Beruflich Dokumente
Kultur Dokumente
DigitalCommons@ILR
CAHRS Working Paper Series Center for Advanced Human Resource Studies (CAHRS)
9-1-1987
Boudreau, John W., "Utility Analysis: A New Perspective on Human Resource Management Decision Making" (1987). CAHRS Working Paper Series. Paper 452. http://digitalcommons.ilr.cornell.edu/cahrswp/452
This Article is brought to you for free and open access by the Center for Advanced Human Resource Studies (CAHRS) at DigitalCommons@ILR. It has been accepted for inclusion in CAHRS Working Paper Series by an authorized administrator of DigitalCommons@ILR. For more information, please contact jdd10@cornell.edu. Please take our short DigitalCommons@ILR user survey.
DRAFT:
DO NOT QUOTE
OR CITE WITHOUT
AUTHOR'S
PERMISSION
UTILITY HUMAN
ANALYSIS:
ON
RESOURCE
MANAGEMENT
MAKING
Working
Paper
87-09
John W. Boudreau
This research was carried out with support from the U.S. Army Research Institute, contract SFRC #MDA903-87-K-0001. The view, opinions and/or findings contained in this chapter are those of the author and should not be construed as an Official Department of the Army policy, or decision. Chapter to appear in L. Dyer (ed.), Human Resource Management: Evolving Roles and Responsibilities, ASPA/BNA Handbook of Human Resource Management, Vol. 1 (Washington: Bureau of National Affairs, 1988).
This paper has not undergone formal review or approval of the faculty of the ILR School. It is intended to make the results of Center research, conferences, and projects available to others interested in human resource management in preliminary form to encourage discussion and suggestions.
DRAFT:
DO NOT QUOTE
OR CITE WITHOUT
AUTHOR'S
PERMISSION
utility
Analysis:
A New Perspective
Page 1
objectives?
in HRM training,
Since
(Milkovich same
costs
& Boudreau,
resources
being
managed
the
rationality
and care
equipment human
even possible
resources, and
organizations systematically?
simply
too ill-defined
unpredictable These
to be managed important
questions
resource most
manager, managers,
or a staff
professional. would
Though
readily of dollar
agree
that human
the most
in terms
expenditures regarded
is often with
little goals
systematic (especially
to organizational performance
to the financial
Indeed,
controversial (Gow,
to corporate
to merit
in a widely-read
professional regarding
journal
1985).
to imagine
the Finance,
or Engineering
competition, manage
from
the U.S.
competitive more
differently,
are paying
to their
HR managers year-to-year
it increasingly
to justify
utility
Analysis:
A New Perspective
Page 2
programs
solely else
because in the
they
are "good
for human
relations," in Japan
"being
done
industry," is being
or "done required
resource
organizations. order to
can a Personnel
manager in an
overhead?"
and outplacement
the organization
to cut costs? programs produce potentially the costs lucrative of such programs decision lost program areas consider
resource
management budgets
visible,
remain without
unmeasured, considering
benefits. (e.g.,
in other
management often
marketing, costs,
accounting, estimates
operations) benefits
of program
as well.
It is
to
to imagine
an engineer process,
proposing without
to spend explaining
a million
dollars
million-dollar of their
consideration
on product
or cost. costs
the decision
to train
employees
can easily
incur
development, often
instructor
time,
go unmeasured. returns,
Lacking
a systematic might
program's
such a million-dollar
HRM program
utility
Analysis:
A New Perspective
Page 3
in otber
management
cost-benefit
explicitly
considers
and benefits
analysis
provides but
formulas it is more
of human
facts,
assumptions
decisions
It
human
or uncertain, it helps
into useful
decision
and
determine utility
information models
is or is not needed. and detailed, It presents have been this several developed
analysis their
can be complex
emphasize
managerial
studies,
illustrating progressively
realistic
it will
algebraic formulas
interested
the findings
discusses systems.
of cost-benefit discusses
general
issues
to applying section
to human applies
resource
decisions. affecting
shows
to programs program
stock
a training models
section using
flows, final
section
and conclusions. Analysis concerned for Human with Resource Management resources Decisions? in HRM programs
how to invest
utility
Analysis:
A New Perspective
Page 4
have
several
types
available (utility)
to support analysis
those in.
decisions.
where
fits
Decision
Costs. Organizations
decisions in every reducing based human on costs, resource
make
their
human
resource them.
attempting
Costs costs
management
so reducing activities.
the number
or scope
of human
Other
typical
techniques
involve
reducing 1987,
costly 1985).
(Cascio,
is only
Reducing but
employee
behaviors
it may
reduce
activities enhance
behaviors
at
may
as occurs
creates
opportunities decision
employees,
but a cost-focused
system
fail might
an explicit
the sheer
a recruitment a selection
program
while use
be assessed it will
be vulnerable
Employment
Opportunity
office.
Lacking
the analysis
personal
influence
or political
organizational Count
of ratios
designed
to analyze
or staff
For
utility
Analysis:
A New Perspective
Page 5
one might
compute
the ratio
of total
employees
or the ratio
of sales
1984).
ratios While
compared be useful
for directing
or costs, If your
than
(or higher
a competitor's), high
what
decision
suggest? ratios
Is it always
or rising
HR staff-toand
Clearly,
provided
for such cost or staff there count is less cause ratios must
within
framework
to be useful approach
Audits.
activities systematic programs limited
(Mahler, and
Sheibar,
audits
detailed,
whether However,
implemented because
whether
appropriate
place. was
consolation
implemented
as planned
and incurred
or that system
A decision
before framework
as well
as providing
studies. studies
approach
involves
conducting resource
of human
include
validation
studies
the statistical
utility
Analysis:
A New
Perspective
Page 6
ratings, levels and experiments between groups alone concerned limited comparing receiving not very
between the
test
scores
statistical
difference programs.
training
However, who
resource
managers,
exploring to
determining benefit.
resources study
a scientific
let alone
report
results
in terms The
cost-benefit
analysis.
cost-benefit
analysis.
The organization
to be considered (such
as dollars),
the benefits
and subtract
to determine
of each
The best
or those standard,
(or returns
exceeding of course,
precisely
and with
certainty.
everything deal
costly. it could
still,
of merit,
be efficiently
Advantages
of a Cost-Benefit approach
A cost-benefit approaches:
offers
1.
Identifying
costs more
beliefs
visible thus
questioned
and corrected,
the chance
incorrect
or counter-productive
information
go undetected.
2.
Consistency.
Human
resource
decisions
are complex.
Without
utility
Analysis:
A New perspective
Page 7
and strategic
planners
to
on a different based
set.
on low cost
(Rynes based
1986),
defensibility to both
these
factors reflects
are relevant
decisions,
of these
factors
of those
hired.
system
encourages
by providing basing
a list
to be explicitly
considered
before
a decision
a few of them. A cost-benefit to many system decisions. promotes Once efficiency developed, because the for each
3.
Efficiency.
re-inventing is unique
each
require
effort,
organizing
commonlyresources to
system
Moreover,
a cost-benefit systems.
approach
to set priorities
information,
as we shall
examples. communication "language" factors and their the data and staff gathering for
4.
it offers
a common
decision
making.
the system
for applying
allows
to coordinate.
It is not necessary
utility
Analysis:
A New Perspective
Page 8
task
every
time,
so decision rather
makers than
more
resources
to making
decisions
identifying organizations
locating (at to
Moreover,
because
other
is improved
dollars.
by cost-benefit HR decision
is expressed
in
When
their
terms
by other
management
Marketing,
Accounting
and Operations),
communication Systems
may be improved.
Decision
in Human
The principal the costs be measured affecting value really approach forecast that that what
drawback
of a cost-benefit be measured
system
is that some
all
cannot
precisely
cannot
to quantify activities
a single But,
system
used
management
the stock
market
or predict
Finance
typically
express
terms. factors,
Their
decisions
on the
isolating
or uncertain risk
and systematically
and explicitly
potential
only human
on costs, resource
adopting
of tradition,
benefits
utility
Analysis:
A New Perspective
Page 9
can have
undesirable
effects.
decision
system
a way out of this bind. a set of variables but about way. it also human It does
situation,
proposing
to describe
resource a great
efficiently
summarizes
resource
programs
measuring
information,
necessary Analysis
utility
human
programs
"utility" explain
decisions.
simply
usefulness,
and improve
the usefulness
Requirements Utility
Analysis like other is desired decision and what systems, requires: (1) A problem, (2)
is currently
achieved;
to address describe
the problem;
(such
as effects function,
(4) a utility
used
to combine usefulness.
an overall
judgment
of each
alternative's
is generally
employed
when utility
are met.
