CRS Report for Congress
Employer-Provided Training
May 3, 2000
Linda Levine
Specialist in Labor Economics
Domestic Social Policy Division

Congressional Research Service ˜ The Library of Congress

It is widely thought that changes which firms have implemented in recent years to improve
their competitiveness have substantially affected the skill requirements of jobs. The upgrading
of skill requirements has, in turn, prompted Congress to spur employer-provided training of
incumbent (i.e., employed) workers through the Workforce Investment Act and the American
Competitiveness and Workforce Improvement Act. This represents a marked departure from
the typical populations served by government training resources. In this report, what is known
about the intensity, content and incidence of companies’ human capital investments is explored
along with the distribution of employer-sponsored training by business and employee
characteristics. Then, the impact on employees and employers of post-school skill acquisition
is analyzed. The report also examines whether employer-supported training of incumbent
workers is underprovided and the policy implications for Congress. This report will not be

Employer-Provided Training
The consensus is that the importance of a skilled workforce to the economic
performance of nations, firms and employees has increased in recent years. To
compete in today’s fast-paced global marketplace, companies have strived to increase
the efficiency of their operations through heightened utilization of technological and
other workplace innovations. It is widely thought that these changes have
substantially affected the skill requirements of jobs. The upgrading of job skill
requirements has, in turn, prompted Congress to spur employer-provided training of
incumbent (i.e., employed) workers through the Workforce Investment Act (P.L. 105-

220) and the American Competitiveness and Workforce Improvement Act (P.L. 105-

277, Title IV). This represents a marked departure from the typical populations
served by government training resources (i.e., low-income and dislocated workers).
Employer-provided education/training lacks a clear definition. It may be tailored
to the needs of a given firm (i.e., specific training) or impart skills that are useful
across firms (i.e., general training). Training may be informal (e.g., watching or
asking others) or formal (e.g., attending on-site classes or vocational schools). These
and other variations can differentially affect the costs and benefits of incumbent
worker training. The latest available estimates, which are for the mid-1990s, suggest
that companies’ expenditures on formal training ranged between $42 billion and $52
billion. In terms of benefits, employer-supported training raises employees’ wages or
improves their job stability and increases productivity or decreases turnover at firms.
It appears that a case can be made for government to promote general
(transferable) skills training because firms could well underinvest in this activity.
However, as much remains unknown about which kinds of firm-supported training are
most effective and how much formal and informal training already is being provided,
policymakers seeking to stimulate the training of incumbent workers may want to
tread warily in this new area. The possibility of a formal training subsidy prompting
firms to substitute federal monies for their own or substituting formal for unsubsidized
informal training rather than increase the total quantity of training also implies the
need for caution. And, while the currently low unemployment rate may have
contributed to policymakers’ focus on skill development for individuals who already
have jobs, incumbent workers are virtually guaranteed some degree of employer-
provided training. In contrast, unemployed persons have only their own presumably
scanty resources and, traditionally, federal training funds to draw upon to learn new
or upgrade existing skills.
To date, it appears that Congress has carefully entered the field of incumbent
worker training by devoting relatively small sums to this purpose (less than $80
million since 1999), including matching requirements in some instances and by
prohibiting the use of government resources for training that companies would have
undertaken anyway. The recency of the grants awarded to stimulate employer-
provided training makes it unlikely that evaluations will be available shortly to shed
light on the federal initiative, however.

What Is Employer-Provided Training?................................3
The Intensity, Content and Incidence of Training........................5
Trends in Training Costs......................................6
Training Content by Hours of Participation........................7
Basic Employability Skills.................................8
The Incidence of Training.....................................8
Informal Compared to Formal..............................8
By Firm Size...........................................9
The Distribution of Training......................................10
Who Provides Training?......................................10
Who Gets Trained?.........................................12
The Impact of Training..........................................14
Returns to Employees.......................................14
Earnings and Earnings Growth.............................14
Employment Stability....................................16
Returns to Employers.......................................17
Productivity ........................................... 17
Employee Turnover.....................................18
Some Policy Implications.....................................18
Is Employer-Supported Training Underprovided?......................20
List of Tables
Table 1. Wage and Salary Costs of Training and Hours of Training by Establishment
Size ...................................................... 9

Employer-Provided Training
Education and training are two examples of human capital investments, which
are expenditures that increase the resources embodied in people. Through the
expansion of people’s knowledge and skills, human capital investments raise
productivity and income in future years.1
Knowledge and skill formation contribute to a nation’s economic performance.
To the extent that income growth per capita is caused by increases in land and
physical capital per worker, the future rate of growth will slow because additional
assets result in diminishing returns. However, the development and application of
people’s scientific and technical knowledge enable economic growth to continue by
raising the productivity of labor and other factors used in the production of goods and
services.2 The adoption, adaptation and diffusion of technological and other
workplace innovations, in turn, rely on a nation’s having a well-educated and well-
trained labor force.3
The consensus is that the importance of a knowledgeable, flexible workforce to
the economic performance of nations and firms has increased in recent decades. The
industrial age was marked by incremental changes which had minimal impact on the
skill requirements of most jobs. To compete in the fast-paced global marketplace of
the post-industrial age, firms have strived to increase the efficiency of their operations
through heightened utilization of computer technology, reconfiguration of the
corporate structure and reorganization of production processes. These quantum
changes in the way business now is being conducted are widely thought to have
substantially affected the skill requirements of jobs. For example, factory workers
today might be required to operate sophisticated computerized equipment or to
participate in problem-solving teams. Firms sometimes have found that they first must
overcome deficiencies in these workers’ basic employability (e.g., computation and
communication) skills — which previously had not been much in demand — before
providing them with the specific skills necessary for implementation of workplace
innovations. 4

1 Becker, Gary S. Human Capital: A Theoretical and Empirical Analysis, with Special
Reference to Education. Chicago, University of Chicago Press, 1993. (Hereafter cited as
Becker, Human Capital: A Theoretical and Empirical Analysis.)
2 Productivity is the amount of output produced per unit of input (i.e., land, labor and capital).
In the case of labor productivity, it is a measure of the goods and services produced per
worker or hour worked.
3 Mincer, Jacob. Studies in Human Capital. England, Edward Elgar Publishing Limited,

1993. (Hereafter cited as Mincer, Studies in Human Capital.)

4 Hollenbeck, Kevin. Classrooms in the Workplace. Kalamazoo, MI, W.E. Upjohn Institute

The accumulation of knowledge and skills has come to play a more important
role in the labor market prospects of workers, as well. The premium employers pay
to workers with greater educational attainment has increased considerably in the past

2 decades: men (women) with bachelor’s degrees went from earning 50% (41%)

more than men (women) with high school diplomas in 1979, to 92% (76%) more in
1998.5 As a consequence of this trend, some policymakers are interested in upgrading
workers’ skills not only for efficiency but also for equity reasons. It is hoped that
additional training will mitigate the growth in wage inequality which began in the

1970s. 6

Although educational attainment is an often-used proxy for skill level, the nature
of recent workplace changes may have sped the rate at which formal education
becomes obsolete. Observers thus assert that post-school skill acquisition is now of
greater importance to adult workers who want to remain employable or to improve
their earnings. Demographic changes are believed to bolster the case for lifelong
learning as well: the aging of the labor force means that, for growing numbers of
workers, the period of formal schooling has long since passed; and the increasing
share of minorities “who, on average, experience lower high school completion rates,
lower educational attainment scores, and greater literacy problems” means more
workers could be unprepared for the labor market’s faster growing, higher paying
jobs. 7
Investments in the nation’s human resources are made when students attend
secondary or post-secondary educational institutions, unemployed workers enter
government-sponsored training programs and when employees participate in work-
based learning activities. This report focuses on the latter, that is, on the training
investments firms make in incumbent workers who no longer are attending school8
Although nearly all of the government’s training resources remain focused on
low-income or dislocated workers, Congress has shown increased interest since the
late 1990s in expanding the amount of training that firms provide to their current
employees. Passed in August 1998, the Workforce Investment Act (WIA, P.L. 105-

