The Scottish Government

« Previous | Contents | Next »

Listen

ANALYSIS OF HISTORICAL CONSTRUCTION COST MOVEMENTS IN SCOTTISH SOCIAL HOUSING FINAL REPORT

8.0 REGIONAL VARIATIONS

8.1 Local Authority Housing Management and Maintenance Expenditure

The Scottish Executive collects data on local authority housing management and maintenance (MM) expenditure for each Local Authority and publishes expenditure per dwelling figures in Scottish Executive Statistical Bulletins.

MM expenditure is made up of expenditure on supervision and management (mainly staffing costs) and the repair and maintenance of the housing stock. The average MM expenditure of Scottish local authorities in 2002-03 was 1326 per unit, ranging from 955 per unit in East Ayrshire to 1722 per unit in Glasgow City.

8.2 Repair and maintenance expenditure

In 2002-03, local authority housing repair and maintenance expenditure averaged 877 per unit ⁴, ranging from 555 per unit in Clackmannanshire to 1209 in Glasgow City.

Chart 8.1 illustrates levels of repair and maintenance expenditure for six regions (three rural and three urban), plotted with the average for Scotland, between 1996-97 and 2002-03.

Chart 8.1 Local Authority Housing Repairs and Maintenance Expenditure

The housing stock of Glasgow City (average repair and maintenance expenditure 1209 per unit in 2002-3) has since been transferred to community ownership, as has that of Dumfries and Galloway (821 per unit) and Scottish Borders (595 per unit), both with below average RM expenditure.

The results show that, as expected, the city areas are generally more expensive per dwelling than rural areas to maintain.

The variation in local authority expenditure will be largely the result of property type, stock condition and tenant profile but also the result of varying construction costs around the country.

8.3 Construction cost variations

Both the Building Cost Information Service (BCIS) and the Scottish Executive's Building Division produce a series of relative location factors, intended to provide a general indication of construction pricing differentials within different areas of Scotland. The figures are derived from the analysis of construction prices within a large number of construction projects throughout Scotland over a number of years. The figures represent differences in construction tender prices (Chapter 4) as opposed to notional building costs (Chapter 2). The methodologies employed by BCIS and SE are, for all intents and purposes, identical.

The Scottish Executive's figures are derived from an analysis of 'over 600 Scottish public sector contracts', the majority of which are less than five years old. The BCIS figures are derived from analyses of just under 800 public and private sector projects from 1980 onwards.

Table 8.1 lists the latest figures (February 2003) produced by the two organizations. All figures are shown as plus or minus percentages compared to the Scottish average. Sample numbers are shown with the BCIS data, to help judge the relative reliability of individual figures. Sample numbers for individual authorities relative to the SE data are unavailable but the sample is understood to adequately cover all 32 Scottish Local Authorities.

For most locations there is quite strong agreement amongst the two sets of data but there are some exceptions where the published figures show considerable divergence. The BCIS data generally show a wider spread of figures than the SE data. The total SE sample may be slightly lower than that used by BCIS but it covers a much shorter and more recent time span. As such, where conflicting data occurs, it may be wise to place greater credence on the SE figures, coupled with consideration of sample size where that information is known.

The SE data show that construction costs can vary within mainland Scotland by up to 8% above (City of Edinburgh) or 9% below (Aberdeen City, Aberdeenshire and Moray) the Scottish mainland average cost. There are several possible explanations for these differences. First, when raw materials have to be transported to more remote regions, an additional cost will generally be incurred. Second, in more remote locations, construction work will not be the only source of income for particular individuals. In order to incentivise the local labour supply to engage in construction work, it may therefore be necessary to offer to pay them at a higher hourly rate than would be available to them elsewhere in the local economy. Third, if insufficient labour is available in the area in question, it may be necessary to provide transportation and accommodation to workers coming from elsewhere. In some cases these workers may also require payment for travel time. Finally, if there are few competitors in the market then construction businesses will have little incentive either to make efficiency gains or to cut margins.

Table 8.1 Regional Cost Variations

Chart 8.2 plots this data graphically and serves to highlight where the differences between the two sets of data lie.

Chart 8.2 Comparison of BCIS and SE Construction Cost Location Factors

The factors are effectively long term averages: for instance, the latest SE figures are derived from the analysis of tenders received during the 5 year period 1997-2002. This enables a sufficiently large sample to be collected to achieve a reliable indication of cost variations. However it means that the figures are not quickly responsive to changes in local market conditions e.g. a sudden increase in workload in a particular location, which might result in above trend tender inflation, would not be fully reflected in the location factor for 2 to 3 years.

The BCIS data covers a much longer period, back to 1980, though statistical techniques have been employed to give greatest credence to the latest figures.

8.4 Islands

Expenditure on repairs and maintenance in the Shetland Islands in 2002-03 was 1086 per house, the second highest of all the local authorities. One reason for this relatively high expenditure is the fact that construction costs on the islands are generally significantly higher than on the mainland.

Chart 8.3 shows the latest construction cost factors published by BCIS and the Scottish Executive's Building Division for the islands.

