A POPULATION policy is generally conceived at the national level. However, in a large country like India with conspicuous spatial variations, a case has often been made for region or state specific policies and some state governments have recently announced population policies. But there are other dimensions of diversity, especially ethnicity, religion and caste. On account of social, cultural and historical factors, the aspirations of groups, and hence immediate objectives, may differ, if not the ultimate goals. Besides, strategies appropriate for one group may not be so for another, even when the objectives are the same.

The Swaminathan Committee suggested the ‘concept of unity in population goal and diversity in implementation strategies’ (Expert Group on Population Policy, 1994, p. 3). However, when interests of various sections differ, it is possible that a policy might give precedence to the concerns of dominant groups at the cost of disadvantaged and minority groups.

It is imperative that the national goals and objectives incorporate aspirations and needs of various groups, including smaller and weaker sections of society, strategies be designed to meet these, and further that national policies and strategies do not adversely affect the interests of various groups, intentionally or otherwise. Formulation of policies separately for social groups may not be feasible since such groups do not form units or levels of governance. However, it is essential that the perspectives of special population groups – ethnic, religious and caste – be taken into account in the formulation of national or state policies.

In principle, a national policy ought to be designed towards achieving collective national goals. But when there is stratification on the basis of race, ethnicity, religion or caste, group interests could possibly differ and conflict with national interests. In the sphere of population, though lowering the population growth rate may be a desirable national goal, many groups would like to ensure that their share does not fall during the process of demographic transition since share in power depends on share in population. This is an issue well recognised in the context of religious diversity in India (Pai Panandiker and Umashankar, 1994; Jain, 1998). Clearly, religions form special groups that could legitimately have concerns about the size and share of their population.

 

 

Moreover, the scheduled castes, which suffered due to oppression and denial of opportunities for generations, and the scheduled tribes that were secluded and excluded from the process of development for a long time, constitute special groups. This paper focuses on the perspectives of religion and caste groups in India in matters of population policy. Specifically, how do religious minorities, the scheduled castes and the scheduled tribes fare in population and related aspects? Do some special groups need policy interventions? The demographic profile and changes in it could give some idea of the needs of special population groups. In particular, an assessment of differentials in growth rates and changes in population share is useful.

This calls for an examination of differentials in the components of population changes, namely, fertility, mortality and migration. Further, population policies in recent years have given greater attention to aspects of reproductive and child health. Do the special groups receive appropriate services in these? Finally, does the National Population Policy 2000 of India address the needs of special groups?

Population growth: The changes in the relative size of population attract public attention because of political overtones. In India, over 99% of the population was recorded as Hindu, Muslim, Christian, Sikh, Buddhist, or Jain in the 1991 Census (India, Registrar General, 1995). The Hindus constituted 82%,¹ the Muslims are the largest minority with 12% of the population, followed by the Christians and the Sikhs with about 2% each (Table 1). The Buddhists and the Jains also have substantial populations, over six and three million respectively. In addition, many other religions and persuasions are listed in the census. Of these, Zorastrianism and Judaism are well recognised, but in 1991 these two religions had populations of only 76,382 and 5,271 respectively.

The share of various religions in the population has not changed much since 1961. There has been some marginal rise, less than two percentage points, in the share of Muslims and a nearly equal fall in the share of Hindus.² Very small changes have occurred in the shares of Christians and Sikhs. The growth rate has been higher than average among the Muslims, by about half a percentage point (annual growth rates are presented in the lower panel of Table 1).

The Christian growth rate was higher in the past but has dropped since 1971. Growth rate has also fallen for Sikhs and Jains. A rise is seen in the case of the Buddhists, but this is more likely to be due to the adoption of Buddhism by many, especially the scheduled castes, rather than due to a higher natural increase. The growth among the Jains is particularly low. Among the other religions, the Zorastrians (Parsees) show a negative growth. The population declined from 111,800 in 1951 to 76,382 in 1991 and there has naturally been concern among the community about this (Visaria and Visaria, 1999).

 

 

The scheduled castes and tribes together account for about a quarter of India’s population; at the time of the 1991 Census, 16.5% were classified as scheduled castes and 8.1% as scheduled tribes (Table 2). The share of the scheduled tribes did not change much between 1981 and 1991 but that of the scheduled castes increased by about one percentage point. The scheduled castes experienced higher growth during 1981-1991, 2.70% annually compared to 2.13% for the entire population. However, this is at least partly due to a change in classification (Visaria and Visaria, 1999).

