Niranjan, S. ; Sureender, S. and Rao, G. Rama. : Family Structure in India Evidence from NFHS. Demography India. 27(2). 1998. P. 287-300. Family Structure in India - Evidence from NFHS S. Niranjan, S. Sureenderand G. Rama Rao Introduction With the increase in the urbanization and industrialization, the concept of family in India, which once was to create and maintain a common culture among the members of the family, is undergoing changes.

However according to Beteille (1964), inspite of socio-economic and political changes, family life and family structure have remained as an integral part of Indian society with the 'spirit of family solidarity' as the sustaining power. Ross (1961) found that many Indians went through changes in the type of family in which they lived in various sequences: large joint family, small joint family, nuclear family, and nuclear family with dependants.

D'Souza (1971) argues that, the Indian family has been subjected to stress and strain, and inspite of resistance to change over the centuries, is slowly undergoing a process of change significantly. According to Cohen (1981) "households have reputedly been shrinking in size for ten thousand years or more, right up to the present, and this is a result of an evolving technology that requires fewer co-operating people to secure food, rear children, and look after the sick".

Though it is generally felt that joint families, whose members were bound together by ties of common ancestry and common property dominate in the past, there are diverging views regarding the same. Gore (1968) says, "the fraternal or collateral joint family was never the most common form". Goode (1968) asserts that the large joint family was not common at any time in India perhaps because of the great forces of fission, initially between daughters-in-law and later between brothers.

In a study of three villages located in three different districts in Karnataka state, two-thirds of the families were nuclear and the rest were different forms of joint family (Rao, Kulkarni and Rayappa, 1986). Although it can be argued that over the years joint family is slowly giving way to nuclear families, a number of studies reveal that despite the fact of living in the nuclear family set-up many functional relationships are maintained with the nonresidential family members (Agarwala, 1962; Desai, 1964; Kapadia, 1969; Gore, 1968).

As still in India most of the marriages are arranged by the parents, marital life begins in the parents family and later depending on the situation, a dwelling unit is arranged by the parents or other older members of the family (Richard et 1 al. , 1985). Hence, it is possible that the decisions taken by the members of the nuclear family are guided by their parents and relatives. However, this to an extent depends on the different types of family structure. Generally, family types are classified conveniently in many of the studies as nuclear and joint families.

In such a case, it would be rather difficult to conclude meaningfully about the significance of these family types on its family members. To be precise, family types are classified differently by various scholars. Kapadia (1969) have identified two broad family types namely; nuclear and joint/extended, while Richard et al. (1985) and Caldwell et al. (1988) have classified into nuclear, stem, joint, joint-stem and others. To understand this issue further it would be necessary to understand the changes in the family structure at the macro level i. . , India over the years, which to an extent has been attempted in this paper. Objectives The specific objectives of this paper are: (i) to understand the change in family structure at two points of time i. e. , in 1981 (census) and 1992-93 (NFHS), in different states of India, (ii) to study the differentials in family structure by different socio-economic characteristics of the head of the family at the all-India level. Sources of Data The data for this paper is obtained from National Family Health Survey, which was conducted during 1992-93.

The primary objective of the survey is to provide national and state-level data on different demographic and socio-economic determinants in respect of family planning, maternal and child health indicators. The survey also collected the information at three levels-Village, Household and Individual levels. The data for this paper is obtained from the Household questionnaire, which contains information, related to age, sex, martial status, education, occupation and relationship to the head of the household for all usual residents as well as for the visitors who slept last night in the house.

In addition, the household questionnaire also included information on housing conditions, such as the source of water supply, type of toilet facility, land owning, type of house and various consumer durable goods and characteristics of the head of the household such as religion, caste and place of residence. A total of 88,562 households were interviewed in India over all, of which two-thirds are from rural areas. All these 88,562 households contains more than 5,50,000 persons, of which 4,99,369 are only the usual residents in the survey.

Hence, for the present 2 paper the above mentioned particulars are analyzed only for the usual residents in the family. For the comparative purpose, the information on Household Structure in India pertaining to 1981 census published by the Registrar General of India is used (Chakravorty and Singh, 1991). Although the two sources of information are different i. e. , census and survey, and hence not to be compared, due to the lack of other sources of information, this attempt is undertaken.

