_______________________________________________________________________
Part II
Department of Labor
_______________________________________________________________________
Employment Standards Administration, Wage and Hour Division
_______________________________________________________________________
29 CFR Parts 1 and 5
Procedures for Predetermination of Wage Rates; Labor Standards
Provisions Applicable to Contracts Covering Federally Financed and
Assisted Construction and to Certain Nonconstruction Contracts;
Proposed Rule
[[Page 17442]]
DEPARTMENT OF LABOR
Employment Standards Administration, Wage and Hour Division
Office of the Secretary
29 CFR Parts 1 and 5
Procedures for Predetermination of Wage Rates; Labor Standards
Provisions Applicable to Contracts Covering Federally Financed and
Assisted Construction and to Certain Nonconstruction Contracts
AGENCY: Wage and Hour Division, Employment Standards Administration,
Labor.
ACTION: Proposed rule.
-----------------------------------------------------------------------
SUMMARY: This document is a proposal resulting from the reexamination
by the Wage and Hour Division, Employment Standards Administration,
U.S. Department of Labor (Wage and Hour) of regulations previously
issued to govern the employment of ``helpers'' on federally-financed
and assisted construction contracts subject to the prevailing wage
standards of the Davis-Bacon and Related Acts (DBRA).
Based on the Department's experience both prior to and during
implementation of the suspended regulations, and a reexamination of the
reasons and data underlying promulgation of the suspended helper
regulations, Wage and Hour proposes to amend the regulations to
incorporate its longstanding policy allowing use of helpers only where
their duties are clearly defined and distinct from journeymen and
laborer classifications in the area.
DATES: Comments are due June 8, 1999.
ADDRESSES: Submit written comments to John Fraser, Deputy
Administrator, Wage and Hour Division (ATTN: Government Contracts
Team), Employment Standards Administration, U.S. Department of Labor,
Room S-3020, 200 Constitution Avenue, N.W., Washington, DC 20210. Any
commenters desiring notification of receipt of comments should include
a self-addressed, stamped post card.
FOR FURTHER INFORMATION CONTACT: William W. Gross, Director, Office of
Wage Determinations, Wage and Hour Division, Employment Standards
Administration, U.S. Department of Labor, Room S-3028, 200 Constitution
Avenue, NW, Washington, DC 20210. Telephone (202) 692-0062. (This is
not a toll free number.)
SUPPLEMENTARY INFORMATION:
I. Paperwork Reduction Act
This rule does not contain any new information collection
requirements and does not modify any existing requirements. Thus, the
rule contains no reporting or recordkeeping requirements subject to the
Paperwork Reduction Act of 1995.
II. Background
The Department's longstanding practice regarding the issuance of
helper classifications, apart from the periods, as discussed below,
when the suspended ``helper'' regulations were implemented, has been to
allow the use of helpers on construction projects covered by the labor
standards provisions of the Davis-Bacon and Related Acts only where (1)
the duties of the helper are clearly defined and distinct from those of
the journeyman or laborer, (2) the use of such helpers is an
established prevailing practice in the area, and (3) the term
``helper'' is not synonymous with ``trainee'' in an informal training
program.
On May 28, 1982, Wage and Hour published revised final Regulations,
29 CFR Part 1, Procedures for Predetermination of Wage Rates, and 29
CFR Part 5, Subpart A--Davis-Bacon and Related Acts Provisions and
Procedures (47 FR 23644 and 23658, respectively), containing the
following four new provisions intended to allow contractors to expand
their use of helpers on Davis-Bacon covered projects at wages lower
than those paid to skilled journeyworkers:
<bullet> A new definition of the term ``helper,'' allowing a
helper's duties to overlap with those of a journeylevel worker:
A helper is a semi-skilled worker (rather than a skilled journeyman
mechanic) who works under the direction of and assists a journeyman.
Under the journeyman's direction and supervision, the helper performs a
variety of duties to assist the journeyman such as preparing, carrying
and furnishing materials, tools, equipment, and supplies and
maintaining them in order; cleaning and preparing work areas; lifting,
positioning, and holding materials or tools; and other related, semi-
skilled tasks as directed by the journeyman. A helper may use tools of
the trade at and under the direction and supervision of the journeyman.
The particular duties performed by a helper vary according to area
practice. (29 CFR 5.2(n)(4), 47 FR 23667.)
<bullet> A provision allowing a helper classification to be
included in the wage determination if it was an ``identifiable'' local
practice. 29 CFR 1.7(d), 47 FR 23655.
<bullet> A provision limiting the number of helpers to two for
every three journeyworkers. 29 CFR 5.5(a)(4)(iv), 47 FR 23670.
<bullet> A provision allowing the addition of helper
classifications on contracts containing wage determinations without
helper classifications. 29 CFR 5.5(a)(1)(ii)(A), 47 FR 23688.
These regulations were challenged in a lawsuit brought by the
Building and Construction Trades Department, AFL-CIO, and a number of
individual unions. On December 23, 1982, the U.S. District Court for
the District of Columbia held that the new helper regulations
conflicted with the Davis-Bacon Act and enjoined DOL from implementing
the regulations. See Building and Construction Trades Department, AFL-
CIO, et al. v. Donovan, et al., 553 F. Supp. 352 (D.D.C. 1982). The
court held that the regulations improperly defined the helper
classification in terms of the level of supervision instead of in the
traditional terms of the tasks performed. Id. at 355.
On appeal, the U.S. Court of Appeals for the District of Columbia
affirmed in part and reversed in part. Building and Construction Trades
Department, AFL-CIO, et al. v. Donovan, et al., 712 F.2d 611 (D.C. Cir.
1983), cert. denied, 464 U.S. 1069 (1983). The court upheld the
Department's authority to allow the increased use of helpers and
concluded that the Secretary's regulatory definition of a helper was
``not clearly unreasonable.'' Id. at 630. However, the court struck
down the regulation allowing for the issuance of a helper wage rate
where helpers were only ``identifiable.'' Id. at 624.
On remand, the district court lifted the injunction as it applied
to the helper definition, but maintained it as to the remaining helper
regulations. The district court added that the Secretary ``may,
however, submit to this Court reissued regulations governing the use of
helpers, and if these regulations conform to the decision of the Court
of Appeals, they will be approved.'' 102 CCH Labor Cases para.34,648,
p. 46,702 (D.D.C. 1984).
In accordance with the district court's order, DOL published in the
Federal Register (52 FR 31366, August 19, 1987) proposed revisions to
the helper regulations to add the requirement that helpers must prevail
in an area in order to be recognized. After analyzing the comments on
this proposal, the Department, on January 27, 1989, published a revised
final rule governing the use of semi-skilled helpers on
[[Page 17443]]
federal and federally assisted construction contracts subject to the
Davis-Bacon and Related Acts (54 FR 4234).
On September 24, 1990, the district court vacated its injunction,
and on December 4, 1990, Wage and Hour published a Federal Register
notice implementing the helper regulations, effective February 4, 1991
(55 FR 50148).
In April 1991, Congress passed the Dire Emergency Supplemental
Appropriations Act of 1991, Public Law 102-27 (105 Stat. 130), which
was signed into law on April 10, 1991. Section 303 of Public Law 102-27
(105 Stat. 152) prohibited the Department of Labor from spending any
funds to implement or administer the helper regulations as published,
or to implement or administer any other regulation that would have the
same or similar effect. In compliance with this directive, the
Department did not implement or administer the helper regulations for
the remainder of fiscal year 1991.
After fiscal year 1991 concluded and subsequent continuing
resolutions expired, a new appropriations act was passed which did not
include a ban restricting the implementation of the helper regulations.
On January 29, 1992, Wage and Hour issued All Agency Memorandum No.
161, instructing the contracting agencies to include the helper
contract clauses in contracts for which bids were solicited or
negotiations were concluded after that date. On April 21, 1992, the U.
S. Court of Appeals for the District of Columbia invalidated the
regulation that prescribed a ratio of two helpers for every three
journeyworkers as being without sufficient support in the record, but
upheld the remaining helper provisions. Building and Construction
Trades Department, AFL-CIO v. Martin, 961 F.2d 269 (D.C. Cir. 1992). To
comply with this ruling, on June 26, 1992, Wage and Hour issued a
Federal Register notice removing 29 CFR 5.5(a)(4)(iv) from the Code of
Federal Regulations. 57 FR 28776. Further advice regarding
implementation of the helper regulations in light of the lifting of the
appropriations ban and the court action was given in All Agency
Memorandum No. 163, dated June 22, 1992, and All Agency Memorandum No.
165, dated July 24, 1992.
Subsequently, Section 104 of the Department of Labor Appropriations
Act of 1994, Public Law 103-112, enacted on October 21, 1993,
prohibited the Department of Labor from expending funds to implement or
administer the helper regulations during fiscal year 1994.
Accordingly, on November 5, 1993, Wage and Hour published a Federal
Register notice (58 FR 58954) suspending the regulations governing the
use of semi-skilled helpers on DBRA-covered contracts, and reinstating
the Department's prior policy regarding the use of helpers. The
Department of Labor Appropriations Act for fiscal year 1995 again
barred the Department from expending funds with respect to the helper
regulations. Section 102, Public Law 103-333. That prohibition extended
into fiscal 1996 as a result of several continuing resolutions. There
was no such prohibition in the Department of Labor's Appropriations
Acts for fiscal 1996 and 1997, Public Law 104-134, enacted on April 26,
1996 and Public Law 104-208, enacted on September 30, 1996.
On August 2, 1996, Wage and Hour published in the Federal Register
(61 FR 40366) a proposal to continue to suspend the implementation of
the helper regulations while additional rulemaking procedures are
undertaken to determine whether further amendments should be made to
those regulations. On December 30, 1996, a final rule was published in
the Federal Register (61 FR 68641) continuing the suspension. Pursuant
to that final rule, the November 5, 1993 suspension of the helper
regulations continues in effect until Wage and Hour either (1) issues a
final rule amending (and superseding) the suspended helper regulations;
or (2) determines that no further rulemaking is appropriate, and issues
a final rule reinstating the suspended regulations.
By decision dated July 23, 1997, the U.S. District Court for the
District of Columbia upheld the Department's December 30 final rule
continuing the suspension of the helper regulations until the
completion of rulemaking proceedings. Associated Builders &
Contractors, Inc. v. Herman, C.A. No. 96-1490, 1997 WL 525268 (D.D.C.
July 23, 1997). The Associated Builders and Contractors had filed suit
challenging the Department's failure to immediately reinstate the rule
when the appropriations ban was lifted. The district court dismissed
the suit, ruling that any error in failing to act immediately to issue
a new effective date for the rule was mooted by the suspension
rulemaking completed in December. The court observed that the
Department was not required to ignore changed circumstances in the two-
and-a-half years since the rule was last implemented, and went on to
hold that the December rule was a valid rule, supported by the record,
and consistent with the requirements of the Davis-Bacon Act.
III. Discussion
During the period following the passage of the appropriations act
for fiscal year 1996, Wage and Hour has carefully considered whether
the suspended regulations governing the use of helpers should be
modified. Seventeen years have passed since Wage and Hour first
promulgated the regulations, and more than five years have passed since
the Department's last attempt to put a revised version of those
regulations in effect was curtailed by legislative action. The final
helpers rule, which first became effective on February 4, 1991, was
originally proposed and adopted because it was believed that it would
result in employment practices on federal construction projects that
more closely mirrored the private construction industry's practice of
using helpers, which was assumed to be widespread, and would at the
same time effect significant savings in federal construction costs. It
was also believed that the expanded helper definition would provide
additional job and training opportunities to unskilled workers, in
particular women and minorities.
Implementation of the suspended helper definition and development
of enforcement guidelines proved, however, to be more difficult than
was anticipated, particularly in light of the court-ordered abandonment
of the ratio provision.
Furthermore, the Department's experience with surveys conducted to
implement the regulation and information from the surveys, and other
data sources which were previously unavailable or not examined,
indicated that the use of helpers was not as widespread as previously
thought. Wage and Hour was also concerned about the possible negative
impact of the suspended regulation on formal apprenticeship and
training programs. These concerns, and the controversy evidenced by the
rule's long history of litigation and by Congressional action over the
1989 final rule, led Wage and Hour to reexamine the basis and effect of
the semi-skilled helper regulations.