4 (...continued)
for Employment Research, 1993. (Hereafter cited as Hollenbeck, Classrooms in the
5 CRS Report 95-1081, Education Matters: Earnings by Highest Year of Schooling
Completed, by Linda Levine.
6 The increased return to education since the 1970s has prompted speculation that the
distribution of training may have contributed to the growth in wage inequality. According to
Jill Constantine and David Neumark (Training and the Growth of Wage Inequality. Industrial
Relations, v. 35, no. 4, October 1996), the increased incidence of training among more
educated workers was not large enough to have substantially affected the wage structure.
7 U.S. Department of Labor. Involving Employers in Training: Literature Review, Research
and Evaluation Report Series 97-K. Washington, D.C., 1997. p. 13.
8 In this report, the terms “incumbent” or “employed worker” and “employee” are used
interchangeably as are the terms “business,” “company,” “employer” and “firm”.

220) allows states to use some of their allocations for “innovative” incumbent worker
training programs (Section 134(3)(A)(iv)(I)) and the Secretary of Labor to carry out,
through grants or contracts, demonstration and pilot projects to develop and test the
effectiveness of various training approaches directed at diverse target populations,
including upgrading skills among employed persons living and working in enterprise
communities or empowerment zones (Section 171(b) and (c)). The American
Competitiveness and Workforce Improvement Act (ACWIA) was included as Title
IV of the FY1999 omnibus appropriations act (P.L. 105-277), which was enacted in
October 1998. In Section 414(c), additional funding was provided for WIA’s Section
171(b) demonstration programs that afford technical skills training to both employed
and unemployed workers. More recently, the Administration has requested $30
million for a new incumbent worker program to be initiated in FY2001. The
competitive grant program would focus on demonstrating innovative approaches to
training/skill upgrading among non-management incumbent workers.9
What Is Employer-Provided Training?
The complexity of the activities that comprise firm-sponsored education/training
makes this question difficult to answer. It is a complicated subject to study for the
following reasons:
!To begin with, it lacks a clear definition. Training may be tailored to
meet a business’ particular needs, but it also may impart knowledge
that would be useful to employees regardless of their place of
employment. Employers may require participation in post-school
skill development or employees may voluntarily undertake it.
Employment-based programs may be of shorter duration and have a
more remedial focus10 or a more occupational focus than school-
based programs that confer formal academic credentials.
Alternatively, some corporate educational programs grant degrees.
Perhaps the only common feature is the involvement of the firm, in
some manner, with the post-school knowledge acquisition of its

9 The meaning of incumbent worker training varies somewhat. One solicitation for grant
applications for an incumbent worker demonstration program stated that:
While in general the term “incumbent worker training” may be used to denote any existing
efforts on the part of employers to provide training to currently-employed workers in order to
help keep these employees employed, the term will be used in the solicitation to describe
efforts to keep firms and workers competitive by keeping workers employed, averting layoffs,
upgrading workers’ skills, increasing wages earned by employees, and improving employees’
In the FY2001 budget request, incumbent worker training not only is invoked as a means of
averting layoffs/plant closings due to trade or technological innovations but also as a synonym
for lifelong learning that enables firms to build career ladders for their employees.
10 “Basic skills” or “workplace literacy” instruction focuses on developing reading and writing
English, math, English-language communication and interpersonal skills.

employees.The nature of employer involvement in training provision
— financial and otherwise — varies. Companies may themselves
develop and have in-house personnel lead training programs or they
may pay others to perform all or part of the training function. At the
same time, they may subsidize their employees’ skill formation
through paid time-off from work or through tuition reimbursements.
The manner of government involvement in employer-sponsored
training also varies. Firms themselves may be eligible for financial
support from state governments that provide training grants or tax
credits as part of their economic development efforts, for example.
They also may be eligible for assistance through WIA, which permits
federal funds to be spent — for the first time — on the training of
employed persons (where no public announcement of a closing has
been made). In addition, for-profit firms generally may deduct
training expenses when calculating their federal income tax11
!The subject is further complicated because company investments in
incumbent workers are of different types, are delivered in various
ways and are of wide-ranging content. Employer-sponsored
education and training is informal (e.g., an experienced employee
showing a new-hire how to perform a task) or formal (e.g., an
employee attending a class or seminar). While informal training
typically occurs at the workplace, formal training may take place
there or off-site (e.g., at a community college or commercial trade
school). In the case of on-site formal training, a class may be led by
an instructor located at the firm or broadcast via satellite.
Alternatively, employees may learn from interactive, multimedia
technologies. And, the content of the training may relate to the
performance of particular jobs (e.g., management skills training) or
it may have broader applicability (e.g., interpersonal skills training).
Differences in such variables as the type, location, delivery mode and
content of training could well have disparate effects on the earnings
and productivity of incumbent workers — two commonly used
outcome measures.Lastly, employer-supported education and
training is neither transparent nor centralized. Rather, it is an
amalgam of not readily observable actions being undertaken largely
on an independent basis by billions of employers, other training
providers and employees. The lack of a uniform accounting method
for business’ investments in its employees has hampered the12
collection of high-quality data. Moreover, information needs to be

11 Businesses generally may expense training investments (i.e., account for them immediately)
rather than amortize them over time. These investments include compensation of in-house
trainers, compensation of trainees including educational assistance benefits and payments to
outside training providers.
12 “Firms do not keep good or standardized data on their training expenditures. When asked
to estimate the amount spent on training, some firms will estimate their actual program costs
while others will compute program costs plus the costs of the employee’s time spent in the

obtained from two groups (i.e., employers and employees) in order
to have comprehensive statistics on formal and informal training.
But, evidence suggests that the parties sometimes provide different13
responses to survey questions.
Thus, despite widespread agreement on the importance of employment-oriented
skill development among incumbent workers who no longer are full-time students,
large gaps remain in our understanding of the issue. The sections below set forth
what we know, first examining the intensity, content and incidence of companies’
human capital investments. Then, the distribution of training by business and
employee characteristics is explored. Next, the impact on employees and employers
of post-school skill acquisition is analyzed. The report closes with an examination of
whether employer-sponsored training of incumbent workers is underprovided and the
policy implications for Congress.
The Intensity, Content and Incidence of Training
It is difficult to measure accurately the cost of and time involved in employer-
provided training, as well as its content and prevalence, because much of it occurs
informally (i.e., learning by doing, watching or by talking with others). Formal
training (i.e., learning by attending a planned activity with a defined agenda that is
conducted in a structured setting) at least has a clear start and finish, which should14
make it easier to measure the cost of the activity. (The employees’ hourly
compensation must be multiplied by the time spent in training to derive the
opportunity cost of training to employers, i.e., value of the output forgone while
employees are being trained.) Because it is difficult to determine how much time is
spent learning informally, cost information usually is limited to formal training. It also
is problematic to ascertain the incidence of informal training because employees may
not readily recall these activities or not regard them (e.g., getting advice from co-
workers) as training per se.15