Prior to 1990, scheduled castes were recorded only among Hindus and Sikhs, but the Constitution (scheduled castes) Order (Amendment) Act, 1990 for the first time recognised scheduled castes among Buddhists as well (India, Registrar General, 1994). Since the late 1950s, many persons belonging to the scheduled castes have adopted Buddhism; though not counted as scheduled castes in the earlier censuses, the 1991 Census classified them as scheduled castes. The growth rate of the scheduled caste population is quite high in Maharashtra³ (6.7%), the state that had a large number adopting Buddhism. Overall, there is no cause for concern about the population sizes of the scheduled castes or scheduled tribes as a whole being too small or declining. However, this is not necessarily true for individual tribes. Some tribes are very small in size and need special attention.

 

 

Child mortality: Do the special groups suffer from higher than average mortality? Life tables by religion or by caste are generally not computed. However, the 1981 Census and the two National Family Health Surveys (NFHS), NFHS-1 and NFHS-2, have given estimates of early childhood mortality by religion and caste (India, Registrar General, 1988; IIPS, 1995; IIPS and ORC Macro, 2000). The NFHS-1 was conducted during 1992-93 and the NFHS-2 during 1998-99, the estimates refer to the 10 year period prior to the respective survey. Two key indicators, the Infant Mortality Rate (IMR) and the Under-Five Mortality rate (U5MR), which is the proportion of children dying before completing five years of age, are given in Table 3.

The census and NFHS estimates of IMR and U5MR show child mortality to be slightly lower than average among Muslims and substantially lower among Christians and Sikhs; the NFHS-2 estimates of IMR are 59, 49, and 53 for Muslims, Christians and Sikhs respectively compared to the average of 73. Mortality is also relatively low among Jains (IMR is 47) and Buddhists (54). On the other hand, infant mortality is higher than average among the scheduled castes (83) and the scheduled tribes (84). Thus, the populations belonging to the minority religions are not disadvantaged in terms of mortality but the weaker sections, the scheduled castes and the scheduled tribes certainly are.

The religious composition of population varies across states. The proportion of Christians is large in the southern states, especially Kerala, that of Muslims in some southern and some central-eastern states, and that of Sikhs in the northwestern states. Since there are large regional variations in the level of mortality, it is conceivable that some of the inter-religion variations observed are attributable to regional variations in religious composition. This calls for an examination of differentials within states.

In Kerala, the only state that has a large (that is, large enough to allow computation of child mortality estimates) Christian population, mortality among Christians is not much lower than average.⁴ In Punjab, mortality among Sikhs is only marginally lower than that for Hindus. The Muslims have higher mortality than the Hindus in Kerala and Assam but lower in many states; the differences are generally not large.

Thus, the large Hindu-Christian and Hindu-Sikh differences seen at the national level are not conspicuous within states. The observed lower mortality among Christians and Sikhs is on account of the low mortality in the regions in which large populations of these communities reside, the states of Kerala and Punjab. However, the scheduled castes and tribes have as above average infant and child mortality in most of the states.

 

 

Fertility: Estimates of fertility measures for major religions and for the scheduled castes and tribes are available from some recent censuses and surveys.⁵ Estimates of the Total Fertility Rate (TFR) and Cumulative Fertility (CF computed as mean children ever born to women of age 40-49 at survey) from the NFHS-1 and NFHS-2, are presented in Table 3. Generally, fertility is higher than average among Muslims and lower among Christians and Sikhs. The NFHS-2 estimates of TFR for Muslims, Christians and Sikhs are 3.6, 2.4, and 2.3 respectively compared to the average of 2.9. Though the TFR for Muslims is higher than average, the difference does not exceed one point, that is, one additional child per woman.

The scheduled castes and tribes show only marginally higher fertility than average; the NFHS-2 estimates of the TFR are 3.2 and 3.1 respectively. If the TFR is lower than the CF, recent fertility decline is indicated since the CF measures cohort fertility and the TFR measures current fertility during the reference period of three years prior to the survey. It can be seen that there is a decline in fertility in all communities but not necessarily at the same pace.

In Andhra Pradesh, Kerala, and Tamil Nadu, TFR for Christians is not much different from that of Hindus. Similarly, fertility level among Sikhs is nearly identical to the level among Hindus in Punjab and Haryana.⁶ Thus, in states where Sikhs and Christians have substantial populations, they do not seem to differ from Hindu populations in fertility. The TFR for Muslims is higher than average in many states, the differences are wide in West Bengal, Assam, Karnataka, Maharashtra, Rajasthan, Kerala, Bihar and Uttar Pradesh, but narrow or negligible in Gujarat, Andhra Pradesh, Madhya Pradesh and Tamil Nadu.