As the Circuit Court of Appeals noted in its 1983 decision
upholding the Secretary's authority to adopt a new definition of
helper, it is within the Secretary's province to alter or overturn
administrative rulings upon reconsideration of relevant facts. See
Building and Construction Trades Department, AFL-CIO v. Donovan, 712
F.2d 611, 629 (D.C. Cir. 1983). The court
[[Page 17444]]
also made clear the authority of the Secretary to choose from among
various regulatory programs the one he or she believes will best serve
the purpose of the statute. As the Court of Appeals acknowledged, the
Secretary is especially entitled to deference when his or her
``decision turns on the enforceability of various regulatory schemes.''
Donovan, 712 F.2d at 629. An important factor to consider in making
that choice is whether a particular regulatory scheme is sufficiently
capable of practical and efficient administration and enforcement to
achieve the statutory goal.
Wage and Hour has preliminarily concluded, after a full review of
the suspended rule and all available information, that it is likely
that the suspended rule cannot be enforced effectively. Furthermore, a
key underpinning of the rule, that helper use is widespread, has been
seriously undermined by an examination of all available data sources.
Wage and Hour also believes that the suspended helper rule, if fully
implemented, could have a negative impact on apprenticeship and
training.
Wage and Hour therefore carefully considered a number of
alternative approaches, focusing particularly on consistency with the
purposes of the Act, enforceability, administrative feasibility, and
ease of compliance. Although not a primary consideration, Wage and Hour
also considered the potential impact of the various alternatives on
employment and training opportunities for unskilled workers, including
women and minorities. A necessary consideration was also consistency
with the Department's ``reinvention'' efforts to revise and improve the
Davis-Bacon wage determination process.<SUP>1</SUP>
---------------------------------------------------------------------------
\1\ Wage and Hour is currently considering two potentially
viable options:
(1) Through procedural changes and the application of
technology, reengineer the current wage survey system to make it
more efficient and to produce more accurate and timely wage
determinations.
(2) Use redesigned and expanded BLS survey instruments--the
Occupational Employment Statistics (OES) Survey and the National
Compensation Survey (NCS, formerly known as ``Comp 2000''), when
these are available, and modified as may be needed--for Davis-Bacon
prevailing wage/fringe benefits determination purposes. (The OES
survey would use government-wide Standard Occupational
Classification (SOC) definitions, which are currently undergoing
review. See 60 FR 10998 (February 28, 1995), 60 FR 52284 (October 5,
1995), and 62 FR 36338 (July 7, 1997).)
---------------------------------------------------------------------------
After a thorough review, Wage and Hour has preliminarily concluded
that the current, longstanding practice of recognizing helpers only
where they are a separate and distinct class with clearly defined
duties is the sole alternative considered that is both capable of
effective enforcement and administration, and at the same time fully
consistent with the purposes of the Act.
Comments are invited on the regulation proposed, as well as the
other alternatives considered, including the Department's analysis and
conclusions thereon.
Problems With the Suspended Helper Definition
1. The Suspended Helper Definition Would Be Difficult To Administer and
Enforce
Wage and Hour has preliminarily concluded that the suspended
regulation poses significant administrative difficulties, and cannot be
effectively enforced in a manner consistent with the goals of the
statute. The Department's experience in trying to develop enforcement
guidelines to implement the helper regulations during the period they
were in effect (from February 4, 1991 to April 10, 1991, and from
January 29, 1992 to October 21, 1993) has led Wage and Hour to conclude
that a supervisory-based, semi-skilled helper definition would be
difficult to administer and enforce consistent with the purpose of the
statute, namely to identify and preserve the locally prevailing wage
for construction job classifications.
The suspended regulation defines a helper, not by the traditional
test of the specific tasks performed by the worker, but as ``a semi-
skilled worker'' who ``may use tools of the trade at and under the
direction and supervision of the journeyman.'' The suspended helper
definition is the first and only instance of determining a Davis-Bacon
classification solely on the basis of the worker's skill level and
work-site supervision. Furthermore, the definition is internally
inconsistent in that the examples given of the types of assistance the
helper might provide to a journeyworker are not semi-skilled but rather
are largely unskilled duties commonly performed by
laborers.<SUP>2</SUP> Thus, the suspended definition specifically
allows extensive overlap with duties performed by both journeylevel
craft workers and laborers, instead of providing an objective means for
distinguishing between helpers and other classifications.
---------------------------------------------------------------------------
\2\ E.g., ``preparing, carrying and furnishing materials, tools,
equipment, and supplies and maintaining them in order; cleaning and
preparing work areas; lifting, positioning, and holding materials or
tools. * * *'' 47 FR 23667.
---------------------------------------------------------------------------
During the period the suspended regulation was in effect, Wage and
Hour tried to develop enforcement guidelines to implement the
regulation. A fundamental problem that emerged was how to make a
meaningful distinction between semi-skilled and skilled workers under
the suspended definition. Wage and Hour has traditionally identified
and differentiated among job classifications on the basis of the tasks
performed by each classification. Among the issues Wage and Hour
struggled with in trying to develop enforcement guidelines were: (1)
What it means to be semi-skilled; (2) how to identify the line between
a semi-skilled and skilled journeyworkers; (3) whether at some point a
semi-skilled helper could acquire sufficient skills to qualify as a
skilled worker, and how to determine when that had occurred; (4)
whether a skilled worker could accept a position as a semi-skilled
helper--and therefore be paid the lower helper wage rate--without
violating the regulation or the intent of the Act; and (5) whether
hiring as a semi-skilled helper a skilled worker who failed to disclose
his skill level would violate the regulation or the Act.
The supervision aspect of the suspended helper definition likewise
provides little assistance in distinguishing a helper from other
classifications of workers. The definition states that a `` `helper' *
* * works under the direction of and assists a journeyman. Under the
journeyman's direction and supervision, the helper performs a variety
of duties to assist the journeyman * * *.'' Supervision by a
journeyworker is not a practical standard for distinguishing semi-
skilled helpers from others on the worksite, as even laborers and
journeylevel construction workers may work under the ``direction and
supervision'' of other journeyworkers. The definition does not indicate
the nature or amount of direction and supervision that helpers must
receive to distinguish them from others on the worksite. The definition
similarly provides little meaningful guidance for distinguishing
between a ``semi-skilled helper'' who uses the tools of the trade, and
a journeyworker with little experience, thus increasing the instances
in which journeyworkers may be misclassified as helpers.
In addition, the definition's allowance of significant overlap
between the duties of helpers and those of laborers increases the
difficulty of identifying helpers as a distinct classification.
Although the definition states that a helper must be ``semi-skilled,''
the unskilled tasks listed in the definition as examples of a helper's
duties are
[[Page 17445]]
commonly performed by unskilled laborers. Thus, it would be difficult
to distinguish between a laborer and a helper when a worker is
performing only unskilled work. It may theoretically be possible for a
helper under this definition to be distinguished from a laborer if the
helper directly assists a particular class of journeyworker(s) and uses
the tools of the trade. However, based on a further review of the
duties of laborers who assist craft workers, together with the
Department's experience in conducting conformance surveys during the
brief period the suspended regulation was in effect, and the low wages
paid helpers in the Current Population Survey (CPS), Wage and Hour now
believes--contrary to its earlier assumptions--that many laborers also
assist journeylevel workers and that laborers sometimes use tools of
the trade to perform certain limited duties (e.g., demolition/removal
of materials, building of scaffolding or forms). The overlap of duties
therefore increases the likelihood that helpers will displace laborers,
or that laborers will be misclassified as helpers. For example, a
laborer working under the supervision of a journeyworker could be
classified as a lower-paid ``helper'' simply by adding to his or her
duties a few relatively low-skilled tasks.<SUP>3</SUP>
---------------------------------------------------------------------------
\3\ As set forth in the economic impact analysis set forth
herein, the Current Population Survey (CPS) indicates that average
earnings for helpers are less than the average earnings received by
laborers.
---------------------------------------------------------------------------
Wage and Hour recognized the subjectivity of the suspended
definition when it first proposed the helper regulations in 1981, and
sought ``to protect against possible abuse'' by proposing to establish
a maximum ratio of helpers to journeyworkers. Wage and Hour originally
proposed a 1:5 ratio, then settled on a ratio of 2 helpers for every 3
journeyworkers in the final regulation. (46 FR 41456, August 14, 1981;
47 FR 23658, May 28, 1982). While not a guarantee against
misclassification in any particular case, the ratio would at least have
decreased the likelihood of widespread misclassification between
journeyworkers and helpers and provided one objective measure for
compliance and enforcement. As the Court stated in its 1983 decision,
the ratio ``increased the likelihood that gross violations will be
caught, or at least that evasion will not get too far out of line.''
712 F.2d. at 630. In rejecting the 2:3 ratio in its 1992 decision on
the ground that the rulemaking record lacked adequate support for that
particular numeric ratio, the Court of Appeals deprived Wage and Hour
of the mechanism designed to mitigate the possibility of abuse.
What remains is a vague standard that Wage and Hour has
preliminarily concluded is not amenable to effective enforcement. Thus,
Wage and Hour believes that the suspended regulation does not define
helpers in a manner sufficient to differentiate readily between semi-
skilled helpers and journeyworkers or laborers, as a practical matter,
in day-to-day compliance and enforcement. Contractors would likely find
it difficult to apply the regulation in classifying their workers and
could find themselves unwittingly in violation of prevailing wage
requirements due to misclassification. It would also be difficult to
prevent unscrupulous contractors from taking advantage of the
uncertainties created by the definition by intentionally misclassifying
large numbers of workers.
The definitional problems discussed above are compounded by
evidence that the term ``helper'' has multiple, quite different
meanings within the construction industry. A review of comments
received in response to the Department's rulemaking proposal to
continue the suspension of the helper rule (61 FR 40366) disclosed that
some contractors use the term ``helper'' to refer to skilled workers
who are less experienced, i.e., those who use tools of a trade to
perform some tasks, but have not been trained in the full range of
journeylevel work. Others use the term to refer to workers who perform
unskilled laborer duties that are related to the work of skilled
journeyworkers, as a short-term entry level job, or as a longer-term
specialized worker to perform a limited range of work duties that
somewhat overlaps those of the craft journeylevel worker. Still others
use the term helper to refer to employees with little or no experience
in the construction industry, i.e., untrained entry level workers. Wage
and Hour believes that these variations in the use of the term helper
may exist in any given local area where use of helper classifications
is prevalent. Direct assistance to, or supervision by, a
journeyworker--the central component of the suspended regulatory
definition--does not appear to be an important consideration for
commenters in distinguishing helpers from other workers. Thus, it
appears that the suspended definition, and perhaps any regulatory
definition of helpers, does not adequately reflect the actual and
varied practice in the construction industry as a whole or even in any
particular area. However, Wage and Hour is interested in obtaining
further evidence regarding how helpers are in fact used by contractors,
particularly any data regarding whether there is in fact a generally
recognized definition of helpers that is capable of being objectively
identified.
Wage and Hour also believes it would be difficult for it to conduct
a meaningful wage determination process concerning helpers in light of
the likelihood that contractors responding to area wage surveys would
ascribe very different meanings to the term ``helpers.'' Thus, contrary
to basic principles of the Davis-Bacon Act, it is assumed that workers
who perform quite different work would likely be grouped together for
purposes of determining prevailing wage rates for a single class of
``helpers'' within a given area. Moreover, Wage and Hour believes that
some contractors may report workers as helpers, whereas other
contractors might report the same type of worker as a laborer or craft
journeyworker. Such data would not provide a meaningful basis for
determining prevailing wage rates for the affected classifications, as
required by the statute.
2. Helpers Are Less Widespread Than Previously Believed.
The belief that a distinct class known as ``helpers'' was in
widespread use in the construction industry was a key assumption
underlying the Department's development of the helper regulation.
Indeed, in the preamble to the proposed rule published in 1987, the
Secretary projected that helpers would be determined to be prevailing
in two-thirds to 100% of all craft classifications. 52 FR 31366, 31369-
370 (August 19, 1987).<SUP>4</SUP> The Department's actual experience
with the helper regulation reflects a different picture.