12 (...continued)
program while still others will impute an overhead rate to cover fixed costs (facilities, training
staff, etc.).” Osterman, Paul. Skill, Training and Work Organization. Industrial Relations,
v. 34, no. 2, April 1995. p. 133. (Hereafter cited as Osterman, Skill, Training and Work
13 Barron, John M., Mark C. Berger and Dan A. Black. How Well Do We Measure Training?
Journal of Labor Economics, v. 15, no. 3, part 1, July 1997.
14 Brown, Charles. Empirical Evidence on Private Training in: Commission on Workforce
Quality and Labor Market Efficiency. Investing in People: A Strategy to Address America’s
Workforce Crisis. Washington, D.C., U.S. Government Printing Office, September 1989.
(Hereafter cited as Brown, Investing in People.)
15 The National Longitudinal Survey of the High School Class of 1972 and the Current
Population Survey training supplements (1983 and 1991), which query individuals or
households, yield much lower incidences of informal training than the surveys of the Small
Business Administration and the Employment Opportunity Pilot Projects, which query

Trends in Training Costs
Annual approximations of formal training expenditures for the 1980s generally
cluster between $12 billion and $30 billion.16 The American Society for Training and
Development (ASTD) produced a more recent estimate by combining data from the
U.S. Bureau of Labor Statistics’ Survey of Employer-Provided Training, the National
Household Education Survey, and the ASTD Benchmarking Forum. In 1995, firms
in the private sector incurred $25.2 billion in direct costs and $27.1 billion in indirect
costs related to formal training for a total of $52.3 billion.17 In contrast, survey data
from the U.S. Bureau of Labor Statistics (BLS) show that during 1994 employers18
spent about $16.6 billion on selected direct costs of formal training. And, between
May and October 1995, employers’ payments for selected indirect costs of formal
training (i.e., trainees’ wages and salaries) were about $12.8 billion.19 Thus, for an
entire year in the mid-1990s the cost of formal training was about $42.2 billion. For
the mid-1990s, then, the latest available estimates suggest that companies’
expenditures on formal training ranged between $42 billion and $52 billion.
Employers appear to have increased their commitment to the formal training of
incumbent workers in recent years. According to a BLS survey of employers in the
private nonfarm sector with at least 50 employees on their payrolls, 69.2% raised their
expenditures on formal training during the early 1990s while just 5.2% lowered them.

15 (...continued)
employers. This discrepancy suggests that workers may not remember informal training very
well or they may not consider informal practices to be training while employers do. Barron,
John M., Mark C. Berger and Dan A. Black. On-The-Job Training. Kalamazoo, MI, W.E.
Upjohn Institute for Employment Research, 1997. (Hereafter cited as Barron, Berger and
Black, On-The-Job Training.)
16 Mangum, Stephen L. Evidence on Private Sector Training in Commission on Workforce
Quality and Labor Market Efficiency, Investing in People.
17 ASTD’s measure of direct costs includes in-house expenses for curriculum development,
salaries and benefits for training personnel and contractors, purchase and maintenance of
equipment and the space used for training, as well as outside expenses for tuition
reimbursement, contributions to union- and trade association-sponsored training funds,
travel/living expenses of employees while attending off-site training and payments to outside
training providers. ASTD’s measure of indirect costs includes the wages/salaries and fringe
benefits of employees while receiving training. Bassi, Laurie J., with Anne L. Gallagher and
Ed Schroer. The ASTD Training Data Book. VA, ASTD, 1996. (Hereafter cited as Bassi,
Gallagher and Schroer, The ASTD Training Data Book.)
18 BLS’ measure of direct training costs includes the following: wages and salaries of in-house
trainers, payments to outside trainers, tuition reimbursements and contributions to union- or
trade association-sponsored training funds. Unlike ASTD’s measure of direct costs, it
excludes payments for equipment, supplies, space, travel time and the benefits of in-house
19 While ASTD’s measure captures as part of indirect costs the benefits of trainees, BLS’
measure is limited to trainees’ wages and salaries. In addition, BLS’ sample excludes
establishments with fewer than 50 employees while ASTD’s sample includes them. Frazis,
Harley with Maury Gittleman, Michael Horrigan and Mary Joyce. Results from the 1995
Survey of Employer-Provided Training. Monthly Labor Review, June 1998.

Larger firms were more likely than smaller firms to have increased the financial
resources devoted to formal training over the period.20 Another employer survey, this
one designed by the National Center on the Educational Quality of the Workforce
(EQW), similarly found that 57% of private for-profit firms with at least 20 employees
expanded their training activities during the first half of the 1990s.21 Companies
continued to raise the level of training investments during the second half of the 1990s
as well.22
Training Content by Hours of Participation
Employees averaged 10.7 hours in formal training activities during the May-
October 1995 period, two-thirds of which involved job-skills development.
Computer-related instruction absorbed the most time (2.1 hours per employee),
followed by professional and technical skills training (1.3 hours per employee) and
production- and construction-related skills training (1.1 hours per employee).
Although management skills training was the most prevalent type of job-skills training
that establishments provided,23 it accounted for comparatively few hours (0.8 per
The development of more broadly applicable skills took up the remaining one-
third of formal training time. Much of this time was spent in communications,
employee development and quality training (1.4 hours per employee) and occupational
safety training (1.2 hours per employee). Not only did employees attend more
activities related to occupational safety training (0.6 activities per employee) than any
other type of formal training, but many firms also offered instruction on this subject
(72%). In contrast, orientation training was as prevalent among firms as safety
training according to the BLS employer survey, but it accounted for much less of an
employee’s work time (0.6 hours and 0.1 activities per employee).
The content of informal training closely reflects that of formal training.
Employees reported that production/construction training and computer training
accounted for more hours than other kinds of informal job-skills training.
Occupational safety and communications/employee development/quality training

20 U.S. Bureau of Labor Statistics. BLS Reports on the Amount of Employer-Provided
Formal Training. USDL 96-268, July 10, 1996. (Hereafter cited as BLS, BLS Reports on
the Amount of Employer-Provided Formal Training.)
21 National Center on the Educational Quality of the Workforce. First Findings from the
EQW National Employer Survey. Philadelphia, PA, University of Pennsylvania, 1995.
(Hereafter cited as National Center on the Educational Quality of the Workforce, First
Findings from the EQW National Employer Survey.)
22 McMurrer, Daniel P., with Mark E. Van Buren and William H. Woodwell, Jr.. The 2000
ASTD State of the Industry Report. VA, ASTD, 2000.
23 About 67% of private nonfarm establishments with at least 50 employees offered formal
training in management skills in May-October 1995, according to BLS Reports on the
Amount of Employer-Provided Formal Training.

accounted for more hours per employee than other kinds of informal general-skills
training. 24
Basic Employability Skills. Somewhat surprisingly in light of oft-heard25
accounts about the inadequacy of workers’ basic skills, relatively few employees
surveyed by the BLS said that they received or spent much time participating (either26
on a formal or informal basis) in workplace literacy training. Similarly, results from
the BLS and EQW employer surveys show that practically no time or funds were27
expended on formal remedial instruction. And, smaller firms appear to be even less
likely than larger firms to offer workplace education programs.28 While
acknowledging such findings, ASTD nonetheless found it “somewhat significant” that
any businesses teach employability skills because “such training covers the very basic29
types of skills that public education is intended to provide.”
The Incidence of Training
Informal Compared to Formal. Almost all incumbent workers at private
nonfarm establishments with 50 or more employees in the mid-1990s had received
informal training while with their current employers (95.8%). Formal training was30
less prevalent, with 84.4% of incumbent workers having participated.
Results from other sources similarly show that the incidence of informal training
greatly surpasses that of formal training. A 1992 U.S. Small Business Administration
(SBA) survey of private nonfarm employers found that just 6.9% of newly hired
employees received off-site formal training and 20.5% received on-site formal
training. In contrast, 90.6% of new-hires received informal training from management