Given a
and a value,
attributes
usefulness
utility
Analysis:
A New Perspective
Page 10
The Unit
of Analysis: analysis
Human models
Utility programs.
on decisions (sometimes
A human
resource
an "intervention" is simply
by industrial
psychologists,
or procedures tests,
selection
training
courses,
techniques, provide
compensation a vital
plans,
Decisions
about
such programs
resource resource
by human
specific issues
human
strategies areas
address
as staffing
to emphasize,
strategy
implementation
a strategy
intended among
to increase competing
manufacturing in selection,
requires
choices
programs
and job design. decisions about human about resource individual programs encompass (such
decisions train,
employees Each
promote
or reward). resource
of these
of human to hire
For
a framework recruitment
or promote (such
of programs
generating
staffing
information
(such
as
inventories). Program resource Useful? programs Quantity, Quality and Cost effects,
a Human
Resource human
Decisions subject
about
to the scrutiny
of many
constituents.
system
utility
Analysis:
A New Perspective
Page 11
effects,
but utility
analysis
focuses
on three value
and cost.
have
behaviors reductions)
over
improvements work
(or avoid
of those develop,
behaviors,
and when
they minimize
required
implement three
the programs. to those typically used in other Marketing, value (such to the
are similar
functions,
and this
similarity
Operations' large
programs
quantities
of productivity
as sales as or
or units reduced
produced), costs
quality
(such
cost
investment.
models
chapter
reflect
resources in programs
in human
resource
management areas.
and investing
resources
from other
management
quantity,
quality
and cost,
it is important different
to recognize of
constituents
with
aspects
factors.
managers
and operating
expect
to be promoted with
shortly),
financial
may be concerned
statements,
flexibility these
Utility implies
analysis
different program
one
a different analysis
Utility programs
identify
the mechanisms
through
which
HR
affect
the organization.
For example,
organizations
facing
Utility
Analysis:
A N@w Perspective
Page 12
increasing output,
product
demand
employee
productivity apply
to increase
while
costs.
each of these
the analyst
to reflect
effect.
outcomes have
such outcomes
analysis
clearly
omit
factors.
They
outcomes,
other
potentially
as employee
attitudes,
relationships,
relations, decision
They
of HR decision decisions about
makers.
any more whether
resource
addresses utility
models
a great
to these
if We Can't Though
Measure
in utility
analysis
This
is (or cannot and
incorrect
imPression cannot
analysis
precisely
can be measured
are often
uncertain some
to change that
Finally, very
variables measure.
expensive
to
Utility
Analysis:
A New Perspective
Page 13
These limitations resource inability analysis and costs are ways Chapters to correct
and
not prevent
programs
about
or the
to precisely of financial
measure
preferences
prevents
strategies.
Simply
put,
decisions
that would
without
benefits;
and
not outweigh
the expected
benefit
decisions. words, information resource gathering program is itself effects an investment decision. makers the
In other Uncertainty
about
human
should
systematic
analysis, how
decisions, from
information. only
differs
information,
decisions
Subsequent resource
sections decisions
illustrate systematic
analysis
models
in the face of uncertainty. Resource above Management Programs and Effects human resource resource programs; program affects
of Human noted
(costs
and benefits;
and cost)
depending
the program
Employee
stocks.
Programs
affecting
employee
stocks
(such
as training,
utility
Analysis:
A New Perspective
Page 14
performance
feedback (such
and employee
involvement)
as skills, current
abilities
or motivation)
their
In terms
decisions
affecting large
stocks
(1) Affect
of employees of employee
(2) Cause
average
increases
Thus,
decisions in their
behaviors
assignments. flows. Employee flows occur when employees move into, through
Employee
and out of an organization demotion affecting large and transfer. employee flows
through
selection,
promotion, decisions
of quantity, productivity
when
numbers
of employee
increases and
in the value
research
utility to other
decisions, flows
employee
(such
or incentives Programs
employees flows
to apply
affecting into,
"work"
by improving
of movements
through
so that
are placed
Employee
models
feedback
in press; Schmidt,
1987;
Hunter
& Pearlman,
also be applied
to other
programs
affecting
employee
stocks
Utility
Analysis:
A New Perspective
Pag@ 15
as compensation
and employee
involvement).
To illustrate
utility
analysis
Though
application
example
on an actual
in 1986. Situation involved a choice between delivering system, staff training for
The Decision
a traditional While
classroom
or through
a sophisticated
network.
some training
believed
investment,
1 describes
in terms
quantity,
quality
----------------------------------------
program
would
cost
$451,035 network
over
five cost
staffing
would
(with the
in up-front
of the
would
it. for training added was the 200 currentlyyear for the next training period. thus could
population
engineers,
every
years. only
constraints,
Classroom
per year,
or 200 over
the 5-year
target per
Audio-Video
could
accommodate
up to 200 persons
year,
Utility
Analysis:
A New Perspective
Pag~ 16
fully
train
the
incumbent
work
force
year,
accommodate
20 new hires
future
year,
five years. its typical trainee cost practice, the organization costs by the
Classroom
training
training
cost
$3,683
trainee
figures
suggested
training
much
per-trainee Discussions
effective.
whether that
no definite dollars
conclusions.
spending
over
one not
on a training
delivery Moreover,
could
and complex
Computation.
on its production
How much
productivity
be affected
of trained
engineers period
1 (because
tenure
averaged delivery
no turnover.) (e.g.,
Because
Audio-Video
are productive
it affects
person-years trains
etc.).
Classroom
training
at a time,
so the work
force
doesn't
until
utility
Analysis:
A New Perspective
Page 17
even
Classroom
training This
affects
600 person-years
of productivity and
40 plus
leverage
demonstrates Leverage
faster
increase programs
occurs
because
human
who affect
productivity little
Typical
to help
us estimate
quality,
and certainly
of improved
performance. productivity
To explicitly increase
sYmbolize
average
per-trainee,
P1 for Option
1 (Classroom
Training)
2 (Audio-
Training). Payoff formulas. quality, Even without the cost knowing the effects information in Table u1) goes 1. of either proved program useful
on employee
quite
in constructing (usefulness) one-dollar the $451,035 u2) goes quality functions to make value
the payoff
formulas training
up by $600 with
in average Similarly,
employee
quality
offsetting (i.e.,
the utility
of Audio-Video increase
up by $1,200
employee payoff
that quite
modest
effects
might effects
be sufficient on employee
worthwhile quite
training
would
produce
returns
an average total
productivity
increase from
utility
of $148,965 This
training
and
$168,853
training. training,
represents
Utility
Analysis:
A New Perspective
PagE! 18
Video
training.
At higher
average
productivity
increases,
the relative
advantage
training Of course,
little
information
was available
increase However,
value embark
this variable,
we divided
by the
to equal 1.
(or "break-even"
The Classroom
training (i.e.,
costs
be covered while
$451,035/600), was
training that
program
these
values
are much
earlier.
Relatively appeared
modest
training large
effects necessary
had originally
to be a very
investment. Program it treated the question training break-even comparisons. each training While option the break-even independently, the more analysis was enlightening, did not address
of whether
to substitute
However, What
Classroom
the total
utility
from
producing as shown
reflecting
utility
Analysis:
A New
Per6p~ctive
Page 19
U1)
to zero,
we obtained
a formula
be
necessary Classroom
to make program
to the P2
a certain implied
for P1 [i.e.,
(.5
The decision
rules 1.
by this break-even
at the bottom
of Table
to $2,000, only
exceed
$1,484
(.5 X 2,000)
Audio-Video increases
training. greater
If Classroom than
training
productivity
(i.e.,
to invest average
in the faster
Audio-Video
productivity formula
effect
than
The break-even
provides investment.
equation
each program
simplified 8).
personal
computers
Chapter without
measuring
effectiveness, to better
focused
the decision
issues. effects
Instead
enough
to it.
done by these
undisputed Video
expenditure could
be justified.
programs
investment.