Methodology To know the distribution of family structure in India, following classifications are considered: Type of family Single Member Nuclear Broken Nuclear Supplemented Nuclear Definition The respondent who is alone This type of family includes Nuclear pair i. e. , Head and spouse with or without unmarried children Head without spouse but with unmarried children It includes three types of families a) Supplemented Nuclear : Head and spouse with or without unmarried children but with other relations who are not currently having spouses. ) Broken Extended Nuclear : Head without spouse but with other relations of whom only one is having spouse c) Supplemented Broken Nuclear: Head without spouse with or without unmarried children but with other unmarried/separated/divorced/widowed relation It includes both lineally extended and collaterally extended families a) Lineally extended family : Head and spouse with married son(s)/daughter(s) and their spouses and parents with or without other not currently married relation(s) (OR) Head without spouse but with at least two married son(s) and daughter(s) and their spouses and/or parents with or without other not currently married relations b) Collaterally extended family : Head and spouse with married brother(s)/sister(s) and their spouses with or without other relation(s) [including married relation(s)] (OR) Head without spouse but with at least two married brothers/sisters and their spouses with or without other relations Joint Family 3 Source: Chakravorty, C. and A. K. Singh. , 1991, Household Structures in India, Census of India 1991, Occasional Paper No. 1, Office of the Registrar General of India, New Delhi. Although there are varying definitions for classifying family structure, the above mentioned definitions are considered in this paper mainly to have a comparison with the 1981 census.

The various background characteristics considered in the analysis to describe the differentials in family type are: (1) Educational level of the head of family: (Illiterate, Literate-upto Primary, Middle complete, High school and above); (2) Place of Residence (Urban, Rural); and (3) Religion of the respondent (Hindu, Muslim and Others); (4) Caste of the respondent (Scheduled Tribe, Scheduled Caste and Others); (5) Sex of the head of family (Male, Female); (6) Landowning Status (Yes, No); (7) Age of the head of family (less than 40 years, 40-60 years and 60 and above); (8) Size of family (Small i. e. 1-3 members, Medium i. e. 4-6 members, Large i. e. 7-9 members, Very Large i. e. , 10 and above).

