The primary purpose of the national projections is to provide an estimate of the future population of the United Kingdom as a consistent framework for use in national planning by the UK government and the devolved administrations in Scotland, Wales, and Northern Ireland. The government actuary produces a new set of projections every two years. Among other things, the projections are used to assess demand for housing and to make long-term projections of the finances of the National Insurance fund.

The most recent, 1998-based, population projections were published in November 1999. Detailed results, including analysis and a description of the methodology used, were published in August this year. The population of the UK is projected to increase gradually from an estimated 59.2m in 1998 to reach 63.6m by 2021. The population is expected to peak at around 65m in 2036. As well as increasing in size, the population is ageing significantly.

Assumptions

The population projected for any future year is a function of the starting population, and the assumptions relating to births, deaths, and migration.

In general, the assumptions are held constant after a certain length of time, as it is not considered possible to predict trends satisfactorily in the long term. Migration has the shortest such horizon (three years), as it is the most uncertain of the components, followed by fertility (10 to 15 years). Mortality, the least unpredictable of the components, is assumed to continue improving throughout the projections, although the rate of improvement is assumed to have fallen to low levels after about 35 years.

In making long-term projections of this sort, the one certainty is that the assumptions will not be borne out precisely in practice. The likelihood of the actual experience matching the assumptions decreases the further into the future we attempt to look. The results must be interpreted with this uncertainty in mind.

Variant projections, using plausible ‘high’ and ‘low’ assumptions for fertility, mortality, or migration are also available. These give an indication of the sensitivity of the results to the assumptions.

Fertility

The average completed family size is the average number of children born to each woman over her lifetime. As it is a cohort measure (a cohort in this case is those women born in a certain year), the eventual figure for a given cohort is known only after about 45 years, when the women born in that year have passed childbearing age.

Average family size has fallen steadily from 2.45 children per woman for women born in the mid-1930s and will almost certainly fall below two children per woman for women born around 1960. The projections assume that the average completed family size will level off at around 1.8 for women born after 1975. This is around 15% below the ‘replacement level’ family size of 2.1 which would lead to the long-term ‘natural’ (that is, ignoring migration) replacement of the population.

Mortality

In formulating the mortality assumptions, the prospects for death rates at different ages are considered separately. It was assumed that the trends apparent during the period 196197 (of falling death rates at most ages) would initially continue at similar rates. However, it is unlikely that current rates of improvement (that is, the reduction in age-specific death rates) can continue indefinitely, and it is assumed that the improvements will gradually move towards 0.5% per year by the year 2032, and thereafter halve every ten years. For any ages where mortality rates are at present increasing, it is assumed that the trend will be reversed.

The assumptions imply an increase in the expectation of life at birth in the United Kingdom from 74.9 years for males and 79.7 years for females in 199899, to 78.5 for males and 82.7 for females in 202021. Note that the difference between life expectancies for the sexes falls from 4.8 years to 4.2 years over this period. Life expectancy is expected to be 79.9 for males and 84.0 for females in 205051.

Migration

The migration assumptions are the hardest to formulate, as net international migration to the UK tends to show considerable fluctuation from year to year. Migration data, generally obtained from surveys, is also less reliable than data on births and deaths, which must be registered.

Net migration into the UK is assumed to fall from the unusually high level of 185,000 in 199899 to a long-term level of 95,000 a year by 200102. This is similar to the average figure for the last ten years. It is significantly higher than the long-term level of 65,000 a year assumed in the previous 1996-based projections. This migration is divided evenly between males and females. Appropriate age distributions are applied to gross flows of migrants into and out of the UK.

Results

As a broad summary, we have a population with below replacement fertility, and increasing life expectancy. Together these imply an ageing population, which will eventually fall in size. Assumed net inward migration partly offsets the decline in total numbers, but has little effect on the ageing of the population.

Population size

Table 1 opposite shows the projected trend in the size of the UK population. The population is projected to increase in size gradually, from an estimated 59.2m in 1998, to reach 63.6m by 2021. This is equivalent to an annual rate of growth of 0.3%. Longer-term projections suggest that the population will peak around 2036 at almost 65m, and then gradually start to fall.

With the single exception of 1976, births have exceeded deaths throughout the 20th century. However, deaths are projected to outnumber births in about 30 years’ time. In the second half of the 2030s, this natural deficit is projected to exceed the assumed net gain from migration and so the population then begins to decline.

Just over half of the projected 4.4m increase in the population between 1998 and 2021 is directly attributable to migration. The remainder is attributable to the excess of births over deaths.

Age structure of the population

The mean age of the population is expected to rise from 38.6 years in 1998 to 41.9 years by 2021. The mean age eventually levels off at 44.2 years in 2044. The number of children aged under 16 is projected to fall from 12.1m in 1998 to 11.4m in 2010 and then to remain broadly at 11.3m over the following two decades. It then begins to fall, reaching 10.8m by 2051.

The number of people of working age (currently defined as between ages 16 to 64 for men and 16 to 59 for women) is projected to rise from 36.4m in 1998 to 38.3m in 2010. Allowing for the planned change in women’s state retirement age from 60 to 65 between 2010 and 2020, the working age population will rise further to 40.1m by 2021. Thereafter the number falls steadily, reaching 37.7m in 2051.

The number of people over pensionable age is projected to increase from 10.7m in 1998 to 11.9m in 2010. Also allowing for the change in women’s state retirement age, the population of pensionable age will rise only slightly further (to 12.2m) by 2021. However, a faster increase will then resume, with longer-term projections suggesting that the number over pensionable age will peak at nearly 16m around 2040. It is projected that by 2008, the number of pensioners will for the first time outnumber children.

The increases are fastest for the very elderly. The population aged 80 and over will grow from 2.3m in 1998 to reach 3.2m by 2021. Longer-term projections suggest it will then grow more rapidly, reaching 4.3m by 2031 and eventually peaking at a little under 6m by the middle of the century.

Dependency ratios for pensioners

Figure 1 shows the projected dependency ratio for pensioners, ie the population of pensionable age expressed as a percentage of the working age population. Note that this definition of the dependency ratio is somewhat arbitrary as retirement actually takes place over a range of ages. The dependency ratio is a measure of the number of retired people who must be supported by the working population. The pensionable age dependency ratio is projected to remain fairly stable until 2020, but then to rise rapidly as the population of pensionable age increases while that of working age falls.

Without the planned change in women’s state retirement age, the proportion of dependants would have risen earlier and further as indicated by the dotted line in Figure 1. At 2021, the projected pensionable age dependency ratio is 304 per 1,000 persons of working age, but would have been 373 per 1,000 had the state retirement age for women remained at 60.

Conclusion

The total population of the United Kingdom is increasing and ageing. Dramatic increases are expected at very old ages (ie 80 and over). However, largely due to the equalisation of state retirement ages for men and women, the dependency ratio for pensioners is expected to stay fairly stable over the next 25 years.

00_10_07.pdf