wage surveys were used by the Department in only 2% of the 372 cases in his sample. It would appear that the department finds it more convenient to rely on readily available union wage rates, the fact that this is a violation of its own administrative procedures notwithstanding.

This unjustifiable practice of complacently relying on readily available union negotiated contracts (which are filed with the Labor Department) to make prevailing wage determinations is undoubtedly seriously undermining our national effort to provide adequate housing for low and moderate income groups through the various interest subsidy programs. Although the homebuilding segment is much less unionized than the construction industry generally, with less than 50% of the labor force carrying union cards, the Gujarati study found that 97% of the determinations for bricklayers, 93% for plasterers, 97% for electricians, 95% for plumbers and 90% for painters were based on union rates. Yet these are the major crafts that effect the cost of home construction and crafts in which there are often significant differentials between union and nonunion wage rates. As a result, Davis-Bacon determinations, in many instances, have almost entirely off-set the reduced cost to homebuyers provided by the interest subsidy.

Professor Yale Bronzen, for instance, recently reported the case of a high-rise apartment developer in Prince George's County, Maryland, who had built two successful apartment complexes for the commercial market and had planned to build a similar complex for moderate income families under the Federal interest subsidy program. However, he was forced to abandon the project after finding that the increased costs due to complying with the Davis-Bacon Act would necessitate higher rentals on the subsidized units than on his earlier commercial units, despite the considerable savings in interest cost. The adverse impact of the Davis-Bacon Act on our low-income housing program thus provides a particularly vivid illustration of the harmful consequences of unnecessary governmental intervention in the market place; in this instance it not only rewards a few relatively well-off producers (union building tradesmen) at the expense of millions of consumers, but also provides one more disheartening example of one arm of government actively undermining the work of the other.

The final way in which the Davis-Bacon Act has been mal-administered is probably not as important as the first two, but is nonetheless significant. The construction industry is separated into at least 17 major crafts, that can perhaps be envisioned as vertical divisions, and also a number of "segments" such as homebuilding, commercial, heavy industrial, highways, and the like, which form horizontal lines of division. Typically each of the scores of squares in this resulting matrix have widely differing wage rates, fringe benefits, and work rules. For this reason, the Act explicitly states that the prevailing wage for a contract shall be that paid to "corresponding classes of laborers and mechanics employed on projects of a character similar to the contract work".

Failure to recognize these differentials can seriously affect the cost of a construction project. Typically, the rates can range from less than $5.00 per hour paid to laborers at the bottom, to sometimes $10 to $12 an hour to operating engineers, who are generally the highest paid craftsmen. Moreover, the differentials according to types of construction-heavy, highway, residential-within a craft may be of similar magnitude. To take one concrete illustration, in southern Alabama during 1971, bricklayers were paid $4.55 on residential construction and $8.12 an hour on industrial construction work; similarly, carpenters were paid $4.00 an hour and $7.33 for residential and industrial construction work respectively, and plumbers $5.50 and $8.72 for the two types of work. Despite the substantial costs differences implicit in these differentials, however, the GAO has found that on repeated occasions the Labor Department's determinations have ignored these important distinctions.

Thus one report stated ". . . as has been shown in this report and in assoclated reports to the Congress, it has been the practice of the department to determine the higher negotiated rates paid on commercial type building construction as the minimum (prevailing) rates for federally financed housing construction instead of the lower rates prevailing for similar private housing construction in the project area." In this particular report, which covered a number of contracts in the New England states, the GAO found that these inappropriate determinations due to the failure to distinguish between types of construction resulted in wage rates that were anywhere from 8% in the case of masons, to 71% in the case of ironworkers, above the rates that actually prevailed for residential construction. On the average, the study concluded, the government was

paying hourly rates on residential construction projects nearly 33% in excess of those that actually prevailed for private residential construction because of the Department's inappropriate use of the significantly higher commercial construction data.

In yet another study issued in July of 1971, the GAO provided an estimate of the cumulative costs of the Department's continuing mal-administration of the Act. Studying 29 major Federal construction projects over the last decade amounting to nearly $88 billion, it concluded that the government had paid $9 billion in excess costs due to faulty administration of the Act alone. Put differently, nearly 10% of the cost of these projects stemmed from improper ap plication of a law that, as the following section will attempt to demonstrate, is of highly questionable validity anyway. At a time when programs of far higher priority are pressing hard for every penny of governmental revenue in sight, it is difficult to find justification for practices that line the already ample pockets of the building trades aristocracy at the expense of so many other Americans who are in far greater need.

The economic case against Davis-Bacon

If the problem were solely one of mal-administration, or of excessive union influence over the "prevailing wage" determination process, new administrative regulations or Congressional clarification of the provisions and objectives of the Act might be the suitable remedy. However, in my view the problem goes much deeper.

In my view, the imposition of "prevailing wage" requirements is fundamentally incompatible with the effective operation of a free market economy; it inevitably leads to inflationary pressures, mis-allocation of resources, protection from competitive forces for the favored producer group (in the case of the building trades union) and a reduction in overall economic output and public welfare.