Thus, though the Muslims have higher than average fertility at the national level, the gap differs considerably over states. Moreover, the gap is not clearly associated with the level of fertility. Among low fertility states, the gap is wide in Kerala, but narrow in Andhra Pradesh and Tamil Nadu and among states with moderate fertility, the gap is wide in West Bengal but narrow in Gujarat.

Fertility is low among Jains, actually below replacement level; the NFHS-2 estimate of the TFR is 1.9 (figures not shown in the Table). The Jains are a highly urbane and advanced community. Fertility among Buddhists is lower than average, but in Maharashtra, the state with a large Buddhist population, the level is not much different from the average. Though NFHS estimates of fertility for Zorastrians are not available, other surveys and investigations provide evidence that fertility is very low among Zorastrians (Rele and Kanitkar, 1974). Overall, though fertility varies by religion, there is no need to be concerned about unusually low fertility, except for religions like Zorastrianism.

It must be noted here that higher fertility among women belonging to a particular religion may not necessarily be attributable to the religion factor per se. There are differences in characteristics such as educational level, residence (rural or urban), occupational distribution and income, many of which have a bearing on fertility. Therefore, some or all of the religious differentials in fertility could be on account of differentials in one or more of these factors (the Characteristics Hypothesis). Appropriate multivariate analysis can indicate the extent to which the Characteristics Hypothesis explains the differentials. This issue is not being addressed in this paper; substantial literature is available on this (Alagarajan and Kulkarni, 1998; Balasubramanian, 1984; Jeffery and Jeffery, 2000; Moulasha and Rao, 1999; Shariff, 1996; Visaria, 1974).

Age at marriage: Since fertility in India occurs primarily within marriage, age at marriage has a large effect on completed fertility. The low age at marriage in India has been cited as one of the main factors responsible for high fertility. Estimates from recent censuses and surveys show that the median ages at marriage of the Hindus and the Muslims do not differ much, but the Christians and the Sikhs marry late. The scheduled castes have a lower age at marriage than average but the difference is not large. The age at marriage for scheduled tribes is not much different from the overall average.

According to the NFHS-2, the median ages at marriage for women are 20.5 for Christians and 20.1 for Sikhs (IIPS and ORC Macro, 2000). Among other religions for which NFHS estimates are available, the median is low for Buddhists (16.9) but not for Jains (18.9). The medians for the scheduled castes and the scheduled tribes are 16.3 and 16.6 respectively, quite close to the median of 16.9 for Hindus and Muslims. Thus, though the age at marriage has risen over time, a majority of marriages are performed before the legally permitted age of 18 years for girls. This is, thus, a broad issue, affecting almost all communities except Christians, Sikhs and Jains.

 

 

Contraceptive practice: This is another important proximate determinant of fertility. The Contraceptive Prevalence Rate (CPR), defined as the per cent of couples of reproductive age using some contraception, has increased in India during the past few decades and was estimated at 45% in 1998 (India, Department of Family Welfare, 1999). The NFHS-2 estimate for 1998-99 is 48%. Most of this, 43 points, is of modern methods, and 36 points, contribution of sterilisation, female or male.

The NFHS-1 and 2 give estimates of the prevalence rate for major religions and the scheduled castes and tribes. The patterns in the surveys are fairly similar and the discussion here is based on differentials as seen in NFHS-2 (shown in Table 4, upper panel). Contraceptive prevalence rate is computed for ‘any method’ (that is, use of any method of contraception), for ‘modern methods’ (use of any modern method including sterilisation) and for ‘sterilisation’. Prevalence for any method is lower among Muslims (37%) and higher among Sikhs (65%) compared to the average (48%). The differences are in the same direction in the case of modern methods.

In per cent sterilised, the figure for Sikhs is close to the average; clearly, the higher CPR among Sikhs is on account of higher use of reversible and traditional methods. Among Muslims, the CPR for sterilisation is much lower than average (20% compared to the average of 36%), but for reversible methods, the prevalence is higher than average (17% compared to 12). Thus, there is some difference in the pattern of method mix for the various communities.

 

 

Sterilisation is less popular among Muslims, and reversible and traditional methods more popular among Sikhs and to some extent among Muslims as well. The prevalence rate among the scheduled castes is close to average. The overall prevalence among the scheduled tribes is lower than average, but the per cent sterilised is close to the average.

Contraceptive prevalence is influenced by two factors: demand for fertility regulation and the use of contraception in case of such demand. Demand would generally be influenced by socio-economic conditions. Factors such as education, income, occupation and place of residence have a bearing on the perceived costs and benefits of children, and couples would demand contraception if low fertility were considered to be beneficial. The programme and other service providers seek to meet this demand.