Results The data for urban areas (Table 1) reveals that in 1992-93; almost half of the urban population in India live in nuclear families and 23 percent, 20 percent in joint family and supplemented nuclear families respectively. The percentage of single member and broken nuclear families together is only 6 percent. As compared to 1981 census, there has been a decline in the single member, broken nuclear and supplemented nuclear families in 1992-93 and the percentage of nuclear and joint families has increased over the years (Graph 1). 4 Table 1: Percentage Distribution of Different Type of Families in States/UTs of India for Urban Areas in 1981 and 1992-93 States 1981 Census Single Broken Nuclear Suppl. Joint Membe Nuclear Nuclear family r 7. 91 4. 24 46. 77 23. 64 17. 08 5. 76 4. 42 52. 46 24. 47 12. 89 -4. 25 3. 25 4. 31 8. 53 4. 30 4. 39 6. 13 3. 8 4. 09 7. 61 7. 94 5. 41 5. 04 4. 53 3. 11 5. 40 -7. 03 3. 86 5. 27 7. 89 6. 21 6. 19 4. 93 3. 74 8. 81 7. 47 -42. 41 48. 43 49. 09 37. 35 48. 79 44. 13 42. 62 44. 91 44. 20 52. 95 43. 42 45. 46 49. 47 49. 17 43. 57 42. 09 -47. 35 47. 12 45. 52 51. 64 56. 33 49. 30 50. 15 51. 09 45. 06 20. 22 -23. 65 24. 04 20. 14 21. 45 20. 55 30. 06 32. 83 25. 08 19. 80 20. 00 31. 23 32. 52 23. 42 19. 63 24. 82 31. 99 -17. 92 20. 98 26. 44 21. 70 5. 10 21. 64 22. 00 20. 88 26. 78 53. 53 -20. 09 17. 98 18. 16 7. 94 20. 94 15. 06 14. 35 16. 81 21. 07 15. 11 5. 78 2. 60 10. 21 19. 41 19. 72 5. 60 -17. 30 20. 47 12. 94 5. 32 1. 95 8. 33 13. 46 12. 97 7. 35 5. 6 1992-93 NFHS Single Broken Nuclear Suppl. Joint Membe Nuclear Nuclear family r 3. 2 3. 3 49. 8 20. 2 23. 4 1. 5 2. 6 54. 3 20. 3 21. 3 3. 5 1. 1 5. 1 1. 5 5. 3 1. 5 2. 2 1. 5 3. 0 2. 6 1. 4 2. 0 -4. 5 1. 2 3. 6 -2. 8 3. 5 3. 2 5. 0 -7. 6 --5. 1 4. 1 -3. 9 1. 3 3. 5 2. 0 3. 5 3. 3 3. 3 4. 7 2. 7 3. 6 5. 2 3. 0 4. 8 3. 0 2. 8 2. 2 -3. 5 5. 2 3. 2 4. 0 -2. 8 --2. 4 5. 2 -43. 6 44. 3 47. 9 52. 0 56. 2 48. 1 46. 2 50. 6 45. 9 47. 2 54. 2 41. 1 75. 8 55. 0 51. 3 54. 3 -50. 8 48. 5 51. 3 40. 7 -55. 6 --50. 0 52. 3 -21. 3 18. 7 19. 0 18. 0 16. 2 18. 9 24. 8 22. 6 20. 5 20. 5 18. 8 28. 7 9. 7 20. 0 20. 3 17. 2 -25. 7 21. 4 18. 5 22. 3 -20. 1 --17. 6 21. 8 -27. 7 34. 24. 6 26. 5 18. 7 28. 1 23. 4 20. 6 27. 9 26. 1 20. 3 25. 2 9. 7 17. 5 24. 4 22. 6 -17. 2 21. 4 23. 8 28. 0 -13. 9 --24. 9 16. 6 -Total No. of cases 28747 1093 1227 1082 1359 1033 1035 987 1449 1213 1457 1753 345 202 227 1290 937 1096 -1445 229 2302 1080 -144 --3371 1827 -- India Andhra Pradesh Assam -Bihar 9. 60 Gujarat 6. 30 Haryana 8. 25 Himachal 24. 73 Pradesh Jammu & 5. 14 Kashmir Karnataka 5. 74 Kerala 4. 04 Madhya 9. 54 Pradesh Maharashtra 7. 87 Manipur 4. 07 Meghalaya 11. 56 Nagaland 14. 01 Orissa 11. 76 Punjab 6. 92 Rajasthan 8. 74 Sikkim 14. 92 Tamil Nadu -Tripura 8. 