Skip to content
Back to search results

Bongaarts Proximate Determinants

Introduced by
John Bongaarts
Year
1978
Type
Multiplicative decomposition of fertility into behavioural and biological factors
Family
Fertility determinants framework
Font size

What does it do?

The Bongaarts framework of the proximate determinants of fertility decomposes a population's fertility into a biological maximum reduced by a small set of directly fertility-inhibiting factors: the proportion of women in sexual unions, contraceptive use, induced abortion, and postpartum infecundability.

By expressing observed fertility as total fecundity multiplied by four indices between zero and one, it quantifies how much each behavioural and biological channel suppresses fertility below its potential ceiling.

The intuition behind the method

Every distal cause of fertility — income, education, religion, culture, policy — can only affect childbearing by working through a handful of immediate, biological-behavioural valves: whether women are exposed to intercourse, whether they use contraception, whether pregnancies are aborted, and how long breastfeeding suppresses ovulation after a birth. Bongaarts reasoned that fertility starts from a biological maximum of roughly 15 children per woman and is throttled down by these four valves in sequence, each multiplying the rate by a factor between zero (complete suppression) and one (no effect). The framework is an accounting device that turns the question 'why is fertility this level?' into 'which valve is doing the throttling, and by how much?'

History

John Bongaarts, a demographer at the Population Council, introduced the framework in 'A framework for analyzing the proximate determinants of fertility' in Population and Development Review in 1978, building on Kingsley Davis and Judith Blake's 1956 enumeration of the intermediate fertility variables but reducing their long list to four quantifiable indices. It was refined in subsequent work by Bongaarts and Robert Potter, including their 1983 book on fertility biology and behaviour, and was widely applied to Demographic and Health Survey data through the 1980s and 1990s to explain fertility transitions across the developing world. Later revisions by Bongaarts and others updated the indices to address non-marital exposure and contraceptive measurement. The framework is treated as a standard analytical tool in demographic texts such as Preston, Heuveline, and Guillot's Demography (2001).

When to use it

Use the Bongaarts proximate-determinants framework when you want to explain why fertility is at a given level, or why it has changed or differs between populations, in terms of the immediate behavioural and biological mechanisms rather than distal socioeconomic causes. It is the standard tool for attributing fertility decline to rising contraceptive use versus changing marriage patterns versus shortening breastfeeding. It requires survey-based measures of the proportion in unions, contraceptive prevalence and effectiveness, abortion rates, and postpartum infecundability, typically from sources like the Demographic and Health Surveys. It assumes the four indices are the dominant proximate channels and that they act multiplicatively and roughly independently. It is weaker where abortion is severely underreported, where exposure outside formal marriage is common (requiring sexual-union rather than marriage definitions), and where data on breastfeeding or effectiveness are poor.

Limitations and considerations

  • Relies on a constant total-fecundity ceiling and fixed empirical constants (e.g., 1.08, 0.4) that may not hold across all populations.
  • Induced abortion is frequently underreported, so the abortion index is often the weakest and most uncertain component.
  • The marriage index assumes childbearing is confined to unions, which understates exposure where non-marital childbearing is common.
  • Treats the four indices as independent and multiplicative, masking interactions (e.g., between breastfeeding and contraception) that the simple product ignores.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Reduces the explanation of fertility levels to four interpretable, measurable channels, each scaled cleanly between zero and one.
  • Multiplicative structure makes it easy to attribute fertility differences and decline to specific determinants and to quantify their relative contributions.
  • Anchored in a near-universal biological maximum, enabling meaningful comparison of inhibition across very different societies.
  • Bridges socioeconomic and biological perspectives by separating distal causes from the proximate mechanisms through which they must act.

Applications

  • Attributing fertility decline in developing countries to rising contraceptive use versus later marriage versus changing breastfeeding.
  • Comparing across countries which proximate determinant accounts for differences in fertility levels.
  • Projecting the fertility impact of family-planning programs by modelling changes in the contraception index.
  • Analyzing the role of induced abortion in fertility regulation where reliable abortion data exist.
  • Linking socioeconomic development to fertility by tracing its effects through the four proximate channels.

How it works

How it works

1

The framework begins from total fecundity (TF), the average number of births a woman would have if she were continuously married throughout her reproductive years, used no contraception, had no induced abortions, and did not breastfeed. Empirically this biological ceiling is remarkably stable across populations at roughly 15 to 15.3 births per woman, providing a common starting point against which inhibition is measured.

