Urban housing in India

FWLE COPY Urban Housing in India Devendra B. Gupta SWP730 WORLD BANK STAFF WORKING PAPERS Number 730 FILE COPY   WORLD BANK STAFF WORKING PAPERS Number 730 Urban Housing in India Devendra B. Gupta ;=r,nzfuTrojlTj j 4, ; A.PP1 193 The World Bank Washington, D.C., U.S.A.  Cotyreh,' (D 'fAja The Ante:nationai Bank for Reconstruction and )IC&ve'opraene H/ TE WORLD BANK 1818 H Sweet, N.W. Washingt-on, D.C. 20433, U.S.A. All rights .-?scrved Manufaci'.-ed in the United States of America First printing April 1985 This is , l vorking docum-ent published informally by the World Bank. To present the results of researcli w:th the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank acccpts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The Worid Bank does not accept responsibility for the views expressed herein, which are those of the aui-hors and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. lThe designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates conceming the legai status of any country, territory, city, area, or of its authorities, or concerning the delimitation of its boundaries, or national affiliation. The full range of World Bank publications, both free and for sale, is described in the Catalog of Publications; the continuing research program is outlined in Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Sales Unit, Department T, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from the European Office of the Bank, 66 avenue d'lena, 75116 Paris, France. Devendra B. Gupta, a consultant to the World Bank, is professor of economics at the Economic Growth Center, Delhi, India. Library of Congress Cataloging in Publication Data Gupta, Devendra B. Urban housing in India. (World Bank staff working papers ; no. 730) Includes bibliographical references. 1. Housing--India. 2. Housing policy--India. 3. Housing--India--Finance. 4. Slums--India. 5. Squatter settlements--India. I. Title. II. Series. HD7361.A3G78 1985 363.5'0954 85-6289 ISBN 0.-8213-0554-9  ABSTRACT Evidence indicates that the situation of urban housing in India has been poor over the past thirty years and may have even deteriorated in many important aspects over this period. There is a high degree of crowding which has been getting worse over the years; by now there are an average of over three persons per room in Indian towns and cities. Despite increasing urban population the share of housing investment has been declining as a proportion of total investment in the country and is much less than in many other countries. One consequence of the lack of housing supply is the increasing proportion of people living in slums and squatter settlements, now estimated to be over a third of total urban population. This paper documents and brings together a range of scattered information not hitherto accessible to shed light on this neglected area of economic policy in India. It evaluates the existing housing stock in the country and the role of housing in the national economy. It analyses the components of increasing housing demand and identifies the series of constraints that impair the expansion of supply. It concludes that there is a clear need for a major overhaul of many of the government policies and regulations such that housing supply may be more responsive to demand from all income levels. Prime candidates for revision include urban land policy, rent control, housing standards set by urban land zoning and building by-laws, the provision of infrastructure and security of tenure to the poor, and a streamlining of the approval process for home building. Several policy interventions are proposed in the field of housing finance in order to expand significantly the currently low level of available  - ii - housing finance. It is clear from the analysis that economic demand exists for such an expansion of housing finance which will therefore be easily utilized. however, many innovations will be necessary in its delivery system so that much of existing housing stock can be upgraded in addition to new starts, and that housing finance can also reach the poor.  CONTENTS SUMMARY AND CONCLUSIONS vi- I. HOUSING IN THE INDIAN ECONOMY 1 II. HOUSING DEMAND AND SUPPLY 13 II.1 Demand for flousing 15 II.2 Structure of the Housing Stock 27 11.3 Quality of Housing 54 11.4 housing Amenities 74 III. SLUMS AND SQUATTER SETTLEMENTS 90 III.1 Conditions in Slum Areas 90 III.2 Government Policies Towards Slums and Squatter Settlements 99 III.3 Bustees of Calcutta 102 III.4 Unauthorized Colonies in Delhi 106 III.5 Madras Slums and Low Income Areas 111 III.6 Summary 117 IV. SUPPLY CONSTRAINTS IN URBAN HOUSING 119 IV.1 Land Ownership Constraints 119 IV.2 Rent Control and Urban Housing 150 IV.3 Construction Constraints 165 IV.4 Housing Finance 176 IV.5 Fiscal Incentives and Housing Investment 200 APPENDICES I. Estimation of GNP and Capital Formation 215 II. Some Census Definitions 219 III. Slum Areas (Improvement and Clearance Act) 221   SUMMARY AND CONCLUSIONS A. Housing in the National Economy The contribution of the housing sector to national income in India has been between three and four percent throughout the decades of the sixties and seventies respectively. This share is lower than many other countries and indicates the relatively tardy progress made by India in the housing sector. The pace of urbanization has also been slow but there has been a steady increase in urban population of between 3 and 4 percent a year over the same two decades. Since urban housing is, in general, more expensive than housing in rural areas, the overall share of housing might have been expected to increase over this period. What is more disturbing is that the rate of growth of gross capital stock in housing has been only about 1.5 to 1.6 percent per annum. The total investment in housing had declined from about 4 percent of GDP in 1950-51 to less than 2.5 percent by the mid-seventies. These figures are for both urban and rural housing but the trends for urban housing alone are no more encouraging. There are some problems underlying the data on which these estimates are based but the broad trends should be regarded as reliable, given other corroborating evidence on urban housing in India. The public sector has not been involved heavily in the provision of housing with the bulk of housing coming from the private sector. Organized private sector activity is also small, the large part of housing being provided for themselves by private households. Investment in housing as a percentage of total investment in the economy has fallen from about 34 percent in the early fifties to about 8 percent  in the late seventies. The share of housing in planned public investment has fallen from 16 percent in the First five-Year Plan to about 1.6 percent in the Sixth Plan. Some of these declining trends in the proportion of housing investment to total investment should be expected in an economy undergoing industrial diversification as India has been. But the evidence of housing problems in urban areas in the country is strong enough to warrant a close new look at the components of housing, at housing policy, and at the underlying data. The overall demand for urban housing can be expected to arise from increases in urban population and from increases in income. There has clearly been a steady increase in urban population. Though separate accounts on urban income are not available, it is clear that urban income has also been increasing and probably a little faster than national income. There are no good estimates of the income elasticity of housing demand in India but judging from the evidence from other countries it should be positive and a little less than unity. Interestingly, average urban household size appears to have increased and so the growth in the number of urban households has been lower than that of urban population. This, however, may be because the supply of housing has not kept up with demand and there may be a relationship between household formation and availability of dwellings. It is quite likely, for example, for single member households to merge with relatives and delay their own household formation. There is evidence that the average number of persons per dwelling has increased, as has average number of persons per room resulting in a decline in space per person in urban areas. It is therefore clear that crowding has increased in urban India. More than 70  percent of households have less than 300 sq. ft. of floor space. One would expect this to happen if the shadow price of housing has been increasing which itself would result from constraints in housing supply. The one positive trend in housing in urban India over the past quarter century has been a clear improvement in quality. The proportion of houses which are "kutcha" (made of temporary materials like mud walls and straw roofs) has shown a continuing decrease and the proportion of brick or other permanent type structures has increased. This finding would be consistent with a rise in unit housing price and higher crowding. There does not appear to be any difference in the quality of housing according to mode of tenure. Overall, about half of households are renters and the other half owners, but the proportion of renters is considerably higher in the largest cities. It is also clear that the provision of public services has improved. There have been improvements all round in the access of households to drinking water, toilet facilities, cooking facilities, bathing facilities and lighting. The average level of these services continues to be very low and there is therefore considerable room for improvement. Much of the access to such services is still shared among many households. Contrary to popular impression, the quality and quantity of services is much higher in larger cities as compared to the smaller towns. The improvement in these services has essentially been due to widespread government activity in these areas.  B. Slums and Squatter Settlements The constraints in the supply of housing have resulted in large number of slums and squatter settlements that have continued to grow in Indian towns and cities. While comparative time series information is hard to get, it is quite probable that the proportions of people living in slums have increased over the years, and particularly so in the larger cities. There can clearly be no hard and fast definition of slums but the existence of slums is usually associated with dilapidated structures or with the lack of basic services. In India they are covered by a legal definition which declares areas as slum areas under various slum (Clearance and Improvement) Acts. The existence of slum is, at the same time, a solution to inadequately met housing demand as well as an indicator of inappropriate housing policy. Government policy has evolved slowly over the years from one of slum clearance in the early years to slum improvement now. The early idea was merely to remove slums and to rehabilitate the uprooted dwellers in newly constructed government contracted structures. It was realized that this was clearly impracticable given the costs involved. A more realistic policy has been followed over the last decade. This has been to improve the environment of slums by making provision for basic services--water supply, drainage, sewerage, pavements and lighting. This has been successful to a limited degree. The main drawback is that this policy does not touch the issue of security of tenure nor of the improvement of existing dwellings in slums. In this respect it is striking that more than half of the area covered by slums in urban areas is government owned land.  Rationalizing security of tenure is therefore very much within the power of the government. Unlike other less successful measures, improving the access of slum dwellers to land they are already living on, would do much to improve the distribution of land ownership. While it is not procedurally.easy to implement it is clear that much improvement could take place if the tenure status of many slum dwellers would be regularized. It would be expected, for example, that they would make greater investments in improving their own structures if security of tenure was assured. Considerable progress has, however, been made in the environmental improvement of slums in many cities, the programs in Calcutta and Madras being striking examples where World Bank projects have had a prominent role. The wholesale relocation of the poor to the outskirts of Delhi provides a contrasting example where slum3 were cleared and people moved out. Once they were moved to those far-flung locations the principle of sites and services was frequently used and the resulting development of these areas is generally successful. Given these varying experiences around the country, there should be a new look at policy towards slums such that the environmental improvement program is strongly supplemented with measures providing security of tenure and assistance resulting in higher housing investments. C. Government Interventions in the Supply of Urban Land The major supply constraint in housing has been the supply of developed urban land. The policy response to this problem has mainly consisted of two kinds of measures, both of which have probably had counterproductive effects. The first, best exemplified by urban land  policy in the development of New Delhi, has consisted of schemes for large-scale acquisition, development and disposal of land. The idea is for the public sector to acquire in advance large tracts of land in and around developing cities; to develop the tracts and then to sell parcels at essentially cost-plus prices to selected and deserving sections of the population and at auction prices to others. The objective is to control unwarranted increases in land prices. The private sector is then excluded from land development. The effect of such programs is often counterproductive. There are two problems. The first is that it is difficult for a public authority to develop land fast enough to keep up with demand because of both financial as well as physical implementation constraints. A queue for developed land therefore develops. Second, it is not easy to devise equitable procedures for the allocation of land such that the poor actually have improved access. The effect of the first problem is an increase in the black market or effective price of land and the effect of the latter is a worsening in the access to land of the poor. In New Delhi, for example, the proportion of people living in slums and squatter settlements is the highest among the largest Indian cities, a fact contrary to popular impression. The second major land policy enunciated in India has been the imposition of the urban land ceiling in 1975. The idea was to acquire all excess land held by individuals over stated limits to permissible holdings. These limits varied by city size, the limit being 500 sq. meters in the largest cities. It was expected that the public sector would be able to acquire large tracts of land by this procedure, develop them and allot plots to the poor at affordable prices. This procedure would help in promoting equity in land holdings and help in moderating the  rise in urban land prices. The actual effects have been quite the opposite. First, there simply was not enough surplus land for a major redistribution even if all the implementation could be done efficiently. Second, there have been immense physical problems in terms of lot size, lot shape, etc., for plots acquired to be used effectively. Third, the imposition of the acts froze the supply of urban land in the private market and froze all land transactions. All such transactions were now subject to approval by a government authority and this has led to obvious bureaucratic red tape and corruption. Fourth, not much of the officially declared surplus land has actually been acquired and so there has not been much to develop, let alone allot to the poor. The result has been a major spurt in urban land prices since the imposition of urban land ceilings. A solution to this problem is yet to emerge although attempts are being made to tighten the provisions in the Acts. D. Rent Controls and Other Regulations In addition to the constraints in the supply of developed urban land, there are other constraints in the supply of housing. The paper has offered a theoretical analysis of the effects of rent control. As might be expected, this analysis shows that rent control unequivocally reduces supply unless the government counteracts this effect by subsidizing supply. In addition to transferring wealth from landlords to tenants, imposition of rent control also involves a dead weight social welfare loss. The analysis also shows how rent control induces a landlord to reduce maintenance and repair expenditures. The overall result is that limiting the rent below the market level results in the decline of new as well as replacement investment in housing, unless the government intervenes to overcome the distortions introduced.  