Chart 8.3 Construction cost factors for the Scottish Islands

The chart illustrates fairly substantial differences between the Scottish Executive and BCIS factors in respect of construction costs in the islands. However the BCIS data is based on small samples covering a long time span. No figure is available from BCIS in respect of the Western Isles because of insufficient sample. As such it would seem prudent to place greater reliance on the Scottish Executive data, which suggest that construction costs in the islands may be 8% (Western isles) to 26% (Orkney) higher than average mainland costs.

However, one result here seems to be counterintuitive - that of Orkney. The Orkney location factor, according to SE data, is 26% above mainland costs, despite the fact that it is considerably nearer the Scottish mainland (for transportation purposes) than Shetland (14% above). Further research may therefore be warranted into relative cost differences for the islands.

8.5 Changes over time

The figures shown in the tables and charts above are those most recently published or produced by the Scottish Executive and BCIS and represent the current position regarding construction cost variations within Scotland, subject to the caveat in paragraph 8.3. The factors are not permanently fixed and can vary over time when construction workload, industry capacity or other economic factors change.

Chart 8.4 shows how construction cost factors for some regions have moved over the last 7 years, based on Scottish Executive data.

Chart 8.4 Change in Construction Cost Factors

The figures suggest that construction costs in Argyll and Bute, Angus, Perth and Kinross and Dundee City have declined relative to costs in other areas over the period considered, while costs in the Scottish Borders and Dumfries and Galloway have increased. Cost factors for most other regions over this period have remained fairly stable.

A change in the mix of the national average figures could result in apparent but unreal changes in the factors for individual regions. A check by comparing the trends of the regions in Chart 8.4 against factors for Glasgow City, instead of the national average, confirms the changes in direction observed in the chart.

Examination of the historical cost factors for the Scottish Islands also suggests that differential movement has occurred. Chart 8.5 illustrates how construction cost factors derived for the islands have changed over the last seven years.

Chart 8.5 Changes in Construction Cost Factors for Scottish Islands

G48

The factors produced by the Scottish Executive for the Island Authorities are based on a ten year sample of construction cost data from suitable projects, which rolls forward on an annual basis. Therefore the factors for 2003, for instance, mainly relate to projects that have been tendered in the period 1992-2002.

Interestingly, the construction cost factors for all three Island Authorities shown have been declining relative to the mainland average cost for at least the last four years (and for the whole six year period displayed in the chart in the case of the Shetland Islands). Given the time span of the data composing these figures, it is possible that the narrowing trend indicated in the chart may have continued further and that the figures above may overstate the current premium in construction costs in the islands.

There are a number of possible explanations for the trend observed above. One explanation could be the introduction of improved production techniques, such as the increased use of prefabrication. Other factors involved could be improved transport infrastructure and improved information and communication technology, necessitating fewer site visits and thus increased efficiency. Finally, in places like Shetland labour could have become more plentiful as other local industries, such as oil and fishing, have declined.

8.6 Management cost variations

In 2002-03, local authority housing supervision and management expenditure averaged 449 per unit, ranging from 293 per unit in Eilean Siar, Comhairie nan to 646 in West Lothian. These figures hide myriad reasons for the differences but staffing costs (responsible for the majority of SM costs) do vary throughout Scotland. Local authority salaries may be subject to pay scales but private sector wages and salaries are more dependent on local market levels. Chapter 6 also indicated how property costs may vary in different locations.

Table 8.2 shows how average earnings throughout Scotland varied in 2002 as plusses and minuses compared to the Scottish average.

Table 8.2 Average Earnings in Scotland 2002

Source: ONS New Earnings Survey (2002)

These figures may give a broad indication of how management costs may vary throughout Scotland though individual figures, such as Aberdeen, may be more strongly influenced by the mix of jobs making up the sample survey e.g. considerable numbers still work in the oil sector which is traditionally a high pay sector. As such this data should be treated with caution and should only be used as a very broad guide.

8.7 Conclusions

The location of a project is clearly important when judging its probable cost. It is recommended that further more detailed investigation be carried out into regional differences when a particular region is under consideration. DLE's experience indicates that in normal market conditions construction location factors can be a good guide to regional variations. However with the introduction of variables such as specialist work or a buoyant market then they may prove inaccurate. In rural or remote areas of Scotland the advent of one large intensive project using up vast amounts of local resources would create the need to import resources and therefore drive up prices.

The available data suggests that, in remote rural or over-pressurised urban areas, cost uplifts of between 1% and 8% should be applied to average Scottish mainland costs. Similarly, in less remote rural areas or less pressurised urban areas, a reduction of up to 9% can be applied. In island authorities, cost uplifts of between 8% and 26% should be applied, though evidence appears to suggest that these cost differences are narrowing. More research into this area would be beneficial in building up evidence as to why this may be the case.

The factors reflect average price levels and variations can occur within each area, especially in island areas or where an area contains a large centre of population and more remote rural communities. The factors are intended to provide a general indication of pricing differential. Attention should also be given as to how these figures might vary from year to year.

« Previous | Contents | Next »