The
demonstrated
that a costly
effectiveness analysis)
(earlier
thought
to be essential
for applying
cost-benefit
utility
Analysis:
A New Perspective
Page 20
to improve
to training affecting
programs. employee
Similar
stocks,
applications
are possible
programs
and employee
involvement.
Moreover,
cost-benefit
analysis
uncertainty Break-even
effects.
Uncertainty
analysis its
of addressing
explicitly, Rich
and reducing
effects several
on decision other
methods
of addressing
affecting
Employee that
Flows employee
affect
is useful, resources
employee
to improve
selection
tests,
management, expense?
and internal
staffing
offer
sufficient
As noted
earlier,
employee
flows
occur (e.g.,
when
people
into,
through
and
positions/jobs
selection, such
turnover, flows
promotions, as affect
affect
as well that
stock
However, decisions
decisions affecting
Whereas
employee positions,
of employees work
in their
current
decisions will
by affecting
which
individuals
occupy
We can consider
general
processes
through
which
individuals
Utility
Analysis:
A New Perspective
Pag@ 21
flow 1988,
through Ch.
the work
force
(Boudreau
& Berger,
1985b;
Milkovich which
& Boudreau,
10-13): a pool
(1) External
involves will be
attracting hired
of job applicants
of them
separation/retention, of employee
which that
involves affect
separations and
employees involves
(3) Internal
staffing, movements
which between
positions
the organization.
Though
managed,
evaluated closely
independent, each of
quite
managing
quality
and cost
for employment
candidates leaves
will
employment
vacancies,
affect
stays
---------------------------------------The following employee first adding flows. section The models solely develops proceed utility from models for decisions to more with complex. affecting
simpler
The
models flow
models
focus
on external
selection,
enhancements Table
to reflect 2 provides
the other
processes. added
of the decision
addressed
focuses
of external
job
#2 incorporates
useful
in making
typically
applied
Model
to reflect
the effects
utility
Analysis:
A New
perspective
Page 22
applying
selection
programs
over
time.
Model
staffing
utility
of employee
separations
turnover,
utility employee
Table
study
explore
utility this
models study
flows
using based
a case
study.
case
information readers
on published it useful
However,
may
find
to produce point
illustrations
to them.
themselves, analysis,
systems
will be used
addressed. behind
though
complex,
analysis
methods burden.
Boudreau,
reduce
the computational
---------------------------------------Consider programmers, 3 contains from a large and 1,000 organization data system employing managers 4,404 one entry-level above computer
level
them.
Table
a description Hunter,
(adapted 1987b).
Schmidt,
McKenzie
Utility
Analysis:
A New Perspective
Pi5ge 23
leave
the entry-level
job each
year
and are
experiences
100 separations
manager
job, and has a promotionfrom the entry-level 100 new hires by these to 718. to make how jobs. provide
100 employees
are promoted
programmer
in this
are called
upon
are selected,
how they
turnover
be managed, whether
determine
of these
returns
to justify could
organizational
investment
be justified?
resources
be allocated
activities
such
external
selection
analysis
addresses
issues
1.
the current
test? 2.
Should quality the recruitment applicants? the pattern of employee performers, separations and how much be changed would to retain program be changed to attract higher-
3.
Should more
of the good
such a change
be worth?
4.
Should system
center
be used would
to promote such
programmers
to
a program
be worth?
Analysis question
The first
strategic
utility
Analysis:
A New Perspective
Page 24
employees
job:
"Should Selection
employment
interview
be augmented
test?"
development
can be an expensive may be called to address Preliminary utility was only the early to develop of these Cascio,
process, Several
resource been
to justify
proposed
this question. Attempts analysis that to Develop models they Selection Utility enjoy Models a long history, However, but it
dollar
values.
models
shows
why researchers
were
analysis
models
detailed
models Hunter
Boudreau,
& Boudreau,
index
of selection
the correlation
is a statistical device
scores
on a selection
levels. indicating
coefficient
relationship
and higher
values
positive one
linear
relationships. value,
aspect
of a test's
affected, and
as dollars), usefulness
ignores
index
originally
"success
ratio,"
or the probability
on a predictor Under
would certain
turn out to be at least assumptions, by setting and when the success higher only passing about
on the job.
is improved
by higher
validity being
coefficients, more
on the predictor
(e.g.,
selective),
utility
Analysis:
A New Perspective
Page 25
half
pool
would
be successful
predictor. validity,
ratio
incorporates
as a cost-benefit
the quantity
affected,
costs.
achieve Model
Moreover, minimally
distinguishes
only
performance
who do not.
#1, The One-Cohort These early utility tangible models models units
External
Selection
the need
for
selection in more
the value
investments job
dollars)
and reflect
performance
doesn't
simply
study
utility
analysis
address
issues. works. The first utility hired model analysis model focuses using
(or "cohort")
of employees utility
with
(B-C-G) 1950;
(Brogden, 1965).
1946b,
Cronbach
& Gleser,
of person-years
of productivity group
affected
by a selection average
multiplied system
by their
dollar-valued with
per-person,
per-year
and those
who would
be selected
without costs
program
is equal
and to apply
it to one
of job applicants. Quality is the product between of three factors: (1) the validity coefficient
the validity
coefficient
of the proposed
utility
Analysis:
A New Perspective
Page 26
system
versus
without test
prediction selected
standard
(Naylor
1965);
(3) the
improvement" Value
in new-hire (usually
value.
as SDy),
the dollar
applicants other
average
performance
is one
deviation
than
applicants. SDy measurement become a controversial This controversy. topic The SD of Applicant psychology, because Value (SDy) has some based and
variable
coefficient
both
the selection
of job behaviors
predicted
in standard
units.
validity
computation
above,
standardized
test
equals
in standard
units
between it.
the device
and those
without
To
the standard
deviation
differences represents
dollars
are multiplied
the dollar
value
difference expect
In general, is greatly
we would affected
(e.g., jobs
discretion
out their
decisions), when
of those
decisions
are important
employees
handle
and when
of job-relevant
characteristics
in the applicant
utility
Analysis:
A New
Perspective
Page 27
as when graduates
job contains
recent
college
Most
research Schmidt,
Some
have measured
them
to estimate
a person
is better who
is better
and should
(e.g.,
productivity, adopted
salary methods
among based
have
detailed
(e.g.,
Cascio
1986).
reviewed techniques
measurement indeed
research,
produce
different
SDy values,
none offered
a convincing
case
for greater
accuracy
or validity. between
Considering existing
in measuring it
As we shall analysis
this
----------------------------------------
Utility
Analysis:
A New Perspective
PagQ 28
that
is the the
and pencil
Aptitude needed
4 shows model.
to apply
The
focuses
programmers. each
of testing
applicant
(i.e.,
(1,236).
of employees
by multiplying to produce
tenure
5,988
person-years. in average
standardized
selection times
.80 with
a 50% selection
ratio,
the Naylor-
tables) times
coefficient
(.76, estimated
the dollar
of a one-standard-deviation by Schmidt,
(calculated
$6,331
analysis, by using
we assume
that
in validity
it (i.e.,
not using
in essentially to reflect
selection). between
Of course,
be modified validity.
two selection
different (i.e.,
computation. (i.e.,
Multiplying 5,988
quality
$6,331
per personbenefit
by quantity
person-years)
produces
testing than
The $37.9
million
divided
by the $12,360
represents
a return
of 306,634
percent!
Not surprisingly,
utility
Analysis:
A New Perspective
Page 29
these because
results
caught
much
uncertainty.
values
certainly
prediction represent
value
and thus
sources
and error.
analysis
of Table
uncertainty
should
decision
models.
We can
the and
break-even
analysis
example
as well.
Considering expect
surrounding about
measures
of SDy, we might
uncertainty function
its value.
