Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
CHAPTER
FOUR
ECONOMIC
IMPACT
ANALYSIS
METHODOLOGY
4.1
OVERVIEW
OF
ECONOMIC
IMPACT
ANALYSIS
METHODOLOGY
This
chapter
presents
EPA
 
s
methodology
for
analyzing
the
economic
impacts
of
the
proposed
erosion
and
sediment
control
(
ESC)
regulations
for
the
construction
and
development
(
C&
D)
industry.

EPA
has
employed
a
number
of
different
methods
for
assessing
the
economic
impacts
of
the
proposed
rule.

These
include
models
that
analyze
impacts
at
the
level
of
the
individual
construction
project,
the
individual
firm,
national
construction
markets,
and
the
national
economy
as
a
whole.
The
analysis
considers
impacts
on
the
firms
in
the
C&
D
industry
who
would
be
complying
with
the
regulations,
on
those
who
purchase
the
output
of
the
C&
D
industry,
and
on
those
who
would
be
responsible
for
implementing
the
proposed
rule.

The
analysis
is
based
upon
engineering
cost
estimates
developed
by
EPA.
The
engineering
costs
reflect
the
costs
to
comply
with
requirements
related
to
erosion
and
sediment
controls
(
ESCs)
employed
over
a
relatively
short
period
(
generally
less
than
one
year)
during
which
land
is
being
converted
from
an
undeveloped
to
a
developed
state.
The
engineering
costs
also
include
the
costs
associated
with
meeting
any
paperwork
requirements
triggered
by
the
proposed
rule,
including
any
requirements
related
to
the
permitting
of
construction
and
development
projects,
and
incremental
inspection
and
certification
requirements
for
ESCs.

The
outline
of
the
chapter
is
as
follows:

 
Section
4.2
presents
EPA
 
s
analysis
of
the
impacts
of
the
proposed
rule
on
model
C&
D
projects.
Here
EPA
develops
pro
forma
financial
analyses
for
representative
projects
and
analyzes
the
impact
of
the
incremental
regulatory
costs
on
project
viability.
The
section
includes
a
description
of
the
model
projects,
model
project
analysis
methodology,
data
sources,
and
assumptions
used
in
the
model
project
analysis.
The
model
project
analysis
results
are
presented
in
Chapter
Five,
Section
5.2.

 
Section
4.3
presents
EPA
 
s
analysis
of
the
impacts
of
the
proposed
rule
on
model
C&
D
firms.
This
section
uses
data
on
the
financial
condition
of
representative
firms
to
examine
the
impact
of
the
incremental
compliance
requirements
on
the
model
firm
 
s
financial
4­
1
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
condition.
This
section
also
describes
how
the
model
firm
analysis
is
used
to
evaluate
economic
achievability
and
barrier
to
entry
considerations
for
the
proposed
rule,
and
to
conduct
the
firm
closure
analysis
and
small
entity
impact
analysis.
This
section
includes
a
description
of
the
model
firms,
model
firm
analysis
methodology,
data
sources,
and
assumptions
used
in
the
model
firm
analysis.
The
model
firm
analysis
results,
including
those
from
the
economic
achievability,
barrier
to
entry,
closure,
and
employment
loss
analyses,
are
presented
in
Chapter
Five,
Sections
5.4
through
5.6.

 
Section
4.4
presents
EPA
 
s
methodology
for
estimating
the
national
compliance
costs
of
the
proposed
rule.
These
costs
are
estimated
starting
with
the
per­
acre
compliance
costs
estimated
by
EPA.
The
per­
acre
costs
are
applied
to
national
estimates
of
the
amount
of
land
converted
to
developed
status
annually.
National
compliance
cost
estimates
are
presented
in
Chapter
Five,
Section
5.3.

 
Section
4.5
describes
EPA
 
s
partial
equilibrium
market
model
analysis.
This
section
considers
the
impact
of
the
incremental
compliance
requirements
on
consumers
of
the
construction
industry
 
s
output,
in
particular
the
impacts
on
home
buyers
and
on
housing
affordability.
The
section
includes
a
description
of
the
market
model
methodology,
data
sources,
and
assumptions
used
in
the
market
models.
The
market
modeling
results
are
presented
in
Chapter
Five,
Section
5.6.

 
Section
4.6
expands
the
analysis
to
consider
the
net
impacts
of
the
proposed
rule
on
the
national
economy.
While
the
compliance
costs
would
reduce
output
in
the
construction
industry
there
may
be
an
offsetting
increase
in
spending
related
to
ESCs
and
inspection
and
certification.
EPA
uses
input­
output
analysis
to
trace
the
implications
of
these
spending
shifts
on
the
national
economy.
The
result
is
an
overall
estimate
of
the
impact
on
macroeconomic
variables
such
as
output
and
national
employment.
The
results
of
the
national
economic
impact
analysis
are
presented
in
Chapter
Five,
Section
5.7.

 
Section
4.7
considers
the
impacts
on
governmental
units
associated
with
establishing
or
modifying
permitting
programs
to
reflect
the
requirements
in
the
proposed
rule
as
well
as
new
or
increased
costs
related
to
permit
processing.
The
results
of
the
government
cost
impact
analyses
are
presented
in
Chapter
Five,
Section
5.8.

4.
1
.
1
Compliance
and
Baseline
Assumptions
In
this
analysis
EPA
assumes
that
the
proposed
rule
would
impact
markets
that
have
already
fully
implemented
existing
regulations
related
to
storm
water
controls
for
C&
D
activities.
EPA
assumes
that
all
states,
tribal
lands,
and
territories
comply
with
the
existing
regulations
or
have
equivalent
programs.
These
programs
are
assumed
to
include
all
of
the
requirements
affecting
C&
D
activities
that
were
part
of
the
national
storm
water
Phase
I
and
Phase
II
NPDES
storm
water
regulations.
Since
the
Phase
II
regulations
are
not
scheduled
to
be
fully
implemented
until
2003,
however,
EPA
acknowledges
that
4­
2
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
current
market
conditions
may
not
fully
reflect
the
baseline
that
would
apply
at
the
time
the
proposed
rule
comes
into
force.
Specifically,
EPA
notes
that
the
baseline
market
conditions
assumed
in
this
analysis
(
including
baseline
financial
conditions
for
affected
firms)
may
not
fully
reflect
the
implementation
of
Phase
II
NPDES
storm
water
requirements.
For
this
reason,
EPA
has
conducted
a
supplemental
analysis
that
reflects
less
than
100
percent
implementation
of
the
Phase
II
NPDES
storm
water
rule
in
the
baseline.

The
supplemental
baseline
analysis
is
presented
in
Appendix
5C
of
this
report.

4.1.2
Cost
Pass
Through
Assumptions
EPA
has
incorporated
into
each
of
the
impact
analyses
described
below
specific
assumptions
about
the
incidence
of
the
compliance
costs.
This
section
describes
generally
EPA
 
s
conclusions
about
cost
incidence
for
the
proposed
regulation
and
then
outlines
the
specific
assumptions
made
for
each
impact
analysis.