24 BLS. BLS Reports on the Amount of Formal and Informal Training Received by
Employees. USDL 96-515, December 19, 1996. (Hereafter cited as BLS, BLS Reports on
the Amount of Formal and Informal Training Received by Employees.)
25 Firms reported that nearly 36% of job applicants whom they tested in 1998 for literacy
and/or math skills lacked the basic qualifications for the positions sought. American
Management Association. 1999 AMA Survey on Workplace Testing: Basic Skills, Job Skills,
Psychological Measure. NY, 1999. Between 25% and 40% of hourly paid workers had some
basic skills deficiency in the early 1990s. Hollenbeck, Classrooms in the Workplace.
26 About 3% of employees spent just 0.2 hours per employee in informal basic skills training.
Under 7% of employees spent less than 0.1 hours per employee in formal basic skills training.
BLS, BLS Reports on the Amount of Formal and Informal Training Received by Employees.
27 For example, just 0.1 hours per employee or 1% of formal training hours were spent
learning basic skills according to the 1995 BLS employer survey.
28 In 1991, less than 5% of firms with under 20 employees said they had established an
education program compared to 20%-30% of firms with 200-499 employees. Bassi, Laurie
J. Smart Workers, Smart Work. Washington, D.C., The Southport Institute for Policy
Analysis, 1992. (Hereafter cited as Bassi, Smart Workers, Smart Work.)
29 Bassi, Gallagher and Schroer, The ASTD Training Data Book, p. 55.
30 BLS, BLS Reports on the Amount of Formal and Informal Training Received by

or supervisors, 60.5% from co-workers and 64.5% from watching others. Among
other things, the lower incidence of formal training in the SBA compared to the BLS
survey suggests that employers are less likely to provide training that entails
substantial outlays (e.g., on tuition or transportation) to new-hires than to employees
with longer tenure who thereby have demonstrated an attachment to their firms.31
Table 1. Wage and Salary Costs of Training and Hours of Training by
Establishment Size
(May-October 1995)
Establishment size
Costs and hours50-99100-499500 or more
All employees employees employees
Total wages & salaries
paid to employees while in37.15.716.814.6
training ($ in billions)
Formal training12.81.3 5.5 5.9
Informal training24.24.311.3 8.7
Per-employee wage &
salary costs of training647462654754
Formal training224110215308
Informal training423352439446
Per-employee hours of44.540.148.042.6
Formal training13.4 8.213.516.6
Informal training31.131.934.526.0
Source: U.S. Bureau of Labor Statistics. BLS Reports on the Amount of Formal and Informal
Training Received by Employees. USDL 96-515, December 19, 1996.
Data on hours of training also reveal that most employer-provided training is
informal. As shown above in Table 1, the typical employee spent much more time
engaged in informal than in formal training during a 6-month period in 1995 (31.1
hours and 13.4 hours, respectively). According to the SBA survey, newly hired
workers averaged considerably more time in informal than in formal training as well.
By Firm Size. Smaller businesses offer much less formal training than larger
businesses. On average, incumbent workers at establishments with 50-99 employees
were engaged in formal training one-half as long as employees of establishments with

31 Barron, Berger and Black, On-The-Job Training. Note: A nationally representative sample
of both small and large establishments were surveyed.

at least 500 employees. (See Table 1.) Smaller firms also spent one-third as much
per employee as firms with at least 500 employees in 1994 on such direct costs of
formal training as the wages and salaries of in-house trainers, payments to outside
trainers, tuition reimbursements and contributions to outside training funds.32
In contrast, smaller firms especially utilize informal training. At employers with
50-99 workers, employees reported that they averaged 31.9 hours of informal training
in May-October 1995, or almost 4 times the amount of formal training. (See Table
1.) At establishments with 100-499 employees, the 34.5 hours of informal training
per employee was little more than 2½ times the amount of formal training. And, at
companies with 500 or more employees, the 26.0 hours of informal training was an
even smaller multiple of formal training.
As shown in Table 1, informal training accounted for about two-thirds ($24.2
billion) of employers’ spending on the wages and salaries of participants in both
formal and informal training ($37.1 billion) during a 6-month period in 1995. While
smaller firms’ expenditures for both informal and formal training were less than those
of larger firms, the gap was widest for formal training. This again indicates the
considerable use of informal training among smaller firms.
The Distribution of Training
The surveys from which the following findings were derived differ in a number
of respects, and for that reason among others, their results sometimes disagree.
Different populations are surveyed (e.g., employers as opposed to employees, all
establishments versus only those with at least 50 employees or a nationally
representative sample in contrast to one that disproportionately includes low-wage
firms or is confined to young adults). Questions are phrased in varying ways (e.g.,
how long does it typically take an employee to become fully trained and qualified for
a particular job versus the length of training that an employee actually engaged in).
And, the time period referred to differs (e.g., training undertaken while at the current
employer or within the first 3 months on the job). Nonetheless, there are several
points on which the empirical literature agree.
Who Provides Training?
It is well-established that larger firms more often provide training, particularly33
formal training, than smaller firms. The positive relationship between formal training

32 BLS, BLS Reports on the Amount of Employer-Provided Formal Training.
33 According to a 1993 BLS survey of formal training of private nonfarm establishments, for
example, just 69% of the smallest companies (i.e., 0-49 employees) provided formal training
compared to 98% of firms with 50-249 employees and 99% of firms with at least 250
employees. The EQW survey includes incidence data for both formal and informal training:
on average, 81% of private for-profit firms with 20 or more employees offered formal training
in 1994, while the same was true for just 75% of firms with 20-49 employees; the gap in

and firm size, even after other variables are taken into account, may be due to larger
companies’ enjoying economies of scale. Put another way, up-front expenses could
be less of a barrier to training for larger firms because they have more employees
across which to spread those costs that are little changed as the number of trainees
increases (e.g., the cost of course development or the salary of an in-house
instructor). Perhaps for the same reason, small employers that are part of multi-
establishment firms are more likely to provide formal training. That is to say, multi-
establishment businesses could spread fixed training expenses over a large number of
plants/offices and thereby keep down their smaller locations’ human capital costs.
Another interpretation of the direct relationship between training and firm size is that
smaller firms may be more fearful of employees being hired away by competitors.
Moreover, the production losses incurred when an employee participates in a formal
training program held off-site may be greater at smaller firms.34 Similarly, larger firms
may find it easier to free-up co-workers to informally train others without
experiencing substantial output losses.35
Firms that experience comparatively high employee turnover tend to less often
sponsor training. Training also is diminished at companies that are more sensitive to
downturns in the business cycle or that are in areas with persistently high
unemployment. A potential explanation for the inverse relationship between training
and these three variables — employee turnover, volatile product demand and high
unemployment — is that employers are reluctant to invest in employees when they
know chances are great that the worker will be laid off or quit before the cost of their36
investment can be recouped.
In contrast, industries that experience rapid technological change are more likely
to provide formal on-the-job training to their incumbent workers generally and to do
so for a larger share of their employees. By implication, technological change may
make knowledge and skills obsolete thereby promoting greater company training. But,
because the training gap between more and less educated is estimated to narrow at
firms experiencing rapid technological change, it appears that the general skills of
highly educated workers may enable them to adjust to new innovations with less