Rent restrictions were first introduced in India (in Bombay) in 1918. But rent control acts became pervasive in almost all states during and after World War II. A key feature of most acts, in addition to limiting rents, are procedures making it difficult, if not impossible, to evict a sitting tenant. Except for new construction where a holiday is provided from rent control provisions for usually 5 years, rents are virtually frozen for postwar construction. In addition to the deleterious effects of rent control acts suggested above, there are other adverse consequences as well. First, it is likely that home owners withhold letting their premises even when they do not need the space because of fear of permanent occupation by the tenant. This can apply to portions of self-occupied premises as well. Second, the revenues of local authorities are affected since they are based on annual value of holdings--and these cannot obviously be higher than the legislated rent control value. Third, a black market is created whereby transacted rents are higher than the legislated rents. Fourth, there is little evidence that the equity argument is justified since tenants are often found to be better off or at least as well off as the landlords. In summary, the rent control acts must be adjudged as serious impediments to the expansion of housing supply in urban India and urgent measures must be taken to amend them. There are also other constraints in the expansion of housing supply in India. Until recently, when the supply of cement was controlled, as was that of steel, it was difficult to obtain these key construction materials, except in the black market at heavily inflated prices. Constraints in the supply of these materials along with the increase in energy prices which has made bricks more expensive, has made  the price of construction increase at a faster rate than the overall wholesale price index. With increasing deforestation problems, the price of timber has also increased significantly. There is little hope of these prices mioderating unless supplies are increased significantly in the future through expansion in domestic production or in imports of materials. The effect on construction cost of price rises is exacerbated by the exercise of antiquated building by-laws and regulations, which discourage the introduction of cost reduction techniques. These laws do not encourage the use of new materials, lack uniformity and are more specification oriented than purpose oriented. The resulting minimum costs of construction are far too high in relation to the income of the Indian urban population. The vast majority of the poor have then little choice but to live in dwellings which cannot be legalized under the existing by-laws. Although an amended National Building Code taking account of these problems has been recommended at the national level, it is yet to be implemented by local authorities. Separate guidelines for housing for the poor are also being suggested now. These recent suggestions must be implemented early if there is to be any likelihood of the poor getting access to legal housing in urban areas in the foreseeable future. E. Housing Finance The last issue addressed in the paper is that of housing finance. Under the system of rationed credit in India, the availability of housing finance has been highly restricted. There is no equivalent of a building society or Savings and Loan system. Banks are not permitted to lend for housing except for a small amount restricted to the poor at  heavily subsidized rates. The banks have found it difficult to deliver even the latter. There are few sources of housing finance available although the situation has been improving in recent years. Each state has housing boards and other housing authorities who undertake new construction and finance individuals at relative low morgage rates. The refinancing agency for these activities is the national Housing and Urban Development Corporation (HUDCO). The volume of this construction is still low and the unit cost of public sector constructed structures typically too high for the people they are meant for. Other agencies supplying housing finance are the Life Insurance Corporation, housing cooperative societies, the relatively recent Housing Development and Finance Corporation (HDFC) and employers funds (including the government). The overall volume of these sources is small in relation to probable demand and much can be done to expand the supply of housing finance. It will be necessary to innovate so that procedures can be found to extend small amounts to the poor for improvements as well as new construction. Funds currently spent on expensive public sector construction could perhaps be better spent in the provision of housing finance to individuals who are better at economizing and at meeting their own needs. This paper documents the slow growth of the housing sector in India since about 1950. The key components of both demand and supply are identified. Considerable detail is given on the condition of existing housing stock, on the evidence of deteriorated supply conditions as well as improvements in quality. Problems in urban land policy, rent control acts and the lack of housing finance have been identified as the key  constraints in limiting the supply of urban housing which is clearly not meeting expanding demand resulting from urban growth. Urgent measures are needed in these areas to ameliorate the current urban housing situation in India.   I. HOUSING IN THE INDIAN ECONOMY The importance of the housing sector in the national economy can be measured in terms of its contribution to national income, capital formation and employment. This section, therefore, contains a brief description of trends in the share of housing in these aspects of India's economy. But before analyzing these trends, we need to define rental income and capital formation in the housing sector. 1/ Dwellings produce housing services which form a part of national product. The income accrues to the owner in the form of rent. It is immaterial whether the owner gets the rent from a tenant, or the owner himself lives in the house and thereby consumes the income generated out of the housing sector. 2/ In India, the contribution from the housing sector to national product is reckoned as net national income originating in the ownership of residential dwellings. Accordingly, the net rental figures are derived from gross rentals by subtracting current expenses on repairs and maintenance, and depreciation charges. The gross rental is estimated as a product of average gross rental per dwelling and the number of census dwellings, and includes inputed rent of owner occupied houses. Appendix 1 contains the details of the procedure employed for obtaining these estimates. It should be noted that the census definition of house makes no distinction between a house in the formal and informal sectors. 1/ For purposes of national income accounts in India, the housing sector covers "ownership of dwellings" in the shape of occupied residential houses. Outlays on such activities as maintenance of utility systems, streets and schools etc. are not included. 2/ The logic of including imputed rents on owner occupied houses arises from the fact that the service is performed not by the owner himself but by the capital he owns.  Table 1.1 Contribution of Housing* to National Income (Rs. crores) National Income Percentage share of Year Income from Housing (G.D.P. at housing in Current Prices 1970-71 prices factor cost) National Income Rural Urban Total Rural Urban Total Current 1970-71 Current 1970-71 prices prices prices prices 1960-61 349 229 578 553 362 915 14071 25534 4.11 3.58 1965-66 472 342 814 670 473 1143 22030 29023 3.69 3.94 1970-71 749 608 1357 749 608 1357 36736 36736 3.69 3.69 1975-76 1053 1050 2103 804 704 1508 65993 42633 3.19 3.54 1976-77 1131 1227 2358 816 734 1550 71616 43248 3.29 3.58 1977-78 1274 1502 2776 828 762 1590 80040 46750 3.47 3.40 * Relates to residential buildings only. Source: Handbook of Housing Statistics, 1980, NBO and National Accounts Statistics, CSO, Series 1960-61 - 1972-73, 1960-61 - 1974-75 and 1970-71 - 1975-76.  - 3 - The contribution of the housing sector to the national income of India is shown in Table 1.1. During the period 1960-61 to 1977-78, the share of the housing sector in national income, whether in current prices or at constant prices, has been around three to four percent, with some marginal fluctuations. But in absolute terms, a significant increase in the amount of aggregate income from housing occurredduring this period; the increase is, however, less sharp when housing income is considered at constant prices. On the whole, the aggregate housing income in urban areas has grown at an average rate of nearly 4.2 percent per annum (compound at constant 1970-71 prices) during the period 1960-61 to 1977-78, a rate faster than the overall growth of housing income in the country. Comparative figures indicate that the share of national income from housing is lower in India than in more developed countries. 1/ Table 1.2 shows that the share of expenditure on rent in private final consumption expenditure has not shown any change during 1950-51 and 1977-78. The amount of money a nation spends on housing is perhaps an appropriate indicator of its commitment to housing, and indeed of the quality of its stock of houses. At present, India spends about two to three percent of its GNP on housing. This is much lower when compared to developed countries. Table 1.3 illustrates the importance of the housing sector in a number of developed countries, and shows a rising trend in the proportion of housing investment as a percent of GNP during the 1970s as compared to the 1960s. 1/ For instance, France (8.7 percent); Greece (7 percent); Japan (6.8 percent) and West Germany (5.6 percent).  - 4 - Table 1;2 Percentage private final expenditure on Total final consumption rent to total at at 1970-71 prices Year- 1970-71 prices (Rs. crores) 1950-51 3.3 14708 1955-56 3.2 17540 1960-61 3.6 21202 1965-66 3.5 23836 1970-71 3.5 29838 1975-76 3.6 33454 1977-78 3.5 36002 Source: Central Statistical Organization.  Table 1.3 Residential Investment as Percentage of GNP Country 1961-69 1970-7& Japan 6.0 7.6 France 5.8 7.1 West Germany 7.2 6.5 Belgium 5.9 6.2 Canada 4.5 5.7 Holland 5.0 5.6 Italy 6.6 5.4 United States 4.2 4.4 Britain 3.5 3.6 Source: OECD quoted from the Economist, January 17, 1981 in Munjee, Nasser M., "Profile on Housing", Indian Merchants' Chamber, Bombay, 1982, p. 16.  - 6 - By contrast, the national income accounts indicate that housing investment as a proportion of gross capital formation in India has shown a steady decline from about 30 percent in 1950 to only 9.4 percent in 1975-76 (Table 1.4). In 1977-78 it rose to about 15.3 percent. To a large extent, this is to be expected in an economy experiencing considerable diversification with massive investment in industry. What is more disturbing, however, os that the rate of growth of gross capital stock in housing (in terms of real value) has been only about 1.5 to 1.6 percent per annum. The growth of housing, as indicated by the above trends, is not very encouraging, especially when compared to the developed countries and even some of the developing countries. This is true as well of urban housing, although the situation is relatively more satisfactory. These figures, however, are only broad magnitudes, as the quality of data, both in respect of the income generated in the housing sector as well as capital formation in housing, is somewhat uncertain. For instance, the income generated in the housing sectors ignores for intercensal periods that part of the informal sector housing stock which is left out of the municipal house tax net. Another source of underestimation is the use of rateable value of the property for purposes of estimating rentals of residential dwellings. These values are grossly underestimated: first, because a lot of "black' money is known to be concealed in land and buildings, and second, in order to effect savings on the stamp duty. it is, however, difficult to estimate the extent of such underestimation. Since the capital formation estimates themselves are based on the gross rental values, clearly these estimates are likely to be underestimates. Yet another source of error is that neither the expenditures  Table 1.4 Capital Formation in Housing* Gross capital GDP at factor GCF in residential Year formation (GCF) - cost (GDP) dwellings 1970-71 1970-71 prices 1970-71 prices prices (GCFR) GCF/GDP GCFR/ GCFR/ GCFRU/ (Rs. crores) (Rs. crores) Urban Total GCF DGP GCFR (GCFRU) (GCFR) (1) *(2) (3) (4) (5) (6) (7) (8) 1950-51 2379 17501 285 698 13.59 29.34 3.99 40.83 1955-56 3323 20924 218 501 15.88 15.08 2.39 43.51 1960-61 4523 25603 296 553 17.67 12.27 2.16 53.53 1965-66 6170 29086 380 782 21.21 12.67 2.69 48.59 1970-71 7177 36736 491 962 19.54 13.40 2.61 50.11 1975-76 8653 43633 425 810 19.83 9.36 1.86 51.04 1977-78 10790 46750 - 1650 23.08 15.29 3.53 - * Relates to residential buildings only. Source: Handbook of Housing Statistics, 1980, NBO.  -8- incurred on improvements of the existing housing stock nor the demolitions of the houses are usually reported, and hence do not get reflected in these estimates. If, however, the extent of underestimation is assumed to be more or less constant from one period to the other, the trends, both in the gross value of rentals as well as in capital formation in housing, can be inferred as constant. It would be interesting to examine the relative contribution of private and public sectors in housing investment. Unfortunately, the CSO does not provide such sectorwise data about the investment in housing. Except for the planned expenditure on housing schemes, no data on housing investment in teh public sector are available on a regular basis, The Government's role in this field has been limited, however, dur to both general resource constraints and the lower priority given to investment expenditure which does not contribute directly to further economic growth. In the main, the role of Government in the past has been restricted to providing subsidized housing in the public sector for industrial workers and economically weaker sectors of the community. In addition, finance was also provided to State Governments and Central and State Housing Boards for undertaking projects of land acquisition and housing construction to address the needs of the low and middle income groups. Also, provisions were made by the public sector undertakings for accommodating their own employees. If we take into account the total investment of the public sector in housing, we find its contribution to be rather small. The Planning Commission itself has pointed out that investment in housing as a percentage of total investment in the economy has fallen from 34 percent in the First Plan to about eight percent in the Sixth Plan (Table 1.5). Further, total public sector outlay in housing has been falling as a proportion of total public sector investment from 16 percent in the First Plan to about 1.6 percent in the Sixth Plan.  