V.. .... lHI"WIN!IllHllHiUIHUI!Ill!ll'lllllUHHIHIIINIHI 1293 00788 3626 “Effififlv Michigan Santa . University 1 L a» “IN THE?I O This is to certify that the dissertation entitled Migration and Fertility in India. presented by Surendar Singh Yadava has been accepted towards fulfillment of the requirements for Doctoral degree in Sociology Major professor Date July 26, 1989 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 02 k/ if (F/ J / PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES rotum on or before date due. DATE DUE DATE DUE DATE DUE Ml GI‘QQ. 0 3:3 \ DEL; W 9" “£5? MSU Is An Affirmative Action/Equal Opportunity Institution cMMma-pJ MIGRATION AND FERTILITY IN INDIA By Surendar Sinch Yadava A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1989 93% @050 ABSTRACT MIGRATION AND FERTILITY IN INDIA By SURENDAR SINGH YADAVA The main focus of this dissertation is to investigate the relationship between internal migration and the overall fertility of developing countries. In developed countries, rural fertility rates are higher than urban fertility rates. Fertility rates of urban residents who are migrants from a rural area are lower than the fertility rates of the residents of the region of their origin and, in some places, even lower than the fertility rates of the urban areas in which they now live. If the move from country to city reduces the fertility of migrants, then other things being constant, the main effect of rural to urban migration is to lower the overall fertility of a country. Because of the differences between rural and urban fertility rates, the overall fertility of a country'at a.given time is largely a function of the rural-urban distribution of its population, and since the main strewn of migration is typically from rural to urban areas, the future fertility rate of a country is importantly a function of the rates of rural to urban migration. These causal patterns have been well established by empirical research in developed countries such as the United States and those in Europe. But because of lack of data, we have not known whether the same causal patterns hold true in developing countries like Sri Lanka, India, Bangladesh, and countries of Africa. Relatively recently data have become available for some developing countries, and India is one of them. This dissertation will examine the relationship between migration and fertility in India, and provide an empirical test of some of the explanations of these relationships. With increasing economic opportunities, rising expectations, social integration, development of transportation and communication facilities, the volume of migration in developing countries is likely to increase in future. In developing countries, where a majority of people still live the rural areas, migration of people to urban areas will be encouraged by availability of more and better jobs and improved economic conditions in the urban areas. And since desperate efforts are being made in the developing countries to reduce population pressures, the subject of migration and fertility is of considerable interest. Copyright by SURENDAR SINGH YADAVA 1989 ACKNOWLEDGEMENTS I feel privileged while expressing gratitude to my advisor Dr. Tom Conner, Professor of Sociology at flichigan State University. I am highly thankful to him for his constantly motivating me and the pain he took during the preparation and completion of the present manuscript, and during the course of my doctoral program. My knowledge of Sociology enhanced through my association with him. I am extremely thankful to the members of my committee, Professors Frank Camilleri, Chris Vanderpool, and Alan Beegle for their valuable advice throughout my degree program and for their careful examination of the whole manuscript. I remain indebted to my parents whose high hopes and sacrifices facilitated and motivated me in achieving my educational goals. All the assistance I received from the Michigan State University and its department of sociology is duly acknowledged. Surendar Singh Yadava iv CHAPTER 1 CHAPTER 2 CHAPTER 3 CHAPTER 4 APPENDIX I APPENDIX II TABLE OF CONTENTS THEORY AND LITERATURE PROBLEM STATEMENT THEORETICAL BACKGROUND METHODOLOGY PROCESS OF ENUMERATION TABULATION PROGRAM SAMPLING & SELECTION PROCEDURES ESTIMATION PROCEDURES RELIABILITY OF ESTIMATES VALIDITY OF CENSUS DATA MEASURE OF FERTILITY MEASURE OF MIGRATION RESULTS AND FINDINGS DESCRIPTIVE STATISTICS TESTS OF HYPOTHESES SUMMARY AND DISCUSSION SOME SOCIO-ECONOMIC IMPLICA- TIONS OF POPULATION MOVEMENTS INDIAN CENSUS: AN OVERVIEW THE CENSUS OF INDIA: HISTORICAL PERSPECTIVE MODERN CENSUS: METHODOLOGY DATA FROM 1971 CENSUS OF INDIA LIST OF REFERENCES l 1 3 24 25 27 29 31 31 32 39 41 55 56 59 81 87 92 92 94 111 132 TABLE 1: TABLE 2 TABLE 3: TABLE 4 TABLE 5: TABLE 6 LIST OF TABLES CORRELATION COEFFICIENTS (URBAN GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) PARTIAL CORRELATION COEFFICIENTS (URBAN GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) CORRELATION COEFFICIENTS (RURAL GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) PARTIAL CORRELATION COEFFICIENTS (RURAL GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) CORRELATION COEFFICIENTS (STATE GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) PARTIAL CORRELATION COEFFICIENTS (STATE GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) vi 62 63 7O 71 78 79 CHAPTER 1 THEORY AND LITERATURE W The main focus of this dissertation is to investigate the relationship between internal migration and the overall fertility of developing countries. In developed countries, rural fertility rates are higher than urban fertility rates. Fertility rates of urban residents who are migrants from a rural area are lower than the fertility rates of the residents of the region of their origin and, in some places, even lower than the fertility rates of the urban area in which.they’now live. If the move from country to city reduces the fertility of migrants, then other things being constant, the main effect of rural to urban migration is to lower the overall fertility of a country. Because of the differences between rural and urban fertility rates, the overall fertility'of a country'at a.given time is largely a function of the rural-urban distribution of its population, and since the main strewn of migration is typically from rural areas to urban areas, the future fertility rate of a country is importantly a function of the rates of rural to urban migration. These causal patterns have been well established by empirical research in developed countries, such as the United States and those in Europe. But because of lack of data, we have not known whether the same causal patterns hold true in developing countries like Sri Lanka, India, Bangladesh, and the countries of Africa. Relatively recently data have become available for some developing countries, and India is one of them. This dissertation will examine the relationship between migration and fertility in India, and provide an empirical test of some of the explanations of these relationships. W? Nith.increasing economic opportunities, rising expectations, social integration, development of transportation and communication facilities, the volume of migration in developing countries is likely to increase in future. In developing countries, where a majority’ of people still live~.in the rural areas, migration of people to urban areas will be encouraged by availability of more and better jobs and improved economic conditions in the urban areas. And since desperate efforts are being made in the developing countries to reduce population pressures, the subject of migration a fertility is of much current interest. W? The reasons for studying the relationship between migration and fertility in India are the following: 1. Availability of relevant data. Among the developing countries, India has the best data available in terms of both quality and quantity. 2. India is a very large country and changes in the population of such a large country influence the whole world. 3. My own interest in and knowledge about India is an advantage in understanding the complexity of the association between migration and fertility caused by various socio-economic factors which are peculiar to India. 4. Limitations of resources. W: Demographic transition theory states that both fertility and mortality of a population will decline from high to low levels as a result of economic and social development. Accordingly, industrialization and migration (through urbanisation) have been observed as the crucial correlates of fertility decline. As mentioned before, among other factors, urbanization is caused by substantial rural to urban migration. These rural to urban migrants form a selected group with respect to age, 4 sex, marital status, and family status. They include an excess of adolescents and young adults as compared to nonmigrants, and among these migrants there are more males than females. In addition, many of these young adults are married and their migration to urban areas temporarily separates them from their spouses. This incidence of physical separation of spouses from each other is an important factor tending to depress the fertility of the rural migrants, depending partly on the frequency of visits to the villages of origin or the .duration of continuous separation (Zachariah, 1968). Zachariah, in his study of Greater Bombay, India, pointed out that migration might affect fertility through factors such as changes in the proportion ever marrying, changes in the age at marriage, separation of husbands from their wives, and changes in .marital fertility for reasons other than the physical separation of spouses. On the basis of an analysis of 1961 census data, Zachariah examined some of these variables and found that the average age at marriage and the percentage never marrying were higher among the women who came to Bombay prior to marriage compared to the women at the place of origin. Zachariah estimated a 30 percent reduction in fertility on these counts for the migrants who came to Bombay before marriage (Zachariah, 1968). 5 Though the migration differentials and the selectivity issues were addressed by Zachariah in quite detail, he was unable to separate the recent migrants from the total number of migrants. Therefore, it was not possible to compare the characteristics of the new migrants with those of the old migrants or the urban natives. The data gathered by Zachariah were not suitable for this purpose. Moreover, his estimates of the reduction in fertility due to the separation of migrant husbands from their wives were based on the assumption of periodicity of return visits and the frequency of coitus for which only very scanty data are available. But now, due to the availability'ofrmore>data, we can further examine Zachariah's assumptions. As pointed out by Zachariah, it is true that the physical separation of spouses has a depressing effect on fertility, but that seems to have only a short term effect because many of the rural migrants eventually do take their spouses (families) with them at a later date after achieving some degree of economic stability. This separation and eventual reunion of families establishes nuclear families in the urban areas. In these families the rearing of children requires more parental time and‘more market goods and services. Since the goods and services which were formerly'provided by the extended 6 family have now to be purchased, the demand for children is reduced. Moreover, the new context in which the migrants live in the form of new kinds of jobs, new surroundings, and better educational and income opportunities that become available, creates a new outlook on the world. When this outlook is accompanied by a flow of modern goods, it generates gradually rising consumption standards. These processes not only weaken the influences favoring high fertility but also set up steady pressures towards rising consumption standards that inhibit people from having large families (Stark, 1978). The above discussion pertains to the rural migrants who are married. But not all the rural migrants are married and separated from their spouses due to migration. Many of them are not married yet. Migrants in the young agergroups are relatively better educated than those who do not migrate. More educated people have a greater tendency to migrate largely because they can earn relatively higher incomes in the urban areas and for them the rural-urban differentials in income are>greater than for the less educated ones. These migrants also try to delay their marriages (Bulsara, 1964: Rele, 1964). While the majority of researchers have found or assumed that there exist rural-urban fertility 7 differentials, Visaria (1971) has asserted the virtual absence of rural-urban fertility differentials in India. He mentions that on the one hand the high fertility of the recent migrants might inflate the overall urban fertility, while on the other the lower fertility of the rural women whose migrant husbands had separated from them may lower the overall rural fertility levels. These two possible forces operating in opposite directions may be responsible for reducing the overall fertility differentials in the rural and urban areas, and even mask the lower fertility norms and behavior of the urban natives and long term migrants to urban areas. Visaria has cited the National Sample Survey (NSS) data in support of his assumption of the absence of the rural- urban fertility differentials. In order to examine the effect of husband's separation on the wife's fertility, Visaria collected the data from 23 villages in two of the districts of Gujrat and Maharashtra states of India. In both these districts the fertility of the women whose husbands were absent for more than six months was found to be lower than those with the husbands present. Information about the women who had migrated from the surveyed villages to Greater Bombay was also collected from their relatives living in the surveyed .villages and this information was verified by B interviewing a sample of these women in Bombay. Visaria observed some discrepancies in the reported age of women, and number of children born & surviving. The analysis revealed that the fertility of migrant women living in Bombay was lower than those who had not migrated from the surveyed villages. But Visaria attributed these differences to the separation of spouses, and a higher educational level of the migrant women. Thus, according to Visaria, place of origin (rural or urban) does not seem to make a difference in the fertility. The weakness of Visaria's study is that he did not control for any of the socio-economic variables. Herely, on the basis of the NSS data, Visaria contended that there was virtual absence of the rural-urban fertility differentials in India. Though Visaria's study is an improvement over Zachariah's, thus far there is no conclusive evidence to his basic premise of presence or absence of fertility differentials between the urban and the rural areas. He did not compare the fertility of the rural migrants with that of the urban natives at the place of destination. His information about the fertility of the migrants was based on the information obtained from their nonmigrant relatives, but the information provided by the nonmigrant relatives is quite likely to be different from that 9 provided by the migrants themselves. In the present investigation we have relevant data available to study the rural-urban fertility differentials by controlling for several socio-economic variables. Arguing against Visaria's study, Rele and Kanitkar have reported another study suggesting the presence of fertility differentials with regard to place of residence. Rele and Kanitkar took a step forward in the direction of studying the relationship between migration. and fertility' by taking' into account the duration of residence and some socio-economic variables. Because of the simultaneous operation of both the 'push' and the 'pull' factors, the Indian metropolitan areas gain population. Moreover, the selection at both ends of the socio-economic scale results in the migrant population containing the’members from the lower as well as the higher socio-economic groups. Thus, there is a possibility that fertility differentials by former place of residence may exist in metropolitan areas. Rele and Kanitkar have tried to locate the factors causing such fertility differentials by analyzing the fertility of currently married women in Bombay in relation to their former residence. Based on the 1966 fertility survey carried out among the 7,872 currently married women in Bombay, Rele 10 and Kanitkar analyzed the relationship between fertility and residential status. They grouped these women into three categories: rural migrants migrating to Bombay from rural areas; urban migrants, migrating to Bombay from other urban areas; and non-migrants, originally born and raised in Bombay. In other words, they grouped these women into three groups according to the place where they had spent most of their lives prior to their coming to Bombay. Their analysis revealed the fertility of the rural migrants to be the highest and that of the urban~ migrants to be the lowest. The fertility of non-migrants was between these two extremes. However, controlling the effect woman's education considerably reduced these differentials. Thus, in this study the educational background of women was found to be more important than migration in explaining the fertility differentials (Role, and Kanitkar, 1974). Bele and Kanitkar's study is an improvement over other studies reported previously in as much as the fertility rates of rural as well as the urban migrants were compared with those of the non-migrants at the place of destination. But besides education no other socio- economic variables were used as controls. Their sample included a considerable number of Parsee (related to Persian) women who are characterized by considerably 11 lower fertility. This could have been responsible for the low fertility of the non-migrants. Instead, a more representative sample should have been taken and more socio-economic variables should have been used as controls. In yet another study, El-Badry studied the fertility of different kinds of migrants. This study was also done in Bombay city in India. Bl-Badry analyzed the information contained in the birth certificates for a sample of 1,000 women between the ages of 15-49 years who gave births during the year 1060. His analysis revealed the existence of fertility differentials according to the place of origin of the husbands of these women. Women whose husbands migrated from South India had the lowest fertility while those with husbands from the North had the highest fertility. El-Badry attributed these differences to the possible variations in their socio- economic and educational status. But he lacked sufficient data to control for these variables. Interestingly enough, his analysis revealed a consistent increase in age-standardized parity averages with an increase in the duration of stay in Bombay irrespective of the region of origin. His explanation of this phenomenon was that in Bombay there was little assimilation among the migrants coming from different parts of India, and that these 12 i-igrants were very different from the native population of Bombay (El-Badry, 1967). The drawbacks in El-Badry's study were that, like Visaria, he also didlnot have enough data to control for many variables. The results of his study show an increase in wife's parity level according to the husband's place of birth. These observed differences in the parity in relation to the geographic origin of the migrants are apparently related to the differences in educational and occupational patterns, with the»migrants from the South having considerably higher education and better occupation than those from the North. But El-Badry did not have access to any statistics on such socio- economic characteristics of the migrants. The physical separation of spouses is another important factor affecting the fertility behavior of the migrants, but El-Badry paid no attention to this factor. Therefore, it is not possible to give a clear interpretation of the fertility trends from South to North India. It may either mean a relatively higher fertility level in the areas of in-migration because of an improvement in the living conditions, or a tendency for immigrants to have smaller families at the time of migration as compared to the urban natives. Subsequent researchers have tried to clarify some of these ambiguities. 