consecutive years. The membership weights in both of the aggregates used for each year-to-year comparison were those reported for the second year. The total of the current year's aggregates was then divided by the total of the previous year's aggregates and the relative so obtained was applied to the previous year's index number. The result is the current index number. index numbers were constructed on this basis in order to minimize the influence of shifts in union membership which might obscure the real changes in wages and hours. It should be noted that the data do not reflect bonus payments or rebates.

Other Components of Building Costs

The

Although the cost of building materials and labor accounts for the bulk of the total cost of residential building, certain other items represent significant elements. The purchase of land, its improvement, the provision of utilities, and contractors' and subcontractors' overhead and profit all figure in the final cost to the purchaser. The distribution of these costs will, of course, vary widely for different classes of dwellings. The general relationships between these elements of capital costs which would be applicable to the typical house within a reasonable range of $5,000 were shown in the NHA bulletin, Housing Costs," referred to above.

Housing Costs, National Housing Agency, December 1944.

748221-48

These data cannot be accepted as accurately reflecting postwar cost distribution or levels, since they relate to a period prior to the imposition of wartime controls on manufacturing, distribution, and construction industries; the widespread shortages, shifts in material use patterns, and general dislocation of the building industry which accompanied the war have no doubt effected important and persistent changes in the distribution of costs. It is believed useful, however, to present one table (No. 28) from that study, for the light it throws on the types of costs which enter into total dwelling unit costs. This table shows the distribution in 1940-41 of costs and profits which enter into the delivered price of materials, as well as the cost of labor, the cost of an average lot and improvements in an urban area and contractors' and subcontractors' overhead and profits. Each item in the table is expressed as a percentage of the total cost of house and land.

The relationship between the delivered price of material and labor, overhead and profit, and land, was developed by the Federal Housing Administration, and represents a composite of typical houses selected in various parts of the country. In preparing this material FHA collected detailed cost data on a Nation-wide basis for the elements entering into residential construction. The data for each area were weighted by the estimated total number of houses built in the area from 1935

31

Source: National Housing Agency, "Housing Costs," December 1944.

to 1940. Hence, the composite cost distribution is believed to reflect an appropriate weighting of the regional differences in construction practices throughout the country.

Data relating to the breakdown of material costs were obtained from unpublished studies of the Office of Price Administration, and, in general, represent 1940-41 conditions. These data were checked with information from the Bureau of Labor Statistics, the Federal Trade Commission, and other sources, both public and private. The distribution of material costs shows, for each of 14 groups of building materials and equipment (cooking ranges and refrigerators are excluded), the delivered price at the site of construction of all materials and equipment, the cost of manufacture, excluding the manufacturers' selling expense and profit; the cost of distribution, including the selling expenses of manufacturers, wholesalers, and retailers; the cost of transportation including freight or truck to the retail outlet; and finally, the combined profits, including profits of the manufacturers, wholesalers, and retailers but not profits on transportation.

The cost of manufacturing lumber, for example, is shown to be 4.19 percent of the total cost of the property, whereas the cost of lumber delivered at the building site is 11.85 percent of the total cost. The total delivered price of all materials, repre

senting 45.7 percent of the cost of property, is composed of (1) the cost of manufacture which amounts to 20.9 percent of the final cost; (2) the cost of distribution which amounts to 14.8 percent; (3) the cost of transportation which amounts to 3.9 percent; and (4) the combined profit which represents 6.1 percent.

The Average Cost of the Dwelling Unit

A continuing stream of building materials and construction labor goes into the production of dwellings. It is the aggregate cost of these items, together with the overhead costs and builders' profits, which result in the dollar volume of privately financed residential construction discussed in chapter II. The magnitude of this figure for any given period depends primarily on the number of dwelling units being constructed, but it is influenced by many other factors such as the prices paid for labor and materials. The quantities of the various materials likewise vary, as does the amount of labor required to incorporate them in the structure. Depending upon local practices, upon shifting standards of housing and builders' response to consumers preferences and ability to pay, there may be varying amounts of particular materials used. By the same token site labor will also vary in amount.

The Bureau of Labor Statistics attempts to record the effect of these changes by computing the average valuation of dwelling units as stated by builders on the building permit records issued by municipal building departments. These averages (ch. II, tables 12 and 13) are a byproduct of the regular BLS permit reporting system and are derived by dividing the number of dwelling units reported into the total of the stated valuations.

The figures are subject to several limitations which restrict their use in the study of building cost trends. First, the averages relate only to permit issuing and reporting places; they therefore do not reflect the effect of the substantial amount of building in outlying, nonreporting places. This omission can be significant because the types and costs of structures erected in such places may be quite different from those in more organized urban places, where building codes and lot use restrictions are imposed. are imposed. Second, the figures represent permit values and as indicated in chapter II, these are generally understated. The understatement itself varies from time to time

depending upon builders' ability to forecast prices and wage costs.

Finally, the usefulness of the figures is distinctly limited because no break-downs are available to show the distribution of cost between labor, materials, and other costs.

These limitations are serious for the purpose at hand, and it is not likely that they can be overcome at this state in the development of the permit reporting system. This is unfortunate, because average costs based on actual records would have an obvious advantage in reflecting the composition of costs for dwellings actually being built.

Indexes of Residential Construction Costs

In contrast to the average value series discussed above, are the indexes of residential construction costs prepared by E. H. Boeckh & Associates, Inc., and the National Housing Agency. These indexes attempt to eliminate the effect of changing levels of building activity as well as of changes in the type or characteristics of the units. They

The Boeckh organization has been interested primarily in providing aids to local appraisal, and has not compiled a national index. (The NHA, however, has compiled a national index from Boeckh data; this is discussed below.)

The data shown in tables 29, 30, 31, and 32 cover, for 20 cities, all residential properties ranging from single-family residences to 6-family apartment buildings. The basic items included in the Boeckh general index grouping are: Brick, lumber, cement, steel, common labor, brick masons, carpenters, structural-iron workers, plasterers, and State or local taxes, if any. The index takes into consideration the relative importance of the items

which are included in various types of structures.

Each month the Boeckh organization obtains basic cost data on materials from local building material dealers and from local building trade papers. Prevailing wage rates for both common and skilled labor are obtained primarily from local contractors and building trade associations. An arbitrary labor-efficiency correction factor is used, based on the organization's study of labor conditions in each area. The index is computed and reported as of the first of each month.

A feature of the index is the use of a common base for all cities; the index figures thus provide comparisons between cities as well as between different periods of time. The effect of location in relation to material prices and construction costs can be illustrated by comparing the cost of building lumber in Seattle, Wash., and Pittsburgh, Pa., in August 1935. In Pittsburgh, the price was $46.25, while in Seattle the same lumber could be bought for $20, thus making a difference of $2,175 in the cost of identical frame houses, provided all other prices were the same.