In an ideal competitive market the function of a price (or wage) is to clear the market or establish an equilibrium between supply and demand. However, a true equilibrium price (which signifies optimal use of resources) will not emerge if either supply is artificially limited or if significant buyers on the demand side are not permitted to offer the lowest bid that would attract the necessary supply. As a consequence of either or both of these impediments, the price or wage level will be artifically high, resulting in either lower output or excessive use of resources in the market involved to the detriment of over-all economic activity. In the construction industry these impediments are present on both the supply and demand side of the market for skilled labor. On the supply side, tight union control of apprenticeship and referral programs sharply limits the amount of manpower available, thus driving up wages above the equilibrium level. If the government were free to place contracts with the lowest bidder, some of this supply distortion would be off-set because reduced demand for high wage union labor would either induce some union members to work for lower wages in the non-union sector (which would presumably get a significant portion of government contracts) or draw in new entrants from outside the construction labor force. In either case, the average construction wage rate would be lowered relative to that which prevails under present market conditions. In turn, lower wage costs would either allow for an expansion of output relative to the current construction wage bill, or for a reduction of the current construction wage bill while maintaining constant levels of output. Both cases or some combination would represent real gain to the economy.

Under the existing organization of the construction manpower market, however, these impediments on each side of the market reinforce each other to produce an inordinately high and artificial average wage level. The restriction on supply makes it easier for unions to push up wage levels and the restriction on government demand-which accounts for 30% of total construction demandto the highest union wage rates tends to lock-in these artificially high rates and perhaps even lift them higher.

Using a model developed by Professor John P. Gould of the University of Chicago, I have attempted to develop an estimate of the dollar cost to the economy of the Davis-Bacon prevailing wage requirement. While this model is based on a number of simplifying assumptions that would not likely satisfy an exacting professional econometrician, I think it nonetheless provides a reasonable "ball park" approximation of the inflationary impact of the Davis-Bacon Act. The basic variables used in the model are: a) the degree of unionization of the construction labor market; b) the average wage differential between the union and the non-union sector; c) the shift in construction demand from the union to the

non-union wage sector that would result from repeal of the Act; and d) the average construction industry hourly wage rate. On the basis of various Labor Department studies and professional monographs the following values were used for the above items: 65% for the unionization variable, 33% for the differential between the union and non-union wage sector, 22.5% for the degree of total construction industry demand shift as a result of rescinding Davis-Bacon, and the average construction wage rate of $5.22 an hour for the year 1970, a sample year used throughout the calculations which follow.

The above variables were first fitted to a simple algebraic equation in which the 1970 average hourly construction wage rate equals the percentage of unionization multiplied by the first unknown variable signifying the average hourly union wage rate (Wu) plus the percentage of the labor force in the non-union sector multiplied by the second unknown variable signifying the average nonunion wage rate (W.). Solving the equation produces a value of $5.71 an hour for the union sector and $4.30 an hour for the non-union sector.

In order to determine the upward shift in the average construction wage rate due to the prevailing wage law, the equation was reformulated to account for the shift of demand from the union to the non-union wage sector (g) of the magnitude indicated above which would occur if the Davis-Bacon Act were repealed. This yielded an average construction wage rate of $4.90 an hour under non-DavisBacon market conditions.

(II):

=

AW (U-g)· Wu+[(1-U) +g]·Wn
AW=0.425($5.71)+0.575 ($4.30)

AWb=$4.90

The difference between the actual construction hourly wage rate in 1970, $5.22 an hour, and the rate that would have prevailed had not the Davis-Bacon Act been operative, $4.90, is 6.5%. Since the construction industry accounted for over 6.6 billion man-hours during that year, the approximate cost to the economy of the Act amounts to over $2.1 billion. In my view, this figure provides strong evidence that the Davis-Bacon Act is indeed a powerful engine of inflationary pressures. The $2.1 billion not only represents a substantial cost to immediate construction consumers (of which the government is the largest), but a cost that may well be magnified as it winds its way through the entire economy.

The need for change

As I have argued previously, the market is a delicate mechanism that can be easily skewed and impaired to the detriment of national economic health and public welfare by unwise or inappropriate governmental intervention. Sometimes the objective to be served by intervention is so compelling that the "costs" must necessarily be accepted. But in the case of Davis-Bacon I can see no such overriding public interest. It primarily benefits a segment of the work-force that in 1971 received average weekly earnings that were fully 75% above the average for the private economy generally and a segment in which new wage increases during the past three years were more than two times, and often three times, greater than the national average. Yet while conferring substantial benefits on this selected group of "producers", it exacts huge costs from consumers, especially housing consumers, injects an inflationary bias into the economy that along with many similar market restraints reduces our ability to approach full-employment without unacceptable inflationary pressures, and skims off substantial amounts of scarce Federal revenues that could otherwise be made available for more pressing purposes.

Senator TOWER. The next witness is Mr. Robert Smith, president of the National Association of Building Manufacturers.

