Version classiqueVersion mobile

Essays on population and space in India

 | 
Christophe Z. Guilmoto
, 
Alain Vaguet

Part I. Demographic trends

1. Demographic Transition in India

Jacques Véron

Texte intégral

1As a country long characterized by persistence of a high fertility rate despite very early adoption of a population policy (the first birth control programmes date from 1952), India since the 1970s has seen a marked reduction in the average number of children that a woman brings into the world. In 1990, the total fertility rate was 3.4 children per woman, which puts India only slightly above the world average.

2The decline in fertility is generalized, but its dimension varies in different states : even today, there are very great regional differences. In Goa or Kerala, the fertility model is that of the developed nations, with an average of two or less children per woman, while in Uttar Pradesh the rate is still almost five.

3This drop in fertility accompanies a marked lowering of mortality, which however does not seem to be the sole explanation for it, contrary to the theoretical postulates of the theory of demographic transition in its strictest form. Improved education for women, a good indicator of their status, explains the reduction in fertility better than the lower mortality rate.

4It is difficult to discern direct effects of population policies upon birth and fertility rates, but regular revision of the targets originally fixed, in view of observed changes, suggests a limited effectiveness. However it is still difficult to assess the significance of social change and availability of means of contraception (or the influence of messages about the desirability of birth control disseminated by the media) as a cause of smaller families.

A late but definite drop in fertility

  • 1 Fertility estimates are taken from the Sample Registration System (SRS) For the period 1990-1992 th (...)
  • 2 The figures quoted by Agartala (1967) for the 1950s and 1960s give fertility estimates of more than (...)

5The Indian fertility rate1 estimated at almost 6 children per woman at the 1 end of the last century (5.76 in 1881-1891), continued to fluctuate around this figure right up to the 1960s (Table 1.1 and Figure 1.1). There even seems to have been a slight increase in fertility over this period, but considering the quality of the data it is difficult to interpret this apparent increase.2

Table 1.1. Changes in fertility over the last century

Year

Total fertility rate

1881-1891

5.8

1891-1901

5.7

1901-1911

5.7

1911-1921

5.7

1921-1931

5.8

1931-1941

5.9

1941-1951

5.9

1951-1961

6.1

1961-1971

5.9

1971

5.2

1972

5.2

1973

4.9

1974

4.9

1975

4.9

1976

4.4

1977

4.1

1978

4.4

1979

4.4

1980

4.4

1981

4.5

1982

4.5

1983

4.5

1984

4.5

1985

4.3

1986

4.2

1987

4.1

1988

4.0

1989

3.9

1990

3.8

1991

3.6

1992

3.6

1993

3.5

1994

3.5

1995

3.5

1996

3.4

Source : SRS and Mari Bhat’s estimates.

6Family size begins to show substantial reductions towards the end of the 1960s and the beginning of the 1970s. In 1970 Indian fertility stood at around 5 children per woman, 15 years later it was close to 4, and the latest data — those of the SRS— show a rate of 3.4 in 1996.

7Although the drop in fertility is general, the situation remains quite varied, as is shown by comparison of the total fertility rates of the different states of the Indian Union (Table 1.2 and Figure 1.2). Fertility is lowest in Goa (1.9 children per woman) and highest in Uttar Pradesh (4.82 children per woman).

8In general, fertility is lower in the south, with levels significantly below 3 children per woman in 1990-1992 in the four states of Andhra Pradesh, Karnataka, Kerala and Tamil Nadu. The highest fertility rates are found in one northern state (Haryana : 3.99 children per woman), two central ones (Madhya Pradesh : 3.90 and Uttar Pradesh : 4.82), one eastern one (Bihar : 4) and one north-eastern state (Arunachal Pradesh : 4.25).

9The importance of the present regional contrasts should not lead us to forget that everywhere, even in Bihar or Uttar Pradesh, fertility has dropped significantly. Thus between 1972 and 1990-92, the average number of children per woman dropped by more than 30 per cent in Uttar Pradesh (from 7.3 children per woman to 4.8 : a decrease of 2.5 children per woman). During this period a number of states have experienced a lessening of their current fertility of 40-50 per cent. In Gujarat, the drop reached 52 per cent in 20 years, and in Kerala, it was more than 56 per cent (while the fertility in this state was already relatively low in 1972). It should be noted that Kerala remains an outstanding case : in 1972 its fertility rate was equal to that of Karnataka, while today it is 0.85 child per woman lower than that of Karnataka.

