Bongaarts’ Model: Key Determinants Affecting Fertility Levels

Fertility doesn’t just rise or fall on its own. Behind every change in a country’s birth rate, there are specific behaviours and biological factors at work – when people marry, whether they use contraception, how long they breastfeed, and how often pregnancies end in induced abortion. In 1978, demographer John Bongaarts gave us a brilliantly simple way to measure exactly how much each of these factors pulls fertility down from its biological maximum. His framework, known as the proximate determinants model, remains one of the most widely used tools in population studies even today.

What are proximate determinants of fertility? 🔗

Demographers split the factors that influence fertility into two categories. Background determinants are the broad social and economic forces – education, income, urbanisation, religion, gender norms. Proximate determinants are the direct biological and behavioural channels through which those background factors actually translate into births. You can have rising female education in a state, but it only affects fertility if it changes when women marry or how they use contraception.

The idea originated with Kingsley Davis and Judith Blake in 1956, who listed eleven intermediate fertility variables. Bongaarts simplified this into a more workable set of four major proximate determinants – marriage or cohabitation, contraception, induced abortion, and postpartum infecundability. He argued these four explain most of the variation in fertility across populations.

The four key determinants in Bongaarts’ model 🔗

Marriage and sexual exposure 🔗

Bongaarts treated marriage as a proxy for regular sexual activity that could lead to pregnancy. If women in a society marry late, or if a large share remain unmarried during their reproductive years, fertility falls automatically – simply because the exposure to pregnancy is reduced. The original model assumed sexual activity happened mainly within marriage, which fit conservative societies well but has been refined over time as extramarital sexual activity has become more common in many populations.

Contraception 🔗

This determinant captures both the prevalence of contraceptive use among couples and the effectiveness of the methods they choose. Modern methods like sterilisation, IUDs, and hormonal pills are highly effective at preventing pregnancy, while traditional methods like withdrawal or rhythm are less so. As contraceptive prevalence rises in a population, fertility drops sharply.

Induced abortion 🔗

Even with contraception, some pregnancies are unintended and end in induced abortion. The abortion determinant measures how many births are averted because women choose to terminate pregnancies. Its impact depends not only on the abortion rate itself but on the average reproductive time “saved” per abortion, which Bongaarts and later researchers estimated carefully.

Postpartum infecundability 🔗

After childbirth, women experience a period of reduced fecundity because of lactational amenorrhea – breastfeeding suppresses ovulation. Cultural practices around postpartum sexual abstinence add to this. In societies where mothers breastfeed for long periods, the natural spacing between births can be quite wide, which significantly lowers total fertility even without contraception.

How the model works: indices and the equation 🔗

Bongaarts expressed total fertility as the product of a series of indices, each capturing the fertility-reducing effect of one proximate determinant. The classic equation is:

TFR = Cm × Cc × Ca × Ci × TF

Here, TFR is the observed total fertility rate, and TF is total fecundity – the theoretical maximum number of children a woman could have in the absence of all inhibiting factors. Bongaarts estimated total fecundity at roughly 15.3 children per woman on average. Each index sits between 0 and 1.

The index of marriage (Cm) 🔗

Cm measures the share of fertility lost because not all women of reproductive age are in a sexual union. It is calculated using the proportion of women currently married in each five-year age group, weighted by age-specific marital fertility rates. If every woman aged 15-49 were married, Cm would equal 1. The later women marry, or the more who stay single, the closer Cm gets to 0. In the most recent national analysis of India, marriage emerged as the single most important determinant, contributing about 36% of the reduction in fertility from the biological maximum.

The index of contraception (Cc) 🔗

Cc captures the fertility-reducing impact of contraception. It is computed from the proportion of married women using contraception and the use-effectiveness of each method. The formula adjusts for the fact that sterilisation prevents nearly all conceptions while traditional methods only reduce them partially. When contraceptive use is universal and methods are highly effective, Cc approaches 0; when no one uses contraception, Cc equals 1.

The index of induced abortion (Ca) 🔗

Ca expresses the ratio of the observed TFR to the TFR that would have occurred without induced abortion. The fewer abortions in a population, the closer Ca is to 1. Calculating this index is often the hardest part because reliable abortion data are scarce in many countries, including India, where unsafe and unrecorded abortions complicate the picture.

The index of postpartum infecundability (Ci) 🔗

Ci compares the average birth interval in the actual population to the interval that would exist without any postpartum infecundity (around 20 months versus a baseline of about 18.5 months without breastfeeding effects). Where women breastfeed for extended periods, Ci falls well below 1, meaning longer birth spacing reduces fertility considerably.