33 (...continued)
informal training is much narrower, with 97% of firms employing at least 20 employees
offering informal training versus 96% of firms with 20-49 employees. Frazis, Harley J.,
Diane E. Herz and Michael W. Horrigan. Employer-Provided Training: Results from a New
Survey. Monthly Labor Review, May 1995; and National Center on the Educational Quality
of the Workforce, First Findings from the EQW National Employer Survey.
34 Lynch, Lisa M., and Sandra E. Black. Beyond the Incidence of Employer-Provided
Training. Industrial and Labor Relations Review, v. 52, no. 1, October 1998. (Hereafter
cited as Lynch and Black, Beyond the Incidence of Employer-Provided Training.)
35 Barron, Berger and Black, On-The-Job Training.
36 Lillard, Lee A., and Hong W. Tan. Private Sector Training: Who Gets It and What Are Its
Effects? Research in Labor Economics, v. 13, 1992. (Hereafter cited as Lillard and Tan,
Private Sector Training: Who Gets It and What Are Its Effects?)

additional training than is needed by workers with fewer years of schooling.37 Other
organizational transformations have been found to have similar effects on employer-
provided training. For example, firms that have implemented innovative workplace
practices (e.g., requiring employees to exercise responsibility for quality control or for
problem solving) not only exhibit a greater probability of offering formal training but
also are more likely to train a greater share of their workforces. The same findings
hold, all else being equal, for businesses that are relatively capital-intensive and have
comparatively well-educated workforces. Employer-supported training thus seems
to complement rather than substitute for investments in physical and human capital,
which suggests that concerns about businesses replacing high-skilled workers with
new technologies or low-skilled workers may not be well-founded.38
Companies that try to actively maintain a long-term relationship with their
employees appear to sponsor more training as well. According to one empirical
analysis, the hours of formal training are greater at establishments that offer more
benefits (e.g., employee assistance plans, employer-financed child care, health
insurance, pensions or paid family leave) and that have more innovative workplace
practices (e.g., pay for knowledge, job redesign or rotation, quality circles, or
teams).39 Another study similarly found that off-the-job training is more likely at firms
that feel it is important to help improve the well-being of their employees’ personal
and family lives (e.g., through the provision of family friendly benefits) and that have
instituted innovations characteristic of “high-performance workplaces” (e.g., quality
circles or total quality management).40 In contrast, various studies have come to41
different conclusions about the effect of unionization on employer-provided training.
Who Gets Trained?
Educational attainment is a significant determinant of who gets trained, even
after holding other variables constant. The probability of formal training receipt
increases with years of schooling, that is to say, education and training are
complements rather than substitutes for one another. Because individuals who enter
the labor force with relatively limited education thus face little prospect of additional

37 Bartel, Ann P., and Machum Sicherman. Technological Change and the Skill Acquisition
of Young Workers. Journal of Labor Economics, v. 16, no. 4, 1998.
38 Lynch and Black, Beyond the Incidence of Employer-Provided Training.
39 Frazis, Harley, Maury Gittleman and Mary Joyce. Correlates of Training: An Analysis
Using Both Employer and Employee Characteristics. Industrial and Labor Relations Review,
v. 53, no. 3, April 2000. (Hereafter cited as Frazis, Gittleman and Joyce, Correlates of
40 Osterman, Skill, Training and Work Organization. Note: The positive relationship between
organizational change and the incidence of formal training also is demonstrated in research
that utilizes a survey of individuals rather than of firms. Leigh, Duane E., and Kirk D.
Gifford. Workplace Transformation and Worker Upskilling: The Perspective of Individual
Workers. Industrial Relations, v. 38, no. 2, April 1994.
41 See, for example, Frazis, Gittleman and Joyce. Correlates of Training; and Lynch and
Black, Beyond the Incidence of Employer-Provided Training.

structured learning once employed, company training effectively widens the skills gap
between less and more educated workers.42
Some demographic features tend to depress the chance of training receipt. All
else being equal, the probability of non-white males obtaining training was estimated
to be significantly lower than that of other males. Racial differences in the incidence
of training appear to be less evident among younger than older men. The effect of
race on training is not apparent among women or among low-income workers,
according to one analysis. It also concluded that economically disadvantaged workers
(regardless of race) are less likely to get post-school training.43
Studies have come to mixed conclusions about the impact of gender on training
receipt, in part because they have measured different things (e.g., the intensity as
opposed to the incidence of training) or have disaggregated the data in different ways
(e.g., all women as opposed to women by race or marital status). Although some
studies estimate that men and women spend about the same amount of time in
training, one determined that women were more likely to fill positions that require less
training.44 The latter finding may result from differences in job turnover by sex (a
reflection of the weaker attachment to the labor market of women generally and of45
married women especially) or from discrimination. While there may or may not be
much of a gender gap in the overall incidence of training, the sources of training do
seem to vary by sex even after controlling for other variables: young men are more
likely to participate in company training and in apprenticeships and young women in46
off-site or off-the-job training (e.g., vocational schools or seminars outside of work).
Agreement is widespread in the economic literature that the probability of
training receipt rises with employee tenure. Consequently, those employees who
accumulate little tenure with a firm because they change jobs often or enter/exit the
labor force frequently are less likely to be offered training. At some point in the
seniority or age spectrum, however, training tapers off. This probably is, in part,
because the payoff period for the firm’s investment shortens.
There are several reasons why firms might want to delay offering training to
recently hired workers, although doing so means that they forgo the return to training
that would have occurred during the early part of the employment period. Employers
might put off the training of new-hires because they are uncertain at the outset of the
employment relationship about the likelihood of employees’ leaving before the

42 Lynch and Black, Beyond the Incidence of Employer-Provided Training.
43 Lillard and Tan, Private Sector Training: Who Gets It and What Are Its Effects?
44 Veum, Jonathan R. Gender and Race Differences in Company Training. Industrial
Relations, v. 35, no. 1, January 1996; and Barron, John M. with Dan A. Black, and Mark A.
Loewenstein. Gender Differences in Training, Capital and Wages. The Journal of Human
Resources, v. 28, no. 2, spring 1993.
45 Royalty, Anne Beeson. The Effects of Job Turnover on the Training of Men and Women.
Industrial and Labor Relations Review, v. 49, no. 3, April 1996.
46 Veum, Jonathan R. Training Among Young Adults: Who, What Kind, and For How Long?
Monthly Labor Review, August 1993.

companies’ investment can be recouped. Alternatively, employers might postpone
training beyond the first year of employment if they think individuals are better able
to learn after they have become acclimated to their job and to the work environment.
Companies also might delay training until they determine who are the likely candidates
for promotion and then concentrate their investments on them. Regardless of the
reason, training apparently is not limited to recently hired workers; instead, it is an
ongoing process provided to employees throughout much of their tenure at a firm.47
The Impact of Training
Workers undertake training to raise their earnings and improve their job security
over otherwise comparable individuals. For their part, employers invest in training to
increase labor productivity and to decrease costly employee turnover.
Returns to Employees
Earnings and Earnings Growth. Empirical studies unanimously confirm a
major tenet of human capital theory, which is that training plays an important role in
wage determination.48 Trained workers earn higher wages and their wages rise more
rapidly than those of comparable employees, according to numerous nonexperimental
analyses. 49
The timing and size of the payoff to training could vary with the kind of training.
Specific training — defined as the development of skills that enhance the productivity
of a worker only at the firm that offers it — likely provides a fairly immediate return
to the worker. (Examples of specific human capital are knowledge of the personnel
policies or organization of work processes at a particular employer.) General
training — defined as the development of skills that enhance the productivity of a
worker at many employers — may provide a somewhat delayed return. (Examples
of general human capital are the knowledge of computer software or automotive
repair.) General skills development also may have a larger payoff because of its
greater portability across firms. In other words, employers presumably are able to
pay smaller wage premiums to retain employees who have had specific training