13 In a study by Majumdar and Das (1972) in West Bengal state of India some aspects of adoption of family planning practices by migrant and old-settlers in five villages were studied. The study revealed that migrants had a more favorable attitude towards family planning practices than the old‘settlers and that migrants adopted institutionalized sources of family planning more than the old settlers. Host past research supports the existence of rural-urban fertility differentials. But in order to study the relationship between migration and fertility, it is necessary to consider the socio-economic and demographic factors which influence fertility even though we do not have a firm theory here to generalize the influence of, such factors, since the strength and relevance of factors which affect the relationship between migration and fertility will vary from one social setting to another. However, It is reasonable to assume that the fertility differentials do exist and these differentials widen with migration. Fertility rates of women living in the rural areas are higher than those living in the urban areas. The rural people have greater faith in and attachment to large joint family life. They believe that the power and strength of the family and brotherhood can only be bad if there is a common kitchen. _h' ,. héu 14 Furthermore, a lack of variety of occupations also leads rural people to live in joint families which inhibits migration and hence their fertility remains high. To sum up, since the migrants are usually younger in age and many of them migrate when single and delay their marriages, and those who are married leave their families behind and bring them only when they have accumulated some money and acquired a suitable place of living, the husband-wife separations result in reduced fertility levels (Bulsara, 1964). It is, therefore, held that the fertility of migrants as a class is lower than that of the.urban natives. The above discussion is su-arized in the following diagram and the arguments following the diagram: RURAL URBAN Immigrants (—————+ Original residents Orig-l Low potential residents ‘ inal Res. I High potential residents L--—--> Immigrants Returned migrants 6—————d The above diagram‘depicts various characteristics of the migrants. Based on these characteristics, following assumptions can be made: 15 1. Fertility of low potential rural migrants (those migrants who are less educated, are older in age, and have low socio-economic status and, therefore, less likely to migrate) is similar to the fertility of high potential rural migrants (those who are relatively younger, are more educated, and belong to a higher socio- economic status and, therefore more likely to migrate). 2. Fertility of urban migrants to rural areas is lower than the fertility of original urban residents. 3. Fertility of rural returned migrants is the same as the fertility of urban immigrants. 4. Fertility of rural returned migrants is lower than the fertility of either low or high potential rural residents. Since the rural migrants to urban areas acquire a higher standard of living and become more knowledgeable during their stay in the urban areas, they are more likely to have a positive impact on their rural friends and relatives after returning to the rural areas of their origin. Rural people who did not migrate are more likely to adopt the life style and advice of these returned migrants. Similarly, urban migrants to rural areas are held in higher esteem and treated with greater respect by the rural people by way of seeking advice from them, listening'to their experiences in the city, and following 13 them with an intention of improving their own life style. In summary, it can be stated that: A. Returned migrants persuade the rural others (those who remain in the rural areas) to have lower fertility. B. Urban immigrants also persuade the same rural people to have lower fertility. Therefore, 1. The greater the number of urban immigrants, the lower will be the resultant rural fertility and the greater the number of persuaders who persuade the original rural residents, the lower will be the resultant rural fertility. Hence, the total rural fertility will decrease. 2. The fewer the urban i-igrants, the higher the resultant urban fertility. The fewer the persuaders for process (A), the smaller the decrease in the fertility of low and high potential migrants. 3. Greater the number of returned migrants, more the persuaders for process (B). Hence, the fertility of both the low as well as high potential types of residents will decrease, assuming that both the returned migrants and the i-igrants persuade the rural residents to have lower fertility. Therefore, the rural to urban as well as urban to rural migration streams are hypothesized to result in lower 17 fertility. The combined effect of these migration streams will be a reduction in the overall fertility levels. In addition, it should be recognized that fertility is likely to be affected not only by the magnitude of migration but also by certain other socio-economic variables. We will, therefore, include some intervening variables such as literacy level, age at marriage, female labor force participation rate, and the per capi ta income to be used as controls. On the basis of the this discussion and assumptions we can state our first hypothesis as: W8 The higher the flow of rural migrants to urban areas of a state, lower its total urban fertility level. In order to test this hypothesis, we will examine the relationship between the total urban fertility rate and the following variables: i. Rural to urban migration rate for both sexes combined. ii. Rural to urban migration rate, males. iii.Rural to urban migration rate, females. Each of these variables will be examined controlling for the following variables: 1. Urban literacy rate, both sexes combined. 2. Urban literacy rate, males. 19 3. Urban literacy rate, females. 4. Percentage of females ever married, age 15-19 years, urban areas. 5. Percentage of females in labor force, urban areas. 6. Per capita income, urban areas. According to the above hypothesis, urban fertility rates are expected to be negatively correlated with migration variables i, ii and iii. But this relationship may be affected by various socio-economic variables. Control variables 1 through 6 listed above‘are some of the ilportant socio-economic variables. It is possible that these variables affect the fertility behavior of urban residents, but the inverse relationship between urban fertility rates and rural to urban migration rates is expected to persist even after we hold the effects of these socio-economic variables constant. In other words, the partial correlations between urban fer'tility and rural to urban migration rates are expected to be significant even after removing the effects of the control variables 1 to 6. Though the main stream of migration in most of the developing countries is from rural to urban areas, there is another stream of migration and that is the migration from urban to rural areas. This is primarily return migration in which the urban migrants are either 19 drawn back to or they are driven back by a shortage of housing and employment in the urban areas. The states which have high rural to urban migration rates, typically also have higher rates of urban to rural migration. These returned migrants not only have lower fertility themselves, but also spread the low fertility norms in the rural areas. They usually return with new ideas, attitudes and values acquired during their stay in the cities, and people at their native places often seek their advice and listen to them with respect. These permanent returnees are the catalysts of change. It is also the case that family welfare workers sometimes seek help from these returned migrants in changing the behavioral patterns of the villagers. Finally, there are those people who are originally the residents of the cities but due to, say housing problems, or high cost of living or a lack of adequate employment opportunities decide to migrate to the rural areas. They often migrate to rural areas with the intention of starting some sort of business and also to avail themselves of cheaper housing. This migration takes a large number of people, with an urban background and outlook, from urban to rural areas and like the returned migrants, these migrants also spread low fertility norms. Therefore, it is reasonable to think of 20 the existence of an inverse relationship between urban to rural migration rate and rural general fertility. Based on these arguments we can state our second hypothesis as: W8 The higher the rate of urban to rural migration in a state, the lower will its overall rural fertility levels be. In order to test this hypothesis, we will examine the relationship between general rural fertility rate and the following variables: i. Urban to rural migration rate for both sexes combined. ii. Urban to rural migration rate, males. iii.Urban to rural migration rate, females. Each of these variables will be examined controlling for the following: 1. Rural literacy rate, both sexes combined. 2. Rural literacy rate, males. 3. Rural literacy rate, females. 4. Percentage of females ever married, age 15-19 years, rural. 5. Percentage of females in labor force, rural. 6. Per capita income, rural. According to the above hypothesis, rural 21 According to the above hypothesis, rural fertility rates will be negatively correlated with migration variables i, ii and iii. But this relationship may be affected by various socio-economic variables. Control variables 1 through 6 listed above are some of the important socio-economic variables. It is possible that these variables affect the fertility behavior of rural residents, but the inverse relationship between rural fertility and urban to rural migration is expected to persist even after we.hold the effects of these socio- economic variables constant. In other words, the partial correlations between rural fertility rates and urban to rural migration rates are expected to be significant even after removing the effects of control variables 1 to 6. Since both rural to urban and urban to rural migration streams are hypothesized to result in lower fertility, the combined effect of these two streams of , migration will be a reduction in the overall fertility levels of'a.state. Therefore, a third hypothesis, derived from the above two hypotheses, is stated as follows: W3 The higher the combined rate of rural to urban and urban to rural migration, the lower the general fertility levels of a state. 22 In order to test this hypothesis, we will examine the relationship between total fertility rate of a state and each of the variables i, ii, and iii listed below: i. Rural to urban and urban to rural migration rate, both sexes combined. ii. Urban to rural and rural to urban migration rate, males. iii.Urban to rural and rural to urban migration rate, females. Each of these three variables will be examined controlling for the following: 1. State literacy rate, both sexes combined. 2. State literacy rate, males. 3. State literacy rate, females. 4. Percentage of females ever married, age 15-19 years, in entire state. 5. Percentage of females in labor force, entire state. 6. Per capita income, entire state. According to the above hypothesis, the fertility rates of the entire state will be negatively correlated with migration variables i, ii and iii. But this relationship may be affected by various socio- economic variables. Control variables 1 through 6 listed 23 above are some of the important socio-economic variables. It is possible that these variables affect the fertility behavior of the residents of a state, but the inverse relationship between the overall fertility of a state and the urban to rural a rural to urban combined migration rate is expected to persist even after we hold the effects of these socio-economic variables constant. In other words, the partial correlations between the fertility of a state and its combined urban to rural a rural to urban migration rates are expected to be significant even after removing the effects of the control variables 1 to 6. mm: This chapter discusses the methodological issues associated with the study of the relationship between migration and fertility. The main source of data for studying this relationship is the 1971 Decennial Census of India. The census. data of India is a very valuable and comprehensive source for such purposes because India possesses an advanced data system for assessing the relationship between migration and fgrtility. In fact, very few developing countries have the range of statistics possessed by India. In India, a census is conducted every ten years. A population census is conducted with reference to a specific date known as the Lgtgggn£§_da1§. The 1971 census of India was taken with April 1, 1971 as the reference date. Obviously, counting and collecting personal information on 548 million people (India's population in 1971 was 548 million) in a single day is an extremely difficult task. Therefore, a period of three weeks preceding the reference date is fixed as the enumeration period. Although the entire period is used 24 25 for enumeration, the data are collected in such a way that they relate to the reference date only. The enumeration period for 1971 census was from March 10, 1971 to larch 31, 1971. During this period the census workers throughout the country visited every household in their respective areas to record the census information. In order to relate the information to the reference date of April 1, the enumerators revisited all the households in their areas between April 1-3 to ask questions about births and deaths that might have occurred after their last visit and before April 1,1971. Necessary changes were made in their records for such births and deaths before calculating the population as on April I, 1971. The homeless people were counted in one night, that is, the night of March 31, 1971. W: Statistics on migration and fertility have traditionally been collected in the Indian census with reference to place of birth since the first census was taken in 1881. In the 1971 census the classification of migrants was done on the basis of place of birth as well as the place of last residence. The question put for collecting‘migration statistics on place of birth was as follows: a. place of birth 26 b. Rural/Urban c. District d. State/Country The above information was collected in respect of every person if he or she had another place of normal residence irrespective of his or her place of birth before coming to the present place of enumeration. Even if a person was born at the place of enumeration but had to migrate due to the nature of his work or for educational purposes, and returned to the place of enumeration, the person was deemed to have had another place of residence prior to enumeration. The last place of previous residence is relevant only if the person had been outside the village or town of enumeration and not simply in another house or locality in the same place. Where a person had merely gone out to another place or had been shifting from place to place purely on tour or pilgrimage or for temporary business purposes, he/she was not considered to have had another residence different from the place where that person or his/her family normally resided. But if the person had the normal residence elsewhere at any time prior to enumeration at his place of last residence, irrespective of where he was born, such a place of previous residence was recorded. 27 W: The census does not end with the compilation of enumeration data. In fact, it marks the beginning of another arduous operation of tabulating the enumerated information. After the enumeration in 1971, millions of W had to be collected from different corners of each state. Each of these slips had to be sorted and resorted so that the information could be taken out for preparing the tables. After tabulation, the 1971 census data were made available in the form of eight Tables described briefly as follows: A SERIES: General Population Tables (5 in number) showing the size of population, decidual variations in population since 1901, villages classified by population, towns classified by size and their growth rate since 1901 and particulars of the standard urban areas. B SERIES: Economic Tables (19 in number) giving various economic characteristics such as the industrial and occupational classification of workers cross-classified by age, educational level, and employment status for rural and urban areas separately. C SERIES: Social and Cultural Tables (10 in number) giving the composition of households, distribution of population by age, sex and educational levels, mother- tongue, and any subsidiary language, religion, and caste. 28 D S-IES: Migration Tables (6 in number) giving the break-up of migrants by place of birth and place of last residence, cross-classified by characteristics such as age, educational levels, duration of residence and marital status of the migrants. E SERIES: Establishment Tables (6 in number) giving the distribution of establishments of broad types such as manufacturing, trade or other establishments, size of employment, types of fuel or power used for industrial establishments, and household industries. F SERIES: Fertility Tables (5 in number) giving the distribution of married women, and the related births during the last one year, cross-classified by characteristics such as educational level, present age, age at marriage, duration of marriage, and religion. 6 SERIES: Special Tables of Degree Holders and Technical Personnel (12 in number) tabulating fairly detailed information on the distribution of the degree holders and technical personnel by subject fields, employment status, sector of employment, emoluments, the unemployed by duration of unemployment, and the level of education. H SERIES: Housing Tables (4 in number) giving the distribution of census houses by the uses they are put to,.material of wall and roof, size of households and the number of rooms occupied. 