as guaranteeinig a nutritional minimuLm could be seriously misleading.5 Suffice to say that this level of expenditure represents an extremely low level of living and one that has been widely accepted as a 'minimum level' in the policy debate. TABLE I Rt'RAl. POVERTY LINFS; CONSUMPTION PER PERSON FOR 30 DAYS (RS. IN (C'RRENT PRI('CE.S) 1956-57 1957-58 195'-6(J 1960-61 1961-62 1963-64 1964-65 1965-66 1966-67 Jti - I'S 1968-69 19 t- 1973-74 Andhra Pradesh 141 141 154 1 5-5 15-7 16i2 194 21-2 24.3 25-O 25.9 37-5 Assam 1sf) 17(0 15-8 163 161 18-4 214 23-6 310 35-7 33-9 x 42-9 Bihar 15-6 17-1 16-1 1> 163 18-6 23-7 28-3 36-2 39's 29-5 32-5 53-2 Gujarat 16-5 165 158 16X8 171 17-8 22-3 22-9 257 26-9 26-9 29-1 41-3 Karnataka 13-9 139 15-4 15-6 156 17-1 22-8 26-7 27-9 29-2 284 29-3. 42-9 Kerala 15-9 15-2 16-3 161 17-1 17-6 21-3 24-2 25-9 27-7 31-4 34-5 44.4 N1adhNj Pradesh 142 14-8 14-2 141 14-5 16-8 19-6 22-0 28'5 30-9 27-5 27-9 43-6 Maharashtra 157 158 16-5 16( 15.? 17-7 24-2 25-4 28-0 29-3 28-3 30(7 44-2 Orissa 13-8 14(1 144 14-5 14X8 19() 20-7 23-5 27-7 3(1-3 31-6 31-7 40-9 Punjab & Harvana 15-7 16-0 164 15-9 16-S 18-2 22-1 21-9 27? 30-7 3(0- 30-8 43-4 Rai,isthin 14-1 13-5 14-7 13-8 152 19-4 20-6 24-6 25-4 26-8 '5-4 41-7 Tamil Nadu 16-1 16-1 17 - 16-4 18-5 198 22-6 23-5 28-7 28-2 29-0 28-5 39-7 Uttar Pradesh 14-6 15-6 14-9 14-5 15-2 19-3 23-8 23-9 30-6 34-2 26-0 26-5 43-9 West Bengal 18-1 19-5 19-9 18-I 19-0 24-0 -6 25-3 30-2 43-6 36-0 37-3 50-0 ALL INDIA 14-6 15-1 153 15-0 15I-5 17- 21-3 23-3 28-5 30-9 27-8 28-8 42-9 Sources: The Consumer Price Indices for A gricu I tu r I Labourers used in prepLi ri n r hibi table relate to the agricultural year July-June and are taken from : the Indian Labour Journal. This source does not report estimates for some of the states (Tamil Nadu. Uttar Pradesh and RPjaNch I n) for years z preceding 1964-5. For these states we have used estimates prepared bv Jwoe [/9-4j supplemented by Lal [197T]. Price indices for 1957-8 were H obtained b- ,i%eraging calendair )eyir- 1957 and 1958 from the Intilitt Labour Journal except in the case ofGuaraLt. Maharashtra, Tamil Nadu. and Uttar Prides,h. "here we use averages of 1956 -7 and 1958 -9 as reported by Lal [1976]. i-i RURAL POVERTY AND AGRICULTURAL PERFORMIANCE IN INDIA 301 Since commodity prices vary significantly across states, the same real consumption level requires different levels of consumer expenditure across states. Bardhan [1971] estimated the level of consumption expenditure for each state in 1960-1, which is equivalent to consumption expenditure of Rs. 15 at all-India rural prices. We have adopted these estimates as our base year estimates of the poverty line in each state. The second step in our analysis is the definition of equivalent poverty lines for different years in terms of consumer expenditure in the current prices for each year. This requires identification of a suitable price index for the rural poor. Price indices for the 'average' consumer are clearly not suitable since the poor spend a much greater proportion of their budget on items whose prices displayed very high inflation rates in the 'sixties (e.g. food and especially coarse grains).6 In the absence of price indices specially designed for the rural poor. we have followed Bardhan [1971] in using the Consumer Price Indices for Agricultural Labourers (CPIAL), prepared by the Labour Bureau, as the most appropriate for our purpose. These indices are available for India as a whole and separately for each state. Applying the CPIAL to the base year estimates of the poverty line in terms of consumption expenditure per person in 1960-1 prices, we can calculate equivalent poverty lines for each of the years for which NSS consumption distributions are available. The resulting estimates of the poverty line for each state, and for India as a whole, for fourteen years spanning the period from the late 'fifties to the early 'seventies, are shown in Table 1. These estimates are obviously subject to all the limitations arising from the use of the CPIAL as the price index. Ideally, we should use separate price indices for different groups comprising the poor, especially distinguishing landless l-bourers relying on wage income from subsistence farmers who rely upon own consumption.,8 More seriously, it can be argued that the very approach of using a base weighted price index is flawed since it cannot reflect the impact of changing relative prices upon the commoditN conmposition of consumption,9 In defence of our estimates we can only state that they are certainly in line with past practice, and are probably the best estimates possible given available data. III. TRENDS IN RURAL POVERTY 1956-7 TO 1973-4 The poverty lines presented in Table 1 have been used in conjunction with the NSS consumption distributions to estimate our two alternative poverty measures: the percentage of the rural population below the poverty line and the Sen Poverty Index. These estimates are obtained from a two-stage procedure. First, we estimate the Lorenz curve of the consumiiption distribution for each year using the method of Kakwani and Podder [1976]. The estimated parameters of the Lorenz curves are then used to obtain point estimates of the two poverty measures."' This procedure has been used to estimate the incidence of poverty over time for rural India as a wvhole as well as for each state separately, (a) All-India Results Our estimates of the two poverty mneasures for rural India as a whole are presented in Table 2. The NSS data permit two different estimates of the TABLE 2 NSS BASE)F S I If AlII S 0)F Rl.RAI POVL:RTY IN INNDIA Penrc entage of' Rutra*l Povety Si ze ot Pol'e'lrty- Population Populationll il Povl-ilu 1(1ujlrnx Derived from Derived froml Ehslinuet I Eiv)tiniate /I Etinimte I Evtinote II 1956- 7 541 n.a. 0-23 181-0 n.a. (Sub-sample 1) (53-5) (0-22) (Sub--,ample 2) (54-7) (0-24) 1957 8 50-2 53*4 0-22 171-0 182-0 (Sub-s;ample 1) (48-6) (021) (Sub-N.ample 2) (51-7) (0-)3) 1;58 9 46-5 n.a. 0.19 162-0 n.a. (Sub-sample 1) (47-9) (0-20) (Sub-sample 2) (44-9) (0.19) 199 6Hi 444 48-7 0-17 158-() 1730 (Sub-sample 1) (463) (0I18) (Sub-sample 2) (42-5) (0-16) 1960- 1 38-9 42-0 0(14 141.0 152-0 iSub-sdirple I) (384) (((14! (Sub-samIlple 2) (39-3) (0-15) 1961 2 39-4 42-3 0-14 .46-0 157-0 (Sub-sample 1? (41-5) (015) (Sub-sample 2) (39-8) (0-14) (Sub--Jniple 3) (36-9) (0-13) 1963 4 445 49-1 0-16 1710 189.(0 (Sub-sample 1) (460) (0-17) (Sub-sample 2) (45-7) (0-17) 1964 5 46-8 50-4 O. (X8410 198-C0 (Sub-sample 1) (46-8) (0-18) (Sub-sample 2) (468) (0 17) 7 1965 6 539 511 0)21 216-0 205-0 z (Sub-samiple 1) (54-7) (0-22) (Sub-sample 2) (53-0) (0-21) 1966 -7 56-6 57-4 (0-24 231() 235-0 1967- 8 56-5 57-9 0-24 235-0 241-0 1968 9 51-0 53-S 0-20 2 17-0 227-0 1970- 1 47-5 49-1 (.18 210-0 217-0 1973-4 46-1 47-6 0-17 214-0 21-0 *As explained in note 3, this index ranges from 0 to I1 RURAL POVFRTY AND AGRICIULTURAL PERFORMANCE IN INDIA 303 percentage of the rural population in poverty in India as a whole. Estimate I is obtained by applying the all-India poverty line for various years (see Table 1) to the NSS consumption distribution for rural India. Estimate II is obtained as a weighted sum of the estimated percentages in poverty in individual states, obtained from the NSS distributions for individual states and the state specific poverty line. As there were substantial interstate differences in prices in the base year, and furthermore, inflation occurred at different rates across states, it can be argued that Estimate II, which is based on state specific poverty lines, is a better estimate of the percentage of the rural population in poverty. The most important feature of the results presented in Table 2 is the marked fluctuation over time in the extent or incidence of rural poverty. The percentage in poverty declines initially from over 50 per cent in the mid-'fifties to around 40 per cent in 1960-61, rises sharply through the mid- 'sixties, reaching a peak in 1967-8, and then declines again, The Sen Index also displays the same pattern. Since this index reflects not only the percentage below the