00 Uttar Pradesh 7. 47 West Bengal 9. 82 Union Territories A. & N. slands 13. 45 Arunachal 24. 77 Pradesh Chandigarh 14. 54 D. & N. Haveli 9. 46 Delhi 9. 34 Goa, Daman & 11. 39 Diu Lakshadweep 12. 92 5 Mizoram 5. 64 7. 43 42. 46 36. 54 7. 93 1. 8 5. 7 49. 6 27. 5 15. 5 561 Table 2: Percentage Distribution of Different Type of Families in States/UTs of India for Rural in 1981 and 1992-93 States 1981 Census Single Broken Nuclear Suppl. Joint Membe Nuclear Nuclear family r 5. 15 4. 58 42. 79 26. 19 21. 18 5. 85 4. 88 48. 82 24. 03 16. 42 -4. 72 3. 35 4. 28 8. 35 4. 42 5. 29 7. 14 3. 67 4. 67 6. 91 10. 11 8. 84 5. 66 4. 55 3. 24 7. 94 -5. 99 4. 12 4. 17 4. 62 7. 74 3. 18 3. 75 2. 98 11. 15 -40. 73 43. 92 42. 32 33. 8 42. 11 41. 79 46. 06 37. 83 42. 64 59. 63 56. 04 59. 82 47. 56 45. 03 36. 65 48. 32 -53. 05 39. 51 49. 07 51. 21 61. 73 42. 65 48. 98 43. 13 41. 79 -25. 47 25. 92 24. 20 31. 43 24. 62 29. 92 30. 01 28. 58 19. 77 17. 30 20. 66 17. 98 24. 58 23. 80 32. 64 25. 40 -18. 55 27. 57 25. 58 19. 65 4. 06 22. 80 21. 84 23. 54 28. 42 -24. 84 21. 98 25. 79 16. 97 24. 76 17. 71 13. 36 23. 09 27. 08 13. 05 6. 68 4. 85 16. 96 22. 25 22. 41 9. 65 -16. 07 23. 29 16. 74 7. 75 9. 93 12. 24 18. 94 22. 29 8. 68 1992-93 NFHS Single Broken Nuclear Suppl. Joint Membe Nuclear Nuclear family r 2. 5 3. 2 46. 3 20. 7 27. 4 2. 3 3. 0 47. 2 22. 4 25. 1 2. 9 2. 1 3. 3 0. 9 3. 1. 2 2. 7 2. 2 2. 6 2. 5 1. 4 1. 1 0. 2 2. 9 2. 1 1. 6 -4. 2 2. 4 2. 4 1. 7 -3. 7 --13. 3 4. 7 5. 1 1. 8 2. 8 1. 7 3. 4 3. 2 3. 9 4. 8 1. 9 2. 3 6. 4 4. 2 5. 1 3. 3 2. 3 1. 5 -5. 4 4. 3 2. 2 3. 2 -3. 8 --1. 0 7. 3 49. 8 40. 6 46. 4 43. 7 44. 4 49. 9 43. 4 48. 6 41. 5 43. 1 58. 6 54. 9 73. 8 47. 7 50. 6 43. 4 -53. 6 46. 9 41. 0 48. 0 -49. 7 --49. 7 45. 9 23. 7 18. 3 19. 8 18. 4 20. 5 18. 1 24. 6 21. 9 18. 6 23. 0 16. 2 19. 3 9. 4 21. 6 18. 4 18. 6 -22. 3 26. 7 20. 4 22. 6 -19. 1 --15. 7 25. 5 18. 6 37. 1 27. 7 35. 3 28. 2 27. 6 25. 3 22. 5 35. 5 29. 1 17. 4 20. 4 11. 5 24. 6 26. 6 34. 9 -14. 5 19. 7 33. 9 22. 4 -23. 7 --20. 3 16. 6 Total No. f cases 59534 3106 2021 3627 2509 1702 2074 1850 2813 3162 4391 2306 740 788 828 3288 2276 3901 -2837 908 7738 3141 -815 --300 1888 India Andhra Pradesh Assam -Bihar 4. 24 Gujarat 4. 83 Haryana 3. 33 Himachal 9. 47 Pradesh Jammu & 3. 59 Kashmir Karnataka 5. 08 Kerala 3. 43 Madhya 6. 39 Pradesh Maharashtra 5. 80 Manipur 2. 92 Meghalaya 6. 45 Nagaland 8. 51 Orissa 4. 90 Punjab 4. 16 Rajasthan 5. 00 Sikkim 8. 69 Tamil Nadu -Tripura 4. 26 Uttar Pradesh 5. 47 West Bengal 4. 42 Union Territories A. & N. islands 16. 77 Arunachal 11. 03 Pradesh Chandigarh 19. 13 D. & N. Haveli 6. 49 Delhi 6. 40 Goa, Daman & 9. 54 Diu 6 Lakshadweep 9. 24 Mizoram 3. 60 11. 27 6. 72 17. 22 49. 85 61. 07 27. 45 1. 20 12. 38 -1. 9 -3. 2 -61. 5 -21. 7 -11. 6 -525