Figure 1.1 : Changes in Indian fertility

Figure 1.1 : Changes in Indian fertility

Women still give birth at a very early age

10An examination of fertility curves by age, based on data from the Sample Registration System and the National Family and Health Survey allows us to evaluate fertility by age group, and changes in fertility by age group over time (Table 1.3 and Figure 1.2). Note that the estimates of fertility rates by age group supplied for the period 1990-92 by the Sample Registration System and the National Family and Health Survey are consistent as a whole, but diverge very considerably for the age group 15-19 years. The reason seems to be a difference in definition of the reference population, which is de jure in the case of SRS but de facto in the case of the NFHS. If, for example, the young women return to their parents’ home to give birth to their first child, birth of a normal resident may not be registered by the SRS.

11Alterations in fertility by age group have taken place mainly in the higher age groups : the drop in fertility of the last twenty years is essentially a drop in the marital fertility convex in 1972, the curve of fertility by age group then becomes concave, which is generally considered indicative of a more widespread use of contraceptives.

  • 3 According to NFHS figures the proportion of menopausal women —self-defined as such, or who have had (...)

12However, use of contraception is thought not to be the only factor involved. Relatively low fertility rates in the higher age groups are attributed by NFHS officials to various reasons : a high proportion of women who have been sterilized or passed the menopause3 and cessatand cessation of sexual relations between a married couple when a daughter reaches reaches puberty, one of their child marries, or when one of their children produces a child (International Institute for Population Sciences, 1995).

13Fertility schedules seem to differ widely within India. Very youthful fertility is particularly marked in Madhya Pradesh, Andhra Pradesh, Haryana and Maharashtra, where the fertility rate among the 15-19 is more than 140 per thousand, whereas in Mizoram, Manipur, Goa and Kerala it is below 50 per thousand.

14The other traditional component in reduced fertility — higher age at marriage— does not seem to play a great role, since fertility rates in younger age groups remain high. Thus in India fertility remains an early phenomenon, because of the persistence of significant fertility amongst young age groups (absolute effect) and decreasing fertility amongst older ones (relative effect).

15Early fertility does not necessarily signify large families : in Madhya Pradesh, the fertility rate among the 15-19 is markedly higher than in Uttar Pradesh (153 per thousand as against 113 per thousand), but the total fertility rate is much lower (3.90 as against 4.82). In Maharashtra a woman’s childbearing begins early and ends early : the fertility rate among the 15-19 is as high as in Uttar Pradesh, but the number of children per woman is almost 2 children less.

Figure 1.2 : Variation in fertility by age group

Figure 1.2 : Variation in fertility by age group

Figure 1.3 : Total fertility rates, 1990-92, by state

Figure 1.3 : Total fertility rates, 1990-92, by state

Source : NFHS

16To evaluate trends of decreasing fertility in conformity with the paradigm of demographic transition, it is interesting to compare the change in the number of children per woman with that in mortality, since this too has undergone a marked decrease.

Mortality and fertility : a simultaneous decrease, without much interaction

17All mortality indicators have shown a decrease in India over the course of the last century (Table 1.4). From 1911 to 1991, life expectancy at birth rose from 23 to 57 years (Figure 1.4), which however still leaves India amongst the countries where mortality remains high. Mortality rates have decreased by more than half in 90 years, yet the rate of infant mortality in 1991 remain 96 per thousand (16 times that in France, where the infant mortality rate in 1994 was less than 6 per thousand).

Table 1.2 : Total fertility rates in the Indian states (number of children per woman)

state

1972

1990-1992

Andhra Pradesh

4.7

2.59

Arunachal Pradesh

-

4.25

Assam

5.8

3.53

Bihar

5.2

4.00

Delhi

-

3.02

Goa

-

1.90

Gujarat

6.2

2.99

Haryana

7.4

3.99

Himachal Pradesh

5.5

2.97

Jammu and Kashmir

-

3.13

Karnataka

4.6

2.85

Kerala

4.6

2.00

Madhya Pradesh

7.0

3.90

Manipur

-

2.76

Maharashtra

4.9

2.86

Meghalaya

-

3.73

Mizoram

-

2.30

Nagaland

-

3.26

Orissa

5.0

2.92

Punjab

5.7

2.92

Rajasthan

7.0

3.63

Tamil Nadu

4.3

2.48

Tripura

-

2.67

Uttar Pradesh

7.3

4.82

West Bengal

-

2.92

Source : Srinivasan (1995) and the International Institute for Population Sciences (1995).