In
general,
EPA
believes
that
developers
and
builders
faced
with
an
increase
in
costs
due
to
new
ESC
requirements
would
have
an
incentive
to
pass
on
all
or
some
of
the
increased
cost
to
the
project
owner.
(
This
is
referred
to
as
cost
pass
through,
or
CPT)
.
The
extent
to
which
the
costs
can
be
passed
through
in
practice
would
depend
on
market
conditions.
The
demand
elasticity
of
the
project
owner
(
i.
e.
,

the
sensitivity
of
the
purchase
decision
to
incremental
changes
in
price)
would
be
influenced
by
two
main
factors:

 
The
magnitude
of
the
cost
increase
relative
to
the
overall
cost
of
the
project.
For
example,
on
a
large
office
project
or
even
a
high
end
single­
family
home,
the
buyer
may
put
up
little
resistance
if
the
cost
increase
is
small
relative
to
the
overall
cost
of
the
project.

 
The
availability
and
price
of
substitutes.
If
the
cost
increase
affects
all
suppliers
and
all
substitutes
equally,
then
the
project
owner
is
less
likely
to
resist
an
incremental
price
increase.

Since
the
proposed
rule
would
be
national
in
scope
and
the
compliance
costs
would
be
similar
within
a
given
geographic
region
(
assuming
similar
sites)
,
the
compliance
costs
should
affect
the
buyer
 
s
alternate
suppliers
roughly
equally.
This
suggests
that
if
the
costs
are
small
relative
to
the
total
cost
of
4­
3
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
the
project,
demand
should
be
relatively
inelastic
and
the
builder
would
be
able
to
pass
all
or
most
of
the
cost
increase
on
to
the
buyer.

Another
factor
facilitating
cost
pass
through
for
builders
is
that
project
owners
often
plan
for
unexpected
cost
changes
by
building
contingencies
into
their
budgets.
A
common
mechanism
in
new
residential
construction,
for
example,
is
for
the
home
buyer
to
absorb
an
unexpected
cost
increase
at
one
stage
of
construction
by
reducing
costs
on
a
later
stage.
This
might
be
done,
for
example,
by
selecting
less
expensive
flooring
material,
deferring
finishing
of
a
basement,
or
opting
to
build
a
garage
at
a
later
date.

This
line
of
reasoning,
which
suggests
demand
is
generally
inelastic,
presumes
that
the
 
good
 
the
buyer
is
purchasing
is
 
new
construction.
 
In
most
markets,
however,
the
owner
can
also
elect
to
buy
from
an
inventory
of
existing
homes,
office
or
retail
space,
or
industrial
facilities
available
for
sale,
or
to
rent
from
a
corresponding
inventory
of
rental
properties.
To
the
extent
that
existing
construction
and
rental
property
serves
as
a
perfect
or
even
partial
substitute
for
new
construction,
the
buyer
 
s
demand
elasticity
would
also
be
influenced
by
conditions
in
the
existing
construction
and
rental
markets.

Existing
homes
and
existing
office,
retail
or
industrial
space
would
not
be
affected
by
the
proposed
regulation.
Cost
increases
that
differentially
affect
new
construction
may
cause
some
buyers
to
choose
existing
construction
over
new,
i.
e.
,
they
could
elect
to
buy
or
rent
rather
than
build.
The
strength
of
demand
for
new
relative
to
existing
construction
depends
on
the
relative
availability,
suitability,
and
price
of
each
type
of
construction.
Buyers
choosing
new
over
existing
construction
often
do
so
for
reasons
related
to
location,
the
ability
to
match
their
specific
needs,
expected
length
of
tenure,
and
greater
certainty
about
a
structure
 
s
condition
and
future
maintenance
requirements.
Demand
for
new
construction
is
also
highly
influenced
by
the
availability,
quality,
and
age
of
existing
construction.
In
geographic
areas
or
market
segments
where
the
existing
inventory
is
weak
or
unsuitable,
demand
for
new
construction
would
be
stronger.

Evidence
from
the
literature
suggests
that
in
residential
construction,
regulatory­
related
costs
are
usually
passed
on
to
consumers
(
e.
g.
,
Luger
and
Temkin
2000)
,
and
this
general
observation
was
echoed
during
EPA
 
s
focus
group
sessions
with
members
of
NAHB.
Industry
literature
points
out
that
in
the
recent
past,
a
variety
of
market
forces
have
shifted
the
new
construction
market
towards
larger,
more
4­
4
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
expensive
homes
(
NAHB
2001a)
.
Other
things
equal,
demand
in
the
higher
end
of
the
housing
market
tends
to
be
more
inelastic.
Efforts
to
model
the
housing
sector
have
shown
that
new
construction
is
more
affected
by
changes
in
household
formation
and
income
than
marginal
changes
in
price
(
Hirsch
1994)
.

Given
this
evidence,
EPA
believes
overall
that
demand
in
the
single­
family
housing
sector
is
relatively
inelastic.

In
the
other
sectors
modeled
(
multifamily
housing,
commercial,
industrial)
,
EPA
believes
demand
to
be
relatively
inelastic
as
well.
In
the
non­
residential
sectors,
interest
rates,
regional
economic
performance
and
outlook,
and
changing
technological
needs
are
important
drivers
of
building
demand.

As
shown
in
the
subsequent
chapter,
the
change
in
costs
relative
to
total
project
costs
in
these
markets
are
relatively
small
and
unlikely
to
influence
the
purchase
decision,
given
the
greater
significance
of
these
other
factors.

EPA
notes
that
under
certain
conditions
developers
might
also
attempt
to
pass
regulatory
costs
back
to
land
owners.
In
a
depressed
market,
builders
may
argue
successfully
that
a
regulatory
cost
increase
would
make
a
particular
project
unprofitable
unless
the
land
costs
can
be
reduced.
For
example,

if
the
land
owner
is
convinced
that
a
residential
subdivision
project
would
not
go
ahead
because
home
buyers
would
not
absorb
an
unexpected
increase
in
sales
price,
they
may
be
willing
to
accept
a
lower
price
per
acre
for
raw
land.
The
ability
of
developers
to
pass
such
costs
back
would
likely
depend
on
the
land
owner
 
s
experience
in
land
development
projects,
their
knowledge
of
the
local
real
estate
market,

and
in
particular
their
understanding
of
the
regulation
and
its
likely
cost.
While
some
evidence
of
cost
pass­
back
to
land
owners
exists
for
fixed
and
readily
identifiable
regulatory
costs,
such
as
development
impact
fees
(
Luger
and
Temkin
2000)
,
it
is
unclear
whether
a
builder
 
s
claim
that
costs
would
be
higher
due
to
the
types
of
requirements
imposed
by
the
proposed
rule
would
induce
land
owners
to
make
concessions.

In
the
sections
below,
EPA
has
made
differing
assumptions
concerning
whether
compliance
costs
are
passed
through
to
buyers,
and
to
what
extent.
In
the
model
project
analyses
in
Section
4.2,
for
example,
EPA
analyzes
results
under
the
extreme
conditions
of
zero
and
100
percent
CPT.
This
enables
EPA
to
examine
the
impacts
under
worst­
case
assumptions
with
respect
to
builders
(
zero
CPT)
,
as
well
as
to
owners
(
100
percent
CPT)
.