We can re-write
the payoff
leaving
parameter
unknown:
utility
Dividing exactly is $3.39 justify $12,360 equal
(3,641 X SDy)
gives the value utility
$12,360
for SDy that would equal make
(1)
benefits
by 3,641
costs
(and total
exactly
to zero).
value
the amount
the quality
fewer
programmers.
might
of $10,413
in value
be less than
shows only
selection to that
is not uncertain,
can address
Suppose
by .76, test
selection
our
would
an average
standardized
Utility
Analysis:
A New Perspective
Page 30
score
of
.80 (which
assumes
will
is without 1985;
& Kehoe
Pearlman
& Hirsh,
We can
effect
on quality (i.e.,
as QUAL, 5,988
formula
$12,360): (2)
of quantity
person-years)
and cost
Utility
The value costs down for QUAL
(5,988 X
at which
QUAL)
$12,360
benefits break even with
is $2.07 to whether
$12,360/5,988). of selecting
improve by an
selection average
decisions of at least
enough $2.07
of those
selected
consider
that
we felt
they might
as $1,000
implying
a total
testing this
cost of $1,236,000.
Equation
2 to reflect
utili ty
(5,988 X QUAL)
for QUAL is increased Thus, even
(3)
to $207 per person100 times higher per person-
value
expected,
be offset because
by rather
selection (i.e.,
effects 5,988
person-year. years)
Model
leverage
affected
#2:
by this
selection
Financial
One-Cohort
Selection
Utility framework
analysis
provides
a systematic
on selection is that
costs
and benefits.
One consequence
of such
it draws
attention
implications functional
of human areas
investments.
Line managers
in other
Utility
Analysis:
A New Perspective
PagE! 31
as Finance,
Accounting considered
frequently
not usually
resource value?
managers. Wouldn't
implications
of increased higher
require to value
costs
to attract increases
them?
Is it
equal to
productivity obtained
increases
resource cost-
unprepared numbers
to address
find
that
their
conclusions model
(Boudreau, addresses
1983a; these
financial model
Financial
analysis
typically
adjusts
of revenue
and costs
increases often
enhance increased
of goods costs"
require
to maintain
those
in addition increased
costs
of the productivity increased received benefits inventories and costs and costs,
(e.g., sales
require
commissions);
incurred because to
are worth
less than
received interest
incurred) received
today later
earlier rate
those
"discount subject
rate"); to income
and
in organizational to pay
taxes
profit External
state utility
or Local Model
of program
and cost
to reflect
utility
Analysis:
A New
Perspective
Page 32
situation. values
Table
5 applies
these additional
derived
earlier. service
Necessary costs
includes
the proportion
of variable
tax rate
the yearly
on investments Applying
1983a as shown
an adjusted
of $12.55 discounted
after-tax,
of
of $6,798 an after-tax
of $12.55
million
smaller
than
of Model to those
Dealing
Equation
2 to reflect
the
financial/economic
5 as follows:
Utility
Where 1,983 reflects
(1,983 x QUAL)
not only the quantity
$6,798
of person-years the discount the $6,798 affected rate by
(4)
testing,
tax rate.
and the
formula
those
programmers
selected
(i.e.,
$6,798/1,983)
This value
is higher that
value
2, because offset
the benefits
of improved taxes,
testing service
not only
costs,
but also
increased
utility
Analysis:
A New Perspective
Page 33
and
interest. Model #2 can be used though to explore these uncertainty in the financial/economic estimated (c.f., by the Rich
factors
themselves,
variables
are typically
financial 1987).
experts,
certain
The point
analysis
allows
uncertainty
addressed
within
framework these
the non-financial
model.
same
be applied
to utility
analyses
affecting
employee
discussed #3:
earlier
(see Boudreau,
Financial Models
External only
Utility
Model
example,
large,
evaluating
group
a new programs
production
on its first
Selection
for several
years. to spend
as we continue
on this costs
testing
as programs
is that
leverage
of selection Model
as selection these
in Table How
to address model
works.
is re-applied applicants
Thus,
of testing once
future
in the one-cohort
newly-hired
cohort
to stay
of time periods
and then
leave.
Thus,
utility
Analysis:
A N@w P@rsp@ctiv@
Page 34
program
is re-applied,
the work
force
contains
larger
employees.
assuming
no one
years
of a selection
better2, and
is joined
by a second group
better-selected in Year 3.
in Year
a third
better-selected model
utility
considers
both years
the increased
selection employees
program
for several
derived
Moreover,
it continues
to reflect
the financial
considerations
earlier.
---------------------------------------Application applying Model to the case situation. case. Table 6 shows force the results size, number tenure, the and those times .76 of of test
The work
testing
number
leaving,
variables still
the PAT
costs
$10 per
to average
$6,331 than
.80) more
per person-year
randomly. cohort
We assume
for seven
that
stays
ten years,
the testing
program
for ten
years.
In each of the first added Because to the work force, stays seven years, 618 better-selected selected new hires without are
replacing
618 employees
the PAT.
each cohort
the number
programmers selected
in the work
force
increases. These
programmers
after
one year.
are joined
by another
utility
Analysis:
A New Perspective
Pag@ 35
year
of 1,236. (i.e.,
In Year 618)
3 there
are
1,854,
and so
Year
7, when
7 times
selected
using
the PAT,
analysis,
of productivity as noted
(i.e., 6.
and
in Table
to the one
in Table leverage
programs.
Moreover, than
the selection of
over
is far greater
the number
selected
(i.e.,
computation.
calculated
by computing summing
for each
and then
adjusting
for taxes, in
and interest
rates.
values
to those
5 illustrates cohorts
subsequent
to a discounted however,
after-tax
$40,000.
benefits, discounted
million effects
years).
cost
million
benefit
is $54.28
after-cost
discounted
randomly). in Table
is substantially because
greater
reported
3, primarily
leverage decision
the selection
program. figure,
compute
a leverage
let alone
its effects
on program
costs
and benefits.
Utility
Analysis:
A New Perspective
Page 36
Dealing with
with
uncertainty. Again,
Once
again, was
break-even some
analysis
uncertainty.
suppose
there
disagreement
uncertainty the PAT. However, likelihood the value programmers between QUAL)
value
produced
it would
one might
question
the assumed
validity
increase
of hiring of those
selected
over
to the work
force.
total
per person-year
as follows:
Utility
where 8,580 equals
(8,580
the total increase
QUAL)
(i.e., $6,331 as well this
$40,000
$54.32 million) divided by
(6)
benefit (i.e.,
force
value
31,282
formula
$4.66
$40,000/8,580) costs.
the $40,000
discounted, value
Once
utility
uncertain, small
break-even to render
suggests
that
enough
investment in new-hire
focuses details
debate behind
on the a particular
to be made, value.
rather
Moreover, computer
because
personal
programs
spreadsheets
of software
can greatly
reduce
computational
requirements.
utility
Analysis:
A New Perspective
Page 37
thus
far has
assumed
that
the
of
be considered programs
selection
operate pool
generated
A stringent if applied
selection
investments colleges),
in improved as well
as recruiting selection.
in better changes
remain pool.
in selection
For example,
polygraph
or drug
reduce
force
quality
applicants When
find
intrusive
or insulting, strategies
actually When
value
through
on applicant
pools?
in improved
recruitment be #4 (see
effective?
integrated Table
to create
combined
2) addresses
questions
External
Works
The quantity model reflects of person-years in the recruitment-selection utility force
the same
Utility selection,
(Model
also
programs
applied
year
Finally,
the quality
of new hires
is calculated
the average
of applicants
in the pool
(the value
that would. be
Utility
Analysis:
A New Perspective
Page 38
produced value
group
plus
added
non-randomly
applicant utility
Whereas
discussed
reflected only
in the three
selection in work
models between
force
value
to randomly, work
value
calculated hired
of those period.
during
the analysis
This
the place.
employees
from
among
applicant
of that
applicant
is in the first
---------------------------------------Applying The be changed the Model second to the Case Situation is: "Should the recruitment Table program the
strategic
question
to attract
applicants?" to estimate
7 applies
the returns
from using
advertising to produce average audience assumed a higher applicants Table recruitment testing
diverse
qualifications advertising
applicant
because
recruitment
little
agency
qualifications is expected
value, them
because
to screen
of new hires,
of applicants,
costs,
average
score,
and financial
considerations).
utility
utility
Analysis:
A New Perspective
Page 39
assuming
the
same
10-year
application
of the staffing
program
$2,500
per hire
(American
Association, an is
Recruitment
advertising one,
to produce
applicant $10,413
pool
similar
to the present
Because
is expected among
to screen them,
applicants reducing
carefully, to
quality.