29 For all the states, except Himachal Pradesh, Jammu a ‘Kashmir, (Manipur, (Meghalaya, Nagaland, and Tripura, the tables are based on the processing of ten percent rural and twenty percent urban individual slips. Later on, the sample data were inflated to 100 percent according to the sampling procedure described below. In case of the above mentioned states, processing was done on a 100 percent basis because of the small population sizes of these states. SAIBLINGLAHD.SILICIIQH_BIQGIDEIIS3 From the individual slips relating to the rural parts of a district, a ten percent sample was selected systematically with a random start. For the urban parts of the district a twenty percent systematic sample was selected with a separate random start. The sample sizes were adopted on consideration of required precision, available resources, and operational convenience. The sample design adopted was a stratified systematic sample of the individual slips. For the purpose of sample selection, each district was divided into three sub-regions as follows: 1. Tehsil or police station (rural) 2. Non-city/Urban district 3. City (urban unit with population of 100,000 or more) 30 These sub-regions were termed operational units. If any of the operational units was unduly large it was divided into operational sub-units of at least 20,000 people. Similarly, small-sized operational units were either merged or added to a contiguous unit so as to form an "effective operational unit" of’aflminimum size. Adoption of these'operational units ensured accuracy'in the sample selection maintaining at the same time the stability in the sampling fraction within the tabulated areas. The procedure adopted ensured effective control over sample selection within each block as well as in the entire operational unit. For the purpose of sampling, the individual slips were sorted out according to eleven broad industrial categories and nine broad age groups for each sex. Thus, the individual slips of a particular operational unit were sorted out in (11)(9) = 99 cells for males and 99 cells for females separately. Sampling was done in each of the operational units independently by taking separate random starts. The slips were then distributed into ten (or five) pigeon-holes for rural (or urban) samples, and then one pigeon-hole was selected at randmm. The slips falling in the selected pigeon-hole constituted the sample. 31 W3 The tables prepared for each of the districts/states from the ten percent rural and twenty percent urban sample slips were inflated to give the final estimation tables. These estimated tables were compared with the population figures available from the primary census analysis tables which were prepared on the full count basis. This way of estimating the cell frequencies ensured column/row totals of the estimated tables tally with that of the full count tables. This is convenient to the user because after adjustment the identical column/row totals are found in tables having the same marginal specifications. The adjustment diminishes the sampling variance to some extent because greater the number of control variables, smaller the sampling variance of the adjusted frequencies. WW: An exact formula cannot be given for measuring the precision of the estimates in view of the systematic sampling design adopted. A rough estimate of the relative standard error on the basis of a simple random sample is given by fl(1-f)(1-p)/np], where f is the proportion sampled (0.10 in case of rural, and 0.20 in case of urban areas), p is the estimated proportion in any cell, n is the sample size. In the 1971 census, the individual slips 32 were sorted by sex, industrial category, and the age groups before sampling. It is expected that the efficiency would be considerably higher than a simple random sample of identical size. Extensive studies conducted have indicated that the systematic sampling of individuals is more efficient than simple random sampling, especially' when the individual slips were presorted for most of the characteristics of the migrants (Lahiri, 1971). Thus, the estimates prepared on the basis of systematic sampling in 1971 census can be treated as fairly reliable. W3 In the Indian census we are faced with several issues regarding the validity of the census data. Some of them are as follows: 1. The level of general literacy of India's population is very low. This reduces the level of response of the population census as well as that of the census taker. This is the basic drawback which makes it difficult for the Indian census bureau to make major improvements in the questionnaire from one census to another. The same reason prevents the Registrar General from adopting’more meaningful concepts, definitions, [methods and classificatory systems in the tabulation. The data on age, marital status, migration, language, and economic 33 activities are broadly, vaguely and approximately described and will not stand scrutiny except at certain levels of aggregation where large numbers pose further problems to the already defective data. 2. Another reason is the ad hoc and rather disheartening quality of the census taker. He is not paid for his work, is insufficiently trained, virtually forced to work for the government, and often no more than barely literate, drawn from a bewildering variety of assignments, few of which demand intellectual rigor. Many of the concepts and definitions so carefully formulated in the printed instructions escape the census taker's comprehension. As a result, the census authorities have to be content with only the broadest and crudest responses to the census schedule. The majority of the answers are given in very general terms, not lending themselves readily to the detailed sub-classification. 3. A lack of planning of computer capacity, time, and personnel adversely affect the validity of the census data. Consequently, a large number of data processing problems are solved in an ad hoc manner as and when they arise at particular time and place. It is not always possible to work out a common program even in the case of complex cross-tabulations. Therefore, the programmer is often left uninitiated in how certain parts of a 34 program should be developed differently from certain other parts. 4. The lack of sufficient funds to co.“ money and machinery well in advance to be commensurate with the task envisaged affects the validity of the data. 5. Lack of printing facilities straight from the computer print-out causes enormous delays in the publication of census reports. Though there is not much one can do to solve the problems of reliability and validity at this point, one can simply recognize the difficulties related to reliability, validity, and«other’weaknesses of the Indian census data. Keeping all these complexities in mind, we are using the data from the 1971 Census of India to study the relationship between migration and fertility. We will do the analysis separately for rural areas, the urban areas and rural a urban areas combined. W? Our .main goal is to study‘ the effect of migration on fertility. For this purpose we have designed three hypotheses. The dependent variable is the fertility rate (rural fertility rate, urban fertility rate, and rural a urban combined fertility rate). The independent variables are the migration rates (rural to urban migration rate, urban to rural migration rate, and rural 35 to urban & urban to rural combined migration rate). In addition, we will use the following variables as control variables: 1. 2. 3. 4. 5. Gender Literacy rate Percentage of women ever married Percentage of females in labor force Per capita income. The first hypothesis specifies the effect of rural to urban migration on urban general fertility. To test this hypothesis, we will calculate the following rates: 1. 2. 3. 4. 5. 7. 10. General fertility rate for urban areas. Rural to urban migration rate for males and females combined. Rural to urban migration rate for males only. Rural to urban migration rate for females only. Literacy rate in urban areas for males and females taken together. Literacy rate in urban areas for males only. Literacy rate in urban areas for females only. Percentage of women ever married in urban areas. Percentage of females in labor force in urban areas. Per capita income in urban areas. 36 Apart from studying the Product-Moment Correlation coefficients among these variables, we will also examine the partial correlations between urban general fertility and each of the variables 2, 3 and 4, controlling for variables 5 through 10, individually. For example, we will calculate the partial correlation coefficient between urban general fertility rate and rural to urban migration rate for both sexes taken together while controlling for literacy rate for urban areas for both sexes combined. Similarly, we will calculate the partial correlation coefficient between urban general fertility rate and rural to urban migration rate for both sexes taken together while controlling for urban literacy rate for males only, and so on. The second hypothesis studies the effect of urban to rural migration on the rural general fertility. To test this hypothesis, we will calculate the following rates: 1. General fertility rate for rural areas. 2. Urban to rural migration rate for males and females combined. 3. Urban to rural migration rate for males only. 4. Urban to rural migration rate for females only. 5. Literacy rate in rural areas for males and females taken together. 37 6. Literacy rate in rural areas for males only. 7. Literacy rate in rural areas for females only. 8. Percentage of women ever married in rural areas. 9. Percentage of females in labor force in rural areas. 10. Per capita income in rural areas. Apart from studying the Product-Moment Correlation coefficients among these variables, we will also examine the partial correlations between rural general fertility and each of the variables 2, 3, and 4, controlling'for variables 5 through 10, individually, For example, we will calculate the partial correlation coefficient between rural general fertility rate and rural to urban migration rate for both sexes taken together while controlling for literacy rate in rural areas for both sexes combined. Similarly, we‘ will calculate the partial correlation coefficient between rural general fertility rate and rural to urban migration rate for both sexes taken together while controlling for rural literacy rate for males only, and so on. Third hypothesis studies the effect of the combined urban to rural and rural to urban migration on the’general fertility of an entire state taking the rural and urban areas together. To test this hypothesis, we will calculate the following rates: 1. General fertility rate for the entire state. 38 2. Rural to urban and urban to rural combined migration rate for males and females taken together. 3. Rural to urban and urban to rural combined migration rate for males only. 4. Rural to urban and urban to rural combined migration rate for females only. 5. Literacy rate in the entire state for males and females taken together. 6. Literacy rate in the entire state for males only. 7. Literacy rate in the entire state for females only. 8. Percentage of women ever married in the entire state. 9. Percentage of females in labor force’in whole state. 10. Per capita income for the whole state. Apart from studying the Product-Mement Correlation coefficients among these variables, we will also examine the partial correlations between total fertility of a state and each of the variables 2, 3, and 4, controlling for variables 5 through 10, individually. For example, we will calculate the partial correlation coefficient between the'general fertility of a state and rural to urban a urban to rural combined migration rate for both sexes taken together while controlling for literacy rate in the state for both sexes combined. Similarly, we will calculate the partial correlation 39 coefficient between the general fertility rate of a state and rural to urban & urban to rural combined migration rate for both sexes taken together while controlling for general literacy rate in the whole state for males only, and so on. W3 Since the birth registration system in India is somewhat defective, the measure of fertility used in this study is the Child-Woman ratio, also called the fertility rate. Let us denote this rate by CUR. CWR is the number of children per 1,000 women of child- bearing age. This rate is obtained by dividing the total number of children who are less than ten years old by the total number of women who are in the age group 15-49 years, and multiplying this quotient by 1,000. CNR is a very useful index whenever vital registration data are lacking or defective. It reflects the fertility performance during the ten year period preceding the census. Another advantage of CWR is that it does not require a special question to be asked in the census questionnaire. In the calculation of CNR, the age group used for the children is 0-9 years rather than the customary 0-5 years age group. This is done so because in India, generally, there is an under-reporting of children in the 40 age group 0-4 years and over-reporting in the age group 5-9 years. This bias in reporting will cancel out by considering the age group 0-9 years. We have calculated CNR for urban areas, the rural areas, and the urban and rural areas combined as follows: Let UCNR denote the child-woman ratio or the fertility rate for urban areas only. UCNR is obtained by dividing the total number of children between the ages 0-9 years in the urban areas by the total number of women between the ages 15-49 years in urban areas and multiplying the quotient by 1,000. Let RCNR denote the child-woman ratio or the fertility rate for the rural areas. RCWR is obtained by dividing the total number of children in the rural areas between the ages 0-9 years by the total number of women in the rural areas between the ages 15-49 years, and multiplying the quotient by 1,000. Let SCNR denote the child-woman ratio or the fertility rate for the entire state taking the rural and urban areas together. SCWR is obtained by dividing the total number of children in the entire state between the ages 0-9 years by the total number of women in the entire state between the ages 15-49 years and multiplying the quotient by 1,000. 41 W: Themmigration considered here is the life-time migration based on the place of birth statistics. Any person whose place of birth was different from the place of residence at the time of census enumeration is considered a migrant. The rates of different streams of migration have been worked out for rural, urban, and the rural urban areas combined, separately for males, for females, as well as for both sexes taken together. These rates are calculated as follows: Total rural to urban migration rate for a state for both sexes combined is given by RUT = (X/Y)(1000) RUT denotes rural to urban migration rate for both sexes taken together: X is the number of those people who, at the time of 1971 census, were living in the urban areas of a state but were born in the rural areas of that state: Y is the total rural population of the state. Total rural to urban migration rate for the male population of a state is given by RUM = (X/Y)(1000) RUM is rural to urban migration rate for males; X is the number of those males who, at the time of 1971 census, were living in the urban areas but were born in rural areas of a state: Y is the total rural male population 42 of the state. Total rural to urban migration rate for female population of a state is given by mm = (x/YHiooo) RUF is rural to urban migration rate for females: X is the number of females who, at the time of 1971 census, were living in the urban areas but were born in the rural areas of a state: Y is the total rural female population of the state. Total urban to rural migration rate for a state for both sexes combined is given by URT = (X/Y)(1000) URT is urban to rural migration rate for both sexes combined: X is the number of people who, at the time of 1971 census, were living in rural areas but were born in the urban areas of a state; Y is the total urban population of the state. Total urban to rural migration rate for the male population of a state is given by URM = (X/Y)(1000) URM is urban to rural migration rate for males; X is the number of males who, at the time of 1971 census, were living in the rural areas of a state but were born in the urban areas of that state; Y is the total urban male population of the state. 43 Total urban to rural migration rate for the female population of a state is given by HR? = (X/Y)(1000) URF is urban to rural migration rate for females: X is the number of those females who, at the time of 1971 census, were living in rural areas but were born in urban areas of a state: Y is the total urban female population of the state. Total rural to urban and urban to rural migration rate for a state for both sexes combined is given by RUURT = (X/Y)(1000) RUURT is the rural to urban and urban to rural combined migration rate for both sexes taken together: X is the number of those persons who, at the time of I971 census, were enumerated at a place different from the place of their birth; Y is the total population of the state. Total rural to urban and urban to rural migration rate for a state for males only is given by RUURM = (X/YHIOOO} RUURM is rural to urban and urban to rural combined migration rate for males only; X is the number of males who, at the time of 1971 census, were enumerated at a place different from the place of their birth: Y is the total male population of the state. 44 Total rural to urban and urban to rural migration rate for a state for females is given by RUURF = (X/Y)(1000) RUURF is rural to urban and urban to rural combined migration rate for females only: X'is the number of those females who, at the time of 1971 census, were enumerated at a place different from the place of their birth; Y is the total female population of that state. (LIIIIA£X_IAIIB8 Literacy rate gives the number of literate per 1,000 population. We have calculated the literacy rates for males and females separately for the rural areas, urban areas, and the rural and urban areas combined for each state. The literacy rate for the entire population of a state, for rural and urban areas combined is given by SLRT = (X/Y)(1000) SLRT’is the literacy rate for a state for males a females combined and rural a urban areas combined; X is the total number of literate peopLe in a state; Y is the total population of the state. The literacy rate for the male population of a state for rural and urban areas combined is given by SLRM = (X/Y)(1000) 45 SLRM is the literacy rate in a state for males only for rural & urban areas combined: X is the total number of literate'males in a state: Y is the total male population of the state. The literacy rate for the female population of a state for rural and urban areas combined is given by SLRF = (X/Y)(1000) SLRF is the literacy rate for a state for females only for rural a urban areas combined; X is the total number of literate females in a state: Y is the total female population of the state. The literacy rate in the urban areas of a state for both males and females taken together is given by ULRT = (X/Y)(1000) ULRT is the literacy rate for urban areas of a state for males & females combined: X is the total number of literate people in the urban areas of a state: Y is the total urban population of the state. The literacy rate in the urban areas of a state for male population only is given-by um! = (x/YNIOOO) ULRM is the literacy rate for urban areas of a state for males only: X is the total number of literate males in the urban areas of a state: Y is the total population of males in urban areas of a state. 