poverty line, but also the average shortfall of this group from the poverty line, it is reasonable to conclude from the range of variation in this index that we are measuring substantial fluctuations in the intensity of poverty and not merely nmaiaginal shifts of large numbers from a position slightly above the poverty line to a positiorn slightly below. It is important to determine whether the observed fluctuations arise solely from the samlpling variation over time in our estimates, or whether they reflect genuine chantges in the incidence of poverty arising from underlying economic factors. Fortunately, NSS surveys are conducted on the basis of interpenetrating sub-samples and differences between estimates of the incidence of poverty based on different sub-samples for the same year provide an indication of the range of variation in sample estimates. Separate tabulations of the data by sub-sample are available for the earlier years (up to 1965-6) and have been used to obtain sub-sample estimates of each of our poverty measures for India as a whole (see figures in parentheses in Table 2). The range of variation between sub-sample estimates for the same year is clearly much smaller than the variations observed over time. This suggests that the measured fluctuation in the incidence of poverty reflects real changes in the severity of the poverty problem over time. The existence of fluctuations over time implies that we cannot generalise about uinderlying trends on the basis of comparisons between selected endpoints. For example, Bardhan [1971] reported a sharp increase in the incidence of rural poverty between 1960-1 and 1968-9 while Lal[1976] has argued that the incidence of rural poverty declined between 1 956--7 and 1970-1.11 Our estiimates for these years (derived by a different estimation method) confirmni the direction of change between these particular endpoints as reported in each study, but they also point to the danger of using such comparisons for any assessment of underlying trends. This can onlv be done on the basis of the time series as a whole. A linear time trend fitted to each of ourtwo estimates ofthepercentage in poverty and our estimates of the Sen Index yields the following results (terms in parentheses are t ratios). 304 JOURNAL. 0[ DEVLL()PNI'IFNT S1T'l)I[ S Estimiate I 46&56 + 021 1 T RI = 004 (15 25) (0(67) Estimate 11 = 47 74 + 0(262 T R2 = 0 06 (14 10) (0'81) Sen Index = (191 + 0(0001 T R =000 (10 (06) (0(04) These results provide no e. idenice for asserting a trend increise or decrease in rural poverty over the period as a whole. The increiasin[g incidenlce of poverty over the 'sixties repor-ted by Bardhan appears as an uipswing in a pattern of cyclical x ariation. We note that our conicilision also differs from M,Tinhias [1970]. whoi reported a decline in the percentage of the rural populaLtion in po\ ert\ over thc period 1956 7 to 1967 8. This difference is due to identifiable differences on some kev issues in estimating rural poverty. 12 The absence of any discernible trend in the incidence of poverty bcahiousLe impilies an increase in the absolute number-s of people in povert; beca .e of the arowth in rural popLIaItion over the period. The .ibsollite size of the poor population also flLctuates ovcr tlhe period, as shio-wrn in Table 2, but in this case, fittine a time trend to each estinmate (Pl and P. re.pepctivel) yields the tolltm in- results: P1 = 15(136 + 4'546 R2 = 0.54 (11 74) (3'76) PI = 153-99 + 4855 T R2 = 060 .( I 1(.6) (3 84) In otlier words, the aibsoluite iiiiniber of the rutral poor has grown sionifilcantl oxer time, increarsing on average by ahbout five million every year. (b) Relts dyfi In- huividhl .Stat's Our estimiates of the twvo po%erty measures for individual states are r eporited in Tables 3(a) and 3(b) respectively. In general, the time pattern of the incidence of poverty in individual statles follo%'.s the pattern of fluctuIation described for) India as a whole. The inicidence of poverty for almnl(ost all states, using either of lhe two poverty mlealsuil-es, declinies up to the earl\ 'sixties and thien begins to rise again,. rcaching a peak in 1967- 8 or 1968 69, and (declininiig again thercafter. Ontce aiain, \\c hiive tested for the existence of an underlying tr-eni(d by fitting a linear time trenid to eaclh of oLr poverty 1neIsuIres for individiual mtate.s. The resuflts are sLuimimiiiarlised in Table 4. Only two states (Assamil aind West Bengal) show a significant trend increase in poverty, according to both our poverty mneasires. Andhlir;i Praidesih and Tamil adlu show a trend decline in the incidence of poverty although in Tamil Nadu this is significant onlyv when we use the Setn Index. For all othiei states there is no significant time trenid in the inicidence of rural poverty using either measure. TABLE 3(a) PERCENTAGE OF RURAL. POPULATION IN POVERTY aY STATFS 195'-8 1959 -60 1960-1 1961--2 1963-4 1964 5 1965-6 1966-7 196 7 V 1968-9 1970-71 1973-4 0 Andhra Pradesh 53-5 48-8 50-1 47-2 45-6 41-5 45-4 47-9 46-0 47-3 41-0 39-8 Assam 28-0 31-4 25-6 29-4 24-4 24-2 31-3 46-8 38-4 47-3 35-3 39.3 > Bihar 59-7 55-7 41-5 49-9 52-3 54-3 59-4 74-4 70-9 594 59-0 584 Z Gujarat * 41-5 31-6 39-7 45-7 49-8 50-7 54-1 50-8 42-8 43-8 35-6 Karnataka 41.3 48.9 39-1 35-4 50-5 55-1 63-9 59-5 56-9 58-8 47-2 46-9 0 Kerala 59-6 62-3 57-8 50-3 52-8 607 70-7 67-1 63-4 64-6 62-0 49-3 S Madhya Pradesh 57-7 46-4 43-8 40-0 43-6 42-1 47-2 58-3 62-3 56-0 52-9 52-3 Maharashtra * 54-5 48-4 43-6 48-2 59-1 57-8 63-2 57-2 54-8 46-6 49-8 Ofissa 66-6 63-4 62-4 49-3 600 61-9 62-1 64-2 64-7 71-2 65-0 58-0 Punjab 2 Haryanat 28-0 24-2 18-8 22-3 29-4 26-5 26,5 29-5 33-9 24-0 23 6 230 o Rajasthan 33-4 n.a. 32-3 33-0 326 31-8 308 37-1 35-9 41-4 41-8 29-8 > Tamil Nadu 67-8 64-4 53-9 51-0 52-0 57-4 59-5 62-7 58-1 60-6 57-3 48-3 Uttar Pradesh 52-3 36-7 37-9 35-4 56-6 53-7 47-1 55-2 60-2 46-4 40-6 47-3 West Bengal 62-3 61-4 40(4 58 3 63-3 64-0 56-5 64-3 80-3 74-9 70-1 66-0 INDIA Estimate 11 53-4 48-7 42-0 42-3 49-1 50-4 51-1 57-4 57-9 53-5 49-1 47-6 Z (Weighted A6erages) * Figures for Gujarat and Maharashta are not available separalely for the year 1957-8 since NSS tabulations for that )ear refer to the old Bombay State including both Maharashtra and Gujarat. The poverty incidence for Bombay State is 56-2 and this figure has been used with the combined weights for U Gujarat and NMaharashtra to calculate the all-India weighted average. t NSS data report a single distribution for the old Punjab State (including lHaryana) up to 1963-4, after which separate distributions are reported for Punjab and Haryana. The poverty incidence for the years after 1963-4 is based on a pooling of the data for the two states, using rural populations of Punjab and 1laryana as recorded in the 1971 Census as weights. It should be noted, however, that parts of the old Punjab States were merged into u Himachal Pradesh and the Union Territory of Delhi. Our procedure ignores this problem but the error is likely to be extremely small. ° 'rABI 1. 3.h) SIN'S PO',LRIY IND)I \ IXOR INDIVIDI Al STAIlS 19Y . 1959) 60 1960 1 1961 2 1Yn,3 4 1964 5 1965 1) 1966 - 196- A Jt 9 3tJ7 -1 1973 4 Andhra Pride%h (0'22 017 1 3 I18 0(18 0(16 0f14 016 0J18 01- 018 0)1 014 A.ssam (7 W 38 0 W 11 0 006 006 0 (I () 13 t1(3 014 00 0-11 Bihar 029 023 0-15 0 19 3)21 0(22 0)25 0(41 0l37, 0)26 3)25 3)24 3 l ')15 312 3)12 (16 I 1 I) 21(1 0)22 023) ()16 (315 01() Karnatakia (3 V 3) 19 01 3)12 0 18 0'1 113)) (-27 0)24 0(25 0717 Kerala 02 " )29 (3)2 3)21 2)21 0)29 13)'3 0)31 () 0(31 0294 t'2() MNidh'a Prid,h t)2' (019 (01- F .14 017 )5 l)17 0126 0t29 0'24 02 1 023) Nlaharashtra 0'21 018 o15 0)1( 0'4 0(1' 0(27 1 2'' 0.'