Comparing estimates of underlying housing demand with actual housing supply is important for understanding developments in housing affordability and identifying potential economic imbalances. For example, past work has suggested that overbuilding in the residential construction sector in the years before 2008 was a significant factor behind the scale of the subsequent recession in Ireland (Whelan, 2013).
A simple estimate of underlying housing demand can be calculated from historical data relating to population size and the average number of persons per household, along with an assumed depreciation rate for the existing housing stock. For the latter, the annual depreciation rate of the housing stock between 1991 and 2002 estimated by Fitzgerald (2005) is taken. It should be noted that such an estimate does not take into account several important factors which are likely to influence future housing demand including the changing demographic profile of the population or changes to household income.
The estimate suggests that housing demand in Ireland rose steadily up until 2007 before a sharp slowdown in population growth, associated with declining net migration at the onset of the crisis, caused housing demand to fall substantially (Figure A.1). The estimate suggests that demand fell from around 70 000 to just over 10 000 dwellings per year between 2007 and 2013, before rising to around 22 000 dwellings in 2016. The recent increase in demand has reflected both stronger population growth and a slight increase in average household size. The latter was uncovered in the 2016 Census and marked a halt in the previous trend decline in the number of people per household. Comparing the estimate of underlying demand with actual housing completions data published by the Central Statistics Office suggests that annual housing supply outstripped demand in the years leading up to the financial crisis, but demand has exceeded supply in the years since 2008. Furthermore, the estimate of housing supply may be overestimated. This is because some housing completions may be double-counted, as these data are based on electricity connections (and there is the potential for reconnections to existing houses).