Table 1.5 Investment in Housing in India - Private and Public Sector Total investment in the Investment in Housing * Investment in housing as economy (in Rs. erores)- (Rs. crores) percentage of total Public Private Total Public Private Total investment in the Plan period economy First Plan 1560 1800 3360 250 900 1150 34 Second Plan 3650 3100 6750 300 lOOQ 1300 19 Third Plan 6100 4300 10400 425 1125 1550 15 Fourth Plan 13655 8980 22635 625 2175 2800 12 Fifth Plan 31400 16161 47561 1044 3636 4680 10 Sixth Plan 97500 74710 172210 1491 11500 12991 7.5 (1980-85) * Does not include investment i& informal housing. Source: India's Five Year Plans, Government of India.  - 10 - Housing, as a labour intensive activity, assumes particular significance in a labor surplus economy like India, since it can provide greater employment per unit of expenditure when compared with many other sectors of the economy. According to estimates prepared by the National Buildings Organization of India (NBO), investment of Rs. ten million in building construction is capable of generating on-site employment of 923 man years and off-site employment of 1,477 man years. According to the 1971 census, nearly 2.2 million workers were engaged in construction activity in India. This represents 1.23 percent of the country's total workers, compared to 1.09 percent for 1961. The proportion of workers engaged in building construction was 0.75 percent of the total workers in 1971, compared to 0.68 percent in 1961. Table 1.6 shows that these proportion are higher for urban than for rural areas. The Sixth Plan envisages a higher share of employment in the construction sector. Thus, though housing presently contributes significantly to national income, capital formation and employment, the trends in housing growth are not particularly encouraging. Since housing expenditure is found to have strong inter-industry linkages in the Indian economy 1/, an increase in housing investment would induce an increase in the output of almost all sectors in the economy. This suggests that higher priority be accorded to housing in the allocation of scarce resources. The study also shows the dominant role of private sector investment in housing. Clearly, if housing is to be encouraged, policies conducive to more private investment should be formulated. Finally, it is useful to consider explicitly the adequacy of the data base on housing. The existence of a substantial informal sector housing in India as well as variety of concepts used in the collection of data on housing 1/ See Dholakia, Bakul, H., Housing in India, NBO (mimeo).  Table- 1; 6 Share of Employment in Housing to Total Employment and Employment in Construction According to 1961 and 1971 Census 1961 1971 Employment Sector Rural Urban Rural Urban 1. Proportion of workers 0.67 3.65 0.74 3.50 in construction to total workers 2. Proportion of workers in 0.32 2.89 0.39 2.43 building construction to total workers 3. Proportion of workers in 46.85 79.15 52.59 69.55 building construction to workers in construction Source:N.B.O.,Handbook of Housing Statistics, 1981.  - 12 - is likely to vitiate some parts of the foregoing analysis. It would, therefore, be useful to undertake detailed studies relating to the role of informal housing sector in the economy. Also scrutiny of the concepts used in the collection of data, especially with regard to trends in share of housing in national product, capital formation and employment, with the object of evolving a reasonable uniformity in the concepts would go a long way in improving our knowledge of housing in India.  - 13 - II. HOUSING DEMAND-ANDYSUPPLY This section describes the trends of various determinants of demand for and supply of urban housing. Housing, one of the most durable and expenaive of assets, is an immobile form of goods which can be considered in two ways - as housing stock and as services supplied by that stock. Housing stock may be defined as the fixed capital stock that is accumulated for the purpose of sheltering the population. Housing services or housing consumption, on the other hand can be thought of as the services implied by the use of the housing stock. The services which are derived from housing stock have two dimensions: the quantity of housing services yielded by housing stock is normally related to some measure of the physical size of a unit of stock; the quality of services is, on the other hand, a vague concept relating to the amount of satisfaction which can be obtained from a given quantity of services. These together can be combined into a single measure and termed as housing services. The distinction between the stock of housing and the flow of services it yields over a period of time provides the basis for two measures of value, commonly encountered in the housing market. These are rent, which is the payment made for a flow of housing services received over a specified period of time, and price, which is the capital value associated with a particular unit of the stock. The aggegate demand for urban housing has a different behavior in the short run than in the long run. In the short run, both the stock and the supply of housing services are relatively fixed, because housing is an extremely durable commodity. In a short period of time, the supply of housing  - 14 - services will be determined largely by the used stock of housing, with an annual rate of growth of housing stock that is only about 1.5 percent. It requires tremendous efforts and involves high costs to add significantly to the supply in a short period. It is clear therefore, that in the short run housing services would be mainly controlled by the demand side of the market, with short run demand largely a reflection of incomes and income expectations as well as of costs and availability of financing. In the long run the aggregate demand for housing is related to the size and distribution of the population as well as to the magnitude and distribution of household income, and to relative prices of housing services. There are numerous estimates of the income elasticity of demand for housing services. The range of these estimates is rather large, extending, for example for the U.S. data, from 0.3 to 1.5. 1/ The British evidence appears to be consistent with that obtained in the United States. 2/ There are hardly any reliable elasticity estimates available for India. According to one estimate both income and price elasticities for housing demand are found to be significantly less than +1 and -1 respectively. 3/ In short, we may assume demand for housing to be a function of income level, relative prices, population size, and distributions of households, household size, and income level. The aggregate supply of urban housing is related to the prices of quantities of inputs of factors of production. One 1/ Most estimates are however in the range 0.3 to 0.5 for renters and 0.5 to 0.7 for owners. See Mayo Stephen K. "Theory and Estimation in the Economics of Housing Demand," Journal of Urban Economics 10, 1981, pp. 95-116. 2/ For a brief description on this, see Robinson, Ray, Housing Ecnomics and Public Policy, MacMillan, London, 1979, pp. 41-49. 3/ Dholakia, Bakul, H., Housing in India, N.B.O. (mimeo).  - 15 - of the most influential of these factors is land. 1/ The long run aggregate supply schedule of residential dwellings appears to be highly price elastic. Building material prices as well as wage rates paid to labor seem to have little relation to the rate of new residential construction. Overall, the housing supply can be assumed to be a function of the cost of land and building construction, tenure, and availability of funds. I1.1 DEMAND FOR HOUSING Three major factors relating to the demand for urban housing will be described below: the trends in urban population, household size and household income. II.1.1 Trends in Urban-Population and-Household Formation One of the important factors determining the demand for housing is the rate of growth of population. Table 2.1 shows the trends in town and urban population growth from 1901 to 1981. The table shows that almost 24 percent of India's population in 1981 lived in urban areas, with 86 percent of the population in towns of more than 20,000 persons. Further the number of towns has been increasing over the period. Although both the rate and level of India's urbanization are clearly low when compared to most other countrie of the world, India is so populous a nation that, except for China, USSR, USA, Japan and Brazil, this increase during 1971-81 is larger than the total urban population of any other country. 1/ Various studies indicate a large difference between the supply elasticity of urban land in the aggregate and that of urban residential land.  Table 2.1 Growth of Urban Population in India Population in towns Total above 20,000 Total urban Percentage population Number persons population Urban Percentage growth of Year (millions) of towns (millions) (millions) of total population urban population 1 2 3 4 5 6 7 1901 232.0 1834 13.5 25.6 11.0 - 1911 246.0 1776 13.8 25.6 10.4 0.00 1921 244.3 1920 15.5 27.7 11.3 0.79 1931 270.7 2049 19.6 33.0 12.2 1.77 1941 309.0 2210 28.7 43.6 14.1 2.82 1951 349.8 2844 43.2 61.6 17.6 3.52 1961 424.8 2330 61.4 77.6 18.3 2.34 1971 528.9 2531 89.6 107.0 20.2 3.26 1981 658.1 3245 134.9 156.2 23.7 3.86 Source: Rakesh Mohan and C. Pant, "The Morphology of Urbanization in India - Some Results from the 1981 Census", Planning Commission, India, 1982, (mimeo).  Table 2.2 The Growth of Large Cities* and Their Hinterland** Annual Growth Rates Percentage per year) 1981 Population ('000) City Hinterland City City Hinterland 1951-61 1961-71 1971-81 1971-81 1 2 3 4 5 6 7 Calcutta 9165 1377 2.26 2.05 2.69 3.04 Bombay 8277 1273 3.42 3.70 3.26 5.09 Delhi 5713 2073 5.08 4.45 4.59 5.83 Madras 4276 972 2.35 5.01 3.04 2.93 Bangalore 2913 1127 4.42 3.27 5.82 3.40 Hyderabad 2528 642 1.03 3.71 3.42 5.08 * City refers to urban agglomeration ** Hinterland for each city refers to all those towns with 1971 population of 20,000 persons and more and within roughly 100 km. radius of the city measured as straight line distance. Source: Rakesh Mohan and C. Pant, The Morphology of Urbanization in India - some results for the 1981 Census, Planning Commission, India, 1982, p. 22.  - 18 - Another feature of the urbanization process is that the increase in urban population has been due not only to an extension in the urban boundaries, but also to a rapid increase in the population of the existing urban areas, especially the large cities. This can be seen from Table 2.2. Apart from population levels, a study of housing requires some idea of the number of households. Because the household is an independent economic unit, the need for housing accommodation and the choice of types and locations are best expressed on a household basis rather than on a population basis. Table 2.3 shows that the number of households in urban areas has also been increasing. Table 2.3 also shows a rising trend in the size of the households, which maybe related to a shortage of housing stock. Unfortunately, at present the data with regard to the number of residential houses is not available for the census year 1981. Since urbanization is closely related to the phenomenon of development, we expect this trend in household size growth to continue for some time. A consequence of this is likely to be strain on various civic services, including housing. Because of the financial constraints it is not easy for housing, including related civic amenities, to keep pave with rising population. The pressure on housing is seen in the growth of slums, especially in big cities, as well as increased crowding. The increased demand for housing is met in several ways: (i) through new constructions, (ii) through more units created from the existing housing stock, with house owners moving in to smaller apartments, and through subtenancy, (iii) through squatting. A major part of the demand for housing caused by the migration of large numbers of rural poor is met through unauthorized constructions on public land. This phenomenon of squatting on public land is rampant in large cities where both land and construction costs are prohibitively high, especially when considered in relation to the low affordability of poor migrants.  - 19 - Table 2.3 Population Household Average Household Census-Year (million) (million) Size 1 2 3 4 1951 62.4 12.8 4.785 1961 78.9 14.9 5.295 1971 108.8 19.1 5.696 1981 156.2 27.42 5.696* * On the assumption that the household size was the same in 1981 as in 1971.  - 20 - II.1.2 Household- Income-and-Housing Expenditure Household income is the major determinant of affordability of a given type of dwelling. Thus, any analysis of housing affordability requires a knowledge of both the trends in the average income of households and the breakdowns among different household income groups. First, analyzing the distribution of urban households and the level of household expenditure, in Table 2.4 we find that about ten percent of the urban households had a per capita expenditure of less than Rs. 28 per month at 1972-73 prices, while about nine percent of households had an expenditure level of about Rs. 100 per month. The remaining households were in the range of Rs. 28 and Rs. 150 per month. The National sample Survey data for 1973-74 indicate similar trends with regard to distribution of households in different expenditure groups. 1/ As might be expected, both these tables show that the household size is generally small for households with high levels of per capita expenditure. Table 2.5, enumerating these trends for households in metropolitan cities like Bombay, Calcutta, Delhi and Madras, shows that the percentage of households in the expenditure group below Rs. 28 is negligible in Bombay and Delhi. In Madras, this percentage is about seven percent, in contrast to the all-India picture where almost ten percent of households had a per capita expenditure of less than Rs. 28 per month. Further, almost 20 to 25 percent of the households in Bombay and Delhi had an expenditure of Rs. 150 and above at 1972-73 prices, with Madras and Calcutta at almost half this percentage. This indicates a generally higher rent paying capacity of households in metropolitan cities, as shown in the last four columns of Table 2.5 and the last column in Table 2.4. And we find that the rent levels in metropolitan 1/ See Table 3.12.  - 21 - Table-2.4 Consumer expenditure on rent per person for a period of 30 days classified by monthly per capita expenditure classes in urban India (1972-73) Monthly per Percentage Estimated Per capita capita expen- estimated number of expenditure diture class number of persons per on rent* (Rs) households household (Rs.) 0 - 13 0.35 4.80 0.10 13 - 15 0.27 5.90 0.26 15 - 18 0.76 6.12 0.22 18 - 21 1.37 6.45 0.26 21 - 24 2.15 6.48 0.28 24 - 28 4.24 6.36 0.39 28 - 34 8.38 6.15 0.57 34 - 43 13.83 5.80 0.83 43 - 55 15.88 5.35 1.36 55 - 75 17.96 4.72 2.18 75 - 100 13.44- 3.82 3.49 100 - 150 12.51 3.12 6.08 150 - 200 4.57 2.70 10.68 200-and-above 4.33 2;42- 21.74 All Classes 100.00 4.72 2.52 * Rents on residential houses, etc. (no imputation of rent is made for owned houses). It does not include expenditure on utilities. Source: Sarvekshana, Vol. II, No. 3, N.S.S.O.  Table 2.5 Consumer Expenditure on Rent* Per Person for a Period of 30 Days Classified by Monthly Expenditure Class (1972-73) Monthly Per Percentage Distribution of Estimated Number of Persons Per Capita Expenditure on Rent* Capita Ex- Estimated Number of per Household in in penditure Households in Class Bombay Calcutta Delhi Madras Delhi Calcutta Bombay Madras Delhi Calcutta Bombay Madras 0 - 13 0.25 0.24 - - - 1.86 5.18 - - - - 13 - 15 0.11 - - 0.25 - - 2.50 11.00 - - 1.00 - 15 - 18 - 0.62 - 0.48 - 3.50 - 6.52 - 1.90 - - 18 - 21 0.16 0.76 0.48 0.86 9.06 4.77 13.97 5.85 0.29 2.07 0.41 0.99 21 - 24 0.13 0.90 0.64 2.00 7.06 6.43 8.50 7.16 0.52 0.89 0.63 0.84 24 - 28 0.17 2.00 1.46 3.77 7.11 6.16 6.00 7.76 0.24 1.76 1.39 1.24 28 - 34 0.75 2.96 4.53 7.44 7.25 5.92 6.23 6.81 0.80 2.14 0.66 1.67 34 - 43 2.35 7.14 8.01 13.18 6.56 5.17 6.13 5.69 1.39 2.23 1.48 2.64 43 - 55 8.64 9.35 12.70 14.62 5.55 5.50 6.52 5.39 1.65 3.12 1.66 4.04 55 - 75 15.73 16.31 16.29 16.91 5.23 3.38 5.51 5.15 2.86 4.14 2.31 4.50 75 - 100 18.38 24.22 15.82 16.36 4.20 2.62 4.24 4.11 4.04 5.36 3.17 8.36 100 - 150 27.73 22.92 19.54 14.16 3.58 2.39 3.27 3.47 8.95 10.69 6.16 10.71 150 - 200 12.69 5.65 9.93 3.63 3.40 3.39 2.78 2.41 12.16 20.49 7.78 27.41 200 and above 12.91 6.93 10.60 6.36 3.04 2.85 2.96 2.48 34.41 33.74 22.73 20.72 All classes 100.00 100.00 100.00 100.00 4.63 3.45 4.10 4.86 6.22 7.36 5.53 5.54 * Rents on residential house, residential land and other consumer goods (no imputation of rent for residential houses owned by the same households was made) Source: Sarvekshana, Vol. II, No. 3, N.S.S.O.  - 23 - cities are indeed generally higher. I/ These two tables also show that the share of expenditure devoted to rent declines as expenditure level rises, suggesting that expenditure elasticity of demand for housing is less than unity. But much more comprehensive data is needed to arrive at meaningful estimates of income elasticity for housing demand. Table 2.6 sets out the distribution of households in slum areas by monthly household income. The figures are available for three cities: Hyderabad, Bangalore and Ahmedabad. First, we find substantial variations in the income distributions in these three cities. In Ahmedabad, for instance, over 75 percent of the households in slum areas had a total household income of Rs. 200 and above, whereas in Hyderabad, the percentage was about 30 percent, and in Bangalore 40 percent. And in Ahmedabad, almost one quarter of the households had a monthly household income of more than Rs. 400 per month. Comparison of average income distributions of households in metropolitan cities with those of slum dwellers reveals clearly the inferior economic situation of slum dwellers. 2/ Analysis of the distribution of households according to expenditure levels, clearly shows that it is the income level rather than the expenditure level that is the more appropriate measure of rent paying capacity. At the same time, both income and wealth status are indicative of the house-owning capacity of people. The NCAER collected some useful data on income, wealth 1/ The average rents are arrived at by dividing the total rent by total number of persons irrespective of whether they are owner occupiers or live in tenanted dwellings. Thus these rents are clearly underestimated. But comparisons of rent levels in metropolitan cities with the all-India level would be roughly valid if we assume that the extent of underestimation is roughly the same in two cases. 2/ There are, of course, some households in slum areas whose dwellings are comparable with those in the non-slum areas. The essential difference usually is the inferior amenities, especially the public services available to slum dwellers.  - 24 - and savings according to income range. Table 2.7 shows that nearly 47 percent of the households belonged to economically weaker sections of the society (on the assumption of Rs. 4800 per annum as boundary household income). Their shares work out to be roughly 19.7 percent of total income 1/, 13.1 percent of wealth and 4.9 percent of savings. In studies carried out by the Housing and Urban Development Corporation, the poor expressed a preference for house ownership, thus suggesting the possibility of encouraging savings among the poor through appropriate housing schemes. Any discussion of strategies to improve urban housing in India must take into account the family network and the custom for family members outside of the immediate household to assist the household in availing itself of major opportunities such as housing. This suggests that purchasing power for major durables is greater than either incomes or savings by themselves would indicate. 1/ Comparison of 1975-76 income distribution with that of 1967-78 shows that the share of bottom 50 percent urban households in total urban household income has increased from about 14.3 percent to 22 percent during the period.  - 25 - Table 2.6 Percentage Distribution of Slum Households by Monthly Income Monthly Household -Percentage of-slum households in Income-(Rs;--) Hyderabad Bangalore Ahmedabad 1 - 49 10.9) 50 - 99 18.1 ) 27.6 ) 100 - 199 39.8 32.9 ) 23.9 200 - 299 16.6 28.3 ) 300 - 399 9.2 9.2 ) 50.6 400 and above 5.4 2.0 25.5 Source: Handbook of Housing Statistics, 1980, N.B.0., Delhi  Table 2.7 Distribution of Households, Average Household Income and Wealth by Income Class - Urban India, 1975-76 Average annual Percentage Average Percentage Average Percentage Gross Percentage income of the Household Income share in Wealth share in Saving Share in household Rupees Total Rupees Wealth Rs. Lakhs Gross Saving Income Below 1200 1.34 874 0.2 4115 0.44 20 0.06 1200 - 2400 11.55 1864 3.0 1732 1.60 1006 0.30 2401 - 3600 18.08 2959 7.0 2846 4.13 4605 1.38 3601 - 4800 16.28 4113 9.5 5301 6.93 10735 3.21 4800 - 6000 13.78 5337 10.4 5874 6.50 18017 5.39 6001 - 7500 10.13 6752 9.7 11116 9.04 21096 6.31 7501 - 10000 11.23 8512 13.5 11416 10.30 36308 10.87 10000 - 15000 8.62 12014 14.6 29163 20.18 52499 15.71 15001 - 20000 4.20 17232 10.2 32831 11.08 49938 14.94 20001 - 25000 2.07 22453 6.6 47917 7.97 38048 11.39 25001 - 30000 1.22 27137 4.7 83213 8.16 27947 8.36 30001 - 40000 0.91) 40001 - 60000 0.37) 47441 10.0 113461 27.32 73929 22.12 60000 and above 0.22) all Source NCAER, Survey of Household Income and its Disposition, 1975-76, New Delhi  - 27 - II.2 STRUCTURE OF THE-HOUSING'STOCK Having analyzed the major demand factors, the study turns now to the amount and kinds of housing we have to meet this demand. The analysis is based on the housing data collected during various decennial population censuses and rounds of the National Sample Survey. II.2.1 Trends in Urban Housing Stock 1/ Table 3.1, which lists the number of occupied residential dwellings for the census years, shows that during the period 1951-71, the number of occupied residential houses increased from 10.3 million to 18.1 million. Also, the net addition of four million dwellings to the housing stock during 1961-71 was slightly higher than for the period 1951-61. Table 3.1 shows that the annual rate of growth of dwelling units has averaged about 2.5 percent. 2/ This relatively slow rate of increase is directly attributable to the durability of housing which results in a stock that is the cumulative total of many years' production; for instance, nearly half of the housing stock in 1965 was built before 1945. These figures, however, should be treated with caution. First, the net addition to the housing stock during the various decades is the result not only of new constructions but also of sub-divisions and mergers of existing dwellings, as well as demolitions and conversion of existing residential dwellings into non-residential houses. There is no data with regard to such conversions. Also, the information about new constructions 1/ By housing stock we mean the number of dwelling units in existence at each decennial census, and flow represents net addition of dwellings in each decade between two censuses. 2/ This is not the same thing as the rate of growth of gross capital stock in housing (in real value terms). The dwelling number may change not only because of new construction but, say, also due to sub-division of existing dwellings.  - 28 - Table 3.1 Households and Housing Stock and Flow - Urban India Census Year Households Occupied Housing Addition to (million) residential gap housing stock houses (million) (million) (million)* 1951 12.8 10.3 2.5 - 1961 14.9 14.1 -0.8 3.8 1971 19.1 18.1 -1.0 4.0 *Includes dwellings, shop-cum-dwellings and workshop-cum-dwellings Source: N.B.O., Handbook of Housing Statistics, 1981, Delhi.  - 29 - is available only in regard to the organized housing sector and there, too, the information is inaccurate because of procedural problems in obtaining completion certificates for new buildings. Comparison of the number of housholds with the available dwellings shows that there are more households than dwellings, a gap that has shown little change during the decades 1951-61 and 1961-71. But such comparisons can often be misleading if they are used to establish housing shortages and surpluses. For one thing, the number of households is not independent of dwellings. To some extent, household formation rates are constrained by dwelling availability. For another, these aggregate indices are of limited use, since they do not reflect local and regional shortages. Further it may happen that more than one household lives in one dwelling. Indeed, census definitions of house and household indicate that such a possibility is real.!/ Though the exact size of the housing deficit is hard to estimate, it is clear that most metropolitan cities. show serious deficits, with the gap particularly large in the case of Delhi (Table 3.2). Table 3.3 also indicates that the number of households in each state is greater than the corresponding number of residential house% with substantial variations between states. Il.2.2 Value of-Housing Stock Table 3.4, which contains the estimated value of the stock of urban residential dwellings, shows that the amount actually spent on constructing the existing stock increased from Rs. 3,848 crores in 1950-51 to Rs. 9,452 crores in 1975-76. In terms of current prices, the corresponding figures are Rs. 5110 crores and Rs. 37,598 crores, and at 1970-71 prices, the estimated values are Rs. 12,889 crores and Rs. 19.066 crores. 1/ See Appendix 2.  Table 3.2 Growth of Population, Household and dwelling Units in Metropolitan Cities in 1971 Population Households Dwelling Units Metropolitan city* 1971 Increase 1971 Increase 1971 Increase (Lakhs) over 1961 (lakhs) over 1961 (lakhs) over 1961 Per- Percentage Percentage centage 1 2 3 4 5 6 7 Calcutta 31.49 - 5.65 - 5.51 - Greater Bombay 59.71 44 10.73 44 10.62 46 Delhi 32.88 59 6.05 65 5.54 93 Madras 24.69 43 3.94 20 3.92 33 Hyderabad 16.07 44 2.06 5 2.02 6 Bangalore 15.41 35 2.46 34 2.38 31 Ahmedabad 15.86 38 2.69 19 2.68 7 Kanpur 11.54 31 2.36 34 2.27 39 Poona 8.56 41 1.57 34 1.56 33 *Municipal corporation Area. Source: i. Census of India 1971, Part II A, General Population Tables. ii. Census of India 1971, Part IV KB, Housing Tables iii. National Building Organization.  - 31 - Table 3-3 Households in Urban Area, and Residential Dwellings in Urban Areas (Figures in thousands) Households Residential Houses 1960-61 1970-71 1960-61 1970-71 1. Andhra Pradesh 1122.6 1482.7 1079.5 1431.9 2. Assam 157.5 222.3 157.0 219.6 3. Bihar 706.4 1000.5 622.6 915.2 4. Gujarat 1037.2 1304.1 1024.8 1296.1 5. Haryana 245.8 307.8 243.6 299.9 6. Himachal Pradesh 44.5 58.0 43.1 55.8 7. Jammu and Kashmir 99.0 134.1 98.2 105.4 8. Karnataka 917.9 1162.2 891.0 1138.1 9. Kerala 385.4 514.0 369.4 494.2 10. Madhya Pradesh 966.6 2268.2 942.3 1151.4 11. Maharashtra 2076.5 2763.4 2050.9 2735.5 12. Orissa 234.4 363.7 215.7 348.4 13. Punjab 464.6 560.7 462.9 543.5 14. Rajasthan 634.4 827.8 637.8 806.8 15. Tamil Nadu 1782.5 2268.0 1588.5 2185.8 16. Uttar Pradesh 1790.2 2136.3 1570.9 1954.1 17. West Bengal 1650.3 1962.9 1634.3 1919.5 18. Other States and Union Territories 525.8 845.0 429.3 776.7 All India 14840.6 19084.6 14061.8 18377.9 *includes shop-cum-dwellings and workshop-cum-dwellings Sources: i. Census of India 1961, Vol. I, Part IV-B, Housing and Establishment Tables. ii. Census of India, Part IV-B, Housing Tables.  - 32 - Table 3.4 Growth of Housing Stock in Urban India Year Estimates value of stock of urban residential dwellings (Rs. crores) Hiitorical. Current Constant cost* Prices 1970-71 Prices 1950-51 3848 5110 12889 1955-56 4012 5698 13341 1960-61 4354 7521 14108 1965-66 4815 10443 14863 1970-71 6t32 17093 17093 1975-76 9452 37598 19066 *Amount actually spent on constructing the existing stock of residential buildings Source: NoB.O.  - 33 - Table 3.4 indicates that during 1950-51 and 1975-76 the value of lhotusing stock at constant 1970-71 prices increased by 67.60 percent, an annual increase of 1.52 percent (compound). The table also shows that on the average, an amount of Rs. 215 crores was invested every year.during the period. II.2.3 Availability-of Living-Space : Room and Floor Area An important index of the adequacy of the community's housing supply is provided by the number of occupants per dwelling unit. If an excessive rate exists, the unit is considered to be overcrowded - one possible criterion of substandard housing. The factor of overcrowding is frequjently related to household income. First, analyzing the availability of rooms to a household, we observe from Table 3.5 that the average number of persons per room increased from 2.62 to 2.77 during 1961-71, with the average number of persons per room in each dwelling size also showing an increase during 1961-71. The proportion of households living in "one-room" tenements and "five or more rooms" tenements shows slight decline. Analyzing inter-state differentials, we observe from Table 3.6 that in all states, except Andhra Pradesh, Jammu, Kashmir and Kerala, the average number of persons per room increased during 1961-71. The extent of crowding as indicated by figures in Table 3.7 is relatively greater in metropolitan cities, especially in Calcutta and Bombay. Comparison of Tables 3.5 and 3.8A further reveals that, while for the country as a whole, nearly 50 percent of households lived in one room tenements, the corresponding  Table 3.5 Average Number of Persons Per Room and Percentage Distribution of Population and Households by Size of Urban Dwellings - 1961 and 1971 Size of Dwelling Percentage of Population Percentage of Households Average Nmber of Unit Persons Per Room 1961 1971 1961 1971 1961 1971 One Room 43.52 41.72 53.05 50.10 4.17 4.57 Two Rooms 26.14 28.08 24.66 26.93 2.69 2.86 Three Rooms 12.55 13.62 10.29 11.42 2.06 2.18 Four Rooms 7.29 7.59 5.35 5.71 1.73 1.83 Five or more 9.73 8.95 5.76 5.64 1.28 1.34 rooms Unspecified 0.77 0.04 0.89 0.20 - - Total 100.00 100.00 100.00 100.00 2.62 2.77 Source: National Building Organization, Handbook of Housing Statistics, 1981, Delhi.  - 35 - Table- 3.6 Average Number of Rooms Per Dwelling and Average Number Persons per Dwelling in Urban India 1960-61 - 1970-71 Average Average Average Average number of number of number of number of rooms per persons rooms per persons dwelling - --per room- dwelling- per room- 1960-61 1970-71 1. Andhra Pradesh 2.050 2.54 1.935 2.39 2. Assam 2.186& 2.37 1.935 2.90 3. Bihar 2.549 2.44 2.521 2.51 4. Gujarat 1.883 2.77 1.851 3.15 5. Haryana + + 2.210 2.60 6. Himachal Pradesh 2.508 1.57 2.048 1.99 7. Jammu & Kashmir 2.974 2.07 3.951 2.02 8. Karnataka 2.131 2.61 2.222 2.63 9. Kerala 2.825 2.30 3.303 2.06 10. Madhya Pradesh 2.068 2.33 2.189 2.46 11. Maharashtra 1.647 3.16 1.639 3.37 12. Orissa 2.478 2.00 2.289 2.18 13. Punjab 2.120 2.48 2.255 2.53 14. Rajasthan 2.269 2.29 2.464 2.31 15. Tamil Nadu 1.943 2.75 1.901 2.84 16. Uttar Pradesh 2.438 3.35 2.310 2.66 17. West Bengal 1.643 3.05 1.700 3.63 India 2.62 2.77 + included in Punjab. Source: National Buildings Organisation, Handbook of Housing Statistics, 1981, Delhi.  Table 3. 7 Axversge Nuraber of Pernona Per Rooa Classified by iuxber of Rooms Occupied in Metropolitan Cities - 1971 Wumber of Persona per oom oceup Netropliiean City One Two Three Four Five or Average Number Room Rooms Rooms Rooms more Rooms of Persons Per Room Ce ct-,a 4.46 3.21 2.57 3.34 1.99 3.40 Greater Bombay 5.26 2.83 2.01 1.69 1.45 3.99 Delhi 4.50 2-93 2.16 1.80 1.40 3.04 Madras 4.23 2.99 2.16 1.70 1.41 3.01 Hyderaba 4.58 3.02 2.32 1.97 1.59 2.81 Bangalore 4.77 2.82 2.15 1.69 1.18 1.58 V.m aedabad 5,14 2,84 2.04 1,.65 1.28 3.22 4.14 2.73 2.24 1.85 1.4C 2.94 4.87 2.85 1,93 1.55 l.17 3.21 C.. Census of India 1971 iiL Katioral 3uiLding Organization, Handbook of H-ousing Statistics, 1981, Delhi  Table 3.8A Percentage Distribution of Households by Size of Dwelling Unit Occupied in Selected Metropolitan Cities - 1971 Metropolitan City -Percentage of Households in One room Two room Three room Four room Five or more Calcutta 67.6 17.2 8.1 3.7 3.3 Greater Bombay 77.4 14.2 5.3 1.9 1.1 Delhi 57.1 25.7 9.5 4.3 3.1 Madras 53.6 24.4 11.0 5.2 5.5 Hyderabad 44.9 28.9 12.8 7.2 6.2 Bangalore 45.4 27.5 11.9 7.4 7.8 Ahmedabad 57.5 27.5 7.8 3.9 3.2 Kanpur 59.6 27.0 7.1 3.3 3.0 Poona 65.1 20.9 7.3 3.6 3.0 Source: i. Census of India 1971, IV-B, Housing Tables. ii. National Buildings Organization.  - 38 - Table 3.8B Percentage Distribution of Households by the Number of Rooms Occuppied Metropolitan Cities: 1961 City Percentage-Size-of-Dwellings 1 room 2 rooms 3 rooms 4 rooms 5 rooms and above Calcutta 75.2 12.4 6.0 3.0 3.3 Bangalore 55.3 21.8 8.6 5.3 5.0 Greater Bombay 72.3 16.5 5.0 2.1 1.7 Madras (M.C). 67.5 17.1 7.2 3.7 4.2 Pune 59.9 23.5 8.8 3.6 3.6 Ahmedabad 65.3 23.0 5.5 2.8 3.2 Hyderabad City 45.8 26.9 12.7 7.3 7.3 Kanpur 62.4 23.7 6.9 3.1 2.9 Source: 1961 Population Census  Table 3.9 Percentage Distribution of Households by Number of Rooms Occupied: Urban India 53-54 N.S.S. Round Number of Rooms 7 10 11 12 15 19 28* Occupied (10/53-3/54) (12/55-5/56) (8/56-1/57) (3-8/57) (7/59-6/60) (7/64-6/65) (10/73-6/74) One 43.64 36.62 39.48 44.00 48.91 51.94 51.05 Two 28.16 34.09 29.87 28.99 28.67 27.21 25.05 Three 12.00 16.55 12.79 12.46 10.29 10.33) Four 5.53 9.01 6.43 5.21 14.25 a/ 10.65 Five and six 4.14 4.77 4.71 4.31 10.52 # Seven and above 5.55 3.07 4.08 3.41 2.61 No room - - - - - 9.65 Total 100.00 100.00 100.00 100.00 100. 00 * Number of living rooms satisfying N.B.O. specification. # Refers to Four-rooms and above. a/ Refers to Three-rooms and above. Source: i. National Sample Survey Rounds, India (Report Nos. 26, 50, 51, 67, 146, 195) ii. Sarvekshana, Vol. 1, No. 2.  - 40 figures for the metropolitan cities of Calcutta, Bombay and Pune were as high as 67.6, 77.4 and 65.1 percent respectively for 1971.1 The National Sample Survey also collected some data on household proportions according to both size of the dwelling unit and the average number of persons per room. Table 3.9 reveals a rising trend in the proportion of households living in one-room tenements. For example, the proportion increased from nearly 40 percent in the mid-fifties to 51 percent in 1973-74. Another one quarter of the hiouseholds lived in two-room tenements. While this trend is similar to one indicated by the Census data, the National Sample Survey data indicates a lower average of the number of persons per room. According to Table 3.10, the average has remained around two persons per room. The size of rooms is another measure of overcrowdig. The National Sample Survey collected data on floor space availability during some of its rounds. The last column of Table 3.10 shows that between 1953-54 and 1973-74, the floor space per person declined from nearly 7.9 square meters to 6.9 square meters. Further, the average availabilityy of floor space in metropolitan cities is lower than the all-India average. Another interesting feature is the high value of the concentration ration of 0.432 in respect to floor space when compared to that of 0.134 in respect to the number of rooms (see Table 3.10), clearly showing the vast disparities in floor space availability. Table 3.11 shows that almost 70 percent of households had, on the average, less than 300 square feet floor area. 1. Comparison of 1961 and 1971 figures show that the proportion of households residing in one room tenements declined in metropolitan cities (see Table 3.8A and 3.8B).  - 41 - Table- 3.10 Estimates of Average Number of Rooms Per Household, Average Number of Persons Per Room and Average Floor Space Per Persons from Different NSS Rounds for Urban India N.S.S. Round Average Num- Average Num- Average Floor ber of Rooms ber of Per- Space (Sq. Mt.) per Household sons per room per person 7. (10/53-3/54) - 7.86* 10. (12/55-5/56) 2.27 2.0 7.69** 11. (8/56-1/57) 2.39 2.1 7.45 12. (3-8/57) 1.47 2.2 6.45 15. (7/59-6/60) 2.01 2.3 6.47 19. (7/64-6/65) 1.97 2.3 - 28. (10/73-6/74) 2.41 2.0 6.89 * In big cities (Calcutta, Bombay, Delhi and Madras), the average floor space available per person was only 6.11 sq. meter. Also nearly 67 percent of households had floor space of less than 4.5 sq. meter per person. * Concentration ratios for the distribution of floor space and number of rooms occupied are 0.432 and 0.134 respectively. Also the availability of floor area increases with he level of household expenditure (range from 165 sq. ft. to 43 sq. ft. per person.).  - 42 - Table 3.11 Percentage Distribution of Households by Floor Space: Urban India Floor Space NSS Round (Sq. Ft.) in the House 7 10 11 12 Oct. 53- Dec. 55- Aug. 56- March-Aug. March 54 May 56 Jan. 57 1957 Below 51 10.59 5.34 6.26 3.96 51 - 100 15.85 14.58 16.42 15.51 101 - 200 24.96 30.10 26.61 30.38 201 - 300 14.19 16.00 17.80 19.51 301 - 400 10.11 11.16 9.06 9.49 401 - 500 5.92 4.69 6.65 6.85 501 - 600 5.28 3.58 4.13 3.40 601 - 800 5.01 4.91 3.66 4.22 801 - 900 2.54 2.40 3.42 2.45 900 and above 7.05 7.24 5.99 4.23 Source: N.S.S Reports 26, 50, 51 and 67  Table 3.12 Distribution of Urban Households and Average Number of Persons and Average Total Area Per Household by Monthly Per Capita expenditure Group: 1973-74 Monthly per Capita Percentage Average Average Floor Area Expenditure (Rs.) Number of Number of Total Area Per Person Households Persons Per (in sq. Meter (Sq. mt.) Household per Household Below 21.00 2.18 6.28 23.82 3.79 21.00 - 27.99 4.90 6.87 28.99 4.22 28.00 - 42.99 24.87 6.05 30.82 5.09 43.00 - 74.99 33.24 5.07 33.84 6.67 75.00 - 99.99 12.83 3.85 31.99 8.31 .100.00 - 149.99 13.47 2.93 34.01 11.61 150 and above 8.34 2.15 36.51 16.98 Net recorded 0.17 2.17 35.19 15.50 All 100.00 4.75 32.64 6.89 Source: Sarvekshana, Vol. 1, No. 2, October 1977.  - 44 - Availability of floor space and rooms according to household expenditure are shown in Table 3012 from data taken in the 28th round of the National Sample Survey for 1973-74. The Table indicates that the household size tends to be low for high levels of per capita expenditure. And, t-yhile the differences in the availability of floor space per household are not substantial, when we consider the household size, the floor space availability per person increases rather sharply with the increase in per capita income, the ration being a little more than one to four. Thus, while the per capita income has an inverse relationship with the household size, it has positive relationship with the per capita floor space availability. As a result wide disparities in the per capita space availability exist based on income. Table 3.12 shows that almost two-thirds of households indicate a per capita income of less than Rs. 75 per month. If we assume Rs. 75 per month at 1973-74 prices to be approximately the upper limit of income for the aconomically weaker sections of the population, we find that the majority can only afford rather modest housing. For tenanted households, Table 3.13 sets out the distribution of households, household size, and floor space availability according to the level of monthly rent in 1973-74. Nearly three-forths of the households residing in rented accommodation paid a rent of less than Rs. 50 per month. Since households at such low rent levels must either have the use of poor housing or be protected by the rent control legislation, it would be instructive to have a breakdown of which households are covered by rent control acts and which are not. This table also indicates a close relationship between the rent level and floor space availability, with total floor area increasing steeply with increases in rent levels. In terms of per capita floor space, the differentials are smaller for the three bottom classes. Further, while the average floor  Table 3.13 Percentage Distribution of Rental Households and Average Number of Persons and Average Total Per Rented Household by Monthly Rent Group - All-India Urban, October 1973 - June 1974 Monthly Rent Percentage of Average Average Floor Space Group (Rs.) Households Number of Total Area Per Person Persons Per (in sq. mt.) (in sq. mt.) Hosehold per Hosehold 0 - 10 19.58 3.29 16.86 5.12 10 - 20 26.14 4.08 18.67 4.58 20 - 50 29.77 5.80 25.99 4.48 50 - 100 9.75 5.26 39.28 7.47 100 - 150 2.34 5.51 53.83 9.97 150 - 250 1.47 5.43 71.26 13.12 250 and above 0.56 5.36 82.49 15.39 Not recorded 10.39 3.11 25.84 8.39 All 100.00 4.17 25.14 6.03 Source Sarvekshana, Vol. 1, No. 2, October 1977.  - 46 - space indicated per person is 6.9 square meters for all urban households in India, it is only 6.0 square meters for the household residing in rented accommodation, showing that people living in their own houses have more floor space per person. The 1971 Population Census data provides some interesting data on tenure status. Table 3.14, on the distribution of households living in rented houses classified by household size, shows, first, that nearly half of households live in rented accommodation, with considerable inter-state differentials. Second, it shows that the proportion of households residing in rented houses is lower in large-sized dwellings, and, conversely, that a high proportion live in small-sized tenements. A similar picture emerges from Table 3.15 in respect to metropolitan cities, although the proportion of households living in rented dwellings is much higher than the all-India average. In summary, the foregoing description of crowding indicates a rising trend in the average number of persons per room, with almost three quarters of the households living in one or two room dwellings and an identical percentage paying a rent of Rs. 50 or less per month. Our analysis also shows wide disparities in the per capita availability of floor space, with much smaller disparities in room availability. Since neither the type of structure nor its condition is clearly known, it is difficult to say whether floor space availability is an adequate indicator of housing. Indeed pucca small houses with adequate housing facilities are preferable to larger dilapidated dwellings with inadequate housing facilities. And, from the given data, it is difficult to say that low  Table 3.14 Proportion of Households Living in Rented Houses Classified by Size of Household - Urban India, 1971 (Per cent households living in rented accommodation) All Number of State House- One Two Three Four five Six Persons holds Person Persons Persons Persons Persons Persons Unspecified 1. India 52.88 70.01 64.24 61.92 58.48 54.43 42.94 80.14 1. Andhra 45.78 53.94 54.59 52.42 49.38 46.39 38.48 68.28 Pradesh 2. Assam 53.22 78.88 74.65 65.73 85.58 53.71 37.09 0.00 3. Bihar 46.23 73.69 60.09 55.43 51.62 46.58 33.94 31.71 4. Gujarat 58.01 66.44 66.26 67.57 65.45 61.54 49.90 96.03 5. Haryana 37.37 58.27 52.82 49.31 45.74 39.29 25.52 82.46 6. Himachal 70.97 88.27 81.52 77.36 72.79 64.58 46.06 99.28 Pradesh -4 7. Kerala 26.35 52.54 39.68 36.64 32.10 28.43 20.45 94.69 8. J and K 25.91 66.81 51.76 40.00 32.36 26.35 15.50 42.86 9. Madhya 53.94 71.02 62.29 62.46 59.21 55.15 41.56 58.06 10. Maharashtra 68.39 76.85 75.00 74.69 72.50 70.51 61.99 92.75 11. Karnataka 55.29 68.03 67.00 65.49 61.61 57.81 46.57 79.63 12. Orissa 68.10 66.41 56.94 53.17 50.62 47.49 36.27 100.00 13. Punjab 39.80 64.31 54.97 51.99 47.34 41.06 27.35 55.56 14. Rajasthan 49.09 69.98 55.82 51.11 45.25 41.09 28.18 72.57 15. Tamil Nadu 43.95 56.98 59.47 58.59 56.43 53.55 47.66 91.64 16. Uttar 45.99 69.58 59.68 55.66 51.01 46.46 34.86 57.71 Pradesh 17. West Bengal 59.87 80.94 77.90 72.92 67.77 61.27 45.26 80.34 Source: 1971 Census  Table 3.15 Percentage Households Residing in Rented Accommodation Classified by Size of the Households - 1971 Size of Household City All One Two Three Four five Six Households Person Persons Persons Persons Persons Persons 1. Ahmedabad 76.33 77.96 79.18 79.44 76.76 73.29 65.58 2. Ahmedabad (M.C) - 76.33 83.00 82.80 80.58 77.33 71.07 3. Bangalore 74.16 70.60 82.19 80.13 76.11 73.40 62.14 4. Banglaore (M.C) - 85.60 85.65 83.02 79.43 76.05 66.01 5. Bombay (M.C) 85.15 86.76 85.21 85.44 84.40 85.38 85.98 41- 6. Calcutta (M.C.) 81.27 91.86 91.72 89.44 86.58 84.00 7.60 7. Delhi (M.C.) 83.83 86.86 78.03 79.40 82.57 84.40 85.02 8. Delhi (NDMC) 55.54 67.75 62.01 61.89 59.89 56.28 48.25 9. Hyderabad 78.31 77.53 76.11 71.06 68.24 61.88 50.35 10. Kanpur (M.C) 91.33 91.33 90.04 89.24 86.24 86.53 75.08 Source : 1979 Census and NBOs' Handbook of Housing Statistics.  - 49 - rentals are necessarily an indication of low quality of housing. It is certainly true that some tenants, being protected under Rent Control Legislation, are paying old, low rents. The other feature which emerges from the preceding description is that a larger proportion of small sized households depend upon rented accommodation, many of whom are likely to be migrants to urban areas. Finally, housing deficiencies seem more serious in metropolitan cities where the crowding appears to be worst. II.2.4 Tenure In any community there are some households who prefer to rent dwellings, while others would choose to live in owned dwellings. A classification of dwelling supply on the basis of tenure is useful to determine how adequately both these demands are being met. Table 4.1 shows that in 1971 almost 53 percent of households lived in rented dwellings, with a slight shift in favor of owned dwellings during 1961-71. Comparisons at the state level indicate differing trends (Table 4.2). For example, during 1971, almost 75 percent of households in Kerala and Jammu and Kashmir lived in owned houses. The percentage was substantially lower for the industrially advanced states of Maharashtra, Gujarat, West Bengal and Tamil Nadu. These proportions were extremely low in cities where only ten to 25 percent of households lived in owned dwellings (Table 4.3). Further, during the period 1961-71, while some cities and states experienced a rise in percent of households living in owned dwellings, others showed a decline. Indeed, the relative stability of the all-India average during the decade 1961-71 is the result of combining these differing trends. To evaluate the effect of rent control  - 50 - Table 4.1 Percentage Distribution of Households by Their Tenure Status in Urban India Tenure Status Population-Census 1961 1971 Owned 46.2 47.1 Rented 53.8 52.9 Source: Census of India 1961 and 1971, Part IV B, Housing Table  - 51 - Table 4.2 Proportion of Total Households Staying in Owned Houses Urban Areas State 1960-61 1970-71 1. Andhra Pradesh 60.55 54.22 2. Assam 47.37 46.70 3. Bihar 56.50 53.77 4. Gujarat 39.-30 41.99 5. Haryana 57.20 62.63 6. Himachal Pradesh 29.89 29.03 7. Jammu and Kashmir 71.43 74.09 8. Karnataka 47.16 44.71 9. Kerala 71.74 73.65 10. Madhya Pradesh 44.08 46.86 11. Maharashtra 30.28 31.61 12. Orissa 61.68 51.90 13. Punjab 54.86 60.20 14. Rajasthan 56.71 58.91 15. Tamil Nadu 48.00 46.85 16. Uttar Pradesh 52.56 54.01 17. West Bengal 33.90 40.13 18. Other States and 33.32 41.50 Union Territories 19. All India 46.22 47.12 Source: Census of India, 1961 and 1971.  - 52 - legislation on the rental housing sector, however, would require a more disaggregative analysis. And the above analysis, admittededly, does not tell us whether the efforts in the two housing markets have on the average been equal, nor does the existing data permit us to carry the analysis further. The increase in number of rental dwellings need not imply more investment in rental housing. For instance, rising prices may prorpt many owner occupiers to rent part of the premises they previously occupied, in order to meet the increased cost of living. Also, it is quite possible that a stagnant or relatively shrinking rental stock is accommodating more households through subdivision or sharing. This is not to imply that such a process is necessarily undesirable, since rising prices may result in more efficient utilization of the housing stock. In summary, we find almost half of urban households staying in rented dwellings, with the percentage substantially higher in relatively industrialized states and in large cities. Further, while at the all-India level, the proportion of households staying in rented dwelling has not shown any appreciable change during the dacade 1961-71, the trends at the state level differ, with some states showing an increase in this percentage while others show a decliine.  - 53 - Table 4.3 Proportion of Total Households Staying in Owned Houses (for million plus cities) SI. No. Cities Urban Urban (Urban) Areas Areas 1961 1971 1. Ahmedabad 17.62 23.67 2. Bangalore 29.96 25.84 3. Bombay 10.00 14.85 4. Calcutta 17.29 18.63 5. Delhi (M.C.) 33.52 16.67 6. Hyderabad 32.99 21.69 7. Kanpur 12.87 8.67 8. Madras 26.06 26.77 Source: Handbook of Housing Statistics, N.B.O.,-1980.  - 54 - II.3 Quality-of-Housing One of the most important characteristics of housing supply is the quality of housing. A significant measure of that quality is the physical condition of the houses. The condition of structure, as well as other characteristics such as type, size, age and access to various amenities, provides the basis for evaluating the overall supply of housing better. The data base for description is the 1961 and 1971 Population Censuses, and the various rounds of the National Sample Survey. II.3.1 Types-of-Structare and-Dwellings We will first consider the type of structure of dwellings by the building materials used in their construction. The data for this purpose comes from the Population Censuses which classify the dwellings according to the kind of building materials used in the construction of walls and roofs, and from the National Sample Survey which classifies all dwelling structures under three distinct categories: pucca, semi-pucca and Katcha. 1/ A similar classification is adopted to present housing data collected in the course of the 1961 and 1971 Population Censuses. 1/ Broad definitions are: Pucca House is one whose walls and roof at least are made of burnt bricks, stone, cement concrete, jack board (cement plastered reeds) or timber. Tiles, galvanished iron or asbestos cement sheets and stone blocks used in construction of roofs are regarded as pucca material. Katcha House is one whose walls and roof are made of unburnt bricks, bamboo, mud, grass, leaves, reeds or of thatch. Semi-Pucca House is one which is neither exclusively pucca nor katcha. Generally a semi-pucca structure comprises walls made of pucca materials, namely, stones, oven burnt bricks, etc. And roofs made of Katcha materials, namely, mud, grass, etc. In some cases, it may consist of walls of katcha materials like unburnt bricks, bamboo, etc. and roofs of pucca material like timber, jack board, etc.  - 55 - First analyzing the Census data for 1971, we observe from columns 4 and 6 of Table 5.1 that over 70 percent of urban dwellings have pucca wall material and about 80 percent have pucca roof material. 1/ Clearly, the proportion of dwellings which are pucca would, in general, be less than either the dwellings which are with pucca walls or with pucca roofs. Table 5.2 classifies Census houses into four categories: Pacca, Semi-Pucca, serviceable Katcha and unserviceable Katcha. A housing unit is termed pucca if both walls and roof are pucca, serviceable Katcha if it has mud walls and thatched roof, and unserviceable katcha when both walls and roof are thatched. The remaining units fall in the category of semi-pucca dwellings. For the country as a whole, the share of pucca dwellings has increased from 46 percent in 1961 to 64 percent in 1971. Moreover, the share of katcha dwellings, serviceable and unserviceable, has declined from 19 to 13 percent. The results of the National Sample Survey confirm this trend. Table 5.3 shows that whereas the share of pocca units increased from about 28 percent in 1953-54 to 65 percent in 1973-74, the share of katcha housing stock declined from 25 to 15 percent during the period. An analysis of the data at the state level shows substantial inter-state variations, although there is a clear trend towards more pucca constructions. Further, Table 5.4 indicates a higher proportion of pucca housing stock in metropolitan cities when compared to the all-India average, although roughly ten to 20 percent of the housing stock is still katcha. 1/ Pucca wall dwelling: Burnt bricks, GI sheets, or other metal sheets, stone, cement concrete, etc. Pucca roof dwelling: Tiles, Slates, corrugated or zinc or metal sheets, ACC, RBC, RCC, brick line, stone, etc.  Table 5.1 Percentage Distribution of Census Houses by Predominant Material of Wall and Roof in Urban Area - 1971 State/Union Territories Predominant Material of Wall Predominant Material of Roof Grass, Leaves, Mud, un- Burnt bricks, Grass, Leaves, Tiles, Slates, reeds or burnt G.I. sheets, reeds, thatch, shingle, G.I. bamboo bricks, other metal bamboo, mud, sheets, bricks, wood sheets, stone, unburnt bricks, lime-stone cement con- wood RCC, RBC crete, etc. 1 2 3 4 5 6 Andhra Pradesh 6.79 30.84 62.37 37.70 62.30 Assam 58.55 6.37 35.08 36.27 63.73 Bihar 6.45 23.92 69.63 10.12 89.88 Gujarat 2.44 12.36 85.20 5.62 94.38 Haryana 0.13 9.38 90.49 43.05 56.95 Himachal Pradesh 0.13 17.37 82.50 15.58 84.42 Jammu and Kashmir 0.12 26.47 73.41 41.51 58.49 Karnataka 4.88 26.52 68.60 24.50 75.50 Kerala 9.90 24.18 65.92 35.10 64.90 Madhya Pradesh 6.04 26.45 67.51 3.07 96.92 Maharashtra 7.44 15.99 76.57 11.48 88.52 Orissa 5.45 38.91 55.64 35.18 64.82 Punjab 0.18 8.89 90.93 32.04 67.96 Rajasthan 0.45 13.39 86.16 10.52 89.48 Tamil Nadu 1.81 34.33 63.86 27.68 72.32 Uttar Pradesh 0.48 16.93 82.59 23.44 76.56 West Bengal 8.87 15.21 75.92 6.57 93.43 India 5.26 22.20 72.54 18.92 81.08 Source: Census of India, 1971, Part IV-B, Housing Tables.  Table 5.2 Percentage distribution cf Residential Houses by Type of Structure in Major States of India - 1961 and 1971 - Urbau 1961 1971 State Pucca Semi- Serviceable Unservice- Pucca Semi- Service- Unservice Pucca Katcha able Katcha Pucca able able Katcha Katcha Andhra Pradesh 29 34 30 7 49 20 24 7 Assam 29 33 - 38 30 29 - 41 Bihar 35 52 4 9 66 24 4 6 Gujarat 48 43 6 3 79 19 - 2 Haryana 30 8 62 - 55 37 8 - Himachal Pradesh 46 44 10 - 71 25 4 - Jammu & Kashmir 30 21 - 49 50 33 7 10 Karnataka 39 49 5 7 51 37 7 5 Kerala 43 7 35 15 49 15 25 11 Madhya Pradesh 38 56 2 4 64 34 - 2 Maharashtra 50 39 3 8 67 28 - 5 Orissa 26 24 42 8 46 21 28 5 Punjab 30 7 62 1 67 24 8 1 Rajasthan 74 12 13 1 80 12 7 1 Tamil Nadu 19 48 31 2 57 16 25 2 Uttar Pradesh 72 21 6 1 65 26 8 1 West Bengal 59 33 - 8 74 20 - 6 India 46 35 14 5 64 23 9 4 Source- Census of India, 1961, 1971.  Table 5.3 Percentage Distribution of Households by Type Structure and House Type of Structure Number 26 Number 195 28 Round % age of Households %age of Households %age of Household Oct. 53-March 54 July 64-June 65 Oct. 73-June 74 1. Pucca 27.64 52.75 64.52 2. Katcha 24.49 29.95 15.84 3. Semi-Pucca 38.97 17.19 19.64 4. Others 8.30 0.11 0.17 co Notes: 1. Semi-Pucca structure has been taken as those structures made withh a. Katcha plinth and wall and pucca roof b. pucca plinth, katcha wall, and roof any type c. pucca plinth and wall and katcha roof. Source: National Sample Survey Reports, 26, 195, and Sarveksnana, Vol. 1, No. 2, October 1977.  Table -5.4 Number of Households, Households Without Separate Houses, Residential Houses and Percentage of Occupied Residential Houses (Pucca, Semi-pucca, katcha) in Metropolitan Cities - 1971 City Number of Total Number of Households Without Percentage of Census Residential Separate Houses Occupied Resi- Households Houses dential Houses Number %age Pucca Semi- Katcha pucca Calcutta (M.C.) 565460 550850 14620 2.58 85.8 0.7 13.5 Greater Bombay 1072785 1062465 10320 0.96 73.4* - - Delhi 728925 599358 129567 17.78 82.3 8.5 9.2 - Madras 394625 391590 3035 0.77 69.0 4.4 26.6 Bangalore 24.5775 238145 7630 3.10 86.2 2.7 11.1 * 26.6 percent would be semi-pucca and katcha Source: Handbook of Housing Statistics, 1981, National Buildings Organisation, New Delhi.  - 60 - Table,5.5 Percentage Distribution of Households by Type of Structure and Type of Dwelling - Urban - India, 28th Round (1973-1974) Type of Structure Percentage-of-Rouseholds-in each Category Total Independent Chawls Flats or Houses Bustees Shared Houses -Pucca 19.4 5.7 39.5 64.6 Semi-Pucca 8.1 3.8 7.7 19.6 Katcha 7.8 3.0 5.0 15.8 Total 35.3 12.5 52.2 100.00 Source: Sarvekshana, Vol. 1, No. 2  - 61 - Table 5.5, based on the 28th round of the National Sample Survey, gives the distribution of households cross-classified by type of dwelling and type of structure. The dwellings are classified into three types : Chawl/Bustee, independent house, and flat. I/ The Table clearly shows the preponderence of flats and shared type of dwellings in the total housing stock. It further indicates that 12.5 percent of the total housing stock comprises bustees and chawls, more than half of which are katcha structures. The overall data on building materials point to improvement in the housing conditions through a clear trend towards pucca constructions. However, 15 to 20 percent of dwellings need to be gradually improved by replacing the existing katcha walls and roofs with pucca. Further, the problem does not seem to be confined to bustees and chawls, for even in the independent type of dwellings, over 20 percent of the stock is katcha. 1/ A Chawl/Bustee is described as a collection of huts or tenements which is generally poorly built with katcha and semi-pucca materials. These Chawls or Bustees are generally found in urban areas, particularly in cities. An independent house is a separate structure with a room or rooms with a separate entrance to the house. In other words, if the dwelling unit and entire structure of the building are physically co-terminous, it is considered as an independent house. A flat is a self-contained dwelling unit with a room or rooms within a house and provided with normal housing facilities like water supply, bath and latrine which are used exclusively by the family residing therein or jointly with other familites. It also includes detached room or rooms with or without housing facilities. It is a distinct category of dwelling as, opposed to the categories, such as, independent house and Bustee type of dwelling.  - 62 - It is clear from Table 5.6 that while nearly two thirds of the pucca structures were occupied by tenants, over 62 percent of the katcha dwellings were owner occupied. This is not unexpected since most katcha housing is usually a makeshift arrangement for shelter by the poor for their own use. Another interesting aspect of type of dwelling structure is the relative costs for their maintenance, provided by figures in Table 5.7 from the 19th round data for 1964-65. The figures show that nearly one third to one half of annual rental income was devoted to repairs. Given the increases in cost of building materials and wages of masons, etc., there is little likelihood that this percentage declined, especially since the rents are frozen for same housing stock. Further, as weather affects katcha housing more than the more durable pucca, the maintenance and repair costs of katcha houses are likely to remain greater as a percentage of rent. Finally, Tables 5.8 A and 5.8 B provide information about the type of dwelling structures which households at various expenditure levels are able to afford. Of the 24.6 percent urban households living in katcha dwellings, nearly 23 percent are indicated as having had a household expenditure level of less than Rs. 100 per month while more prosperous households are generally shown to afford superior dwelling structures. Similar features emerge from Table 5.8 B, relating to households staying in rented dwellings. Expect for expenditure class Rs. 0-8, the proportion of households living in pucca dwellings tends to increase with rise in expenditure levels, although even in the highest expenditure class, almost 30 percent of households lived in katcha and almost 20 percent lived in semi-pucca dwellings.  - 63 - Table-5.6 Percentage Households by Type of Structure and Tenure Status, Urban Indis (1973-74) Type of Structure Ownership-of-Dwelling- Total Owner Rented Occupied Pocca 22.93 41.59 64.52 Semi-Pucca 10.51 9.13 19.64 Katcha 9.79 6.05 15.75 Total 43.23 56.57 100.00 Source: Sarvekshana, Vol. 1, No. 2.  Table 5.7 Average Annual Cost of Repair and Maintenance and Monthly Rent Per Household Residing in Rented Houses - All-Indis: Urban - 1964-65 Type of Structure Monthly Rent Annual Cost Percentage Repair (Rs.) of Repair and and Maintenance Maintenance (Rs.) Cost of Rent Pucca 21.23 88.50 34.74 Semi-Pucca 10.13 55.20 45.41 Katcha 7.89 44.70 47.21 0S Source: Sarvekshana, Vol. 1, No. 2.  Table 5.8 A Percentage Distribution of Household by Structure Type and Monthly Household Expenditure Level: 1957 Monthly Expenditure of Household (Rupees)- Sl. Structure 0-25 26-50 51-100 101-200 201-300 301-400 401 and Total No. above 1. All Katcha 3.83 9.52 9.54 1.54 0.15 - 0.03 24.61 2. Katcha plinth, wall 1.08 3.33 2.48 0.51 0.04 0.25 - 7.69 and fabricated roof 3. Pucca plinth, Katcha 0.27 1.10 2.72 0.48 0.18 0.11 0.04 4.90 wall and roof of any type 0% 4. Pucca plinth, wall 0.62 2.49 4.24 1.48 0.17 0.03 0.03 9.06 and Katcha roof 5. Pucca plinth, wall 0.89 4.15 7.12 3.91 0.92 0.23 0.23 17.45 and fabricated roof 6. All Pucca 0.59 - 1.43 3.04 2.33 0.16 0.17 0.04 7.76 7. Others/not reported 2.21 7.95 9.39 5.62 2.00 0.77 0.59 28.53 8. Total 9.49 29.97 38.53 15.87 3.62 1.56 0.96 100.00 Source: National Sample Survey Report No. 67.  Table 5.8 B Percentage Distribution of Urban Households Residing in Rented Houses and Type of Structure According to Per Capita Monthly Expenditure: 1963-64 Per Capita Expen- Percentage of Pucca Katcha Semi-Pucca Percentage Households diture (Rs.) Households Residing in Rented Houses 0 - 8 37 44.86 41.60 13.54 0.60 8 - 11 24 25.25 52.70 22.05 1.69 11 - 13 30 27.87 52.91 19.80 2.24 13 - 15 29 33.67 43.60 22.54 3.66 15 - 18 31 30.20 47.94 21.85 7.13 18 - 21 34 39.61 38.65 21.59 7.62 21 - 24 36 41.45 37.85 20.59 7.96 24 - 28 43 45.55 36.43 18.02 9.91 28 - 34 46 51.43 30.19 18.28 12.72 34 - 43 50 54.01 28.94 16.93 11.97 43 - 55 55 61.94 22.21 15.68 11.07 55 - 75 56 68.06 18.77 12.87 10.84 75 and above 70 79.37 9.87 10.69 12.59 All 47 52.37 29.95 17.19 100.00 Source: N. S. S. 18th Round.  - 67 - II.3.2 Age-of^Dwellings' The age of dwelling units is significant primarily in relation to other characteristics of housing supply. For example, information about age of dwellings in itself may be of little consequence without the knowledge of their condition, size, type, etc. While the age of an existing dwelling may be a rough index of its condition, age alone will not measure the adequacy of the dwelling or its structural soundness. The quality of initial construction and the level of subsequent maintenance can make all the difference to the life span of a dwelling. In any case, the aging process alters the mix of housing types with the passage of time, as well as changing the quality mix of the existing housing stock. Table 5.9 for 1964-65 shows that nearly 50 percent of houses were built twenty years prior to 1964-65. We can assume (i) that about half of the housing stock was tenanted, and (ii) that in cities, where rent control acts have been enacted, rents of nearly 25 percent of the dwellings would be frozen at around the 1940. / level. This clearly has implications for revenues, especially property tax, which is computed on the basis of annual rental value. Based on the 16th round of the NSS, Table 5.10 presents 1960-61 data on the percentage distribution of total costs of repair and maintenance of existing urban housing stock in relation to both age and 1/ The 16th round of the NSS relating to 1960-61 shows that nearly 44 percent of the households residing in rented houses lived in dwellings built before 1940-41. This reinforces our assumption of 25 percent, of course, if half of total dwellings are assumed to be rented out.  - 68 - Table -5.9 Percentage Distribution of Households by Age of the House - Urban India Age of the House 'N; ;-S;-Round. -'19 Years 1964-65 0 - 5 13.87 6 - 10 14.61 11 - 20 17.83 Above 20 49.25 Not recorded 4.44 Total 100.00 Source Government of India, Handbook of Housing Statistics, 1979, National Building Organization, p. 34.  Table 5.10 Percentage Distribution of Total Costs of Repair and Maintenance by Period Built Since and Condition of Structure - 1960-61 Condition of Period Built Lines (Years) 0-2 2-3 3-5 5-7 7-10 10-15 15-20 20 and All above 1. Excellent 8.08 0.36 0.81 1.41 1.97 2.98 7.15 8.52 31.28 2. Fairly Good 8.12 0.28 1.01 0.32 1.27 17.66 1.14 30.12 59.92 and Needs No Repair 3. Bad, Dilapi- 1.37 0.25 1.13 0.19 0.11 1.28 0.96 3.15 8.80 dated and Impover ished Total 17.57 0.89 2.96 1.92 3.35 21.92 9.25 42.14 100.00 Source N. S. S. Round No. 16, Report No. 137.  - 70 - condition of structure. The Table shows that the share of costs on repair and maintenance was less than ten percent for "bad, dilapidated and impoverished" dwellings, although they constituted nearly 20 percent of the total part of this stock would consist of houses which were constructed to inferior standards (including katcha structures) and hence from the start would have suffered from relatively poor maintenance. There is also the possibility that some of the old housing stock which is covered by the next control legislation is being willfully neglected. Considering the scarcity of housing, if may be worth exploring the fesibility of recovering at least part of such housing stock through repairs. But before doing this, specific studies are needed to determine the relative costs involved in upgrading dilapidated dwellings verses undertaking new constructions. II.3.3 Condition-of-Structare The National Sample Survey Organization provides some additional data on the condition of the structure of the building in terms of its requirements for repair. For this purpose, the housing data was classified into three categories : "excellent," "fairly good and needs no repair," and "bad, dilapidated and inprovised." The housing stock under the last category could indeed serve as a useful indicator of the effort needed to upgrade or replace the existing housing stock. Table 5.11 gives the distribution of the households by condition of structure in urban areas for 1960-61, 1961-62 and 1963-64. It shows that during the mid-sixties nearly onefifth of the housing stock was in dilapidated condition, requiring replacement through new constructions or upgrading through major repairs. Although no recent data are available, there is no reason to expect that the situation has significantly changed. In II.3.2(above) we mentioned the need to carry  - 71 - out a comparative study of the financial implications of upgrading versus replacing such dilapidated housing stock. In fact, there was a HUDCO study of the costs involved in upgrading old buildings in the rental housing sector in Bombay whose average remaining life was six years in 1975-76. They calculated the requirement to be about Rs. 116 per square meter to upgrade these buildings so as to increase their life by 15 years. Thus, when compared to the sums of money required for new constructions, the costs of upgrading the existing old housing stock appear financially more attractive. Of course, there are problems, legal and otherwise, which may obstruct such a step. The problem may indeed by serious where tenanted buildings are covered by rent control legislation, both because of the difficulties in enhancing rents and in persuading the tenants to vacate the premises for major repairs or reconstruction. Analyzing the relative conditions of dwelling structure by tenure, we find from Table 5.12 that distributions are broadly identical for tenanted and owner occupied dwellings. 1/ Even non-rented buildings present problems. For example, if old buildings are replaced by new ones, restrictive building by laws relating to the floor space index may act as a disincentive.  Table 5.11 Percentage Distribution of Households by Condition of Structure in Urban Areas Condition of Structure 16 17 18 (1960-61) (1961-62) (1963-64) 1. Excellent 17.99 19.37 18.28 2. Fairly Good and No Repair 59.88 58.99 63.26 3. Bad, Dilapidated and Improvised 22.13 21.74 18.46 4. All Structures 100.00 100.00 100.00 Source: National Sample Survey Report Numbers, 137, 150 and 170.  - 73 - Table 5.12 Percentage Distribution of Households by Condition of Structure in Urban Areas by Tenure Status - 1963-64 Condition of Structure Rented Owned All Households 1. Excellent 18.97 17.76 18.28 2. Fairly Good and Needed No Repair 63.11 63.37 63.26 3. Bad, Dilapidated 17.92 18.87 18.46 Source: NSS Report No. 170, Round 18, February 1963 - January 1964.  - 74 - II.4 Housing-Amenities Besides the type of structure of the dwelling unit and its physical condition, access to housing amenities such as running water, lighting, toilet and kitchen facilities is the other indicator of housing quality. The following description gives an account of the availability of such amenities in urban India. II.4.1 Drinking-Water It is well known that the quality and adequacy of drinking water contributes importantly both to preventing diseases and to improving the health of the population. Table 6.1 gives a profile of the availability of drinking water facilities by sources for 1953-54, 1959-60, 1964-65 and 1973-74. Where tap water is regarded as safe, we find from this table that dependence on non-tap water sources has declined from about 55 percent in 1953-54 to 23 percent in 1973-75. Figures in Table 6.2, however, show substantial interstate variations in the availability of tap water. The situation in big cities like Calcutta, Delhi, Bombay and Madras is much better, where over 90 percent of the households are indicated to have had access to municipal tap water. In 'big towns', the percentage is 86 percent, while in 'small towns' (with population less than 3 lakhs) it is only 39 percent. The 31st round of the National Sample Survey for the period 1976-77 shows that almost 80 percent of the slum population had drinking water facilities either from tap or tube-wells. This percentage for cities with a population of one million and over is as high as 93.5 percent. For cities with less than one million population, the percentage is nearly 80. It is clear from these figures that towns, both large and small, need greater attention to the provision of safe drinking water.  - 75 - Table- 6.1 Percentage Distribution of Urban Households by Source of Drinking Water Source of National-Sample-Survey Periods Drinking Water* Oct. 53- July 59- July 64- Oct. 73- March 54 June 60 June 65 June 74 Tap 45.42 54.54 60.28 66.97 Well 40.49 32.03 24.26 18.05 Tubewell 5.25 4.92 11.31 12.69 Tanks and ponds 5.41 1.72 1.28 0.76 Rivers, Lakes 3.43 1.92 1.42 0.96 and Spring Others 4.87 1.45 0.42 Not recorded - Total 100.00 100.00 100.00 100.00 * In the big four cities, Calcutta, Delhi, Bombay and Madras, 92 percent of households get municipal tap water. In big towns nearly 86 percent and in small towns about 39 percent households get their drinking water from the municipalities. (Big towns with population over three lakhs, small towns less than three lakhs). Source: National Sample Survey of India, Report Nos., 26, 146, 195, and Sarvekshana, Vol 1, No. 2.  - 76 - Table 6.2 Access to Amenities in Urban Housing by State Percentage of Households Source of drinking Type of lighting Latrine water arrangement State/Area Tap Well & Others Electri- Kero- Others No Some tube city sene latrin latrin Andhra Pradesh 65.97 25.87 8.16 43.81 55.73 0.46 47.67 52.33 Assam 36.34 47.43 16.23 36.39 63.07 0.54 15.66 84.34 Bihar 50.38 33.81 15.81 41.33 58.28 0.39 46.16 63.84 Gujarat 84.81 11.18 4.01 67.92 31.77 0.31 27.52 72.48 Haryana 64.21 5.87 29.92 72.23 27.43 0.35 46.62 53.38 Himachal Pradesh 85.24 10.39 4.31 84.60 14.77 0.63 42.00 58.00 Jammu & Kashmir 90.40 1.98 7.62 88.88 10.17 0.95 18.77 81.23 Karnataka 78.48 20.42 1.10 57.07 42.74 0.19 37.07 62.93 Kerala 41.52 53.69 4.79 47.39 52.37 0.24 39.19 60.81 Madhya Pradesh 63.57 31.56 4.87 47.04 52.51 0.45 39.64 60.36 Haharashtra 85.60 13.53 0.87 64.84 34.53 0.63 26.48 73.52 Manipur 66.50 1.46 32.04 44.66 55.10 0.24 3.89 96.11 Meghalaya 88.31 4.44 7.25 67.29 32.01 0.70 7.71 92.29 Nagaland 80.48 13.21 6.61 55.51 44.49 - 13.79 76.21 Orissa 54.93 38.59 6.48 41.73 57.71 0.56 52.56 47.44 Punjab 41.25 3.39 55.16 77.28 21.57 1.15 12.23 67.77 Rajasthan 70.93 25.37 3.70 50.46 48.86 0.68 42.46 57.54 Tamil Nadu 72.81 17.93 9.26 51.36 48.08 0.56 39.62 60.35 Tripura 35.41 54.05 10.54 20.54 79.19 0.27 7.30 91.70 Uttar Pradesh 55.42 20.51 24.07 44.89 54.31 0.80 27.33 72.68 West Bengal 58.78 40.53 0.69 48.24 51.30 0.46 14.94 85.06 Chandigarh 98.45 1.04 0.51 82.47 16.50 1.03 11.86 88.44 Delhi 70.02 1.71 28.27 66.32 33.26 0.41 19.52 90.46 All India 66.97 23.40 9.93 53.48 45.99 0.53 33.01 66.99 Source: N.B.O., Hand Book of Housing Statistics, 1980.  - 77 - Table 6.3, based on the 28th round of the NSS for 1973-74, gives further breakdowns of the extent to which drinking water facilities were available for the exclusive use of the urban households. Table 6.3 shows that almost 77 percent of the urban households lacked the exclusive use of drinking water facilities. Further, irrespective of the nature of source, only about 40 percent of all urban households had drinking water facilities within their own dwellings. Table 6.4 also shows that almost 30 percent of the households had to traverse more than 50 yards of distance from their dwellings to obtain water. In short, efforts are needed to provide larger number of households with safe drinking water facilities within a reasonable distance from their dwellings.  - 78 - Table 6.3 Percentage Distribution of Urban Households According to Access to Drinking Water Facilities : 1973-1974 Source of For Households's For Commercial Shared with Total Drinking Water Exclusive Use Use Other households (including in the Same others not Building reported Tap 17.3 18.5 29.2 67.0 Other Sources 5.9 7.0 17.1 33.0 Total 23.2 25.5 46.3 100.0 Sources : Sarvekshana, Vol. 1, No. 2.  - 79 - Table 6.4 Percentage Distribution of Urban Households by Distance from Source of Drinking Water National Sample Survey-Round No.s 16 18 Distance from Source 7/60 - 8/61 2/63 - 1/64. (yards) Within premises 35.35 39.84 Up to 50 34.10 30.51 51 - 100 14.89 14.95 101 - 400 10.72 8.77 Over 400 4.74 5.93 Total 100.00 100.00 Source NSS Report Nos. 137 and 170  - 80 - Table 6.5 Percentage Distribution of Households by Type of Lighting in Urban India Type of Lighting N; S;-S;-Round 15 (July'59 - June'60) 28 (Oct.'73 - June'74) Kerosene Oil 71.45 45.99 Electricity 27.90 53.48 Others 0.65 0.53 Source : N.S.S. Report No. 146 and Sarvekshana, Vol. 1, No. 2  - 81 - II.4.2 Energy Source for Lighting Access to proper lighting and the implied greater access to many aspects of modern life, such as the use of appliances, bring obvious social benefits. Table 6.5 shows that the increased pace of power generation has reduced dependence on kerosene oil for domestic lighting. If we analyze inter-state differentials in the type of lighting from Table 6.2, we find wide variations. In such states as Bengal, Punjab, Jammu and Kashmir, Himachal Pradesh and Haryana, almost 70 to 80 percent of the households use electrical lighting. Andhra Pradesh, Assam, Bihar, Uttar Pradesh, on the other hand, indicated a rather flow coverage of urban households by electrical lighting. Although no recent data are available, it appears that despite rapid increase in power generation, a large number of households even in cities and big towns still depend on kerosene oil lamps for lighting. If the situation in slums is taken as a rough indicator of the deprived sections of urban households in cities, 15 to 20 percent of the population remains without any access to electric power for lighting purposes. II.4.3 Toilet Facilities Adequate toilet facilities ensure the proper disposal of human waste and minimize hazards to health by preventing the spread of diseases. Thus, from the viewpoint of sanitation and hygiene, availability of toilet facilities is another important factor in quality of housing. Table 6.6 shows that almost one third of dwellings were without any facility of built-up latrine. Columns (16) and (17) of Table 6.2 indicate wide inter-state variations with regard to this amenity. Table 6.7, which gives the breakdown according to the type of laterine used, shows that the proportion of households using the flush system of laterine has gone up from about 12 percent to 20 percent, and of septic tank system latrine from about five  Table 6.6 Percentage Distribution of Urban Households by Facilities of Built up Latrine National Sample Survey Round Number 7 10* 11 12 19 28 Facility of built up latrine 10/53-3/54 12/55-5/56 8/56-1/57 3-8/57 7/64-6/65 10/73-6/74 1. In use of individual household 21.43 20.23 20.29 18.55 23.93 24.13 2. Shared with others 34.94 36.75 37.61 37.37 36.55 42.68 3. Number built up latrines 43.63 43.02 42.10 44.08 39.04 33.01 4. Not recorded - - - - 0.48 0.18 * In big cities (Bombay, Calcutta, Delhi and Madras), the distribution being 13 percent, 81 percent and six percent. Source: National Sample Survey of India, Report Nos. 26, 50, 51, 67, 195, and Sarvekshana, Vol. I, No. 2  - 83 - Table 6.7 Percentage Distribution of Urban Households by Type of Latrine National Sample Survey-Round No. 15 28 Type of Latrine 9/61-7/62 10/73-6/74 Flush 12.40 20.08 Septic Tank 4.70 13.92 Services - 30.25 Bucket 28.37 - Pit 8.54 - No latrine 36.93 33.01 Others 9.06 2.74 Total 100.00 100.00 Source: National Sample Survey Report No. 146 and Sarvekshana, Vol. 1, No. 2.  - 84 - percent to 14 percent. Further, Table 6.8 shows that of the 65 percent households who had some type of toilet facilities, only 23 percent had the exclusive use of this amenity. The situation with regard to toilet facilities in slum areas in towns and cities appears to be even worse. The results of the 31st round of the NSS relating to survey of slums in 1976-77 indicate that for towns with less than one million population, the percentage of households without any toilet facility is around 40-50 percent and for cities with a population of one million and more it is 82 percent. Further, almost 90 percent of the slum population had to rely on community facilities for this purpose, and only five percent of those who had some latrine facility had its exclusive use. These figures attest to widespread unhygienic conditions resulting from the acute shortage of toilet facilities in urban areas, especially among slum dewllers. II.4.4 Cooking Facilities Table 6.9 sets out the proportion of urban houses having access to kitchen facilities for 1957. 1/ Colomn 8 shows that nearly 44 percent of urban households had the exclusive use of kitchen facilities, six percent shared kitchen facilities with other households, and nearly 35 percent of the households used part of thier living room for cooking purposes. The figures in the last column of the table indicate a less satisfactory situation in big cities like Calcutta, Bombay, Madras and Delhi. Expectedly, a positive relationship exists between the proportion of households with exclusive use of kitchen facilities and the level of household expenditure. 1/ Data relating to post 1957 period is not available.  Table 6.8 Percentage Distribution of Urban Households by Facility of Latrine: .1973-1974 Shared with For exclusive other house- use of the For community holds in the Other or not Type of Latrine household use same building recorded Total + 1 2 3 4 5 6 1. Flush system 7.17 2.74 9.89 0.28 20.08 2. Septic tank system 5.66 1.60 6.36 0.30 13.92 3. Services 10.07 6.43 13.02 0.73 30.25 4. Others/No latrine, etc. - - - - 35.75 5. All 22.90 10.77 29.27 1.31 100.00 Source: Sarvekshana Vol. 1, No. 2.  Table 6.9 Percentage Distribution of Urban Households by Kitchen Type and Monthly Households Expenditure (12th NSS Round - March-August 1957) Facility of Kitchen Monthly Household Expenditure in Rupees Under 26-50 51-100 101-200 201-300 301-400 Above All In big 25 400 Towns Cities 1. In Individual Use 28.97 32.40 44.08 62.98 77.18 76.42 85.00 44.23 27.05 2. In Common Use with other Households 9.47 9.62 3.46 3.62 1.95 11.34 11.92 6.05 6.31 3. Part of Living Room Used as Kitchen 45.12 41.36 35.21 26.91 12.83 4.60 - 35.04 41.20 0 4. Covered Verandah Used as Kitchen 7.62 5.68 8.32 4.98 5.67 7.64 3.08 6.79 12.10 5. No Specific Part of House Used as Kitchen 5.01 6.44 4.98 0.57 2.37 - - 4.51 2.49 6. Individuals Forming a Mess 3.81 3.13 2.59 0.63 - - - 2.40 6.01 7. Others - 1.37 1.36 0.31 - - - 0.98 4.84 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Source: National Sample Survey Report No. 67.  - 87 - II.4.5 Bathing Facilities The figures presented in Table 6.10 show that until 1960 almost two thirds of urban households were without a fixed bath. Of the remaining households, nearly 14 percent had to share a bath. Although no recent data are available, it is safe to infer that increased income of urban households would have somewhat improved the situation with regard to bathing facilities, since a close relationship between expenditure levels and access to bathing facilities is indicated from Table 6.11. II.4.6 Summary The above analysis shows that while the proportion of dwellings lacking basic amenities has declined steadily during the period under study, the situation is still far from being satisfactory, especially with regard to toilet facilities. Further, a substantial proportion of dwellings in slum areas and small and medium towns lacked most of these amenities, while most urban dwellings lacked the exclusive use of selected amenities. Admittedly, given the existing constraints on resources, it will be difficult to provide the exclusive use of these amenities in the near future. It should be possible, however, to provide these amenities at least on a shared basis. Second, the availability of amenities varied widely between different states. The existence of regional differences in the age of housing stock and household incomes are possible factors responsible for such a trend. One basic gap in the present analysis is the unavailability of detailed information about access to various amenities by income levels and tenure status. Such data would be useful to understand the links of dwelling characteristics with income levels and tenure status. Moreover, since public intervention in housing is generally income specific, income may be usefully employed as a criterion for a primary disaggregation of the housing system.  - 88 - Table 6.10 Percentage Distribution of Households by Type of Bathroom: All-India Urban N.S.S. Round 12th (March- 15th (July 59- Type of-Bath August 57) June 60) 1. In Individual Use 16.66 19.40 2. In Common Use 11.60 13.67 3. No Bathroom 71.74 66.93 Total 100.00 100.00 Source: National Sample Survey Report Nos. 67 and 146.  Table66 11 Percentage Distribution of Households by Type of Bathroom According to Household Expenditure in Urban Areas: July '59 - June '60 Type of Bathroom Monthly Expenditure Group (Rs.) 0.25 26-50 51-100 101-200 201-300 301-400 400 and above 1. No Bathroom 88.59 83.02 70.02 54.50 47.08 38.18 12.49 2. Bathroom in Individual Use 6.03 5.58 15.82 33.62 42.62 43.28 77.73 3. Bathroom in Common Use 5.38 11.40 13.47 11.88 10.10 8.54 9.78 - Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 %age Household in Monthly Expenditure Group 9.49 2.97 38.53 15.87 3.62 1.56 0.96 Source: National Sample Survey, Report No. 146.  - 90 - III. SLUMS AND SQUATTER-SETTLEMENTS IN INDIA Slums and squatter settlements are all-too-familiar parts of the urban Indian scene. 1/ According to a special survey of Conditions of Slums in Cities conducted by the National Survey during its 31st round relating to 1976-77, nearly 18 percent of the urban population in cities with over one lakh population was living in slums with extremely poor housing conditions. 2/ The following figures provide some idea of the magnitude and tenacity of the problem. 3.1 Conditions in Slum Areas A more recent survey reveals that in 1980 the same percentage of India's urban population, nearly 18 percent, continued to reside in slums (Table 7.1). The Table, which lists the estimated slum population by size class of urban agglomerations, sets this percentage for cities of one million or more at almost 29 percent. Figures in this table also reveal a close relationship between size of slum population and the total population of the city. In general, large cities contain a higher percentage of slum population, and Table 7.2 further shows that industrially developed cities have a significantly larger share of slum population. An overview of the conditions in slum areas follows, based on the results of a special survey on "Conditions of Slum Areas in Cities" conducted by the National Sample Survey during its 31st round covering the period, 1/ While there is a central legislation in India defining a slum, various state governments have enacted their own legislations for this purpose. Appendix 3 presents a statement showing the definitions of slums laid down in the Central Act and the relevant acts of some states. 2/ National Sample Survey, Draft Report No. 290, Government of India, New Delhi (Mimeo).  Table 7.1 Estimated Slum Population by Size Class of Urban Agglomerations (UA) Cities and Towns in 1980 Number of Percentage of U.A./Cities/ Estimated Estimated Slum Popula- Population size Towns in Urban Popu- Slum Popu- tion to Total of cities/towns 1971 lation Percentage lation Percentage -Population---- 10 lakhs and above 9 37563 25.06 10776 40.65 28.69 5 - 10 lakhs 10 8980 5.99 2211 8.22 24.62 3 - 5 lakhs 25 14486 9.66 2624 9.75 18.11 1 - 3 lakhs 104 23271 15.53 4125 15.33 17.73 50,000 - 100,000 183 17353 11.58 1955 7.27 11.27 Below 50,000 2319 48232 32.18 5216 19.38 10.81 All Classes 2650 149885 100.00 26907 100.00 17.95 Source: National Building Organization, Handbook of Housing Statistics, 1980, New Delhi.  - 92 - Table 7.2 Slum Population in Cities with Population 10 Lakhs and Above as Per 1971 Census (Population in '000) Estimated Populatuion in 1980 Urban Agglomeration Urban Slum Percentage or City- Population Population Sluma Population 1. Calcutta 8450 2965 35.05 2. Greater Bombay 8325 2883 34.63 3. Delhi 5438 1509 27.75 4. Madras 4968 1263 25.42 5. Hyderabad 2504 492 19.65 6. Ahmedabad 2438 602 24.69 7. Bangalore 2217 231 10.42 8. Kanpur 1634 636 38.92 9. Poona 1580 195 12.34 All cities with 37563 10776 28.69 population above 10 lakh Source: National Buildings Organization, Handbook of Housing Statistics, 1980, New Delhi.  - 93 - July 1976 - June 1977. According to this survey, there were 1,321 declared slums and 3,320 undeclared slums in 142 class one cities having 1971 populations of less than one million, and 5626 declared slums in eight big cities having 1971 populations of one million and more. 1/ The density of population in slums is generally much higher than the average density of cities, as shown in Table 7.3. The figures listed in this table also show.. that both density of population and average household size in slum areas differed widely between cities belonging to different size classes and also between eight big cities. 2/ Slum areas in Kanpur city were found to be most densely populated, followed by Delhi and Hyderabad. Major characteristics of slums vary according to city size. To demonstrate this, cities have been classified into three major categories: type A (Population between one lakh - three lakhs); type B (Population between three lakhs and one million); and type C (Population of one million and above). The characteristics in question listed in Table 7.4. First, analyzing tenure, we find that the percentage of households residing in rented houses in slum areas was higher in large cities. The average monthly rent paid varied between Rs. 15 and Rs. 18, with rents slightly lower in eight big cities when compared to smaller B type cities. Clearly, this may have to do with the quality of the housing stock. Data on the dwelling structure in slum areas show that a very large percentage of the population lived in katcha and semi-pucca dwellings. The percentage of slum 1/ An undeclared slum was defined as an area or unit having 25 or more katcha structures mostly of a temporary nature, or 50 or more households residing mostly in katcha structures with practically no private latrine and inadequate public latrine and water facilities. 2/ The slum densities of population are much greater when compared to overall population density (see Section of Land Constraints).  - 94 - Table 7.3 Density of Population and Average Household Size in Slum Areas Average No. of Persons Per acre Per City-group of land Household Cities having 1971 census Population i. 1 lakh or more but less than 151 4.8 3 lakh ii. 3 lakh or more but less than 111 4.8 1 million iii. 1 million or more: (declared slums only) a. All cities 97 4.6 b. Hyderabad 210 5.1 c. Ahmedabad 73 6.1 d. Bangalore 48 5.6 e. Bombay 40 4.5 f. Madras 190 5.0 g. Kanpur 484 4.0 h. Calcutta 88 3.7 i. Delhi 255 4.7 Source: N.S.S. Draft Report 290, Government of India.  - 95 - Table 7.4 Some Characteristics of Slum Dwellers 1976-77 (31st NSS Round) Characteristics A B C % house owned 54.12 50.74 42.47 House neither owned nor rented 8.61 6.20 8.88 Rented house 37.27 43.06 48.65 Average monthly rent (Rs.) -15.11 18.22 -.17.92 Wall % katcha 59.38 54.76 33.18 % Semi-pucca 8.95 7.94 17.46 % Pucca 31.67 37.30 49.36 Roof % Katcha 37.51 30.36 25.55 % Semi-pucca 52.21 52.59 52.23 % Pucca 10.28 17.05 22.22 Latrine % Separate for households (Sanitary) 2.55 2.32 5.18 Separate for households others 7.49 6.46 3.02 No separate latrine 89.96 91.22 91.80 Ownership with Private Sector % Areas 50.99 73.07 35.00 % Households 50.74 60.58 57.80 % Persons 50.94 59.54 56.26 Cities: A 1 lakh'to 3 lakhs population B 3 lakhs to 10 lakhs population C 1 million population and over Source: N.S.S. Draft Report 290.  - 96 - population residing in dwellings with pucca roof and wall is, however, higher for large cities. 1/ About 90 percent of households were found to lack access to a separate latrine for exclusive use of their households. Another interesting feature provided by this table is the extent of public versus private ownership of dwellings. The last three rows of the table show that nearly 65 percent of slum land area in million plus cities is publicly owned and that it housed nearly 44 percent of the population. On the other hand 56 percent of the population resided on the 35 percent of land that is privately owned. The situation is somewhat different in 142 class (B type) where, on 73 percent of privately owned land area, nearly 60 percent of the population resided. 2/ The type of housing and community facilities available to slum dwellers is presented in Table 7.6. 3/ In general, comparison of the magnitude of various housing facilities in different city groups contradicts the common belief that the problem is more acute in big cities. On the whole, we find that big cities have relatively greater access to various housing amenities. However, the table also shows that a substantial percentage of the slum population is still adversely affected by water-logging during the 1/ Table 7.5 corroborates the poor quality of dwelling structure in slums. For instance, 70 percent of slum dwellers in Hyderabad resided in dwellings whose floor material is mud. Also 57 percent of dwellings had mud walls, and nearly one fourth had thatched roofs. The situation is somewhat better in Bangalore and Ahmedabad. 2/ About 80 percent of the slum population were found to reside in slums situated in the residential areas of the cities, and the remaining 20 percent in the slums situated in the industrial commercial and other areas of the cities (N.S.S. Draft Report 290, p. 5). 3/ The percentages reflect the availability and not the adequacy of the facilities. For example, the entire population of a slum inhabited by, say, more than one thousand persons was considered as having the facility of drinking water even with one single tubewell for the use of the entire slum population. This is applicable to other facilities, too.  - 97 - Table 7.5 Housing Condition in Slums of Selected Cities Percentage Distribution of Slum Dwellings by Material of Floor, Wall and Roof %age of slum dwellings in Type of Material Hyderabad Bangalore Ahmedabad 1. FLOOR (i) Mud 69.3 0.4 ... (ii) Stone slashed 19.2 68.8 ... (iii) Cement 11.0 3.2 ... (iv) Mud and Bricks (v) Others -0.5 27.6 ... All Materials 100.0 100.0 2. WALL (i) Mud 56.5 0.3 70.0 (ii) Brick 39.6 63.0 (iii) Metal sheets 1.1 35.7 ... (iv) Bamboo, Palm leaves 2.4 ... 25.4 (v) Lime, Cement ... ... 4.0 (vi) Others 0.4 1.0 1.00 All Materials 100.0 100.0 100.00 3. ROOF (i) Thatch 23.3 0.2 ... (ii) Tiles 37.1 40.2 (iii) Metal sheets 26.0 42.2 53.00 (iv) Cement, R.C.C. 13.4 ... (v) Rags, Straw ... ... 25.00 (vi) Wood, Asbestos ... ... 22.00 (vii) Others 0.2 17.4 ... All Materials 100.0 100.0 100.00 Source: National Buildings Organization, Handbook of Housing Statistics, 1980, New Delhi.  - 98 - Table 7.6 Percentage Distribution of Slum Population by Type of Housing Facility - 1976-77 (NSS 31st Round) Percentage of slum population having the facilities in city group (Population (Population (With 1 Type of facility provided 1-3 lakhs) between 3 million and the slum areas and 10 lakhs above proper) 1. Electricity 82.1 85.4 81.1 2. Approach road other than katcha to reach slum area 78.1 87.1 91.6 3. Slum area not water logged during monsoon 53.7 51.4 65.3 4. Tap or tubewell as source of drinking water 79.9 79.2 93.5 5. Latrine facility 44.8 51.5 81.9 6. Tap or tubewell and latrine facility 40.8 47.3 81.0 7. Underground sewerage system 3.9 10.5 44.8 8. Garbage disposal system 75.7 84.7 89.9 9. With facilities (6) and (7) 3.4 9.9 44.3 10. With facilities (6) and (8) 33.3 43.3 78.0 11. With facilities (6), (7) and 8 2.6 9.7 42.8 12. Minimum needs program 15.7 26.4 59.3 13. Slum clearance program 14.9 25.8 46.9 14. Both (12) and (13) 8.6 18.8 28.4 15. Some development made in slum areas during last 5 years 53.9 63.9 56.2 Source: N.S.S. Draft Report 290.  - 99 - monsoon and by the absence of an underground sewerage system. Further, only about 60 percent of population in million plus cities is covered by a minimum needs program. The percentage of population covered by slum clearance programs is still lower (47'percent). However, almost 60 percent of the population had experienced some development in slum areas within five years prior to the survey-period. Comparison of housing conditions in slums with those in other areas would have been instructive in providing a clearer insight into the conditions of slum dwellers, but, unfortunately, no comparable data are available. In summary, the preceding description reveals the existence of deplorable conditions in the slums of India. These slums, generally characterized by poor housing with inadequate housing amenities, also lack the services of an adequate urban infrastructure. The situation appeared to be worst in smaller cities. Clearly, this reflects government policy, since both slum clearance and minimum needs programs had a wider coverage of slum population in bigger cities. 3.2 Government Policies Towards Slums and Squatter Settlements The government policies towards slums and squatter settlements have been evolved within the broad framework of existing resource constraints. First, there is the policy of general redevelopment 1/ which has proceeded under the auspices of the slum clearance program. Essentially, the concern has been with the demolition and replacement of areas of unfit housing. Under the Slum Clearance/Improvement Scheme of 1956, the government was vested with necessary powers to compulsorily acquire such areas and redevelop them. 1/ The term redevelopment is used here to refer to schemes that involve the demolition of existing housing and its replacement by new dwellings. On the other hand, improvement or renovation would generally refer to the upgrading of existing dwellings.  - 100 - Simultaneously, the emphasis of the program was also on rehabilitation. In implementing this scheme, serious efforts were to be made to rehouse the slum dwellers as far as possible at the existing sites or at sites nearby to ensure that they were not uprooted from their areas of employment. Further, keeping in view the low paying capacity of slum dwellers, this scheme aimed at providing minimum standards of environmental hygiene and essential services rather than construction of any elaborata structures. The scheme also prescribed minimum standards for redevelopment of plots and construction of tenements, as well as laying down ceiling limits for various types of accommodation. The rehousing facility was made available to slum dwellers belonging to the category of economically weakar section