() 0 16 018 032 3)28 0'31 3)23) 0'26 0)26 0'26 (327 ); - 0'33 030 023 PunwTh & H1arvjran 11 II ( 0)07 ()(S 1F 0.10 ( 008 f Il( 00(8 Ol 00 (5 11 (I('06 Ri istahin 1)14 n.a. (1 012 011 0111 3)13) 0)14 (3-14 017 316 -(19 Tamil Nadu 0t34 0-3() 04 0f22 0-'21 0(23 W 0-25 0 26 ((24 (.2S 1)" ' 0(17 Uttar Pradesh 01'21 013 3)14 0(-12 023 ('21 3I, 3)23 1025 0'17 1"13 015 West Benw'al 3)26 025 (014 0)20 0 26 0-26 022 O)'27 0)40 0(33 031 0(31 RURAL POVERTY AND AGRICULTURAL PERFORMANCE IN INDIA 307 TABLE 4 LINEAR TIME TREND FOR POVERTY INCIDENCE BY STATE Percentage of Rural Population Sen 's Povertv Index as in PovertY as Dependent Variable Depentdent Variable Time Constant Time R2 Constant I x 10 2) R2 Andhra Pradesh 52-50 -0671** 064 0204 -038** 0-56 (31P84) (4-25) (18 38) (3'58) Assam 23A47 1 060** 0-38 0 049 0 38** 0*35 (5 29) (2'50) (2'85) (2'30) Bihar 5083 0752 0 17 0214 044 008 (9 15) (1 41) (438) (095) Gujarat 42-18 0-199 902 0 158 002 000 (7 27) (0 37) (5 00) (0 05) Karnataka 42 76 0 800 0(18 0 166 0 35 0 10 (765) (149) (466) (103) Kerala 59.84 0 022 G001 0 276 -0,04 0 00 (1304) (005) (9 10) (0 14) Madhya Pradesh 4495 0559 0 13 0,186 020 004 (939) (122) (5.57) (0*64) Maharashtra 51[85 0 116 001 0 198 003 000 (10 12) (025) (629) (0 11) Orissa 6130 0117 001 0(284 -010 001 (16 64) (0'33) (10 88) (0 39) Punjab & Haryana 2526 0 058 0n00 0 084 -0 07 0 03 ($ 997) (0 22) (6 17) (0 56) Rajasthan 30 91 0 300 0 08 0 1163 0 08 0 02 (8 120) (0 86) (5 62) (0 39) Tamil Nadu 62 59 -0-514 0Q18 0t303 -0 62** 0*46 (17 15) (1V47) (13 66) (2 93) Uttar Pradesh 43-92 0 375 0 04 0 179 0 01 000 (7'68) (0n68) (555) (0 02) West Bengal 5272 1.143* 030 0 190 0Q84** 037 (90'3) (205) (5'20) (241) *Indicates that the coefficient on Time is significant with the sign indicated at the 10 per cent level for a two-tail test. **Indicates significance at the 5 per cent level for a t"o-rail test. Our results for Punjab and Haryana are particularly worth noting in view of the findings of other studies that the incidence of rural poverty has increased in this region, despite the visible success of the Green Revolution in raising output. For examnple, Bardhan [1971] found that the incidence of poverty in Punjab and Haryana increased between 1960-1 and 1967-8 and Rajaraman [1976] reported that the incidence of poverty in Punjab (excluding Haryana) increased between 1960-1 and 1970-1. We find that while both conlclusions stand as long as we compare these particular endpoints, they are misleading in respect of underlying trends, The incidence of rural poverty in Punjab and liaryana appears to have been unusually low in 1960-1 and estimates for this year therefore should not be used as the base for comparison.13 Consideration of the time series as a whole clearly does not suggest a trend increase in the incidence of poverty. The main conclusion to be drawn from these results is that the incidence 308 JoItRNAI. Of: IENL)PNIFINT STUDIES of poverty in rural India does not show any sustained trend ovei the pa:l;t two dCZaLdes. What we observe in most of rural India is a pattern of fluctuation with an increase in the incidence of -overty in one sub-period followed by a decrease in ainother. This pautern obviotisly calls for some ;-plalllatioio in terms o) thie economic flictors, i It.cctin u the rurra! cofnomivl anid this is attempted ini thie niext sectioni. IV. A(GiM ti II RAt PFRl ON( .\AN(CI ANI) i 'RAI. POVRITY Ld,e. tll\e thehSer-\e(ld chan1ges in the incidence of povertyover diffe'rent sub- pei-iods shiouild be exphlaineld inl termns of' some e\xplicit mode! of' the dcterminiamn o rural rpo' erty and the behll, iot i Ir of'thlese determinai,it t;ver the sub-periods. In t c.5 context, it is obviousiv relevanit to onsider the rate ofigricuilkir!il grLk th, the factors dLeteriili itsii istriha lion across farm sizes, its impact onl Lenall ne\ conditions, its effect onl the demand for labour, etc. A complete explor,itior, of the impact of these l.ictors on the economic condition or different socioecononmic groups in the ruiral econoyny is oh.i.,sl\l heyond the scope of this paper. Instead, we shall confine 0ousel\L'.s to exaniininu the relationship between the incidence of poverty and some aggregatki\e inidices of aericllturald performance over the period 1956 7 to 1973 74. Once again the 'inal\-is is presented separately for India as a whole and for the inldividual states. (a) Tl, . I X h i,/i-Inia r, clc i . We begin by postulating thit an imiportant determinant of the extent of rural poverty is thelevel ofauIr-iclIltLunml production relative to the size ofthe rural poml:m1tion.ll. ALricuitiC re is not the on1y sour-ce ofincome in rural areas but it is the doL0ninallmt ou0rce,. and besides, the scale of non-agricultural income generating acti' ity in rural areas almostcertainlv depenids upon the leVel otfag2IricUltLlral produIction1. If there is any vtrickle-down' mechanism at work in the rural econoni\ we sh011 uld eXpeCt increases in agriciiltu ral prodUCt iOn per head to reduce the incidence or absolute pov ert). Does the a'.ailaible evidence supLport this 'iew? The first point to note about the Indian experience of the past two dccadeN is that the grouth of agricultural oLIltpLt has only just kept pace with the vro'.\ th of the rural population. The result is a stagnation in output per head, which can be seen from the fololo.'. inig results reporting linear time trends fitted to thlee (ilffer-.ini ind(lices of output per head of the rural popiilation (terms in p.irenthcses are t ratios). log (FRP) - 554' + 0-0044 T R2 = 0-08 (34-4) (1P26) log (APRIP) 5-563 + 0)0037 T R2 = 0-lu (43-9) (1-38) log (NI)PARP) 5-167 - 0(0021 T R2 = 0(04 (182-46) (0-86) l.I'ere FRP = the index o(f food produiction per lheaid of rural population, APRP the indlex fgzl',riL ulturil prodLictioni per heatd otrural popla.tion, andll NDPARP = NDP in agricultLIe (1960) ' prices) per head of rural TABLE S INCIDENCE OF POVERTY AND AGRICULTURAL GROWTH Estimated Coefficients on Independent Variables (terms in parentheses are t ratios) 0 Dep4-dent Variable Average of Current Alternative Poverty and Previous Year X Measures Constant NDPARP NDPARP Time R2 F 1) Percentage of Poor: Z Estimate I a 106-35 -0-338** 0-43 9-0 o (5-49) (3-00) b 128-791 -0470** 0-49 11-4 (5-40) (3-30) c 136-125 -0-505 -0-141 0-50 5-5 (4-84) (3-20) (0-64) 2) Percentage of Poor: Estimate 11 a 96-302 -0 269** 0-40 6-7 ;0 (5-40) (2-59) O b 126-390 -0+447** 0-62 16-3 (6-70) (4-04) c 127-24 - 0451** 0-021 0-62 7-4 (5-80) (3-64) (0-09) 3) Sen's Poverty Index a 0-540 -0 0020** 0-42 8-7 (4-54) (2-95) b 0-647 -0 0027** 0-42 V-83 (4-20) (2-97) c 0-770 -0-0033** -0-0024 0-52 6-06 (4-61) (3-48) (1-52) "*Indicates that the coefficient is significantly different from zero with the sign shown at the 5% level for a two-tail test. 310 JOURNAL Of L)EVFI.()I LNT STt.'lS population. The extent of stagnation is particularly ev ideCIt in the third equation, which uses NDP in agriculture (in constant prices) as the measure of agricultural income and which is the preferred equation for our purposes since NDP is a measure of value added.'4 This stagnation parallels the observed lack of any trend improvement in the in2idence of rural poverty over the period and canl be argued to be one of its principal causes. The relationship between the incidence of poverty anid agricultural performance over the period under review can be more systematically examined through regression aial%sis. When this is do(ne we find that improved agricultural performance is definitely is.,ociated with reductions in the incidence of poverty. Table 5 presents the results orsoome rearc,;sions along these lines using three alternative dependent variables (Estimnate I and Estirnite 11 of the per cent of ppulaoktion in poverty and the Sen Index) to measure poverty. Initially we hypothesise that the incidence of povert) depends upon the level of NDP in agriculture per head of the rural population (NDPARP). As shown in Table 5 tLhe coerficient on this variable is negative and significant in all cases (Eqs. la, 2a, 3a). An alternative hypothesis is that the inicidence of poverty depends not only on the current year's level of NDPARP but also on the level in the pre% iouis year.15 This is tested by using the average value of NDPARP for the current and previous vear as the indLepenident variable in the regressions. (We use the average value of the current and previous year rather than introduce both as independent variables solely because of' the limited sample size.) As shown in Table 5 (Eqs. lb, 