46 The literacy rate in the urban areas of a state for female population only is given by ULRF = (X/Y)(1000) ULRF is the literacy rate in the urban areas of a state for females only: X is the total number of literate females in the urban areas of a state: Y is the total population of females in urban areas of a state. The literacy rate in the rural areas of a.state for both sexes combined is given by RLRT = (X/Y)(1000) RLRT is the literacy rate in the rural areas of a state for males and females combined: X is the total number of literate people in the rural areas of a state: Y is the total rural population of the state. The literacy rate in the rural areas of a.state for males only is given by RLRM = (X/Y)(1000) RLRM is the literacy rate in the rural areas of a state for male population only: X is the total number of literate males living in the rural areas of a state: Y is the total rural male population of the state. The literacy rate in the rural areas of a.state for females only is given by RLRF = (x/Y)(looo) \‘l. lll'llltllll 47 RLRF is the literacy rate in the rural areas of a state for females only: X is the total number of literate females living in the rural areas of a state; Y is the total rural female population of the state. W: In the current investigation, we are interested in the incidence of early marriage only. In India, the lowest legal age for marriage for women is 19 years. In a way, it is the reco-ended' age for marriage. But. in the villages, many’marriages are performed prior to reaching the age of 19 years since there are no penalties for doing so. Therefore, by early marriage we mean the marriage performed in the age group 15-19. We do not need to worry about marriages before the age of 15 since such marriages are rare. The percentage of women ever married in the age group 15-19 years in a state in the rural and urban areas combined is given by pwswr = (X/YHIOOO) PWDIT is the percentage of women ever married in the rural and urban areas combined of a state: X is the total number of ever married females in a state who are 15-19 years old: Y is the total female population of the state in the age group 15-19 years. 48 The percentage of women over married in the age group 15-19 years in a state in the rural areas of a state is given by PNEMR = (X/YHIOOOJ PNEMR is the percentage of women ever married in the rural areas of a state: X is the total number of ever married females in a state who are 15-19 years old and live in the rural areas: Y is the total rural population of females in the age group I5-l9 years. The percentage of women in urban areas of a state in the age group 15-19 years in a state' is given by rwmu = (X/YHIOOO) PNEMU is the percentage of women ever married in the urban areas of a state: X is the total number of ever married females in a state who are 15-19 years old and live in the urban areas: Y is the total urban population of females in the age group 15-19 years. Labor force participation of women offers to them a higher degree of control over their own lives. A woman who can earn her own living is not entirely dependent on others, economically or psychologically. Female labor force participation in the rural areas in India includes participation by female workers in 49 industries and sectors such as cultivation, agriculture laborers, livestock, forestry, fishing, manufacturing and processing, construction, and other sectors such as working in village health clinics and schools. In the urban areas, females find employment in sectors such as: food, beverage, tobacco, textile, paper and printing, leather, chemical, cottage industries, public administration, medical and health, hotel services, sanitary services, and electricity, water a gas supply. The percentages of females in the labor force are computed by dividing the total number of females in the labor force (figures available from the 1971 census) by the total number of females in the state and multiplying this ratio by 100. The percentage of females in labor force in the entire state, that is, for both rural and urban areas combined is given by PFLFT = (X/Y)(1000) PFLFT is the percentage of females in labor force in a state for rural and urban areas combined: X is the total number of females in labor force in a state: Y is the total female population of the state. The percentage of females in labor force in the rural areas of a state is given by PFLFR = (X/Y)(1000) 50 PFLFR is the percentage of females in labor force in a state in rural areas only; X is the total number of females in a state in labor force who live in rural areas: Y is the total female population in the rural areas of a state. The percentage of females in labor force in the urban areas of a state is given by PFLFU = (X/Y)(1000) PFLFU is the percentage of females in labor force in a state in urban areas only: X is the total number of females in a state in labor force who live in urban areas: Y is the total female population in the urban areas of a state. We will make use of the correlation and partial correlation analysis methods in order to draw inferences regarding the association between migration and fertility in the rural, the urban, and the rural and urban areas combined, of a state. We have 16 variables under consideration in our analysis. These variables are as follows: 1a. Child-woman ratio, urban areas. 1b. Child-woman ratio, rural areas. 1c. Child-woman ratio, rural a urban areas combined. 2a. Rural to urban migration rate, males & females combined. 2b. 2c. 3a. 3b. 3c. 4a. 4b. 4c. 5a. 5b. 5c. 6b. 6c. 7a. 7b. 7c. 51 Urban to rural migration rate, males & females combined. Rural to urban and urban to rural migration rate, males and females combined. Rural to urban migration rate, males. Urban to rural migration rate, males. Rural to urban and urban to rural migration rate, males. Rural to urban migration rate, females. Urban to rural migration rate, females. Rural to urban and urban to rural migration rate, females. Literacy rate for urban areas, males and females combined. Literacy rate for rural areas, males and females combined. Literacy rate for rural a urban areas combined, and both sexes combined. Literacy rate for urban areas, males. Literacy rate Literacy rate males. Literacy rate Literacy rate for for for for rural areas, males. rural and urban combined, urban areas, females. rural areas, females. Literacy rate for rural and urban areas combined, 52 females. 8a. Percentage of women ever married, urban areas. 8b. Percentage of women ever married, rural areas. 8c. .Percentage of women ever married, rural and urban areas combined. 9a. Percentage of females in labor force, urban areas. 9b. Percentage of females in labor force, rural areas. 9c. Percentage of females in labor force, rural and urban areas combined. 10a. Per capita income, urban areas. 10b. Per capita income, rural areas. 10c. Per capita income, rural and urban areas combined. The dependent variables 1a, 1b, and 1c denote the child-woman ratio for the urban areas, the rural areas, and rural a urban areas combined, respectively. Variables 2a through 4c are the independent variables, and variables 5a through 10c are the control variables. we will first obtain three zero-order correlation matrices, one each for the urban, the rural, and urban a rural areas combined, and examine the intercorrela- tions, specifically looking at how the child-woman ratio correlates with the rest of the variables. Then, we will 53 calculate various partial correlations. The partial correlation coefficients are computed as follows: First, we take variable la, the urban general fertility, and correlate it with variable 2a, the rural to urban migration rate for the entire state for urban and rural areas combined and control for the variables which correspond to the urban areas only, namely, 5a, 6a, 7a, 8a, 9a, and 10a, only one at a time. Then we correlate variable 1a with variable 3a, the rural to urban migration rate for the male population of a state, controlling for the same variables, 5a, 6a, 7a, 8a, 9a, and 10a, individually. Similarly, we *will correlate variable 1a with the other variables, 2a through 4c, each time separately controlling for variables 5a, 6a, 7a, 8a, 9a, and 10a. Next, we correlate variable 1b, the child-woman ratio for rural areas of a state, with each of the variables 2a through 4c, by separately holding constant the 'variables 5b, 6b, 7b, 8b, 9b, and 10b, which correspond to the rural areas only. Similarly, calculate the partial correlations between variable 1c, the general fertility of entire state and each of the variables 2a through 4c, individually holding constant the variables 5c, 6c, 7c, 8c, 9c, and 10c, corresponding to the rural & urban areas 54 of a state combined. Controlling for only one variable at a time is necessitated by the relatively small sample size. In the present investigation the unit of analysis is a State and there are 25 states in India. Hence our sample size is 25 which is not large enough for simultaneous control of several variables. CHAPTER 3 RESULTS AND FINDINGS As described in Chapter 2, the separate analyses were done for rural areas, urban areas, and rural & urban areas combined. Fertility is measured by the child-woman ratio. Migration is measured by life-time migration based on place of birth statistics. A person whose place of birth was different from the place of his/her residence at the time of enumeration, is considered a migrant. Tables 1 through 6 on the following pages contain the results of the analysis. These results and findings will be used to examine my hypotheses. Tables 1, 3, and 5 contain three 16 by 16 zero-order correlation matrices. Table 1 corresponds to urban data, table 2 to rural data and table 3 to combined data. Only the upper triangles of the matrices are shown, and the diagonals of 1's have been suppressed. Descriptions of all the variables in each table are given at the. end of the table. Tables 2, 4, and 6 contain the’partial correlation coefficients corresponding to each of the three hypotheses (urban, rural & combined data, respectively). 55 56 As a background to this analysis, however, I want to present first some general descriptive statistics related to the variables used. W3 Descriptive statistics for various migration rates are given in the following table. For variable labels and descriptions, refer to Table 1: Variabl e Mean Standard Minimum Maximum Name Deviation RUT 45 23 15 94 RUM 46 19 20 99 RUF 48 24 18 87 URT 76 33 24 150 URM 61 24 23 121 URF 87 33 24 182 RUURT 42 23 18 95 RUURM 37 20 16 84 RUURF 49 26 19 106 The range of each of the migration rates is large, indicating there are considerable interstate variations in the migration rates. Rural to urban migration rate, combined for both sexes (RUT), ranges from a low of 15 migrants per 1,000 rural population to a high of 94 migrants per 1,000 rural population, with a mean of 45 and a standard deviation of 23. Rural to urban migration rate for males only (RUM), ranges from a low of 20 male migrants per 1,000 rural male population to a high of 99 male migrants per 1,000 rural male population, with a mean of 46 and a standard deviation 57 of 19. Rural to urban migration rate, for females only (RUF), ranges from a low of 18 female migrants per 1,000 rural female population to a high of 87 female migrants per 1,000 rural female population, with a mean of 48 and a standard deviation of 24. Urban to rural migration rate, combined for both sexes (URT), ranges from a low of 24 migrants per 1,000 urban population to a high of 150 migrants per 1,000 urban population, with a mean of 76 and a standard deviation of 33. Urban to rural migration rate, for males only (URM), ranges from a low of 23 male migrants per 1,000 urban male population to a high of 121 male migrants per 1,000 urban male population, with a mean of 61 and a standard deviation of 24. Urban to rural migration rate, for females only (URF) ranges from a low of 24 female migrants per 1,000 urban female population to a high of 182 female migrants per 1,000 urban female population, with a mean of 87 and an SD of 33. Rural to urban and urban to rural combined migration rate, for both sexes combined (RUURT), ranges from a low of 18 migrants per 1,000 population of a state to a high of 95 migrants per 1,000 population of the state, with a mean of 42 and a standard deviation of 23. Rural to urban and urban to rural combined migration rate, for males only (RUURM), ranges from a low of 16 58 male migrants per 1,000 male population of a state to a high of 84 male migrants per 1,000 male population, with a mean of 37 and a standard deviation of 20. Rural to urban and urban to rural combined migration rate, for females only (RUURF) ranges from a low of 19 female migrants per 1,000 female population of a staterto a high of 106 female’migrants per 1,000 female population, with a mean of 49 and a standard deviation of 26. It can be seen from above that, while rural to urban migration rates are similar for male and female migrations, in the urban to rural migration, the female rate is higher. The data about the fertility for rural areas of various states of India show considerable variations in the general fertility rates as represented by child- woman ratios. These fertility rates for rural areas range from 1.08 to 1.71 with a mean of 1.36 and a standard deviation of .16. The corresponding fertility rates for the urban areas of various states, range from 1.02 to 1.41 with a mean of 1.19 and standard deviation of .12. When urban and rural areas are combined and the entire state is taken as a unit, the general fertility rates range from 1.05 to 1.68 with mean and standard deviation equal to 1.31 and .16 respectively. The interstate variations also exist in the combined rates 59 for other socio-economic indices. W The first hypothesis predicts a negative association between overall urban fertility and various migration variables. The zero-order and partial correlations for this purpose are presented in Tables 1 and 2 respectively. We would expect the correlation coefficients between urban fertility rate and various migration rates to be negative and significant. The correlation coefficients in Table 1 reveal a negative association between urban general fertility rate and the rural to urban migration rate for males, females, as well as for both sexes combined. All the coefficients shown in bold face are significant at least at .05 level of significance. The first entry in row 1 of this table is -.68. This is :12 indicating that the correlation between variable 2 (rural to urban migration rate for both sexes combined, RUT) and variable 1 (urban fertility rate, UCNR) is -.68. This correlation coefficient is significant at .001 level of significance. The second entry in row 1 is -.74. This is the correlation r13 between variable 3 (rural to urban migration rate for males only, RUM) and variable 1 (urban fertility rate, UCNR) -.74, and this correlation coefficient is significant at .001 level. The last entry 60 in the first row of Table 1 is -.06. This indicates that the correlation between variable 16 (urban per capita income, PCIU) and variable'1 (urban fertility rate, UCNR) is r1,15= -.06, and this is not a significant correlation. Similarly, the very last entry in Table 1 (last entry in column 16) is .32. This indicates that the correlation between variable 16 (urban per capita income, PCIU) and variable 15 (percent of females in labor force in urban areas, PFLFU) is r1545 = .32, and that the higher the percent of females in labor force, higher the per capita income in urban areas. This coefficient is significant at .10 level of significance. A visual inspection of the coefficients in row'l of Table 1 indicates that the urban fertility rate (UCWR) is significantly related to all, but two, of the variables. Urban fertility rate does not seem to correlate strongly with variables 15 (percent of females in labor force in urban areas, PFLFU) and 16 (per capita income in urban areas, PCIU). Moreover, urban fertility rate (UCVIR) correlates negatively with all variables except variable 14, the percent of women ever married in urban areas (PNEMU). The first-order partial correlation coefficients are shown in Table 2. This table has 9 rows and 6 columns. The first entry in first row of this table is -.55. This means that the partial correlation 61 coefficient between variable 1 (urban fertility rate, UCWR) and variable 2 (rural to urban migration rate, RUT) controlling for variable 11 (urban literacy rate for both sexes combined, ULRT), is -.55, while the zero-order correlation coefficient between variable 1 (urban fertility rate, UCIVR) and variable 2 (rural to urban total migration rate, RUT) was found to be -.68 in Table 1. This partial correlation is significant at .05 level. The second entry in row 1 of table 2 is -.51, indicating that the partial correlation between variable 1 (urban fertility rate, UCWR) and variable 2 (rural to urban migration rate for both sexes combined, RUT) while controlling for variable 12 (urban literacy rate for males only, ULRM) is -.51. This is significant at .05 level . The corresponding zero-order correlation in Table 1 was found to be -.68. Similarly, the last entry in row 1 is -.68, which is the partial correlation between variable 1 (urban fertility rate, UCNR) and variable 2 (rural to urban migration rate for males and females combined, RUT) controlling for variable 16 (per capita income in the urban areas, PCIU), and this is significant at .01 level. The corresponding zero-order correlation coefficient between variables 1 and 2 was found to be - .68. This indicates that the urban per capita income has little or no effect on the relationship between urban 62 fertility (UCWR) and rural to urban migration (RUT). The first entry in last row of Table 2 is -.48. This is the partial correlation coefficient between variable 1 (urban fertility rate, UCWR) and variable 10 (rural to urban & urban to rural combined migration rate for females only, RUURF) controlling for variable 11 (urban literacy rate for males and females combined, ULRT). This partial correlation coefficient is signifi- cant at .05 level. From Table 1 we notice that the zero- order correlation between variables 1 and 10 (9th entry in row 1) is also -.48 indicating that urban literacy rate does not affect the relationship between urban fertility rate and the rural to urban a urban to rural migration rate for females (RUURF). Similarly, the last entry in last row of Table 2 is -.49 which is the partial correlation coefficient between variable 1 (urban fertility rate, UCNR) and variable 10 (rural to urban & urban to rural migration rate for females, RUURF) while holding variable 16 (urban per capita income, PCIU) constant, compared to the zero-order correlation of -.48 between variables 1 and 10 found in Table 1. The level of significance for this coefficient is .05. This analysis lends support to our first hypothesis that there is a positive correlation between rural to urban migration and fertility in a state. 