this
recruited
applicants
to have
costs
(including
to attract
of $40,000
a net applicant
value
applicants
service
applicants
to be $52,065
applicants of $36,445
due to the lack of pre-screening), per person-year, (somewhat lower for a net applicant than
is offset
costs
value
per person-year
.
for agency-recruited
utility
Computation value of each new hire randomly pool, plus is the sum of two values: (i.e., hiring average-value value produced
(l)
incremental
selection
that pool.
the average
value
of those
utility
Analysis:
A New Perspective
Page 40
hired
through
advertising
value
generated plus
applicants
(i.e., value
or $15,620
selection
(i.e., $21,951
.80 times
totalling
the agency-generated
to the average
-
of the agency-generated
(i.e., value
$60,000 produced
.80 times
$8,500,
per person-year),
per person-year. Once precisely quality years, these quality levels 6. are computed, That we compute total utility new-hire
as before
in Table
average
of new hires
in the work
analysis
period,
subtracting
computations software
accomplished to reflect
a personal
spreadsheet
programmed
relationships. are shown (i.e., in Table hiring 7. The after-tax, of average million, Value discounted value) with Value the
Selection
applicants
from
applicant
pool
a recruiting Selection
of Random
(i.e.,
applicants is higher
value)
million)
million.
The value
by testing $54.32
(i.e.,
minus
$40,000
However,
testing
the agency-generated
utility
Analysis:
A New Perspective
Page 41
pool
adds
(because of $35.00
they
pre-screened), testing
incremental
million,
the same
of $40,000. Considering only selection strategy, utility, we can see that a smaller the PAT pays payoff applied off
under
either
recruiting
but achieves
than $54.28
applicants
model higher
the advertising-generated $141.04 million). the advantage When of agency million the
(i.e.,
million
versus
are integrated,
recruiting
is clear. value
Selecting of $207.45
a total
$172.4
advertising (i.e.,
recruits, $54.28
which
produces plus
a total
million
million
$136.48
million).
Sacrificing applicant
effectiveness
in average
quality
the cost-benefit
implications model
decisions, lead
identifies
that
this utility
model
be used
implications 1985).
of the recruitment-selection
& Rynes,
analysis was
offers
a method that
of addressing
Suppose
there
agency-recruited value
really minus
average $52,065)
of $7,935 than
service
those
recruited
through
Agency
utility
Analysis:
A New Perspective
Page 42
recruiting attract recruited you don't by each determine applicants and salary the agency million test's $35.00
costs
more,
agency
recruits
require
higher
service
costs
to
and retain
them,
when
applied
to agencywhen generated
Isn't
to an agency
difference
in value
method? higher
have
to be in order
effectiveness. recruitment
Recruiting costs
(i.e.,
$8.10
incremental million),
value a total
million effect
million
of Random
million
with of
net applicant
of $20,000 million
Selection when
rose by $39.46
$141.04 $20,000
the average
value
rose by $4,380
$15,620)
variables
affecting
the Value
of Random
of Random
in the applicant
pool
Selection
= (9,009 x
in the Value in net value formula
Net
(7)
the change
of Random
divided
per person-year
this
in average Selection
to increase
the Value
of Random
utility
Analysis:
A New Perspective
Page 43
$22.87
million
(enough we obtain
to offset $2,538
the negative
recruitment),
9 , 009 ) .
$15,620 agency
$2,538)
recruiting rise
applicants must
to $40,000 $58,158
be at least
(i.e.,
$18,158)
per person-year
to offset
the negative
effects applicants
agency-recruited
service
of $60,000
producing
recruitment
utility
underscores to be
the
of cost-benefit and
calculations
uncertainty 7 suggest
Moreover, an integrated
the
effects
to recruitment human
programs
may produce
substantially
valuable utility
resource
Analysis
Integrating
Recruitment,
Selection
and
program
decisions.
However,
decisions by employee
selection
programs
affect
selection
produces
sooner
less-qualified
applicants, of improved
nullify
the advantages
should
invest
in programs
performers,
compensation
or improved
benefits?
What
Utility
Analysis:
A N@w Perspective
Page 44
is an "optimum"
level
and how
by
performers?
not just
of employee quits,
acquisitions,
separations
well.
How
#5 (see Table
2) provides
a framework
the Financial
Recruitment,
Selection
each
hired
simply
stays This
for a number
of
leaves. but
is more
on only
cohort,
it is still Hired
of actual
separation
patterns. leave
employees
together. old,
Rather,
good or bad
include
new,
employees. which
and programs
affect
in turn
the value
force. designed
to manage
directly
(Boudreau
on work focuses
is lost,
Consider work
before with
separations a certain
pre-separation value
as a pool When
employees
a subset valuable
of the employees
work
is lower
and vice
we can analyze
utility
Analysis:
A New Perspective
Page 45
affect
work
force
quality
by calculating work
per person-year)
time
period
different
assumptions
is retained.
The
of employee (e.g.,
includes
or severance
as well
designed
to affect
the pattern
of separations affected by
(e.g.,
The quantity
of person-years
is the number
involved valuable
work
value,
and vice
with
we simply
consider
of employees
added
to the work
force,
time period,
of employees of employees
retained added
average
quality,
plus
times
their
average
quality,
minus
of acquiring
(recruitment/selection (separation/retention
costs) costs).
releasing utility
The
values
in each rate.
time period,
adjusting
for taxes,
variable
In concept,
is similar
an applicant
pool
Separation/retention certain retention individuals is quite decides are apparent who leaves
an incumbent
decisions,
where
the organization
Utility
Analysis:
A New Perspective
Page 46
stays.
The concept
also
applies,
although
perhaps of quits
employees those
do the choosing
(e.g.,
who remain
compensation extended
tenure)
intended
to manage
---------------------------------------The third separations would Cohort Model such strategic question to retain asks: more Table "Should the pattern performers, of employee and how much MultipleUtility application of
be changed a change
be worth?"
the Financial
External (Model
Recruitment
Selection
#5 in Table in detail
2) to address
in Boudreau
& Berger,
The number
and separations
is still
4,404.
remain
8 assumes
recruitment
parameters
reflect 7.
advertising
in Table
validity cost,
testing cost,
service
applicant
service
and SDy
of 618 acquisitions
selection
per person-year)
Utility
Analysis:
A New Perspective
Page 47
selection
(i.e.,
$6,331
on a review
of costing
literature,
(reflecting
$2,500
and administrative administrative and severance pay). joining force service for
assistance, regardless
of the quality
of the person
we assume yearly
population yearly
average
is $52,065
and average
incumbent
service
cost
is $36,045
of $15,620 assumptions
per person-year). establish all the information pattern needed except whether the the four
of separations. keeps
determines Table
its better
performers.
selection those
where force;
average
(3) valid
selection value
retaining
retained
have work
yearly
of $2,707 while
than
selection yearly
retained
have
value
the pre-separation
work
force.
Computation of the four 8. selection and retention combinations required a LOTUS are shown
of Table
Though
to generate
are complex,
1-2-3
spreadsheet
reflecting model
(Boudreau,
Computerization
Utility
Analysis:
A New
Perspective
Page 48
calculation
effort,
Under
Option
(validity have
of zero)
a 10-year Under
after-tax, Option
of $200.31
million.
2, where remain
(validity force
value
increases
that
validity more
the best
it will
Finally,
Option
4 shows
that
highly-valid
if the worst
employees
(i.e., by $2,707
the effect
of separations producing
is to lower
value
a low 10-year
force
of $132.50
interaction
between
separation
patterns only
has
some
If decision expect
makers
a $41.79 selection
dollar
to random
$242.10
$200.31
million).