18The figures differ greatly from state to state, for in 1990-91 the infant mortality rate ranged between 15 per thousand in Mizoram and 112 per thousand in Orissa (International Institute for Population Sciences, 1995). The rate in Orissa is confirmed by findings from other sources (SRS). On the other hand the probability of dying during the first year of life seems “abnormally” low in Mizoram, as also in Nagaland (infant mortality rate, 17 per thousand) if we compare their figures with those in Goa (32 per thousand) and Kerala (24 per thousand), which are known to be very advanced on the plane of demographic change.

19Do the actual disparities in infant mortality from state to state explain the differences in fertility ? Looked at “simultaneously” (using current data), the relationship between mortality and fertility is less than might be expected (Figure 1.5). According to Srinivasan (1995), who has undertaken regression analyses of figures from the end of the 1980s, infant mortality explains only 60 per cent of the fertility — that is, differences in infant mortality from state to state do not systematically correspond with differences in fertility, following the same ranking. Thus in Orissa, where infant mortality is much higher than in Haryana, fertility is significantly lower. Fertility rates in Assam and Punjab are comparable, while infant mortality rates are very different (the rate in Assam is 30 points higher than that of the Punjab).

Table 1.3 : Age specific fertility rates (per 1 000 women)

Table 1.3 : Age specific fertility rates (per 1 000 women)

Source : SRS series and IIPS (1995)

20According to the theory of demographic transition in its strict form, decrease in birth rate/fertility rate is seen as a response to a decrease in mortality. Decreased mortality (overall or infant mortality) is supposed to “explain” the decrease in fertility. Changes in the figures for infant mortality and fertility over the course of the century do show a decrease in mortality preceding that in fertility, but the figures are not sufficient to allow satisfactory measurement of any correlation. On the other hand, correlation calculation can be carried out in more acceptable conditions, based on the figures by state (since the information available is more complete).

Table 1.4 : Changes in mortality over the last hundred years

Year

Crude mortality rate

Life expectancy at birth

Infant mortality rate

1911

42.6

22.9

222

1921

47.2

20.0

212

1931

36.3

26.8

176

1941

31.2

31.8

168

1951

27.4

32.1

148

1961

22.8

41.3

139

1971

18.9

45.6

129

1981

15.0

50.5

129

1991

11.1

57.3

96

Source : Estimates quoted by Srinivasan (1995)

21This does not validate the theory of demographic transition in its strict form. However it must be recalled that for testing the theory a cohort analysis, over time, is more suitable than a cross section analysis, and that in addition it is important to allow for some time-gap between changes in mortality and changes in fertility. The enlarged theory of demographic transition explicitly integrates social change. Therefore, it is worth investigating the influence of other variables besides mortality upon changes in fertility— especially the education of women.

Women’s education, an essential variable for change

22In India there are marked disparities not only from a geographical standpoint : significant variations also appear according to place of residence level of education, and religion (Table 1.5).

Figure 1.4 : Life expectancy at birth

Figure 1.4 : Life expectancy at birth

23On an average, rural women give birth to one child more than urban women do. The fertility of women without any education is almost two children more than that of women who have attended at least high school. Muslim women have, on average, one more child than Hindu women and two more than Sikh women. Christian fertility lies between that of Sikhs and Hindus.

Table 1.5 : Fertility differentials according to place of residence, level of education and religion of women (total fertility rate in 1992-93)

Criterion

Number of children per woman

Place of residence

urban

2.70

rural

3.67

Education

illiterate

4.03

literate < middle complete

3.01

middle school complete

2.49

high school and above

2.15

Religion

Hindu

3.30

Muslim

4.41

Christian

2.87

Sikh

2.43

Other

2.77

Source : International Institute for Population Sciences. 1995.

24Do Muslim women have more children because they are Muslim (indicating a strong religious influence), or because they are less often literate than women from other communities (influence of education) ? Cross comparison of religion and educational level reveals a persistence of both effects (Table 1.6). A Muslim woman will give birth to 5 children if she has had no education, while a Sikh woman in the same situation will give birth to only 3.4 (on the other hand in this case fertility amongst Christians is lower than amongst Sikhs). Moreover, whatever her level of education, a Muslim woman has a higher fertility rate than that of a woman from any other religious background.

