Determinants of fertility decline in India

PUN Technical iNotes RES 11 TUE WORLD BANK DETERMINANTS OF FERTILITY DECLINE IN INDIA March 1984 Population, Health and Nutrition Department * This paper is one of a series issued by the Population * * IHealth and Nutrition Department for the information and * * guidance of Bank staff working in these sectors. The * * views and opinions expressed in this paper do not neces- * * sarily reflect those of the Bank. *  A B S T R A C T Indian fertility has declined by at least 25 p-rcent in the last 20 years. The current fertility level is a TFR of 4.8, which is lower than that in any other country at the same or lower level oL soci'o-economic development. Since 1975 however, the birth rate has shown a tendency to stall. This is partly due to an emerging unfavounable age-structure (increase in the proportion of women in child-bearing ages) and partly due to stagnation in the family planning acceptance rate. Nearly 90 percent of the fertility decline could be accounted for by the increase in family planning practice, and the balance by increase in age at marriage. In spite of its dominant role in fertility decline, family planning input variables were less important than socio-economic factors. About 60 percent of the interstate variation in family planning practice (and hence fertility decline) could be attributed to socio-economic differences, 10 percent to differences in family planning inputs, and the remaining 30 percent to their interaction. Our interpretation of the extent, pattern and socio-economic correlates of fertility decline in India is that in the changing socio-economic situation, the increase in the cost of bringing up children creates a situation in which a smaller family size is becoming an economic necessity to an increasing number of couples. The family planning program served not only this emerging demand for small families, but created additional v:mand for contraception through information and incentives. Prepared by: K. C. Zachariah, PHNPR Sulekha Patel, PHNPR March 1984  Table of Contents Page No. T. Introduction ............................................ 1 II. Determinants of Fertility Decline - A General Framework 2 ............................... 2 III. Fertility Trend in incdia by States .................... 1. 4 IV. Proximate Determinants of Fertility .................... 15 - Age at M4arriage and Pruportion M4arried ............. 15 - Family Planning Program ........................... 18 V. Socio-Economic Factors Related Fertility Decline ......... 27 VI. Summary and Conclusions ............................. 39 Statistical Annex ........................................... 51  List of Figures Page No. Figure 1 Crude Birth Rate, India 1971-1980 .................. 5 Figure 2 Demographic Transitions, c. 1960-1980, India and Indonesia ................................ 14 List of Maps Map 1 Total Fertility Rate by States, 178 ................... 12  List of Tables Page No. Table 1 Bir:ih Rates and Total Fertility Rates, India, 1941-51 to 1981 ............................. 6 Table 2 Birth Rates by States .............................. 7 Table 3 Total Fertility Rates .............................. 8 Table 4 Ranks of States by Fertility Rate around 1980 ........................................10 Table 5 Ranks of States with Respect to Measures of Fertility Decline ....................... 10 Table 6 Mean Age at M4arriage of Females ..................... 16 Table 7 Proportion Married (Females) by Age, India .............. 17 Table 8 Fertility Impact of Changes in Proportion Married, 1961-71 and 1971-81 ....................... 19 Table 9 Birth Rates According to SRS and Estimated Birth Rates on the Basis of Births Averted due to Family Planning Program 1971-72 to 1982-83 ............................................20 Table 10 Percent of Couples Effectively Protected, India, 1970-71 to 1981-82 ..........................20 Table 11 Cumulative Sterili-zation, IUD Acceptance Rate and Percent of Couples Effectively Protected ..........................................22 Table 12 Decline in the Birth Rate, Actual, and Estimated on the Basis of Percent of Couples Effectively Protected .... .................. 23 Table 13 Fertility Impact of Official Family Planning Program, 1961-71 and 1971-81 ................... 24 Table 14 Family Planning and Nuptiality Factors in Fertility Decline ...............................25 Table 15 Distribution of Family Planning Acceptors by Age of Wife and Method, 1973-74 and 1980-81 ............................................26  - iv - List of Tables Page No. Table 16 Average Age and Average Number of Living Children of Family Planning Acceptors .................... 