47 Loewenstein, Mark A., and James R. Spletzer. Delayed Formal On-The-Job Training.
Industrial and Labor Relations Review, v. 51, no. 1, October 1997.
48 See, for example, Altonji, Joseph G., and James R. Spletzer. Worker Characteristics, Job
Characteristics, and the Receipt of On-The-Job Training. Industrial and Labor Relations
Review, v. 45, no. 1, October 1991; Lynch, Lisa M. Private-Sector Training and the Earnings
of Young Workers. American Economic Review, v. 82, no. 1, March 1992; and Mincer,
Studies in Human Capital.
49 Employees may not be randomly selected for training, in which case trainees could have had
higher wages or greater productivity than non-trainees even without training. To avoid
overstating the impact of training per se, the studies discussed above identify those factors
available in the particular database being used that are known to effect the chance of being
trained (e.g., education level or firm size) and control for them when estimating the return to
training for employees and employers.

because their skills would not be equally valued and therefore equally rewarded by
other employers.50
Consequently, some studies have analyzed the impact on earnings and earnings
growth of different kinds of formal training. But, very few surveys explicitly inquire
whether training is specific or general. They instead may ask about the source of the
training, with some sources (e.g., the company) thought to offer training with greater
specific content than others (e.g., vocational-technical school). Their results,
discussed below, are sometimes contradictory or inconclusive.
One analysis found that participation in formal company training and attendance
at off-the-job seminars, which may impart specific or general skills, were positively
related to wage levels.51 Participation in company training improved the wage growth
of men and women. Attendance at seminars did so for men only. The rate of earnings
gains also was estimated to be directly related to the length of vocational-technical
training but unrelated to its incidence. This suggests that, in the case of vocational
training, what is important to producing a return on investment is not just
participating in the program but actually completing it. Notably, proprietary
institutions other than vocational-technical schools (e.g., correspondence courses and
business schools) did not appear to raise workers’ earnings growth, at least not within
the 4-year period under study.52
Not only was formal company training estimated to have a significant positive
impact on earnings growth in another study, but so too were apprenticeships and
correspondence schools. In addition, business and vocational-technical school
training as well as apprenticeships appear to have a delayed impact on earnings gains.
These findings seemingly contradict those described in the preceding paragraph,
perhaps because the reference period of research must be longer to pick up the wage
effects from certain training sources. Company training, in contrast, seems to have
a more immediate impact on wage growth but the impact lessens over time. Company
training includes formal programs staffed in-house, vendor training held at the
worksite but run by an outside training provider and outside seminars that employees
leave the worksite to attend. Of the three, only vendor training was found to have no
influence on wage growth.53
Interestingly, the wage effect of training appears to be larger if employers pay
for their employees’ school-based training than if the employees pay for it

50 Becker, Human Capital: A Theoretical and Empirical Analysis.
51 Off-the-job training includes business school, nursing programs, vocational-technical
institutes, barber and beauty schools, flight school, correspondence courses, seminars or
training programs outside of work, vocational rehabilitation centers and other.
52 Veum, Jonathan R. Sources of Training and Their Impact on Wages. Industrial and Labor
Relations Review, v. 48, no. 4, July 1995.
53 Lengerman, Paul Adrian. The Benefits and Costs of Training: A Comparison of Formal
Company Training, Vendor Training, Outside Seminars, and School Based Training. Human
Resource Management, fall 1996, v. 35, no. 3. (Hereafter cited as Lengerman, The Benefits
and Costs of Training.)

themselves.54 One interpretation of this finding is that companies are better than
workers at choosing training that provides skills which are most job-relevant. If so,
government subsidies to promote incumbent worker training may be more effective
at raising employees’ earnings if they are offered to firms rather than to individuals
through, for example, vouchers. But, an alternative explanation of the smaller wage
effect for employee-financed school-based training is that individuals may elect to
acquire skills that improve their career mobility in the future rather than skills that55
immediately raise wages on their current jobs.
Another analysis also found that workers who undertook either formal company
or off-the-job training had higher wages than workers who did not. The return was
greater for company training, and particularly for completed programs. It did not
seem to matter whether the company training took place while the worker was at the
prior or current employer. This result suggests that company training may not be
synonymous with the development of specific skills. Rather, company training56
appears to impart skills that are valued equally regardless of who is the employer.
Other researchers similarly have concluded that, because current employers reward
employees for skills they acquired while at previous firms, the skills largely are
general and are recognized as such by other employers.57
Employment Stability. Here too the issues are not only whether training
benefits workers through reduced unemployment but also whether certain kinds of
training are more effective than others at enabling workers to find and keep jobs.
Studies have most often examined training’s effect on wages, with much less analysis
of training’s impact on job stability or on the other outcome variables discussed
shortly in this report.
According to one study that looked only at young men, training is associated
with a reduced likelihood of unemployment and this effect may continue for 12 years.
Of the sources of training, the impact of company training on the incidence of
unemployment was the most enduring at almost 13 years. Training provided by
business or technical schools was not found to significantly inhibit unemployment. Of

54 In addition to Lengerman, The Benefits and Costs of Training, other studies that make the
same point are Bowers, Norman, and Paul Swaim. Recent Trends in Job Training.
Contemporary Economic Policy, v. 12, January 1994; Grubb, Norton. The Varied Economic
Returns to Postsecondary Education: New Evidence from the National Longitudinal Study of
the Class of 1972. Journal of Human Resources, v. 28, no. 2, spring 1993; and Loewenstein,
Mark A. and James R. Spletzer. Dividing the Costs and Returns to General Training.
Journal of Labor Economics, v. 16, no. 1, January 1998.
55 Lynch, Lisa M. Private-Sector Training and the Earnings of Young Workers. American
Economic Review, March 1992.
56 Parent, Daniel. Wages and Mobility: The Impact of Employer-Provided Training. Journal
of Labor Economics, v. 17, no. 2, 1999. (Hereafter cited as Parent, Wages and Mobility: The
Impact of Employer-Provided Training.)
57 Loewenstein, Mark A., and James R. Spletzer. General and Specific Training. Journal of
Human Resources, v. 34, no. 4, fall 1999. (Hereafter cited as Loewenstein and Spletzer,
General and Specific Training.) and Veum, Jonathan. Training, Wages and the Human
Capital Model. Southern Economic Journal, v. 65, no. 3, 1999.

the types of training, professional/technical training most reduced the probability of
unemployment but its impact lasted less than 12 years. In contrast, the positive
impact of semiskilled manual training on employment stability was determined to
persist for 12.2 years. And, managerial training did not significantly effect the
likelihood of unemployment.58
Returns to Employers
Productivity. The effect of training on labor productivity has been particularly
difficult to empirically ascertain because surveys of individuals that include training
questions do not also include questions about employers’ economic performance.
The dearth of information has led researchers to rely on a very few employer surveys
or to use subjective measures of productivity.
One study of manufacturers found those that implemented new formal training
programs after 1983 experienced significant productivity increases during the 1983-

1986 period. However, the extremely low response rate of surveyed establishments59

tempers the reliability of the analysis.
Another employer-based study that used as a training variable the number of
workers involved in training in 1990 and 1993 did not find a significant impact on
productivity. But, the training variable itself probably contributed to the result as it
did not capture the accumulated training of all workers. In contrast, the study also
determined that the proportion of time spent in off-the-job training was directly
related to productivity in manufacturing establishments. The researchers suggest this
positive effect results from lower output losses because the training occurred outside
work hours or from the advanced nature of this type of training. In the
nonmanufacturing sector, computer skills development was found to have a positive
effect on productivity. Perhaps, then, the productivity effect at some establishments60
is due more to the content than to the incidence of training.
Still other employer-based studies utilized a firm’s subjective rating of its
employees’ productivity. The time spent in formal and informal training was61
estimated to be positively related to productivity growth. In addition, formal and
informal company training received from a previous employer were found to increase
productivity. Off-the-job training seems to generate even more substantial, longer
lasting productivity gains for the worker’s subsequent employers.62

