The National Institute for Regional and Spatial Analysis (NIRSA), a university institute at the National University of Ireland at Maynooth set up in 2001 “to undertake fundamental, applied and comparative research on spatial processes and their effects on social and economic development in Ireland”, has produced a working paper which draws together much work undertaken by NIRSA on Irish ghost estates, overproduction of housing, planning and the Irish response to the banking and property crisis.
The working paper has received widespread media coverage, in particular its call for a specific inquiry into the property bubble and related planning issues. The paper examines the property bubble in particular from 2000 onwards, planning, overproduction, the property boom and bust. NIRSA received much coverage at the start of 2010 when its work on vacant residential property and ghost estates was published. It is noteworthy that NIRSA have revisited their estimates of vacant property and are now estimating just over 200,000 properties vacant (defined as base vacancy plus overhang), which places NIRSA somewhat below other studies such as UCD’s and DKM’s. Apparently the DoEHLG is in the process of counting vacant property and I would not be surprised if the results show that there is much less property that could potentially be made available for sale than the recent studies would imply – – according to research by Federcasa (the Italian Housing Association – used as a source in the recent UCD vacant homes study) in 2005/6, Ireland has had a vacancy rate of 10.2% even back in 1991 (the source is credited to “National Statistical Institutes, Ireland and the DoEHLG”), it may turn out to be the case that the potential supply of vacant property that can be made available for sale might be in the area of 100,000 units – hopefully the DoEHLG will provide comprehensive statistics later in the year.
Of particular interest here is NIRSA’s consideration of NAMA, and NIRSA is cautious about NAMA’s operation and prospects. Although the report doesn’t introduce new data in relation to the operation of NAMA, it does draw together a number of concerns about how NAMA is operating:
Lack of transparency – this is a subject touched on by many previously including the Ombudsman and Information Commissioner, Emily O’Reilly – all calling for NAMA to be included within the Freedom of Information. The thrust of the NIRSA call for transparency however originates from a different angle and NIRSA want NAMA to share information on their assets to assist in the orderly planning of housing and infrastructure. This call echoes recent calls from local politicians and county managers for example in Cork and Westmeath. NIRSA also raise concerns over the way in which NAMA is valuing loan assets, stating “It is also not clear as to how valuations are being made and whether they take into account existing levels of oversupply and evidence-informed, long-term projections of an area’s demography and labour market.”
Risk of losing money – NIRSA highlight the fact that NAMA is controversially paying in excess of what property is worth today (more accurately what it was worth in November 2009) by paying the Long Term Economic Value. NIRSA point out “Land in areas of high surplus housing and/or over-zoning is likely to fall greatly in value and to stay that way for quite some time, limiting the ability of NAMA to realize any profit, especially if it is acquired for too high a value”. It is of course to be hoped that NAMA have conservatively valued property. NIRSA’s own study on vacant property in the State however was published after 10th January, 2010 and according to the NAMA LEV Regulation, any analysis produced after this date cannot be used in valuing the LEV. NAMA may be able to use a DKM study from September 2009 but that study did not indicate the location of oversupply. NIRSA also express a concern that NAMA may pay for property subject to zoning which may then be de-zoned.
The negative impact upon the existing residents of ghost estates by NAMA’s decisions regarding empty or partly built houses – firesales (might worsen negative equity), leasing for social housing (might concentrate social difficulties – “creating new Ballymuns” as one commentator remarked), hoarding (will mean that existing residents continue to live without neighbours). There is an appeal for NAMA to meet the civilized expectations of people trapped in ghost estates.
There is also an examination of the context which gave rise to NAMA and whether it is an “illogical logical” construct and NIRSA question whether in principle an operation like NAMA, which may seek to re-inflate the property bubble and protect what many would call elites, is really what Ireland should be doing at this juncture.
The main part of the paper however doesn’t deal with NAMA, it examines the context for our property crash and does so with clarity with rich supporting statistics. A worthwhile document that deserves widespread media exposure.
NAMA Wine Lake 
“It is noteworthy that NIRSA have revisited their estimates of vacant property and are now estimating just over 200,000 properties vacant (defined as base vacancy plus overhang), which places NIRSA somewhat below other studies such as UCD’s and DKM’s.”
I’m not sure how you have come to this conclusion. The DKM, UCD, NIRSA and Goodbody studies all have strong alignment with regards to vacancy and oversupply estimates (as per Table 1). As stated in the paper: “All four organisations estimate that vacancy including holiday homes is over 300K, that vacancy excluding holiday homes is over 228K, and that the potential oversupply is over 103K ” The 103K is a result of Goodbody using a 7.3% base vacancy rate, NIRSA and DKM use 6% and UCD 5%. Interestingly, these are figures that the DEHLG endorse in their Housing 2009 report.
The vacancy rate during the boom, as recorded in the Census, consistently rose throughout. In 1996 the rate was 8.5%, in 2002 it was 9.8%, and in 2006 it was 15%.