26 Table 17 Percent of Eligible Couples Effectively Protekcted in Rural and Urban Areas by State by Method, 1980-81 .............................. 28 Table 18 Regression Analysis, Program Variables and Socio-economic Variables ............................. 31 Table 19 Total Marital Fertility Rate by Educational Attainment of Mothers .................. 32 Table 20 Age Specific Marital Fertility by Education, 1978 ............................. 33 Table 21 Age Specific Marital Fertility by Level of Education and Urban-Rural Residence, 1972 .......... 33 Table 22 Partial Regression Coefficients of Selected Fertility Related Variables on Years of Schooling (Wife), Kerala ........................ 35 Table 23 Regression between Children Ever-born and Education, Punjab ........................... 36  DETERMINANTS OF FERTILITY DECLINE IN INDIA I. Introduction The Indian birth rate is now around 34 per 1,000 population and the TFR is about 4.8.1/ Prior to 1950 the birth rate was about 45 and the TFR was about 6.5. Thus India has definitely experienced a fairly significant fertility decline in recent years. Recent data show that the decline in the birth rate is not continuing as fast as it did in the past; the rate has stalled somewhat'at a level of about 34. Although such a stalling has been observed in a fewi other developing countries in recent years, it was not expected on the basis of historical experience of developed countries. In the past, at a level of 34, a birth rate usually declined rapidly. For example, international data on birth rates for the 1950's and: 1960's show that there were few countries with birth rates in the range of 30-34; most of the national birth rates were either above 35 or below 25.2/ The Indian birth rate is in this range, and therefore, it should have been declining more rapiuly. Yet the decline is slow; the rate appears to have stalled. There appear to be some basic differences between the factors which moderated fertility in the 1950's and earlier, and those which seem to be operating in India and in some of the other developing countries at present. What are the factors which caused the Indian fertility to decrease in recent years? In what ways do the determinants of the recent fertility decline in India differ from those of other countries in earlier times? This paper is concerned mainly with these questions. The paper begins with a general framework for explaining fertil- ity trend in India. Inasmuch as this framework was developed largely on the basis of the work done by the author on determinants of fertility decline in Kerala,3/ this section gives a brief description of the Kerala framework of fertiTity decline and its relevance in the Indian context. The paper later proceeds to a description of fertility levels and trends by states in India, and relates the Indian experience in fertility decline with the world experience. This is followed by an analysis of the role of the family planning program and increase in age at marriage in the decline in fertility. The last section takes up the interstate variation and socio-economic differentials in fertility and family planning practices and examines the factors related to their differences in a multivariate context. The paper ends with a summary and conclusions. 1/ Th- official estimate of 4.5 for 1978 is biased because of the higher representation of unmarried women in the sample. 2/ UN Population Bulletin No. 7, 1963. Table 1.2, p. 2 3/ K. C. Zachariah. Anomaly of the Fertility Decline in Kerala, PHN Technical Notes, RES2, Population, Health and Nutrition Dept., World Bank, July 1983.  -2- II. Determinants of Fertility Decline - A General Frame.uork In a recent reappraisal of theories of fertility decline, Ronald Freedman4/ has suggested that for less developed countries today, motivation for fertility decline can arise from fewer changes than those that characterized the WZest. Under modern conditions, ideas and aspiration for a different way of life are also important in motivating lower fertility. The concept and means of family limitation within marriage produce an independent effect on fertility decline, once motivation is present. The author's analysis of fertility decline in Kerala suiggests that the concept and means of family planning have independent effects on fertility decline not only when motivation is present, but also when the motivation is weak or even absent. Official intervention in population matters is a relatively