58 Lillard and Tan, Private Sector Training: Who Gets It and What Are Its Effects?
59 Bartel, Ann P. Productivity Gains from the Implementation of Employee Training
Programs. Industrial Relations, v. 33, no. 4, October 1994.
60 Black, Sandra E., and Lisa M. Lynch. Human-Capital Investments and Productivity.
American Economic Review, v. 86, no. 2, May 1996, Papers and Proceedings of the 108th
Annual Meeting of the American Economic Association.
61 Holzer, Harry J. The Determinants of Employee Productivity and Earnings. Industrial
Relations, v. 29, no. 3, fall 1990; and Barron, Berger and Black, On-The-Job Training.
62 Bishop, John H., The Impact of Previous Training on Productivity and Wages in Lynch,

Employee Turnover. In theory, an investment in specific human capital binds
workers to the firm that makes the investment. Expressed differently, training that is
very tailored to the needs of a given firm is expected to discourage quits or layoffs
that end the mutually beneficial relationship (i.e., employees are paid higher wages and
experience greater wage growth than they could get elsewhere because the training
is not portable and employers enjoy productivity gains that exceed the worker’s
wage). An investment in general human capital, in contrast, is not expected to affect
worker mobility because the transferability of general skills means that workers are
equally productive — and therefore can command the same wage levels and wage
growth — at any firm.
As anticipated, employer-provided training reduces the probability of employee63
separations. While this finding supports the idea that company training is specific,
another study estimated the negative effect of employer-provided training on
employee turnover is quite small. The magnitude of the mobility effect, in addition
to the economists’ finding that current employers reward employees for training
received while at prior employers, led them to surmise that much employer-provided
is general.64
In an analysis of workplace education programs, it was found that firms which
offered general training in basic skills were more likely to report improved employee
retention. Rather than raising turnover as workers with readily portable skills leave
for or are poached by other companies, this research suggests that the acquisition of
general skills may actually lower separations.65
Some Policy Implications
Summarizing the findings on the impact of training, employer-supported training
!trainees through higher wages,
!training firms through higher productivity, and
!third parties who, in this case, are firms that hire employees with
prior general training.
The existence of positive externalities — benefits that spillover beyond the parties
immediately involved in an activity — imply that firms may not be offering an optimal
quantity of training. In deciding how much training to offer incumbent workers,
employers consider the costs and benefits of training to them. That is to say, they do

62 (...continued)
Lisa M. (ed.) Training and the Private Sector: International Comparisons. IL, The
University of Chicago Press, 1994. (Hereafter cited as Bishop, The Impact of Previous
Training on Productivity and Wages.)
63 Mincer, Studies in Human Capital; and Parent, Wages and Mobility: The Impact of
Employer-Provided Training.
64 Loewenstein and Spletzer, General and Specific Training.
65 Bassi, Smart Workers, Smart Work.

not take into account the positive or negative impact of their decision on others. To
the degree that having a skilled workforce confers benefits that exceed the private
return to the training firm, employers will not offer the optimal amount of training
(i.e., training is said to be underprovided). The market’s seeming failure to provide
a socially optimal level of training is a rationale for government involvement in
employer-sponsored training.
There is additional evidence which suggests that training not only generates
benefits for other firms but also for society at large. As previously discussed,
individuals with relatively more training have higher earnings and greater job security.
Consequently, they are more likely to pay higher taxes and less likely to utilize
government services (e.g., unemployment insurance or cash welfare and health
benefits). As it thus seems that employer-supported training does produce third-party
externalities and the market fails to offer it at a socially optimal level, “modest
governmental efforts to stimulate general OJT [on-the-job] and employer-sponsored66
formal off-the-job training would appear to be in order.” (See the final section of
this report for further elaboration on this point.)
Although employer-sponsored education and training has been found to benefit
both employees and employers, the research results demonstrate that very little is
known about which kinds of training (e.g., formal versus informal training or formal
on-the-job versus off-the-job training) are most effective for employees and firms
overall, or for different kinds of employees (e.g., minorities or low-wage workers) and
firms (e.g., small employers). Accordingly, policymakers seeking to stimulate
incumbent worker training may not want to be too prescriptive about its content,
delivery mode or target population.
For equity (i.e., distributional) reasons, however, Congress may want to focus
training resources on less educated employees whose relative disadvantage upon
entering the labor force is exacerbated by the current pattern of employer-provided
formal training. Indeed, demonstration programs under WIA (or its predecessor, the
Job Training Partnership Act) and under ACWIA that focus on dislocated worker,
incumbent worker or skills shortages training sometimes have targeted groups that
historically have needed assistance in overcoming employment barriers (e.g.,
minorities, disabled and low-skilled workers) and they have included a mixture of67
basic employability and occupation- or industry-related skills development.
However, a federal subsidy targeted at less-educated or otherwise disadvantaged
workers may not induce additional training. Instead, it may prompt companies to
substitute formal for informal training of these groups. This could be the outcome of
a broad-based training incentive as well. The result of a training subsidy or mandate

66 Bishop, The Impact of Previous Training on Productivity and Wages, p. 194.
67 See, for example, the Employment and Training Administration’s press release on H-1B
technical skills training grants at: [http://www.wdsc.org/sga/awards/99-019finalsum.htm].

thus may be to raise expenditures on formal training, lower the amount of informal
training and leave employees’ earnings unchanged.68
This discussion assumes that policymakers would be interested in subsidizing
only formal training because of the difficulty in measuring expenditures on informal
training (i.e., distinguishing it from normal work and supervision). This assumption
is borne out by a statement in solicitations for grant applications for dislocated or
incumbent worker training demonstrations that expressly prohibits the use of federal
funds to pay participants’ salaries. While trainees’ salaries are indirect costs of both
formal and informal training, they are likely to account for more of the expenditures
on informal training which does not entail certain expenses associated with formal
training (e.g., payments to outside training providers for curriculum development and
instruction; tuition reimbursement; transportation, lodging or meals; and purchase or
maintenance of equipment and classroom space). By putting off-limits the salaries of
employees engaged in training, the language in the grant solicitation favors the
provision of formal over informal training.
As previously mentioned, however, it is not known whether the payoff to formal
training is greater than to informal training. And, it is possible that certain types of
skills may be less costly to learn (i.e., more efficiently provided) through informal
means. Consequently, the availability of a formal training subsidy may reduce the
efficiency of human capital production by encouraging the use of more expensive
methods. Further, since smaller firms are more likely than larger firms to train
informally, “a subsidy of formal training programs ... is an implicit decision to
subsidize larger firms.”69 Perhaps to compensate for this bias, the Labor Department
in one instance geared a portion of incumbent worker demonstration grants to assist
in the training of employees of small and medium-sized companies (i.e., those with70

500 or fewer employees).

