Known unknowns: how many houses are we really building?

Cantillon: We need an accurate measure of housing output if we are to address crisis

Ten years on from the crash and we still don’t have an accurate measure of housing output. The one metric that might give us an understanding of the scale of the current crisis and how to remedy it is essentially an unknown.

Traditionally, the Department of Housing uses electricity connections as a proxy for counting new builds, but this method has proved entirely unreliable.

This is because new connections can be triggered by formerly vacant units – think of the myriad of ghost estates left over from the boom – coming back on stream.

Even so, a figure of 19,300 completions for 2017, based on this questionable method, has been adopted.


The State's main economic think tank, the Economic and Social Research Institute (ESRI), sees no reason to demur, using it for its various forecasts, while the Central Bank seems content to tinker with it without really adopting a more reliable approach.

In its latest quarterly bulletin, the bank examines ways of bolstering the electricity connections figure with other datasets, most notably new house registrations and new home loans.

By doing this, it forecasts housing completions will rise 23,500 this year and to 28,500 next year, which don’t really differ from other projections, which take the 19,300 figure at face value.

However Lorcan Sirr, lecturer in housing policy at Dublin Institute of Technology, has painstakingly gone to each of the State's local authorities to tot up the figures for 2017.

While the process is not yet complete, he estimates the 19,300 figure may be overestimating the level of new house building by up to 50 per cent.

The Department of Housing had set up a housing data analytics group to address the problem, but this has so far failed to come up with a solution.

So we blindly proceed to remedying the problem of housing without a clear notion of the dose we’re administering.