A more reasonable explanation is that the often-cited restricted entry practices and other supply control policies of the building trades unions have intensified. While its full impact is difficult to evaluate, the rapid expansion of hiring halls during the past decade has certainly increased the unions' control of the workforce. In the wake of a National Labor Relations Board ruling in the late 1950's and a subsequent Supreme Court decision clarifying the status of hiring halls, many craft unions pressed hard for their adoption. Following a number of strikes over the issue, hiring halls have become relatively widespread. Although, technically, few are wholly "exclusive," as a practical matter, they afford the unions substantial control over the flow of workers to the job.

Concurrent with the unions' increased bargaining power, the contractors' bargaining power has declined as interest rates have risen. The carrying costs of a construction strike, and hence the price an owner and his contractor are willing to pay to avoid it, have risen significantly. (Nevertheless, contractors have accepted many costly strikes during the past year rather than agree to initial wage demands.*) With normal competitive market forces being weakened and the bargaining balance tilted, the phenomenon of accelerating wage settlements is more comprehensible.

The Problem Intensifies

The construction labor cost situation is already acute, but it will almost certainly get worse over the next two years. During the year ended December 31, 1969, wage rate increases (excluding fringes) for members of building trades unions averaged 9.4%. Approximately one-third were covered under new collective bargaining agreements and received first-year increases averaging 14%. The remaining two-thirds received increases averaging 7.4% under old contracts. (Table 2) This year, old contracts call for three-fifths of those under collective bargaining agreements to receive increases averaging 10.3%. If the remainder, who will negotiate new contracts in 1970, obtain no more than last year's average of 14%, the total increase in union rates will be 11.6%. Unlike the manufacturing wage contracts which are usually front-loaded, the contracts negotiated in the building trades in 1968 and 1969 scheduled large deferred wage increases. This will make it extraordinarily difficult to keep the overall increase in the trades unions rates under 13% during 1971, even presuming that the economy-wide inflationary pressures are brought under control.

1 Bureau of Labor Statistics survey of 7 major building trades in 100 cities.

NOTE: Includes most of the workers covered by those contracts involving 1,000 or more workers.
Source: Bureau of Labor Statistics and estimates.

4 During periods of tight construction labor markets, workers technically on strike have no difficulty obtaining another job during the period of strike. Consequently, while the economic pressure of a strike on the contractor is extraordinary, the losses to the union worker are often insignificant.

5 A contract is front-loaded when the wage increases in the early years of the contract are higher than in the subsequent years.

It is almost certain that during the next two years the premium which building trades unions receive over industrial unions will escalate. In 1969 manufacturing production workers covered by major collective bargaining agreements had wage increases amounting to only 5.8%. This resulted from an abnormally low proportion of workers negotiating new contracts last year. The vast majority received deferred increases amounting to 4% plus cost-of-living adjustments. In 1970 there will be a million more workers negotiating new contracts. If the pattern of front-loading continues and first-year costs average as high as 9%, wage rates for major unions in manufacturing will rise by 7.2% this year. The increase in 1971 is apt to be less. Thus, while union construction wages increased 3.6% more than manufacturing wages in 1969, by 1971 they will be increasing at a 6% faster rate.

Although only three-fifths of construction workers are under collective bargaining agreements, non-union average hourly earnings in construction have generally paralleled the trend of union wage rates. In addition, wage gains of construction office and clerical employees have apparently kept pace with those of construction workers. The union increases during 1970 and 1971, however, are unlikely to be fully matched by the white collar workers in construction. Nevertheless, the total construction payroll will still be dominated by union wage gains.

Thus, unless productivity in contract construction improves at a far more rapid rate than in recent years, 7 unit labor costs will accelerate their rate of increase. This means that the cost of site fabrication (which is 75% labor cost) should rise by more than 9% this year and by more than 10% during 1971. Final prices of structures should rise somewhat less owing to the far slower acceleration of construction materials costs. Construction prices, however, should continue to rise far faster than industrial prices.

The Mid-1970's

What of 1972 and beyond? Will increases in construction wages and prices continue to accelerate relative to industry? What will happen if they do, and what is required to prevent it?