2b, 3h), the explanatory po% er of the equations improves substantially anid the regression coefficients are again necative and highly sionificaint in all cases. It can be argued that while NDPARP is inverselv related to the inicidence of poverty, there are other factors operating in the rural economny leading to an increase in povertv incidence over time. This hypothesis can be crudely tested by including time as an additional explana tory ariaible as in Eqs. Ic, 2c, 3c. We find that the coefficient on this variable is not significant in any equation, s1Igge.tin1 that there is no undeilying time trend in the incidence of poverty after allowing for changes in poverty incidence associated with changes in NDPARP. What can we legitimnattely infer r, omnl these results? There is clear evidence of an inverse relationship between rural poverty and agricultural performance. As shown in Figure 1, fluctuation in poverty inicidence simply mirrors the movement in atgricultural produtiction per head and this inverse relationship is even more firmly identiifible when account is taken of the lags involved. Such empirical rclationship,, are at best a crude basis for dra\ing inferences about comlplex causal nlcchanismiis. but taken at face value they do suggest that there is some trickllinlg down of benefits from increases in agricultura1l produLCtion. (b) The Evidlenc lfor Indh%dhiual .Stelhl. Does the inverse rclationship between poverty and agricultural growth also hold at the level of individual states? Testing for such a relationlship at tlle state level is subject to two important limnita lions. The first is conceptual: the rural area of a single state is a more 'open1' econoimiic system than rural India as a whole and therefore all-India relationslhips may not hold for -60 t U1- PPOM Ba3fl3nXUOZ IOU ae --qI sneA nu.uauujdaj nwod ii9uuoi 01 pa-n a 1e saui Ua-O-- -1- 9L-1C £L-ZU ZL-It IL-9L 01-69 69-89 89-L9 L9-99 99-99 59-t9 b9-C9 C9-Z9 Z9-19 19-09 09-69 65-89 8ES-LS S-96 I...-..-.-...-.---. I I I I I I I I I I I w aje-IS 146t0j u a6S uolinjndad 1 1\ I ai1W1 Z ~ ~ ~ ~ ~ ~ - IPJ «oPeaH I.d \ apeI < 1fl-n3U U1 dO3N A.AJand 09t 0 O95 - SS sit OS0 ga 11 alu!l3- O av.uappui \ 0 A Y leJnS1°- p MH Jad ; -111-5ljV -1daN HIMOU3 lvunlNno189t (INV 3ON301ON1 AlU3A0d oalU 312 JOU'RNAL OF D)EVELOPMENT STUDIES individual states. For example, the effect of adverse agricultural conditions in a single state (especially in a particular year) on rural poverty in that state may be mitigated by temporary migration both to urban areas and to rural labour markets in other states. These 'cushioning' possibilities do not exist at thc all-India level. The only way migration can mitigate the adverse effects of poor agricultural performance for the country as a whole is through migration to urban areas and that too is limited by the fact that a generally poor agricultural performance t) pically also has an adverse effect on non-agricultural employment opportunities. The second limitation is of data. There is no time series of NDP in agriculture for individual states over the period under review. Our measure of agricultural output is therefore limited to an index of agricultural production for all crops constrLucted by A. V. Jose of the Centre for Development Studies (Trivandrum). We have used this index to construct an index of agricultural production per head of the rural population (IAPPH) for each state and used this as the explanatory variable measuring agricultural performance in our regressions.' 6 Since this is a gross output measure, which takes no account of the increased input intensity of agriculture over time, it probably overstates agricultural growth in terms of net value added. These limitations are likely to make the empirical results at the state level somewhat weaker and this is indeed the case. There is some support for the hypothesis that the incidence of poverty is inversely related to agricultural output per head, but the corroboration is not complete. Indeed, there are interesting differences between the all-India and state-level resl]ts, which call for further investigation. An important difference between the all-India experience and the experience for individual states is that whereas in the former case agricultural output per head of the rural population stagnated, this is not true for individual states. Table 6 presents the estimated (exponential) growth rates in IAPPH for individual states. Six states (Kerala, Orissa, Punjab and Haryana, Tamil Nadu, Uttar Pradesh and West Bengal) show significant growth in output per head. Yet, as shown in Table 4, none of these states show a significant trend decline in the incidence of poverty (except for Tamil Nadu on the Sen Index) and West Bengal actually shows a significant trend increase. The absence of a trend decline in the incidence of poverty in states that have experienced growth in agricultural output per head is clearly disturbing and calls for further investigation. Following the approach adopted for the all-India analysis, we have estimated regression equations for individual states testing the hypothesis that the incidenice of poverty depends on the level of agricultuLrIal outputper head of the rural population."' Table 7 presents the results of two regressions for each state, using the percentage of the rural population in poverty and the Sen Index respectively as the independent variables. The explanatory variable measuring agricultural performance is a two-year average of IAPPH where the averaging reflects the existence of lagged effects. We have also included timc as an additional explanatory variable to test for the existence of underlying time trends in poverty incidence generated by factors operating independently of changes in the level of output per head. Our results can be summarised as follows. RURAL POVERTY AND AGRICULTURAL PERFORMANCE IN INDIA 313 TABLE 6 EXPONENTIAL TIME TRENDS IN AGRICULTURAL OUTPUT PER HEAD OF RURAL POPULATION 1956-57 - 1972-73 (terms in parenthcse, are t ratios) Cqj w!11/ (c '1/u'nI oil Time R2 Constant Ix 0 2 Andhra Pradesh 4h608 -0'62 12 (106&88) (1-46) Assam 4'54 -0(13 '02 (194 38) (0'58) Bihar 4h64 1'34 I11 (45 44) (1'35) Gujarat 4 85 0(86 04 (41'93) (0(77) Karnataka 4 63 0'76 '11 (79 42) (134) Kerala 4 54 0-86** 66 (278'17) (5'42) Madhya Pradesh 453 -1*08 '10 (53-62) (1-32) Maharashtra 4X80 3-27** 48 (53'94) (3'72) Orissa 4'64 2'65** '38 (52'09) (3 05) Punjab & Harvana 446 3-16** .73 (86'93) (6'30) R,tj;isthm rl 45S -0'29 '01 (46f66) (0'30) Tamil Nadu 4.59 101 ** '58 (20102) (4'52) Uttar Pradesh 454 138** 59 (90 80) (2'83) West Bengal 4-56 0.79** '27 (133 36) (2'36) **Indicates that the coefficient is significant at the 5 per cent level for a two-tail test. (i) There is clear evidence of a significant negative relationship between agriculturall output per head and the incidence of poverty in seven states: Andhra Pradesh, Bihar, Karnataka, Madhva Pradesh, NMaharashtra, Tain]il Nadu, and Uttar Pradesh. Although only seven states of the fourteen conform to this pattern, it is important to note that thesc states accotunt for 56 per cent of the rural population of India and about three-quarters of the rural poptliltion in poverty. Of the other states, it is worth noting that Orissa and West Bengal have negative co-efficients on the IAPPH variable with t ratios that are fairly high, although not high enough to ensure significance at the ten per cent level for a two-tail test. (ii) Where the state level results difrer suhstitntially from the all-India results is in the estimated coefficients on the time term, At the all-India level we found no significaLnt time trend in poverty incidence operati:ng independently of the effect of agricultural output per head. The results for TABLE 7 RURAL POVERTY AND AGRICULTURAL GROWTH BY STATES (TERMS IN PARENTHESES ARE T RATIOS) Percentage of Rural Population in Poverty as Dependent Variable Sen's Poverty Index as Dependert Variable A verage* of Average of Current and Current and 4 Previous Year Previous Year IAPPH Conslant IAPPH Time RX Constant Cx 10-2) Time R2 Andhra Pradesh 96-38 - 0.479** -0.607** 0-74 0-46 -0-26* -0-37** 0-63 (5-58) (2-62) (3-86) (3-36) (1[87) (3-04) Assam -2-05 0-25 1-358 * 0-38 --04 0-09 0-44 0-29 (0-02) (0-28) (1-96) (0-12) (0-26) (1-61) Bihar 95-82 -0-42** 1.286** 0-78 0-64 -0-40** 091** 0-82 (8-58) (4-47) (3-44) (7-57) (5-54) (3-21) Gujarat 35-71 -0-00 0-956 0-27 0-23 -0-08 0-57 * 0-33 (1-59) (0-02) (1-50) (2-08) (0-99) (1-82) Karnataka 86-90 -0-474* 1-716** 0-56 0-52 -0-37** 096** 0-59 (3-31) (1-87) (3-07) (3-39) (2-51) (2-95) 0 Kerala 83-70 -0-298 0-829 0-22 0-33 -0-09 0-41 0-14 C (1-35) (0-46) (1-37) (0-78) (0-20) (0-98) z Madhya Pradesh 88-72 -0-454** -0-052 0-51 0-49 -031** -0-20 0-45 > (18-83) (2-45) (0-09) (3-66) (2-38) (0-52) r Maharashtra 114-25 -0-548** -0-818 0-48 0-60 -0-35** -0-59 0-50 (4-16) (2-40) (1-16) (3-62) (2-55) (1-39) Orissa 76-41 -0-167 1-063 * 0-31 0-39 -0-12 0-54 0-22 m (6-43) (1-53) (1-86) (4-53) (1-48) (1-30) Punjab & Haryana 38-74 -0-175 0-811 0-29 0-14 -0-06 0-22 0-15 o (4-09) (1-62) (1-76) (2-76) (1-21) (0-91) Rajasthan 21-34 0. 