However,
if improved $109.59
selection could
is combined be realized
retention, million
an additional $242.10
million
minus
million).
token,
patterns
can disrupt
selection, with
illustrated worst
employees
force these
million
selection
effects
on a specific programs
of assumptions, affecting
suggest
that
resource
selection
and retention
substantial
organizational
Utility
Analysis:
~ New Perspective
Page 49
benefits. Dealing with Uncertainty in Table when 8 shows the best that the highest value from improved force million
is achieved
employees
Work
million
when
the separation
$351.69
separation separation
effect effect
is $2,707 is -$2,707
million force
changes This
by $109.6 suggests
million
for each
of $2,707 between
Effect.
separation Work where dollar work
relationship value:
Separation
effect force
and total
force
Value $109.6
40,488 x
million
Effect
(8)
Simply value work put, every
40,488
equals
divided between
increase
force
$40,488
in discounted, managers
after-tax, disagreed
the likely
effect
pattern. with
If the current
selection
respect
cause
employees likely
However, due to better force between enough the selection $242.10 random. force per
employees could
to leave
this
of the work
improved (i.e.,
an additional $200.31
million),
assuming
remains
Equation
necessary
to reduce
Work
by -$41.79
a break-even
Separation
Effect
of -$1,032
Utility
Analysis:
A New perspective
Page 50
(i.e.,
-$41.79 pattern
selection is greater
a separation
produce
a higher Thus,
work
random require
selection
retention.
Model
#5 does Effect.
not
In fact,
it enables by making
precisely
the
the relationships
recruitment,
and separation/retentions. Clearly, for selection with others it is not enough to appear valid, for recruitment to simply fill all vacancies,
in terms
of quantity, model
and cost
integrated patterns
utility
provides
a framework
can produce
substantial
force
value
Analysis
Integrating
Internal utility
integrates but
recruitment important
Decisions
affect
selection
in one
job often
affect
jobs within
the organization.
select
also make
them poor
performers on their
jobs?
is it better jobs
to select rather
employees their
based
than
qualifications
your best
technical
performers
into management,
Utility
Analysis:
A New Perspective
Page 51
work
force
value
by reducing value
valuable strong
or building and
organizational staffing
with
internal
linked.
candidates
is partially may
and involves
Downsizing
involve
redeployment jobs.
or rebalancing internal
by moving
employees of
between external
Moreover,
is important substantial
independent
staffing.
Organizations
time
to promotions,
demotions
even
if no external
#6 in Table that
2) presents affect
of decisions force,
the work
but how
they move
the Financial
Recruitment,
Selection,
employees
between
jobs within
the
includes Human
demotions has
(Boudreau
1985a).
planning
movements
between
(Milkovich
models
to predict 1984;
(e.g., 1983).
Milkovich
& Tsui,
Rosenbaum, between
Stewman
identifies jobs.
desired
in various
recognized resources
affect 1971;
a framework
consequences movements
employee
of
but on their
(Boudreau,
utility
Analysis:
A New Perspective
Page 52
movement pool"
is very
much
like
consists
of candidates in part by
be determined
current
Whereas
future
external
selection
might
use test
internal
selection center
decisions scores.
might
consider
seniority,
or assessment supply
employees,
of internal that
movements of
are
is determined layoffs
staffing staffing
decisions, decisions
of by emphasize
or quits.
in the receiving
with
qualified on
as we shall candidates
of internal
movements
supply
can be as serious
as employee
separations, job.
and may
not always
be offset model,
by improved
performance
of the work
force
is a function when
times
of
retained times
external
of employees
(i.e.,
times
the quality
internal
of external
internal
selection
activities.
establishes
that value
to reflect
of internal
and external
employee
movements.
The resulting
utility
Analysis:
A New Perspective
Page 53
force
values
analyzed
during
the
time
adjusted
for taxes,
rates,
and variable
costs.
---------------------------------------Application Question promotions to the Case Situation investing Table in an assessment center for internal
#4 asks whether
is cost-effective. Multiple-Cohort
9 presents
the Financial
Internal/External Model
and Separation/Retention 9, the external as before, separations, separations year analysis For into staffing that
to the current
example.
except
to replace to replace
100 promotions.
considerations
of internal
promotions
a Data
Manager
job containing
upper-level
employees,
exists,
separation cost
$8,000,
the $7,000
of those $8,000
retained.
Each
promotion
from Programmer
to Manager
(including
relocation, with
orientation
selection
is accomplished
an assessment
applicant
(Cascio
& Silbey,
producing
million
to assess
all 3,786
promotion
candidates. considered
the symmetry
the external
staffing
variables
Utility
Analysis:
A New perspective
Page 54
staffing
variables
considered
in of 3,786 place
available minus
in each year
after
separations applicants
618 equals
3,786).
With
100 job
choosy,
SD (standard
assumes
all Programmers
but one
easily are
consider
situations
a limited
eligible
or tested.
differences
consequences promotion
in the Manager
candidates
is $11,454 applicants).
(about
higher
than
the Manager
job involves
more
discretion value
service with
of Programmers
to Manager
random
promotions)
than
by ten percent
reflecting is increased
work
external
also
Thus,
benefit
decisions
produce
promotion also
jobs,
and vice
versa.
The
reflect
relationships
between
Programmer
performance.
Computation of Table 9 shows the effects shown of different represent internal and
patterns.
The values
the after-tax,
utility
Analysis:
A New Perspective
Page 55
discounted the
value
and Manager
work
10-year
analysis utility
While
is explicit with
to be programmed The values (Boudreau, Option a system, internal $249.86 PAT with this
a personal
computer using
spreadsheet
in Table 1987a).
9 were
generated
such a spreadsheet
1 depicts
random value
external of each
staffing. remains
Under constant
such as of the
force
movement
occur,
producing valid
a total
10-year
value using
million. a validity
Option of
external
selection
is enhanced,
and
enhances
force
of $296.90
Option
the
internal
staffing
manner. equal to
It acknowledges
increases
to $302.51
million.
Option
reducing
Programmers
producing
utility future
of $278.68
the assessment
predicts
for Managers,
staffing
This
centers
are always
poor
investments,
it illustrates
of a decision
framework
utility
Analysis:
A New Perspective
Page 55
incorporating supply
staffing that
decisions
on the
jobs
that
employees
employees. that
It also
only the
the limits
of internal count
staffing levels.
consider that
quantities
or head
the fact
were force
constant value
across
emerged.
decisions
affecting
internal
flows,
decisions force
between and
internal systems
size,
Dealing
With
Uncertainty the most integrative utility framework, selection factors. and and Obviously,
Model encompassing
and external
recruitment, investment
as well
as financial require
on such a model
estimates,
be uncertain analysis
or variable.
As with
the earlier
utility
systematically
assessment-center-based value
a reduction million
in average
Programmer
advantage #3 minus
of better $296.90
internal million
selection in Option
in Option losses
in the Programmer
job.
However, argue
the $625
Some might
that while
center
of the 3,686
Programmers
year.
How much
less would
the productivity
reduction
the decision
in favor
of the Assessment
Center?
Utility
Analysis:
A N@w Perspective
Page 57
9 suggest in average
that value
the the
work
force
after
force
Value
changes
Therefore, a change .
change Value
Promotion $23.83
in Total Knowing
Work
force
this,
we can find
Promotion of improved
exactly
offset
by dividing Effect
-$5.61
by $38,128.
This produces
of -$147
value
would
obtained does
rather
example).
the analysis
question work
of how Assessment
promotions
affect
force.