4­
5
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
In
other
parts
of
the
impact
analysis
EPA
introduces
more
realistic
assumptions
about
actual
market
conditions.
For
example,
it
is
generally
thought
that
the
long
run
supply
of
new
construction
is
almost
perfectly
elastic,
as
resources
can
shift
easily
into
the
industry.
When
empirical
elasticity
estimates
are
used
to
estimate
actual
cost
pass
through,
the
combination
of
inelastic
demand
and
highly
elastic
supply
results
in
relatively
high
cost
pass
through
rates,
on
the
order
of
85
percent.
In
the
model
firm
and
closure
analysis
(
Section
4.3)
,
EPA
analyzes
the
impacts
under
conditions
of
zero
CPT
(
worst­

case)
as
well
as
under
the
most
realistic
estimates
of
actual
CPT.
In
the
market
models
(
Sections
4.5
and
4.6)
EPA
uses
only
the
estimates
of
actual
CPT.

4.1.3
Operation
and
Maintenance
Costs
In
order
to
remain
effective
all
of
the
ESCs
should
be
maintained.
The
engineering
costs
for
ESCs
include
costs
for
operating
and
maintaining
the
controls.
The
controls
used
during
the
active
phase
of
construction
are
assumed
to
be
in
place
for
one
year
and
therefore
should
be
maintained
throughout
the
period.

4.1.4
Impacts
Associated
With
NSPS
Under
Option
2,
EPA
is
proposing
to
define
a
 
new
source
 
under
Part
450
as:
:
 
any
source
of
storm
water
discharge
associated
with
construction
activity
that
results
in
the
disturbance
of
at
least
five
acres
total
land
area
that
itself
will
produce
an
industrial
source
from
which
there
may
be
a
discharge
of
pollutants
regulated
by
some
other
new
source
performance
standard
elsewhere
under
subchapter
N.
 
1
This
definition
would
mean
that
the
land­
disturbing
activity
associated
with
constructing
a
particular
facility
would
not
itself
constitute
a
"
new
source"
unless
the
results
of
that
construction
would
yield
a
"
new
source"
regulated
by
other
new
source
performance
standards.
For
example,
construction
activity
that
is
intended
to
build
a
new
pharmaceutical
plant
covered
by
40
CFR
439.15
would
be
subject
to
new
source
performance
standards
under
§
450.24.
.
At
the
same
time,
EPA
is
seeking
comment
on
whether
1
All
new
source
performance
standards
promulgated
by
EPA
for
categories
of
point
sources
are
codified
in
subchapter
N.

4­
6
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
no
sources
regulated
under
Option
2
should
be
deemed
 
new
sources
 
on
the
grounds
that
construction
activity
itself
is
outside
the
scope
of
those
activities
intended
to
be
covered
by
Section
306
of
the
Clean
Water
Act
(
CWA)
.
2
Under
the
proposed
definition,
EPA
believes
that
the
NSPS
standards
could
trigger
a
National
Environmental
Policy
Act
(
NEPA)
review
process
for
those
C&
D
activities
permitted
by
EPA.
To
assess
the
potential
impact
of
such
a
result,
EPA
examined
NPDES
construction
permitting
data
for
19
states
fully
or
partially
administered
by
EPA.
In
2000,
the
number
of
permits
administered
by
EPA
was
8,563.

EPA
believes,
however,
that
by
the
time
the
proposed
C&
D
rule
is
finalized
the
states
of
Florida,
Maine,

and
Texas
(
currently
fully
administered
by
EPA)
will
have
assumed
permitting
authority
for
construction
activities.
In
2000,
the
number
of
permits
administered
by
EPA
excluding
these
three
states
was
1,454.

The
NPDES
permitting
data
does
not
include
sufficient
detail
to
indicate
the
number
of
sources
that
could
be
new
sources
covered
by
CWA
Section
306.
EPA
notes,
however,
that
in
a
1999
study
of
14
jurisdictions,
slightly
under
one
percent
of
construction
permits
were
for
industrial
facilities
(
EPA,
1999;

see
Table
4­
15)
.
Based
on
this,
EPA
believes
that
the
number
of
construction
permits
for
new
sources
(
regulated
under
Subchapter
N)
that
would
be
administered
by
EPA
is
likely
to
be
small.
At
this
time,

therefore,
EPA
has
not
estimated
any
potential
costs
for
NEPA
review
as
part
of
this
economic
analysis.

4.2
IMPACTS
ON
MODEL
PROJECTS
EPA
has
analyzed
the
impacts
of
the
proposed
rule
by
developing
financial
models
of
representative
C&
D
projects.
These
models
evaluate
whether
the
additional
costs
of
complying
with
the
proposed
regulation
would
make
the
model
project
unprofitable
and
vulnerable
to
abandonment
or
closure.
In
the
absence
of
an
industry
survey,
the
economic
models
are
based
on
EPA
 
s
best
available
data
and
assumptions
concerning
construction
project
characteristics,
and
are
designed
to
depict
as
accurately
as
possible
the
change
in
cash
flow
resulting
from
compliance
with
the
proposed
rule
for
typical
projects,
representative
of
the
type
required
to
comply
with
the
proposed
rule.
The
models
developed
reflect
the
range
of
C&
D
projects
typically
undertaken
by
industry
participants.

2
"
The
term
'
new
source'
means
any
source,
the
construction
of
which
is
commenced
.
.
.
"
33
U.
S.
C.
sec.
1316(
a)
(
2)
(
emphasis
added)
.

4­
7
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.2.1
Description
of
Model
Project
Approach
EPA
selected
the
model
project
types
by
analyzing
data
on
the
output
of
the
C&
D
industry.
The
industry
output
reflects
both
the
diversity
of
the
industry
itself
and
the
diversity
of
the
U.
S.
economy
overall.
To
illustrate
this
diversity,
EPA
notes
that
the
Census
of
Construction
assigns
construction
projects
to
one
of
17
building
and
32
nonbuilding
construction
categories
(
see
Appendix
2A,
Table
2A­
3)
.

In
terms
of
economic
value,
building
construction
projects
accounted
for
$
371.4
billion
(
97.3
percent
of
total
construction
revenues)
in
1997,
while
nonbuilding
construction
projects
accounted
for
only
$
5.9
billion
(
1.5
percent)
.
3
The
largest
single
category
of
construction
activity,
accounting
for
$
150.5
billion
(
39.4
percent
of
the
total)
,
was
single­
family
home
construction.
This
was
followed
by
office
buildings
at
$
40.3
billion
(
10.6
percent
of
the
total)
,
all
other
commercial
buildings
at
$
36.5
billion
(
9.6
percent
of
the
total)
,

manufacturing
and
light
industrial
buildings
at
$
26.2
billion
(
6.8
percent
of
the
total)
,
educational
buildings
at
$
25.1
billion
(
6.6
percent
of
the
total)
,
and
multifamily
housing
at
$
19.6
billion
(
5.1
percent
of
the
total)
.
Based
on
this
review,
EPA
developed
models
for
four
types
of
development
projects
that
reflect
the
range
of
projects
undertaken
by
the
industry
and
that
would
fall
within
the
ambit
of
the
proposed
rule.