But the Indian programme, like many other programmes, also seeks to influence (or generate) demand. The extension component of the Indian family planning programme was designed for this purpose. Plausibly, both demand and acceptance of contraception in case of demand could vary by community. Perceptions of costs and benefits of children could differ. Access to contraceptive services could vary. Besides, cultural factors may influence acceptability of the idea of fertility regulation and of specific contraceptives.

An indicator of the total demand for family planning (TDFP) is the percentage of women who did not want another child at all or did not want one within two years, information on which was obtained in the NFHS. Differentials in TDFP are similar to those in CPR, higher level for Sikhs (74%) and lower for Muslims (59%) compared to the average of 64% in NFHS-2 (Table 4). However, the differences in TDFP are narrower than in CPR. For example, the CPR for Muslims is 11 points lower than average but TDFP is only five points lower. Similarly, for Sikhs, the CPR is 17 points above average but the TDFP only 10 points. Among the scheduled tribes, demand for contraception is lower than average (55%).

A measure of met need (MN) for contraception is the percentage of women who (or whose husbands) were using any contraception among those with demand for family planning. Overall, 75% of the need for contraception is met, but this is lower among Muslims (63%) and very high among Sikhs (88%) compared to others (Table 4). In terms of met need, Christians, scheduled castes, and scheduled tribes are close to the average.

Within states, the pattern of differentials is generally similar to the national pattern, with some departures.⁷ The CPR is lower than state average among Muslims in most states except Gujarat, Madhya Pradesh and Tamil Nadu. Though the CPR for Sikhs is higher than average nationally, this is not the case in Punjab. The scheduled tribes have lower CPR than average in Assam, West Bengal, Andhra Pradesh, Bihar, Madhya Pradesh, and Rajasthan, but not in Gujarat. Again, TDFP follows a pattern similar to that of CPR in most states but with narrower differences. The proportion of need for contraception that is met is lower than average among Muslims in most states.

 

 

Reproductive and child health: For quite some time, maternal and child health care services have been integrated with the family planning programme in India. This was justified on a number of grounds. First, the service delivery network could cater to the needs of birth control and maternal and child health, especially preventive health care such as immunisation, and pregnancy and delivery care. Second, improvement in child survival was considered essential to achieving a fertility decline. Third, integration of maternal and child health with family planning gave the programme the image of being a ‘family welfare’ programme rather than a mere ‘birth control’ programme.

 

 

The scope was further broadened in the 1990s with the introduction of the ‘reproductive and child health’ programme. All women are expected to receive care during pregnancy that includes physical check-up, tetanus-toxoid injections, supplementary nutrition, and advice about health care. Broadly, this is called antenatal care. Professional care during delivery and after (post-natal health care) is also important. Similarly, for children, a regime of immunisations is recommended during the first year.

The NFHS obtained data on pregnancies that resulted in live births during a reference period of three years before survey and on child immunisation. Data on birth weight were also collected. But for a majority of births, weight was not recorded, and using the available data on birth weights (that is, on births for which weight was recorded) to examine differentials would not be appropriate because of the problem of selection bias.

Table 4 (lower panel) gives differentials in four indicators of reproductive and child health care. These are: (i) ANC: per cent of pregnancies in the reference period during which the woman received antenatal check-up, whether from public or private providers; (ii) Institutional Delivery: per cent of deliveries during the reference period that were conducted in health institutions; (iii) Professional Assistance: per cent of deliveries during the reference period that received professional assistance (from a doctor or a nurse/midwife); and, (iv) Child Immunisation: per cent of children of age 12-23 months at survey who had received all the recommended doses of vaccinations.

The coverage for antenatal care is higher among Christians (84%) and Sikhs (75%) and lower among the scheduled tribes (57%) than average (65%); the levels for Muslims and the scheduled castes are close to average. The extent of institutional care for delivery is relatively high for Christians (54%) and Sikhs (47%), low for the scheduled castes (27%) and very low for the scheduled tribes (17%).

 

 

A majority of Sikhs and Christians were successful in securing professional assistance during delivery (69% Sikhs and 64% Christian). But only 42% of Hindus, 39% of Muslims and 37% of scheduled castes could obtain such assistance. The most disadvantaged are the scheduled tribes, only 23% of deliveries received professional assistance. In child immunisation, the Sikhs are well covered (70% children receive all recommended immunisations), as are the Christians (61%) but not the scheduled tribes (26%) and the Muslims (33%).