9 0710** 0-53 0-05 0-05 035 0-33 r (2-00) (0-0) (2-76) (0-66) (0-54) (1-78) Z Tamil Nadu 165-68 - 1-050** 0-577 0-62 0-94 -0-64** 0-03 0-69 H (5-52) (3-52) (1-61) (4-62) (3-18) (0-11) H Uttar Pradesh 123-64 -0-864** 1-674** 0-75 0-63 -0 49** 0 80** 0-79 (7-03) (4-71) (3-75) (7-14) (5-36) (3-56) West Bengal 110-94 -0-631 1-966** 0-49 0-67 -050 1-33** 055 co (2-46) (1-37) (2-77) (2-42) (1-79) (3-09) *Indicates coefficient is significant at the 10 per cent level for a two-tail test. "Indicates coefficient is significant at the 5 per cent level for a two-tail test. RURAL. POVERTY AND AGRICULTURAL PERFORMANCE IN INDIA 315 individual states show that the coefficient on time is significant in a number of states, and in all these cases (except Andhra Pradesh) it is significantly positive. If we accept the argument that the time term picks up the net impact of variables excluded from our analysis, these results suggest that in these states-Assam, Bihar, Gujarat, Karnataka, Orissa, Rajasthan, Uttar Pradesh, and West Bengal-there may be factors at work in the rural economy which by themnselves tend to increase the incidence of rural poverty. Identifying these factors is clearly crucial for understanding the causal mechanisms determining rural poverty. Unfortunately, our data provides no basis for developing and testing specific hypotheses along these lines. However, it is interestinig to note that this group includes all the states of the eastern region where conditions of tenancy are most adverse for small farmers. (iii) For most of the states for which the coefficient on the time term is positive, there is also a significantly negatic., coefficient on the IAPPH variafle. This suggests that while there were factors operating in the rural economy which tended to increase the incidence of poverty, agricultural growth leading to higher output per head tended to offset the adverse impact of these factors. Can we conclude from this that increases in agricultural output per head are in themselves always beneficial? The key assumption underlying this argument is that suchl increases can be achieved, within the existing institutional structure, without affecting the other factors which operate to increase the incidence of poverty. This is clearly a strong assumtpcion. Some of the factors whose effect is captured by the time term obviously operate independently of any attempt to increase agricultural output. Increasing population pressure on land leading to a -duction in size of landholdings is an obvious example of such a process which is particularly relevant in rural India. However, there are other processes which may have an adverse effect on poverty and which may be speeded up if we attempt to accelerate the pace of agricultural growth. These include technological change in agriculture which has a net labour displacing effect and which may therefore \ eaiken the econo mnic position of both landless labour and small farmers x ho rely in part on wage employment. The scope for reducing the incidence of poverty through raising. agricultural output per head in the future is therefore crucially dependent upon whether this can be done without technological changes that are excessively labour displacing. The reeressionis reported above cannot address this questioin ;atisfactorily since the observed variation in agricultural output over the period is donminated by weather-induced fluctuazitionis and the impact of such variations on poverty may be different from the impact of a more rapid expanision in agricultural output over tinle. (iv) Finally, the most disqluieting feature of our results is the evidence f'rom Punjab and H-Iarymna which does not sulpport the hypothesis that improved agriclftulrall perIforl-1ManCe will help reduce the incidence of poverty. This region has experienced a dramnatic growth in agricultural output per rural person but there is no evidence of a downtrend in the incidence of poverty. Nor is there a significantly negative coefficienit on the output per head variable in the regressions reported in Table 7. The 316 JOt'RNAL OF DEVELOPMENT STUDIES Punjab-Haryana experience may have been wrongly described by some authors as a case of increasing incidence of poverty despite rapid agricultural growth, when the evidence shows no significant time trend, but it is scant reward for the most successful agricultural performance of any state if the poorest 25 per cent of the rural population experienced stagnant levels of real consumption. How do we account for this outcome? One approach is to attribute it to the particular nature of the agricultural growth experienced in this region. Stagnation in real consumption levels at the lower end of the scale could arise from strongly labour-displacing technological changes accompanying agricultural expansion, exemplifying the malign effects of some types of growth. However, there is also a benign explanation, attributing the lack of any reduction in poverty incidence to the fact that there has been heavy migration into rural Punjab in response to the growth of labour demand. It can be argued that the poorest quarter of the rural population in Punjab and Haryana contains an increasing proportion of in-migrants, and although real consumption levels of the poorest quarter appears stagnant over time, this group consists increasingly of individuals whose consumption is higher than it would have been if they had not migrated. In other words, 'trickle-down' benefits have taken the form of increased employment benefiting migrants from other states, rather thani increased wages benefiting the pre-existing poor. Such supplementary evidence as is available certainly points to a substantial growth in labour deinand and in- migration into the region together with an increase in average real wages, developments which are difficult to reconcile with the malign view of the impact of agricultural growth."8 A further aspect of the Punjab results which must be kept in mind is that poverty in this region is closer to being a problem of low end poverty than in other states. As such, the insignificant coefficients on the IAPPH variable may arise from the fact that even if trickle-down processes do exist in the rural economy, they may not reach all the way down to the very poor. It has long been recognised that for such groups, poverty alleviation will require special programmes of assistance and support and not merely a general improvement in productivity. The evidence provides some support for this view. If we redefine the poverty line for this region to be 20 per cent higher, and use the resulting estimates of the incidence of poverty as dependent variables in our regression equations, we find a statistical improvement over the results reported in Table 7. When the dependent variable is taken as the percenluge poor, we find a positive coefficienit on time and negative coefficient on the IAPPH variable with both coefficients significant at the five per cent level. When the Sen Index is used, the coefficients have the same sign but they remain insignificant at the ten per cent