Is Employer-Supported Training Underprovided?
The question of whether there is a training deficit among incumbent workers
actually is concerned with whether companies are providing the optimal level of
general training.71 Firms find it more attractive to provide specific than general

68 Barron, Berger and Black, On-The-Job Training; and Bishop, John H. What We Know
About Employer-Provided Training: A Review of the Literature. Center for Advanced
Human Resources Studies Working Paper 96-09. NY, Cornell University, School of
Industrial and Labor Relations, July 1996. (Hereafter cited as Bishop, What We Know About
Employer-Provided Training.)
69 Barron, Berger and Black, On-The-Job Training, p. 114.
70 U.S. Department of Labor. Employment and Training Administration. Job Training
Partnership, Title III, Demonstration Program: Incumbent Worker Demonstration Program.
Federal Register, v. 63, no. 240, December 15, 1998, p. 69103-69116. Information on
awards is available at: [http://www.wdsc.org/sga/awards/99-002award.htm].
71 See pages 14 and 18 for a discussion of this and other concepts that are used in this section

training because the former does not impart portable skills (i.e., skills that are useful
throughout the labor market). Put another way, firms are willing to share with their
employees in the costs of specific skills development because they are fairly sure of
benefitting from it: as specific skills raise productivity only at the training firm,
employees earn higher wages than they would be able to command elsewhere; specific
training consequently reduces employee turnover and allows the firm to recoup its
training expenses by paying wages that are less than the workers’ productivity in the
post-training period. In contrast, employers are unwilling to share in the costs of
general skills development: as general skills are equally useful to many companies,
the training firm must pay employees wages equal to their productivity in the post-
training period; if the training firm fails to give employees the entire return to training,
they will leave for firms that will pay them a wage commensurate with their
transferable skills. Thus, businesses do not have an incentive to pay for general
training while incumbent workers do.72
The human capital model predicts that employees will pay the full cost of general
training by borrowing money, earning a lower wage during training periods or
accepting a lower starting wage for jobs that offer general training. If these things
occur, there theoretically should not be underinvestment in training. However,
employees cannot readily borrow funds to pay for general training and employees
could be reluctant to self-finance training because they are uncertain of its rewards.73
So, if incumbent workers actually had to pay the full cost of general training,
underinvestment would be likely. Further, employers may not be able to pay workers
a lower wage conditioned on training receipt because of such things as minimum wage
legislation, collective bargaining agreements or competition from other firms.74
Indeed, there is empirical evidence which suggests that the starting wages of workers
who receive training are not depressed (or barely so) and that post-training wage75
growth is smaller than productivity improvements. In addition, there are abundant
examples of companies that do not reduce their employees’ wages while paying the
expenses associated with participation in literacy programs, problem-solving classes
or other general skills instruction. All of this implies that — contrary to standard76
training theory which is predicated on a perfectly competitive labor market —

71 (...continued)
of the report.
72 Becker, Human Capital: A Theoretical and Empirical Analysis.
73 Ritzen, Jozef M.M. Failure for General Training, and Remedies, in: Stern, David, and
M.M. Jozef Ritzen (eds.) Market Failure in Training? Germany, Springer-Vergal, 1991.
(Hereafter cited as Stern and Ritzen, Market Failure in Training?)
74 Bishop, What We Know About Employer-Provided Training.
75 Barron, Berger and Black, On-The-Job Training; and Bishop, John H. On-The-Job
Training of New Hires in Stern; and Ritzen, Market Failure in Training?.
76 In a perfectly competitive labor market, its numerous firms are wage-takers (i.e., they do
not have control over the employee’s wage). “The market” (i.e., firms collectively) determines
the wage, which is equal to the marginal product of labor. Alternatively, in a non-competitive
labor market characterized by one or a few firms, employers are not wage-takers. Rather than
the market determining the wage, these monopsonistic firms have power over the wage and

employers pay a large share of the costs and garner a large part of the returns to
training. 77
It has been suggested that imperfections in the labor market enable firms to reap
some benefit from and therefore to pay toward general training as they do toward
specific training. For example, a firm may be able to offer employees with
transferable skills a below-market wage in the post-training period and still retain
them because other businesses are unwilling to equally reward new-hires for skills
whose quantity and content they cannot accurately judge. The fact that employer-
provided skill development typically is not accredited permits training firms to gain
from this asymmetry of information in the labor market. Further, the termination of
an employment relationship is not costless either to the employee who must search for
another job or to the employer who must hire a replacement. The presence of search
and hiring costs thus may serve to temper employee turnover. Moreover, general
skills may complement specific skills so employers are willing to provide instruction78
in the former to enhance the value of the latter to the firm. In addition, each worker
may have a bundle of general skills — in effect, a firm-specific skill mix — that is
more useful to the current employer than to other employers and thereby reduces
worker mobility.79
But, if employees do not foot the complete bill for general skills development
and employers instead share in its cost, the likely outcome — given the possibility of
employee turnover — is underprovision of general training. When deciding how
much general training to offer incumbent workers, a profit-maximizing firm will
compare its current costs for employee training with its future productivity gains. The
training firm will not factor into its decision any benefits that might accrue to
subsequent employers, to the employees who change employers or to society-at-large.
But, as demonstrated by the research previously reviewed, employees who change
companies do benefit through higher wages from the training acquired while at
previous firms and similarly, companies do benefit through higher productivity from
those employees who received training while at previous firms. The presence of these
spillover effects, which training firms do not consider, means that employers likely err
on the side of offering too little general skills training.80

76 (...continued)
are able to set it at less than the marginal product of labor (i.e., at less than the wage that
would prevail in a competitive labor market).
77 Acemoglu, Daron, and Jorn-Steffen Pischke. Beyond Becker: Training in Imperfect Labor
Markets, The Economic Journal, v. 109, February 1999. (Hereafter cited as Acemoglu and
Pischke, Beyond Becker: Training in Imperfect Labor Markets.)
78 Acemoglu and Pischke, Beyond Becker: Training in Imperfect Labor Markets.
79 Bishop, What We Know About Employer-Provided Training.
80 An alternative way to determine whether employer-provided training is underprovided is to
compare its rate of return to that of other investments (e.g., in education or in physical capita).
There is a dearth of high-quality data from which to develop the rate of return to employer-
sponsored training, however.

This suggests that there is a role for the federal government to play in employer-
sponsored training. In light of the limited information available on what works and
what doesn’t and on how much total (formal and informal) training currently is being
provided to employees, the government may want to tread warily into this new venue
for its limited training resources. The possibility that firms could substitute federal
monies for their own resources or substitute formal for informal training rather than
increase the total quantity of incumbent worker training also implies the need for
caution. And, while the historically low unemployment rate in recent years may have
contributed to policymakers’ new focus on skill development for individuals who
already have jobs, incumbent workers are virtually guaranteed some degree of training
from their employers. In contrast, unemployed persons have only their own
presumably scanty resources and, traditionally, federal training funds to draw upon
to learn new or upgrade existing skills.
Up to this point, it appears that Congress has cautiously ventured into the area
of employer-provided training. In the case of WIA, the incumbent worker training
initiatives are small-scale demonstrations and the grant solicitations explicitly state
that federal funds should not to be used for training that employers otherwise are
capable of undertaking or that would have been provided in the absence of the grant.
While there is no matching requirement for WIA funds going toward incumbent
worker training, ACWIA does include such a requirement which mitigates the total
replacement of company resources with government resources on a given project.
Between the two funding sources, less than $80 million likely will have been81
committed to incumbent worker training through 2000. In addition, the requirement
that incumbent worker projects involving WIA or ACWIA funds are to be evaluated
hopefully will inform any further legislative proposals in this policy area (e.g., the
request for a new incumbent worker program in the FY2001 budget). The recency
of many of the awards involving employer-provided training makes it unlikely that
evaluations will become available shortly, however.82

81 Grants totaling $33.1 million were awarded in 1999 and 2000 under Title III of the Job
Training Partnership Act (WIA’s predecessor) and under ACWIA for the training of
incumbent or unemployed workers ($20.7 million and $12.4 million, respectively). Grant
solicitations issued in March 2000 are expected to result in awards totaling $47.2 million for
the training of incumbent or unemployed workers ($7.2 million and $40 million, respectively).
82 In April 2000 a request for proposals was issued to evaluate the program year 1998
incumbent worker demonstration program under Title III of the Job Training Partnership Act.