Some of the differentials observed at the national level are not seen within states. In particular, though Christians show higher than average level of reproductive and child health care coverage at the national level, in Kerala, Tamil Nadu, and Andhra Pradesh this is not the case.⁸ Similarly, in Punjab, hardly any Hindu-Sikh difference is seen. However, delivery care is poorer for the scheduled tribes than average in almost all the states with substantial scheduled tribe populations. To a smaller extent, this is true of the scheduled castes as well. In the case of Muslims, large differences (coverage for Muslims poorer than average) are seen in some northern and eastern states but not in the southern and western states.

 

 

The available data show that religious minorities and weaker sections such as the scheduled castes and tribes differ in certain demographic aspects from the average or the majority population. The rate of growth has fallen among Christians but is not so low as to be of concern. The fall has been steeper among Jains. This is considered to be an elite community and appears to be well ahead of the other major communities in demographic transition with fertility having fallen below replacement level. However, the current growth rate is still positive.

On the other hand, the Zorastrians have experienced a decline in population size. The populations of some tribes are also low enough to cause concern though the scheduled tribes as a whole continue to show positive growth. Child mortality is high especially among the scheduled castes and tribes. The scheduled castes and tribes face handicaps in income and education and hence a high level of mortality is not unexpected. However, the public health programmes are designed so that the disadvantaged populations can be catered to. Clearly, the programmes have not been fully successful. Women from the weaker sections are not as well covered as the majority populations in delivery care.

This is particularly true of the scheduled tribes though the scheduled castes also do not fare well. Similarly, child immunisation is less common among the scheduled tribes. Though Muslims do not experience higher than average child mortality, Muslim children in northern-eastern states are not as well covered as others. Delivery care is also poorer for Muslim women compared to other women in some states. Contraceptive needs of most sections are met to a large extent; about 75% of those not desiring an additional child soon (or ever) do use contraception. But among Muslims, there is relatively greater unmet need.

 

 

The core issue is whether population policies in India have addressed the needs of special groups. The thrust of the National Population Policy 1976, the first explicit population policy in India, was on intensifying the programme and there was little mention of special groups. But in the context of compulsory sterilisation it was stated that, if brought about by states, it be made ‘uniformly applicable to all Indian citizens resident in the state without distinction of caste, creed or community’ (India, Department of Family Welfare, 1977: p. 174). This was apparently designed to emphasise uniformity.

The policy statement of 1977, primarily rejecting the element of compulsion in family planning, made no mention of special groups (for reviews of the policies, see, Mitra, 1978; Srinivasan, 1982; Raina, 1988; Visaria and Chari, 1998). At the international level, the International Conference on Population and Development, (ICPD) 1994, recognised that indigenous populations have a distinct and important perspective on population and development relationship (U.N., 1994). One of the objectives stated by the ICPD was to ensure that indigenous people receive population and development related services that they deem socially and culturally appropriate.

The National Population Policy 2000 (NPP 2000) explicitly recognises the special needs of tribal communities and hill area populations. The policy statement notes the problems of low literacy, poor nutrition and high childhood mortality among these sections of population (India, Department of Family Welfare, 2000: para 25). The policy acknowledges that these groups are under-served on account of poor access and suggests strategies to overcome the problem. Significantly, the NPP 2000 also makes the point that ‘many tribal communities are dwindling in numbers and may not need fertility regulation. Instead they may need information and counselling in respect of infertility’ (ibid: p. 23).

 

 

Thus, there is recognition of the fact that a population policy need not always be reduced to one of fertility control. Health issues receive primary attention for these special populations. It must be stated that though the earlier population policies did not note the special needs of tribal populations, health programmes have always done so and prescribed better than average health centre to population ratios for tribal and hilly areas.

The NPP 2000 does not take note of the special needs of the scheduled castes. The data show that the scheduled castes are not well served in delivery care. Poverty is undoubtedly a factor operating in this matter. But are public services also loaded against the scheduled castes? A point to be noted is that in antenatal care the scheduled castes are as well served as other groups. This is a task performed by the female health worker (in some states called an auxiliary-nurse-midwife or public health nurse).

To their credit, these workers have been doing a good job of reaching all sections of the society and providing services to pregnant women. But delivery care falls in a different class. Accessing a health facility or a health professional at the time of need is not easy for women from the weaker sections and the existing programmes have not been able to overcome this handicap.

 

 

Religion as a factor does not find place in the NPP 2000 document. Recognition of religious differentials in a policy document is presumably considered politically incorrect. In fact, as noted earlier, the 1976 policy stressed that there should be no distinction by community. Yet certain needs do vary by religion. The low or negative growth of some religions does not attract much public attention perhaps because populations of these are quite small and upper class. It is probably felt that they are capable of addressing these issues themselves and not in need of policy support.