In rural areas, while 46 percent and 27 percent of families belong to nuclear and joint types respectively in 1992-93, and the percentage of families in supplemented nuclear, broken nuclear and single member type are 21, 3 and 3 respectively. As compared to 1981 census, the pattern of change in urban areas in different family types is almost the same as in the rural areas in 1992-93 (Graph 2). As compared to urban areas, the single member households are less frequent in rural areas. It is obvious that persons who migrate to urban areas have to stay single for quite a long period of time, hence this type of families is found to be slightly more in the urban areas. In different states of India, in 1992-93, the percentage of nuclear families in the urban areas, is high in Nagaland (76 percent) and low in case of Bihar (44 percent) and Uttar Pradesh (51 percent).

In the less developed states like Bihar and Uttar Pradesh, a high proportion of joint families could possibly indicate the traditional nature of the society and the lack of significant occupational mobility found among the people of the state. In case of Nagaland, it could be that the tradition, which normally demands newly, weds to set up as a separate family and also along with the absence of large landholding result in higher proportion of nuclear families. Although, there is an increase in joint families in 1992-93 as compared to 1981, it is found to be more pronounced in urban areas than in rural areas in most of the states of India. In the states of Tamil Nadu, Kerala, Karnataka, Orissa, Goa, Daman & Diu, Meghalaya, Mizoram and Arunachal Pradesh, the proportion of supplemented nuclear type of families are higher than the joint type of families in 1992-93.

A few plausible reasons for the same are given below: 7 (a) Migration is an important component which forces the members of the natal family to stay together usually revolving around one married couple and sharing the economic assets and income. For instance, when a person migrates for work he leaves his wife, children behind with his parents, hence in such a society the supplemented nuclear families are expected to be more. Similar is the situation, in case of a person who migrates to urban areas and is forced to stay with his relatives, due to lack of housing. (b) The increase in the socioeconomic development of the society also leads to increase in the status of women.

And as the status of women in the family and in the society being high, it is expected that the chances of divorce/separation would also be high in case of marital incompatibility between the partners. This is found to be true in case of Kerala where the status of women and the divorce rates are high (Sureender et al. 1992). In this situation too, there is a possibility of the supplemented nuclear families to be more in the society. In general, there has been a decline in single member, broken nuclear and supplemented nuclear type of families, while an increase is observed in nuclear and joint families as compared to 1981 in most of the states of India.

While the decline in the proportion of single and broken nuclear families could be largely attributed to improvement in the health conditions of the population over the years, whereas the following reasons could be thought of for the increase in nuclear families. • The decline in the interest of the individual towards satisfying the groups (families) interest; The problems related to housing especially in case of migrants to urban areas; The lack of adjustment between the family members especially when a bride/groom enters into the family. According to Caldwell et al. (1996), the ultimate reason for the break-up of the joint family is the friction between mother-in-law and daughter- in-law and between daughters-inlaw themselves.

In the case of increasing joint families, problems related to housing could be cited as an important factor especially in urban areas, which force the couples to stay in joint families. Further, low age at marriage among girls also force the couples to stay with the parents till they attain economic independence. This implies an addition of married women in the same family, which results in the classification of the family as a joint family. Joint family also gave security to widows, physically handicapped, economically non-productive and other insecure members • • 8 of the society. They were thus, the ideal type of household meeting all the requirements of society in the past (Chakravorty and Singh, 1991).

According to Mandelbaum (1970), people tend to remain in joint families longer when economic factors favour such families. He also argues that the poorest and the lowest groups tend to have fewest joint families, but even at these social levels, most families become joint for at least for some time after son marries. Further it was argued that even if a whole society strives towards ideal of joint families, there will inevitably be a very considerable proportion of simpler families at transitional stages in a demographic cycle because of deaths among the older generation, the departure from the joint family of surplus married brothers and other factors.

Hence, even a considerable proportion of nuclear families in a population is evidence neither of change nor of the forming of that type of families. In additions to the above-mentioned ones there are obvious socioeconomic characteristics of head of the family which result in changes within the family structure. Some of these characteristics which are analysed with the family structure in this paper at the all India level are: Education, Age, Religion, Residence, Caste, Sex, Land owning status, and the size of the family. Differentials The results in Table 3 show that there is not much difference in family structure for both urban and rural areas in India as a whole.