Although the already huge construction wage settlements will exert some pull on union agreements in the rest of the economy, there is no reason to suppose that the industrial trade unions will be able to match the settlements achieved by the building trades. Recent bargaining agreements in manufacturing have shown no evidence of any closing of the gap. This is certainly not the result of any lack of interest in larger settlements on the part of the industrial unions. It is rather a function of the more balanced bargaining power of the two parties in industry.

Moreover, since industrial bargaining agreements are generally front-loaded, wage rate increases will abate far faster in manufacturing than in construction where front-loading is minimal. It is thus likely that when the economy-wide inflation finally subsides, the construction wage premiums will accelerate for a while.

Parenthetically, the very existence of evenly spaced wage increases in construction underscores the gross differences in bargaining power balance between construction and manufacturing. Contract construction encompasses many firms, the bulk of them quite small, confronting powerful unions able to exert significant control over the supply of workers. Many of the contractors, being small and having limited capital, are necessarily concerned primarily with the jobs in progress or immediately prospective. Accordingly, they are often pressed to trade off somewhat lower immediate wage increases which would reflect themselves in the current operations against escalations in the second, third and sometimes fourth year of a contract. Contractors cannot readily pass on wage increases on work already contracted for or in progress as can, for example, a company manufacturing a product. The unions, on the other hand, can take a longer view and need not heavily discount gains negotiated for the later years of contracts.

This is the reverse of the situation typical in manufacturing where the company is the more permanent institution, programming its profits over the long-term and willing to trade off higher short-term wage rates against lower increases in the future. Their unions, on the other hand, are apparently far more interested in achieving short-term gains. Thus, with construction wages accelerating and no prospect of any major offset from productivity, unit labor costs should continue to rise sharply relative to those in the industrial and service areas. This must inevitably mean that the price of structures will continue to advance far faster than the price of goods and services generally.

6 Cost-of-living escalator adjustments in construction are rare.

7

The data on productivity in construction are very poor, but approximations indicate that there has been very little increase in recent years. (See Appendix Table A-20)

The Consequences

Already the modest priced single-family home is beginning to disappear as the costs of construction and developed-land price it out of the market. The cost of housing has risen sharply in recent years and f building wage cost trends are not altered the prospect for any major slowing in the years immediately ahead is poor. Such a trend would inevitably accelerate the growth in factory built homes and/or mobile homes. Americans might also forego their usual emphasis on high quality housing and allocate their resources to other segments of the consumer markets where their money would buy relatively more.

The residential site construction market conceivably could dry up to such an extent that rising unemployment in the building trades might finally induce a more realistic wage policy on the part of the trade unions. 8

In the area of public construction

dams, highways and schools - rapidly rising costs are being borne by the taxpayers who are thus indirectly subsidizing the building trades unions.

The effects of rising industrial plant costs, though complex, are highly important. If the price of plants rises faster than the price of the goods produced by such plants, the rate of return falls, depressing the incentive to initiate technological improvements. Over the long-run this would slow the growth in productivity and living standards. It would also induce a greater emphasis on building plants abroad.

Table 3. Differences between Union Construction Scales and Straight-Time Average Hourly Earnings
of Maintenance Workers, 3 Trades in Selected Metropolitan Areas, 1955 and 1966
Excess of construction rates over maintenance straight-time average hourly earnings for-
Carpenters

Region and metropolitan

Electricians
Percent excess
1955 1966

Painters

We have asserted that the building trades unions have apparently exercised a significant degree of monopolistic control over certain major segments of the construction labor market. What is the evidence that such monopolistic power exists and what is its source?

The folklore on the building trades' restrictionist practices is, of course, voluminous. The existence of many of these abuses, alleged or otherwise, very often hinges on hearsay evidence or fanciful reporting. But we need not appraise case by case the alleged instances of restrictionism, featherbedding or abuses of

8 An ostensible alternative to the eroding of residential site construction would be a massive federal subsidy which, in effect, would amount to partial subsidization of the building trades unions' high wages. This is scarcely solution, however, since it is fighting the symptoms of the problem. Moreover, the subsidy costs would soon become so extraordinarily high as to be fiscally impractical.

power. Even if most, if not all, were verified, we would still have the problem of tracing what consequences, if any, these actions had on the labor market.