63 TABLE 1: PARTIAL CORRELATION COEFFICIENTS CORRELATION ULRT ULRM 5 ULRF PWEMU PFLFU PCIU VARIABLES r(1,2) -.55 -.51 -.57 -.59 -.71 _-.68 r(1,3) -.62 -.60 -.64 -.69 -.78 -.75 r(l,4) -.60 -.57 -.61 -.62 -.75 -.72 r(1,5) -.48 -.44 -.51 -.52 -.55 -.56 r(1,6) -.55 -.53 -.57 -.59 -.65 -.65 r(1,7) -.34 -.23 -.39 -.36 -.27 -.28 r(1,8) -.48 -.35 -.54 -.38 -.50 -.46 r(1,9) -.48 -.36 -.54 -.40 -.55 -.49 r(l,10) -.48 -.37 -.53 -.38 -.49 -.49 The variables are defined as follows: (1) UCWR : child-woman ratio, urban areas. (2)“ RUT : rural to urban migration rate, males and females. (3) RUM : rural to urban migration rate, males. (4) RUF : rural to urban migration rate, females. (5) URT : urban to rural migration rate, males and females. (6) URM : urban to rural migration rate, males. (7) URF : urban to rural migration rate, females. (8) RUURT : rural to urban and urban to rural migration, males and females combined. (9) RUURM : rural to urban & urban to rural migration rate, males. (10) RUURF : rural-urban and urban-rural migration rate, females. (11) ULRT : urban literacy rate, males and females combined. (12) ULRM : urban literacy rate, males. (13) ULRF : urban literacy rate, females. (14) PWEMU : percent of women ever married, urban areas. (15) PFLFU : percent of females in labor force, urban areas. (16) PCIU : per capita income, urban areas. 64 TABLE 2: PARTIAL CORRELATION COEFFICIENTS CORRELATION ULRT ULRM ULRF PWEMU PFLFU PCIU VARIABLES F(1,2) -e55 -e51 -e57 ‘e59 '—e71 -e68 l‘(l,3) -e62 "u60 -e64 -e69 ’e78 -e75 r(194) -e60 -e57 -e61 "'e62 -e75 -e72 r(1,5) -.48 -.44 -.51 -.52 -.55 -.56 I‘(1,6) "e55 -e53 'e57 -e59 "'e65 "e65 r(1,7) -e34 -e23 ‘e39 -e36 "e27 "ezs [(1,8) -e48 -e35 "es‘ -e38 -e50 “e46 l‘(1,9) -e48 -e36 -e54 "e40 “e55 -e‘9 I‘(1,10) “'e48 -e37 "e53 -e38 -e49 ‘e‘g The variables are defined as follows: (1) UCWR : child-woman ratio, urban areas. (2) RUT : rural to urban migration rate, males and females. (3) RUM : rural to urban migration rate, males. (4) RUF : rural to urban migration rate, females. - (5) URT : urban to rural migration rate, males and females. (6) URM : urban to rural migration rate, males. (7) URF : urban to rural migration rate, females. (8) RUURT : rural to urban and urban to rural migration, males and females combined. (9) RUURM : rural to urban & urban to rural migration rate, males. (10) RUURF : rural-urban and urban-rural migration rate, females. (11) ULRT : urban literacy rate, males and females combined. (12) ULRM : urban literacy rate, males. (13) ULRF : urban literacy rate, females. (14) PWEMU : percent of women ever married, urban areas. (15) PFLFU : percent of females in labor force, urban areas. (16) PCIU : per capita income, urban areas. 65 The second hypothesis predicts a negative association between overall rural fertility and various migration variables. The zero-order and partial correlations for this purpose are presented in Tables 3 and 4 respectively. We expect the correlation coefficients between rural fertility rate and various migration rates to be negative and significant. The correlation coefficients in Table 3 reveal a negative association between rural general fertility rate and the urban to rural migration rate for males, females, as well as for both sexes combined. All the correlations which are significant at least at the .05 level are shown in bold face in these tables. The first entry in row 1 of this table is -.32. This is r", and it indicates that the correlation between variable 2 (rural to urban migration rate for both sexes combined, RUT) and variable 1 (rural fertility rate, RCWR) is -.32. This is significant at .10 level. The second entry in first row is -.49. This is denoted by r13, and it indicates that the zero-order correlation between variable 3 (rural to urban migration rate for males, RUM) and variable 1 (rural fertility rate, RCNR) is -.49 which is significant at .05 level. Let us look at the effect of urban to rural migration for both sexes combined, URT, on the overall 66 rural fertility, RCNR. The 4th entny in first row of Table 3 is -.38. This is the correlation coefficient between variable 1 (rural fertility rate, RCIVR) and variable 5 (urban to rural migration rate for both sexes combined, URT) and it is significant at .10 level. Similarly, the 5th entry in row 1 of table 3 is -.47 which is the correlation between rural fertility rate and urban to rural migration rate for males (URM). This is significant at .05 level. The last entry'in the first row of Table 3 is -.18. This indicates that the correlation between variable 16 (rural per capita income, PCIR) and variable 1 (rural fertility rate, RCWR) is rLJ¢= -.18. This is not significant at .10 level. Similarly, the very last entry in Table 3 (last entry in column 16) is -.13. This indicates that the correlation between variable 16 (rural per capita income, PCIR) and variable 15 (percent of females in labor force in rural areas, PFLFR) is rlglg = -.13. This is also not a significant association at .05 level. A visual inspection of the coefficients in row 1 of Table 3 indicates that the rural fertility rate (RCNR) is moderately significantly related to all, but two, of the variables. Rural fertility rate does not seem to correlate strongly with variables 15 (percent of females in labor force in rural areas, PFLFR) and 16 (per capita income in rural areas, PCIR). Moreover, urban fertility 67 rate (RCWR) correlates negatively with all variables except variable 14, the percent of women ever married in rural areas (PNEMR). The first-order partial correlation coefficients are shown in table 4. This table has 9 rows and 6 columns. The first entry in first row of this table is -.13. This means that the partial correlation coefficient between variable 1 (rural fertility rate, RCVIR) and variable 2 (rural to urban migration rate, RUT) controlling for variable 11 (rural literacy rate for both sexes combined, RLRT), is -.13, and this coefficient is not significant at .10 level. The zero-order correlation coefficient between variable 1 (rural fertility rate, RCVIR) and variable 2 (rural to urban total migration rate, RUT) was found to be -.32 in Table 3. This shows that rural literacy rate significantly affects the association between rural fertility (RCNR) and rural to urban migration rate (RUT). The second entry in row 1 of Table 4 is -.11, which indicates that the partial correlation between variable 1 (rural fertility rate, RCWR) and variable 2 (rural to urban migration rate for both sexes combined,.RUT) while controlling for variable 12 (rural literacy rate for males only, RLRM) is -.11, which is not significant at the .10 level. The corresponding zero-order correlation of -.32, indicating 68 that the literacy rate of males significantly affects the association between rural fertility and the rural to urban migration rate. Similarly, the last entry in row 1 of Table 4 is -.31, which is the partial correlation between variable 1 (rural fertility rate, RCNR) and variable 2 (rural to urban migration rate for males and females combined, RUT) controlling for variable 16 (per capita income in rural areas, PCIR). This coefficient is significant at .10 level. The zero-order correlation coefficient between variables 1 and 2 was found to be - .32. This indicates that the urban per capita income has almost no effect on the relationship between rural fertility (RCHR) and rural to urban migration (RUT). Now let us examine the partial correlations between rural fertility and urban to rural migration rates. The first entry in row 4 of table 4 is -.26. This partial correlation between variable 1 (rural fertility rate, RCNR) and variable 5 (urban to rural migration rate for males and females combined, URT) controlling for variable 11 (rural literacy rate for both sexes combined, RLRT) is weak and not significant at .10 level. The corresponding zero-order correlation from Table 3 is - .38, suggesting that rural literacy rate does affect the relationship between rural general fertility and urban to rural migration rate. The last entry in row 4 is -.42 69 indicating that the partial correlation between variable 1 (rural fertility rate, RCWR) and variable 5 (urban to rural total migration rate, URT) holding variable 16 (rural per capita income, PCIR) constant is -.42. This is significant at .10 level. The corresponding zero-order correlation from table 3 is .38 indicating that rural per capita income does not influence the association between rural fertility and urban to rural migration rate. Similarly, the first coefficient in last row of Table 4 is -.25, and it is not significant at .10 level. This is the’partial correlation between variable 1 (rural fertility rate, RCNR)-and variable 10 (rural to urban and urban to rural combined migration rate for females only, RUURF) controlling for variable 11 (rural literacy rate for males and females combined, RLRT). From Table 3 we notice that the zero-order correlation between variables 1 and 10 (9th entry in row 1) is -.31 indicating that rural literacy rate does affect the relationship between rural fertility rate and the rural to urban a urban to rural migration rate for females (RUURF). Similarly, the last entry in last row of Table 4 is -.28 which is the partial correlation coefficient between variable 1 (rural fertility rate, RCHR) and variable 10 (rural to urban & urban to rural migration rate for females, RUURF) while holding' variable’ 16 (rural per capita. income, PCIR) 70 constant. This partial correlation coefficient is not significant at .10 level. The zero-order correlation of - .31 between variables 1 and 10 in Table 3 indicates that rural per capita income does not affect the association between rural fertility and rural to urban a urban to rural migration rate for females. With regards to the effect of urban to rural reverse migration on the rural fertility level, the zero- order correlation coefficients in‘Table 34reveal that the urban to rural migration rates for males, females and both sexes combined, are slightly negatively related to overall rural fertility. These correlations remain unchanged after the per capita income is held constant. Therefore, our second hypothesis that 'the higher the rate of urban to rural migration in a state, the lower its overall rural fertility level ' gets support from the results of this analysis. 71 TABLE 4: PARTIAL CORRELATION COEFFICIENTS CORRELATION RLRT RLRM RLRF PWEMR PFLFR PCIR VARIABLES r(1,2) -.13 -.11 -.13 -.17 -.33 -.31 I'(1,3) .‘e31 -e3o -e29 -e34 -.50 -e49 I'(I,4) -e16 -e15 "e15 -e19 -e37 -e36 r(1,5) -e26 -025 -026 "e41 “"038 “-.42 r(1,6) -e33 —e32 -e31 -e36 -e47 -e‘9 r(1,7) -.17 -.16 -.16 -.33 -.25 -.30 I'(1,8) "e33 -e31 -e34 -e33 “e40 -e38 r(l,9) -e36 -e34 -e37 -e35 -e‘7 -e‘z ‘ r(1,10) -.25 -.23 -.25 -.23 -.3o -.28 The variables are defined as follows: (1) RCWR : child-woman ratio, rural areas. (2) RUT : rural to urban migration rate, males & females. (3) RUM : rural to urban migration rate, males only. (4) RUF : rural to urban migration rate, females only. (5) URT : urban to rural migration rate, males and females. (6) URM : urban to rural migration rate, males only. (7) URF : urban to rural migration rate, females only. (8) RUURT: rural to urban and urban to rural migration, males and females combined. (9) RUURM: rural to urban & urban to rural migration rate, males. (10) RUURF: rural-urban and urban-rural migration rate, females. (11) RLRT : rural literacy rate, males and females combined. (12) RLRM : rural literacy rate, males only. (13) RLRF : rural literacy rate, females only. (14) PWEMR: percent of women ever married, in rural areas. (15) PFLFR: percent of females in labor force, rural areas. (16) PCIR : per capita income, rural areas. | Q I . I . . . . 5 u). 72 TABLE 4: PARTIAL CORRELATION COEFFICIENTS CORRELATION RLRT RLRM RLRF PWEMR PFLFR PCIR VARIABLES r(1,2) -.13 -.11 -.13 -.17 -.33 -.31 r(1,3) -.31 -.30 -.29 -.34 -.50 -.49 r(1,4) -.16 -.15 -.15 -.19 -.37 -.36 r(1,5) -.26 -.25 -.26 -.41 -.38 -.42 r(1,6) -.33 -.32 -.31 -.36 -.47 -.49 r(1,7) -.17 -.16 -.16 -.33 -.25 -.30 r(1,8) -.33 -.31 -.34 -.33 -.40 -.38 r(1,9) -.36 -.34 -.37 -.35 -.47 -.42 r(1,10) -.25 -.23 -.25 -.23 -.30 -.28 The variables are defined as follows: (1) RCWR : child-woman ratio, rural areas. (2) RUT : rural to urban migration rate, males & females. (3) RUM : rural to urban migration rate, males only. (4) RUF : rural to urban migration rate, females only. (5) URT : urban to rural migration rate, males and females. (3) URM : urban to rural migration rate, males only. (7) URF : urban to rural migration rate, females only. (8) RUURT. rural to urban and urban to rural migration, males and females combined. (9) RUURM: rural to urban & urban to rural migration rate, males. (10) RUURF: rural-urban and urban-rural migration rate, females. (11) RLRT : rural literacy rate, males and females combined. (12) RLRM : rural literacy rate, males only. (13) RLRF : rural literacy rate, females only. (14) PWEMR: percent of women ever married, in rural areas. (15) PFLFR: percent of females in labor force, rural areas. (16) PCIR : per capita income, rural areas. 73 The third hypothesis predicts a negative association between overall fertility in a state (rural and urban fertility' combined) and various migration variables. The zero-order and partial correlations for this purpose are presented in tables. 5 and 6 respectively. We expect the correlation coefficients between general fertility rate of an entire state (rural and urban areas combined) and various migration rates to be negative and significant. The correlation coefficients in table 5 reveal a negative association between general fertility rate of a state and all the migration variables. The first entry in row 1 of table 5 is -.50. This is r" and it indicates that the correlation between variable 2 (rural to urban migration rate for both sexes combined,‘RUT9 and variable 1 (state fertility rate, SCWR) is -.50. This coefficient is significant at .01 level. The second entry In first row is -.58. This is denoted by r”, and it represents the correlation between variable 3 (rural to urban migration rate for males, RUM) and variable 1 (state fertility rate, SCWR). This is significant at .01 level. Let us look at the effect of urban to rural migration for both sexes combined, URT, on the overall state fertility, SCWR. The 4th entry in first row of table 5 is -.40. This is the correlation coefficient 74 between variable 1 (state general fertility rate, SCWR) and variable 5 (urban to rural migration rate for both sexes combined, URT). This is not significant at .05 level. Similarly, the correlation between state fertility rate and urban to rural migration rate for males (URM) is -.51. This is indicated by 5th entry in row 1 of table 5. This coefficient is significant at .01 level. The last entry in the first row of Table 5 is -.30. This indicates that the correlation between variable 16 (state per capita income, PCIT) and variable 1 (state fertility rate, SCWR) is r1J¢= -.30, which is not significant at .05 level. Similarly, the very last entry in Table 3 (last entry in column 16) is -.47. This indicates that the correlation between variable 16 (state per capita income, PCIT) and variable 15 (percent of females in labor force in the entire state in both rural and urban areas combined, PFLFT) is r15," = -.47, and this association is significant at .05 level. The 7th entry in row 1 of table 5, is the correlation coefficient between the state general fertility and rural to urban a urban to rural combined migration for both.sexes taken together. This correlation is equal to -.46 and it is significant at .05 level. Similarly, the 9th entry in row 1 is -.47. This is the correlation between state general fertility and rural to 75 urban a urban to rural combined migration rate for females only and it is significant at .05 level. A visual inspection of the coefficients in row 1 of Table 5 indicates that the state general fertility. rate (SCWR) is very significantly related to all of the variables 2 through 16. Moreover, the state general fertility rate (SCWR) correlates negatively with all the variables except variable 14, the percent of women ever married in the entire state (PWEMT). The first-order partial correlation coefficients are shown in Table 6. This table has 9 rows and 6 columns. The first entry in first row of this table is -.21. This means that the partial correlation coefficient between variable 1 (state fertility rate, SCWR) and variable 2 (rural to urban migration rate, RUT) controlling for variable 11 (state literacy rate for both sexes combined, and both rural and urban areas combined, SLRT), is -.21, while the zero-order correlation coefficient between variable 1 (state general fertility rate, SCWR) and variable 2 (rural to urban total migration rate, RUT) was found to be -.50 in Table 5. This means that the partial correlation between state general fertility and rural to urban migration rate is not significant and that the state literacy rate significantly affects this association. The second entry 76 in row 1 of Table 6 is -.18, indicating that the partial correlation between variable 1 (state fertility rate, SCWR) and variable 2 (rural to urban migration rate for both sexes combined, RUT) while controlling for variable 12 (state literacy rate for males only, SLRM) is -.18 and is not significant at .05 level. The corresponding zero- order correlation from table 5 is -.50 indicating that the literacy rate of males in a state significantly affects the association between state fertility and the rural to urban migration rate. Similarly, the last entry in row 1 of Table 6 is -.45, which is the partial correlation between variable 1 (state fertility rate, SCWR) and variable 2 (rural to urban migration rate for males and females combined, RUT) controlling for variable 16 (per capita income in the entire state, PCIT). The zero-order correlation coefficient between variables 1 and 2 was found to be -.50. This indicates that the per capita income of a state has little effect on the relationship between state fertility (SCWR) and rural to urban migration (RUT). Now let us examine the partial correlations between state general fertility and urban to rural migration rates. The first entry in row 4 is -.28. This is the partial correlation between variable 1 (state general fertility rate, SCWR) and variable 5 (urban to 77 rural migration rate'for'males and females combined, URT) controlling for variable 11 (state literacy rate for both sexes combined, and rural & urban areas combined, SLRT). The corresponding zero-order correlation from Table 5 was -.40. The last entry in row 4 is -.45 indicating that the partial correlation between variable 1 (state general fertility rate, SCWR) and variable 5 (urban to rural total migration rate, URT) holding variable 16 (state per capita income, PCIT) constant is -.45 as compared to the zero-order correlation of -.40. Similarly, the first coefficient in last row of table 6 is -.32. This is the partial correlation between variable 1 (state general fertility rate, SCWR) and variable 10 (rural to urban and urban to rural combined migration rate for females only, RUURF) controlling for variable 11 (state literacy rate for males and females combined, SLRT). From table 5 we notice that the zero-order correlation between variables 1 and 10 (9th entry in row 1) is also -.47 indicating that state literacy rate does affect the relationship between state fertility rate and the rural to urban & urban to rural combined migration rate for females (RUURF). Similarly, the last entry in last row of table 6 is -.40 which is the partial correlation coefficient between variable 1 (state fertility rate, SCWR) and variable 10 78 (rural to urban & urban to rural combined migration rate for females, RUURF) while holding variable 16 (state per capi ta income, PCIT) constant, compared to the zero-order correlation of -.47 between variables 1 and 10. From tables 5 and 6, we notice that the general fertility of a state is significantly correlated with rural to urban migration. This association is significant at 5 percent level of significance. The association between state general fertility rate and the urban to rural migration rate is negative, but weak. The state general fertility is, however, significantly correlated with the combined rural to urban & urban to rural migration rate in the state. This association is found significant at 5 percent level of significance. Similarly, the combined rates of rural to urban and urban to rural migration streams for males is significantly associated with state general fertility. These effects remain significant even after holding the per capita income constant. Therefore, it is asserted that higher the combined rate of rural to urban and urban to rural migration, lower the overall fertility of a state. 