level (although gaininig in significance compared to Table 7). (c) The Evidence on Relative Inequialityl Thus, far, the distributional implications of increases in agricultural output per head have been examined solely in terms of the impact on absolute poverty. It is also useful to examine the evidence on relative inequality in consumption (and changes in inequalit) over time) in order to determine TABLE 8 t- RELATIVE INEQUALITY OF CONSUMPTION: GINI COEFFICIENTS O 196--7 1957-8 1959-60 1960-1 1961-2 1963-4 1964-5 1965-6 1966-7 1967-8 1968-9 1970-1 1973-4 Andhra Pradesh 0-33 0-30 0-3 0-32 0-31 0-31 0-31 0-28 0 28 0-29 0-28 0-30 > Assam 0-26 0-29 0-24 0-23 0-22 0-20 0-21 0-25 0-19 0-20 0-19 0-22 z Bihar 0-32 0-29 0-40 028 0-29 0-29 0-31 0-32 0-31 0-28 0-27 029 t Gujarat 0-33 0-26 0 27 0-30 0-30 0-29 0-30 0-29 0-29 0-28 0-24 > Karnataka 0-38 0-32 0-29 0-37 0 29 0-28 0-31 0-31 0-30 0-32 0-29 0-28 Kerala 0-35 0-34 0 -33 0-33 0-30 0-34 0-30 0-30 0-32 0-42 0-33 0-32 Madhya Pradesh n-a- 0-41 0-34 0-30 0-34 0-35 0-31 0-32 0-29 0-32 0-33 0-32 0-29 Maharashtra * 0-29 0-29 0-28 0-29 0-28 0-29 0-29 0-26 0-29 0-26 0-28 Orissa 0-32 0-31 0-33 0-30 0-28 0-27 0-28 0-25 0-30 0-29 0-29 0-30 C Punjab & > Haryanat 0-32 0-30 0-37 0-35 0-30 0-32 0-33 0-31 0-30 0-28 0-30 0-29 Ra ja-lhari 0-41 0-36 0-32 0-37 0-31 0-32 0-32 0-35 0-34 0-40 0-33 0-29 Tamil Nadu 0-32 03. 0-31 0-31 0-31 0-30 0-29 0-28 0-28 0-29 0-27 0-28 IIllar Pradesh 0-3() 0-3() 0-3() 0-32 0-30 0-30 0-29 (-?8 0-28 0-31 0-29 0-25 0 West Bengal 0-27 0-27 0-26 1) 0-27 0-24 0-27 0-26 0-25 0-23 0-27 0-30 0 ALL Z INDIA 0-33 0-34 0-32 0-33 0-32 0-30 0-30 0-30 0-30 0-29 0-31 0-29 0-28 Q *The NSS tabulations for 195' -8 report the distribution for the old Bombay state which yields a Gini coeffieient of 0-297. Separate distributions for Z Maharashtra and Gujarat are not available. tAfter 1963-4, the conmunipion diktribulion is available only separately for Punjab & Haryana. Our estimates for these years are obtained after pooling distributions for Punjab & Haryana using weights proportional to rural population size in the 1971 census. 318 JOU'RNAL OF DEVELOPMENT STUTDIES whether it supports the benign or malign view of the distributional impact of agricultural growth in India. The view that agricultural growth within the present institutional constraints does not contribute to poverty alleviation, and indeed may even generate absolute impoverishment for the poor in the sense of declining real incomes, implies that we should also see increased relative inequality in the distribution of consumption. This would be the case as long as the consumption function is monotonic, since income increases for the non- poor would be reflected in increased consumption for these groups while stagnant (or reduced) incomes for the poor will lead to stagnant (or lowered) consumption levels.19 TABLE 9 LINEAR TIME TRENDS IN CiINI COEFFICIENTS OF CONSUMPTION DISTRIBtUTION (TERMS IN PARENTHESES ARE T RATIOS) Ev lnultui CoIic wilh % ol rR: Time ( C)onstatll ( X I1)' Andhra Pradesh 0(33 0295** 05 (35-51) (3-3]) Assam 0-27 -(0461** ()51) (17-48) (3 15) Bihar 0(33 -0.291 (17 (15 54) (1-43) Gujarat 0 31 -) 2f63 0 24 (18-39) (1 68) Karnataka 0-35 -0410** 0°34 (18-81) (2-29) Kerala 0 33 0 001 0(00 (15-14) (0 002) Madhya Pradesh 0 37 - 0i45* 0(43 (21-23) (2-74) Maharashtra 0-29 -0 -123 0-19 (32-20) (1-47) Orissa 0-31 -0(192 0-18 (22'65) (1-47) Punjab & Haryana 0 35 -0-345** 0 34 (23-6) (2-45) Rajasthan 0 37 -0 334 0 20 (16-74) (157) Tamil Nadu 0 33 -0-384** 0 82 (56 84) (6-85) Uttar Pradesh 0(32 - 00254** 0(44 (32-18) (2-78) West Bengal 0(26 -(1 055 0 02 (i0200) (0-45) All India 0-34 -0,324** 0-82 (73-23) (7-06) Indicates that the coefficient is significant at the 10 per cent level for a two-tail test. "Indicates significance at the 5 per cent level for a two-tail test. RURAL P0VERIY ANt) AGRi('t1TI'RAL PFRI 0101 \NCF. IN IND)IA 319 The NSS data provide the basis for 11C documentinL trends in the inequality of consuLmlption expenditUlre ill nomioil.i terms. Ideally, we should examine trends in the distribution of real conimnmption since clhallges in the distribution of nominal consuLmnlptioni may- reflect no imloIrc than d ilferenliai mioveents in the price in(ices. However. in the absence ot ffractile specific price indices \ e can onl\ conmpare ilieCLquallit in nomlinal terms. Suchl comparisons remain of initerest since the a\ailiable evidence s.uggcsts that while they may exaggerate the extenit of the change in ineLuLa.lit., they ne%er-theless point in the right dirLction.9"' Table 8 presents the (iini coefficients of the distr ibutioni of'rural consumption expenditure for India as a whole and for the individl.;zI sttates. We have tested for the existence of a linear time trend in each case and the r esults are presenited in Table 9. Far from finding an incretase in relativc inCeLuLlallity, we find that the evidence points in the opposite direction. There is a siunificant h1'c-c'ww in relative inequaility for India as a whole and for se\ e n of the fourteen st;at es. Of the other states, all except Kerala have a negative co-eofficient on the time term, although none of these coerficienits is tiatisticallv significant. It is particularly worth noting that the Gini coefficient for Punjab and Har \ ilanal shows a .t.atisticall sivnificanL decline in inellallity over the period as a whole. The albsence ofa nmarked inncre.ase in relative inei.qualit\ ais meaSu red by the NSS is yet anotlher rca-on for LILIestilniffii thle view thiat agrictil tiiral growth has been aiccompLanied by absolute inipo\erishmnint bor the rural poor. As sho%x n in Table 6. six states (Kerala. Orissa, PuLijalb and Hrana. Tamil Nadu, Uttar Pradeshl and \N'est Bengal) slhow ai siginificant trenld increase in output per head of the rratL populat Iion. We n0ote that none of these states e\perienced the increase in relative inequality whiich should ha,ve occurred if -ro\With was aiccomnpanied by absolute imipoverishmllent. Indeed three of them (Punjab and Har) ana. Tamil Nadu and Uttar PrLdeslh) show a siiniificilat Ldccli ie in e( il Ull lpioll iitia ;lliti\ V. CONCUTSIONS The twin objectives of this paper weretoLdocLInent trends in rural po\ el-ut in India and to examine the relaitions,hip hbemeen rural poverty and agricu.lturail performance. Otur principal emipir icil finding.-s and the caveats acconmpanying them can be aslianma ried a1 l0llox\s: li) The evidence :evie\ved pro\ ides a rliirlv firm basis for doCLum1enlinlgL trends in rural poverty in India. We find that the Indian e\perience over the past t\ o decades cannot be chi racter-ised assliot ing ai trenid incr ease in the incidenice of poverty in India a1s a wh(ole. The sameconchlusion holds tor all the indi\ idL,ual Stltes except Assam and Bengali. '\ hiTh shtm it aignificalnt trend increase in poverty. In general. the titime Xseries lho\xs a 11a1ttel-n Of fluctuation. witlh the incidence of' or p ert\ falIling in periods of' good a gnrcu Itu Li l perkl'rma nce anvvid rising in pe.riodsoOf p oor perf'ormnance^. C,ivenl the inmport a ne of eCatllel-i idLicCL ilari at0io1ns in n11dian11 agrniculItuLr, there can be little doubt aihbout the inlportlance of' suich fimtmnaitiorns and it is crucial to keep these in mind in tissessiing tndcrling tr-ends. (ii) The ceidence ona the relationship bct cen ruiral po\ert% and agricultUlrill perforina nce is more ditheult to e% Za I;tC a for two reasollns. In 320 JOUIRNAL OF DVEVLOPMENT S U)DIES the first place, the evidence itself is sonime%hlt mixed. Much dellendis upon the level ofag,gregatiori at which the ann.Ex sis is conductetL with the all-India results presentinp a somexx hat dirf'erenit picture from that obtained at the level of individual states. Furthermnore. the ex idence necessaril) is difficUlt to interpret since we are i-iplicitlh searLchilg ft r causal1 relaltion'ships in what are at best observed correlations. The all-India evidence is entirelv consistent with the hypothesis that the incidence of rural poxerty is inversely related to agricultural perl'roinilce mie;nicasured in ternms of agricultural NDP per rural