These
included:

 
A
residential
development
of
single­
family
homes
 
A
residential
development
of
multifamily
housing
units
 
A
commercial
development
(
enclosed
shopping
center)

 
An
industrial
development
(
industrial
park)

Furthermore,
for
each
class
of
project
,
EPA
has
developed
models
that
correspond
to
a
range
of
project
sizes.
In
each
case,
there
are
versions
of
the
model
for
projects
of
1,
3,
7.5,
25,
70,
and
200
acres.

The
combination
of
four
project
types
and
six
project
size
classes
results
in
a
total
of
24
model
projects.

EPA
 
s
models
for
these
projects
assess
their
vulnerability
to
shutdown
or
closure
by
predicting
the
cash
flow
changes
that
would
result
from
the
incremental
costs
that
project
developers
would
incur
in
3
Another
$
4.2
billion
(
1.1
percent
of
the
total)
was
not
specified
by
kind.

4­
8
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
complying
with
the
proposed
rule.
The
models
establish
the
baseline
financial
conditions
for
each
representative
project
and
assess
the
significance
of
the
change
in
cash
flow
that
results
from
the
incremental
compliance
costs.
The
model
project
characteristics
are
based
on
best
available
data
and
reasonable
assumptions
about
development
activities
and
project
financing.

4.2.2
Treatment
of
Nonbuilding
Construction
Projects
As
noted
above,
an
estimated
$
5.9
billion
in
nonbuilding
construction
is
put
in
place
each
year.

This
total
includes
highways,
roads
and
streets
(
$
1.6
billion)
;
sewage
and
water
treatment
facilities
(
$
1.7
billion)
;
bridges,
tunnels,
and
elevated
highways
(
$
587
million)
;
sewers
and
water
mains
(
$
211
million)
;

power
and
communication
lines
and
towers
(
$
160
million)
;
and
private
driveways
and
parking
areas
(
$
100
million)
.
While
considerable
in
absolute
value,
such
nonbuilding
construction
activity
represents
less
than
two
percent
of
the
total
value
of
construction
completed.
Estimates
of
the
land
area
disturbed
as
a
result
of
nonbuilding
construction
activity
are
not
available.

EPA
has
not
developed
engineering
costs
applicable
to
nonbuilding
construction
projects,
due
to
the
diversity
of
the
activities
covered
under
this
category
and
the
relatively
small
share
of
overall
construction
activity
it
accounts
for.
4
By
way
of
analysis,
EPA
has
developed
a
reduced
form
model
project
for
highway
construction
and
analyzed
the
likely
magnitude
of
the
costs
and
impacts
using
the
highway
model.
This
analysis
is
presented
in
Section
4.2.7.

4.2.3
Description
of
Model
Projects
To
develop
the
model
projects,
EPA
focused
first
on
the
single­
family
residential
model
project.

As
noted
above,
single­
family
residential
construction
represents
the
highest
value
category
of
construction,
and
information
about
the
construction
and
development
process
for
single­
family
homes
is
4
The
national
costs
of
the
proposed
rule,
however,
do
account
for
the
costs
borne
for
these
types
of
projects.
See
Section
4.4.

4­
9
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
readily
available.
5
EPA
was
able
to
develop
a
relatively
detailed
model
for
the
single­
family
development
and
then
adjusted
the
model
parameters
as
necessary
to
reflect
differences
in
the
other
project
categories.
In
general,
EPA
believes
that
projects
in
the
other
categories
follow
a
roughly
similar
development
path,
and
has
thus
used
a
similar
general
structure
for
all
of
the
models.

Since
many
of
the
data
elements
and
modeling
assumptions
are
based
on
the
single­
family
residential
model,
this
model
is
discussed
in
detail
below.
Many
of
the
assumptions
and
data
elements
defined
for
this
model
were
applied
directly
or
modified
only
slightly
for
use
in
the
other
models.
The
discussion
of
the
other
three
project
types
focuses
primarily
on
those
assumptions
or
methods
that
differ
from
those
employed
in
the
single­
family
residential
model.

4.2.3.1
Residential
Single­
family
Development
The
model
single­
family
residential
project
or
site
is
an
undeveloped
parcel
zoned
for
single­

family
residential
housing.
The
number
of
housing
units
built
would
depend
on
the
size
of
the
model
project.
6
The
location
of
the
site
is
unspecified,
and
for
this
reason
EPA
has
used
national­
level
data
wherever
possible.
In
this
case,
the
site
is
assumed
to
be
controlled
by
a
developer­
builder
(
sometimes
referred
to
in
the
industry
as
merchant
builders
or
operative
builders)
.
The
developer­
builder
is
responsible
for
all
aspects
of
the
project,
from
land
acquisition
through
permitting,
subdivision
of
the
parcel,
installation
of
any
ESCs,
and
construction
and
marketing
of
all
completed
housing
units.
EPA
recognizes
that
there
are
many
variations
on
how
a
particular
site
may
be
developed,
but
believes
this
model
project
to
be
representative
of
a
large
number
of
projects
actually
undertaken
each
year
in
the
U.
S.
7
5
For
example,
EPA
was
able
to
obtain
input
to
the
single­
family
residential
model
from
representatives
of
the
National
Association
of
Home
Builders
(
NAHB)
,
a
prominent
C&
D
industry
association.
Input
from
NAHB
assisted
EPA
in
identifying
cost
elements
associated
with
each
stage
of
project
development.

6
Model
projects
were
developed
for
sites
of
1,
3,
7.5,
25,
70,
and
200
acres.

7
Other
common
scenarios
involve
the
developer
selling
all
or
some
of
the
finished
lots
to
builders.
The
developer
may
or
may
not
retain
lots
in
the
development
to
complete
and
sell.
See
Figure
2­
4,
for
example.

4­
10
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
The
starting
point
for
the
project
is
the
acquisition
of
the
parcel,
which
is
assumed
to
be
purchased
or
optioned
from
another
land
owner.
8
The
development
and
construction
process,
as
modeled,
is
assumed
to
proceed
through
three
phases,
characterized
as
follows:

 
Land
acquisition
 
The
developer­
builder
puts
together
the
necessary
financing
to
purchase
the
parcel.
When
lenders
are
involved,
they
may
require
certain
documentation,
such
as
financial
statements,
tax
returns,
appraisals,
proof
of
the
developer
 
s
ability
to
obtain
necessary
zoning,
evaluations
of
project
location,
assessments
of
the
capacity
of
existing
infrastructure,
letters
of
intent
from
city/
town
to
install
infrastructure,
environmental
approvals,
etc.
To
satisfy
these
factors,
the
developer
may
incur
costs
associated
with
compiling
this
data.

 
Land
development
 
The
developer­
builder
obtains
all
necessary
site
approvals
and
prepares
the
site
for
the
construction
phase
of
the
project.
Costs
incurred
during
this
stage
are
divided
among
soft
costs
for
architectural
and
engineering
services,
legal
work,
permits,
fees,
and
testing,
and
hard
costs
,
such
as
land
clearing,
installing
utilities
and
roads,
and
preparing
foundations
or
pads.
The
result
of
this
phase
is
a
legally
subdivided
parcel
with
finished
lots
ready
for
construction.