---------------------------------------------------------------------------
\4\ This was amended by the statement (without quantification)
in the final rule that this would be reduced somewhat to the extent
that collectively bargained rates were recognized as prevailing and
did not provide for use of a helper classification. 54 FR 4242.
---------------------------------------------------------------------------
During implementation of the suspended regulations, Wage and Hour
collected data and determined whether helpers prevailed in various
areas, in accord with the Court's ruling and the requirements of the
now-suspended rule. Thus, implementation of the suspended regulations,
albeit brief, did provide some data and insight into whether the use of
helpers is, in fact, widespread in the construction industry.
The data Wage and Hour received in implementing the regulations
failed to substantiate the prior assumption that the use of helpers is
widespread.
[[Page 17446]]
Whether analyzed by individual classifications covered or by surveys
completed (each of which would include various classifications), the
survey data showed a substantially lower rate of helper use than was
anticipated. For example, a review of the wage schedules issued based
on the 78 prevailing wage surveys completed during the period the rule
was in effect,<SUP>5</SUP> revealed that the use of helpers prevailed
with respect to only 69, or 3.9 percent, of the 1763 classifications
included in wage schedules. Of the 69 classifications in which helpers
prevailed, only 48, or 2.7 percent of the 1763 classifications, were in
the non-union sector.<SUP>6</SUP> This is particularly noteworthy
because it had been assumed in the past that helpers would almost
always be found to prevail for classifications in the non-union sector.
---------------------------------------------------------------------------
\5\ Not included in the 69 helper classifications are instances
where the number of helpers actually used or the number of
contractors using helpers was not enough to provide an adequate
basis for determining a prevailing wage rate. (Wage and Hour
procedures at the time these surveys were conducted required that
there be at least 6 workers employed by at least 3 employers if the
contractor-response rate to the survey was less than 50 percent, and
at least 3 workers employed by at least 2 employers if the response
rate was 50 percent or more.)
\6\ Fifteen of the 21 union-sector helpers classifications were
elevator constructor helpers--a classification historically
recognized nationwide in the union sector of the constructor trade.
---------------------------------------------------------------------------
Furthermore, use of helpers was not prevailing in any
classifications in 43 of the 78 surveys conducted, covering 229 of 328
counties surveyed. The 78 surveys included two in which the resulting
wage schedules contained only collectively bargained rates, ten surveys
in which the schedules contained only open shop rates, and 66 mixed
schedules, 51 of which contained 50 percent or more open shop rates. In
13 of the 35 surveys where a helper classification was issued, the only
helper classification found to prevail was a union helper. A total of
only 48 open shop helper classifications were found to prevail. Thus,
in only 20 of the 78 surveys conducted, covering only 52 of 328
counties surveyed, were any open shop helper classifications found to
prevail. See 61 FR 68644-68645.
The conclusion that helpers are less widespread than had been
expected is also supported by the Economic Impact and Flexibility
Analysis. The 1996 Current Population Survey (CPS), compiled and
published by the Bureau of Labor Statistics (BLS) and the Bureau of the
Census, which Wage and Hour believes is most likely to be
representative of the distribution of employment of helpers in the
construction industry, shows that helpers account for only 1.2 percent
of total construction industry employment. Data from the Occupational
Employment Statistics (``OES'') program, which formed the basis for
earlier analyses of helper employment, shows that helpers comprise 8.7
percent of the total construction workforce--higher than the CPS data
but a much lower incidence than the Department's economic impact
analysis in 1987 and 1989 would suggest. Furthermore, as is discussed
more fully in the Economic Impact Analysis, infra, the OES figure is
based on a helper definition that appears to correspond to what is
commonly considered to be laborers' or tenders' work and does not
appear to envision that helpers use tools of the trade--an important
component of the definition in the suspended regulation. For this
reason Wage and Hour believes that the OES figure significantly
overstates the use of helpers in the construction industry.<SUP>7</SUP>
---------------------------------------------------------------------------
\7\ As discussed in the Impact Analysis, there are strengths and
weaknesses to both the CPS and the OES data sources. For example,
CPS is a household survey and it may be that a carpenter's helper
would self-report his or her duties and occupation as a carpenter.
The Impact Analysis also contains an alternative estimate of the
number of helpers, utilizing the percentage of laborers in the CPS
workforce to adjust the OES data. Under that methodology, described
further in the Impact Analysis, helpers constitute 3.4% of the total
construction workforce.
---------------------------------------------------------------------------
3. The Suspended Regulation Could Have a Negative Impact on Formal
Apprenticeship and Training Programs
Wage and Hour has long been of the view that formal structured
training programs are more effective than informal on-the-job training
alone. Workers enrolled in formal apprenticeship training programs are
more likely to achieve journeylevel status, and to do so more quickly,
than workers trained informally, who may become stuck in low-paying
jobs. Apprenticeship programs are also more likely to produce better
skilled, more productive and safety-conscious workers.<SUP>8</SUP>
---------------------------------------------------------------------------
\8\ Indicative of the lesser efficacy of informal training is
the report issued by the Business Roundtable, which found that more
than 60 percent of its member respondents said they could not find
adequate numbers of skilled workers, and 75 percent said the trend
had accelerated in the past ten years. 203 Daily Labor Report (DLR)
A-9 (Oct. 21, 1997). The report associated the problem with the
``lack of a unified approach to training nonunion trades workers,''
which surfaced 14 years ago, and ``the lack of a consistent delivery
method and commitment to training by other than a small minority of
major contractors.'' Significantly, the Bureau of Apprenticeship and
Training reports almost three times as many union as non-union
apprentices (77,163 union apprentices, compared to 28,542 non-union
apprentices, out of data reported for 36 states (14 states and the
of Columbia do not maintain data byP union affiliation)).
---------------------------------------------------------------------------
Although not its primary concern in this rulemaking, Wage and Hour
is concerned about the potential impact of the suspended regulations on
formal apprenticeship and training programs. An acknowledged goal of
Wage and Hour when it proposed the suspended helpers definition was to
encourage training for unskilled and semi-skilled workers, including in
particular, women and minorities,<SUP>9</SUP> (47 FR 23647 (May 28,
1982)) and to that end Wage and Hour encourages formal training and
work advancement to assure that workers--particularly young, minority,
and female workers--are not frozen into low paying, low skilled jobs.
Because the Department's experience suggests that some contractors may
establish apprenticeship programs to take advantage of the lower wages
which can be paid apprentices and trainees on Davis-Bacon
projects,<SUP>10</SUP> Wage and Hour believes that the suspended helper
regulations could undermine effective training in the industry if
contractors use helpers, who may never become journeylevel workers, in
lieu of apprentices and trainees participating in formal programs.
---------------------------------------------------------------------------
\9\ Wage and Hour has no data to support or refute the
proposition that employment of helpers leads to an increase in
minority and female skilled employment in the non-union sector.
\10\ Effective training for targeted under-represented or
economically disadvantaged workers who are not qualified for
apprenticeship programs can be designed under the existing
regulations. For example, the Step-Up Program developed by the
Department of Housing and Urban Development (HUD) provides
disadvantaged workers with training necessary for them to move on to
other more skilled jobs or into a formal apprenticeship program.
---------------------------------------------------------------------------
The Proposed Rule--Helpers as a Separate and Distinct Class with
Clearly Defined Duties Which Do Not Overlap With Laborer or Journeyman
Classifications
Wage and Hour proposes to amend the regulations to reflect the
longstanding policy of recognizing helpers as a distinct classification
on DBRA-covered work only where Wage and Hour determines that (1) the
duties of the helpers are not performed by other classifications in a
given area, i.e., the duties of the helper are clearly defined and
distinct from those of the journeyworker and laborer; <SUP>11</SUP> (2)
the use of such helpers is an established prevailing practice in the
area; and (3) the term ``helper'' is not synonymous
[[Page 17447]]
with ``trainee'' in an informal training program.<SUP>12</SUP>
---------------------------------------------------------------------------
\11\ For example, roofing subcontractors, like other specialty
subcontractors, often do not hire laborers, and might employ helpers
to perform duties such as bringing materials to the roof and
removing the old roof.
\12\ Where Wage and Hour has determined that this standard is
met, the helper classification will be listed on the wage
determination. Where no helper is listed on the wage determination,
a contractor who believes that use of a helper classification
meeting the criteria is prevailing in the locality may request an
additional classification in accordance with 29 CFR 5.5(a)(1)(ii).
Like other classifications, the particular duties such a helper may
perform are determined by area practice.
---------------------------------------------------------------------------
This approach retains the duties-based classification distinction
that provides an objective basis for administration and enforcement. It
provides clear criteria to facilitate compliance. It is also consistent
with the intent of the Davis-Bacon Act to assure that workers employed
on federal and federally-assisted construction work be paid at least
the wages paid to workers doing similar work on similar construction in
the area. Lack of overlapping duties should also discourage contractor
misclassification and/or abuse. This approach also encourages
contractors to establish or participate in structured training programs
leading to journeylevel status if they want to pay subminimum rates to
entry-level or less skilled workers.
Unlike some of the other alternatives considered, this policy
concerning helpers does not require Wage and Hour to make a fact-bound
inquiry in each case to assess a worker's skill level and the nature of
work-site supervision to determine whether the worker will be
recognized as a ``helper'' for Davis-Bacon prevailing wage compliance
and enforcement purposes. The requirement that helpers be separate and
distinct from journeylevel workers and laborers should also facilitate
collection of wage data to establish the prevailing wage rates to be
paid on DBRA-covered construction work.
Although this proposal could be said to disregard local area
practices in those instances where there may be a prevailing practice
of employing ``helpers'' who do not meet the regulatory test set forth
above, it appears that there is wide variation in how helpers are used,
such that change in practices by contractors would be likely under any
definition. Wage and Hour has been unable to identify a generally
accepted definition of helper that corresponds to industry practices.
Similarly, Wage and Hour has been unable to find a practical method of
determining prevailing practice regarding how helpers are in fact
utilized in an area.
Discussion of Other Alternatives Considered
1. Add a Ratio Requirement to the Suspended Helper Definition
Wage and Hour recognized that the broad scope of the helper rule's
definition created the potential for abuse when it originally proposed
to amend the regulations to allow the expanded use of helpers. The rule
as proposed in 1981, as well as subsequent modifications, sought ``to
protect against possible abuse'' by establishing a maximum ratio of
helpers to journeyworkers. In 1992, the Court of Appeals ruling
nullified the ratio of two helpers to every three journeyworkers
because that specific numeric ratio had not been justified in the
rulemaking record. As noted in the foregoing discussion, the inherent
definitional problems regarding the suspended ``helper'' rule were
compounded by elimination of the ratio provision, which was intended to
ameliorate the possible overuse of helpers.
Since the Court of Appeals ruling does not prevent Wage and Hour
from implementing a ratio, provided it has support in the rulemaking
record, implementation of a new ratio was the first alternative
considered. Implementation of a ratio provision would be essential if
the suspended rule were implemented, since it would reduce the
potential for abuse. However, adoption of such a provision would not
address or resolve the inherent definitional problems discussed above,
which make it extremely difficult under the suspended rule for
contractors, as well as Wage and Hour and contracting agencies, to
identify helpers for Davis-Bacon enforcement and wage determination
purposes.
Furthermore, determination of an appropriate ratio standard --
either a single nationwide ratio or local ratios--would be difficult.
While a nationwide ratio would not accord with local practices, local
ratios would present significant administrative and enforcement
concerns, and would require substantial resources for implementation.
2. Change the ``Helper'' Definition To Emphasize the Semi-Skilled
Nature of the Classification
The intention of Wage and Hour in promulgating the suspended rule
was to allow the expanded employment on Davis-Bacon covered projects of
helpers who are ``semi-skilled,'' in other words, they perform some
journeylevel duties, but not the entire range of journeylevel work.