The relatively high fertility among Muslims has been a touchy issue for some time. While it is true that demand for contraception is lower than average among Muslims, sweeping statements such as Muslims are against birth control are not supported by evidence. The NFHS-2 revealed that 37% of Muslim couples used some contraception and 20% were sterilised. These figures are not insignificant and in fact, contraceptive prevalence among Muslims in India as a whole is higher than the level for general population in major states like Bihar and Uttar Pradesh.

But an important issue that emerged out of the data presented is the high unmet need for contraception among Muslims. Possibly, there are reservations about specific methods, particularly sterilisation; the prevalence of reversible methods is not lower among Muslims compared to the general population. If the cafeteria approach were seriously pursued allowing couples greater choice in reversible methods, the contraceptive needs of Muslims could also be met to a larger extent.

 

 

Finally, the comparative view of the special groups discussed in the paper, namely religious minorities and the scheduled castes and tribes, reveals that the principal differences are in areas of health. The scheduled castes and especially the scheduled tribes are poorly served in maternal and child health care. This calls for strategies to cater to the needs of these weaker sections. The NPP 2000 has recognised these issues for the scheduled tribes, but special strategies could also be developed to cater to the needs of the scheduled castes.

The NPP 2000 proposes special schemes for urban slums, this would take care of needs of some weaker sections. However, a majority of the scheduled castes live in rural areas and programmes need to be developed for this large population. In contraceptive practice, the large unmet need among Muslims is a matter of concern. Moreover, though the NPP 2000 explicitly recognises that some tribes are dwindling in number, it is not clear what specific efforts are contemplated to support these tribes. In the matter of low fertility caused by infecundity, specific reproductive health programmes can help. But population policy may not be the appropriate instrument to overcome the problem of low growth caused by late marriage and voluntary low fertility, observed in some sections, since international experience shows that pro-natalist policies have rarely been successful.

The available data also indicate that certain issues are relevant in some states and not necessarily in others. The NPP 2000, designed to operate at the national level, may not be an ideal instrument to address these. State policies are normally designed to achieve state level goals and plan strategies appropriate to conditions in the state. Given the large inter-state variations, this is a desirable approach. But in doing so, it is essential that the policies for individual states give adequate attention to the needs of special groups within these.

 

TABLE 1

Population Size, Share and Growth Rates by Religion in India

 

Year/ Period	Religion All	Hindu	Muslim	Christian	Sikh	Buddhist	Jain
Population $ (in thousands)
1991	838,584	687,647	101,596	19,640	16,260	6,388	3,353
Percentage Share*
1991	100.00	82.41	11.67	2.32	1.99	0.77	0.41
1981	100.00	83.09	10.88	2.45	1.96	0.71	0.48
1971	100.00	83.51	10.37	2.56	1.94	0.70	0.49
1961	100.00	84.27	9.84	2.41	1.83	0.73	0.47
Average Annual Exponential Growth Rate* (per cent)
1961-71	2.17	2.08	2.69	2.78	2.75	1.73	2.48
1971-81	2.23	2.18	2.72	1.57	2.34	2.22	2.10
1981-91	2.13	2.05	2.83	1.56	2.27	3.07	0.43
1961-91	2.18	2.11	2.75	1.97	2.46	2.34	1.67

$: Excluding Jammu and Kashmir.

*: Excluding Assam and Jammu and Kashmir.

Note: The column ‘All’ includes ‘other religions and persuasions’ and ‘religion not stated’ as well.

Source: 1991 Population and Percentage Shares: India, Registrar General, 1995.

Growth rates: Kulkarni, 1996.

 

TABLE 2

Population Size, Share and Growth Rates for Scheduled Castes, Scheduled Tribes and Others

 

Year/Period	All	Scheduled Castes	Scheduled Tribes	Non-Scheduled Castes/Tribes
	Population $ (in thousands)
1991	838,584	138,223	67,758	632,602
	Percentage Share *
1991	100.00	16.73	7.95	75.32
	Average Annual Exponential Growth Rate * (per cent)
1981	100.00	15.81	7.83	76.36
1981-91	2.13	2.70	2.28	2.00

$: Excluding Jammu and Kashmir.

*: Excluding Assam and Jammu and Kashmir.

Source: Computed from: India, Registrar General, 1993.