 
Construction
 
The
developer­
builder
undertakes
the
actual
construction
of
the
housing
units.
A
substantial
portion
of
this
work
may
be
subcontracted
out
to
specialty
subcontractors
(
foundation,
framing,
roofing,
plumbing,
electrical,
painting,
etc.
)
.
Marketing
of
the
development
generally
begins
prior
to
the
start
of
this
phase,
hence
the
developer­
builder
may
also
incur
some
marketing
costs
at
this
time.
Housing
units
may
come
under
sales
agreement
at
any
time
prior
to,
during,
or
after
completion
of
construction.

While
the
length
of
each
phase
and
the
overall
length
of
the
project
may
vary
considerably,
EPA
assumes,
for
modeling
purposes,
that
the
time
elapsed
from
acquisition
of
the
parcel
through
development
and
construction
totals
36
months.
Focus
groups
with
NAHB
in
Dallas
provided
estimates
that
ranged
from
13
to
63
months.
While
acknowledging
there
will
be
wide
variation
in
the
duration
of
each
phase,

EPA
further
assumes
that
each
phase
 
land
acquisition,
development,
and
construction
 
takes
12
months.

EPA
presently
lacks
detailed
data
on
the
exact
timing
of
ESC
installation
during
project
development.
EPA
assumes
that
ESCs
installed
to
control
runoff
during
the
active
phase
of
construction
8
Options
involve
payments
from
the
developer
to
a
land
owner
to
secure
the
rights
to
develop
the
land
for
a
specified
period
of
time,
usually
while
a
more
complete
assessment
of
project
viability
is
undertaken.

4­
11
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
are
put
in
place
early
in
the
development
phase
and
are
maintained
throughout
the
construction
phase.

Thus,
the
capital
costs
for
such
ESCs
would
generally
be
incurred
early
in
the
project,
and
the
structures
would
be
maintained
in
place
for
the
duration
of
the
project.
The
costs
for
removing
the
ESCs
would
be
incurred
at
project
completion.
9
These
general
assumptions
aside,
in
this
analysis
EPA
has
used
the
simplifying
assumption
that
the
costs
for
all
ESCs
are
incurred
at
the
beginning
of
the
project.
EPA
acknowledges
that
capital
costs
would
actually
be
incurred
some
time
after
the
start
of
the
project,
and
that
as
a
result,
the
costs
would
be
discounted
back
to
their
present
value.
In
making
this
assumption,

EPA
is
thus
overstating
the
magnitude
of
the
true
costs
incurred,
since
costs
incurred
in
the
future
would
have
a
lower
present
value.

EPA
understands
that
land
development
projects
involve
significant
cash
outflows
early
on
to
finance
land
acquisition,
development,
and
construction,
with
revenues
generally
received
only
after
completed
houses
are
sold
to
buyers.
For
this
reason,
EPA
assumes
that
the
integrated
developer­
builder
assumed
here
would
be
motivated
to
have
several
projects
underway
at
one
time.
Cash
inflows
from
the
sale
of
completed
units
in
one
development
can
offset
cash
outflows
associated
with
the
earlier
stages
of
development
on
another
project.
For
simplicity,
EPA
assumes
that
the
developer­
builder
involved
in
the
model
project
has
three
projects
underway
so
that
in
any
given
year
the
developer­
builder
incurs
all
of
the
costs
 
and
earns
all
of
the
revenues
 
associated
with
completing
the
land
acquisition,
development,

and
construction
phases
of
a
project,
even
though
these
may
occur
on
different
projects.

Additional
assumptions
and
sources
for
data
used
in
the
model
project
analysis
are
presented
below.
The
model
project
is
developed
using
assumptions
about
the
types
and
magnitude
of
costs
incurred
during
various
phases
of
the
project,
the
sources
for
these
funds
(
i.
e.
,
the
amounts
borrowed
versus
the
amounts
provided
from
the
developer­
builder
 
s
equity)
,
and
the
expected
profit
margins
earned
by
the
developer­
builder
from
each
phase
of
the
project.
EPA
is
seeking
comments
on
these
assumptions
as
well
as
any
additional
data
that
may
enable
the
Agency
to
more
accurately
model
such
impacts
at
the
project
level.

9
In
practice,
some
ESCs
installed
to
control
runoff
during
the
construction
phase
that
are
then
converted
to
permanent
BMPs
to
control
post­
construction
flows.
These
structures
would
not
need
to
be
removed.

4­
12
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Assumptions
regarding
the
various
cost
elements
incurred
during
each
phase
of
the
residential
single­
family
development
are
described
in
detail
in
Section
4.2.5.

4.2.3.2
Residential
Multifamily
Development
The
model
multifamily
residential
development
is
an
apartment
building
or
complex.
The
project
is
assumed
to
be
developed
in
a
similar
fashion
to
the
single­
family
model
development
described
above:
a
single
developer­
builder
is
responsible
for
all
site
acquisition,
site
preparation,
construction,
and
marketing
of
the
project;
the
project
timeline
is
similar,
i.
e.
,
three
years
from
start
to
finish;
and
the
project
proceeds
through
the
same
project
phases.
Similarly,
the
developer­
builder
is
assumed
to
have
several
projects
underway
to
help
balance
cash
flows.
This
assumption
makes
it
possible
to
examine
the
impacts
of
a
three
year
project
on
a
single
year
 
s
cashflow
for
the
affected
business.
Data
sources
and
inputs
specific
to
the
model
multifamily
development
are
discussed
in
Section
4.2.5.

4.2.3.3
Commercial
Development
The
commercial
development
is
assumed
to
be
an
enclosed
retail
shopping
or
office
area.

Depending
on
the
size
of
the
model
project,
it
could
range
from
a
small
stand­
alone
retail
outlet
to
a
large,
enclosed
mall
or
office
complex.
As
with
the
residential
projects,
a
single
developer­
builder
is
assumed
to
be
responsible
for
all
site
acquisition,
site
preparation,
construction,
and
marketing
of
the
project.
The
project
timeline
is
assumed
to
be
the
same
as
for
the
residential
projects,
i.
e.
,
three
years
from
start
to
finish,
and
to
proceed
through
the
same
project
phases.
Similarly,
the
developer­
builder
is
assumed
to
have
several
projects
underway
to
help
balance
cash
flows.
This
assumption
makes
it
possible
to
examine
the
impacts
of
a
three
year
project
on
a
single
year
 
s
cashflow
for
the
affected
business.
Again,
the
particular
data
sources
used
and
inputs
to
this
model
project
are
discussed
further
in
Section
4.2.5.

4­
13
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.2.3.4
Industrial
Development
The
industrial
development
is
assumed
to
be
an
industrial
park
or
a
stand­
alone
manufacturing
facility.
As
with
the
residential
and
commercial
projects,
a
single
developer­
builder
is
assumed
to
be
responsible
for
all
site
acquisition,
site
preparation,
construction,
and
marketing
of
the
project.
The
project
timeline
is
assumed
to
be
the
same
as
for
the
residential
and
commercial
projects,
i.
e.
,
three
years
from
start
to
finish,
and
to
proceed
through
the
same
project
phases.
Similarly,
the
developer­
builder
is
assumed
to
have
several
projects
underway
to
help
balance
cash
flows.
This
assumption
makes
it
possible
to
examine
the
impacts
of
a
three
year
project
on
a
single
year
 
s
cashflow
for
the
affected
business.
A
detailed
discussion
of
data
sources
and
inputs,
which
are
similar
to
those
used
for
the
model
commercial
development,
follows
in
Section
4.2.5.