79 TABLE 6: PARTIAL CORRELATION COEFFICIENTS (STATE GENERAL FERTILITY, MIGRATION AND SOCIO-EOONOMIC VARIABLES) VARIABLES HELD CONSTANT CORRELATION SLRT SLRM SLRF PWEMT PFLFT PCIT VARIABLES (11) (12) (13) (14) (15) (16) r(1,2) -.21 -.18 -.21 -.34 -.49 -.45 r(1,3) -.31 -.28 -.32 -.45 -.58 -.60 r(1,4) -.22 -.19 -.22 -.35 -.53 -.49 r(1,5) -.28 -.26 -.29 -.46 -.40 -.45 r(1,6) -.35 -.33 -.35 -.44 -.53 -.56 r(1,7) -.13 -.10 -.14 -.28 -.18 -.24 r(1,8) -.36 -.32 -.37 -.39 -.45 -.42 r(1,9) -.40 -.36 -.41 -.42 -.47 -.47 r(1,10) -.32 -.30 -.32 -.39 -.51 -.40 The variables are defined as follows: (15) (16) SCNR RUT RUM RUF URT URM URF RUURT RUURM RUURF SLRT SLRM SLRF PNEMT PFLFT PCIT co to ee 0e ee 0e Co Co child-woman ratio, rural rural rural urban urban urban rural males rural to to to to to to to urban urban urban rural rural rural urban migration migration migration migration migration migration and urban the whole state. males and females. males only. females only. males and females. rate, males only. rate, females only. to rural migration, rate, rate, rate, rate, and females combined. to urban & urban to rural migration rate, males. rural-urban and urban-rural migration rate, state literacy rate, males and females combined. state literacy rate, males only. state literacy rate, females only. percent of women ever married, entire state. percent of females in labor force, entire state. per capita income in the state as a whole. females. 80 TABLE 6: PARTIAL CORRELATION COEFFICIENTS (STATE GENERAL FERTILITY, MIGRATION AND SOCIO-ECONOMIC VARIABLES) VARIABLES HELD CONSTANT CORRELATION SLRT SLRM SLRF PWEMT PFLFT PCIT VARIABLES (ll) (12) (13) (14) (15) (16) r(1,2) -.21 -.18 -.21 -.34 -.49 -.45 r(1,3) -.31 -.28 -.32 -.45 -.58 -.60 r(1,4) -.22 -.19 -.22 -.35 -.53 -.49 r(1,5) -.28 -.26 -.29 -.46 -.40 -.45 r(1,6) -.35 -.33 -.35 -.44 -.53 -.56 r(1,7) -.13 -.10 -.I4 -.28 -.18 -.24 r(1,8) -.36 -.32 -.37 -.39 -.45 -.42 r(1,9) -.40 -.36 -.41 -.42 -.47 -.47 r(1,10) -.32 -.30 -.32 -.39 -.51 -.40 The variables are defined as follows: (1) SCWR : child-woman ratio, the whole state. (2) RUT : rural to urban migration rate, males and females. (3) RUM : rural to urban migration rate, males only. (4) RUF : rural to urban migration rate, females only. (5) URT : urban to rural migration rate, males and females. (6) URM : urban to rural migration rate, males only. (7) URF : urban to rural migration rate, females only. (8) RUURT : rural to urban and urban to rural migration, males and females combined. (9) RUURM : rural to urban A urban to rural migration rate, males. (10) RUURF : rural-urban and urban-rural migration rate, females. (11) SLRT : state literacy rate, males and females combined. (12) SLRM : state literacy rate, males only. (13) SLRF : state literacy rate, females only. (14) PWEMT : percent of women ever married, entire state. (15) PFLFT : percent of females in labor force, entire state. (16) PCIT : per capita income in the state as a whole. CHAPTER 4 SMART AND DISCUSS ION An attempt has been made in this dissertation to explore the relationship between migration and fertility in India on the basis of data from 1971 Census of India with state as the unit of analysis. Both rural to urban and urban to rural migration were considered as well as migration no matter what the origin or destination. The first finding coming out this research is that there is an inverse relationship between rural to urban migration rates and the urban fertility levels. However, the impact of migration depends not only on the rate of migration, but also on various characteristics of migrants. The relation is stronger for urban areas where the urban general fertility and various migration rates are all significantly negatively correlated at .01 level of significance. It is widely assumed that in the developing countries, overcrowding in agriculture and poor living conditions in the rural areas push people out of the 81 82 villages towards the cities. Since, these people generally have a higher fertility, their influx to the urban areas inflates the overall urban fertility levels. But, in India, migrants from rural areas are generally the most active, most dynamic and most enterprising and they belong to higher socio-economic and educational groups. The rate of unemployment among the migrants has also been found to be less than the urban natives, and the migrants in the cities are economically better off than the original residents of the city. Thus, rural to urban migration in India is primarily of a "pull" rather than "push" type. The main indicator of the social status of these migrants at the place of their origin has been their family background, but at the place of destination the social position and prestige are determined by occupational achievement, income, consumption‘standards, and educational attainment. Since the children are perceived as a disadvantage to their social and economic well being, parents try to restrict the number of children thry produce. Furthermore, the.most perceptible effect of migration is the breakdown of joint family which has resulted in freeing the young migrant couples from the traditional family fertility norms. The means 83 to limit the number of children are more easily and readily available in the urban than in the rural areas. In the face’of higher cost and higher standards of living in the cities, the migrants as well as the original residents are more amenable to family planning practice. Furthermore, rural to urban migration in India is highly selective of young single people. These people try to delay their marriages and those who are married at the time of migration usually leave their families behind. This is evident from the extremely low sex ratios in the urban areas of India. These delayed marriages and husband-wife separations depress the fertility of the migrants. It appears that all these factors tend to lower the fertility of the rural to urban migrants to the levels considerably lower than those of urban natives so as to significantly affect the overall urban fertility. Our second finding is that rural general fertility is lower in states which have higher rural to urban migration rates and in states which have higher urban to rural migration rates. The former may seem to conflict with common sense. If rural to urban migration in India.is highly selective, then common sense reasoning would suggest that the drift of the enlightened people from rural areas should raise rather than lower the 84 overall rural fertility levels since these people are in the prime period of fertility as they are young, and they are economically more sound and can afford to have more children. But the things are not as simple as they look on the surface. The migration situation in India is very complex. One of the important features of Indian migration is a very high degree of return migration. The life time migration statistics of the Indian census do not provide the magnitude of return migration, but we can assume that the states which have high rates of rural to urban migration should also have higher return migration rates. These return migrants not only have lower fertility themselves but also spread these norms in the rural areas. They usually return with new ideas, attitudes and values acquired during their stay in the cities and the people at their native place often seek their advice and listen to them and follow their advice with respect. Also, the other migrants who do not permanently return, maintain close kinship ties with their native place and often make visits and bring fresh ideas from the cities. These permanent returnees and temporary visitors prove as catalysts of change, and it is possible that the family welfare workers seek their 85 help and assistance in moulding the behavioral patterns of the villagers. Furthermore, in contrast to those who in the first instance migrate alone and subsequently take their families, there might be many who would like to raise their children in the villages. These women with comparatively lower fertility contribute to the overall lowering of rural fertility levels. These are some of the possible explanations for the inverse relationship between both rural to urban and urban to rural migration and rural general fertility. The relationships, however, are not very strong because the number of return migrants, as a proportion of the total rural population, is very small. The fertility patterns of the Indian population are quite different in rural and urban areas of a state. In general, families in the urban areas desire as well as have a smaller number of children than the ones in the rural areas. Studies in other settings have shown that migrants to urban areas have somewhat higher fertility than natives but lower fertility than rural stayers. In the rural areas, out-migrants and returned-migrants have lower fertility than non-migrants. Moreover, migrants tend to marry at a later age and separation has the effect of lowering their completed family size. 86 Increasing migration, therefore, is likely to reduce fertility as well as overall rate of population growth in the country. So, the states with high rate of internal migration will exhibit overall lower fertility rate of its residents. Education, both of males and females, is a strong factor in reducing fertility. From Table 1 we see that literacy' is ‘very' significantly’ correlated ‘with fertility rate in the urban areas. The correlation is weaker in rural areas. Therefore, mass education needs to be emphasized with special interest in health and family planning. The increased public investment in education would bring about reduced fertility. In order to decide the number of children both in the urban and rural areas, parents do pay attention to the advantages and disadvantages of having children. In the urban areas the educational and income opportunities are generally greater, health facilities are better and easily available, and the communication channels are relatively more effective. Therefore, more urban than rural people adopt the patterns of lower fertility. Some researchers have also shown that the act of migration itself causes a decline in fertility and therefore, contributes to the reduction in the over-all 87 fertility. Therefore, our analysis suggests that in India, the states with larger population movements have lower fertility as compared to those having smaller movements. There is a strong possibility that internal migration in India may have undergone a structural change in the context of the development activities of the Five Year Plans, with greater importance than before for urban-to-urban and rural-to-rural migration. This can be related to a shift of the urban labor force from unorganized to organized industries in urban areas and the rural labor force to sites of large scale construction projects in the rural areas. There is a possibility of population movement from smaller to bigger towns without altering the proportion of the population that is urban. However, the available data do not always seem to support this hypothesis (Agarawala, S., 1968). In order to get a clearer picture of the population situation caused by migration, it is necessary to have some knowledge about the qualitative aspects of internal migration in addition to knowing its quantitative trends. A closer examination of the Indian Census data suggests that there are several ways in which 88 the internal migration in India can influence the existing social framework. Because of the social framework within which it is occurring, migration has a peculiar significance in India. This has reference to the closer ties which the migrants maintain with their places of origin (Ea-es, E., 1967). It has encouraged return migration which is quite widespread in India. The economic significance of migration in India is low especially because of the preponderance of women among migrants over short distances. For most of these women marriage is the sole reason for migration. Among men the major reason for migration is a lack of job opportunities in the villages. Thus, obligatory and non-economic reasons cause females to leave the village, while voluntary and economic reasons are prominent in the out- migration of males from villages. Although the total volume of migration comprises only about 3.5 percent of the population, the absolute number is very substantial. The net inflow of migrants has been largest in the states of West Bengal and Maharastra, followed by Madhya Pradesh and Punjab. These states also have an above average per capita income. But states like Karnataka, Kerala, and Himachal Pradesh also have per capita income quite close to the 89 national average, while the out-migration rates in these states are not very high. Therefore, there does not appear to be any close relationship between low per capita income and high rate of out migration. It emphasizes the lack of importance of the "push" factor as a cause of internal migration. It is true that in- migrants find better employment opportunities and often reaeh a higher standard of living in the in-migrating state. People of Northern India have a greater tendency'to move out of their home state than those from other parts of India. This is evidenced by the fact that the proportion of out-migration to total population is very high in Punjab, Rajasthan and Himachal Pradesh. On the other hand, the inhabitants of Eastern India, those of Assam, Orissa, and West Bengal are far less prone to move out of their native places. About South India, the ratio of out-migrants to the total population is higher than the all India figures for Kerala and Karnataka, while that for Andhra and Tamil Nadu is lower. Among the people of Wetsern India, Gujratis are more likely to move out of the home state for better opportunities than people from Maharashtra. Thus, differences in climate, social outlook, and educational level have induced 90 people to react differently to almost similar economic conditions about leaving their home states in search of better employment and higher income. It may be that the. people of North India have a relatively higher achievement motivation than those of the East. Looking at the per capita income of states (Table 8, Appendix II) it is seen that the five states, namely, Delhi, Maharashtra, West Bengal, Punjab and Gujrat have the highest per capita income and they absorb about 50 percent of the inter-state migrants in India. But five comparatively poor states, namely, Bihar, Rajasthan, Orissa, Madhya Pradesh and Andhra Pradesh received less than 30 percent of the in-migrants. Hence, it might be said that the general tendency of the in-migrants is to move into the relatively prosperous regions of the country. Therefore, as we mentioned before, it is reasonable to hold that the "pull" factor has been more powerful in giving momentum to the internal migration in India. In other words, the economic conditions are primarily responsible for migration. In general terms, internal migration has an impact on migrants as well as residents. The most important impact on the migrants is the creation of nuclear rather than joint families. This reduces the role 91 of joint families as economic institutions. Along with this disintegration of the joint family system, family background makes less of a contribution to social status. From the standpoint of controlling population growth, free internal migration should be encouraged because the policies to reduce rural to urban migration are more likely to increase the rate of population growth for the nation as a whole unless steps will be taken simultaneously to introduce lower fertility norms in the rural areas and to reduce the rural-urban fertility differentials. But since the rural-urban fertility differentials are largely due to the disparities in economic and social development between the two areas, the thrust of population policies ought to be on narrowing these disparities. This can be achieved only if population policies are considered as an essential component of country's overall development strategy. APPENDICES APPENDIX I INDIAN CENSUS: AN OVERVIEW W A population census is an administrative operation organized by a government to count the heads and collect various personal informations of every man, woman, and child in a country with reference to a specified date. The practice of census-taking dates back to ancient times. The objectives of the early censuses, however, were very limited. But in modern times, the need for reliable data on different characteristics of the population is felt for the purposes of demarcation, policy making, economic planning, health, education, and other fields. Thus, modern census assumed the character of’a scientific statistical operation to collect accurate data. India took its first regular*modern census in 1872. The next census took place nine years later in 1881 and since then a census has been conducted regularly every ten years. The census of 1971 marked a hundred years of census-taking in India. It needs to be mentioned that the process of tabulating and printing data in India is very very slow and it takes about 10 years from the date of 92 93 census-taking before. the data become available for further analysis. The 1971 census had for the first time used electronic computer facilities. But a modern computer technology without a matching printing technology has created a situation which has no parallel to it. The census organization has a few computers but no printing press at its con-and and because of the bureaucratic procedures, the census authorities have to depend almost wholly on government printing presses. Any census report or table, which is printed ten years after the census enumeration, is almost sure to be shelved by the reader in anticipation of the next census. The Indian census is a fascinating blend of intellectual curiosity, encyclopaedic scholarship, administrative skills, imperial policy and the wit and whim of Indian masses. Today census is taken for granted and very'few'people would question the need for a census. However, this was not true a hundred years ago when the first systematic census was launched. People were naturally curious to know why the government wanted to collect all kinds of data about each and every person in the country. What they dreaded most was the imposition of taxes. Surprisingly enough, even after hundred years of census-taking in India, it is the experience of field 94 workers in villages that the most common misgiving about any census or sample survey is the fear of imposition of taxes. From this point of view, it would appear that a span of 100 years has not made much difference in the attitude of the people towards statistical enquiries imposed on them by the government or any other agency. Perhaps it is a manifestation of the stagnation of India's economy and the poverty of the rural masses who live in constant fear of a further deterioration in their depressed level of living. Besides, a deep-distrust of the. 