This attempt to define helpers as similar to but less skilled than a
journeyworker resulted in a helper definition that is internally
inconsistent, since the specific tasks listed as within the scope of a
helper's duties are commonly performed by unskilled workers. Wage and
Hour therefore considered possible modifications to the helper
definition to emphasize the semi-skilled nature of helpers, elaborate
on the supervisory relationship of the journeyworkers with the helper
and the craft-specific assistance provided, and expressly limit the
unskilled work the helper may perform.
This approach to the definition would help assure that the
``helper'' classification would be a true ``semi-skilled''
classification rather than a broad catch-all classification that can
perform everything from laborer duties to an undefined assortment of
skilled tasks overlapping the work of the journeyworkers. Such a
definition would therefore aid in distinguishing helpers from laborers.
However, this alternative would not resolve the administrative and
enforcement problems that stem from the overlap of duties between
journeyworkers and helpers. Furthermore, Wage and Hour is concerned
that this type of definition, with its emphasis on semi-skilled duties,
may result in helper classifications being used to replace, rather than
supplement, the use of apprentices and trainees registered in bona fide
training programs.
3. Define ``Helpers'' Based on the Bureau of Labor Statistics,
Occupational Employment Statistics (OES) Dictionary of Occupations,
Which Focuses on Unskilled Duties and the Worker's Interaction With
Journeylevel Craft Workers
The Bureau of Labor Statistics Occupational Employment Statistics
(OES) Dictionary of Occupations classification scheme includes a broad
category titled ``Helpers, Laborers, and Material Movers, Hand,
Exclud[ing] Agriculture and Forestry Laborers.'' The work of helpers so
defined in the construction industry is currently described generally
as follows:
Help workers in the construction trades, such as Bricklayers,
Carpenters, Electricians, Painters, Plumbers and Surveyors. Perform
duties such as furnishing tools, materials and supplies to other
workers; cleaning work areas, machines, and tools; and holding
materials or tools for other workers.
Use of this approach would provide for definitional consistency
with other uses of the OES data and would take advantage of a standard
definition that could be easily followed and understood by contractors
from whom data is collected for various purposes,
[[Page 17448]]
including Davis-Bacon prevailing wage surveys. The OES definitions
would focus on the role of the helper in assisting the journeyworker,
in accord with the Department's intention that such a role be a key
component of any definition selected.
These definitions, which would eliminate the ``semi-skilled''
characterization from the definition and highlight unskilled duties,
could provide a more practical basis for distinguishing helpers from
journeyworkers. On the other hand, laborers may often perform the same
work encompassed within the OES helper definition, thereby raising
significant problems in conducting wage and area practice surveys and
in enforcement. It may be difficult for contractors to determine
whether workers performing similar or identical duties are ``laborers''
or ``helpers'' when submitting Davis-Bacon survey data and in
classifying workers on Davis-Bacon projects. In turn, Wage and Hour
believes it would likely be difficult for it to determine whether
contractors have properly classified workers paid as helpers as
distinguished from laborers on Davis-Bacon projects, and therefore
whether contractors have submitted accurate wage data in regard to
helpers.
4. Explicitly Delineate the Semi-Skilled Tasks Performed by Each Helper
Classification
The ``job family'' concept is currently employed for certain
occupations under the McNamara-O'Hara Service Contract Act. An employee
who performs only lower level duties that are associated with a
particular job family may be classified and paid at the lower level
helper rate; however, an employee who performs some lower level duties
and some higher level duties must be paid the higher level journeylevel
rate for all of the employee's work time.
In effect, this approach would allow for the expanded use of
helpers, with differentiation based on the skill and knowledge required
to perform particular duties. Once the duties or tasks that the helpers
could perform were clearly defined, wage data could be collected on
that basis, and contractors could reasonably be expected to comply with
the wage requirements for the various classifications employed on their
contracts, thereby facilitating administration and enforcement.
However, developing clear definitions of the duties or tasks that
helpers to each journeylevel craft worker would be allowed to perform
would be very difficult. It would require extensive occupational
analyses and further rulemaking to promulgate helpers duties
descriptions. Furthermore, this alternative--like other alternatives
considered--presumably would result in uniform, nationwide definitions,
departing from the principle that classifications are determined based
on local area practice.
IV. Executive Order 12866; Sec. 202 of the Unfunded Mandates Reform
Act of 1995; Small Business Regulatory Enforcement Fairness Act
Wage and Hour has determined that this proposed rule should be
treated as ``economically significant'' within the meaning of Executive
Order 12866 and as a major rule within the meaning of the Small
Business Regulatory Enforcement Fairness Act. This proposed rule would
continue the status quo which has been in effect since November 1993,
and therefore it would have no economic impact compared to current
practices. However, various alternatives considered would result in
potential savings which could be in excess of $100 million per year.
Therefore a full economic impact analysis has been prepared.
However, for purposes of the Unfunded Mandates Reform Act of 1995,
this rule does not include any federal mandate that may result in
increased annual expenditures in excess of $100 million by state, local
and tribal governments in the aggregate, or by the private sector. The
requirements of the Unfunded Mandates Reform Act, 2 U.S.C. 1532, do not
apply here because the proposed rule does not include a ``Federal
mandate.'' The term ``Federal mandate'' is defined to include either a
``Federal intergovernmental mandate'' or a ``Federal private sector
mandate.'' 2 U.S.C. 658(6). Except in limited circumstances not
applicable here, those terms do not include an enforceable duty which
is ``a condition of Federal assistance'' or ``a duty arising from
participation in a voluntary program.'' 2 U.S.C. 658(5)(A)(I) and
(7)(A). A decision by contractors to bid on Federal or Federally-
assisted construction contracts is purely voluntary in nature, and
their duty to meet Davis-Bacon requirements are ``conditions of Federal
assistance'' which arise ``from participation in a voluntary Federal
program.''
Similarly, the proposed rule is not an ``unfunded mandate'' within
the meaning of Executive Order 12875 since it does not create any
unfunded mandate not currently required by the Davis-Bacon and Related
Acts and regulations thereunder. Furthermore, most of the funds
necessary to pay the direct costs incurred by State, local and tribal
governments under projects subject to the Davis-Bacon and related Acts
are provided by the Federal Government.<SUP>13</SUP> Thus, any
additional savings to States if the proposed rule increased use of
helpers allowed on Davis-Bacon projects would not be significant.
---------------------------------------------------------------------------
\13\ It is significant that no such entities commented on the
proposed rule published in August 1996.
---------------------------------------------------------------------------
V. Economic Impact and Flexibility Analysis on Davis-Bacon Helper
Regulations
Summary
This document presents an Economic Impact Analysis comparing the
proposed rule governing the use of helpers under the Davis-Bacon and
Related Acts to the suspended rule. The basic process utilized to
estimate the potential savings impact of the suspended regulation is to
compare the occupational distribution of workers with, and without
helpers. The alternative occupational employment patterns are then
assessed in terms of their costs, based upon the annual earnings of the
workers in the occupations affected by the suspended regulation:
journeyworkers, apprentices, laborers, and helpers. The total wage bill
with the suspended regulation in force is then subtracted from the wage
bill estimated without the regulation. The difference, then, is the
estimated savings.
The principal finding of the analysis is that any impact which
would result from the increased use of helpers under the suspended
rule, or any of the other alternatives considered, would be relatively
modest. Potential savings are estimated to be from $72.8 million
(utilizing Current Population Survey--CPS data) to $296.0 million
(utilizing Occupational Employment Statistics--OES data). A methodology
that is OES-based, but utilizes CPS data to estimate the number of
laborers and helpers in the OES, provides an estimate of $108.6 million
in possible savings. This alternative OES estimate was developed to
compensate for the likelihood that OES data overestimate the number of
helpers.<SUP>14</SUP> In any case, for reasons discussed below, Wage
and Hour believes that the potential savings are likely to be closer to
$72.8 million than to $296.0 million.
---------------------------------------------------------------------------
\14\ As explained in detail below, OES has no distinct
classification for laborer. This characteristic of the OES program,
in combination with the helper OES definition that includes workers
who would normally be classified as construction laborers, inflates
the OES helper total.
---------------------------------------------------------------------------
Relative to total construction expenditures covered by the Davis-
[[Page 17449]]
Bacon and Related Acts, these potential cost savings are very small,
ranging from 0.2 percent to 1.0 percent. As discussed below, the
estimated savings are far less than previously believed. For the most
part, changes in the savings potential resulted from the use of
improved data, including information derived from experience
administering the suspended regulations, which temporarily expanded the
use of helpers.
A. Introduction
Over the years, Wage and Hour has prepared and updated regulatory
impact and flexibility analyses in connection with proposed and final
regulations governing the use of semi-skilled helpers under the Davis-
Bacon and Related Acts. Specifically, cost savings derived from the
increased use of helpers were estimated in the August 14, 1981 proposed
rule (46 FR 41456); the May 28, 1982 final rule (47 FR 23644); the
August 19, 1987 proposed rule (52 FR 31366); and the January 27, 1989
final rule (54 FR 4234). Wage and Hour is now updating its cost
estimates in connection with the proposed rule being published today,
as set forth above.
This latest economic impact analysis has the advantage of utilizing
information not previously available. For example, for the first time,
survey data are available from a limited period when the regulations
expanding the use of helpers were actually being implemented. Other
data sources, utilized for the first time in such an analysis, include:
<bullet> Estimates of apprentice employment, based upon information
provided by the Bureau of Apprenticeship and Training (BAT) from its
Apprentice Information System (AIMS).
<bullet> F.W. Dodge construction reports.
<bullet> Detailed published occupational information and
unpublished Bureau of Labor Statistics (BLS) tabulations from the
Current Population Survey (CPS).
<bullet> National Occupational Employment Statistics (OES) Program
data.
B. Assumptions and Data Sources
1. Assumptions
a. There is a strong positive correlation between the value of
construction and the level of construction employment. This assumption
is derived from the fact that labor costs generally are considered to
constitute a significant proportion of total construction expenditures.
b. Under the suspended rule, helpers would replace laborers,
apprentices, and journeyworkers in proportion to the number of workers
in each of these occupations. The previous helper impact analysis
assumed that helpers would only replace journeyworkers, and measured
only the wage differentials from this replacement effect. This
exaggerated the estimates of possible cost savings from the expanded
use of helpers. Since wage rates generally reflect skill levels, the
relative closeness of average annual earnings for helpers, laborers,
and apprentices, compared to journeyworkers, strongly suggests that
this assumption was incorrect. These wage data suggested that helpers
(at $9,008 per year) are more likely to assume the duties of laborers
(at $15,907 per year) and apprentices (at $12,564 per year) than
journeyworkers (at $23,007 per year).<SUP>15</SUP> In fact, had the
redistribution of employment been strictly in accordance with
occupational wages, savings estimates would have been reduced
significantly (see Estimating Process, Step 2).
---------------------------------------------------------------------------
\15\ Source: 1996 BLS/CPS.
---------------------------------------------------------------------------
The assumption that helpers would perform tasks previously
performed by laborers and apprentices, as well as journeyworkers, is
also based upon comments made by general contractors surveyed during
the processing of helper conformance requests during the period
February 1992 to October 1993. These comments indicated that the job
title ``laborer'' was often applied to those performing the work of a
``helper'' (as defined in the suspended regulations). In order to take
the middle ground for this analysis, it is assumed that when a helper
classification is added, the jobs which would be performed by helpers
were previously those of laborers, apprentices, and journeyworkers, in
the same proportion as their relative occupational employment.
c. Utilizing the decision rules specified in Section 1.7(d), 29 CFR
of the suspended regulations (see Section C, Part 2, Estimating
Process, Step 3, below), helpers would be likely to ``prevail'' for a
limited number of classes in areas that represent about half the
construction employment covered by the Davis-Bacon and Related Acts.
This estimate is based on the findings of prevailing wage surveys
conducted during the period from February 1992 to October 1993. This is
generally consistent with the small number of helpers relative to total
construction employment found in the CPS, OES, and adjusted OES
databases, only 1.3 percent, 8.7 percent, and 3.4 percent of
construction employment, respectively.<SUP>16</SUP>
---------------------------------------------------------------------------
\16\ Based upon the results of the methodology utilized, if the
suspended regulations were in effect, the proportion of helpers to
total employment would increase from 1.3 to 1.4 percent (CPS), 8.7
to 9.2 percent (OES) and 3.4 to 3.5 percent (Adjusted OES, hereafter
``AdjOES'').