4.2.4
Cost
Pass
Through
Assumptions
For
modeling
purposes,
EPA
has
analyzed
the
impacts
of
the
regulatory
options
on
each
model
development
project
under
two
extreme
alternatives:
100
percent
cost
pass
through
(
CPT)
and
zero
percent
CPT.
As
explained
in
Section
4.2,
this
allows
EPA
to
show
the
impacts
under
worst­
case
conditions
for
builders
(
zero
percent
CPT)
and
worst­
case
conditions
for
owners
(
100
percent
CPT)
.

Under
the
100
percent
CPT
scenario,
a
fixed
percentage
is
assumed
for
the
developer­
builder
 
s
profit
margin
and
the
model
calculates
the
final
sales
price
that
each
buyer
would
be
asked
to
pay
after
the
compliance
costs
have
been
passed
through.
Under
the
zero
CPT
scenario,
a
fixed
percentage
is
assumed
for
the
developer­
builder
 
s
profit
under
baseline
conditions
and
the
change
in
profit
is
calculated
under
each
regulatory
option,
with
the
sale
price
of
each
housing
unit
remaining
the
same.
Section
4.2.5
contains
further
details
on
the
assumed
profit
levels
and
other
inputs.

4.2.5
Inputs
to
the
Model
Project
Analysis
As
noted
above,
the
representative
projects
take
place
in
three
phases:
land
acquisition,
site
development,
and
construction.
The
process
of
obtaining
options
on
land
to
be
developed
(
a
common
but
not
universal
step
that
occurs
in
the
very
early
stages
of
development)
,
has
been
combined
with
the
4­
14
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
 
land
acquisition
 
activities
for
simplicity.
.
Assumptions
regarding
the
various
costs
that
are
incurred
during
each
phase
of
the
project
are
summarized
in
Table
4­
1
below.

Table
4­
1.
Costs
Incurred
at
Various
Stages
of
a
Residential
Construction
Project
Project
Phase
Cost
Elements
Land
Acquisition
 
Raw
land
(
purchase
or
option)
 
Interest
on
land
acquisition
loan
 
Opportunity
cost
of
capital
Development
 
Engineering
 
Due
diligence
 
Land
development
 
Storm
water
controls
 
Contingency
 
Impact
fees
 
Interest
on
development
loan
 
Opportunity
cost
of
capital
 
Overhead
Building
Construction
 
Lot
cost
(
if
sold
to
a
builder;
includes
land
acquisition
and
development
costs
plus
profit
to
the
developer)
 
Construction
cost
 
Builder
overhead
 
Interest
on
construction
loan
 
Opportunity
cost
of
capital
 
Real
estate
and
marketing
fees
Overall,
EPA
has
used
more
than
two
dozen
different
modeling
parameters,
although
not
all
project
types
encompass
all
of
these
parameters.
Since
the
project
location
is
not
specified,
national
estimates
are
used
where
possible.
Participants
in
the
NAHB
focus
group
meetings
in
Chicago
assisted
EPA
with
identifying
ranges
for
a
number
of
cost
elements
for
the
hypothetical
residential
construction
project,
developing
estimates
for
raw
land
costs,
engineering
costs,
and
construction
costs,
among
others.

Some
of
the
estimates
proposed
during
the
NAHB
Chicago
meetings
are
used
in
the
model
project,

especially
where
actual
national­
level
data
has
not
yet
been
identified,
and
may
reflect
market
conditions
in
that
part
of
the
country.
Table
4­
2
presents
the
assumptions
used
in
the
single­
family
residential
model,
along
with
the
data
source(
s)
used.
Appendix
4A
contains
a
similar
table
outlining
the
data
parameters
and
sources
for
all
four
model
project
types.
A
more
detailed
discussion
of
selected
parameters
and
data
sources
used
for
the
project
models
is
contained
in
Appendix
4B.

4­
15
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
2.
Model
Parameters
and
Data
Sources
Model
Parameter
Source
1,
3,
7.5,
25,
70,
and
200
size
of
parcel,
in
acres
EPA
assumption
$
40,000
cost
of
raw
land,
per
acre
Estimate
from
NAHB
Chicago
focus
groups,
based
on
experience
of
the
Chicago­
area
participants.
See
Appendix
4B
for
further
discussion.

0.33
size
of
lot,
in
acres
Census
Report
C25
(
Characteristics
of
New
Housing,
1999)
reports
a
mean
lot
size
for
new
single­
family
homes
sold
of
12,910
square
feet,
which
represents
a
density
of
close
to
3
lots
per
acre
(
evenly
distributed
with
1/
3
acre
lots)
.
(
The
median
lot
size
is
8,750
square
feet,
which
implies
a
density
of
almost
5
lots
per
acre.
)

2.67
approximate
density
(
number
of
lots
per
acre)
Calculated
based
on
impervious
surface
ratios
from
 
Chesapeake
Bay
Watershed
Impervious
Cover
Results
by
Land
Use
Polygons,
 
to
account
for
impervious
surface
area.
Total
number
of
lots
(
density
x
site
size)
is
rounded
to
nearest
whole
number.

$
2,500
due
diligence
costs,
per
acre
Based
on
$
100,000
in
total
due
diligence
costs
for
a
hypothetical
40­
acre
development
discussed
by
the
NAHB
Chicago
focus
group
participants.
Participants
considered
the
costs
associated
with
all
necessary
environmental
and
engineering
assessments,
usually
done
prior
to
land
acquisition.
During
these
assessments
the
developer
works
to
identify
any
potential
future
problems
or
liabilities.
See
Appendix
4B
for
further
discussion.

$
25,000
land
development
costs,
per
lot
Estimate
from
NAHB
Chicago
focus
groups.
This
figure
includes
any
construction
activities
related
to
land
development
(
e.
g.
,
infrastructure
costs)
.

6%
engineering
costs,
as
percent
of
land
development
costs
Estimate
from
NAHB
Chicago
focus
groups.

10%
overhead
costs,
as
percent
of
development
costs
Estimate
from
NAHB
Chicago
focus
groups.

10%
contingency,
as
percent
of
land
development
costs
(
before
impact
fees)
Estimate
from
NAHB
Chicago
focus
groups.

$
15,000
impact
fees,
per
lot
Estimate
from
NAHB
Chicago
focus
groups.
See
Appendix
4B
for
further
discussion.

7%
real
estate
and
marketing
fees,
as
percent
of
house
sales
price
Estimate
from
NAHB
Chicago
focus
groups.

2,310
average
square
footage
of
new
house
From
Census
Report
C25,
the
average
size
of
new
single­
family
homes
sold
in
1999
and
conventionally
financed
was
2,310
square
feet
$
53.80
cost
of
house
construction,
per
square
foot
From
NAHB
 
s
website,
construction
costs
for
a
generic
single­
family
house
are
$
124,276.
$
124,276
÷
2,310
sq.
.
ft.
=
$
53.80
per
sq.
ft.
(
NAHB
2001b)
.
See
Appendix
4B
for
further
discussion.

