Population size alters slowly from year to year, but cumulative changes over decades can add up to big numbers. Governments encourage the study of longer-term demographic change – get it wrong and you end up with inadequate infrastructure, especially in slow-to-build sectors like water supply.

For any geographical territory, the change in population is the demographic identity B-D, plus or minus NM, where B is births, D is deaths and NM is net migration. If births roughly equal deaths and there is little migration, the population will be more or less constant.

While the birth and death rates do not fluctuate too much year to year, for small countries, net migration can at times be large, can swing from positive to negative quite rapidly and population forcasting is hazardous.

ADVERTISEMENT

The magnitudes for Ireland in recent years have been births at around 55,000, deaths about 35,000, so the population would rise by around 20,000 if net migration was zero. This gives an annual growth rate around one-third of 1% compounding each year, so the population would grow by about 4% every decade.

This is not fast enough to produce serious infrastructure gaps. In Ireland in recent decades the excess of births over deaths (called the natural increase) has been fairly easy to predict, but it has been dwarfed by the scale, and volatility, of net migration. Net migration has exceeded 100,000 on a few occasions and has also turned strongly negative – the population actually fell every year from 2010 to 2014 because the net outflow exceeded the natural increase.

This extreme volatility has meant that longer-term forecasts for Ireland have been a poor guide, including forecasts of only modest population growth made back in the 1990s by various people, the present writer included.

There was such a surge in net inward migration that population grew much more rapidly until the banking crash, then fell and is only recently back in a high growth pattern.

This is really a small country problem – for the world as a whole, net migration is zero, pending the exodus to Mars envisaged by Elon Musk, and predicting population change only requires forecasts of births and deaths.

The United Nations demography unit reckons the world’s total population at around 8.2bn currently and expects the figure to grow, at a declining rate, to a peak of around 10.4bn 60 years from now, despite steady falls in the birth rate and smaller families.

The reason is that death rates are falling even faster, especially in developing countries where growth has been rapid, reflecting improved healthcare and the gradual elimination of killer diseases.

The fastest growth is expected outside Europe, whose population is ageing as it is in Japan and other mature economies. But middle-income countries, notably China, also seem to have reached peak population.

The UN projections are produced every two years and are widely reported. They are also widely criticised, since extrapolations of births and deaths into the distant future are subject to very wide margins of error.

The UN experts acknowledge this and have recently been paring back their projections: successive revisions are producing lower UN figures for peak world population, and the year in which the peak is reached are not as distant as had earlier been expected.

What happens after the peak is reached? It is quite easy to manipulate assumptions to deliver a falling population after the peak – just assume that fertility rates continue to decline and that there is a limit to the life-extending achievements of medicine and healthcare. One recent study has extrapolated for several centuries and concluded that world population could, a few hundred years from now, shrink back towards well under a billion.

Since he is believed to have fathered at least 14 children, one of whom he brought along to a televised cabinet meeting in the Oval Office, there is a calculation doing the rounds amongst demographers that the world’s population in the year 2500 could consist largely of people descended from Elon Musk.

The Irish Government is revising the National Planning Framework and will need to make a stab at figures for future population given the implications for infrastructure planning and housing provision. On past experience the figures will be too high or too low, perhaps by a big margin, the further out the forecast horizon.

Given the dominant influence of net migration, impossible to predict over any extended period, it would be best not to rely on forecasts at all. Existing backlog, in housing, electricity, water supply and other areas, dwarfs the demand likely to arise from population growth over any sensible forecast horizon.

The backlog is well understood at this stage and its elimination requires policy changes. If these are prioritised, population growth will look after itself.