Thursday 27 October 2016

How big data can help the housing crisis

Ronan Lyons

Published 25/09/2016 | 02:30

Amazingly, Ireland still relies on connections to the electricity grid to measure new dwellings being built. (Stock picture)
Amazingly, Ireland still relies on connections to the electricity grid to measure new dwellings being built. (Stock picture)

Imagine a company that didn't know what assets it owned. No doubt there are plenty of sole traders who might not formally have a balance sheet, and similarly, there are plenty of companies - including big ones - that find it difficult to put a value on intangible assets like their brand. But it is hard to imagine any company that wouldn't know what real estate it owned. Particularly if that company was a multi-billion euro company.

  • Go To

Yet that, amazingly, is the situation Ireland finds itself in when it comes to property. True, there are national accounts that estimate the total value of various types of assets, including housing. For example, we know there are roughly two million properties in the country and we know that the average value of a property is a little above €200,000. So the total value of residential property in Ireland is those two numbers multiplied, so roughly €400bn.

But this number should be calculated from the bottom up, not from the top down. Most developed countries have what is known as a cadastre, a formal register of who owns what - and in some countries, who lives in what. This information is used for all sorts of purposes. Examples include which neighbourhoods need new schools or more hospital beds or greater provision of housing for older people.

In Ireland, for various historical reasons, the system of owning property was traditionally done by what is termed priority - ie, you prove in court by having the best bundle of documents that you own a plot or building - rather than title, where the State itself takes a position on who owns what. Recently, Ireland has recognised just how out of date such a system is and is switching to a title-based system.

However, this switch is happening at a snail's pace as it takes a transfer of property for title to be required. More importantly, it will not solve other limitations when it comes to Ireland managing its housing stock, including what is built when (and where) and what is vacant or under-used.

Amazingly, Ireland still relies on connections to the electricity grid to measure new dwellings being built. This has led to the farcical situation of one of Ireland's official statistics (the Department of Housing data on completions) suggesting that there were more than 50,000 properties built between 2011 and 2016, whereas another (the Census) indicates that less than 20,000 were built.

Some of the difference may be down to properties falling into disrepair, but it is likely the vast bulk is due to a gap between when properties were built and when they were hooked up to the network. In many cases, this gap may have been substantial, with properties lying idle in ghost estates throughout the 2007-2012 crash.

This state of ignorance about what exactly has been built over the last 10 years means we are also effectively ignorant about what is being built now. How many "completions" are actually homes built in the bubble finally coming on stream? And if we don't know how much (or little) building is taking place, how can we diagnose what is wrong? This problem is multiplied many times over if we want to know not just what buildings are where, but which are vacant.

The picture is somewhat grim. However, earlier this week the CSO released a new residential property price index. In terms of price trends, it revealed little in the way of surprises. It did show that property prices fell by more between 2006 and 2012 than the previous CSO index suggested - although the new peak-to-trough fall is effectively the same as the index of list prices, which uses the same method.

The importance of the index is not really in its results but rather its methods. To calculate this index, the CSO brought together not only the Property Price Register but also Eircodes and the register of Building Energy Ratings. As a result, the CSO can now publish average prices by area, as well as trends by region. And with official averages, policy-makers and financial institutions can make projections about housing supply, social housing, negative equity and many other important measures.

This is exactly the kind of contribution so-called 'big data' can make. The team at the CSO is to be praised for making Ireland's housing market that little bit more resilient. But the to-do list in the sector remains huge. "A lot done, more to do" as they say!

  • Ronan Lyons is assistant professor of economics at Trinity College Dublin and author of the Reports

Sunday Independent

Read More

Promoted articles

Don't Miss

Editor's Choice