SUIIIna.ry
and Implications
Summary
This chapter has illustrated how cost-benefit effects (Table utility analysis resource utility flows, models
analyze
of human
program affecting
1) illustrated
#1 through 3 through
#6 in Table 9) showed
study not
how cost-benefit
utility
Analysis:
A New Perspective
Page 58
only
simple
one
cohort,
but the
separations
and internal
to integrate
stocks
demonstrated commonly
precise The
believed. variables
important could
task affect
values
for those
that
resource
assess effort
their
implications with
can proceed
a definite explicit
goal.
the
of utility human
analysis
is its ability
to make
relationships
linking
resource
decisions assumptions
outcomes,
the critical
Systems
Planning models
utility number
analysis
of productivity-related evaluated.
otherwise
As we have programs
summaries already
to evaluating a framework
they offer
resource
programs
resource
are planned,
communicated by
to affect
throughout analysis
they
implemented. and
offers
that
focuses
explicitly
consequences
resource makers,
Several
constituents
be considered with
concerned
productivity
utility
Analysis:
A New perspective
Page 59
but
employees,
often
(Milkovich
& Boudreau,
summarize to more
productivity easy
and other
less tangible
to balance
million-dollar legal
return
against
possible animosity,
effects task
on employee
or union
productivity will
makers
be operating important
data
address
organizational
concerns.
is No Excuse function
operates
uncertainty,
are routinely
considered
problems"
up to rigorous presented
a convenient that
it, but
incorrect
decisions. resource
that human
"soft"
Utility risk
or uncertainty
effects variable.
on decisions, Techniques
than
attempting even
to precisely
as break-
analysis
"what as they
As we have
of uncertainty
enhance
utility
Analysis:
A New perspective
Page 60
that
only
very
unlikely
events
could
make
human
resource
unprofitable. Decisions Are Management Decisions of human resource decisions that "our
Resource Though
the contributions
and importance
acknowledged
abstractly
statement
programs
consider
only leverage
of programs resource
decisions ignored.
large
of employees
and time
is often
If there here,
message decisions
from
the
illustrations a difference
presented to
it is that human
can make
and that
these
decisions
decisions
This
and
managers, within
they must
demonstrate framework,
resource
a management
provides provide a
information.
analysis
Ignoring decision
resource
or adopting
convenient
but
faulty
systems
is a dangerous or surprising
This is
of
conclusion, within
consequence
resource applies
decisions
of quantity, cases,
quality simply
to any management
In many
utility
Analysis:
A New perspective
Page 61
framing
human
resource
issues toward
in these
terms
clarifies
their
importance
managers Actual
better will
decisions involve
(Florin-Thuma
decisions
different
assumptions
to further tools
decision actual
support
using
computers,
such models
managerial
decisions
(Boudreau,
Dyer
1986).
It is hoped decisions using that HRM decision makers will analyze and present their with
cost-benefit
and utility
concepts; make
better HRM
supervisors, decisions;
productivity concepts
and no one
may have
supervisors,
line managers
continue
to ignore
resource
management's
contribution.
utility
Analysis:
A New Perspective
Page 62
References American Management report. Association New York: (1986). American Hiring strategies and costs:
Management A. (1981).
organizational Bierman,
of Management w. H.
(1981). Ill:
analysis Boudreau,
Homewood,
Irwin.
the utility
J. W.
considerations improvement
in estimating programs.
of human
resource
productivity
Personnel
resource
productivity
programs.
Journal
of Applied
Psychology, Boudreau, J. W.
68, 396-407. (1984). Decision theory contributions 23, 198-217. program NY: for quantifying to HRM research
Industrial
Relations.
(1985).
EXTMOV:
movement
external Boudreau,
employee
Ithaca,
Author.
resource management
J. W.
(in press).
Utility
in human
of Industrial-Organizational
MOVUTIL:
A spreadsheet employee
program
the utility
of internal
and external
movement.
Author.
Boudreau, J. W. (1987b). Utility analysis applied to internal framework. School and external
employee paper,
movement: Center
An integrated Human
theoretical Resource
Working
of Industrial
for Advanced
Studies,
and Labor
Relations,
Cornell
University.
Utility
Analysis:
A New Perspective
Page 63
Boudreau,
C. J. (1985a). employee
utility of Applied
to external
[Monograph],
movement
& G. R. Ferris
in personnel
and human
management
(pp. 31-53).
Greenwich,
analysis
for manpower
decision processes
of their
effects
and outcomes.
Research
Institute,
G. T. (1988).
Human
Personal
(PC) exercises
TX:
management. Boudreau,
Business (1985).
S. L. Journal On the
in staffing
utility Brogden,
of Applied
70, 354-366.
H. E.
interpretation
efficiency.
An approach
to the problem
of differential
prediction. Brogden, H. E.
Psychometrika,
(1949).
When
, 171-183. Brogden, H. E. & Taylor, E. K. concept (1950). The dollar criterion--Applying Personnel
to criterion
construction.
(1987).
utility
Analysis:
A New Perspective
Page 64
behavior Cascio,
in organizations. R.
(2nd ed.)
Boston,
MA:
Kent.
of a new terms.
W. F. & Ramos,
(1986).
Development
and application
in behavioral/economic
71, 20-28. utility of the assessment Psychology, 64, tests center 107-118. and personnel Press. for as
W. F. & Silbey
V.
(1979). Journal
a selection Cronbach,
device.
of Applied
Psychological University
decisions Cronshaw,
of Illinois
S. F. & Alexander,
(1985).
One answer
Selection Behavior M.
utility
and Human
P. & Piore,
(1971). MA:
Internal
labor
analysis.
Lexington,
Fitz-Enz,
J. (1984).
How to measure
McGraw-Hill. Flamholtz, E. (1985). Human resource accounting (2nd ed.). San Francisco,
CA:
Florin-Thuma, feedback
on managerial
decision
processes.
Psychology.
Gow, J. F. ( 1985 ) .
Human
resource
managers
must
remember
the bottom
line.
Personnel Hunter,
Journal,
April, F. L.
J. E. & Schmidt,
of personnel
selection
on national Human
productivity.
In M. D. Dunnette (Vol.
(Eds.),
performance
and productivity
F. J., Farr,
J. L. & Jacobs,
R. R.
(1982).
utility
concepts
in
utility
Analysis:
A New Perspectiv@
Page 65
Organizational
Behavior
and Human
W. R. (1979).
PAIR.
In D. Yoder
Jr., DC:
ASPA
of personnel of National
and industrial
relations.
in supervisory Journal,
skills:
A time-based
J. C. (1982).
systems.
In K. M. Rowland
(Eds.),
Personnel
Boston:
Allyn J. W.
Personnel/human
Plano, TX:
resource Publications,
management:
approach.
Business
Inc.
Naylor, J. C. & Shine, criterion L. C. (1965). score obtained A table by using for determining a selection the increase
in mean
device.
Journal
of Industrial Rich,
Psychology, J. W.
1, 33-42. (1987). The effects A Monte Carlo of variability analysis and risk
and comparison.
Psychology, (1984).
Rosenbaum,
J. E.
mobility
in a corporate
hierarchy.
New
York:
Rynes, S. L.
Academic
& Boudreau,
Practice,
and research
implications.
Psychology,
39, 729-757. N., Tenopyr, questions M. L., Kehoe, validity N. & Zedeck, S. (1985). and meta-
Commentary
about
generalization
Utility
Analysis:
A New Perspective
Page 66
analysis.
Personnel
Psychology, J. E. (1983).
Schmidt,
F. L. & Hunter,
An empirical procedure
utility.
of Applied
Psychology,
J. E., Pearlman,
K. & Hirsh,
H. R. (1985).
Forty
validity
generalization
and meta-analysis.
Personnel
38, 697-798. J. E., McKenzie, procedures R. C., & Muldrow, on work-force T. W. (1979).
selection
productivity.
of Applied Hunter,
Psychology,
F. L.,
impact
of personnel
on work-force
productivity.
Psychology,
35, 333-347. practices March. Careers and organizational behavior. labor American review: a personnel audit
(1974).
Personnel
activity. Stewman,
Personnel S. L.
S. & Konda,
Demographic of Sociology,
of organizational
40, 298-321. J. T. (1939). The relationship of tests Psychology, of validity in selection: 23, 565-578.
H. C. & Russell,
effectiveness of Applied
Journal
(1980).
inequality:
York:
Mechanisms Books.
of distribution
in the U.S.
economy.
Basic
utility
Analysis:
A New
P@rspective
Page 67
Table
1.
utility and Break-Even Analysis Applied to Decisions Employee stocks: A Training Delivery Decision. Decision
Affecting
Should training for engineers in basic production processes be delivered through: (1) Classroom-based training, with twenty half-day sessions, class sizes of 40, conducted by experienced engineering employees, training up to 40 students per year; or (2) An Audio-Video network, consisting of one broadcast studio and 10 remote conference rooms (one-way video, two-way audio), allowing up to 200 to be trained per year Option 1 Classroom Delivery Option Audio-Video 2 Delivery
Costs:
$451,035 over five years (no start-up costs because system already exists). Train 40 students per year
$1,031,147 over five years (very high startup costs). Train 200 incumbents in the first year, and 20 new hires in Years 2 through 5. in the Work Force (Leverage)
Quantity:
Number
of Trained
Employees
1 2 3 4 5
Dollar Value of the increase in productivity due to classroom training, per person-year (Symbolized P1). Payoff Formulas
u1
(600 X P1)
$451,035
Break-Even
u2 = (1,200 X P2)
Analysis
- $1,031,147
P1
$451,035/600
$752
P2
$1,031,147/1,200
= $860
utility
Analysis:
A New Perspective
Page 68
Table
1.
utility and Break-Even Analysis Applied to Decisions Affecting Employee stocks: A Training Delivery Decision. (Concluded)
Payoff
Formula
U2
- U1
(1,200
Decision
1. 2. 3.