person. Taken at face xalue. this correlation suggests that faster agricultural growth, by raising agricultural NDP per rural person, might have led to a reduliced incidenecc of poverty. The absence of any adverse underlying time trends fuirther supports this view. (iii) The state level analysis presentsi a :,onmcx ha t different picture. On the one hand w-e find a signiificanllt inverse relationship in at least seven states accounting for three-quarters of the rural poor. On the other halndl the state level analysis also shows that there may be processes at work in the rural economiy which tend to increase poverty over time. These r.esulits are open to the interpretation that agriCultural growth offsets the adverse impact to other factors so that if onlv i1griCu1ltUre can grow fast en1o0hll. it is likely to reduce the incidence of ruiral poxertx. However, this interpretation rests crucially on the assumptioni that inCreaNsedl agriCtURaNIl 0:tpLut caln be obtained without exacerhating those uinidenitified( lfactors xMiMI ield to increase rural poxcertl!, and xx hich are rellected( in the time termii in our regressions. It is in this context that the e\ idence frm(n Punjbh aind :Ilar% ana is disquieting, although. again. there are a numilhber of reasons Why this evidence may be mnisleading. It is a familiar feature of empirical research that the results of related investigations do not alxvays point in the same direction. Clearly a great deal of judgment must enter into weighing the d if frent pieces of evidence before presenting a composite pietuire. The role of judLigmenet is further enlarged when we are trinig to inferthe natuLeof dc0mplexCausa1l prOCesses through ohserxed associations between aguregative xariables. In our v iexx the all-India results should he given -ubstantial Nweight. if only because of the greater reliability of the data series used, and these results do point to a beneficial impact on aibsolute poverty fron increases in agricLiltural output per head. In other words, there is exidence of some 'trickle down' associated with agrieLlltLiral gromxth. At the same time, the state level evidence on the existence of underlking forces xithin the rural econonly which tend to increase the inci(dence of poverty is extremely imnportant. What are these forces, do the) vary across state,. and to what extent can they be mitigaLted?' These are important. indeed kev. questions for policy. Aggregative analysis of the type attempted in this paper cain help to raise these questions but it cannot hope to aniwer them. NOTES 1. For a sampling of the Indian debate on this question, see Bardhan [191]. B\re. [19-], Dandekarand Raih[/9711. Lal[/976l. Raj[/9'6] and R.md i,.ivn /I'5. Fora mOre gencral assertion ofthe operation of this process in a numberofdeC\clfpinlgL(ollunlrics. see (riffin and Khan [1976g. Rt 'RAI 1)()VF,RT'IY A\D1 A( RkICULYftRAXI P1 I II()R\I \,N('I IN INDIA 321 2. Minhias'studx was based onai conibination of NSS daitai and iiiittoinai accounts daita. For a discussion of' the imrplications of uisingt national accotunts datat we AhluwaNitlia f ItJ'WJ. 3. I'ratLakL iLLsil e sn/Y'J h ne a ex itna f t ( I tiGo vwhere x is thle percentage of' thle population beltolx the pox rt> linie t:,ct. is thle mieani consumrption of thie poor. and ( i i thle ( ilink-i celiLII itOf :1 te thu lii' 1111.i nit 'oile't the Poor. Trhe inidex ranges from (1 toi I 4. See Planning ( ominssion t1%(21. 5. Noit onhK is it extremlxI dfilfiliult toi measIl e the actunal cailo rs intatke from thie countimer expendituire lexel as r-ecoirded In thle NSS. it is Alsk arguable thait the tis of 2 iIca Itoi Is" -As mitinimum reL imiiencim Iis iiiappropi i.t OwI se. IPril examnple. Bliss and Sterni f 11Y'.i I I oir a discussionl of the difficulties in translating thle notion of a nutritiotial mnimiitimii inito at powerh line in terms of COn1StiMpti0on 2\11Cd IR i C. see siikhI aiinc [/9'7I 6. For ani attemipt to) conrstrujct tractile s,pecihe pr ice indice, FOr the rural potpulatioii in itie sixties see V. d sa hari aii N 4J Vaidx anlat ian binds that thle price indeix I or thle poort rose shlihtlx fa1ster thanii For the ili(ti Poor 7. Seefbootnote to ITable I trdtisoidaa iOlcsued and sexeratl asstnnptiotis mide to obtain estimiates for inii-inu %eatrs N. TIhe dlistinction is important itornlx because it i, Incorporated in the witiuation procedures used bN the NSS. Consumption out of horne-arotxx n lotck is Naltied at litrml gate prices while consuitnptioiin ouit of wNages whklether paid in mritne> ot in kind) tIs treated ii5 purchased see Minhas [IWr1. 9. This problemi can onix~ b' rc:,olwtd ilt the poxert% linie ini curtrent prices i,~ determiiiied Fotr each year sepaLratL11 ats thle sola'i-Itn to the Ii.iimn ittbk'iliii o defiiiing the miiiinimumi cost budget needed to) achiexe om 1cii ' iii peb iLmimiuniiii WicttuiieCiiieIIIs in nutiltionail andl oilier terms. subect toi an'N desired set oIb taste .id piit'iiii C01citisraints. NvEHL- t es O sa'~. no uc exrerise yielding po%eri lines t o%cr timie hais been ettducted for IndLiai or' to ait> 01 Ili1 ii Cs It). This procedure has an imptort.tnt adantagew in aifii is ti%Old II liHieetd Forinter'po1lation InI n.i i n~ liepci cii ie kI ''A ie i xcit' in xlien thle linle taI11ls etween tile limi1ts ofonle of, the discrete expenditure classes tisedi b% the \ 'S to, report the distribution of 'the population For a miore detailied discussiton see Ahlulwalia I,. I11. La Is fi nd ings a re b itseLd,IIL it dta It ,I I Is\ c t.! i ..s t ti nl. anibt I i tIIt lie nIt I' Ire. on I thfe InIIcId idece tIf pv \xert w i th in Ie I sctit-called wkea kecr seLti t IIt1s of the pit ptuila t Ion. ThIIIs comInpatri s on I s s omew \khait qlues t i onaiblIe fo r t he pItIpr Is lo 1 d etei tI I I I . III II . tIe at ItheI icII I idenice ofI -1 rurliI Po ve rtII sice It is pos si ble t hat theI .I ic IdetI I ctI I~' cI InI thI ce xeike,;:i sec t Ion, max III declInIIIe I wIlIe thfe I incId enct:e t t poe rtI> tin thIIe ruiralI populaitijonI as a xx 1 lttIle ma x II1 tucicas wI1 IhisC cai la 1 ~Ilt If pc I III e pillI po IIII tio oti t he ruiralI pttp1I Il;t It I InoIII1 th11e 'weamk er sectit I n, I t I. Ix ises II1 .Ii ii, en ox er C thIICfel peI tid 1 2. \IinlIaIzIs' re sults t WIt 11iw t %!Ii tinII ts t~oi Is,tInItIotI I II In the lei sI II I1,Iace, IC,,IIII letibies, theI N SS estimate of the distribution of rural co.nsunipiion x iili an estiniate tof thi: lexel of' rural consumption per person xhInchi is lased ttn a national actcounits consumption series. Since his national accounts data showk hit-1ci ocfstt.t'i*iigi' in laiter wcars compared to the NSS. this tenids to reduce the inicidenice oft po%ert% ttvei time. \uch miore cruiciail fosr the outcome. howexer. is Minhas' choice ob 'the (Di) '.:1. r as thie relex ant price inidex for conistruceitng equixalent ptt rit% lines. Since the ( P1 X shows, a mutch Ilaster inflation rate thlan the (iD)P letlai1or. use of the mii ir> ieldslioxxei estimatesol ptosri> In latter xears. F,or adiscussion ofitlie results obtatiinedI tisinlg 'fitlnha' appittachi xxit[ titur d, asee Afhilux aba I IYNj. 13. Onle pi iss1bt R'iiSOlIi br thisl,I isthe smal su/e itt the stilt ptk I hec sainple si/ fOr Pltintabt an1d Ilartaiia in I 9611 1 wasj oil>I 1412 househ0olds k,iuq AIrcid w iihi 24; in I19)S) 611 and 224 ini 1961 2. 14. It is \xell knowi& thiat thit. inde\ ott 'itodgrains production and ilcttles wdt\o pi oductt.an ot all, crops c aL~itcthe tiw% II HI tiiint x itlu aIded beLause tihe> are gi ss otpujtli measurestel xuc hic o niot taik-e acitLCunt t4 t1he t11iiiie'Id iiiptilt iiitel)iisitxoi 'ti.iili iii- t roll) the late 'sixties onxx ards. 15. A latggedl effect otl this sio t is. 0ii in cl pltaaibkc lei'. the niatuicl tit' access to thle capitail tiairket. In a bad agricultural willt small I tii nier and I i ttI.laboiirers niax a ic nipt IOi cushioni i lie ads\ci e impact on their etousuiptiiti 'ii h , i i.t i tlit I' i iti I These loans atre t%I1ie.tlI\ a%ailable at ser% high interest rates antI ma'.1 base hie repaid In III,: niesl year so thatt bad auilricULlfiurat %eardtepiesscs ilie IsL.