65%
percent
of
total
land
cost
that
a
developer
can
finance
for
land
acquisition
Loan­
to­
value
ratio
as
written
in
the
Real
Estate
Lending
Rules.
See
Appendix
4B
for
further
discussion.

4­
16
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
2.
Model
Parameters
and
Data
Sources
Model
Parameter
Source
75%
percent
of
total
development
costs
that
a
developer
can
finance
for
this
stage
Loan­
to­
value
ratio
as
written
in
the
Real
Estate
Lending
Rules.
See
Appendix
4B
for
further
discussion.

80%
percent
of
total
building
construction
cost
that
a
builder
can
finance
Loan­
to­
value
ratio
as
written
in
the
Real
Estate
Lending
Rules.
See
Appendix
4B
for
further
discussion.

7.5%
loan
interest
rate
for
builder/
developer
EPA
estimate.

3
term
of
land
acquisition
loan,
years
EPA
assumption.
Assumes
that
the
land
acquisition
loan
is
paid
off
over
the
life
of
the
project,
which
in
this
case
is
3
years.

1
term
of
development
loan,
years
EPA
assumption.
Assumes
that
the
land
development
loan
term
is
equal
to
the
length
of
the
development
phase
of
the
project,
which
in
this
case
is
1
year.

1
term
of
construction
loan,
years
EPA
assumption.
Assumes
that
the
construction
loan
term
is
equal
to
the
length
of
the
construction
phase
of
the
project,
which
in
this
case
is
1
year.

10%
assumed
baseline
profit
on
land
development
NAHB
Chicago
focus
group
estimated
12­
14
percent;
10
percent
is
an
EPA
assumption.
See
Appendix
4B
for
further
discussion.

10%
assumed
baseline
pre­
tax
profit
on
construction
NAHB
Chicago
focus
groups
estimated
8
to
12
percent
pre­
tax
at
time
of
sale.
R.
S.
Means
also
uses
10
percent
as
a
profit
assumption
in
their
Cost
Data
series.

4.2.6
Model
Project
Analysis
Approach
The
model
project
defines
the
baseline
financial
performance
of
the
residential
subdivision
project
prior
to
the
promulgation
of
the
proposed
rule.
The
baseline
case
is
assumed
to
incorporate
the
costs
of
full
compliance
with
the
existing
Phase
I
and
future
Phase
II
NPDES
storm
water
regulations.

The
model
is
set
up
to
then
assess
the
incremental
impact
of
additional
requirements
imposed
under
the
proposed
effluent
guidelines.

4.2.6.1
Baseline
Model
Project
Performance
Table
4­
3
presents
the
model
project
analysis
under
baseline
conditions,
that
is
prior
to
adding
in
compliance
costs
associated
with
the
proposed
regulatory
requirements.
The
model
estimates
the
final
4­
17
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
sales
price
per
housing
unit
using
the
assumptions
discussed
above.
The
model
incorporates
built­
in
targets
for
profit
margins
on
both
the
development
and
construction
portions
of
the
project,
as
well
as
other
assumptions
that
affect
the
target
sales
price
for
each
unit.
As
seen,
using
the
assumptions
discussed
here,
the
calculated
sales
price
for
each
unit
is
$
283,093.
EPA
notes
that
this
is
higher
than
the
national
mean
sales
price
for
conventionally­
financed
new
single­
family
housing
units,
which
was
$
234,900
in
2000
(
FHFB
2001)
.
EPA
attributes
the
difference
to
assumptions
in
the
model
that
may
reflect
higher­
priced
housing
markets.
Despite
this
likely
bias,
EPA
believes
that
the
model
is
sufficiently
well­
calibrated
to
allow
comparison
of
the
impacts
of
alternative
storm
water
control
costs
on
the
model
project
financials.

It
is
important
to
note
that
while
the
model
recognizes
that
projects
are
developed
over
time,
the
model
does
not
fully
account
for
the
time
value
of
money.
Assumptions
have
been
made
regarding
the
duration
of
each
stage
of
development
in
order
to
determine
the
period
for
any
loans
taken
on
by
the
developer,
i.
e.
,
three
years
for
land
acquisition
loan,
one
year
for
development
loan,
one
year
for
construction
loan.
These
assumptions
influence
the
debt
carrying
costs
incurred
by
the
developer.
What
the
model
does
not
account
for,
however,
is
the
fact
that
some
costs
are
incurred
in
years
two
and
three
(
e.
g.
,
construction
costs
are
incurred
three
years
out)
and
therefore
should
be
discounted
back
to
the
base
year,
which
is
the
year
the
project
starts.
The
discount
factors
for
costs
incurred
two
and
three
years
in
the
future
are
0.873
and
0.816,
respectively,
assuming
a
seven
percent
discount
rate.
This
means
that
any
adjustments
made
to
reflect
the
time
value
of
money
would
reduce
the
overall
project
costs,
but
to
a
fairly
limited
degree.
10
4.2.6.2
Results
of
Model
Project
Analysis
The
model
incorporates
the
costs
of
incremental
regulatory
costs
via
the
shaded
line
item
shown
in
Table
4­
3.
These
engineering
cost
estimates
are
specific
to
both
the
type
of
project
and
project
size.

As
these
costs
are
added
to
the
other
costs
incurred
during
development,
the
financing
requirements
in
the
development
stage
also
increase.
Table
4­
4
shows
the
baseline
project
data
and
illustrates
how
the
10
These
comments
apply
to
the
baseline
costs
incurred
for
project
development,
but
do
not
apply
to
the
incremental
regulatory
costs.
EPA
has
discounted
all
regulatory
costs
that
wold
be
incurred
in
the
future
back
to
the
baseline
year,
in
accordance
with
EPA
and
OMB
guidance
for
conducting
regulatory
impact
analysis.

4­
18
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
project
financials
change
in
response
to
the
regulatory
costs
associated
with
Option
1
under
the
proposed
regulation.

As
seen,
the
incremental
controls
for
the
option
shown
in
the
example,
$
483,
would
raise
the
calculated
sales
price
on
each
housing
unit
from
$
283,093
to
$
283,137,
a
difference
of
$
44.
This
represents
0.02
percent
of
the
baseline
sales
price.
When
the
$
44
per
lot
cost
passed
on
to
the
buyer
is
compared
with
the
contractor
 
s
per­
lot
cost
of
controls
(
i.
e.
,
$
483
÷
20
lots
=
=
$
24.15)
,
the
calculated
cost
 
multiplier
 
for
this
model
project
is
in
the
range
of
1.814.
.
The
cost
multiplier
is
determined
by
taking
the
calculated
increase
in
house
sales
price
(
over
baseline)
and
dividing
it
by
the
actual
per­
lot
cost
of
storm
water
controls
incurred
by
the
builder.
In
this
example,
all
costs
are
passed
on
to
the
buyer
(
100
percent
CPT)
.
In
Chapter
Five,
EPA
presents
the
results
for
all
combinations
or
regulatory
options
under
both
the
100
percent
and
zero
CPT
assumption.
Under
the
zero
CPT
assumption,
the
builder
would
absorb
the
$
24.15
in
compliance
costs
on
each
lot.
The
impact
would
be
reflected
in
a
decrease
in
the
builder
profit,
and
the
sales
price
of
the
housing
unit
would
remain
the
same.

