18 October 2016, Bedford Hall, Dublin Castle  

‌

The seminar was attended by approximately 80 participants from public bodies, government departments, academic institutions, brokerage firms, real estate companies, housing organisations, economic consultants and other private firms in the construction sector.

There were four speakers on the day which were as follows:

• Gregg Patrick, Statistician, CSO
• Ronan Lyons, Trinity College Dublin
• Brendan Murphy, Statistician, CSO
• David Duffy, Property Industry Ireland

The powerpoint presentations from each are included here:

Introduction - Joe Treacy, CSO (PPT 974KB)

New Residential Property Price Index - Gregg Patrick, CSO (PPT 778KB)

The Availability of Housing Statistics in Ireland: Gaps and Future Needs - Ronan Lyons, TCD (PPT 280KB)

Census 2016: Counting Ireland's Housing Stock - Bendan Murphy, CSO (PPT 2,935KB)

Housing Policy in Ireland - David Duffy, Property Industry Ireland (PPT 189KB)

Attendee List (XLS 11KB) ‌

Note on Proceedings:

Joe Treacy (Director of Business Statistics, CSO) opened the seminar and introduced the theme of the day as an open policy debate on the housing statistics required to inform good policy decision making.  He presented the various statistics currently available from both the CSO and the Department of Housing and outlined that one of the goals of the seminar was to identify areas where further data gaps existed.

Gregg Patrick and Ronan Lyons then made their presentations following which Joe Treacy invited questions from the floor.

Comments and questions from the floor:

• Is it possible to use the Building Energy Ratings (BER) dataset as source of information on plot size for the RPPI? In relation to completions, why is more use not being made of the data contained in the Building Control Management System (BCMS) which is controlled by the Local Government Management Agency (LGMA)? This system should contain detailed information on commencements and completions.
• There was little mention by the speakers of building a statistical system to capture the social aspect of house prices e.g. housing affordability. Can more be done on this?
• The value of a house is also influenced by the capital investment (such as renovations) into residential property. Would it be possible to capture this in the price model?
• Is there any source of information on the value of land which residential property is built on?
• Are the relevant bodies that produce housing statistics all moving in the same direction?
• How often does the deprivation index used in the RPPI get updated? Who else uses the double exponential smoothing approach that was introduced in the recent update of the RPPI?  Is it possible to publish the raw RPPI data without smoothing the data?

Responses from speakers and other contributors:

• It was agreed that adding income to the house price information would provide more insight into the affordability issue.
• CSO carried out some work using a theoretical builders’ model that could be used to decompose house prices into the structure and land components.  However, this model would require additional information on plot size and more time to develop.
• The point was made that making the Eircode mandatory on the Stamp Duty returns would have a big impact on the ability to produce good quality statistics.
• CSO emphasised the importance of the Eircode within the National Data Infrastructure (NDI) and is also a core part of the National Statistics Board (NSB) strategy.
• The point was made that the smoothing techniques implemented in the RPPI were necessary to avoid excessive volatility in the index.  Trust in the data could be undermined if the index was fluctuating every month.  CSO would be willing to provide the unsmoothed data to specialist users if requested.
• The deprivation index gets updated every five years which results in changes to the absolute indices. However, the relative indices remain quite stable over time.

After the coffee break, Brendan Murphy and David Duffy made their presentations following which Joe Treacy invited questions from the floor.

Comments and Questions:

• What is the reason for providing the preliminary Census data at a more aggregated level i.e. electoral divisional level rather than small area level?
• Would it be possible to provide median prices for residential property rather than average prices which are influenced to a greater extent by outliers?
• What is the effect of reduced housing mobility on the economy?
• Can CSO provide more information on approx. 40,000 updates made to the Geodirectory during the processing of Census information?
• Can CSO provide more information on the analytic exercise that is proposed to provide more information on vacant properties in Census 2016?
• Are there any plans for CSO to take over the production of statistics for housing commencements and completions?
• Are there any plans to carry out the analytical work on vacant properties for the 2011 Census as well?
• When will the Census 2016 small area statistics be available on the CSO website?
• Do the speakers have any concerns about the Action Plan in relation to it being a one size fits all for the country? Should there be a different plan for Dublin?

Responses from speakers and other contributors:

• CSO stated that the preliminary Census results are produced at a more aggregate level (electoral division). However, the final results will be published during April 2017 and will be available at small area level.
• CSO stated that its intention is to expand the additional indicators of the RPPI to include median prices.
• CSO provided more detail on the changes that were required to be made to the Geodirectory during Census 2016.  These were largely to do with internal processing rather than incorrect information.
• CSO outlined the typology that it hoped to use to provide more information on vacant properties e.g. for sale, between rental contracts, occupant in nursing home etc.
• CSO stated that there were no plans for CSO to take on the production of statistics on commencements and completions.
• A representative from the Department of Housing outlined the sources of information for the compilation of their current statistics on commencements and completions. The completions data is based on ESB connections and the commencements data is sourced from the Building Control Management System (BCMS).
• CSO made the point that CSO are continually looking to improve how we inform users about the information that is available. 
• It was agreed that the availability of more granular level data from CSO may allow for the development of more appropriate housing policy at a regional level going forward.