A more direct approach is to ask first whether there is in fact evidence within the construction labor market itself of severe market distortions. Here the issue is unequivocal: extraordinary differences exist between hourly union wage rates and non-union wage rates by craft in construction, and between construction wage rates and manufacturing wage rates for relatively comparable maintenance crafts.

Table 4

Average Hourly Union and Non-Union Wages in Selected Building Trades Occupations, 1936

We estimate that average hourly earnings for union members in contract construction run approximately 50% greater than for the comparable crafts in the non-union sectors of contract construction. The Bureau of Labor Statistics in a study for the year 1965 showed that contractors paid an average of $4.95 per working hour in wages and benefits to building construction workers in establishments

where the majority of the workers were covered by collective bargaining agreements. The payment in establishments where none or a minority of workers where covered amounted to only $3.12 an hour; that is, union compensation ran nearly 59% above non-union. Since union members received far higher fringe benefits, the gap in payroll hourly earnings was only 52%. Our analysis indicates that part of this huge discrepancy results from the fact that there is a disproportionate number of lower paid crafts and skills in the non-union establishments. More laborers and fewer skilled crafts are represented. Even when this adjustment for craft and occupation bias is made, the union premium still runs close to 50% over the open shop.

10

Moreover, these gaps are not the result of some peculiar regional averaging. Recent studies suggest large wage differences in New York City between union and non-union painters and plumbers. Apparently very substantial differences also exist for all trades in a number of Canadian cities.11Evidence of major discrepancies between union and non-union wage rates by occupation within localities is not particularly new. A 1936 study by the Bureau of Labor Statistics evidenced very much the same results we are seeing today. (See Table 4)

To test the stability of these differences over time, we have attempted to disaggregate the data on construction worker employment and earnings, distributing the contractor establishment employment by occupation and union participation. (See Appendix for details.) The building trades craft wage levels were applied to the union portion and we assumed that non-union wages by craft maintained the same percentage relationship to union wage levels as prevailed in the 1936 study. The derived average hourly earnings for all construction workers by the various sub-industry divisions of contract construction were surprisingly close to the reported actual levels. Thus, the dimension of the gap has apparently not materially altered since 1936.

Similar differences exist between union scales in the building trades and the comparable crafts in maintenance work. For example, a recent study by the BLS of carpenters, electricians and painters showed an excess in this regard of roughly 40% for construction carpenters in 1966 and the major cities across the nation. The average premium for electricians was even higher, while the differential for painters varied city by city but was nevertheless quite substantial. In general, these percentage gaps have increased since an earlier 1955 study, confirming the acceleration of building trades wage rates relative to nonconstruction trades (See Table 3).

Is it possible for such extraordinary differences to exist without implying restrictive practices on the part of the building trades unions? One may validly argue that a disproportionate percentage of the most skilled workers within each craft are members of unions and hence would command some differential even if all of contract construction were open shop.

Moreover, it is often maintained that the very substantial gap in hourly wages between construction trades and other industries results from the high seasonality and intermittency of construction work such that annual earnings of construction workers are barely at parity with industrial craftsmen. The typical construction worker is unemployed part of the year as he shifts from one job to another and the high hourly rates are a market compensation for this phenomenon. There is, of course, no question that this explains a part of the high average construction wage relative to industry in general. It also partly explains the differentials between construction and maintenance wages. It does not explain however, the acceleration of the differences in recent years or the union/non-union differentials within construction. 12 To Sum Up

After all the conceivable qualifications one can imagine, it remains very difficult to dismiss the hypothesis that some extraordinary market imperfections exist.

The virtually complete control of new construction projects in many major cities by union establishments has been taken for granted for decades. Very little has been done to analyze its effects on

10 See reference in forthcoming BLS study on Compensation in the Construction Industry: Employment Patterns, Union Scales, and Earnings.

11

12.

See Appendix Footnote 6.

The ability of non-union wages in construction to keep pace with union rates probably reflects a considerable amount of open shop work by union members. There appears to be enough worker "arbitrage" to keep the differences relatively uniform.