---------------------------------------------------------------------------
d. The proportion of employment by occupation would be consistent
in all areas, and therefore the average national proportion of helpers,
apprentices, laborers, and journeyworkers would be the same in areas
where helpers prevail and where they do not. One could, of course,
contend that a proportion higher than the national average should be
used for helpers in the half of Davis-Bacon construction in which it is
assumed that some helpers would prevail. However, some helpers would
also be employed in the much larger group of classifications in which
helpers would not be determined to prevail. Furthermore, an analysis of
helper employment from Davis-Bacon surveys during the period when the
suspended Regulation was in effect, found that in areas where helpers
prevailed for one or more classifications, versus those where no
helpers prevailed, the percent helpers were of total employment was
almost identical (1.8 percent vs. 1.7 percent). Therefore, it is
reasonable to assume that, on average, the level of helpers employed in
areas where helpers prevail would be consistent with the level of
helper employment overall.
e. Approximately one-third of public, non-Federal construction
projects receive Federal assistance. This estimate is based upon the
extensive experience of the Office of Federal Contract Compliance
Programs (OFCCP) with F.W. Dodge data (which classifies public projects
into Federal and public non-Federal classifications) to select
construction sites for compliance inspections (only Federal and
Federally-assisted projects are inspected by OFCCP). However, since not
all types of Federal assistance trigger Davis-Bacon and Related Acts
coverage, recent prevailing wage surveys were used to determine the
average proportion of public, non-Federal construction covered by the
Davis-Bacon and Related Acts. Based upon a study of 34 prevailing wage
surveys, approximately 23 percent of the value of public non-Federal
construction is covered by Davis-Bacon.
f. Except for the specific requirements of Davis-Bacon, such as
those concerning helpers, primary characteristics of the labor force,
i.e., occupational distribution, work assignments, etc. under Davis-
Bacon are
[[Page 17450]]
comparable to those of the labor force not covered by prevailing wages.
2. Data Sources
Databases from which estimates were developed include:
<bullet> The BAT AIMS Reporting System (number of apprentices).
<bullet> The BLS/Bureau of the Census CPS (total construction
industry employment, distribution of employment by selected occupation,
and total annual earnings by occupation).
<bullet> The BLS OES Program (total construction industry
employment and distribution of employment for selected occupational
combinations).
<bullet> F.W. Dodge Construction Reports (construction value by
ownership).
<bullet> Wage and Hour Division Regional Survey Planning Reports
(RSPR) (public construction value by wage determinations reflecting
union, open shop, and mixed wage rates).
<bullet> Information gained through conduct of Wage and Hour
Division wage surveys.
There are significant differences in the CPS and OES data, some of
which are due to the way the data are collected. The CPS is a household
survey and relies on information provided by residents, whereas the OES
is an establishment survey, with data usually provided by employers'
personnel offices. The most apparent difference is in the total number
of construction workers. In the CPS survey, the total number is much
higher than in the OES, in part because the OES does not count the
self-employed. (See tables in Section C.1.b., below.)
Although they constitute the best available data on occupational
employment and wages in the construction industry, neither the CPS nor
the OES is ideal for the purpose of this analysis. In fact, there are a
number of differences between the two surveys that are of particular
importance to this analysis. Specifically, each has strengths and
weaknesses that impact the helper savings derived from database use.
For the purpose of estimating the impact of the proposed helper
regulation, the Current Population Survey has the following strengths:
<bullet> The CPS survey includes those workers not covered by State
unemployment insurance--primarily self-employed workers. This latter
group is particularly important since the construction industry
includes a significant number of workers (e.g., painters, carpenters,
and plumbers) who are independent contractors, and therefore self-
employed. Davis-Bacon prevailing wage requirements extend to every
laborer and mechanic working on a covered project, regardless of
contractual relationship, including the self-employed (independent
contractors). Thus, the CPS number of construction workers,
particularly the skilled workers who are more likely to be self-
employed, reflects the universe of construction employment that is
relevant to this analysis.
<bullet> The CPS provides separate employment totals for helper,
apprentice, laborer, and journeyworker classes, all of which are needed
to conduct this impact analysis.
<bullet> The CPS provides annual average earnings for the above
classes. These data are also essential to estimating the impact of
implementing the suspended rule.
For purposes of this analysis, the CPS survey program also has the
following weaknesses:
<bullet> CPS is a household survey, rather than an establishment
survey. In general, household surveys are likely to produce less
accurate and consistent wage and classification information than
establishment surveys. Self-reporting can result in some workers
exaggerating their level of responsibility or wages. For example, a
carpenter's helper may self-report his or her duties and occupation as
carpenter.<SUP>17</SUP>
---------------------------------------------------------------------------
\17\ A 1973 comparison of CPS earnings reported by those
surveyed versus corresponding IRS records indicated that
exaggeration is minimal. See Herriot, Roger A., and Spiers, Emmet
F., ``Measuring the Impact on Income Statistics of Reporting
Differences between the Current Population Survey and Administrative
Sources,'' Unpublished, 1973.
---------------------------------------------------------------------------
<bullet> The CPS data on annual earnings include wages earned
outside construction, although construction is the industry of longest
employment for each worker.
<bullet> Apprentice data, other than four separately identified
classes, are combined with data for the associated journeyworkers. This
has the effect of inflating journeyworker employment totals and
lowering journeyworker wages.
<bullet> CPS responses can be provided by the worker's spouse or
adult child if the worker is unavailable.
For the purpose of estimating potential savings, the OES Program
exhibits three particular strengths:
<bullet> The OES Program utilizes standard occupational
definitions, describing those workers who should be reported in each.
<bullet> Establishment (i.e., employer) personnel staff usually
provides the survey data requested. This, together with the standard
definitions, is likely to result in more accurate and consistent
assignment of occupational classes, and more accurate wage reporting
than that characteristic of surveys with self-reporting of workers,
such as the CPS.
<bullet> The OES sample size (1.2 million employers) is larger than
the CPS sample size (50,000), with one-third of the 1.2 million
establishments surveyed each year. This large sample increases the
number of participating establishments and reduces sampling error.
Weaknesses of the OES survey program, for purposes of this
analysis, include:
<bullet> The OES does not provide a specific employment total for
``laborer.'' Instead, laborers appear to be combined with craft
helpers, as well as in an OES occupational category titled ``All Other
Helpers, Laborers, and Material Movers, Hand.'' Since it is assumed
that some helpers would replace laborers under the suspended
regulation, separate laborers and helpers totals are required for
development of an accurate savings estimate.
<bullet> The OES definitions of the various helper classifications
are very similar to unskilled laborers who provide assistance to
journeyworkers. (Helpers ``perform duties such as furnishing tools,
materials and supplies to other workers; cleaning work areas, machines,
and tools; and holding materials or tools for other workers.'') Thus,
the OES craft helper may often be an unskilled worker (and thus a
laborer) rather than the semi-skilled worker required in the suspended
regulation. As a result, laborer employment in the OES likely is, to a
great extent, included with helper employment, thereby overstating the
number of helpers.
<bullet> The OES survey collects only hourly wage data and does not
collect annual hours worked data. At the same time, OES counts jobs
rather than employees. As a result, if one person holds a job at more
than one establishment, each one of those jobs will be counted,
providing a total that exceeds the number of employees. Since labor
costs are computed by multiplying the number of employees times annual
CPS wages, the OES jobs count acts to overestimate costs.
<bullet> The OES excludes those who are self-employed (independent
contractors) and those not covered by State unemployment insurance,
thus significantly understating the total number of construction
workers and the number of construction workers in the Davis-Bacon
workforce.
<bullet> All apprentice data are combined with the journeyworker
data, thus overstating the number of journeyworkers.
[[Page 17451]]
Based principally on the fact that at this time the OES does not
have a separate classification for laborer, together with the fact that
OES does not collect data on self-employed individuals, Wage and Hour
believes that the CPS data are more likely than the OES data to be
representative of the distribution of employment in construction by
occupation for helpers and laborers. However, given that neither
database is ideal for this purpose, and the fact that OES data are also
relevant, both CPS and OES will be used to develop a range of possible
savings estimates.
3. Measuring Helpers and Laborers
The major difficulty in developing an impact analysis to estimate
potential savings from the expanded use of Davis-Bacon helpers is the
dearth of data that reasonably represent the employment of helpers as
defined by the suspended regulations, and of laborers. As noted in the
Data Sources section, above, the Current Population Survey (CPS) does
have separate categories for helper and laborer. However, the survey
does not contain standard occupational definitions. Therefore, there
can be no assurance that the number of those reported as helpers truly
corresponds to the definition in the suspended regulation. For example,
it is believed that some helpers--defined by the regulation as semi-
skilled workers who may use tools of the trade--may actually be
reported in the CPS as journeyworkers. On the other hand, many laborers
may be reported as helpers.
Also, as noted above, although the Occupational Employment
Statistics (OES) program includes the use of standard occupational
definitions, it has no distinct category for laborers and the helper
definitions used in the survey are quite different than the definitions
in the suspended regulations. In fact, the OES helper definition likely
includes many laborers who primarily work with or assist
journeyworkers. Also, many of those reported as helpers under OES may
not be semi-skilled at all, but unskilled workers who perform ``duties
of lesser skill'' and do not have the knowledge and abilities necessary
to use tools of the trade. Therefore, Wage and Hour believes that the
OES employment totals for helpers likely include many laborers and
unskilled helpers.
Since the OES database has no distinct class for laborer, the
methodology to estimate potential savings using OES data requires
development of a methodology for separating laborers from helpers.
Therefore, OES classes were identified that by their terms appeared to
primarily include laborers. The classes selected for that purpose
included Helpers, Mechanic and Repairer; Helpers, Extractive Workers;
Freight, Stock, and Material Movers, Hand; Vehicle Washers and
Equipment Cleaners; and Other Helpers, Laborers, & Material Movers,
Hand. On the one hand, given these job titles, some workers other than
laborers would be included in these totals. On the other hand, the
total number of laborers derived from this process (262,310) is well
below what would be expected, leading one to believe that many laborers
are included in the OES craft helper employment totals.
Corroborating evidence that this approach to the OES data without
further adjustment overestimates the number of helpers, underestimates
the number of laborers, and therefore overestimates potential savings,
when utilized in a helpers impact analysis, may be found in both
Decennial Census and CPS data. The Decennial Census estimates 949,000
construction laborers (a ratio of 1 laborer for every 5 journeymen),
the CPS estimates 988,000 (1 laborer for every 4 journeymen), and OES
estimates just 262,310 (1 laborer for every 10
journeymen).<SUP>18</SUP>
---------------------------------------------------------------------------
\18\ Since the OES universe is 60.7 percent of the CPS universe,
one would expect the OES laborers total to be about 600,000.
---------------------------------------------------------------------------
One way to compensate for this likely undercount of laborers and
overestimate of helpers is to determine what percent laborers
constitute in the CPS universe, and apply that to the OES data. The
laborer category is chosen for this purpose because it is the least
likely to suffer from error due to the reporting workers exaggerating
their duties. In the CPS, the laborers constitute 18.8 percent of the
total for journeymen, apprentices, laborers, and helpers. Multiplying
that percent times the comparable OES total provides an adjusted number
of OES laborers. Subtracting the adjusted laborer total from the
laborer-helper combination (called helper) in OES yields an adjusted
number for helpers. While this figure (and therefore the potential
savings estimate) is probably an improvement over the unadjusted OES
helper total, one cannot be certain of problems that may have
inappropriately affected the resulting estimates, since two dissimilar
databases have been combined.<SUP>19</SUP>
---------------------------------------------------------------------------
\19\ As discussed below, it is also necessary to utilize CPS
wage data, thereby combining dissimilar data bases with attendant
problems.
---------------------------------------------------------------------------
In light of these problems, it is advised that the estimates
included in this impact analysis be considered with caution. All the
figures provided should be treated as very rough measures that provide
a general range within which possible savings could fall.