Thank you for your comment. Firstly congratulations on the publication of your working paper. I hope that certain themes are carried forward and that the paper continues to attract deserved media attention. Personally I was aghast at Minister Cuffe’s suggestion on Prime Time that the 80% development tax had solved issues around land speculation when it seems we are regularly creating further multimillionaires by paying huge sums for land for infrastructure such as roads, – see https://namawinelake.wordpress.com/2010/07/09/the-government-continues-to-create-property-millionaires-by-failing-to-implement-the-kenny-report/
With respect to my comments on NIRSA’s work with vacant housing, I was drawing attention to statements in your working paper on page 19 in relation to Table 1 which compares the breakdown of the four studies into vacant property. You state
“d) NIRSA has employed two different methods in its analysis. The first method (not in brackets) used house completion, address database and new mortgage data to estimate vacancy, but not potential oversupply. The estimate of 302K includes undercounted holiday homes but did not estimate how many there were. A subsequent analysis to calculate potential oversupply used a projection of population household growth based on 96-06 household change, factoring in
obsolescence and holiday home rates, so as to be able to estimate county rates (new mortgage data is not available at county level). The two different methods lead to slightly different estimates of vacancy excluding holiday homes due to the estimation of their number (see note
e), but overall vacancy including holiday homes is about the same.
e) Using the second approach, NIRSA estimated the total number of holiday homes at the end of 2009 as c.86K (using the DKM method of 5% of total stock in 2006 based on the 2005 Household Budget Survey [73K], and 5% of potentially available stock between Apr 06-09 [13K]).
f) This is not the base vacancy, but the excess vacant stock in the 2006 Census (216,533) above an expected 6% base rate of total stock (106,177) minus 23,000 undercounted holiday homes (5% total 2006 stock would be 73,000 not 49,789).”
I had understood that NIRSA’s well publicized work on vacant housing in January, 2010 had indicated 302,625 vacant homes PLUS holiday homes of 49,789 to give a total of 352,414. You stated back in January 2010 “This figure includes vacant houses available for sale, vacant houses available for rent, vacant houses that are not on the market, under-counted second and holiday homes, and abandoned properties, but does not include 49,798 holiday homes recorded in 2006 census. This is not a measure of availability but vacancy (some of which is available to the market now, and some of which will become available when prices rise/demand returns, some of which will not become available).”
Looking back at your IrelandafterNAMA blog I see that you introduced in March 2010 what I have referred to as the revisit of your January 2010 work. And at that time as opposed to your current working paper, you revised your estimates of holiday homes and introduced a figure for obsolete properties. Your base vacancy and overhang estimate then became 207,604 which is lower than the corresponding figures from UCD (c270k), DKM (c228-253k) and Goodbody (c229-270k). So that is what I was referring to as the revisit of your estimates and that your estimate for vacant property (*defined as base vacancy and overhang*) is greater than the other three studies.
With respect to Ireland’s historical vacancy rate. I was referring to Federcasa’s study (hyperlink in the post above) which seems to be well thought of – it was used as a source for aspects of the UCD study. They state on page 63 that in 1991 our vacancy rate was 10.2%. You refer to the CSO’s census and state “In 1996 the rate was 8.5%, in 2002 it was 9.8%, and in 2006 it was 15%”. Could you help me with the source for these numbers? I would have expected these numbers to be obtainable from the CSO’s Housing Census reports. Let’s take 2002 as an example where you have 9.8%. Referring to the CSO report available here, could you point to how you derive 9.8%?
The CSO document I referred to was Vol 13 – Housing http://www.cso.ie/census/Vol13.htm
Okay, I think I see where you are coming from with this. The difference between our figures as you’ve expressed them and DKM, UCD, Goodbody is the obsolescence rate – dwellings that might be expected to fall out of the housing stock due to change of use or state of repair. DKM and Goodbody do not estimate this rate and UCD have a very low estimate. If they were to estimate then I would expect the base vacancy+overhang (but not inc. obsolescence) to align as they would be undertaking the same calculation. The only difference would perhaps relate to the obsolescence rate used. We used 6/1000, which is generous given the age of the housing stock.
As for the census data on oversupply we have either calculated that ourselves using the CSO data – households to housing units – or its in one of their published documents (I’d need to talk to one of the team).
I would indeed be interested in how you arrive at vacancy rates in 2002 and 1996 – a few months ago I did try to pursue this with the CSO but didn’t have the time to follow it up – DKM make reference to a vacancy rate in their report last year but say they obtained it from the CSO. In response to your reply, the number of households in the 2002 census was 1287k – do you have a source for the number of houses for 2002? If I take home completions for 2002-2009 and subtract that from 2m and divide that into households in 2002 I get 17% vacancy for example. That would be an approximation for lots of reasons but I would be interested how you would support your statistics of 9.8% for 2002 and 8.5% in 1996.
To lay my cards on the table, I also tried to research further vacancy rates in other countries and because the US is a relatively open and transparent society, I found a lot of up to date information at the census. The most recent housing figures suggest the vacancy rate is 14.4% in the US at Q2, 2010 – see here:
http://www.census.gov/hhes/www/housing/hvs/qtr210/files/tab4.xls
And there does not seem to be any great concern about vacant housing in the US. In fact in the last week when the US housing start data was released for June 2010, the concern was that it was on the low side with a figure of 549,000. See here
http://www.businessweek.com/news/2010-07-20/housing-starts-in-u-s-slide-more-than-forecast-as-credit-ends.html
And although it is completely unscientific, I am familiar with a town in the Midlands where some 15 estates were erected during the boom. And based purely on drive-by through the town and estates and observation (lights on, cars outside, curtains, stickers still on windows, for sale signs) and knowledge of emigrants who are seeking to rent and extrapolating the [(empties in the town divided by total housing in the town) multiplied by 2m (estimate of houses in the State)] would indicate vacant housing potentially available for sale or rent of just over 100,000 – completely unscientific of course in the sense that I have no idea how representative the town is of the State. But supportive nonetheless.
So combining the information above means I have a hunch that the actual amount of completed housing that is vacant and can potentially be made available for sale or rent may be some ways off the 200-270k that your study and others have indicated. And if I were a betting man I would say the results of the Great Count by the Planning Minister later this year will be lower that the four studies you refer to. On the other hand I think our population for the next few years is not going to demanding anything near 30,000 new homes each year – the figures for 2009 would indicate a demand for 2,000 homes versus over 10,000 being built and a large amount of overhang.
In 2008, number of households = 1.59m (CSO), dwellings = 1.934m (DEHLG) which is a large difference (340K). Given that 215K houses have been built since last census in 2006, and household growth has slowed considerably, it seems very unlikey that the vacancy rate has fallen since 2006. The key thing is oversupply as opposed to vacancy. DKM/DEHLG, NIRSA and UCD all estimate this is 120K+ (based either on pop projections or number of new mortgages). The other vacant properties are expected to be vacant (6%) or holiday homes or have become obsolete. 120K is a big overshoot (about 3.5 years if pop had continued to go up at the same rate as 96-06), but is geographically uneven affecting some places much more than others.
I would expect the DEHLG house count to come in well under the 120K rate (probably nearer the CIF rate). It only has reference to post-April 2007 housing estates where there is vacancy above 10%. It therefore a) excludes estates where vacancy is below 10%, b) excludes unoccupied/unfinished one-off housing, c) it assumes that estates started prior to April 2007 are both finished and have occupancy levels above 90%. There are lots of houses in pre-2007 estates that are empty; the same with one-offs. Irishpropertywatch.com reports about 115K houses for sale in state and 20K for rent. In other words, it’s a partial survey and we’ll need to wait for census 2011 for a fuller picture.
In terms of houses that have never been sold this is somewhere between the oversupply figure and the 40K of CIF, depending on how many house have been bought for cash, though remain empty. There are lots of investment property, bought to flip, that remains empty. One estimate, as noted in the report, is that there are c.40K empty apartments in Dublin alone.
My feeling is that, however you want to look at it, we have a problem – we’re spending a fortune through NAMA, bank recapitalisation, etc to sort that problem out and its difficult to see how supply and demand will be harmonised in the short term in many places.
That’s interesting information on the DoEHLG count and leaves me wondering what value it has – it won’t tell you the number of vacant properties in the State, it won’t tell you where the ghost estates are (assuming there are pre-2007 estates with low occupancy). From the point of view of planning it would seem to provide only a partial picture. From the point of view of demand:supply housing (and consequences on prices) it will definitely provide an incomplete (and possibly misleading) picture. What value does it have?
I’m afraid I am a little confused about what you say about irishpropertywatch.com – they appear to show on their home page that there are 54,200 properties for sale (and that’s for 32 counties isn’t it?) and 21,164 for rent (again isn’t that for 32 counties). Do these two categories include multiple listings for the same property? Do these statistics include property that is both for sale and for rent? There appear to be 62,740 non-site properties for sale on DAFT this morning in the 26 counties, these most definitely include multiple listing properties where there is more than one estate agent. Myhome.ie has 40,950 properties this morning, again including multiple listings and indeed some properties will be on both daft and myhome. Some (if not many) of these properties will be occupied and any sale will be part of a chain.
I would be grateful if you could let me know how you arrived at vacany levels in 2002 and 1996. These look low compared with Federcasa’s report for 1991 and compared with 2010 vacancy rates in the US.
But overall I don’t think there is any disagreement that we have a problem with an oversupply of property. I think this will be exacerbated by the short to medium term stagnation or even decline in population. All of which will give rise to planning challenges because as you say, what we do know is that there has been an oversupply of housing in some areas, esp the Upper Shannon, where the indigenous population is not going to occupy the houses already built. It seems like there has never been a more needy time to have an institute like NIRSA to help inform the debate.
[…] Minister Ciaran Cuffe’s long-awaited review of some ghost estates is to be published later today. Remembering that the review was set up to examine estates built after 2007 where there is more than 10% vacancy […]