Table 1.6 : Fertility by religion and educational level of women (number of children per woman in 1992-93)

Education Religion

Illiterate

Literate <Middle complete

Middle school complete

High school and above

Hindu

3.93

2.93

2.45

2.07

Muslim

5.03

3.61

3.05

2.97

Christian

3.30

2.86

2.50

2.79

Sikh

3.43

2.80

2.06

1.68

Others

3.57

2.59

2.59

2.14

Total

4.03

3.01

2.49

2.15

Source : International Institute for Population Sciences, 1995.

25Thus the influence of religion is not an effect of the influence of education ; but women sharing the same religious background show fertility rates that very clearly vary according to their level of education : the disparity between the extremes (without education, and level corresponding to at least high school) is about 2 children per woman amongst Hindus and Muslims.

26Education is a fundamental variable of demographic change, in India as elsewhere. It gives a better explanation than mortality for the differences in fertility between states : 72 per cent of the variance in the average number of children per woman at the end of the 1980s is explainable by literacy rates, as against 60 per cent by infant mortality rates (Srinivasan, 1995).

27The figures for 1990-92 confirm an inverse relation between fertility and women’s education, but there is a not inconsiderable scattering around the average values : within a given level of women’s education family size may vary by as much as one child (Figure 1.6).

28A study carried out at district level in India, based on data from the 1981 Census, showed that the influence of education on fertility could be broken down into a direct and an indirect effect (Sharma and Retherford, 1990). The higher the women’s literacy level, the less children they bore (direct effect). The indirect effect worked through a higher age at marriage, and a decreased infant mortality rate. This study demonstrated that when the variable of place of residence (urban or rural) was added, the model did not gain in explanatory value. No independent “urbanization” effect existed, apart from education, age at marriage or infant mortality.

29In Kerala, where fertility is very low today, urban and rural fertility rates are little different (1.8 children per woman in urban areas in 1990-92, as against 2.1 in rural areas), and fertility curves by age group show similar forms. Fertility varies little according to the womens educational level, but the total fertilility rate amongst Muslim women remains more than one child per woman higher than that of Hindus and Christians. The low fertility of Kerala is however linked with a low rate of urbanization (urban dwellers 28 per cent of the total population) and a very high level of women s education (proportion of female literates, 84 per cent).

Figure 1.5 : Fertility and infant mortality in 1992-93, by state (NFHS)

Figure 1.5 : Fertility and infant mortality in 1992-93, by state (NFHS)

30In developing countries, women’s education is a distinguishing variable, since the attitudes and expectations of educated women differ from those of illiterates in many respects : marital behaviour, freedom to decide the number and spacing of births, access to contraceptives, cost of pregnancy, etc. (Cochrane, 1988). More generally, the educational level of women is an indication of their status. This means that its influence on behaviour cannot be reduced to purely mechanical effects.

31Since India was the first country in the world to adopt a population policy, it is appropriate to inquire, in the final analysis, about the role played by birth control programmes upon observed fertility and modes of contraception.

Birth control by female sterilization

  • 4 According to the NHFS study, 82 per cent of the women who resort to sterilization have never used a (...)

32Fertility correlates very closely with contraceptive practices. In India, the prevalence of contraceptives remains quite weak, since only 41 per cent of women between 13 and 49 years use any method of contraception. By taking age into account, we can gain a more detailed evaluation of behaviour. Thus, it seems that in the age group between 30 and 44 years, more than half of all Indian women practice some form of contraception. In fact, women of this age limit their fertility by mass recourse to sterilization.4 41 per cent of all women aged 30 to 34 have undergone sterilization, and 45 per cent of those aged 35 to 39. Then the proportion of sterilized women decreases (40 per cent at 40-44 years, and 31 per cent at 45-49 years) which can probably be explained by a cohort effect superimposed upon an age effect. Male sterilization is far less widespread than female sterilization. It reaches its peak among in couples where the wife is between 45 and 49 : 11 per cent of men married to women in this age group have been sterilized. Taking all age groups together, female sterilization is thus by far the most common method of limiting offspring (it involves 27 per cent of all couples practising contraception — who constitute 41 per cent of the total number of couples). Second in order of importance come traditional methods (4.3 per cent of these couples), and third comes male sterilization (3.4 per cent).

33How far is increased use of contraception, and above all reduced fertility, the result of the population policies successively introduced in India over the past forty years ? The first five-year Plan (1951-1956) presented to Parliament in 1952, made reference to “family limitation and population control.” It made provision, amongst other things, for identification of the causes of rapid demographic growth, and the discovery of techniques enabling easy dissemination of family planning means (Srinivasan, 1995). The third Plan (1961-66) set the objective of making small families the social norm, through education and the involvement of opinion leaders. Amongst the objectives of the eighth Plan (1992-97), containing population increase is mentioned yet again.

