Tempo-Adjusted Fertility
- Introduced by
- John Bongaarts & Griffith Feeney
- Year
- 1998
- Type
- Tempo-and-quantum adjustment of the period total fertility rate
- Family
- Fertility tempo measures
What does it do?
Tempo-adjusted fertility is the Bongaarts-Feeney correction of the period total fertility rate that removes the distortion introduced when the timing of childbearing changes.
When women collectively postpone (or advance) births, the conventional period total fertility rate is artificially depressed (or inflated) even if the number of children women ultimately have is unchanged; the adjustment inflates each birth-order-specific rate by a factor based on the changing mean age at childbearing to recover an undistorted measure of fertility quantum.
The intuition behind the method
Imagine measuring how much water flows through a river in a year by reading the gauge each day. If everyone upstream suddenly decides to hold back water in reservoirs for a while, the gauge reads low this year — not because less water exists, but because its timing shifted. The period total fertility rate is exactly this gauge for births: when a cohort of women collectively delays childbearing, fewer births are recorded in the current year, so the rate dips below the true number of children women will have. Bongaarts and Feeney showed that the size of this distortion depends on how fast the average age of childbearing is moving, and that dividing by one minus that speed restores the reading the gauge would have shown with stable timing.
History
John Bongaarts of the Population Council and Griffith Feeney published 'On the quantum and tempo of fertility' in Population and Development Review in 1998, formalizing the long-recognized intuition that period fertility rates are distorted by shifts in the timing of births. Their tempo adjustment sparked an extended methodological debate through the 2000s, with contributions and critiques from demographers including Norman Ryder (whose translation/demographic-translation work prefigured the issue), Tomas Kohler, Joshua Goldstein, and Wolfgang Lutz, and refinements such as the variance-adjusted version. The episode reshaped how low fertility in Europe and East Asia was interpreted. The quantum-tempo framework is now a standard topic in advanced fertility analysis and demographic texts such as Preston, Heuveline, and Guillot's Demography (2001).
When to use it
Use tempo-adjusted fertility when the period total fertility rate is being driven up or down by changes in the timing of childbearing rather than by changes in the number of children women have, and you want a quantum measure undistorted by tempo. It is most valuable in low-fertility societies undergoing sustained postponement, where the raw period TFR understates underlying fertility. It requires order-specific age-specific fertility rates and reliable mean ages at childbearing by order for adjacent years. The method assumes that the order-specific fertility schedules shift without changing shape and that the rate of mean-age change is smooth; it is unreliable when these assumptions fail, when data are noisy, or when fertility quantum and tempo are changing simultaneously and rapidly. It is a diagnostic complement to the raw TFR, not a replacement for it.
Limitations and considerations
- Assumes order-specific fertility schedules shift without changing shape; real shape changes (e.g., compression of the schedule) violate this and bias the adjustment.
- Highly sensitive to the estimated rate of change in the mean age, which is noisy and can produce volatile or implausible adjusted values.
- Requires good-quality birth-order-specific data, which many statistical systems do not collect reliably.
- Corrects only for tempo in the mean; it does not capture distortions from changes in the variance or higher moments of the schedule.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Separates fertility quantum (how many) from tempo (when), clarifying whether a low period TFR reflects fewer children or merely delayed ones.
- Operates birth-order by birth-order, correctly capturing that postponement differs across first, second, and higher-order births.
- Uses only order-specific rates and mean ages already derivable from standard birth and population data.
- Provides an early signal that a depressed period TFR may rebound as postponement ends, important for population projection.
Applications
- Reinterpreting very low period TFRs in Southern and Eastern Europe and East Asia as partly an artefact of childbearing postponement.
- Estimating the likely rebound in period fertility once a wave of postponement ends, for use in population projections.
- Disentangling whether declining births reflect smaller completed families or merely delayed timing.
- Comparing the quantum of fertility across countries at different stages of the postponement transition.
- Informing pronatalist and family-policy debates by distinguishing genuine fertility decline from timing shifts.