C. Key Data Elements, Estimating Process and Computations
1. Key Data Elements
a. Value of total construction starts, 1996:
Total: $321,736,705,000
Federally owned: $10,799,923,000
Public-Non-Fed: $87,122,347,000
b. Construction industry employment, and average annual earnings,
1996:
Table 1.--Construction Industry Employment and Average Total Earnings, Total and Selected Occupations, CPS
Database, 1996
----------------------------------------------------------------------------------------------------------------
Average total
Data source and occupation Total Percent of annual
employment * total earnings **
----------------------------------------------------------------------------------------------------------------
CPS.......................................................... 9,333,000 100.000 N.A.
Construction Trades, Except Supervisors and Apprentices ***.. 3,958,000 42.409 $23,007
Apprentices.................................................. 192,000 2.057 12,564
Helpers...................................................... 124,000 1.329 9,008
Laborers..................................................... 988,000 10.586 15,907
Other Occupations ****....................................... 4,071,000 43.619 N.A.
----------------------------------------------------------------------------------------------------------------
* CPS data include the incorporated self-employed.
** Total average annual earnings data are for workers who reported their longest job during the year to be in
the construction industry. The data are for 1996. Compensation for non-construction work by these workers is
included.
[[Page 17452]]
*** The CPS figure for four classes of apprentices is 48,000, while the BAT/AIMS total for all occupations is
192,000. For this purpose, the BAT apprentice figure was utilized, with the 144,000 ``additional'' apprentices
subtracted from the CPS construction trades total, based on the assumption that a number of apprentices self-
reported their occupation to be journeyworkers. AIMS data are generated as part of the national apprenticeship
program and represent active apprentices at the end of the year. Since several states do not report these
data, BAT staff estimated the U.S. total based upon the percent of construction employment represented by the
missing States.
**** Other occupations include Executive, Administrative, and Managerial positions; Technical, Sales, and
Administrative Support; etc., and others, such as those in Service occupations.
Table 2.--Construction Industry Employment and Average Total Earnings, Total and Selected Occupations, OES
Database, 1996
----------------------------------------------------------------------------------------------------------------
Total Percent of
Data source and occupation employment * total Average total annual earnings **
----------------------------------------------------------------------------------------------------------------
OES......................................... 5,666,150 100.000 N.A.
Construction Trades Except Supervisors and 2,064,900 36.443 N.A.
Apprentices ***.
Apprentices................................. 192,000 3.389 N.A.
Helpers ****................................ 495,600 8.747 N.A.
Laborers *****.............................. 262,310 4.629 N.A.
Other Occupations........................... 2,651,340 46.793 N.A.
----------------------------------------------------------------------------------------------------------------
* Excludes self-employed and those not covered by UI.
** Data on wages provided in hourly rates only.
*** Since all apprentices are combined with OES journeyworkers, the BAT apprentice total of 192,000 was
subtracted from the OES journeyworker total.
**** Likely includes significant numbers of unskilled helpers and laborers who primarily work with or assist
journeyworkers.
***** Figure taken from a catchall classification that includes ``All Other Helpers, Laborers, And Material
Movers, Hand,'' plus the OES classifications of Helpers, Mechanic and Repairer; Helpers, Extractive Workers;
Freight, Stock, and Material Movers, Hand; and Vehicle Washers and Equipment Cleaners.
Table 3.--Construction Industry Employment and Average Total Earnings, Total and Selected Occupations, Adjusted
OES Database, 1996
----------------------------------------------------------------------------------------------------------------
Total Percent of
Data source and occupation employment * total Average total annual earnings **
----------------------------------------------------------------------------------------------------------------
Adjusted OES................................ 5,666,150 100.000 N.A.
Construction Trades Except Supervisors and 2,064,900 36.443 N.A.
Apprentices ***.
Apprentices................................. 192,000 3.389 N.A.
Helpers ****................................ 191,126 3.373 N.A.
Laborers.................................... 566,784 10.003 N.A.
Other Occupations........................... 2,651,340 46.793 N.A.
----------------------------------------------------------------------------------------------------------------
* Excludes self-employed and those not covered by UI.
** Hourly rates only.
*** BAT figure subtracted from journeyworker total.
**** The OES reports a laborer/helper combination employment of 757,910 (Helpers, Laborers, & Material Movers,
Hand). To separate laborer from helper, the percent that CPS laborers (988,000) are of the CPS employment sum
(5,262,000) for journeyman, apprentices, laborers, and helpers (18.8 %) was multiplied by the comparable OES
employment total (3,014,810). That product (566,784) then was adopted as the OES laborer total, and subtracted
from the laborer/helper combination to yield the OES helper figure (191,126).
2. Estimating Process and Computations
A 5-step estimating process was developed and utilized to
approximate annual savings that might have been realized in 1996 from
the increased use of helpers, if the suspended regulations had been
implemented:
Step 1: Davis-Bacon Employment. Determine the value of construction
covered by the Davis-Bacon and Related Acts. This is achieved by adding
100 percent of Federal construction starts value to 23 percent of the
value of public, non-Federal construction starts. Divide that sum by
the total value of construction starts to obtain the proportion that
Davis-Bacon covered construction is of total construction value.
Multiply the Davis-Bacon proportion times total construction employment
to estimate the share of total construction employment allocated to
Davis-Bacon construction.
Value of DB Construction = ($10,799,923,000) + (0.23 x
$87,122,347,000) = $30,838,062,810
Proportion DB is of Total = $30,838,062,810/$321,736,705,000 = 9.585%
DB Employment =
CPS: 9,333,000 x 0.09585 = 894,568
OES: 5,666,150 x 0.09585 = 543,100
AdjOES: 5,666,150 x 0.09585 = 543,100
Step 2: Occupational Employment, 1996. First, the number of
additional apprentices, estimated by BAT and above the CPS apprentice's
estimate, was added into the CPS construction apprentices total, and
subtracted from the journeyworkers total. The BAT apprentice total was
similarly subtracted from the OES journeyman/apprentice combination,
and established as the OES apprentice total, both unadjusted and
adjusted.
Then, 1996 Davis-Bacon employment for the number of journeyworkers,
laborers, apprentices, and helpers is obtained. (Note that these on-
site construction workers are the only occupations likely to be
impacted by any helper regulation.) This is accomplished for each
occupational group by multiplying their corresponding adjusted CPS/OES
proportions times total Davis-Bacon construction employment. However,
since procedures in effect in 1996 prohibited the use of helpers on
Davis-Bacon work, the number of helpers computed must be allocated
(added) to the number of Davis-Bacon journeyworkers, laborers, and
apprentices. This allocation is made in
[[Page 17453]]
proportion to each occupational group's composition of covered
employment, in order to obtain final estimates of total Davis-Bacon
employment for the selected occupational groups. (As indicated under
Assumption 2, had this employment been distributed based upon closeness
of occupational wage, helper savings would have been significantly
reduced.)
DB Journeyworker Employment =
CPS: 894,568 x 0.42409 = 379,377
OES: 543,100 x 0.36443 = 197,922
AdjOES: 543,100 x 0.36443 = 197,922
DB Laborer Employment =
CPS: 894,568 x 0.10586 = 94,699
OES: 543,100 x 0.04629 = 25,140
AdjOES: 543,100 x 0.10003 = 54,326
DB Apprentice Employment =
CPS: 894,568 x 0.02057 = 18,401
OES: 543,100 x 0.03389 = 18,406
AdjOES: 543,100 x 0.03389 = 18,406
DB Helper Employment =
CPS: 894,568 x 0.01329 = 11,889
OES: 543,100 x 0.08747 = 47,505
AdjOES: 543,100 x 0.03373 = 18,319
Subtotals:
CPS: 504,366
OES: 288,973
AdjOES: 288,973
Table 4.--Data for ``No Helper'' Helper Adjustment
----------------------------------------------------------------------------------------------------------------
AdjOES
Occupation CPS No. CPS % OES No. OES % No. AdjOES %
----------------------------------------------------------------------------------------------------------------
Journeyworker................................. 379,377 0.77034 197,922 0.81966 197,922 0.73127
Laborer....................................... 94,699 0.19229 25,140 0.10411 54,326 0.20072
Apprentice.................................... 18,401 0.03736 18,406 0.07623 18,406 0.06801
-----------------------------------------------------------------
Total..................................... 492,477 0.99999 241,468 1.00000 270,654 1.00000
----------------------------------------------------------------------------------------------------------------
Helper Adjustment
DB Journeyworkers =
CPS: 379,377 + (11,889 x 0.77034 = 388,536
OES: 197,922 + (47,505 x 0.81966) = 236,860
AdjOES: 197,922 + (18,319 x 0.73127) = 211,318
DB Laborers =
CPS: 94,699 + (11,889 x 0.19229) = 96,985
OES: 25,140 + (47,505 x 0.10411) = 30,086
AdjOES: 54,326 + (18,319 x 0.20072) = 58,003
DB Apprentices =
CPS: 18,401 + (11,889 x 0.03736) = 18,845
OES: 18,406 + (47,505 x 0.07623) = 22,027
AdjOES: 18,406 + (18,319 x 0.06801) = 19,652
Step 3: Occupational Employment (52 FR 31368). Determine Davis-
Bacon employment for the number of journeyworkers, laborers,
apprentices, and helpers likely to be employed if the Regulations
published in 52 FR 31368 were in effect throughout 1996. For the
employment half in which helpers do not prevail for any classes, Step 2
proportions are utilized; for the half in which helpers do prevail for
a limited number of classes, proportions reflect average national
employment of helpers.
Employment of DB Journeyworkers (CPS: 0.77034; OES: 0.81996; AdjOES:
.73127) + Laborers (CPS: 0.19229; OES: 0.10411; AdjOES: .20072) +
Apprentices (CPS: 0.03736; OES: 0.07623; AdjOES: .06801) =
CPS: 504,366
OES: 288,973
AdjOES: 288,973
Half DB Selected Occupation Employment (CPS: 252,183; OES: 144,487;
AdjOES: 144,487)
Where Helpers Are Not Likely To Prevail:
Journeyworkers =
CPS: 252,183 x 0.77034 = 194,267
OES: 144,487 x 0.81966 = 118,430
AdjOES: 144,487 x 0.73127 = 105,659
Laborers =
CPS: 252,183 x 0.19229 = 48,492
OES: 144,487 x 0.10411 = 15,043
AdjOES: 144,487 x 0.20072 = 29,001
Apprentices =
CPS: 252,183 x 0.03736 = 9,422
OES: 144,487 x 0.07623 = 11,014
AdjOES: 144,487 x 0.06801 = 9,827
Half DB Selected Employment (CPS: 252,183; OES: 144,487; AdjOES:
144,487) Where Helpers Are Likely To Prevail for Some Occupations
Table 5.--Data for Helper Adjustment, Including Helpers
----------------------------------------------------------------------------------------------------------------
AdjOES
Occupation CPS No. CPS % OES No. OES % No. AdjOES %
----------------------------------------------------------------------------------------------------------------
Journey-worker................................ 379,377 0.75219 197,922 0.68492 197,922 0.68492
Laborer....................................... 94,699 0.18776 25,140 0.08700 54,326 0.18800
Apprentice.................................... 18,401 0.03648 18,406 0.06369 18,406 0.06369
Helper........................................ 11,889 0.02357 47,505 0.16439 18,319 0.06339
-----------------------------------------------------------------
Total..................................... 504,366 1.00000 288,973 1.00000 288,973 1.00000
----------------------------------------------------------------------------------------------------------------
Journeyworkers =
CPS: 252,183 x 0.75219 = 189,690
OES: 144,487 x 0.68492 = 98,962
AdjOES: 144,487 x 0.68492 = 98,962
Laborers =
CPS: 252,183 x 0.18776 = 47,350
OES: 144,487 x 0.08700 = 12,570
AdjOES: 144,487 x 0.18800 = 27,164
Apprentices =
CPS: 252,183 x 0.03648 = 9,200
OES: 144,487 x 0.06369 = 9,202
AdjOES: 144,487 x 0.06369 = 9,202
Helpers =
CPS: 252,183 x 0.02357 = 5,944
OES: 144,487 x 0.16439 = 23,752
AdjOES: 144,487 x 0.06339 = 9,159
Total:
Journeyworkers =
CPS: 194,267 + 189,690 = 383,957
OES: 118,430 + 98,962 = 217,392
AdjOES: 105,659 + 98,962 = 204,621
Laborers =
CPS: 48,492 + 47,350 = 95,842
OES: 15,043 + 12,570 = 27,613
AdjOES: 29,001 + 27,164 = 56,165
Apprentices =
CPS: 9,422 + 9,200 = 18,622
OES: 11,014 + 9,202 = 20,216
AdjOES: 9,827 + 9,202 = 19,029
Helpers =
CPS: 5,944
OES: 23,752
[[Page 17454]]
AdjOES: 9,159
Step 4: Alternative Wage Bills. Since half of Davis-Bacon
employment is estimated to be in areas in which helpers would not be
found to prevail for any classification, such employment would not have
been affected by the proposed regulation change. For the remaining
half, the occupational group totals--both before and after a possible
regulation change--are multiplied by the corresponding annual
salaries.<SUP>20</SUP>
---------------------------------------------------------------------------
\20\ Note that this methodology counts each additional helper
towards potential cost savings. However, results of relevant Davis-
Bacon wage surveys indicate that only a small proportion of helpers
would be in classifications in which helpers prevail, thereby
substantially reducing savings realized.