However, a considerable increase is observed in the proportion of nuclear families in urban areas compared to rural areas. This statement strengthens the hypothesis that the urban respondents are more likely to choose the nuclear family than the rural respondents. Table 3: Percentage Distribution of various type of Families According to the Background Characteristics of the head of Household (1992-93 NFHS Survey) Background Characteristics Type of Family Single Broken Member Nuclear 3. 7 1. 9 2. 0 2. 4 2. 5 1. 9 2. 9 5. 1 2. 6 1. 7 1. 1 3. 3 3. 8 3. 1 Number of househol ds 36067 24267 8562 19385 10587 10759 669635 Nuclear Suppl. Nuclear 21. 0 20. 2 21. 3 19. 7 20. 20. 3 20. 6 Joint Family 27. 6 27. 2 22. 8 23. 5 24. 7 22. 5 26. 9 Education Illiterate Literate-primary Middle Complete High School + Caste Scheduled Caste Scheduled Tribe Other Castes 42. 7 48. 1 52. 2 53. 3 49. 2 51. 5 46. 5 9 Religion Hindu Muslim Sikh Others Place of Residence Urban Rural Sex of the Head Male Female Land Owning Yes No Age of the Head ; 40 years 40-60 years 60 + Marital Status Married Separated Widowed Divorced Never married Family size Small Medium Large Very Large 2. 9 2. 2 2. 1 2. 9 3. 2 2. 5 1. 7 11. 7 2. 0 3. 5 2. 3 2. 1 5. 0 0. 8 21. 4 12. 4 12. 9 23. 2 14. 6 ---- 3. 1 2. 7 4. 7 2. 5 3. 3 3. 2 0. 8 24. 2 2. 4 4. 2. 5 4. 3 2. 5 -34. 3 27. 4 30. 4 1. 5 9. 5 2. 5 0. 8 0. 1 46. 7 47. 3 52. 9 49. 2 49. 8 46. 3 51. 1 16. 1 42. 6 52. 2 57. 7 50. 3 22. 7 55. 0 ----56. 8 59. 5 30. 6 6. 6 20. 7 20. 1 19. 9 19. 3 20. 2 20. 7 18. 7 36. 2 20. 8 20. 3 23. 7 18. 2 20. 6 15. 7 37. 9 49. 4 50. 0 59. 2 15. 5 22. 7 24. 4 9. 9 26. 6 27. 7 20. 4 26. 0 23. 4 27. 4 27. 8 11. 9 32. 2 20. 1 13. 8 25. 7 49. 3 28. 5 6. 4 10. 8 6. 7 15. 5 3. 7 15. 3 44. 2 83. 4 68948 8623 8880 1830 28747 59534 79003 9273 43720 44545 32670 37152 18459 76013 688 9261 194 2110 16596 43274 20123 8288 10 A positive association is found between education of the head of the family and family structure.

When the head of the family is illiterate, only 43 percent families are nuclear, the relative percentage for the heads who are educated upto high school and above, is 53 percent (Graph 3). Similarly more percentage of scheduled tribes stay in nuclear families as compared to scheduled caste and other caste people, i. e. almost 52 percent of nuclear families were found in scheduled tribes compared to 49 and 46 percent in scheduled caste and other caste people. As evidenced, more proportion of low waged population are prevalent in low caste, so always the head of the family tries to push away the married children from his house to make the family with reduced burden.

This could be the plausible reason why the nuclear families are more found in low castes. Srivastava and Nauriyal (1993) also noted in Uttar Pradesh that the joint family system is found to be more popular among the higher castes than the intermediate and lower castes. It is possible that, since the land holding are more among the non-scheduled caste/tribe people, they tend to stay more in joint families compared to scheduled caste/tribe people (Caldwell et al. , 1988). In a study conducted in Karnataka, Caldwell et al. (1984) show that, among those with no land at all, 71 percent are found in nuclear families; with land upto one acre, 65 percent; with land from one to four acres, 58 percent; with over four acres 46 percent.

