Ireland’s plunging birth rate and rising unemployment challenges narrative on economy

The CSO has just published two significant data sets , one relating to  the latest demographics and the other to the labour market in the second quarter of the year. The former was broadly as expected,  but the latter came as something of a shock, challenging the recent narrative on the economy.

The Irish population has been growing strongly again after a softer period following the financial crash and that trend continued in the twelve months to April, with a 65,000 increase (1.3%) taking the total to a fresh high of 4.92m. In general, official estimates and projections have tended to underestimate population growth, in large part because of the signifcance and volatility  of migration in the Irish data. In this case  the scale of emigration was largely unchanged from the previous year, at 55,000, but  offset by  substantial immigration of some 89,000, again similar to 2018, to give a net migration figure of 34,000. The latter is now larger than the natural increase ( births – deaths) which has slowed significantly  of late, to 31,000 from 49,000 at the turn of the decade. Indeed, the Irish birth rate ( births per 1,000 population)  although still very high by EU standards is falling rapidly , to 12.4 from 16.6 a decade ago.

Nonetheless, the robust pace of overall population growth allied to the boom in economic activity has resulted in  the emergence of massive capacity constraints across many areas of the economy, most nobaly housing but also in education, health care, public transport and the  infrastructure  around the main cities. The current make-up of that population growth ( a falling birth rate offset by  high net immigration ) makes it particularly difficult to plan for the future, however, as a period of weaker growth in Ireland might have a very significant impact on migration flows and hence total population.

The boom in the economy had propelled employment  to record highs and prompted much discussion about the level of full employment, as the unemployment rate had fallen to a cyle low of 4.5% in June. In that context the other main CSO release, the quartely Labour Force Survey, was a shocker. Headline employment fell marginally in q2  from the previous quarter but normally rises for seasonal reasons and so the adjusted figure saw a 21,000 decline, spread over half of the fourteen industry sectors categorised by the CSO. The labour force continued to rise  in the quarter so the numbers unemployed rose, as did the unemployment rate, to 5.2% from 5.1%. A marginal change in that context but it did result in a much more significant revision to the previously published monthly estimates, as unemployment in July is now 18,000 higher than previously thought and has been rising for the past four months, taking the unemployment rate to 5.3% instead of 4.6%.

Coverage of the data tended to emphasis the annual growth in headline employment, which is still impressive at 40,000 or 2.0%, although this is the slowest annual growth rate since early 2013. The seasonal  adjusted fall in q2 also followed a strong rise of 49,000 in q1 giving a net positive figure over the first half of the year of 28,000, so one might put all this down to a quirk in the data ( which is based on a household survey) particularly as the numbers claiming unemployment  benefit are still falling, albeit at a much reduced pace.

However, confidence surveys have weakened of late and some of the hard data has pointed to slower domestic activity, notably retail sales which fell by a cumulative 6.5% in the three months to July. One obvious culprit behind this caution is Brexit, which may also be impacting the demand side of  the  housing market. Indeed one other surprising feature of the Labour Force data was that construction employment has flatlined for the past nine months, which may be due to a scarcity of labour but also to caution from builders and developers as well.

All in all, food for thought- has unemployment bottomed in this cycle or is this just a short term hiccup which a Brexit resolution  might help to cure? On population, the Irish birth rate may still be the highest in Europe but is heading rapidly towards the EU average which is under 10.

Are estimates of Irish housing demand far too high?

The consensus view on Irish housing is that demand exceeds supply by a considerable margin and that the gap is closing, which many conclude is the key factor in the clear deceleration in residential property price inflation seen over the past year. Yet the increase in supply is not that pronounced and it may well be that demand is not as  substantial as generally believed, particularly into the medium term..

The latest figures on housing completions from the CSO, covering the second quarter of the year, show a half- year total of 9,100  which implies the full year figure will be above the 18,000 recorded in 2018,  albeit pointing to a outturn below 21,000. Supply is therefore still rising but at a sluggish pace, and certainly still a long way from the 30,000-35,000 widely seen as a good estimate of the annual demand over the medium term.

That figure is largely based on projections for household formation but there is an obvious circularity in that households are defined as occupying a house or an apartment. In other words  housing supply creates household formation so  the latter is not an independent estimate of demand. For example, the number of  households in Ireland rose by just 48,000 in the five years to the 2016 census, or by less than 10,000 per annum, but the numbers living in those households increased by 166,000, indicating a rise in the average household size. Housing demand projections  generally assume that the size of households will fall over the medium term.

So household formation averaging 30,000 or more a year implicitly assumes  housing supply around that figure , and lower supply would mean lower household formation. A look at the latest population projections by the CSO also gives food for thought. On the assumption of unchanged fertility and an annual net migrant inflow of 30,000 per year the population between 25 and 44 increases by just 8,000 in total by 2025, with the numbers between 30 and 40 years old declining sharply. If migration was much lower, at 10,000 per annum, and fertility declined, the CSO projection shows a 100,000 fall in the 25-44 age group which is the cohort one generally associates with house purchase and household formation.

If supply is the main determinant of demand  the recent data on planning permissions, showing a fall in the annual figure to below 29,000, casts doubt on whether supply will exceed 30,000 a year and , if so , household formation will be much lower than the received wisdom. The average number per household may in fact continue to rise .