What the model reveals about India’s fertility transition 🔗

India provides a fascinating case for applying Bongaarts’ framework. The country has moved from a TFR of nearly 6 in the 1950s to 2.0 in 2019-21 according to the National Family Health Survey-5, dipping below the replacement level of 2.1 for the first time.

Marriage: still the dominant factor 🔗

Despite social change, marriage age remains the most powerful brake on fertility. NFHS-5 data show that 23.3% of women still get married before the legal age of 18, down from 26.8% in NFHS-4. As young women increasingly stay in education and enter the workforce, the median age at first marriage has risen, pushing Cm downward and substantially reducing fertility.

Rising role of contraception 🔗

Contraceptive use has grown significantly over the past two decades. Modern contraceptive use among married women rose from 48% in NFHS-4 to 56% in NFHS-5, with female sterilisation remaining the most popular method at 38%. The demand for family planning among married women aged 15-49 climbed from 66% to 76% in the same period. This is why studies applying the Bongaarts model to Indian data find that contraception’s contribution to fertility decline has been growing steadily.

Postpartum infecundability and abortion 🔗

Breastfeeding traditions remain strong across much of India, which keeps Ci low and helps space births. Induced abortion plays a smaller but real role; the Medical Termination of Pregnancy framework has been in force since 1971 and was expanded in 2021. Research applying the model finds that abortion’s contribution varies significantly by state, with higher impact in states like Kerala compared to negligible effects in others.

Applications and policy uses of the model 🔗

The real strength of Bongaarts’ framework is its ability to decompose change. If India’s TFR fell by 0.5 children between two surveys, the model tells policymakers exactly how much of that decline came from later marriages, how much from rising contraceptive use, and how much from changes in breastfeeding patterns.

Comparing regions and subgroups 🔗

The model is widely used to explain why fertility differs between states or social groups. Application of the Bongaarts model to NFHS data has produced state-level and sub-group estimates by age, education, residence, wealth and caste, showing how each determinant operates differently for women in different contexts. For instance, marriage timing matters more in northern states with earlier weddings, while contraception drives differences in the south where modern methods are more widespread.

Designing family planning programmes 🔗

Governments use the framework to target interventions. If contraception is already widespread but unmet need persists, expanding method choice helps. If early marriage is the main driver, raising the legal marriage age and girls’ education matter more. A study in Ethiopia found that contraceptive use was the single most important factor driving the fertility decline between 2005 and 2016, with its fertility-inhibiting effect rising from 15% to 37%, guiding the country to push for even higher contraceptive prevalence.

International comparisons 🔗

Because the indices are standardised and bounded between 0 and 1, the model allows clean comparisons across countries. A demographer can compare the Cm of Bangladesh with that of Pakistan, or Ci in India with that in Nigeria, and pinpoint where the real fertility differences originate.

Limitations and revisions of the model 🔗

No model is perfect. Bongaarts himself revised his framework in 2015 to address weaknesses that had become clearer over time. The original assumption that sexual activity happens only within marriage no longer fits many populations, so the marriage index was reconceptualised as an index of sexual exposure. Estimates of abortion-related birth aversion were adjusted, and the treatment of contraceptive effectiveness was updated to reflect modern data.

Other limitations remain. Reliable data on induced abortion are scarce, especially in countries where the procedure is restricted or stigmatised. Self-reporting of contraceptive use can be inaccurate. The model also focuses on the immediate channels and does not by itself explain why those channels change – for that, we still need theories of socioeconomic development, education, and women’s empowerment.

What do you think? Looking at your own state or community, which proximate determinant – late marriage, contraception, breastfeeding, or abortion – seems to play the biggest role in shaping fertility? And as India’s TFR slips below replacement, should the focus of population policy shift from reducing fertility to addressing the consequences of an ageing society?

References

1. https://pmc.ncbi.nlm.nih.gov/articles/PMC7276374/
2. https://www.demographic-research.org/articles/volume/33/19
3. https://www2.census.gov/software/proxdet/documentation/methodology.pdf
4. https://pmc.ncbi.nlm.nih.gov/articles/PMC8820640/
5. https://dhsprogram.com/pubs/pdf/OF43/India_National_Fact_Sheet.pdf
6. https://www.drishtiias.com/daily-news-analysis/nfhs-5-national-report
7. https://pmc.ncbi.nlm.nih.gov/articles/PMC10657051/
8. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0263532
9. https://pubmed.ncbi.nlm.nih.gov/32581604/