With more resources and a need for more labour, there is more point in keeping a larger family together. A study of 5,200 households throughout Karnataka state, conducted in 1975 by the Bangalore Population Centre, recorded the percentage of different types of families as follows: 57. 3% nuclear, 30. 8% stem, 4. 7% joint, and 3. 4% joint stem. The same picture is found in this study too, i. e. , those who possess land, higher percentage stay in joint families than those with no land (Graph 4). It could be that the requirement of manpower in agricultural families and the practice of property staying with the senior citizen of the family tend to keep the joint families intact. Nimkoff (1959) also writes that in India, he joint family system is traditionally most common among the elite, the higher castes and those with more property. It is often held that joint families are especially appropriate for peasants who cultivate land, that such families, especially those who till their own land, favour large families and favour joint families, because the excess numbers form labour pools (Kolenda Pautine et al. , 1987). The sex of head of the family is having a significant relation in forming a particular type of the family. While 51 percent of the male headed families are found to be nuclear type and only 16 percent of nuclear families have female as head of the family. However, the picture is found different in case of the supplemented nuclear and broken nuclear families.

Female-headed families are found to be more in supplemented nuclear and broken nuclear family types. While, migration of males in search of jobs could be one of the reasons which forces the females to head the supplemented nuclear families, the death of the 11 husband and to an extent the increasing divorce rate (especially in urban areas) could be some of the plausible reasons in case of more females heading the broken nuclear families. Religious differentials clearly indicate that more percentage of Sikhs (53 percent) are living in nuclear families compared to all other religions (i. e. , 46. 7, 47. 3 and 49 percent from Hindu, Muslim and other religions respectively).

It has been observed in the analysis that the proportion of illiterate heads of the family in Sikh religion is considerably less as compared to the other religions. Also, it was supported that the education of the head of the family has a positive association with family structure as the education increases, the proportion of nuclear families increase. Hence, it could be one of the plausible reasons; the nuclear families are more in Sikh religion. Though, it has been found in the analysis that in India as a whole, the proportion of nuclear families are more in Sikh religion compared to other religious groups. Independently, in Punjab and Haryana the proportion of Sikh religion is more, but it comprises only 25 percent of India's Sikh population.

Our results support the contention of Kingsley Davis; "Sikhs took more seriously to education, as they are more literate than either the Hindus or the Muslims. Their high percentage in the Indian army has doubtless helped their literacy". Also a district wise analysis of selected states in India by Kolenda Pauline et al. (1987) reveals that high joint family districts have more Hindus and substantially fewer Christians than the low joint family districts. These figures according to the authors suggest that Hindus have a preference for joint family living compared to other religions. As expected, age of the head of family is having a significant association with family structure.

Joint families are found to be more among the older ones where the age of the head of family is over 60 years (49. 3 percent). While only 25. 7 percent of the middle aged beads maintain joint families, and the corresponding percentage among younger ones is 14. It is 12 felt that, always the old persons prefer to maintain their family as joint type, because to fulfil their psychological satisfaction through the youngest in the family. This finding is found to be similar to an earlier study conducted by Driver (1962) in Nagpur district of Maharashtra. A study by Morrison (1959) reveals that nuclear families are generally small and medium in size, whereas joint families are large and very large.

It is equally obvious that there is considerable overlapping in the medium and large size categories so that it is not possible to say that medium sized families are always nuclear and large sized are always joint in composition. Similar type of results were noticed in this study too, that, nuclear families are generally small and medium in size and joint families are large and very large in size. To be precise, while percentage of small and medium size in nuclear families are 56. 8 and 59. 5 respectively and among the joint families are 3. 7 and 15. 3 respectively. On the other hand, the percentage of large and very large size families in nuclear type are 30. 6 and 6. 6 percent respectively as compared to 44. 2 and 83. 4 among joint families.

These results clearly show that broken nuclear families and nuclear families are usually small and medium in size, whereas joint families are large in family size. Conclusion Keeping in view that the changes in family structure are inevitable partly as a result of continuing demographic change, this paper examines the changes in family structure from 1981 to 1992-93 in India. The results reveal that over the years, there has been an increase in the nuclear and joint families, although nuclear families are leading in both the rural and urban areas. On the other hand, a decline is observed in the single member, broken nuclear and supplemented nuclear families. Nuclear families are found to be more in case of Nagaland as compared to rest of the states of India.