new phenomenon. This haR altered somewhat the pattern of fertility decline and its relationship with socio-economic factors. In evaluating the determinants of fertility decline in India, it is necessary to take into consideration not only the socio-economic factors which are involved in demographic transition, but also this new element of official involvement and financial incentive for family planning practice. Emperical support for such a reappraisal of the demographic transition theory is given by the-Kerala study. Kerala Framework of Fertility Decline5/ Kerala has experienced an unexpectedly rapid decline in fertility under seemingly unfavorable circumstances--low per capita income, low degree of urbanization, industrialization, etc. At the same time, the population is well-developed with respect to education and mortality decline, two very important factors in the classical transition theory. The theoretical framework assumed to explain fertility decline in Kerala includes not only these elements from the transition theory but also the new element of the influence of state policies and programs on fertility. The fertility decline in Kerala is a result of historical developments as well as recent policy interventions. Historical developments over the past 100 years have eroded the economic value of inherited characteristics (caste, religion, etc.) and inherited wealth (land), and enhanced the economic value of personal characteristics (education and health). Fueled by inter-communal rivalry in developing human capital, the State experienced a sharp decline in mortality and an accelerated improvement in educational attainment. Mortality has declined so much that 90 percent of children born today survive to adulthood, and education has improved so much that nearly all women in the childbearing ages are literate. The low mortality conditions and high educational attainment provide the necessary milieu for a 4/ Ronald Freedman. "Theories of Fertility Decline: A Reappraisal", Social Forces, 58(1):1-17, 1979. 5/ For an elaboration of the Kerala framework of fertility decline, see Zachariah, op.cit., pp. 230-243.  -3- fertility decline. The low mortality enables couples to attain their desired (surviving) family size with fewer children ever-born. The high educational attainment enables parents to make rational decisions about family size, taking into consideration mortality risks and their given socio-economic perspective; it has enabled couples to utilize family planning services effectively and limit or postpone births as and when they feel desirable; it has increased the cost of bringing up children as education has become a necessity for economic survival, and moreover, educated parents tend to give an even better education to their children; it has increased the opportunity costs of women and reduced the economic benefits from children; and it has increased the risk of uTnemployment especially at critical younger ages. While these long-term socio-aconomic changes produced the necessary milieu and caused the age at marriage to increase, the sharp decline in marital fertility was precipitated by more recent policy interventions--land reforms, agrarian reforms, and the official family planning program ia particular. The land reforms and other redistributive policies operated on fertility in different ways. The economic realignment has caused some socio-economic groups to suffer loss of land, status, income, etc., with no hope of immediate recovery, forcing them to make demographic adjustments toward later marriages and fewer children. The increase in wages has decreased the income of land owners, even those who have not lost any land, forcing them also to adopt fertility control. Wage increases have boosted the cash income of female daily wage earners, raising the opportunity cost of their time spent in bringing up children. The increase in wages also has reduced employment opportunities, forcing children out of the job market and making them economically less valuable. Even the hutment dwellers who undoubtedly gained by the land reforms and wage legislation began to feel the need for birth control, as their children have found it impossible to obtain a place for a ilut of their own. The official family planning program came at an opportune time to serve the manifest demand, strengthen the latent demand, and create new demand for birth control, especially among the poor. Applicability of Kerala Framework to other States of India The socio-economic situation in the other states is somewhat different from that in Kerala, and yet we