The dependency ratio is a demographic indicator that compares the size of the population presumed economically inactive—children and the elderly—with the population presumed to be of working age. Its conceptual basis lies in the age-structure analysis developed by demographers in the mid-twentieth century and standardised through the United Nations Population Division, whose World Population Prospects series remains the authoritative global source. The measure rests on a convention rather than a statute: the working-age band is defined as 15 to 64 years, the young dependent band as 0 to 14, and the old dependent band as 65 and above. National statistical offices, including India's Office of the Registrar General and Census Commissioner and the United States Census Bureau, report the ratio using these thresholds, allowing cross-country comparison while acknowledging that actual labour-force entry and retirement ages vary by economy.
The procedural mechanics are arithmetic and transparent. The total dependency ratio is computed by summing the population aged 0–14 and the population aged 65 and over, dividing that sum by the population aged 15–64, and multiplying by 100. A ratio of 50, for example, indicates that every 100 working-age persons support 50 dependants. The figure is conventionally disaggregated into two components. The child dependency ratio isolates the 0–14 cohort over the 15–64 cohort, while the old-age dependency ratio places the 65-and-over cohort over the same denominator. The two component ratios sum to the total, which permits analysts to attribute movements in the overall figure to either declining fertility, which shrinks the child component, or rising longevity, which inflates the old-age component.
Several variants refine the basic measure. The economic dependency ratio substitutes the actually employed population for the working-age band, capturing unemployment and labour-force non-participation that the purely age-based ratio ignores—relevant where female labour-force participation is low, as in much of South Asia and the Gulf. Some analysts adjust the upper threshold to 60 in countries where statutory retirement falls earlier, which raises the measured old-age burden. The potential support ratio, the inverse of the old-age dependency ratio, expresses how many working-age persons exist per person aged 65 and over and is favoured in pension-system modelling. UN reporting frequently presents all three.
Contemporary figures illustrate divergent trajectories. Japan, with the world's oldest population, recorded an old-age dependency ratio exceeding 50 by the early 2020s, meaning fewer than two workers per elderly person, a pressure that has driven successive reforms by the Ministry of Health, Labour and Welfare. India, by contrast, reported a total dependency ratio that fell below 50 over the 2011–2021 decade as fertility approached replacement level, a configuration NITI Aayog and Census projections cite as the country's window of demographic advantage. Niger sustains a total ratio above 100, driven almost entirely by its child component and persistently high fertility, while the European Union as a bloc, tracked by Eurostat, projects its old-age ratio rising toward 50 by 2050.
The dependency ratio must be distinguished from the demographic dividend, with which it is frequently conflated. The dividend is the accelerated economic growth that can occur when the dependency ratio falls and the working-age share peaks; the ratio is the measurement, the dividend the contingent outcome that materialises only if the working-age population is educated, healthy, and employed. It also differs from the median age and the total fertility rate, which describe age structure and reproductive behaviour respectively without expressing the support burden. Crucially, the ratio is an age-based proxy, not a measure of actual economic dependence: a 16-year-old in full-time education is counted as productive while a 30-year-old outside the labour force is not.
The measure carries known limitations and live controversies. Because it treats everyone aged 15–64 as productive, it overstates the economically active population in societies with extended education, early retirement, or high informal-sector unemployment, and it understates dependency where child labour or elderly self-support persist. Demographers increasingly favour the prospective old-age dependency ratio, advanced by Warren Sanderson and Sergei Scherbov, which redefines old age by remaining life expectancy rather than a fixed 65, producing markedly lower elderly burdens in long-lived populations. The COVID-19 pandemic's mortality among older cohorts and its disruption of fertility also complicated short-run projections issued after 2020.
For the working practitioner, the dependency ratio is a first-order input into fiscal, pension, and development policy rather than an academic abstraction. Desk officers assessing a country's medium-term stability read a rising old-age ratio as a signal of mounting pension and healthcare liabilities and constrained labour supply, while a falling total ratio flags a potential growth window that hinges on employment policy. Diplomats negotiating migration, trade, and development frameworks use the indicator to identify structural complementarities—ageing capital-rich economies seeking labour against young labour-surplus economies. Read alongside fertility, participation, and education data, the ratio remains an indispensable shorthand for the demographic pressures shaping national budgets and foreign-policy posture across the coming decades.
Example
In the early 2020s Japan's Ministry of Health, Labour and Welfare reported an old-age dependency ratio above 50, meaning fewer than two working-age people supported each resident aged 65 and over.
Frequently asked questions
Sum the population aged 0–14 and the population aged 65 and over, divide by the population aged 15–64, and multiply by 100. A result of 50 means 50 dependants for every 100 working-age persons. It splits into a child component and an old-age component that add to the total.
Keep learning