'outsider' is a persistent phenomenon of Indian rural social life. W We discussed important issues regarding the methodology in chapter 2. Here, we will further comment on the censuses conducted .after India. attained independence, that is, censuses of 1951, 1961, and 1971. The 1951 census removed the question on caste from the census questionnaire_and limited it to scheduled caste and scheduled tribes because of constitutional requirements. The government of India took this decision of not to collect data because it was feared that collection of such data tended to make people more and more caste conscious. In the absence of data on caste, 95 the crusaders for caste have been disarmed and this is indeed a happy development. The 1951 census took an important decision not to cross-tabulate’all important tables by religion as was the practice in the British days. There was only one table on religion and it gave just the distribution of population by different religions in India. The question on religion has only heightened religious animosities and has helped certain political parties which keep on harping on the increasing proposition of certain minorities in the total population. Since India's elections are not based on religion, there is no need to collect such data on religion and in a secular State there is no need to collect data on religion of every individual. The major contribution of the three censuses in Post-Independence India is really in the fields of internal migration, urbanization, and fertility. For the first time in the history of census, the 1951 census gave the rural-urban breakdown in all the tables. In the earlier censuses, there were some detailed tables for a few selected cities in India, but there was no way of getting a complete rural-urban breakdown. 96 The 1961 census introduced a new sub-question on the place of birth, namely "whether rural or urban". This brought about a tremendous improvement in the analysis of data on internal migration based on the place of birth data. As a result, for the first time in the history of census of India, it was possible to identify the different migration streams, namely, rural to rural, rural to urban, urban to urban, and urban to rural. The 1971 census has introduced another sub- question which will give direct evidence on migration without referring to the place of birth. The question asked was "place of last residence". There was also a new question on the place of work which will throw light on commutation, especially to the big cities. Therefore, in the post-independence period, the censuses have emphasized economic data. Today the census is regarded as an essential tool for planning and policy- making, while during the British period, the census emphasized caste, tribe, religion, and language, and the census data, in combination with British diplomacy and imperial strategy, were used to foster disharmony through the policy of "divide and rule". Conducting a census every ten years in a country like India is a very difficult task. Even with 97 a sufficiently long period of enumeration (three weeks), counting every man, woman, and child in a vast country is a stupendous task. Thus, Indian census in considered to be one of the greatest administrative operations in the world. A good deal of planning, preparation, and organization is necessary to ensure proper coverage and accuracy in such a huge operation. The first preparatory step in Indian census is to find out every human habitation whether a village, town or a remote forest or hilly area and then to locate every house in such habitation. For this purpose the census was preceded by a housing census. First, a complete and exhaustive list of all towns and villages and other human settlements is prepared. The villages, towns and settlements are then formed into small areas called blocks covering, on an average, 120 to 150 households. A census worker, often a primary school teacher, called the enumerator is placed incharge of each such block. He goes round his block and prepares a detailed map showing the locations of all the houses in it. He then paints a number on each house on his block. This procedure which is called house numbering is necessary to avoid any possible omissions of a house in a block. The enumerator then collects some simple 98 information about each house and household, like material of wall and roof, and number of members in the household. Such housing census is conducted well before the final population count. The housing census is an essential preliminary step towards the final count. It not only gives useful information about the housing conditions of the country, but also provides rough estimates of population before the final census. It is with the help of the maps prepared and the informations gathered in the housing census that the enumerator takes up the more important task of actually counting and collecting personal information on each person. For the 1971 census, this enumeration was done all over India in March-April, 1971. During this phase, the enumerator visited every household and recorded the responses to various questions on each man, woman, and child in a form called Individual Slip. He also recorded the information of the houseless persons. The Individual Slip contained 17 questions which are reproduced below: (1) Name (2) Sex (3) Age (4) Marital status (6) For currently married women only: (7) (8) (9) (10) (11) (12) (13) (l4) (15) (16) 99 a. Age at marriage b. Any child born in the last one year Birth place: a. Place of birth b. Rural/Urban c. District d. State/Country Last residence: a. Place of last residence b. Rural/Urban c. District d. State/Country Duration of residence at the village or town of enumeration. Religion Scheduled caste or scheduled tribe Literacy Educational level Mother tongue Other languages Main activity: a. Broad category; whether worker or non-worker (with description of type of worker or non-worker) 100 b. Place of work c. Name of establishment d. Nature of industry, trade, profession or service. e. Description of work f. Class of worker (17) Secondary work: a. Broad category; whether worker or non-worker (with description of type of worker or non-worker). b. Place of work c. Name of establishment d. Nature of industry, trade, profession or service. e. Description of work f. Class of worker Question 6 regarding fertility covers only currently married women and does not cover widows or divorced persons. This has been done keeping in mind the convenience of canvassing as to evoke the most reliable responses. The responses to this question when tabulated according to variables such as age of mother, duration of marriage, educational level, and religion, are expected to yield useful data on current fertility. 101 Questions 7 and 8 relating to migration are expected to yield useful tabulations with reference to place of birth as well as place of last residence. The rural/urban flows are likely to be better reflected with reference to place of birth. Questions 7, 8 and 9.relate to the birth place, last residence and duration of residence at the village/town of enumeration, respectively. Answer to Question 7 had to be filled with reference to the place of birth of the persons enumerated and the particulars recorded under the four sub-items of this question. Enumerators followed the following instructions: Question 7 (a): Place of birth Write 'PL' for a person born in the village or town where he is being enumerated. Where 'PL' is noted against this question, put X against items b, c, a d . For those born outside the village or town of enumeration write the actual name of the place against sub-item (a) and fill the other details against sub-items b,c and d. Question 7 (b): Rural/Urban For those born outside the village or town of enumeration ascertain if the‘place of birth was a village or town at the time of person's birth. To enable a person to determine whether the place was a town or a village 102 the enumerator may be required to indicate the status of the place of birth in comparison with a known town in the neighborhood of the place of enumeration. You may mention some important urban characteristics to enable the person to make out if the place of his birth was rural or urban. Question 7 (c): District For a person born outside the village or town of enumeration but within the district of enumeration, write '0'. For a person born in another district of the State of enumeration, write the name of the district. If the person cannot name the district, write 'not known'. Question 8: Last residence Answer to this question had to be filled in respect of every'person if he had another place of normal residence irrespective of his place of birth, before he came to the present place where he is enumerated. Even if a person were born at the place of enumeration, but by the nature of his work or for studies, he had shifted subsequently to another village or town and had come back again to the place of enumeration, he should be deemed to have had another place of residence prior to his residence here. The last previous residence is relevant only if he had been outside the village or town of 103 enumeration and not simply in another house or locality in the same place. Question 9: Duration of residence at the place of enumeration Note here the period of the existing continuous residence, in completed years, in the village or town where the person is being enumerated. This will apply even to a person born at the place of enumeration. If he had left this village or town and had lived elsewhere for sometime, then the duration of residence to be noted against this question is the period of the last continuous residence. But if a person had been away on a temporary visit or tour, that should not be taken as a break in the period of his continuous residence here. The economic Questions 16 and 17 have made some basic departure from the past. Each person was required to declare what his/her main activity was, that is, how he engaged himself mostly. A person was treated as a worker for his main activity if he participated in any economically productive work by his physical or mental 'activity. No specific question on unemployment was asked at the 1971 census because the previous experience had shown that unemployment and underemployment required enquiries in depth to get a realistic estimate and this 104 was not possible in a quick massive operation such as the census. The correct recording' of answers to these questions in the Individual Slip, however, called for an army of trained personnel. All throughout the country the census information was collected'uniformlya This required long hours of training for the census workers to learn to ask, explain and record answers to the census questions in a uniform manner throughout the country. For such a training and supervision a hierarchy of census personnel had to be built up. At the lowest level were the enumerators who were appointed mainly from among the primary school teachers and government office staff. There were nearly half a million enumerators for the final enumeration task. All the enumerators, supervisors, and superintendents were under the overall supervision of the Deputy Co-issioner of a district. The head of the census organization for the whole country was the Registrar General of India. But merely an army of census officials, however well-trained and organised, does not make a census. The cooperation of the millions of people to be enumerated is equally important. It is only when the person enumerated furnishes the answers to the questions 105 honestly and truthfully the census workers collect correct and accurate data. There is a census law which gives legal powers to the census officials to put the census questions to the people. It also requires all citizens to answer the questions truthfully and prescribes penalties for violations of these legal requirements. The census law also guarantees that the personal information collected in the course of a census will be kept strictly confidential and cannot be used as evidence even in a court of law. Fortunately, Indian census has always owed its vsuccess to the willing cooperation of the citizens on the one hand and dedicated services rendered by the vast army of the census workers on the other. The Indian census is not content with providing merely skeletal statistical picture of the population of the country and its administration units. It presents detailed analytical reports which shed considerable light on the various facets of India's life and economy, and therefore, census is the most fruitful single source of information on the people of the country. W In India the only main source providing migration data is the Census. Census organization in 106 India has been collecting data on migration since the first systematic census was conducted in 1881. Prior to India's independence, need for memeographic studies was never felt by the government or scholars. Moreover, the internal migration in India was very much limited. There were no large scale developments, the means of transport were limited, and the masses were orthodox and had the belief that "it is better to have bread without butter at home place than to shift to outside normal place of residence for getting bread and butter". The Census Organization, therefore, continued to collect and tabulate the limited data on the place of birth till 1931. In 1951, data were processed for smaller units also, that is, migration within and between districts was also presented. On account of the influx of refugees from Pakistan more detailed data on the displaced persons for the year of migration were also collected and published. Large scale industrial programs throughout the country and specially the introduction of five-year development plans attracted many persons to the places of industrial development for better employment. It was greatly responsible for migration from rural to urban areas. The 1961 census took care of it and recorded rural-urban classification from thelplace of birth in addition to the 107 names of place, district and the State for those having born in India but at a place other than the place of enumeration. Though 1971 census was rightly commended for providing the internal migration data, it had the following drawbacks: For the purpose of Table D-I, a migrant is a person who has been enumerated at a place other than his place of birth. In the remaining tables, the term migrant refers to a person who had his last residence at any place outside the village or town of enumeration. Migration, whether based on place of birth or last residence, includes all types of movements whether temporary, seasonal or long-term, arising from a variety of social and economic causes like employment, education, and marriage. During the 1971 census enumeration, persons who were away from their normal residence throughout the enumeration period were enumerated at places where they were found during the enumeration period. Though small in numbers, such accidental cases were also included among migrants. In India, due to the cultural reasons, most of the womenlgo to their parents during pregnancy and remain there for about 3-6 weeks after the delivery of the child. On account of the lack of medical facilities a 108 good number of women from the villages go to the nearby towns having a hospital or maternity health center where they remain for a few days. Information based on the 'Place of birth' of child born in such cases inflates the figures of’migration as it is not the real migration. After retirement or due to some other reasons, most people return to their original home villages/towns which are'generally their places of birth even though for a good period of their life they have lived at another place. Migration in such cases is hidden. Those born outside the place of enumeration were asked to state whether the place of birth was rural or urban at the time of their birth. DEFINITIONS OF WORKERS AND NON-WORKERS: In the 1971 census, a worker was defined as a person whose main activity was participation in any economically productive work through his physical or mental activity. Work involved not only actual work but also the effective supervision and direction of work. In case of regular work in trade, profession, service or business, a person was considered as a worker if he had participated in such regular work on any one of the days during the week preceding the date of enumeration and that work had been returned as his main activity. A person participating in 109 regular work, but absent from his work during the last week preceding the date of enumeration owing to illness, travel, holiday, strike, or temporary breakdown, was treated as a worker. In case of work such as cultivation, livestock, plantations, or some type of household industries, a person was treated as a worker after ascertaining whether he had participated in any work at any time during the last one year, even if he was not economically active in the week prior to the enumeration. These were mainly the seasonal workers. Persons under training as apprentices, with or without a stipend, were recorded as workers. A public or social worker engaged actively in public or social work or a political worker actively engaged in furthering the political aims of a party were also regarded as a workers. Under-trials or persons who were in jail but were not convicted by a court of law were recorded as workers if they were engaged in any of the activities which were considered as work by definition, before thewaere‘arrested. Similarlyy persons who were admitted to the hospitals or sanatoria were considered as workers if they were engaged in activities which were considered as work before they were admitted. A person who had secured an employment or was offered a job but had not actually joined duty was not 110 treated as a worker. A person whose primary duties were the household duties such as cooking for one's own household, or a boy or girl who was primarily a student attending an institution, even if such a person helped in the family's economic activity but not as a full-time worker, was not treated as worker for the’main activity. A person who merely received income such as rent or pension and who did .not have to work for receiving that income, was not treated as worker unless he’ had returned some economic nactivity' and if that activity was returned as the’ main activity of the individual. Non-workers are those who are not mainly engaged in any economically productive work. They comprise seven broad groups: (1) Those basically engaged in unpaid household duties and doing no other work, (2) Full-time students, (3) Retired persons, not employed again in full-time work, (4) All dependents such as infants, children attending schools, and disabled, (5) Beggars, vagrants, (6) Inmates of penal or mental institutions, (7) All non-workers who may not come under any of the above six categories, but are seeking work. APPENDIX II DATA FROM 1971 CENSUS OF INDIA 111 TABLE 1: DATA ON AGE DISTRIBUTION FOR COMPUTING FERTILITY RATES TOTAL POPULATION -NAME OF AGE T=TOTAL PERSONS MALES FEMALES STATE GROUP R=RURAL U=URBAN ANDHRA ALL T 43502708 22008663 21494045 PRADESH R 35100181 17698247 17401934 U 8402527 4310416 4092111 0-9 T 12380457 6202403 6178054 R 10079298 5041212 5038086 U 2301159 1161191 