feel that the fertility decline in these other states can also be explained in the same general framework, namely; - increase in cost of briniging up children in relation to the benefits from them, and - family planning education and services. In Kerala, the cost increase came about because of increase in the proportion of children surviving in a family, the economic necessity to invest in children's education and health, and through a series of policies related to land distribution, wages, etc. These changes are relevant in the other states also, but the degree of change and their relative  -4- importance vary significantly from state to stateg It is our hypothesis that interstate variation in the extent of fertility change in recent years can be explained by the differences in the family planning program and in the changes in the socio-economic factors which determines the cost of bringing up children in these states. The principal among these socio-economic factors are mortality, education, agricultural density, female wages, and land redistribution. III. Fertility Trend in India by States Fertility Trend in India. There is some uncertainty about the fertility level and trend in India. Recent data are more reliable and they indicate that the birth rate is about 34 and the total fertility rate is about 4.8. Working backwards on the basis of expected decline due to family planning program and changes in proportion married, an estimate of 6.5 can be taken as the TFR for India as a whole before 1960. These figures indicate a fertility decline of 26 percent in recent years. Substantial declines took place before and after 1971. Rough estimates are: a decline of 0.8 in TFR before 1971 (12%) and a decline of 0.9 during 1971-81 (16%). Since 1976 the decline in birth rate has slowed dowai. In fact, birth rate seems to have stalled (Table 1 and Figure I), although the fertility rate continues to decline, but at a slower rate. An approximate decomposition of the decline in birth rate during 1961-71 and 1971-81 is given below: Factors 1961-71 1971-81 1961-81 1. Total decline: 4.0 (100)1/ 6.0 (100)1/ 10.0 (100)1/ 2. Due to age-sex compositional change: 2.14 (54) - 1.13 (-19) 1.01 (10) 3. Due to changes in marital status composition: 0.61 (15) 0.76 (+13) 1.37 (14) 4. (2) and (3) together: 2.75 (69) - 0.37 (-6) 2.38 (24) 5. Due to fertility reduction within marriage: 1.25 (31) 6.40 (106) 7.64 (76) 1/ 4 = 44 - 40; 6 = 40 - 34; 10 = 44 - 34. It shows that during 1961-81, the birth rate declined by about 10 points. Nearly three-fourths of this decline was due to declines in marital fertility. The age-sex composition contributed 10 percent to the decline and marital status composition 14 percent. Thus the three factors reinforced each other.  -5- Figure I: Crude Birth Rate, India 1971-1980 C B R 40. ........ - -------------- ---2-- 3- 74 75 76- 77 Ii _-----7, 8 7J  - 6" Table 1. Birth Rates and Total Fertility Rates, India, 1941-51 to 1981 Birth Rate TFR Period Reported Correctedc/ Reported Correcteda/ 1941-51 39.9 43.0 1951-61 40.9 44.0 1961-71 41.1 41.1 1970 36.8 1971 36.9 40.0 1972 36.9 38.6 5.17 5.75 1973 34.6 37.4 - 1974 34.5 36.4 4.88 1975 35.2 35.5 - 1976 34.4 34.9 4.68 1977 33.0 34.4 - 1978 33.3 34.1 4.50 4.90 1979 33.7 34.1 - 1980 33.7 34.1 1981 33.9 - 4.80b/ Sources: Reported-:. 1941-51 to 1961-71. Census of India estimates (reproduced from Visaria p. 10); 1970-1978 Sample Registrationi data (from Visaria p. 19 and 1979-81 from Sample Registration Bulletin XVI, (2) December 1982, p. 34. a/ "Preliminary Estimates of Fertility Decline in India during the 1970's," A.K. Jain and A.L. Adlakha. b/ World Bank estimate. C/ SRS Bulletin, June 1983. This was however not the case in the most recent decade. In 1971-81 changes .in age-sex distribution was highly unfavorable to a decline in birth rate (i.e., the proportion women in child bearing ages increased). Had there been no change in marital status distribution and marital fertility, the birth rate would have increased by 1.1 points because of the age-structure effect. On the other hand, -in the earlier period, 1961-71, changes in age-sex composition were highly favorable to a decline in the birth rate. Had there been no change in the marital status and marital fertility the birth rate would still have declined by 2.1 points. The shift in the direction between 1961-71 and 1971-81 is very instructive. It shows that 1971-81 could be the beginning of a longer period in which increasing proportion of women in child bearing ages would tend to dampen the decline in the birth rate or make it increase even when fertility falls. The "age effect" has to be much stronger towards the end of the decade than in the beginning (evident from the change in the sign). This is one explanation of the stalling of the Indian birth rate since 1976. There could, of course be other explanations also.  -7- Fertility Trend by States The States are very heterogeneous with respect to levels and trends in the birth rate. Table 2 indicates that in five of the major states, the crude birth rate was below 30. These states are inhabited by 190 million people, more than the population of Indonesia. Such a significanit decline among such a large population is worth noting while comparing the Indian family planning program with other countries in Asia. Table 2. Birth Rates by States U. Crude Birth Rate per 1,000 Population States Census SPS 1951-61 1961-71 1970-72 1979-81 Andra Pradesh 39.7 39.2 35.4 31.5 Assam 49.3 48.4 37.9 32.7 Bihar 43.4 41.9 40.9a/ 38.4 Gujarat 45.7 41.6 40.4 35.1 Haryana 44.2a/ 44.5 39.6 36.8 Jammu and Kashmir 42.2a/ 41.5 32.5 31.3 Karnataka 41.6 39.9 32.0 28.2 Kerala 38.9 37.5 31.3 26.2 Madhya Pradesh 43.2 46.6 39.2 37.4 Maharashtra 41.2 41.0 32.0 28.3 Orissa 40.4 41.3 35.8 31.8 Punjab 44.7 36.9 34.2 29.6 Rajasthan 42.7 42.7 41.1 36.8 Tamil Nadu 34.9 36.8 31.3 28.2 Uttar Pradesh 41.5 42.5 44.5 39.5 West Bengal 42.9 44.3 36.6a/ 31.5 India 41.7 41.2 37.2 33.2 a/ Estimated by author. Sources: 1951-61 and 1961-71: Economic Situation and Prospects in India. World Bank Report No. 3872 IN. 1970-72: Family Welfare Programme in India Yearbook, 1981-82, Government of India, New Delhi, Table B.5. 1979-81: Sample Registration Bulletin, Vol. XVI, tio. 2, Dec. 1982. Ten years ago there was no state in India with a birth rate below 30, and twenty years ago there was none with a birth rate below 35.  -8- In comparison to these relatively successful states are several other states, six among the major ones, with a combined population of 314 million and a birth rate above 35 even in 1982. These states have also experienced some decline in the birth rate, but the effect of the family planning program was not large enough to make a major impact on the birth rate. Table 3. Total Fertility Rates Urban Rural Combined States* 1972 1978 1972 1978 1972 1978 Andra Pradesh 4.5 2.7 4.8 4.1 4.7 3.8 Assam 3.7 2.8 5.3 4.0 5.2 3.9 Bihar 3.9 3.4 5.1 4.2 5.0 4.1 Gujarat 4.5 3.6 6.0 4.6 5.6 4.3 Haryana 4.7 3.2 7.0 4.6 6.6 4.3 Jammu and Kashmir 3.1 2.8 5.2 4.8 4.8 4.4 Karnataka 3.5 3.0 4.5 3.7 4.3 3.5 Kerala 3.6 2.3 4.2 2.8 4.1 2.7 Madhya Pradesh 4.4 3.8 6.2 5.3 5.9 5.0 Maharashtra 3.8 2.9 4.8 3.9 4.5 3.5 Orissa 4.0 3.5 4.7 4.3 4.6 4.2 Punjab 4.1 3.2 5.7 4.1 5.3 3.9 Rajasthan 5.4 3.9 6.4 5.5 6.2 5.2 Tamil Nadu 3.0 2.9 4.4 3.5 4.0 3.3 Uttar Pradesh 4.8 4.1 6.9 5.9 6A6 5.6 West Bengal 3.6a/ 2.5 4.5a/ 3.5 4.3a/ 3.2 * States with a population of 5 million or more in 1981. a! Estimated by the author. Source: Vital Statistics Division, Office of the Registrar General. Levels, Trends and Differentials in Fertility, 1979. Ministry of Home Affairs, New Delhi, November 1981. Table 3 above gives the reported fertility rates by states in 1972 and 1978 for urban and rural areas separately. These rates are not very accurate, but can be used for comparative purposes.6! The heterogeneity shown by the birth rate is also evident in the fertility rate, both in the urban and rural areas. There are a few other sources of fertility data for inter-state comparison. These are from the 1971 and 1981 census. The principal sources are child-women ratio (children 5-9 years/women 20-49 years) and average number of children ever-born. On the basis of all these we intend 6/ Some of these rates are clearly underestimates although the number of births used in the calculations are fairly accurate. The total marital fertility rates (TMFRs) are more reliable.  -9- to make an assessment of the relative rank of the various states with respect to (i) fertility level around 1980 and (ii) fertility decline during 1970-80. Table 4 gives the ranks of the states by.the fertility indices. The various souarces are not entirely consistent, but the average ranks are likely to give the best comparison. These are given below in Column A. A comparison of fertility declines according to available measures is given in Table 5. The independent sources are not entirely consistent, but the inter-state patterns shown by the SRS data (1972 and 1978 and the child-women ratio (1971 and 1981 for the group 5-9 years) are remarkably close. On that basis we get the ranks of the states given under Column B. Column A Column B Fertility