---------------------------------------------------------------------------
Alternative Wage Bills (1996)
CPS: (388,536 x 23,007) + (96,985 x 15,907) + (18,845 x 12,564) =
8,939,047,752 + 1,542,740,395 + 236,768,580 = 10,718,556,727
OES: (236,860 x 23,007) + (30,086 x 15,907) + (22,027 x 12,564) =
5,449,438,020 + 478,578,002 + 276,747,228 = 6,204,763,250
AdjOES: (211,318 x 23,007) + (58,003 x 15,907) + (19,652 x
12,564) = 4,861,793,226 + 922,653,721 + 246,907,728 = 6,031,354,675
Suspended Regulation
CPS: (383,957 x 23,007) + (95,842 x 15,907) + (18,622 x 12,564) +
(5,944 x 9,008) = 8,833,698,699 + 1,524,558,694 + 233,966,808 +
53,543,552 = 10,645,767,753
OES: (217,392 x 23,007) + (27,613 x 15,907) + (20,216 x 12,564) +
(23,752 x 9008) = 5,001,537,744 + 439,239,991 + 253,993,824 +
213,958,016 = 5,908,729,575
AdjOES: (204,621 x 23,007) + (56,165 x 15,907) + (19,029 x
12,564) + (9,159 x 9008) = 4,707,715,347 + 893,416,655 + 239,080,356
+ 82,504,272 = 5,922,716,630
Step 5: Estimated Annual Savings. Subtract the Davis-Bacon wage
bill computed assuming helper employment from the comparable wage bill
with no helpers employed. The difference is an estimate of potential
1996 savings. Divide that total by the value of Davis-Bacon
construction to obtain savings as a percent of 1996 Davis-Bacon-covered
construction starts.
Short-term Annual Savings:
CPS: 10,718,556,727 - 10,645,767,753 = $72,788,974
OES: 6,204,763,250 - 5,908,729,575 = $296,033,675
AdjOES: 6,031,354,675 - 5,922,716,630 = $108,638,045
Savings as a Proportion of the Value of 1996 Davis-Bacon
Construction Starts =
CPS: $72,788,974/$30,838,062,810 = 0.00236 or 0.236 percent;
OES: 296,033,675/$30,838,062,810 = 0.00960 or 0.960 percent;
AdjOES: $108,638,045/$30,838,062,810 = 0.00352 or 0.352 percent.
D. Findings
Given the above assumptions, data, process, and computations,
several key findings are established concerning the economic impact of
the suspended regulation:
1. Davis-Bacon Employment. The workforce on construction projects
covered by the Davis-Bacon and Related Acts is estimated to be under 1
million workers (CPS: 894,568; OES: 543,100; AdjOES: 543,100).
Occupational Employment (No Helpers). Davis-Bacon employment for
relevant occupations was estimated without the employment of helpers.
Under this scenario, employment for those occupations impacted directly
by the helper regulation was as follows:
Journeyworkers--CPS: 388,536; OES: 236,860; AdjOES: 211,318;
Laborers--CPS: 96,985; OES: 30,086; AdjOES: 58,003; and Apprentices--
CPS: 18,845; OES: 22,027; AdjOES: 19,652.
Occupational Employment (Helpers). In this case, Davis-Bacon
occupational employment in areas where it is assumed helpers would
prevail for at least one classification was as follows:
Journeyworkers:
CPS: 383,957
OES: 217,392
AdjOES: 204,621
Laborers:
CPS: 95,842
OES: 27,613
AdjOES: 56,165
Apprentices:
CPS: 18,622
OES: 20,216
AdjOES: 19,029
Helpers:
CPS: 5,944
OES: 23,752
AdjOES: 9,159
Wage Bills and Savings. Total earnings for each of the two
employment patterns described above were estimated as follows:
Without helpers:
CPS: $10,718,556,727
OES: $6,204,763,250
AdjOES: $6,031,354,675;
With helpers:
CPS: $10,645,767,753
OES: $ 5,908,729,575
AdjOES: $ 5,922,716,630
Therefore, possible savings are estimated to range from $72.8
million (CPS) or 0.236 percent of the value of 1996 Davis-Bacon
construction starts, to $108.6 million (AdjOES) or 0.352 percent, to
296.0 million (OES) or .960 percent. However, it should be noted that
these short-term savings realized through increased use of helpers
could be partially offset in the long run by higher journeyworkers'
wage rates.
This follows from the fact that helper use has been most extensive
among contractors who traditionally do not sponsor formal
apprenticeship and training programs. As increased helper use on Davis-
Bacon contracts might lead to contract gains for such employers,
reduced use of apprenticeship programs might lead to a somewhat smaller
supply of journeyworkers. This could cause a modest increase in
journeyworkers' wage rates, in the long run.
These findings indicate that previous Department of Labor estimates
of savings that could be attributed to the expanded use of helpers have
been greatly overstated. For example, while the current analysis places
possible annual savings from $72.8 to $108.6 to $296.0 million, earlier
estimates (1982 and 1989) placed such savings at $687.1 million and
$760.5 million (all 1996 dollars). While the current estimates' ratios
of savings to the value of Davis-Bacon construction starts are only
0.00236 to 0.00352 to 0.00960, estimates of the comparable 1982 and
1989 savings ratios would have been over twice what today's data
indicate. In addition, some State laws restrict the use of helpers on
public construction, thereby further reducing potential savings from
those estimated for Federal regulations that expand the use of helpers.
Several factors appear to be responsible for the wide variation in
savings estimates:
<bullet> The value of construction covered by the Davis-Bacon and
Related Acts, as a proportion of total construction value, about 9.6
percent, is significantly less than was previously assumed. Earlier
estimates of 18 percent and higher appear to have been based upon the
assumption that all non-Federal public construction is covered.
However, examination of available information does not confirm that
assumption. For example, experience working with F.W. Dodge information
indicates that the majority of city, county, and State-owned
construction has no Federal assistance. Specifically, by identifying
non-Federal public construction projects through F.W. Dodge reports,
and then determining their Davis-Bacon coverage through completed wage
survey forms for those projects, it
[[Page 17455]]
becomes clear that the majority of such construction is not covered.
<bullet> The previously utilized assumption that helpers would
prevail for 67 percent to 100 percent of the trades (and on projects
representing 67 to 100 percent of the Davis-Bacon employment) is not
confirmed by survey experience under the previously proposed
regulations or by other relevant information. For the 78 wage surveys
conducted under the new regulations, rates were recommended for helpers
in one or more classifications in just 35 of these data collection
efforts. Although all 12 open shop areas surveyed found one or more
helper classifications to prevail, they prevailed for only 20 percent
of the classes represented. For the 64 mixed area surveys, helpers were
found to prevail in 30 surveys, but for only 6 percent of the classes.
No helpers were found to prevail in the two area surveys that found all
prevailing rates to be union. Furthermore, 30 percent of the helper
classifications that were found to prevail were union helpers--
especially elevator constructor helpers, a classification negotiated
nationwide in that trade. Therefore, the assumption in this analysis
that one or more helper classifications prevail in areas that represent
half of Davis-Bacon covered employment is probably inflated in terms of
estimating the actual prevalence of helpers.
<bullet> Previous estimates of the proportion that helper
employment is of total construction employment appears to have
overstated that classification's workforce standing. For example, the
1976-77 compensation study, upon which many of the early helper savings
estimates were based, found that helpers comprised just 3.2 percent of
the survey universe. Because of the survey's concentration in
metropolitan/union areas and the fact that enough helper data were
found to publish for only four construction trades, that proportion was
doubled and tripled when developing alternative savings estimates.
Later estimates of helper employment proportions assumed that 15
percent of total construction employment fell into that classification.
As noted above, CPS, AdjOES, and OES estimates are approximately 1.3,
3.4, and 8.7 percent, respectively.
<bullet> The assumption that helpers will replace journeyworkers
exclusively was not supported by experience during implementation of
the suspended regulation. For example, personnel who processed helper
conformance actions have indicated that often construction contractors
surveyed reported that workers meeting the definition of helper in the
regulations were classified by the contractors as laborers. Similarly,
the low wage rates paid helpers are indicative of their lower skill
level, increasing the likelihood of substitution for laborers.
Recognizing helpers may perform work of laborers and apprentices, as
well as journeyworkers, narrows the differential between the wage bills
incurred before and after helper expansion. In fact, in the short run,
helpers may disproportionately assume work of laborers and apprentices.
In the longer run, supply problems in obtaining quality skilled
journeyworkers may well appear, as helpers displace apprentices, and
subsequently, apprentice-trained journeyworkers.
E. Possible Economic Impact of Helper Alternatives
A number of different approaches were considered in developing the
proposed regulation to define the circumstances in which helpers may be
used on Davis-Bacon projects. In addition to the proposal that helpers
only be permitted where the prevailing practice is to use helpers with
duties that do not overlap with those of a journeyworker or laborer,
Wage and Hour considered four other alternatives: (1) Add a ratio
requirement to the suspended helper definition; (2) change the helper
definition to emphasize the semi-skilled nature of the classification;
(3) define helpers in accordance with the OES definition which focuses
on unskilled duties; and (4) delineate the semi-skilled tasks performed
by each helper classification.
Section D of this Impact Analysis estimated helper use under the
suspended rule in areas where helpers would prevail. Alternatives 1-4
involved changing the helpers definition or their use. Each alternative
would likely result in greater use of helpers than under the proposed
rule, but less than under the suspended rule. Similarly, the economic
impact of the alternatives would presumably yield some portion but not
all, of the savings anticipated under the suspended rule.
Given that each alternative encompassed many possible variations
and outcomes, and that there is no data source that would provide
appropriate information on these variations and outcomes, it is not
possible to provide detailed estimates of the economic impacts of the
four alternatives. However, discussed below are the factors likely to
affect the economic impact of the alternatives.
Proposed Rule--Helpers Used in Accordance With Current Practice
The proposed rule would reflect the longstanding, and current,
practice of recognizing helpers only where helper duties are separate
and distinct from those of journeyworkers and laborers. As it would
continue a practice that has been in effect for many years, the
proposed rule is expected to have no economic impact.
Alternative 1--Add a Helper to Journeyworker Ratio Requirement to the
Suspended Rule
Adding a ratio, whether one ratio that applies nationally or a
number of local ratios, to the suspended rule would have the effect of
limiting the number of helpers allowed on Davis-Bacon sites, as
compared to the number that could be utilized under the suspended rule
alone. Where the practice of employers under the suspended rule without
a ratio would result in the use of more helpers than allowed under a
ratio cap, the economic impact would be lower savings with the cap than
without it. On the other hand, allowing helpers to be used under a rule
that combined the suspended rule with a ratio would allow greater
helper use than exists currently and would likely result in savings.
The amount of savings to be achieved would depend on the ratio chosen.