1139968 15-49 T. 20299394 10255448 10043946 R 16091067 8059598 8031469 U 4208327 2195850 2012477 ASSAM ALL T 14957542 7885064 7072478 R 13630561 7126453 6504108 U 1326981 758611 568370 0-9 T 5025780 2509087 2516693 R 4668093 2325849 2342244 U 357687 183238 174449 15-49 T 6497882 3503617 2994265 R 5818693 3088346 2729347 U 679189 414271 264918 BIHAR ALL T 56353369 28846944 27506425 R 50719403 25728987 24990416 U 5633966 3117957 2516009 0-9 T 17117138 8757333 8359805 R 15540742 7943902 7596840 U 1576396 813431 762965 15-49 GUJRAT ALL 15-49 HARYANA ALL 15-49 HIMACHAL ALL PRADESH 0-9 15-49 JAMMU ALL 6 KASHMIR 0-9 awe a”a awe one awe cwfi awe awe awe awe awe awe 112 25702320 22901858 2800462 26697475 19200975 7496500 7948223 5909990 2038233 12243150 8529203 3713947 10036808 8263849 1772959 3176215 2685962 490253 4251228 3411141 840087 3460434 3218544 241890 972792 916146 56646 1566912 1432789 134123 4616632 3758411 858221 1392309 1152734 239575 12918843 11315714 1603129 13802494 9842483 3960011 4098030 3041878 1056152 6317190 4326000 1991190 5377258 4420225 957033- 1682890 1423881 259009 2229962 1771669 458293 1766957 1628623 138334 491130 461868 29262 775430 695893 79537 2458315 1996864 461451 710919 586713 124206 12783477 11586144 1197333 12894981 9358492 3536489 3850193 2868112 982081 5925960 4203203 1722757 4659550 3843624 815926 1493325 1262081 231244 2021266 1639472 381794 1693477 1589921 103556 481662 454278 27384 791482 736896 54586 2158317 1761547 396770 681390 566021 115369 KARNA- TARA KERALA MADHYA PRADESH MAHARA- SHTRA 15-49 ALL 15-49 ALL 15-49 ALL 15-49 ALL awe axe awe awe c”ha awe one am“ awe ema awe awe 113 2133353 1709420 423933 29299014 22176921 7122093 8579652 6631460 1948192 13350686 9822476 3528210 21347375 17880926 3466449 5731377 4853933 877444 10158943 8430786 1728157 41654119 34869352 6784767 13262380 11264276 1998104 18544224 15303410 3240814 50412235 34701024 15711211 14656252 10638544 4017708 1137797 906488 231309 14971900 11249209 3722691 4316084 3332568 983498 6871757 4981134 1890623 10587851 8852350 1735501 2899362 2454863 444499 4970165 4103986 866179 21455334 17823411 3631923 6718583 5690199 1028384 9600727 7819563 1781164 26116351 17482020 8634331 7413252 5364606 2048646 995556 802932 192624 14327114 10927712 3399402 4263568 3298874 964694 6478929 4841342 1637870 10759524 9028576 1730948 2832015 2399070 432945 5188778 4326800 861978 20198785 17045941 3152844 6543797 5574077 969720 8943497 7483847 1459650 24295884 17219004 7076880 7243000 5273938 1969062 15-49 MANIPUR ALL 0-9 15-49 MEGHA- ALL LAYA 0-9 15-49 NAGALAND ALL 15-49 ORISSA ALL awe awe awe awe axe awe awe awe awe and awe awe 114 23654230 15352340 8301890 1072753 931261 141492 313051 274730 38321 491443 422725 68718 1011699 864529 147170 316200 278478 37722 472384 394335 78049 516449 465055 51394 134652 123025 11627 258163 224206 33959 21944615 20099220 1845395 6471668 5944836 526832 12371231 7610083 4761148 541675 470231 71444 157563 138338 19225 248801 213709 35092 520967 441533 79434 158796 139947 18849 244739 200070 44669 276084 241171 34913 67914 62119 5795 141559 115838 25721 11041083 10041023 1000060 3195612 2928696 266916 11282999 7742257 3540742 531078 461030 70048 155488 136392 19096 242642 209016 33626 490732 422996 67736 157404 138531 18873 227645 194265 33380 240365 223884 16481 66738 60906 5832 116604 108368 8238 10903532 10058197 845335 3276056 3016140 259916 PUNJAB RAJAS- SIKKIM TAMIL NADU 15-49 ALL 15-49 ALL 15-49 ALL 15-49 ALL and awe awe awe awe awfi awe awe awe awe awe awe 115 9927163 8999116 928047 13551060 10334881 3216179 3696235 2866446 829789 6110876 4522158 1588718 25765806 21222045 4543761 8043472 6710458 1333014 11410919 9287555 2123364 209843 190175 19668 53407 48119 5288 108164 97477 10687 41199168 28734334 12464834 10810687 7621182 3189505 5011714 4485180 526534 7266515 5533475 1733040 1956833 1519421 437412 3229803 2373888 855915 13484383 11060995 2423388 4179270 3483642 695628 5925182 4776995 1148187 112662 100959 11703 25501 22741 2760 62516 55704 6812 20828021 14438727 6389294 5456933 3842388 1614545 4915449 4513936 401513 6284545 4801406 1483139 1739402 1347025 392377 2881073 2148270 732803 12281423 10161050 2120373 3864202 3226816 637386 5485737 4510560 975177 97181 89216 7965 27906 25738 2528 45648 41773 3875 20371147 14295607 6075540 5353754 3778794 1574960 TRIPURA UTTAR PRADESH WEST BENGAL ANDAMAN NICOBAR ISLANDS 15-49 ALL 15-49 ALL 15-49 ALL 15-49 ALL axe one and one a”8 awe and awe awe awe awe awe 116 20422475 14010007 6412468 1556342 1393982 162360 477620 436805 40815 678048 598182 79866 88341144 75952548 12388596 26105403 22626315 3479088 39689644 33779057 5910587 44312011 33344978 10967033 13210502 10718256 2492246 20422878 14509003 5913875 115133 88915 26218 32284 26328 5956 10270480 6954771 3315709 801126 717227 83899 239942 219398 20544 349593 307347 42246 47016421 40214012 6802409 13724165 11914955 1809210 20768206 17444062 3324144 23435987 17173552 6262435 6580691 5300403 1280288 11194319 7626226 3568093 70027 53195 16832 16309 13281 3028 10151995 7055236 3096759 755216 676755 78461 237678 217407 20271 328455 290835 37620 41324723 35738536 5586187 12381238 10711360 1669878 18921438 16334995 2586443 20876024 16171426 4704598 6629811 5417853 1211958 9228559 6882777 2345782 45106 35720 9386 15975 13047 2978 15-49 ARUNACHAL ALL PRADESH CHANDI- GARE . DELHI GOA DAMAN 8 D10 15-49 ALL 15-49 ALL 15-49 ALL awe awe axe awe awe awe awe use one and a”8 use 117 64234 47734 16500 467511 450223 17288 133727 129826 3901 240201 228712 11489 257251 24311 232940 62952 6580 56372 146937 12498 134439 4065698 12498 134439 1066371 131146 935225 2112916 188686 1924230 857771 630997 226774 224982 169573 55409 42983 31412 11571 251231 239369 11862 67335 65243 2092 134289 125759 8530 147080 14444 132636 33260 3548 29712 87077 7749 79328 2257515 7749 79328 559368 70204 489164 1204314 103717 1100597 431214 310909 120305 114552 86165 28387 21251 16322 4929 216280 210854 5426 66392 64583 1809 105912 102953 2959 110171 9867 100304 29692 3032 16660 59860 4749 55111 1808183 4749 55111 507003 60942 446061 908602 84969 823633 426557 320088 106469 110430 83408 27022 PONDI- CHERY 15-49 ALL 15-49 awe ema awe awe 118 418434 297372 121062 471707 273419 198288 131573 77974 53599 223946 128218 95728 215023 147743 67280 237112 137783 99329 66377 39158 27219 111257 63517 47740 203411 149629 53782 234595 135636 98959 65196 38816 26380 112686 64701 47988 119 TABLE 2: DATA FOR COMPUTING RURAL TO URBAN MIGRATION RATES (AREA OF BIRTH = RURAL) STATE ENUMERATED IN URBAN AREAS OF BIRTH PERSONS MALES FEMALES ANDHRA PRADESH 2041299 967997 1073302 ASSAM ' 200253 111777 88476 BIHAR 1268304 625042 643262 GUJRAT 1467511 697425 770086 HARYANA 194870 89294 105576 HIMACHAL PRADESH 61294 38225 23069 JAMMU 8 KASHMIR 68650 33598 35052 RARNATAKA 1224317 599615 624702 RERALA 513324 216408 296916 MADHYA PRADESH 1133182 526081 607101 MAHARASHTRA 3258664 1752357 1506307 MANIPUR 12223 5512 6711 MEGHALAYA 15278 7143 8135 NAGALAND 9316 5175 4141 ORISSA 515508 264883 250625 PUNJAB 505274 239965 265309 RAJASTHAN 652492 282208 370284 SIRKIM 1743 1013 730 TAMIL NADU 2281196 1081535 1199661 TRIPURA 6135 3083 3054 UTTAR PRADESH 2040150 998643 1041507 WEST BENGAL 690564 336201 354363 ANDAMAN 395 210 185 ARUNACHAL 2605 1488 1117 CHANDIGARH 1196 829 367 DELHI 8546 3675 4871 GOA-DAMAN 47166 21159 26007 PONDICHERY 8452 3782 4670 120 TABLE 3: DATA FOR COMPUTING URBAN TO RURAL MIGRATION RATES. (AREA OF BIRTH = URBAN) STATE ENUMERATED IN RURAL AREAS OF BIRTH PERSONS MALES FEMALES ANDHRA PRADESH 768410 282930 485480 ASSAM 97990 44640 53350 BIHAR 371682 80832 290850 GUJRAT 508074 175297 332777 HARYANA 87191 24490 62701 HIMACHAL PRADESH 35591 14329 21262 JAMMU 6 KASHMIR 47901 18449 29452 RARNATARA 613246 235168 378078 RERALA 382330 165120 217210 MADHYA PRADESH 508597 189023 319574 MAHARASHTRA 1139960 446273 693687 MANIPUR 11816 4787 7029 MEGHALAYA 8366 5002 3364 NAGALAND 1253 818 435 ORISSA 176840 67391 109449 PUNJAB 163280 52000 111290 RAJASTHAN 343232 94710 248522 SIKKIM 621 426 196 TAMIL NADU 937733 359041 578692 TRIPURA 11388 5663 5725 UTTAR PRADESH 852491 195023 657438 WEST BENGAL 247293 99233 148060 ANDAMAN 1719 910 809 ARUNACHAL 2735 1634 1101 CHANDIGARH 217 131 86 DELHI 12627 4947 7680 GOA DAMAN 33942 14589 19353 PONDICHERY 6551 3042 3509 121 TABLE 4: DATA FOR COMPUTING URBAN TO RURAL AND RURAL TO URBAN MIGRATION RATES (PLACE OF BIRTH : RURAL OR URBAN) STATE ENUMERATED IN RURAL OR URBAN OF BIRTH AREAS PERSONS MALES FEMALES ANDHRA PRADESH 2809709 1250927 1558782 ASSAM 298243 156417 141826 BIHAR 1639986 705874 934112 GUJRAT 1975585 872722 1102863 HARYANA 282061 113784 168277 HIMACHAL PRADESH 96885 52554 44331 JAMMU 6 KASHMIR 116551 52047 64504 RARNATARA 1837563 834783 1002780 KERALA 895654 381528 514126 MADHYA PRADESH 1641779 715104 926675 MAHARASHTRA 4398624 2198630 2199994 MANIPUR 24039 10299 13740 MEGHALAYA 23644 12145 22499 NAGALAND 10569 5993 4576 ORISSA 692348 332274 360074 PUNJAB 668564 291965 376599 RAJASTHAN 995724 376918 618806 SIKKIM 2364 1439 925 TAMIL NADU 3218929 1440576 1778353 TRIPURA 17525 8746 8779 UTTAR PRADESH 2892641 1193666 1698945 WEST BENGAL 937857 435434 502423 ANDAMAN 2114 1120 994 ARUNACHAL 5340 3122 2218 CHANDIGARH 1413 960 453 DELHI 21174 8622 12551 GOA DAMAN 81108 35748 45360 PONDICHERY 15003 6824 8179 122 TABLE 5: DATA FOR COMPUTING LITERACY RATES FOR RURAL AND URBAN AREAS COMBINED STATE T=TOTAL POPULATION LITERATES =MALES F=FEMALES ANDHRA T 37485068 10689665 PRADESH M 19003247 7303297 F 18481821 3386368 ASSAM T 12413545 4295564 M 6619005 2932566 F 5794540 1362998 BIHAR T 48131372 11237613 M 24704179 8839395 F 23427193 2398218 GUJRAT T 22838394 9555238 M 11832989 6363748 F 11005405 3191490 HARYANA T 8459151 2699179 M 4556018 2005424 F 3903133 693755 HIMACHAL T 2964464 1105825 PRADESH M 1516749 763192 F 1447715 342633 JAMMU 6 T 3952135 857964 KASHMIR M 2120546 657660 F 1831589 200304 KARNATAKA T 25077251 9235127 M 12844451 6231363 F 12232800 3003764 KERALA T 18491971 12898072 M 9146093 7054096 F 9345878 5843976 MADHYA T 34971312 9223084 PRADESH M 18100306 7016655 F 16871006 2206429 MAHARASHTRA MANIPUR MEGHALAYA NAGALAND ORISSA PUNJAB RAJASTHAN SIRKIM TAMIL NADU TRIPURA UTTAR PRADESH WEST BENGAL W38 W36 MID-i W38 "1:6 "138 N38 "le "430-3 "le MID-3 'NKO-i 123 43157442 22443187 20714155 917845 464403 453442 850833 439603 411230 451806 243836 207970 18819069 9501641 9317428 11791168 6347946 5443222 21770367 11440534 10329833 184099 100205 83894 35809975 18112317 17697658 1331995 687556 644439 75379871 40371404 35008467 37856092 20239066 17617026 19752608 13331079 6421529 353090 249383 103707 298312 177772 120540 141518 96681 44837 5745399 4227806 1517593 4562123 2934281 1627842 4914293 3875435 1038858 37230 28579 8651 16256393 10783783 5472610 482082 322017 ‘ 160065 19173970 14812311 4361659 14711739 10031891 4679848 ANDAMAN NICOBAR ISLANDS ARUNACHAL CHANDIGARH DELHI GOA DAMAN 8 DIU PONDICHERY ‘30-! W3") W38 1130-3 1130-3 W38 124 98126 61478 36648 398085 217066 181019 224849 130037 94812 3536396 1982459 1553937 745700 374023 371677 406599 204302 202297 50191 36160 14031 52791 44776 8015 158371 98495 59876 2301605 1438268 863337 383864 234178 149686 217058 135851 81207 125 TABLE 6: DATA roa coupurruc LITERACY RATES roa.uanan annas. STATE T=TOTAL POPULATION LITERATEB u=uanzs (URBAN) (URBAN) r=rauanzs ANDHRA T 7287463 3955667 PRADESH n 3746351 2469737 0 3541112 1485930 assau T 1162057 778835 M 674663 489576 F 487394 289259 BIHAR T 4883646 2530541 M 2735655 1728205 F 2147991 802336 GUJRAT T 6509627 4115843 M 3453984 2532672 8 3055643 1583171 HARYANA T 1535441 904254 M 832356 565810 0 703085 338444 HIMACHAL T 213210 146450 PRADESH M 123579 92356 F 89631 54094 JAMMU 6 T 752316 327610 KASHMIR M 406238 215025 9 346078 112585 KARNATAKA T 6173920 3662937 8 3243884 2248588 8 2930036 1414349 KERALA T 3027897 2298663 M 1514313 1249309 F 1513584 1049354 MADHYA T 5784024 3361666 PRADESH n 3120768 2195840 8 2663256 1165826 MAHARASHTRA MANIPUR MEGHALAYA NAGALAND ORISSA PUNJAB RAJASTHAN SIKKIM’ TAMIL NADU TRIPURA UTTAR PRADESH WEST BENGAL W38 W38 W38 W38 W38 W38 W38 W38 W38 W38 W38 W38 126 13707516 7615897 6091619 122872 62197 60675 128923 70472 58451 45064 31802 13262 1585856 869334 716522 2815030 1523043 1291987 3886800 2085224 1801576 17124 10374 6750 10880453 5590416 5290037 143467 74393 69074 10696693 5932061 4764632 9853848 5694663 4159185 9123909 5774697 3349212 75334 47012 28322 95985 55552 40433 31241 23088 8153 904215 599479 304736 1688220 1014695 673525 1974945 1345562 629383 9081 6259 2822 7024806 4265565 2759241 103933 60756 . 43177 5404708 3542936 1861772 6133785 3783254 2250531 ANDAMAN NICOBAR ISLANDS ARUNACHAL CHANDIGARH DELHI GOA DAMAN 6 DIU PONDICHERY W38 W38 W38 W38 W38 W38 127 23113 15276 7837 15068 10757 4311 203874 117376 86498 3185958 1788751 1397207 198512 105835 92677 171921 85910 86011 16132 11265 4867 8724 7032 1692 150949 92885 58064 2149915 1325855 824060 127729 77468 50261 111505 65386 46119 128 TABLE 7: DATA FOR COMPUTING LITERACY RATES FOR RURAL AREAS. STATE T=TOTAL POPULATION LITERATES =MALES (RURAL) (RURAL) F=FEMALES ANDHRA T 30197605 6733998 PRADESH M 15256896 4833560 F 14940709 1900438 ASSAM T 11251488 3516729 M 5944342 2442990 F 5307146 1073739 BIHAR T 43247726 8707072 M 21968524 7111190 F 21279202 1595882 GUJRAT T 16328767 5439395 M 8379005 3831076 F 7949762 1608319 HARYANA T 6923710 1794925 M 3723662 1439614 F 3200048 355311 HIMACHAL T 2751254 959375 PRADESH M 1393170 670836 F 1358084 288539 JAMMU 6 T 3199819 530354 KASHMIR M 1714308 442635 F 1485511 87719 KARNATAKA T 18903331 5572190 M 9600567 3982775 F 9302764 1589415 KERALA T 15464074 10599409 M 7631780 5804787 F 7832294 4794622 MADHYA T 29187288 5861418 PRADESH M 14979538 4820815 F 14207750 1040603 MAHARASHTRA MANIPUR MEGHALAYA NAGALAND ORISSA PUNJAB RAJASTHAN SIKKIM TAMIL NADU TRIPURA UTTAR PRADESH WEST BENGAL W38 W38 W38 W38 W38 W38 W38 W38 W38 W38 W38 W38 129 29449926 14827290 14622536 794973 402206 392767 721910 369131 352779 406742 212034 194708 17233213 8632307 8600906 8976138 4824903 4151235 17883567 9355310 8528257 166975 89831 77144 24929522 12521901 12407621 1188528 613163 575365 64683178 34439343 30243835 28002244 14544403 13457841 10628699 7556382 3072317 277756 202371 75385 202327 122220 80107 110277 73593 36684 4841184 3628327 1212857 2873903 1919586 954317 2939348 2529873 409475 28149 22320 5829 9231587 6518218 2713369 378149 261261 116888 13769262 11269375 2499887 8577954 6248637 2429317 ANDAMAN NICOBAR ISLANDS ARUNACHAL CHANDIGARH DELHI GOA DAMAN 6 DIU PONDICHERY W38 W38 W38 W38 W38 W38 130 75013 46202 28811 383017 206309 176708 20975 12661 8314 350438 193708 156730 547188 268188 279000 234678 118392 116286 34059 24895 9164 44067 37744 6323 7422 5610 1812 151690 112413 39277 256135 156710 99425 105553 70465 35088 131 TABLE 8: PER CAPITA INCOME (IN RUPEES) NAME OF STATE PER CAPITA INCOME COMBINED RURAL URBAN ANDHRA PRADESH 600 542 1407 ASSAM 1000 748 1652 BIHAR 900 490 2726 GUJRAT 1200 736 2121 HARYANA 1500 1117 1903 HIMACHAL PRADESH 674 473 1077 JAMMU 6 KASHMIR 880 563 1299 KARNATAKA 1100 748 1725 KERALA 1330 1197 1200 MADHYA PRADESH 1185 650 1721 MAHARASHTRA 1573 649 2437 MANIPUR 950 448 1180 MEGHALAYA 759 444 1002 NAGALAND 1000 601 1239 ORISSA 1208 685 1941 PUNJAB 1499 1214 2003 RAJASTHAN 995 517 1786 SIKKIM 801 398 997 TAMIL NADU 1300 792 2044 TRIPURA 609 476 1100 UTTAR PRADESH 1256 765 1820 WEST BENGAL 1204 940 2116 ANDAMAN NICOBAR 777 450 902 ARUNACHAL 1282 314 1290 CHANDIGARH 2538 1243 2902 DELHI 2400 933 2681 GOA,DAMAN, 6 DIU 1100 989 1459 PONDICHERY 1300 1505 2114 L I ST OF REFERENCES REFERENCES Agarwala, S.N. 1968. "Socio-Economic and Demographic Characteristics of the Rural Migrants and Non -migrants." Journal of Institute of Economic Research, III(2). Bose, A. 1968. "Migration Streams in India." Population Review, March, 1968. Bulsara, J.F. 1964. Problems of Rapid Urbanization in India. Popular Prakashan, Bombay. Census of India, 1971. Vol.1, Part II(C): Migration Tables. Davis, R. 1951. The Population of India and Pakistan. New Jersey, Princeton University Press. Davis, K. 1962. "Urbanization in India:Past and Future." In Turner, Roy (ed), India's Urban Future, Berkeley, University of California Press. Desai, I.P. 1964. "Small Towns - Facts and Problems." Economic Weekly, XVI(16). Eames, E. 1967. "Urban Migration and Joint Family in a North Indian Village. ” Journal of Developing Areas, December, 1967. Easterlin, R.A. 1978. "Economics and Sociology of Fertility: A Synthesis." In Historical Studies of Changing Fertility by C. Tilly. Princeton, Princeton University Press. El-Badry, M.A. 1967. "A Study of Differential Fertility in Bombay." Demography,IV(4). Goldstein, S. and Goldstein, A. 1981. "The Impact of Migration on Fertility: An Own-Children Analysis for Thailand." Population Studies, 35. Gopi, K.N. 1975. "Demographic and Social Structure of a Metropolitan Fringe Community." The Journal of Social Work, XXXVI. 132 Greenwood, M.J. 1981. "A Regression Analysis of Migration to Urban Areas of a Less Developed Country- The Case of India. Journal of Regional Science, II(Z). Hiday, V. 1982. "Migration, Urbanisation and Fertility in Philippines." International Migration Review, 12(3). Kleinman, 0.8. 1974. "Fertility Variation and Resources in Rural India. " Economic Development and Cultural Change, 21(4). Lee, 8.8. and Farber, S.C. 1984. "Fertility Adaptation by Rural-Urban Migrants in Developing Countries: The Case of Korea." Population Studies, 1(38). Macisco, J. and Bouvier, L. 1969. "Migration Status, Education, and Fertility in Puerto Rico." The Milbank Memorial Fund Quarterly, vol.47, No.1. Macisco, J. Weller, R. and Bouvier, L. 1982. "Some General Consideration on Migration, Urbanisation and Fertility in Latin America." In The Family in Transition; published by National Institute of Health. Majumdar, A.K. and Das, K.D. 1972. "Some Aspects of Adoption of Family Planning Practices in Indian Villages." The Journal of Family Welfare, 18(3). NSSO, National Sample Survey Organization, Government of India, 13th Round, 1955-58, Number 53 - Tables with notes on Internal Migration. Paydarfar, A. 1980. "Differential Life-Styles Between Migrants and Non-migrants: A Case Study of Shiraz City in Iran." Demography, II(3). Rele, J.R. 1969. "Trend and Significance of Internal Migration in India." Sankhya, Series B, 31. Rele, J.R. and Kanitkar,T. 1975. "Residence Background and Fertility in Greater Bombay." Population Studies,28. Robinson, W.C. 1960. "Urban-Rural Differences in Indian Fertility." Population Studies,XIV(2). 133 Sabagh, G. and Yim, S. 1970. "The Relationship Between Migration and Fertility in an Historical Context: The Case Study of Morocco. " International Migration Review, 14(2). Sovani, N.V. 1966. Urbanisation and Urban India. New York, Asia Publishing House. Stark, O. 1975. "Desired Fertility and Rural-to-Urban Migration in LDC's: The Positive Connection." The David Horowitz Institute for the Research of Developing Countries, Tel Aviv University. Suprounovich, B.P. 1971. Migration and Changes in the Geography of Rural Population and Agriculture in India. Census Centenary Monograph, No.7. Visaria, P.M. 1971. "Urbanisation, Migration and Fertility in India." In Campbell Arther et.al.(eds), The Family in Transition, Fogerty International Center. Zacharia, X.C. 1968. Migrants in Greater Bombay. London, Asia Publishing House. Zarate, A.O. and Zarate, A.U. 1980. "On theuReconcilation of Research Findings of Migrant-Nonmigrant Fertility Differentials in Urban Areas." International Migration Review,9. 134 H. "00000ES