For 14 years after 1980 the Republic's birth rate fell rapidly. By 1994 the number of births annually had dropped by more than a third to below 48,000. Some experts believed it would fall further, to 45,000. Instead, it is now well over 53,000, and rising.
Demographic forecasting is very tricky. On the one hand, once people have been born, the proportion likely to survive to various ages can be predicted fairly accurately because death rates change only slowly - although medical advances could increase significantly, as will the number who will survive beyond the age of 80. On the other hand the marriage rate and the fertility rate are very hard to predict; for long-established trends in both can be suddenly reversed.
I believe such a reversal has just taken place here. We have recently passed a turning-point in relation to births, as well as emigration. But before discussing this birth-rate development, something should be said about the implications of the recent change in our external migration pattern.
In the past 30 years we have seen not just wild fluctuations in net emigration, but three actual reversals of the direction of this migration. Thus, in 1971 we moved into a period of net immigration, with Irish workers and their families returning in significant numbers. Then, in 1982, this was reversed and net emigration resumed, reaching a significant level between 1984 and 1990. More recently, there has been a further reversal.
Over the first four years of this decade the number of new emigrants was balanced by returning Irish people bringing with them children born abroad. But since the summer of 1995 there has been a big and growing net inflow, the bulk resulting from returning emigrants with children. I estimate that something approaching one-quarter of new jobs being created are now being filled by these returning emigrants.
This historic change in the pattern of Irish migration, which looks like being sustained, carries considerable implications for the economy. I doubt if these have yet been properly incorporated into our public policy decision-making.
To track the scale of this immigration, other types of flow must be distinguished. For, even leaving tourists aside, there is a constant movement back and forth of people who stay away for a year or more, seeking education, training and experience (some 16,000 Irish young people are currently studying in the UK, including Northern Ireland).
Such students starting and completing their courses, together with people going to or returning from relatively short periods of work to gain experience, seem recently to have accounted for an annual flow of almost 20,000 in both directions.
Setting aside this two-way movement - which, because it is balanced should have no implications for housing demand - we find that in the 12 months ended March last year, the outflow of young people seeking more or less permanent jobs elsewhere was significantly lower than in 1992-94. Simultaneously, those of working age settling here have more than doubled since 1994 to something like 17,500 a year. These two factors combined have led to the sudden emergence of substantial net immigration in the past two years.
A small proportion of these recent immigrants will have been refugees, and some others may, of course, have been foreigners bringing special skills, including language skills. But, excluding the refugees, and allowing for the fact - evident from immigration data in the 1996 census - that over half of these returning emigrants are single (although some of these younger, single returnees may have gone back to live with their parents) the additional housing demand generated by immigrants must have been well over 10,000.
This is probably double the additional annual demand derived from this source from 1992 to 1994.
Theoretically, some of this additional demand could be met by dwellings vacated by the net outflow of younger emigrants. But most of these younger emigrants are probably leaving from their parents' homes and thus do not vacate accommodation. In any event, the number of such emigrants is now some two-fifths lower than in 1992-1994, thus reducing whatever small supply of accommodation that may in the past have become available from this source.
It is clear, therefore, that a significant part of our present housing crisis derives from the emergence and growth of immigration, which mainly takes the form of emigrant Irish workers returning to work here. And, in the short term at least, this source of additional housing demand seems more likely to grow than fall.
This new immigration pattern also has other policy implications. In the past seven years the child population has been added to by an inflow of almost 35,000 children born elsewhere and brought back to Ireland by their immigrating parents. This inflow has already significantly slowed the numerical decline in schoolgoers.
But future school-student numbers will also be affected by another crucial demographic development of the last couple of years - the recovery in the birth rate that started in early 1995 and has gone on since the summer of 1996. In the 12 months between then and last summer, the number of births jumped by about a tenth.
Half of this 12-month rise was accounted for by a further increase of over 15 per cent in non-marital births. These have risen throughout the whole of the past 36 years at an average annual rate of 7 per cent. Today's level is 14 times greater than in 1961.
What is new is that since mid-1996 the marital birth rate has also started to rise again, for the first time in almost 20 years. And only a fraction of this increase was accounted for by a rise in the number of women in the key child-bearing age-groups. This means last year there was a sharp rise in marital fertility, radically reversing the trend of decades.
It may be said too much should not be read into data relating to such a short time span. But given this development has now extended over five successive quarters, and also given its scale - a jump of almost 6 per cent in marital births over the latest 12-month period for which figures are available - I do not think this happening should be seen as a flash in the pan, although it may not last indefinitely.
While up-to-date details of the age of mothers at birth in this most recent period are unavailable, earlier figures suggest this increase in births has probably been largely concentrated in the 30-39 age group and represents births postponed by working mothers. This is a process that could continue for quite a few years.
Something similar did happen throughout most of northern Europe in the latter part of the 1980s, when there was a recovery in the marital birth rate lasting five or six years in most cases. In Denmark and Norway it seems to have gone on for a decade or more.
If our birth rate remains for some time to come at its current high, and if the present level of net immigration continues, this combination of new demographic factors would soon slow and then halt the fall in child population numbers - 175,000 in the past 12 years. In that event, by 2011, the number of children could, I believe, be some 75,000 higher than was estimated in last year's medium-term review population projections.
These new trends, of course, could end. But it is equally possible they will accelerate - with net immigration and the birth rate increasing even further in years ahead. It is only possible to say at this stage that pending further developments it would be unwise to base policy decisions on an assumption that our child population will fall by another 100,000, as projected in the medium-term programme.
There is anecdotal evidence of a recent recovery in the marriage rate which, if correct, could derive from a postponement of marriage in earlier years. Up to mid-1997 there was no sign of such a development, but a significant increase in the year's third quarter may yet represent the first stage of such a change. But we must wait a while before we can know whether, as in the case of emigration and the birth rate, we have just passed a milestone in our marriage pattern.