34An evaluation of the effectiveness of population control programmes can be made on the basis of an analysis of the targets in terms of numbers of births set by the population policies under the various Plans. The target of a birth rate of 25 per thousand by 1972, set in 1962, was increased in 1969 to 32 per thousand (for the period 1975-85). The aim of 25 per thousand by 1984, set in 1974, was reconsidered in the sixth Plan (1985-90) and replaced by a target birth-rate of 29 per thousand for around 1990. The target of the most recent Plan is a rate of around 26 per thousand for 1996-97, and 22 per thousand by 2006-07.

35However, the fact that targets in terms of birth rate have had to be regularly revised, does not prove that the population programmes have had no effect. It is difficult to reach an exact assessment, but criticism of these policies has been directed mainly at the organization of family planning, which has been handled in a very uniform manner, and at the too great place given to sterilization, thus giving preference to a total prevention of births, rather than to lowering the number of births at an early stage in fertile life (since Indian women give birth mostly at a young age).

Figure 1.6 : Fertility and women’s education in 1992-93 by state (NFHS)

Figure 1.6 : Fertility and women’s education in 1992-93 by state (NFHS)

36Demographic transition is indeed taking place in India, but mortality remains high, and fertility has decreased slowly over the last 20 years. Regional diversity remains extremely high, for some states have only recently started the transition —in terms of a move towards low fertility— (Uttar Pradesh, Arunachal Pradesh, Bihar) while others have already achieved it or are on the way to doing so (Kerala, Mizoram, Tamil Nadu).

37A rise in the status of women, as reflected by literacy rates, largely explains fertility decline. This goes along with increasing use of contraceptives, which has however not yet reached very high levels in India. Birth control programmes do not seem to have taken sufficiently into account the geographical, social and religious diversity existing in the country.

38Since fertility is decreasing in the Indian Union far less rapidly than in China, it is not surprising that present projections show India as the nation with the highest population in the world by 2050.

Notes

1 Fertility estimates are taken from the Sample Registration System (SRS) For the period 1990-1992 they are supplemented with those of the National Family and Health Survey (NFHS), which supplies a wealth of information, both current and retrospective

2 The figures quoted by Agartala (1967) for the 1950s and 1960s give fertility estimates of more than six or even seven children per woman in both urban and rural areas

3 According to NFHS figures the proportion of menopausal women —self-defined as such, or who have had no period for at least six months— varies by age-group as follows : 3 per cent at 30-34 years, 7 per cent at 35-39 years, 17 per cent at 40-41 years. 25 per cent at 42-43 years, 38 per cent at 44-45 years, 56 per cent at 46-47 years, and 71 per cent at 48-49 years (International Institute for Population Sciences. 1995).

4 According to the NHFS study, 82 per cent of the women who resort to sterilization have never used any other method of contraception beforehand.

Table des illustrations

Titre Figure 1.1 : Changes in Indian fertility
URL http://books.openedition.org/ifp/docannexe/image/9798/img-1.jpg
Fichier image/jpeg, 15k
Titre Figure 1.2 : Variation in fertility by age group
URL http://books.openedition.org/ifp/docannexe/image/9798/img-2.jpg
Fichier image/jpeg, 28k
Titre Figure 1.3 : Total fertility rates, 1990-92, by state
Crédits Source : NFHS
URL http://books.openedition.org/ifp/docannexe/image/9798/img-3.jpg
Fichier image/jpeg, 107k
Titre Table 1.3 : Age specific fertility rates (per 1 000 women)
Crédits Source : SRS series and IIPS (1995)
URL http://books.openedition.org/ifp/docannexe/image/9798/img-4.jpg
Fichier image/jpeg, 27k
Titre Figure 1.4 : Life expectancy at birth
URL http://books.openedition.org/ifp/docannexe/image/9798/img-5.jpg
Fichier image/jpeg, 19k
Titre Figure 1.5 : Fertility and infant mortality in 1992-93, by state (NFHS)
URL http://books.openedition.org/ifp/docannexe/image/9798/img-6.jpg
Fichier image/jpeg, 27k
Titre Figure 1.6 : Fertility and women’s education in 1992-93 by state (NFHS)
URL http://books.openedition.org/ifp/docannexe/image/9798/img-7.jpg
Fichier image/jpeg, 33k

© Institut Français de Pondichéry, 2000

Conditions d’utilisation : http://www.openedition.org/6540

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search