Mr. Smith, your statement will be printed in its entirety in the record. If you would like to proceed in a summary fashion, you may do so, or you may read your statement, as you choose.

STATEMENT OF ROBERT SMITH, PRESIDENT, NATIONAL ASSOCIA TION OF BUILDING MANUFACTURERS, AND GENERAL MANAGER OF BOISE CASCADE MANUFACTURED HOUSING GROUP, ACCOM. PANIED BY RICHARD L. BULLOCK, EXECUTIVE VICE PRESIDENT, AND DAVID FALK, FROSH, LANE & EDSON, GENERAL COUNSEL

Mr. SMITH. Mr. Chairman, I am the president of the National Association of Building Manufacturers, headquartered in Washington, as well as the general manager of the Boise Cascade Manufac tured Housing Group, a company that has been active in the building industry.

I am here today, to testify on behalf of the National Association of Building Manufacturers, a trade association of over 400 companies, which collectively, produce about half of the factory-manufactured, or industrialized housing produced in the country.

Today, I am accompanied by Mr. Richard L. Bullock, executive vice president of NABM, and David Falk, of Frosh, Lane, and Edson, Washington, D.C., general counsel to our organization.

First, I wish to express the gratitude of our industry to your committee for holding hearings today on the important measures introduced by Senator Brock, for, as will be stated in the testimony, both S. 3373 and S. 3654 are of singular importance to the industrialized housing industry.

In 1970, this industry accounted for nearly 300,000 code complying housing units. Last year this figure increased to about 400,000 units. Similar growth is expected for 1972.

I might add these numbers do not include mobile housing units. Impressive as these figures are, they would be even more so, but for the problems of unwarranted restraints on industrialized housing.

The multiplicity of building codes prevailing in most local jurisdictions around the country presents a major problem for production of industrialized housing. The economic shipping radius of one factory may be several hundred miles.

This may embrace several States and hundreds of local political jurisdictions, each with an individual set of building codes, and a regulatory body for enforcement of these codes. The result is that the industrialized housing producer is faced with a wide array of differ ing and even contradictory standards and requirements in a single market area, making it impossible to design a housing unit that will meet the code-requirements of every jurisdiction.

Consequently, the manufacturer must constantly adjust his production processes to meet the requirements of the particular jurisdiction that he is supplying. The economics of industrialization are dissipated by this lack of uniformity.

Secondly, the manufacturer sometimes finds requirements in jurisdictions which he cannot effectively meet at all. In some jurisdictions, only one type of building material is permitted, to the exclusion of all others, without regard to their performance in actual use. Or, only one design of a building element is permitted. Or, to cite another typical situation, the building code permits the local code enforcement official to accept materials or techniques that are equivalent

in performance to those specified in the code, but the code enforcement officials are unwilling to exercise their authority in this respect.

Elsewhere, the codes specify inspection practices by code enforcement officials that interfere with the industrialized process. Most frequent are provisions that have the effect of prohibiting closed wall construction at the factory.

The situation I have described is wasteful. It deprives the producer, the developer, and the consumer of the full economic benefits inherent in industrialization in the building process.

Let me give you some examples of what I mean: In one mediumsize, midwestern city, Kalamazoo, a study showed that a new building would have to meet five different State codes, four different city codes; building, plumbing, electrical, and mechanical, pass inspections by seven different agencies, and get seven licenses, plus seven different State and city permits.

In Massachusetts, there is a plumbing board's directive requiring that housing built in a factory must have the plumbing exposed over every inch of its length so that it can be inspected on-sites. In such cases, there would be no sense in including the plumbing as an integral part of the manufacturing process, thus obviating an advantage of factory-produced housing.

In one case of which I am particularly aware, involving Kingsberry Homes, within our Boise Cascade Manufactured Housing Group, we had to bring a very costly Federal court action against the commissioners of a Georgia county who had denied us building permits because our manufactured homes differed from the local building code in two very inconsequential respects, such as whether roof sheathing should be three-eighth or one-half inch.

In order to win our case, we had to undertake extensive tests of our product, and to produce and present expert testimony. On hearing our evidence, the court ruled that the local codes were both unreasonable and unenforceable.

However, this specific example lasted almost 3 years, $10,000 in legal fees, not to mention involvement of management in untold time. In the labor area, problems are equally vexing. We are aware of many instances where labor contracts or work practices impede the use of factory manufactured housing or add such additional expenses, that our inherent cost efficiencies disappear.

An example of the problem is frequently seen where a union may refuse to handle or work with manufactured components at all, or unless they are first dismantled and rebuilt by employees at the job site.

Under certain circumstances, the National Labor Relations Board may find that such practices are illegal, secondary boycotts. However, the time involved in reaching these decisions make them mute in that the contractor cannot wait for the courts to so declare.

Another area of concern is the demand by a union for what amounts to "double coverage" on tasks which have been simplified by a new building code, into jobs requiring fewer men.

We believe that S. 3373 would be most beneficial in remedying some of the problems we have discussed, by prescribing unnecessary restraints of the use of industrialized housing in HUD programs.