Formula
\text{TF} \approx 15.3 \ \text{(potential births absent all inhibitors)}
2

The index of marriage Cm captures the fertility-reducing effect of women not being in a sexual union. It is the ratio of the age-specific marital fertility actually 'exposed' to the marital fertility that would occur if all reproductive-age women were married, computed by weighting the proportions married m(a) by the age pattern of marital fertility g(a). Cm equals one if all women of childbearing age are in unions and falls toward zero as exposure declines.

Formula
C_m = \frac{\sum m(a) g(a)}{\sum g(a)}
3

The index of contraception Cc measures the reduction from contraceptive use among married women, as a function of the prevalence of use u and the average use-effectiveness e, with the 1.08 constant adjusting for the sterilizing effect of contraceptive sterilization and for non-users who are sterile. Cc equals one when no one contracepts and approaches zero as effective contraceptive use rises toward universal coverage.

Formula
C_c = 1 - 1.08\, u\, e
4

The index of induced abortion Ca quantifies the births averted by induced abortion, relating the observed total fertility rate to the total abortion rate TA, where each abortion is assumed to avert roughly 0.4(1+u) births because it removes a pregnancy that would otherwise have led to a birth after some months of exposure. Ca equals one in the absence of abortion and decreases as the abortion rate rises.

Formula
C_a = \frac{\text{TFR}}{\text{TFR} + 0.4(1+u)\,\text{TA}}
5

The index of postpartum infecundability Ci captures the suppression of fertility by the period of anovulation and amenorrhea following a birth, driven mainly by breastfeeding intensity and duration. It relates the average birth interval with breastfeeding to the interval that would prevail without it, using the average duration of postpartum infecundability i (in months); long breastfeeding lowers Ci substantially, while no breastfeeding leaves it near one.

Formula
C_i = \frac{20}{18.5 + i}
6

Observed total fertility is reconstructed as the product of total fecundity and the four indices: TFR = Cm · Cc · Ca · Ci · TF. Because the indices multiply, the framework cleanly attributes the gap between potential and actual fertility to the combined throttling of marriage, contraception, abortion, and lactational infecundability, and comparing the indices across populations or time shows which determinant drives a fertility difference or decline.

Formula
\text{TFR} = C_m \cdot C_c \cdot C_a \cdot C_i \cdot \text{TF}

Common pitfalls

  • Using the marriage index based on formal marriage in settings with widespread cohabitation or non-marital sexual activity, biasing exposure downward.
  • Trusting the abortion index where induced abortion is illegal or stigmatized and therefore badly underreported, distorting the whole decomposition.
  • Treating the empirical constants as universal truths rather than calibrated averages that can drift across populations and eras.
  • Forgetting the indices interact — for example, the decline of breastfeeding and the rise of contraception partly offset each other — so reading them as fully separable can mislead.

Frequently asked questions

What are the four proximate determinants and why only four?

Bongaarts identified four channels that account for almost all variation in fertility across populations: the proportion of women in sexual unions (exposure), contraceptive use, induced abortion, and postpartum infecundability driven by breastfeeding. Davis and Blake had listed eleven intermediate variables in 1956, but Bongaarts showed empirically that these four dominate, with the remaining factors (such as spontaneous intrauterine mortality and natural sterility) folded into the total-fecundity ceiling. Reducing the framework to four measurable indices is what made it practical for routine analysis of survey data.

Why is total fecundity treated as roughly constant across populations?

Total fecundity is the number of births a woman would have absent all the inhibiting behaviours — no contraception, no abortion, continuous union, no breastfeeding. Because it reflects underlying reproductive biology rather than behaviour, empirical estimates cluster tightly around 15 to 15.3 births per woman across diverse populations. Holding it fixed gives the framework a common biological yardstick, so that all observed differences in fertility are attributed to the four behavioural and lactational indices rather than to biological variation, which is small by comparison.

How does this framework relate to socioeconomic explanations of fertility?

Distal factors like income, education, urbanization, and culture cannot affect fertility directly; they must operate through the proximate determinants — for instance by changing when women marry, whether they contracept, or how long they breastfeed. The Bongaarts framework therefore complements socioeconomic analysis: socioeconomic models explain why the proximate determinants take the values they do, while the proximate-determinants framework translates those values into a fertility level. Together they form a two-stage causal chain from social context through immediate mechanisms to observed childbearing.

Related methods

Sources

  1. Bongaarts, J. (1978). A framework for analyzing the proximate determinants of fertility. Population and Development Review, 4(1), 105–132.

    Open publication reference
    Copied.
  2. Preston, S. H., Heuveline, P., & Guillot, M. (2001). Demography: Measuring and Modeling Population Processes. Blackwell.

    ISBN: 9781557864512
    Copied.