Alternative 2--Emphasize Semi-Skilled Nature of the Helper
Classification
Changing the suspended rule to emphasize the ``semi-skilled''
nature of the helper classification would likely result in less use of
helpers than there would be under the suspended rule, but more than
under the rule currently in effect. The extent of helper use would
depend on the scope of duties allowed under such a helper
classification. Thus, some savings would be achieved, but less than
would be expected under the suspended rule. The amount of savings would
also be impacted by how such a definition affected the relative
substitution of helpers for laborers and journeyworkers. As it could be
expected that emphasizing the semi-skilled nature of the helper
classification would result in little or no substitution for laborers,
the decrease in savings as compared to the suspended rule would be less
dramatic.
Alternative 3--Emphasize Unskilled Duties
As with Alternative 2, defining helpers by limiting their duties to
unskilled duties would also result in less use of helpers than there
would be under the suspended rule, but more than under the rule
currently in effect.
[[Page 17456]]
While some savings would be achieved, this amount would be less than
expected under the suspended rule. Again, the effect of the rule on the
substitution of helpers for laborers versus journeyworkers would impact
the degree of savings. Under this alternative, it could be expected
that few, if any, helpers would replace journeyworkers, resulting in
greater savings than would be expected under Alternative 2.
Alternative 4--Delineate Semi-Skilled Tasks for Each Helper
Classification
The extent of savings, as compared to current practice, under this
alternative would depend on the scope of the tasks allowed to be
performed by helpers assisting in each craft. Again, savings would be
expected relative to current practice, but in an amount less than would
be achieved under the suspended rule. As in Alternative 2, limiting
helpers to semi-skilled duties would likely result in less substitution
for laborers, and the decrease in savings as compared to the suspended
rule would be less dramatic.
F. Benefits
Wage and Hour originally believed that the primary benefits to be
gained from promulgation of the suspended helper regulation would be a
construction workforce on Federal construction projects that more
closely mirrored the private construction workforce's widespread use of
helpers, and significant cost savings in Federal construction costs. As
is more fully explained previously in this document, Wage and Hour now
believes that the use of helpers is less widespread than originally
thought and that the cost savings would be a small fraction of the
amount originally computed.
On the other hand, this proposal would allow Wage and Hour to
arrive at a definition of helper that would be capable of effective
administration and enforcement consistent with the purpose of the
Davis-Bacon Act. The alternatives considered would lessen the overlap
with other classifications, and would also provide a more objective
means by which both government agencies and contractors can distinguish
between helpers and other classifications, consistent with the
underlying purpose of the Davis-Bacon Act. All of the alternatives
would to varying degrees ameliorate the potential for misclassification
and abuse of helper classifications, thereby providing fairer
competitive bidding on Federal and federally-assisted construction
projects. Finally, Wage and Hour believes that this proposal could help
preserve effective training in the construction industry. A discussion
of the possible benefits provided by each of the specific proposed
alternatives immediately follows.
The proposed rule would continue the current practice which
requires that helper duties be separate and distinct from those of the
journeyworker and laborer. By retaining the traditional duties-based
classification distinction, it would provide clear criteria that can be
objectively administered and enforced, and that facilitate contractor
compliance. Because classifications would not have overlapping duties
under this alternative, there would be less opportunity for contractor
misclassification and abuse. Wage and Hour also believes that this
approach would encourage contractors to establish or participate in
structured training programs that would aid workers in achieving
journeylevel status.
Alternative 1, which would provide use of a national ratio, or a
number of local ratios, would reduce to some extent the potential for
abuse of the helper classification by contractors seeking to gain an
unfair competitive advantage, whether implemented in conjunction with
the suspended helper definition or with one of the other proposed
alternatives.
Alternative 2 would change the helper definition to emphasize the
semi-skilled nature of the classification by modifying the suspended
definition to emphasize semi-skilled duties. The modified definition
under this alternative might possibly aid in differentiating the helper
from journeyworker and laborer classifications by emphasizing the
``semi-skilled'' nature of the work performed by helpers, the
supervisory relationship between journeyworkers and helpers, and the
craft-specific assistance provided by the helper. This definition would
also expressly limit the unskilled work the helper may perform in an
attempt to distinguish helpers from laborers.
Alternative 3, which would utilize the OES definition of helper,
would provide a more objective definition of helper than the suspended
definition. By focusing on unskilled duties and the helper's
interaction with journeylevel craft workers, this alternative could
provide a more practical basis for distinguishing helpers from
journeyworkers.
Alternative 4, which would in essence adopt the ``job family''
concept currently utilized under the McNamara-O'Hara Service Contract
Act, would allow for the expanded use of helpers, with differentiation
based on the skill and knowledge required to perform various duties.
This would result in clearer definitions of helper classifications on a
craft-by-craft basis, which would facilitate administration and
enforcement.
VI. Regulatory Flexibility Act
Under the Regulatory Flexibility Act, Public Law 96-354 (94 Stat.
1164; 5 U.S.C. 601 et seq.), Federal agencies are required to prepare
and make available for public comment an initial regulatory flexibility
analysis that describes the anticipated impact of proposed rules that
would have a significant economic impact on small entities. Wage and
Hour is of the view that a Regulatory Flexibility Analysis is not
necessary for the proposed rule because the proposed regulation would
not result in any changes in requirements for small businesses.
Furthermore, if Wage and Hour were to propose implementing the
suspended rule or any of the alternatives considered, it would not be
more costly than current regulatory requirements and therefore would
not have a significant economic impact on a substantial number of small
entities. Furthermore, Wage and Hour is of the view, as discussed in
the preamble, that neither the suspended rule nor any of the
alternatives considered would accomplish the objectives of the statute.
Notwithstanding, because of widespread interest in the rule, Wage and
Hour has prepared the following Regulatory Flexibility Analysis, which
compares the proposed rule to the suspended rule and should be
considered in conjunction with the analysis set forth in the preamble
and the analysis under Executive Order 12866.
(1) Reasons Why Action Is Being Considered
In 1982, over fifteen years ago, Wage and Hour published final
regulations which, among other things, would have allowed contractors
to use ``semi-skilled'' helpers on Davis-Bacon covered projects at
wages lower than those paid to skilled journeyworkers. These rules
represented a sharp departure from Wage and Hour's longstanding
practice of not allowing overlap of duties between job classifications.
To protect against possible abuse, a provision was included limiting
the number of helpers which could be used on a covered project to a
maximum of two helpers for every three journeyworkers. This ratio
provision was subsequently invalidated by the U.S. Court of Appeals for
the District of Columbia.
As discussed in greater detail above, during its existence, the
helper rule has been the subject of considerable
[[Page 17457]]
litigation and Congressional attention. The rule has been enjoined by
the district court and modified on two occasions as a result of court
of appeals decisions. It has twice been implemented for short periods
of time. It has also been suspended on two occasions as the result of
Congressional action prohibiting Wage and Hour from spending any funds
to implement or administer the helper rule. On December 30, 1996, the
suspension was continued pending completion of this rulemaking.
The helper rule was originally proposed and adopted because it was
believed that it would result in a construction workforce on Federal
construction projects that more closely mirrored the private
construction's ``widespread'' use of helpers and, at the same time,
effect significant cost savings in federal construction costs. It was
also believed that the expanded definition would provide additional job
and training opportunities for unskilled workers, in particular women
and minorities. The Department's subsequent efforts to develop
enforcement guidelines led it to conclude that administration of the
revised helper rule would be much more difficult than anticipated,
especially in light of the court's invalidation of the ratio provision.
Moreover, new data, including the Department's experience implementing
the helper regulations, indicated that the use of helpers is not as
widespread as previously thought. Wage and Hour is also concerned about
the possible negative effect of the helper regulations on formal
apprenticeship and training programs. These factors, and the obvious
controversy evidenced by the rule's long history of litigation and by
Congressional actions prohibiting implementation of the rule, led Wage
and Hour to reexamine the helper rule and consider several alternative
approaches to govern employment of helpers on DBRA-covered projects.
(2) Objectives of and Legal Basis for Rule
These regulations are issued under the authority of the Davis-Bacon
Act, 40 U.S.C. 276a, et seq., Reorganization Plan No. 14 of 1950, 5
U.S.C. Appendix, and the Copeland Act, 40 U.S.C. 276c. The objective of
these regulations is to establish the most appropriate approach to
governing employment of helpers on DBRA-covered projects. Wage and Hour
believes the proposed rule is the only alternative considered that is
both consistent with the purposes of the Davis-Bacon Act and capable of
practical and efficient administration, enforcement, and compliance.
(3) Number of Small Entities Covered Under the Rule
Size standards for the construction industry are established by the
Small Business Administration (SBA), and are expressed in millions of
dollars of annual receipts for affected entities, i.e., Major Group 15,
Building Construction--General Contractors and Operative Builders, $17
million; Major Group 16, Heavy Construction (non-building), $17
million; and Major Group 17, Special Trade Contractors, $7 million. The
overwhelming majority of construction establishments would have annual
receipts under these levels. According to the Census, 98.7 percent of
these establishments have annual receipts under $10 million. Therefore,
for the purpose of this analysis, it is assumed that virtually all
establishments potentially affected by this rule would meet the
applicable criteria used by the SBA to define small businesses in the
construction industry.
As explained above, however, the proposed rule would cause no
impact on small entities since it does not propose to make any changes
in requirements applicable to small businesses. Implementation of the
suspended rule or any of the alternatives considered would expand the
use of helpers and could result in some savings. The impact would
depend upon the specifications of the alternative relative to current
practice. Even relative to unlimited use, however, possible savings
would be very modest, ranging from 0.239 percent of the value of Davis-
Bacon annual construction starts (CPS), to 0.359 (adjusted OES), and
0.958 (unadjusted OES) percent.
(4) Reporting, Recordkeeping and Other Compliance Requirements of the
Rule
There are no reporting or recording requirements for contractors
under the proposed rule. Nor would there be any such requirements under
the suspended rule or any of the alternatives considered. The
compliance requirements under any rule regarding helpers would merely
require contractors who use helpers to do so in accordance with a
chosen definition and pay helpers at least the appropriate prevailing
wages for helpers as set by the Department.
(5) Relevant Federal Rules Duplicating, Overlapping or Conflicting With
the Rule
There are currently no Federal rules that duplicate, overlap or
conflict with this proposed rule.
(6) Differing Compliance or Reporting Requirements for Small Entities
The proposed rule contains no reporting, recordkeeping, or other
compliance requirements specifically applicable to small businesses or
that differ from such requirements applicable to the Davis-Bacon
contracting industry as a whole. Such different treatment would not
seem feasible since virtually all employers in the industry are small
businesses.
(7) Clarification, Consolidation, and Simplification of Compliance and
Reporting Requirements
The compliance and reporting requirements of the proposed rule, the
suspended rule, and each of the alternatives considered, as well as the
advantages and disadvantages of each, are described in the preamble
above, which discusses issues such as ease of compliance for
contractors.
(8) Use of Other Standards
The Davis-Bacon Act requires the Secretary to determine the
prevailing wages and fringe benefits to be paid to the classes of
workers to be employed on a project. Therefore compliance by
contractors can only be achieved through design standards. The proposed
rule, the suspended rule, and the alternative approaches to employing
helpers on DBRA-covered projects are discussed in the preamble above
and are not repeated here.
(9) Exemption From Coverage for Small Entities
Exemption from coverage under this rule for small entities would
not be appropriate given the statutory mandate of the Davis-Bacon Act
that all contractors (large and small) performing on DBRA-covered
contracts must pay its workers prevailing wages and fringe benefits as
determined by the Secretary of Labor. Further, exclusion of such small
businesses from data collected to determine prevailing wages and fringe
benefits for helpers would be impractical and would distort such
determinations, possibly to the detriment of small businesses.
VII. Document Preparation
This document was prepared under the direction and control of John
R. Fraser, Deputy Administrator, Wage and Hour Division, Employment
Standards Administration, U.S. Department of Labor.
[[Page 17458]]
Signed at Washington, D.C., this 1st day of April, 1999.
Bernard E. Anderson,
Assistant Secretary for Employment Standards.
[FR Doc. 99-8566 Filed 4-7-99; 8:45 am]
BILLING CODE 4510-27-P