 

TABLE 3

Differentials in Indicators of Mortality and Fertility

 

Source	Measure	Religion/Caste/Tribe
		All	Hindu	Muslim	Christian	Sikh	SC	ST
1981 Census	IMR	115	122	92	67	75	NA	NA
	U5MR	152	155	135	97	108	NA	NA
NFHS - 1	IMR	86	90	77	50	47	107	91
1992-93	U5MR	119	124	106	68	65	149	135
NFHS -2	IMR	73	77	59	49	53	83	84
1998-99	U5MR	101	107	83	68	65	119	127
NFHS - 1	TFR	3.4	3.3	4.4	2.9	2.4	3.9	3.6
1992-93	CF	4.8	4.8	5.8	4.0	4.0	5.4	4.8
NFHS - 2	TFR	2.9	2.8	3.6	2.4	2.3	3.2	3.1
1998-99	CF	4.5	4.3	5.7	3.5	3.6	4.9	4.7

SC: Scheduled Castes; ST: Scheduled Tribes.

IMR: Infant Mortality Rate; U5MR: Under Five Mortality Rate = 1000 x 5q0

where 5q0 = probability of death before completion of age 5.

TFR: Total Fertility Rate; TMFR: Total Marital Fertility Rate.

CF: Cumulative Fertility = Mean children ever born for women of age 40-49.

NA: Not available.

Sources: 1981 Census: India, Registrar General, 1988.

NFHS-1: IIPS, 1995.NFHS-2: IIPS and ORC Macro, 2000.

 

TABLE 4

Differentials in Contraceptive Use and Reproductive and Child Health Care, NFHS-2

 

Measure	Religion/Caste/Tribe
	All	Hindu	Muslim	Christian	Sikh	SC	ST
Contraceptive practice, demand and need
CPR -Any	48	49	37	52	65	45	39
-Modern	43	44	30	45	55	40	35
-Sterilisation	36	38	20	39	32	36	32
TDFP	64	64	59	67	74	61	55
MN		75	77	63	78	88	73
71
Reproductive and child health
ANC	65	65	63	84	75	61	57
Institutional Deliveries	34	33	32	54	47	27	17
Professional Assistance	42	42	39	64	69	37	23
Child Immunisation	42	42	33	61	70	40	26

CPR: Contraceptive Prevalence Rate: per cent of couples of reproductive age using contraception.

– Any: current use of any method of contraception.

– Modern: current use of modern method of contraception.

TDFP: Total Demand for Family Planning: per cent of couples of reproductive age not wanting an additional child or not wanting a child in the next two years.

MN: per cent of need for contraception that is met.

ANC: per cent of pregnancies that received ante-natal care.

Institutional deliveries: per cent of deliveries conducted in health institutions.

Professional Assistance: per cent of deliveries conducted by health professionals.

Child Immunisation: per cent of children of age 12-23 months at survey who had received all recommended vaccinations.

Source: IIPS and ORC Macro, 2000.

 

Footnotes

* This paper arose out of discussions with Leela Visaria and Vimala Ramachandran of the HealthWatch Trust. Able technical support was provided by M. Sivakami.

1. The 1991 Census was not conducted in the state of Jammu and Kashmir, and the population figures in Table 1 are only for the enumerated areas. Since the 1981 Census was not conducted in Assam, percentage shares excluding both Assam and Jammu and Kashmir are given in the table for 1961 to 1991 censuses. The data from the 1961 and the 1971 censuses show that exclusion of Assam and Jammu and Kashmir raises the percentage share of Hindus by about 0.8 and lowers that of Muslims by 0.8-0.9. If the same correction were applied to the 1991 data, the share of Hindus would be about 81.6% and that of Muslims 12.5%. The percentages for other religions would change very little.

2. An issue raised frequently in India, especially at the political level, is the growth of the Muslim population. There are some apprehensions that due to higher than average growth, Muslims will soon be in a majority. While it is true that the Muslim growth rate has been higher than the national average in all the recent inter-censal decades, the difference has been small, close to half a percentage point. Over the period 1961-91, the average growth rate for Muslims was 2.75% (except the populations of Assam and Jammu and Kashmir) compared to 2.18% for the national population. Simple calculations show that even if the growth imbalance persists, Muslims would not be the majority population in the next two-three centuries, a point that has often been noted earlier (see, for example, Bhatia, 1990). Moreover, the growth differential is primarily on account of higher fertility among Muslims. But this has also been declining and as the transition progresses the gap in fertility and hence in growth rate would narrow down.

3. Differentials by religion and caste within states are not shown in the tables in order to save space. Tables on differentials in major states can be obtained from the author.

4. See footnote 3.

5. Estimates of the Total Fertility Rate (TFR) and the Total Marital Fertility Rate (TMFR) are also provided by recent censuses. These are generally under-estimates. Estimates adjusted using the Brass method (P/F ratio) are also available but not given here because, since fertility decline was in progress, such adjustments would be inappropriate.

