Design and the Economics of Building

D. Jaggar, R R Morton

• Publisher: Taylor & Francis
• Year: 2003
• eBook ISBN: 9781135823771
• Edition: 1
• Topic: Architecture
• Subtopic: Architecture General

A textbook on design economics for students of architecture, building and quantity surveying, it examines the links between design and the costs of building as well as more general economic issues and their significance for designers and builders.

1: Introduction to the tender price and building cost indices

Index numbers are used in economics mainly to describe changes over time in prices, costs or output, but they can also be used to indicate geographical variation; for example, variations in income level between one part of a country and others—or between different countries. The principle is very simple: one particular point in time or space is taken as the standard (the base) from which others are measured; the relevant measure at the base date or place is given the value 100. So, for example, if an index were being developed for materials prices using January 1990 as the base date, the price of timber in 1990 would be assigned the index number 100; if prices rose 5% by January 1981 the index would then be 105; if it fell by 5% the index would be 95.

An index could be used as in that example just to show a change in price of one particular commodity, but the most usual function of index numbers is to amalgamate a range of different prices into one single indicator; and it is important to remember index numbers are just that—indicators of difference; they are never mathematically accurate measures.

There are two problems which arise in trying to create a good index series for a range of different prices.

Problem 1: prices changing at different rates

Individual prices might be changing at different rates; but some prices might be more important, for one reason or another, than others. This problem is tackled by ‘weighting’ the different constituent prices in an index by some measure of their relative importance.

For example: assume we want a simple index of building costs, that we already have information on the costs of labour, materials and plant, and that investigation has shown that on most projects costs are split 40% labour, 50% materials and 10% plant. Assume the cost of plant rose very much more rapidly (this is unlikely—an exaggerated example is used to show the effect of weighting); the situation might be as follows:

	Year 1	Year 2
	(the base year)
Index of plant cost	100	130
Index of labour cost	100	110
Index of materials cost	100	112

To derive a simple index of all costs for year 2 we could simply average the three numbers in the last column, giving (130+110+112)/3=117.3

But this gives too much importance to the rise in plant costs which only account for 10% of the total. So if we weigh the figures by their relative importance we get the following:

Plant	130×10	=	1300
Labour	110×40	=	4400
Materials	112×50	=	5600
Total (indices×weights)			11 300

To derive the index we now divide by the total weights (100)—for we now have 100 units of measurement, not three; this gives us an index of 113, considerably lower than the first one (117.3) and, because it reduces the effect of the increase in plant costs is a better reflection of the general increase in total costs.

Problem 2: change in significance of items indexed

The second basic problem with index numbers is that the weights used in a base year may not be appropriate in later years; over long periods of time they may become quite inappropriate. The best-known index of all—the retail price index—is particularly sensitive to this sort of change. At one time household coal was an important item in the average household budget; expenditure on cars was not. This has now of course completely changed.

There are no real solutions to this problem—just ways of making it less serious. Usually new base years are selected, with updated weightings; or current year weightings are used—that is the weights are changed as circumstances change. The difficulty with both these solutions is that they strictly invalidate comparisons over long periods; one is no longer comparing like with like.

(These issues are fully dealt with in all basic text statistics and will not be pursued further here.)

The tender and building cost indices

There are a number of published indices developed for both building and civil engineering but the most well-known and most widely used are those published by the BCIS (the Building Cost Information Service).

The BCIS all-in tender price index is based on a random sample of priced bills of quantities for accepted tenders; a project index is calculated for each one in relation to a set of base date prices; then all the projects in the sample are averaged to produce the published tender price index, which, it is claimed, ‘provides an accurate and effective measure of the variations in contractors’ prices over time’. There is a range of other tender price indices focusing on one type of tender, such as the ‘private commercial’ and the public-sector TPIs.

There are several building cost indices published, some for particular types of construction such as the Steel Framed Buildings Cost index but also one general building cost index, which is a weighted average of three of the special ones. The difference between the tender and the cost indices is that the latter are derived from changes in national labour rates, materials prices and plant costs, and are produced by applying the changes in these costs to cost models of average buildings. Some 30 work categories such as excavation, masonry, hardwood joinery, are given percentage weights to reflect their significance in the total cost of an ‘average’ building, then the effect of changing resource prices (also weighted) calculated for each category.

To take one example, for electrical work labour is weighted 40% and material 60%; the weighted index of electrical work costs is then given a weight of almost 10% in the total index.

The final cost index produced is therefore a weighted index of weighted indexes. So much averaging is bound to lead to a very broad ‘indicator’ indeed; a rise of say 5% in the building cost index can give only an approximate indication of the general effect of increases in building prices on the costs of a particular building.

As well as these nationally available series, many large client bodies and quantity surveying practices produce indices for their own purposes; this is because indices produced in-house, based on data which is primary rather than secondary to the user, are felt to be more reliable.

Manipulating the indices

The indices are used by building economists in a number of ways and for a number of purposes. The technique of manipulating them is in principle very simple but the results have to be interpreted very carefully.

Using indices to update building prices for budgeting purposes

We have data for a building of a particular type in a particular place and wish to get some idea of what a similar building might cost at a different time in the same place. The calculation may be set out as follows:

Cost per m^2* of comparison building 1989=£500:

All in Tender Price Index (1985 mean=100)

1989 4th quarter: 134

1994 4th quarter: 110

Therefore the probable cost to the client of a similar building in 1994=£500×110/135=£407.4 per m².

Note how important it is to use the right index. If we had used the BCIS general building cost index the answer would have been very different.

General building cost index	1989:4th quarter: 127
	1994:4th quarter: 154

The calculation therefore would be: £500×154/127=£606.3 per m².

The difference between the two results shows that while tender prices fell as a consequence of severe competition for work, basic resource costs continued to rise. It is not surprising that many contractors went out of business in these circumstances.

Calculation of price fluctuations

The contractual arrangements for building projects specially over 12 months’ duration and of significant value usually allow fluctuations to be accounted for. The traditional way of allowing for fluctuations was to record the basic costs of the various resources needed in the project at the time the contractor priced for them (the tender date) and at the time of incorporation into the project and then paying or retaining any differences. Although this is an arithmetically straightforward process, it is extremely time-consuming. As a result, the National Economic Development Organization initiated an index-based approach reflecting changes in resource costs for a number of works categories. This approach, known as Formula Price Adjustment, has proved extremely effective in dealing with fluctuations during the construction. In this case of course it is the building cost indices which are relevant.

Assessment of changes in cost and price relationships

Indices can be used to estimate the consequences of changes in costs and prices resulting from using different construction methods or building in different locations.

For example the building cost indices show that costs for concrete-framed construction since 1989 have increased less than the costs of either steel-frame or brick construction. It does not of course follow that concrete frame is cheaper; this is a very easy trap to fall into, especially when reading from graphs of index numbers on which the lines inevitably meet at the level representing the base year. Whether one is cheaper than the other depends of course both on what the relationship was at the base year as well as on how prices have moved since.

Extrapolation

Indices can also be used, through extrapolating past trends, to predict future prices. It is a technique which needs to be used very cautiously and can, at points of change in the building cycle particularly, lead to serious errors. For example, predictions made in August 1989 on t...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Foreword
• Acknowledgements
• About the authors
• Introduction
• I: Basic issues in the design and economics of building
• II: The economic context: Materials, labour and physical capital
• III: Economic aspects of design decisions
• IV: Relating design choices to building and its management
• V: Cost limits and values
• 1: Introduction to the tender price and building cost indices
• 2: SfB Classification System: Basic tables
• 3: Example of a cost plan
• Bibliography