The Census 2016 results show that Ireland has two million homes, a net increase of just 9,000 compared to 2011. However, the official data on house completions indicated that about 50,000 new homes were completed over this interval.
Does this mean house completion figures, which are largely based on ESB connections, are wrong, as some have maintained?
A closer examination of the Census data suggests that they are probably broadly consistent with the official housing completions statistics.
The use of the Census to collect data on the stock of housing happened almost by accident. As part of the quality control on the Census data, from 1991 onwards the Central Statistics Office (CSO) asked enumerators to record all habitable houses even if they were unoccupied on Census night. The CSO wanted to ensure that nobody was left out of the Census.
As the potential policy significance of the housing stock data became clearer from 2006 onwards, the CSO instructed its enumerators to pay particular attention to the vacant housing count.
The Census also asks residents to give the age of their home. Unfortunately, in the latest Census there was a significant increase in the number of households that don’t know this. These are usually people in rented accommodation – home owners generally have more reason to know how old their house or apartment is. So it’s hardly surprising that, with a big increase in numbers renting, there was also an increase in these “don’t knows”.
In the 2016 Census, a total of 33,000 households said that they were living in accommodation built since 2011, while the official completion figures suggest around 50,000 dwellings were built over the same period. In Dublin there is an even closer match between what residents say was built since 2011 (8,500), and what the housing completion figures show (11,000).
While there is not an exact match between them, the two datasets do appear to be broadly consistent. The CSO knows the precise location (geocodes) of all dwellings in the two censuses, and it should be able to provide a more detailed reconciliation of the data in the future.
Dilapidation
New home-building in any period is, of course, offset by any losses of housing units over that same time – obsolescence. Such losses can take place through dilapidation, through demolition for replacement, or the conversion of houses containing bedsits to single-family homes.
Using the Census data on the age of the housing stock it is generally possible to estimate how many units of housing were lost over the period from the previous Census. I did this exercise for the Census periods from 1961 to 2002 in a paper published in 2005, finding an annual obsolescence rate ranging between 0.4 per cent and 1 per cent.
There are a number of factors which would have driven a higher obsolescence rate between 2011 and 2016, providing an explanation for much of the discrepancy between official housing construction data and the net growth in the housing stock recorded by the Census.
Increased dilapidation of houses in “ghost estates” almost certainly meant some reclassification of vacant dwellings from “habitable” in 2011 to “uninhabitable” in 2016, accounting for some increase in the rate of obsolescence
Regulations banning bedsits, which had a four-year lead-in period, finally came into effect in early 2013. Nationally there was a reduction of 2,400 bedsits recorded between Census 2011 and Census 2016, with Dublin accounting for the bulk of the fall. With so many single people homeless in Dublin today, it must be questioned whether 2013 was the right time to ban accommodation, albeit of inferior quality, which catered for single people on low incomes.
Good news
The Census has the good news that the vacant homes rate fell from 14.5 per cent to 12.3 per cent over the last five years, but there are still over 180,000 unoccupied homes other than holiday homes.
The State could usefully examine factors that may give rise to excessive vacancies, particularly in areas of high housing demand.
Other countries have successfully harnessed upper floors in commercial streets for housing – we could learn from their fire safety approaches.
The current absence of bridging finance impedes older people from trading down, releasing larger properties.
Some adjustments to the Fair Deal scheme might encourage the rental of properties whose owners are in nursing homes.
Speeding up the probate process could facilitate earlier sales after an owner’s death.
Understanding why properties stay vacant may bring us nearer to finding workable solutions to the housing shortage.