6. Op cit., footnote 3.

7. Op cit., footnote 3.

8. Op cit., footnote 3.

 

References

Manoj Alagarajan and P.M. Kulkarni. 1998. ‘Fertility Differentials by Religion in Kerala: A Period Parity Progression Ratio Analysis’, Demography India 27(1): 213-228.

K. Balasubramanian. 1984. ‘Hindu-Muslim Differentials in Fertility and Population growth in India’, Artha Vijnana 26(3): 189-216.

P.S. Bhatia. 1990. ‘Population Growth of Various Communities in India – Myth and Reality’, Demography India 19(1): 121-130.

Expert Group on Population Policy, 1994. Draft National Population Policy. New Delhi: Ministry of Health and Family Welfare.

India, Department of Family Welfare, 1977. Family Welfare Programme in India Yearbook 1976-77. New Delhi: Ministry of Health and Family Welfare.

India, Department of Family Welfare, 1999. Family Welfare Programme in India Yearbook 1998-99. New Delhi: Ministry of Health and Family Welfare.

India, Department of Family Welfare, 2000. National Population Policy 2000. New Delhi: Ministry of Health and Family Welfare.

India, Registrar General. 1988. Census of India 1981: Child Mortality Estimates of India. Occasional Paper No. 5. New Delhi: Controller of Publications.

India, Registrar General. 1993. Census of India, Series 1 India, Final Population Totals: Brief Analysis of Primary Census Abstract. Paper 2 of 1992. New Delhi: Controller of Publications.

India, Registrar General. 1994. Census of India, Series 1 India, Primary Census Abstract, General Population, Part II, B(i) Vol. II. New Delhi: Controller of Publications.

India, Registrar General. 1995. Census of India 1991: Religion, Paper 1. Series 1 India, New Delhi: Controller of Publications.

International Institute for Population Sciences (IIPS). 1995. National Family Health Survey (MCH and Family Planning), India 1992-93. Bombay: IIPS.

International Institute for Population Sciences (IIPS) and ORC Macro. 2000. National Family Health Survey (NFHS-2), 1998-99: India. Mumbai: IIPS.

Anrudh K. Jain. 1998. ‘Population Policies that Matter’, in Anrudh K. Jain (ed.) Do Population Policies Matter? Fertility and Policies in Egypt, India, Kenya, and Mexico. New York: The Population Council: 1-18.

Roger Jeffrey and Patricia Jeffrey. 2000. ‘Religion and Fertility in India’, Economic and Political Weekly 35(35): 3253-3259.

P.M. Kulkarni. 1996. Differentials in the Population Growth of Hindus and Muslims in India, 1981-91. Monograph Series. No. 1. New Delhi: Population Foundation of India.

Asok Mitra. 1978. India’s Population: Aspects of Quality and Control. Vol. 2. New Delhi: Abhinav Publications.

K. Moulasha and G. Rama Rao. 1999. ‘Religion-Specific Differentials in Fertility and Family Planning’, Economic and Political Weekly 34(42): 3047-3051.

V.A. Pai Panandiker and P.K. Umashankar. 1994. ‘Fertility Control and Politics in India’, in The New Politics of Population, Jason L. Finkle and C. Alison McIntosh (eds.), Population and Development Review – A Supplement to Vol. 20: 89-104.

B.L. Raina. 1998. Population Policy. Delhi: B.R. Publishing Corporation.

J.R. Rele and T. Kanitkar. 1974. ‘Residence Background and Fertility in Greater Bombay’, Population Studies 28(2): 299-308.

Abusaleh Shariff. 1996. ‘Poverty and Fertility in Differentials in Indian States: New Evidence from Cross-Sectional Data’, Margin 29(1): 49-68.

K. Srinivasan. 1982. ‘Population Policy and Programme’, in Population of India: Country Monograph Series No.10. New York: U.N.-ESCAP.

United Nations. 1994. International Conference on Population and Development: Programme of Action. New York: United Nations.

Leela Visaria. 1974. ‘Religious Differentials in Fertility’, in Ashish Bose et al. (eds.) Population in India’s Development, 1947-2000. Delhi: Vikas Publishing House.

Pravin Visaria and Vijaylaxmi Chari, 1998. ‘India’s Population Policy and Family Planning: Yesterday, Today, and Tomorrow’, in Anrudh K. Jain (ed.) Do Population Policies Matter? Fertility and Policies in Egypt, India, Kenya, and Mexico. New York: The Population Council: 53-112.

Pravin Visaria and Leela Visaria. 1999. ‘India’s Population: Its Growth and Key Characteristics’. Occasional Papers in Sociology, No. 3. New Delhi: Institute of Economic Growth.