level from lowest Fertility Decline from highest to the highest to lowest 1. Kerala 1. Kerala 2. Tamil Nadu 2. Punjab 3. Maharashtra 3. West Bengal 4. Punjab 4. Gujarat 5. Andhra Pradesh 5. Maharashtra 6. Karnataka 6. Karnataka 7. West Bengal 7. Tamil Nadu 8. Gujarat 8. Andhra Pradesh 9. Orissa 9. Orissa 10. Bihar 10. Madhya Pradesh 11. Madhya Pradesh 11. Rajastan 12. Rajasthan 12. Bihar 13. Uttar Pradesh 13. Uttar Pradesh  - 10 - Table 4. Ranks of States (lowest fertility equal to 1) by Fertility Rate around 1980 CWRC/ Average No. of B.R.a/ TFRb/ 1981 Children Ever- States 1979-81 197W (5-9 yrs.) Born (30-34) 1. Kerala 1 1 1 1 2. Punjab 5 7 3 6 3. West Bengal 6.5 2 7 9 4. Gujarat 9 , 10 5 7 5. Maharashtra 4 4.5 4 4 6. Karnataka 2.5 4.5 8 9 7. Tamil Nadu 2.5 3 2 2 8. Andhra Pradesh 6.5 6 6 3 9. Orissa 8 9 9 8 10. Madhya Pradesh 11 11 10 12 11. Rajasthan 10 12 13 13 12. Bihar 12 8 11 5 13. Uttar Pradesh 13 13 12 11 a/ Birth Rate b/ Total Fertility Rate c/ Child-Woman Ratio Table 5. Ranks of the States with Respect to Measures of Fertility Decline States (1) (2) (3) (4) Andhra Pradesh 6 7 10 7.7 Bihar 8 12 12 10.7 Gujarat 4 3 4 3.7 Karnataka 7 2 7 5.3 Kerala 1 5 1 2.3 Madhya Pradesh 11 9 8 9.3 Maharashtra 5 1 5 3.7 Orissa 13 6 9 9.3 Punjab 2 5 2 3.0 Rajasthan 10 10 11 10.3 Tamil Nadu 9 .8 6 7.7 Uttar Pradesh 12 13 13 12.7 West Bengal 3 4 3 3.3 (1) Using SRS total fertility rates in 1972 and 1978. (2) Based on estimated decline due to FP and marital status change. (3) GCange in child (5-9 years) - woman ratio. (4) Average of cols. (1) through (3).  Geographic Patterns: Fertility rates and birth rates are lower in the southern States of India than in the northern States. In the north, the central States have in general higher fertility rates than those in the eas. or west (See Map 1). Assuming that fertility rates were not very much different in the past, the relatively lower rates in the south must imply a relatively more rapid fertility decline in these states. The path of fertility decline in India seems to follow the path of the south-west monsoon. The monsoon reaches India first in Kerala and then spreads northwards. On the way, it splits into two, covering northwest and northeast before reaching the north central states, namely Uttar Pradesh, Bihar, Madhya Pradesh and Rajasthan. These states are yet to experience a good shower of fertility decline. Indian Fertility in the International Context Is a birth rate of 34 or a TFR of 4.8 too high, too low, or just average for a population like India's? In one respect, this is a very low level. When its development indicators are taken into consideration, there is no other country in the world with a fertility rate lower than India's (see Figures A1 and A2 in the Annex). For example, at the level of income per capita in India, the expected TFR is 5.8, but the rate for India is 4.8. Similarly, for its literacy rate, the expected TFR is 6.6, and for its IMR the expected TFR is 6.1. Thus, with respect to each one of these socio-economic indicators, the Indian fertility rate is lower than expected on the basis of international or less developed country experience. Not only that, at these levels of development, no other country has a lower fertility rate. For example, there is no country with an IMR higher than India's but fertility rate lower than India's. Similarly there is no country with a female literacy rate lower than India's, but with lower TFR. Thus, the Indian fertility rate is relatively low when its developmental indicators are taken into consideration. But this is hardly comforting for the Indian planners. The Indian population growth rate is too high from the point of the country's developmental needs. Fertility Trend in India and Indonesia One interesting comparison of Indian fertility trend is with the Indonesian rates. As is shown in Figure II, demographic transitions in India and Indonesia are practically indistinguishable. Fertility and mortality decline in Indonesia are no higher than those in India during the period 1960-80, yet the family planning program in Indonesia is considered to be a "success" while that in India is referred to as a "failure." Ia this connection the following observation by McNicoll and Singarimbun7/ is worth quoting: 7/ Fertility Decline in Indonesia. Analyses and Interpretation Committee on Population and Demography. Report No. 20. National Academy Press, Washington, D.C., 1983, p. 112.  -12 - Total Fertility Rate by States, 1978 I N D I A 14.1 - ; A.Sh'4JL#' * 'Jlre *n'd Stnr' IIoAItItOy ctp,iItril t t *i'. | Jar- '- * .-.. *-Q >th.r Clle e , i'*o