The differentials in family structure reveals that the socio-economic background of the head of the family has a definite role to play in the growth of nuclear families in India. These changes in the family structure calls for the examination of its plausible consequences on the attitude and behaviour of the family members at the micro level. References 1. Agarwala, B. R. , 1962, Nature and extent of social change in a mobile commercial community. Sociological Bulletin, 11. 2. Beteille, A. , 1964. Family and social change in India and other South Asian Countries. Economic and Political Weekly, Annual. XVI: 237-244. 13 3. Caldwell, J. C. ; Reddy, P. H. and Caldwell, Pat. , 1984.

The determinants of family structure in Rural South India. Journal of Marriage and the Family, 46 (1): 215-230. 4. Caldwell, J. C. ; Reddy, P. H. and Caldwell, Pat. , 1988. The Causes of Demographic Change : Experimental Research in South India. University of Wisconsin Press, Madison. 5. Caldwell, J. C. , Reddy, P. H. and Caldwell, Pat. , 1996, The family in South India : Past, present and future. Social Change, 26(2) : 116-129. 6. Chakravorty, C. and Singh, A. K. , 1991. Household Structures in India. Census of India 1991. Occasional Paper No. 1. Social Studies Division, Office of the Registrar General, India. 7. Cohen, Yebudi A. , 1981. Shrinking Households.

Society 48-52. 8. Desai, L. P. , 1964. Some Aspects of Family in Mahuva. Asia Publishing House, Bombay. 9. Driver, E. D. , 1962, family structure and socioeconomic status in Central India. Sociological Bulletin, 11:112-120. 10. D'Souza, A. , 1971. The Indian Family in the Changes and Challenge of the Seventies. Sterling Publisher Private Limited, New Delhi. 11. Gore, M. , 1968. Urbanization and Family Change. Popular Prakashan, Bombay. 12. Goode Willian, J. , 1968. Foreword in M. S. Gore, Urbanization and Family Change. Popular Prakashan, Bombay. 13. Gould, H. A. , 1968. Time-dimension and Structural Change in an Indian Kinship System. In: M.

Singer and B. S. Cohn (eds. ). Structure and Change in Indian Society, pp. 413-42 1. Chicago. 14. Kapadia, K. M. , 1959. The family in transition. Sociological Bulletin, 8 (2): 68-99. 15. Kapadia, K. M. , 1969. Marriage and Family in India. Oxford University, Press, Bombay. 14 16. Khatri, A. A. , 1972. The Indian family: An empirically derived analysis of shifts in size and types. Journal of Marriage and the Family 34 (4): 725-734. 17. Kolenda, Pauline and Haddon, Lorraine. 1987. Marked Regional Differences in Family, Structure in India, In: Pauline Kolenda (ed. ), Regional Differences in Family Structure in India. Rawat Publications, Jaipur. 18.

Krishna Moorthy, S. and Kulkarni. P. M. , 1985-86, Family formation and structure. Journal of Family Welfare. 32 (1). 19. Mandelbaum, David G. , 1970. Society in India - Continuity and Change (Vol. 1). University of California Press, London. 20. Morrison, W. A. , 1959. Family types in Badlapur: An analysis of a changing institution in a Maharashtrian Village. Sociological Bulletin, 8 (2): 45-67. 21. Nimkoff, M. F. , 1959, The family in India. Sociological Bulletin. 8 (2): 32-58. 22. Nimkoff, M. F. and Middleton, R. 1960. Types of family and types of economy. American Journal of Sociology, 66 (3): 215-225. 23. Rao, N. Bhaskara, Kulkarni, P. M. , Rayappa and P.

Hanumantha, 1986, Determinants of Fertility Decline: A Study of Rural Karnataka. South Asia Publishers, New Delhi. 24. Reddy, P. H. and others, 1975, Dual Record System. Population Research Centre, Bangalore. 25. Richard, J. , et al. , 1985, Family type and the aged. The Journal of Family Welfare, 31 (4): 31-38. 26. Ross, A. D. , 1961, The Hindu Family in its Urban Setting. Toronto: Oxford University Press. 27. Srivastava, K. K. and Nauriyal, D. K. 1993, Family structure and child survival among Jamsaris of Uttar Pradesh. Social Change, 23 (2&3): 159163. 28. Sureender, S. et al. , 1992. Divorce in India: A macro level analysis. Social Change, 22 (2). 15