\ fii t ts'!.i iric in tI'. t;l,.t war A siiii.ir 322 JOURNAL OF DEVELOPMENT STUDIES argument applies when assets are run down in a bad year and income of the next year is partly diverted to building them up again. It should be noted that the crucial assumption in this explanation is not the existence of a high interest rate as such but rather a capital market imperfection in which a high rate must be paid for borrowing but a corresponding high rate cannot be obtained for any surpluses generated in good years. Clearly if the capital market were perfect in the sense that borrowing rates and lending rates were equalised we would be much closer to the sNorld of permanent income theorists in which consuirption would be substantially insulated from fluctuations in current income. 16. The index of output per head of the rural population is obtained by dividing the output index for each state by a rural population index which is constructed on the assumption that the rural population in each state grew over the period 1956-7 to 1972-3 at the same compound annual rate as observed between the decennial censuses of 1961 and 1971. 17. This approach to examining the relationship between poverty and agricultural performance is preferable to relating estimated trends in the incidence of poverty to trertds in agricultural output per head since it makes full use of the available information. 18. See, for example. Lal [19761 and Johl [1975]. 19. Note that if'all incomes increase, but incomes of the rich increase more than incomes of the poor, then the inequality of consumption may either.rise or fall depending upon the relative income increases and the shape of the consumption function. Thus a decrease in inequality in consumption is consistent with an increase in inequality in income, but our focus here is not on whether income inequality increased, which it probablv did, but whether agricultural gro%%th produced absolute impoverishment. 20. See, for example, Vaidyanathan [!2-4]. REFE RFNF'FCS Ahluwalia, M. S.. 1978, 'Rural Poverty in India: 1956-57 to 1973--74', World Bank Staff Working Paper, Vw'shingon, D.C. Bardhan. P. K.. 1971, 'The Incidence of Rural Poverty in the Sixties', Economic and Political W'eeklv, February 1973, reprinted in Bardhan and Srinivasan (eds.). Bardhan, P. K. and T. N. Srinvasan (eds.). 1974. Poverty and Income Distribution in Iho/liti, Statistical Publishing Society. Calcutta. Bliss, C. and N. Stern, 1976. 'Economic Aspects of the Connection Between Prod uCli its and Consumption'. University of Essex, Department of Economics, )iscussior Paper No. 67. October. Byres. T. J., 1972, 'The Dialectic of India's Green Revolution', South Asian Reviewv, Vol. 5, No. 2. January. Dflndekar, V. M. and N. Rath. 1971, 'Pover.ty in India: Dimensions and Trends', Economic ant! Political W'eek!r, 2 January. Griffi'i, K. and A. Khan. 1976. Po&erty antd Landlessness in Rutral Asia, ILO (mimeo). Johl, S. S, 1975. 'Gains of the Green Revolution:How They Have Been Shared in Punjab', Joturntal of Derelopment Studies, Vol. 11. No. 3, April. Jose, A. V., 1974, 'Trends in Real Wage Rates of Agricultural Labourers'. Economic and Political [IVeekly, Vol. IX. No. 13, September. KaklAani, N. C. and N. Podder. 1976, 'Efficient Estimation of the Lorenz Curve and Associated Inequality Measures from Grouped Observations'. Econonietrica, Vol. 44, No. 1. Lal, D., 1976, 'Agricultural Growth, Real Wages. and the Rural Poor in India', Economic and Political J1'c'cklc . Rc licwt of Agriculture, June. Minhas, B., 1970, 'Rural Poverts, Land Distribution aDd Development Strategy: Facts', reprinited in Bardhan and Srinivasan [1I7J41. Minhas, B-. 1971, 'Rur.jl Poverty and the Minimum Level of Living: A Reply', Indian Economic Recic . Vol. VI (New Series), No. 1, April. Planning Coniniissioni. 1962, 'Per,pecti'.e of Development: 1961-1976. Implications of Planning for a Minimum Level of Living', reprinted in Bardhan and Srinivasan [1974]. Raj, K. N., 1976, 'Trends in Rural Unemployment in India', Fconomic and Political Weekly, Vol. XI, Nos. 31-33, August. See also comments thereon by Pravin Visaria and S. K. Sanyal in Vol. XII, No. 5, 29 January, 1977. RURAL POVERTY AND AGRICULTURAL PERFORMANCE IN INDIA 323 Rajaraman, I., 1975, 'Poverty, Inequality and Economic Growth: Rural Punjab, 1960 61-1970/71', Journal of Development Studies, July. Sen, A. K., 1973, 'Poverty, Inequality, Unemployment: Some Conceptual Issues in Measurement', Economic andel Political WeeklV, August. Srinivasan, T. N., P. N. Radhakrishnan, and A. Vaidyanathan, 1974, 'Data on Distribution of Consumption Expenditure in India: An Evaluation', in Bardlian and Srinivasan [1974]. Sukhatme, P. V., 1977, 'Malnutrition and Poverty', Ninth Lal Bahadur Shastri Memorial Lecture, June. Vaidyanathan, A., 1974, 'Some Aspects of Inequalities in Living Standards in Rural India', reprinted in Bardhan and Srinivasan [1974]. Country Economic Reports Commonwealth Caribbean: The Integration Experience by Sidney E. Chernick and others, published by The Johns Hopkins University Press, 1978 Ivory Coast: The Challenge of Success by Bastiaan den Tuinder and others, published by The Johns Hopkins University Press, 1978 Kenya: Into the Second Decade by John Burrows and others, published by The Johns Hopkins University Press, 1975 Korea: Problems and Issues in a Rapidly Growving Economy by Parvez Hasan, published by The Johns Hopkins University Press, 1976 Lesotho: A Development Challenge by Willem Maane, distributed by The Johns Hopkins University Pbress, 1975 Nigeria: Options for Long-Term Development by Wouter Tims and others, published by The Johns Hopkins University Press, 1974 Papua New Guinea: Its Economic Situation and Prospects for Development by George Baldwin and others, distributed by The Johns Hopkins University Press, 1978 The Philippines: Priorities and Prospects for Development by Russell Cheetham, Edward Hawkins, and others, distributed by The Johns Hopkins University Press, 1976 Turkey: Prospects and Problems of an Expanding Economy by Edmond Asfour and others, distributed by The Johns Hopkins University Press, 1975 Yugoslavia: Developmnent with Decentralization by Vinod Dubey and others, published by The Johns Hopkins University Press, 1975 World Bank Staff Occasional Papers A Model for Income Distribution, Employmnent, and Growth: A Case Study of Indonesia by Syamaprasad Gupta, published by The Johns Hopkins University Press, 1977 Coffee, Tea, and Cocoa: Market Prospects and Development Lending by Shamsher Singh and others, published by The Johns Hopkins Uni- versity Press, 1977 Malnutrition and Poverty: Magnitude and Policy Options by Shlomo Reutlinger and Marcelo Selowsky, published by The Johns Hopkins University Press, 1976 Economic Evaluation of Vocatioinal Training Programls by Manuel Zymelman, published by The Johns Hopkins University Press, 1976 A Development Model for the Agricultural Sector of Portugal by Alvin C. Egbert and Hyung M. Kim, published by The Johns Hopkins Uni- versity Press, 1975 Other Publications Agrarian Reform7l as Unfinished Business: The Selected Papers of Wolf Ladejinisky edited by Louis J. Waiinsky, published by Oxford Univer- sity Press, 1977 Twenty-five Years of Ecoionoic Developmenit: 1950-1975 by David Morawetz, distributed by The Johns Hopkins University Press; 1977 World Tables 1976, published by The Johns Hopkins University Press, 1976 The Tropics and Econiomiiic Development: A Provocative Inquiry into the Poverty of Nationis by Andrew Kamarck, published by The Johns Hopkins University Press, 1976 Size Distribution of Income: A Comiipilation of Data by Shail Jain, dis- tributed by The Johns Hopkins University Press, 1975 Redistribuitioni with Growth by Hollis Chenery, Montek S. Ahluwalia, C. L. G. Bell, John H. Duloy, arid Richard Jolly, published by Oxford University Press, 1974 THE WORLD BANK Headquarters 1818 H Street, N.W. Washington, D.C. 20433 U.S.A. Euiropeaz Office 66, avenue d'Iena 75116 Paris, France Tokyo Office Kokusai Building 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan