Vacant Dwellinngs in England
The challenges and costs of bringing them back into use
By Maggie Davidson and Kevin White
Published by BRE Trust, 2010, ISBN 978-1-84806-131-6, A4, 20pp. £26.44 (instant download) or £22.50 (hard copy). It is very much to the credit of the BRE Trust that they have commissioned this report. It's the first serious analysis of vacant dwellings since the Department of the Environment report Vacant Dwellings in the Private Sector, published 15 years ago. The report draws heavily on data from the English House Condition Survey (2006) for its amalyses, but also refers to other research such as MORI surveys in North London and Birmingham into why properties were empty. Whilst the focus is on long-term, private sector vacants, public sector dwellings are also discussed and there is a cursory discussion of associated issues such as homelessness The key findings - the ones unique to this study - are those relating to the costs of repairs needed in the empty dwellings. The study found that 20% of private sector long-term dwellings needed little or no repair, whereas 20% needed £11,000 or more. Perhaps the most telling statistic is that £5.8billion would have to be spent to address the urgent repairs needed on all these dwellings - but only £91million would be needed to address the urgent repairs on the less delapidated half. Of course this analysis depends on the questionable if orthodox distinction between transactional and long-term empty dwellings revolving around the "6-month empty" criterion. This terminology has created a fantasy world where it is too easy to assume that so-called long-term empties are empty indefinitely, whereas in reality the majority are probably in an unduly slow process of being brought back into use. (The more accurate phrase "longer-term empty" might usefully replace "long-term empty"). It is a shame that the BRE research did not include a more granular analysis correlating repair costs with different durations of vacancy. The relative brevity of the report does leave gaps in the material. The most significant are around methodological issues. People are becoming increasingly aware of the adage that there are "lies, damned lies and statistics". Close scrutiny of the methodology involved in data collection is essential to counteract this. The discussion of the methodology here is too cursory and astonishingly there is not even a reference to the detailed methodology used in the English House Condition Surveys (which can be found at http://www.communities.gov.uk/documents/housing/pdf/ehcstrp05.pdf). The key findings here are based on 781 empty dwellings drawn from 3 years worth of those surveys, to "maximise the number of vacant dwellings in the sample" (p.3). The report does not state what sampling method was used to identify these 781 dwellings nor the overall size of the sample from which they are drawn; nor does it discuss any possible bias in the construction of the sample or any measures that might have been taken to alleviate that. For the dwellings that make it into the sample, the methodology as described involves interviewing owners and physical surveys including internal inspections. The bottom line is that if dwellings are only included where full internal access and interview are possible, then we can assume that any sample is pretty skewed. If, on the other hand, dwellings are included where full access / interview are not carried out, this would allow a more accurate sample but would potentially undermine the quality of the data gathered. Either way, a full discussion of the issues needed to be included in the report. In the absence of such a discussion, it is possible to work backwards from the information provided in the document cited above to answer some of these questions. This states that "overall, 9,176 occupied and 926 vacant addresses were issued to surveyors for the 2004-05 survey, and 8,864 occupied and 804 vacant addresses for the 2005-06 physical survey." It then goes on to say that "A total of 8,440 full surveys including 308 at vacant properties were achieved in 2004-05, and 8,230 full surveys, including 303 at vacant properties, in 2005-06." So we can assume that the sample of 781 vacant properties is based on properties where full surveys were possible - good quality data but derived from a suspect sample because full surveys were only possible in around a third of the vacant dwellings where addresses were issued to surveyors. Moreover, the guidance seems to suggest that instructions to surveyors were in turn only issued where it had been possible to interview owners - which must already have introduced considerable bias into the sample compared with the fotal population of empty homes. Whilst the guidance referred to above shows that strenous efforts were made to adjust the overall sample for different variables being assessed in the survey, those variables do not for the most part seem to include factors which might be relevant for ability to access empty homes, such as duration of vacancy. In short, there have to be serious questions about how representative the sample of empty homes is. What would have been useful in such circumstances would have been some insight into the circumstances which facilitated or hindered vacant dwellings being included within the survey, coupled with an attempt to assess a building from its external characteristics where internal inspections were not possible. True the quality of data for buildings assessed from external inspection would have been lower, but overall the sample might then be more representative and loadings could be applied to try and adjust for non-survey. There is a good case for the English Housing Survey (successor to the EHCS) developing an independent methodology for assessing vacant homes. One or two errors have crept in. The report discusses data sources and refers to the discrepancy between the Empty Homes Agency figures (based on HSSA data) and the figures derived from the EHCS, and attributes this in part to the fact that "if they [local authorities] are relying on council tax records, this will only include dwellings that have been vacant for at least six months". This is not the case as dwellings empty for under 6 months are easily identifed in council tax data from their exemption code and would be included in both HSSA and EHA figures (though local authorities no longer have to calculate the figure for homes empty less than 6 months, as they were once required to do). The A4 glossy format allows the text to be well-illustrated with colour graphs and tables and whilst £22.50 sounds a lot of money for 20 pages it is cheap compared with the work involved in retrieving the useful information that is here collected together from a wide range of sources - to say nothing of the information about repair costs which is only available here. The report should feature in every empty homes practitioner's library. The publication is available as either an instant download (see http://www.brebookshop.com/details.jsp?id=325436) or in hard copy format (see http://www.brebookshop.com/details.jsp?id=325403). David Gibbens
Published by BRE Trust, 2010, ISBN 978-1-84806-131-6, A4, 20pp. £26.44 (instant download) or £22.50 (hard copy). It is very much to the credit of the BRE Trust that they have commissioned this report. It's the first serious analysis of vacant dwellings since the Department of the Environment report Vacant Dwellings in the Private Sector, published 15 years ago. The report draws heavily on data from the English House Condition Survey (2006) for its amalyses, but also refers to other research such as MORI surveys in North London and Birmingham into why properties were empty. Whilst the focus is on long-term, private sector vacants, public sector dwellings are also discussed and there is a cursory discussion of associated issues such as homelessness The key findings - the ones unique to this study - are those relating to the costs of repairs needed in the empty dwellings. The study found that 20% of private sector long-term dwellings needed little or no repair, whereas 20% needed £11,000 or more. Perhaps the most telling statistic is that £5.8billion would have to be spent to address the urgent repairs needed on all these dwellings - but only £91million would be needed to address the urgent repairs on the less delapidated half. Of course this analysis depends on the questionable if orthodox distinction between transactional and long-term empty dwellings revolving around the "6-month empty" criterion. This terminology has created a fantasy world where it is too easy to assume that so-called long-term empties are empty indefinitely, whereas in reality the majority are probably in an unduly slow process of being brought back into use. (The more accurate phrase "longer-term empty" might usefully replace "long-term empty"). It is a shame that the BRE research did not include a more granular analysis correlating repair costs with different durations of vacancy. The relative brevity of the report does leave gaps in the material. The most significant are around methodological issues. People are becoming increasingly aware of the adage that there are "lies, damned lies and statistics". Close scrutiny of the methodology involved in data collection is essential to counteract this. The discussion of the methodology here is too cursory and astonishingly there is not even a reference to the detailed methodology used in the English House Condition Surveys (which can be found at http://www.communities.gov.uk/documents/housing/pdf/ehcstrp05.pdf). The key findings here are based on 781 empty dwellings drawn from 3 years worth of those surveys, to "maximise the number of vacant dwellings in the sample" (p.3). The report does not state what sampling method was used to identify these 781 dwellings nor the overall size of the sample from which they are drawn; nor does it discuss any possible bias in the construction of the sample or any measures that might have been taken to alleviate that. For the dwellings that make it into the sample, the methodology as described involves interviewing owners and physical surveys including internal inspections. The bottom line is that if dwellings are only included where full internal access and interview are possible, then we can assume that any sample is pretty skewed. If, on the other hand, dwellings are included where full access / interview are not carried out, this would allow a more accurate sample but would potentially undermine the quality of the data gathered. Either way, a full discussion of the issues needed to be included in the report. In the absence of such a discussion, it is possible to work backwards from the information provided in the document cited above to answer some of these questions. This states that "overall, 9,176 occupied and 926 vacant addresses were issued to surveyors for the 2004-05 survey, and 8,864 occupied and 804 vacant addresses for the 2005-06 physical survey." It then goes on to say that "A total of 8,440 full surveys including 308 at vacant properties were achieved in 2004-05, and 8,230 full surveys, including 303 at vacant properties, in 2005-06." So we can assume that the sample of 781 vacant properties is based on properties where full surveys were possible - good quality data but derived from a suspect sample because full surveys were only possible in around a third of the vacant dwellings where addresses were issued to surveyors. Moreover, the guidance seems to suggest that instructions to surveyors were in turn only issued where it had been possible to interview owners - which must already have introduced considerable bias into the sample compared with the fotal population of empty homes. Whilst the guidance referred to above shows that strenous efforts were made to adjust the overall sample for different variables being assessed in the survey, those variables do not for the most part seem to include factors which might be relevant for ability to access empty homes, such as duration of vacancy. In short, there have to be serious questions about how representative the sample of empty homes is. What would have been useful in such circumstances would have been some insight into the circumstances which facilitated or hindered vacant dwellings being included within the survey, coupled with an attempt to assess a building from its external characteristics where internal inspections were not possible. True the quality of data for buildings assessed from external inspection would have been lower, but overall the sample might then be more representative and loadings could be applied to try and adjust for non-survey. There is a good case for the English Housing Survey (successor to the EHCS) developing an independent methodology for assessing vacant homes. One or two errors have crept in. The report discusses data sources and refers to the discrepancy between the Empty Homes Agency figures (based on HSSA data) and the figures derived from the EHCS, and attributes this in part to the fact that "if they [local authorities] are relying on council tax records, this will only include dwellings that have been vacant for at least six months". This is not the case as dwellings empty for under 6 months are easily identifed in council tax data from their exemption code and would be included in both HSSA and EHA figures (though local authorities no longer have to calculate the figure for homes empty less than 6 months, as they were once required to do). The A4 glossy format allows the text to be well-illustrated with colour graphs and tables and whilst £22.50 sounds a lot of money for 20 pages it is cheap compared with the work involved in retrieving the useful information that is here collected together from a wide range of sources - to say nothing of the information about repair costs which is only available here. The report should feature in every empty homes practitioner's library. The publication is available as either an instant download (see http://www.brebookshop.com/details.jsp?id=325436) or in hard copy format (see http://www.brebookshop.com/details.jsp?id=325403). David Gibbens