Fertility, i.e. the basic ability to reproduce is at the root of reproductive health. Both infertility as well as high levels of fertility are bad for health. High prevalence of infertility has important psychological and sociological impact on the community. Very high levels of fertility means that women are devoting greater part of their lives to child bearing, and the exposure to maternal complications is high. In the contemporary Indian context, high levels of fertility is a major problem. Hence reduction in fertility levels is viewed as an indicator of improvements in reproductive health. The three common measures of fertility are; (a) Crude Birth Rate (CBR), (b) Age-specific fertility rates (ASFR), and (c) Total Fertility Rate (TFR). Crude birth rate (CBR) is the simplest of all and is defined as the number of births per annum per 1000 population. However, crude birth rate is influenced by the age-sex structure of the population apart from it's true fertility experience. The Age specific fertility rates, control for age-sex composition of the population. ASFR or age specific birth rate for a given age group is defined as the number of children born to women in the said age group per 1000 women in the same age group. In this males are excluded. So variation due to sex composition is controlled. ASFRs are usually computed by five year age group within the reproductive age. For example, ASFR 15-19, ASFR 20-24, ASFR 45-49, etc. This stratification by age group controls for age composition but there are seven pieces of estimate, to deal with. The TFR summarises the ASFRs and provides a single statistic of the fertility experience. Hence TFR is mostly used by demographers to analyse trends of fertility. TFR is the average number of children a woman would bear throughout her reproductive life, if she were to experience the contemporary age specific fertility rates. In this section we will first present available estimates of crude birth rate and then proceed to TFR to understand the fertility trend in AP. The decennial census, Sample Registration System (SRS), and NFHS-1&2 are the major sources of information on fertility trends. In addition, the Multiple Indicator Survey-2000 also gives estimates of fertility in Andhra Pradesh for the year 2000.

Figure 1 shows SRS of CBR estimates from 1970s till date. In AP the Crude Birth Rate has been steadily declining over the last 30 years, as is the case for the neighbouring states and the all India average. Around 1971-73 the CBR in AP was about 35 per 1000 population. It declined to about 25 per 1000 population during the 1990s. The decline of CBR in AP is slightly better than the all India average, but is less than some other south Indian states like the Kerala and Tamil Nadu. The decreasing trend of CBR is also corroborated by the estimates from the two NFHS conducted in 1992-93 (NFHS-1) and 1998-99 (NFHS-2) respectively. The two NFHS carried out six years apart show a decline in the CBR from 24.2 in 1992-93 to 21.4 in 1998-99. The CBR estimate of 23.2 per 1000 population in AP from the Multiple Indicator Sample Survey in 2000 is also consistent with the estimates from SRS (22.4 for 1998) and NFHS-2 (21.4 for 1998-99).

Figure 2 shows total fertility rate of Andhra Pradesh and other states from SRS. There is a constant decline of TFR from 1970s to 1998. The TFR of AP was 4.5 in 1971-73 and it has decreased to 2.4 in 1998. The southern states in India like Kerala and Tamil Nadu attained much lower fertility levels than AP (1.8 and 2 respectively). All India TFR was 5 in 1971-73 and declined to 3.2 in 1998 which is higher than AP. Madhya Pradesh shows the highest fertility rate among the neighbouring states of AP.

Figure 3 shows the TFR estimates from the two NFHS. The NFHS estimates show a higher level of fertility compared to SRS estimates. This could be either real or on account of the small sample size of NFHS. However, the time trend and the regional variations in TFR estimates from NFHS are consistent with SRS figures. For example SRS estimates show that TFR in Kerala and Tamil Nadu have been consistently lower than the TFR of AP. TFR of Karnataka has been more or less similar to that of AP. The data from two NFHS also give similar picture. The NFHS estimates also show a decline in TFR between NFHS-1 and NFHS-2. Multiple Indicator Sample Survey-2000 Andhra pradesh has collected fertility information. The TFR (calculated on the basis of the births that occurred during one year preceding the survey i.e., 1999-2000) in Andhra Pradesh is 2.3, 2.4, 1.9 for all AP, rural and urban areas respectively.

Figure 4 shows that fertility has been higher in rural areas of AP compared to the urban areas. In both rural and urban areas fertility has declined over the last three decades (1970s,80s and 90s). The sudden jump in rural area crude birth rate estimated for the year 1985 might be a sampling error. The decline in fertility appears to have been more sustained. The variation in fertility estimates for urban area may also be a result of uneven rural-urban migration. Around 1985 the TFR in urban areas went up a little bit and the decline after that appears to have been less than the fertility decline before 1980s. Overall, the rural urban gap appears to have reduced a little bit by the 1990s.

The NFHS estimates in Total Fertility rates in rural and urban areas of Andhra Pradesh show a similar (12 to 13%) decline from NFHS-1 to NFHS-2. The Urban TFR has decreased from 2.35 to 2.07 (12% decline from the NFHS-1 level) and the rural TFR has decreased from 2.67 to 2.32 (13% decline from the NFHS-1 level).

Although the state average of fertility has steadily declined, important regional differences exist with in the state. Starting with the 1981 census, fertility estimates are available by district. Figure 5 shows a plot of TFR in each district against the infrastructure development index computed by the Centre for Monitoring Indian Economy (CMIE). The TFR estimates are from 1991 census and the IDI is for the year 1995. Though the two estimates do not correspond to the same year their time location is close enough considering the fact that changes in TFR as well as IDI generally happens at a very slow rate. Except Hyderabad, which is an outlier, all other districts appear to fall into a pattern. The TFR tends to be lower as the IDI increases. This is consistent with our understanding that fertility tends to decline as socio economic development takes place. Districts like Mahaboobnagar, and Kurnool have lower levels of infrastructure development and high levels of fertility. Districts like Nalgonda, Medak and West Godavari have intermediate levels of infrastructure development but comparatively higher fertility levels. Strangely Hyderabad district has the highest infrastructure development index in the state and has very high levels of fertility. This discrepancy needs further investigation. One conjecture is that higher levels of fertility among the Muslims (Table 1) and presence of sizeable Muslim population in Hyderabad may be working to increase the TFR estimates for Hyderabad. Other possibilities are, (a) immigration of women from rural areas, (b) reporting of births in Hyderabad for families ordinarily residing in surrounding countryside but who might have made short term move to Hyderabad for the child birth, etc.

In Figure 6 we have plotted district wise estimates of TFR from 1991 census and female literacy. Here again Hyderabad is an outlier. All other districts fall into a pattern suggesting association between high literacy and low fertility.

NFHS-2 provides us with individual household level data relating fertility with various socio economic characteristics such as education, standard of living index etc. Table 1 shows that the TFR is 0.41 children higher among illiterate women than among women who have completed middle school. The TFR also varies by religion, being 0.33 children higher among Muslims and Christians than among Hindus. The TFR is 0.75 children higher among ST women and 0.51 children higher among SC women than among women in the other category. The TFR is 0.3 children higher among women living in households with a low or medium standard of living than among women living in households with a high standard of living. Note, however, that the NFHS sampling design was not primarily geared up to generate statistical estimates by religion and various other socio economic characteristics. Hence, some of the estimates may not be precise, if number of persons in the sample are less. 

                              Updated by Samatha Reddy Dated: 18/08/2003