How it works
How it works
The adjustment is applied separately by birth order (first births, second births, and so on) because postponement operates differently at each parity. For each order i, the period total fertility rate TFR_i(t) is computed in the usual way as the sum across ages of the order- and age-specific fertility rates in year t, decomposing total fertility into its parity components.
Formula
\text{TFR}_i(t) = \sum_x f_i(x,t)For each birth order, the pace of timing change is measured by r_i(t), the annual rate of change in the mean age at childbearing for that order, estimated as the centred difference of the order-specific mean ages between the surrounding years. A positive r means childbearing is being postponed; a negative r means it is being advanced.
Formula
r_i(t) = \frac{\mu_i(t+1) - \mu_i(t-1)}{2}Each order-specific period rate is divided by (1 - r_i(t)). When childbearing is postponed (r positive), this factor exceeds one and inflates the rate to undo the artificial depression caused by births sliding into later years; when childbearing is advanced (r negative), the factor is below one and deflates the rate. The adjustment assumes the shape of the fertility schedule for each order stays constant while the schedule shifts bodily along the age axis.
Formula
\text{TFR}'_i(t) = \frac{\text{TFR}_i(t)}{1 - r_i(t)}The adjusted order-specific rates are summed across birth orders to give the tempo-adjusted total fertility rate TFR'(t) for the year. This figure estimates the quantum of fertility — the level that would have been observed in the absence of timing change — and is interpreted as a cleaner indicator of underlying period fertility than the raw TFR during periods of postponement.
Formula
\text{TFR}'(t) = \sum_i \text{TFR}'_i(t)Common pitfalls
- Applying the adjustment to the all-order TFR directly instead of order by order, which mixes incompatible timing shifts and biases the result.
- Trusting a single year's adjusted value, which can swing wildly because the mean-age change is differenced from noisy data — smoothing across years is essential.
- Interpreting the tempo-adjusted TFR as a forecast of completed cohort fertility, when it is a hypothetical period quantum under constant-shape assumptions.
- Ignoring that the constant-shape assumption fails when the spread of ages at childbearing is also changing, leaving residual distortion uncorrected.
Frequently asked questions
Why does postponing births lower the period total fertility rate?
The period total fertility rate sums the age-specific fertility rates observed in a single year, treating them as if a synthetic cohort experienced them. When women across society shift their childbearing to older ages, the births that would have occurred this year are pushed into future years, so the rates observed now are temporarily thinner at every age. The synthetic-cohort sum therefore falls below the number of children women will actually have — a tempo distortion, not a real drop in quantum. When postponement stops, the rate recovers, sometimes overshooting, even though completed family size never changed.
Why must the adjustment be done separately by birth order?
Postponement and its pace differ across parities: the timing of first births may be moving rapidly while higher-order births shift more slowly, and the mean ages and their rates of change are order-specific. Applying a single adjustment factor to all-order fertility would blend these distinct timing shifts and distort the correction. Computing an order-specific change in mean age and adjusting each order's TFR before summing keeps each parity's tempo distortion handled with its own factor, which is central to the Bongaarts-Feeney formulation.
Is the tempo-adjusted TFR the same as completed cohort fertility?
No. The tempo-adjusted TFR is a hypothetical period quantum: it estimates the level the period rate would show if there were no change in the timing of births this year, under the assumption that order-specific schedules shift without changing shape. Completed cohort fertility, by contrast, is the actual average number of children born to a real cohort of women over their lifetimes. The two are related but distinct, and they can differ when the constant-shape assumption fails or when quantum itself is changing across cohorts.
Related methods
Sources
Bongaarts, J., & Feeney, G. (1998). On the quantum and tempo of fertility. Population and Development Review, 24(2), 271–291.
Open DOI referenceCopied.Preston, S. H., Heuveline, P., & Guillot, M. (2001). Demography: Measuring and Modeling Population Processes. Blackwell.
ISBN: 9781557864512Copied.