If both P1 and P2 are less than break-even, do neither program. If P1 exceeds $752, and P2 is less than (.5 X P1) + $484, do Option If P2 exceeds $860, and P2 is more than (.5 X P1) + $484, do Option
1. 2.
utility
Analysis:
A New Perspective
Page 69
Table
2.
Summary
of Cost-Benefit
Decision
Models
for Employee
Flows
Decision
Model
Added
Features
by
Model #1: One-Cohort External Selection utility Model (Schmidt, et al., 1979)
Deciding how to choose which external applicants should be hired in a particular time period.
Model #2: Financial One-Cohort External Selection utility Model (Boudreau, 1983a) Model #3 Financial Multiple-Cohort External Selection utility Model (Boudreau, 1983b)
Effects of taxes, interest rates, and costs of maintaining and improving employee performance. Effects of re-applying the selection program to subsequent applicant
Financial value of deciding how to choose which external applicants should be hired in a particular time period. Financial Value of deciding how to choose which external applicants should be hired in each future time period during which a selection program is applied. Financial value of deciding how to attract the applicant pool, as well as how to choose which external applicants should be hired in each future time period during which recruitment and selection programs are applied. Financial value of deciding how to attract the applicant pool, how to choose which external applicants should be hired, and how to manage employee separations/retentions during each future time period during which recruitment, selection and separation management programs are applied.
groups.
Model #4 Financial Multiple-Cohort External Recruitment and Selection utility Model (Boudreau & Rynes,
1985)
Model #5 Financial Multiple-Cohort External Recruitment, Selection, and Separation/Retention utility Model (Boudreau & Berger,
Effects of employee separation/retention patterns on recruitment and selection, and vice versa.
1985)
utility
Analysis:
A New
Perspective
Pag@ 70
Table
2.
Decision
Models
for Employee
Flows
Decision
Model
Added
Features
by
Model #6: Financial Multiple-Cohort Internal/External Recruitment, Selection, and Separation/Retention utility Model (Boudreau, 1987b)
Effects of recruitment, selection and separation/retention of employees moving between jobs within the organization on external staffing decisions, and vice versa.
Financial value of deciding how to attract the applicant pool, how to choose which external applicants should be hired, and how to manage employee separations/retentions from the organization; as well as how to attract, choose and manage separations when employees move between jobs within the organization, during each future time period in which internal/external recruitment, selection and separation management programs are applied.
utility
Analysis:
A New Perspective
Page 71
Table
3.
Situation
to be Analyzed
Cost-Benefit Information Current Employment Number Separating Number Selected Number Promoted Adapted from:
.
Entry-Level Programmers
Computer
Upper-Level
Data
System
Schmidt,
(1979)
and Boudreau
(1987b)
utility
Analysis:
A New Perspective
Page 72
Table
4.
One-Cohort
External
Selection
utility
Model
Cost-Benefit Information
Entry-Level Programmers
Computer
Current Employment Number Separating Number selected Average Tenure Number of Applicants Average Test Score SO of Applicant Value (SOy)
Abili ty
Validi ty
1,236 .80 SO
$10,413/yr.
Test
.76
$10/applicant Cost Testing
$37.9
million
PAT
Adapted from: Schmidt, Hunter, McKenzie, & Muldrow (1979).
utility
Analysis:
A New Perspective
Page 73
Table
5.
Financial
One-Cohort
Entry-Level
Selection
Utility
Model
Cost-Benefit Information
Entry-Level Programmers
Computer
Current Employment Number Separating Number Selected Average Tenure Number of Applicants Average Test Score SD of Applicant Value (SDy) Variable Costs Corporate Tax Rate Corporate Interest
Abili ty Test
1,236 .80 SD
$10, 413/yr.
Rate
5% 45% 10%
.76
$10/applicant
After-Cost, After Tax, Discounted, One-Cohort utility Increase over Random Selection Without the PAT. Adapted from: Boudreau
$12.55
million
(1983a).
utility
Analysis:
A New Perspective
Page 74
Table
6.
Financial
Multiple-Cohort
External
Selection
utility
Model
Cost-Benefit Information
Entry-Level Programmers
Computer
Current Employment Number Separating Number Selected Average Tenure Number of Applicants Average Test Score SD of Applicant Value (SDy) Variable Costs Corporate Tax Rate Corporate Interest Analysis Period
1,236 .80 SD
$10 ,413/yr.
Rate
5% 45% 10%
10 years
7 years 31,282
Validity Testing Cost After-Cost, After Tax, Discounted utility Increase over Random Selection Without the PAT (Millions) Adapted from: Boudreau
.76
$10/applicant Benefit
-
Cost $.04
$54.32
$54.28
(1983b).
utility
Analysis:
A New Perspective
Page 7S
External Recruitment and Selection
Table
7.
Cost-Benefit Information
Entry-Level Programmers
Computer
Current Employment Number Separating Number Selected Average Tenure Number of Applicants Average Test Score Variable Costs Corporate Tax Rate Corporate Interest
Rate
force
utility
Recruitment
.76
$10japplicant
.60
$10japplicant
Recruitment Cost/Hire Avg. Applicant Service Value Avg. Applicant Service Cost Avg. Net Applicant Value SD of Applicant Value
(sn" )
Value of Random Selection Cost of Random Selection Value Added by Testing Cost Added by Testing Total After-Tax, After-Cost Discounted Work force Value (Millions) Adapted from: Boudreau & Rynes
$190.76
(1985).
$207.45
utility
Analysis:
A New Perspective
Page 76
Table
8.
Recruitment,
Selection
and
Cost-Benefit Information
Entry-Level prograrmners
Computer
Current Employment Beginning Average Service Value Beginning Average Service Cost Number Separating Number Selected Acquisition Cost Separation Cost Number of Applicants Average Applicant Service Value Average Applicant Service Cost Average Test Score SO of Applicant Value (SOy) Testing Cost Variable Costs Corporate Tax Rate Corporate Interest Analysis Period
5%
Rate
45% 10%
10 years
Work force utility Results Staffing Variable Test Validity Separation Effect After-Tax, After-Cost Discounted Work force Value (Millions)
Adapted from: Boudreau
Option 1 0.00 $0
Option 2 0 .76 $0
$200.31
& Berger (1985a).
$242.10
$351. 69
$132.50
utility
Analysis:
A New Perspective
P<lge 77
Table
9.
Recruitment,
Selection,
Cost-Benefit Information Current Employment Beginning Average Service Value Beginning Average Service Cost Number Separating Number Selected Number Promoted Acquisition Cost Separation Cost Promotion Cost Number of Applicants Average Applicant Service Value Average Applicant Service Cost Average Test Score SD of Applicant Value (SDy) Testing Cost Assessment Center Cost Variable Costs Corporate Tax Rate Corporate Interest Rate Analysis Period
Entry-Level Computer Programmers 4,404 $52,065 $36,445 618 718 100 $7,000 $7,000 NA 1,436 $52,065/yr.
Upper-Level Data System Managers 1,000 $57,272 $40,000 100 0 100 NA $8,000 $8,000 3,786 1.10 times average Programmer value 1.10 times average Programmer cost 2.32 SD $11,454/yr. NA $1.44 million/yr. 5% 45% 10% 10 Years
$36,445/yr. .80 SD
$10 ,413/yr. $10/ applicant
Results
Options
HRM Activity Programmer Selection Validity Programmer Promotion Effect Manager Promotion Validity After-Cost, After-Tax, Discounted Total Work force Value (Millions) Adapted from: Boudreau (1987b). 1 0.00 $0 0.00 2 0.76 $0 0.00 3 0.76 $0 0.35 4 0.76 -$625 0.35
$249.86
$296.90
$302.51
$278.68