4­
19
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
3.
Baseline
Economic
Model
of
Hypothetical
7.5
Acre
Residential
Development
Project
Cost
Element
Value
Land
Acquisition
(
7.5
acre
parcel)
Raw
land
Interest
on
land
acquisition
Opportunity
cost
of
capital
Land
acquisition
costs
$
300,000
$
29,955
$
16,129
$
346,084
Land
development
(
7.5
acre
parcel)

Engineering
Due
diligence
Land
development
ESC
engineering
costs
Contingency
Impact
fees
Interest
on
development
loan
Opportunity
cost
of
capital
Overhead
[
a
]

Land
development
costs
$
30,000
$
18,750
$
500,000
$
0
$
50,000
$
300,000
$
50,555
$
16,852
$
78,079
$
1,044,235
Land
acquisition
+
land
development
costs
Profit
on
land
acquisition
and
development
Total
 
Land
acquisition
and
development
$
1,390,319
$
154,480
$
1,544,799
Construction
Costs
(
per
lot)

Finished
lot
cost
Construction
cost
Interest
on
construction
loan
Opportunity
cost
of
capital
Builder
overhead
[
a
]

Total
costs
to
builder
Marketing
fees
Profit
on
construction
costs
House
sales
price
(
calculated)
$
77,240
$
124,276
$
12,091
$
3,023
$
18,338
$
234,968
$
19,817
$
28,309
$
283,093
Incremental
Regulatory
Impacts
Change
in
sales
price
per
lot
Costs
as
percent
of
sales
price
Multiplier
$
0
0.00%

0.00
[
a
]
Overhead
in
both
the
development
and
construction
stages
is
calculated
as
total
overhead
(
(
based
on
10
percent
of
development
or
construction
costs)
less
the
opportunity
cost
of
capital.
This
was
done
to
avoid
double­
counting
the
opportunity
cost
of
capital.
Source:
EPA
estimates.
See
also
Table
4­
2
for
model
parameters
and
data
sources.

4­
20
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
4.
Illustration
of
Impact
of
Incremental
Storm
Water
Control
Requirements
on
Model
Project
Under
Proposed
Rule
Option
1
 
100
Percent
Cost
Pass
Through
Scenario
Project
Cost
Element
Baseline
Option
1
Land
Acquisition
(
7.5
acre
parcel)

Raw
land
Interest
on
land
acquisition
Opportunity
cost
of
capital
Land
acquisition
costs
$
300,000
$
29,955
$
16,129
$
346,084
$
300,000
$
29,955
$
16,129
$
346,084
Land
Development
(
7.5
acre
parcel)

Engineering
Due
diligence
Land
development
ESC
engineering
costs
Contingency
Impact
fees
Interest
on
development
loan
Opportunity
cost
of
capital
Overhead
[
a
]

Land
development
costs
$
30,000
$
18,750
$
500,000
$
0
$
50,000
$
300,000
$
50,555
$
16,852
$
78,079
$
1,044,235
$
30,000
$
18,750
$
500,000
$
483
$
50,000
$
300,000
$
50,582
$
16,861
$
78,121
$
1,044,796
Land
acquisition
+
land
development
costs
Profit
on
land
acquisition
and
development
Total
 
Land
acquisition
and
development
$
1,390,319
$
154,480
$
1,544,799
$
1,390,880
$
154,542
$
1,545,422
Construction
Costs
(
per
lot)
Finished
lot
cost
Construction
cost
Interest
on
construction
loan
Opportunity
cost
of
capital
Builder
overhead
[
a
]

Total
costs
to
builder
Marketing
fees
Profit
House
sales
price
(
calculated)
$
77,240
$
124,276
$
12,091
$
3,023
$
18,338
$
234,968
$
19,817
$
28,309
$
283,093
$
77,271
$
124,276
$
12,093
$
3,023
$
18,341
$
235,004
$
19,820
$
28,314
$
283,137
Incremental
Regulatory
Impacts
Change
in
sales
price
per
lot
Costs
per
lot
as
%
of
baseline
sales
price
Multiplier
[
b
]
$
0
0.00%

0.000
$
44
0.02%

1.814
[
a
]
Overhead
in
both
the
development
and
construction
stages
is
total
overhead
(
(
based
on
10
percent
of
development
or
construction
costs)
minus
the
opportunity
cost
of
capital.
This
was
done
to
avoid
double­
counting
of
the
opportunity
cost.
[
b
]
[
Incremental
regulatory
costs
per
lot
x
number
of
lots
]
÷
[
engineering
costs
]
Source:
EPA
estimates.
See
also
Table
4­
2
for
model
parameters
and
data
sources.

4­
21
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.2.7
Model
Nonbuilding
Project
Analysis
As
noted
in
Section
4.2.2,
nonbuilding
construction
such
as
roads,
highways,
bridges,
etc.
s
a
sizeable
activity
but
overall
represents
less
than
two
percent
of
the
total
value
of
construction
completed
each
year.
To
assess
the
potential
impacts
of
the
proposed
rule
on
such
activities
EPA
has
developed
a
model
highway
construction
project
and
used
this
model
to
assess
the
proposed
rule
 
s
costs
and
impacts.

EPA
believes
the
model
captures
and
reflects
the
likely
magnitude
and
significance
of
the
impacts
of
the
proposed
rule
on
the
nonbuilding
construction
sector
overall.

From
the
highway
engineering
literature,
EPA
assumed
that
the
typical
four­
lane
interstate
roadway
is
configured
as
follows:
two
travel
lanes
of
24
feet
each,
one
20­
foot
median
between
the
travel
lanes,
and
10
foot
buffer
on
each
side
of
the
highway
(
Wright,
1996)
.
EPA
assumed
that
the
combined
width
of
the
road
surface,
median,
and
buffers,
88
feet,
represents
the
typical
disturbed
area
for
new
highway
construction.
One
mile
of
new
highway
would
therefore
represent
10.67
acres
in
disturbed
area.
11
To
develop
representative
baseline
costs
for
the
model
highway
project,
EPA
examined
data
from
the
Federal
Highway
Administration
 
s
(
FHWA
 
s)
Highway
Statistics
publication.
Table
FA­
10
(
 
Obligation
of
Federal­
Aid
Highway
Funds
for
Highway
Improvements
 
)
of
the
Highway
Statistics
series
shows
the
number
of
miles,
federal
funds
obligated,
and
total
cost
for
approved
projects
in
a
number
of
highway
improvement
categories
and
roadway
functional
classifications.
Improvement
categories
include
new
construction,
relocation,
widening,
and
bridge
work,
among
others.
Roadway
functional
classifications
include
arterials,
collectors,
and
local
roads,
both
rural
and
urban.
Arterials
are
further
divided
into
interstate,
other
freeways
and
expressways,
other
principal
arterials,
and
minor
arterials.

EPA
aggregated
the
mileage
and
cost
for
the
following
improvement
categories:
new
construction,
relocation,
reconstruction
with
added
capacity,
and
major
widening.
EPA
further
used
only
data
for
urban
interstates
and
other
freeways
and
expressways,
since
other
functional
classifications
may
11
The
disturbed
area
is
88
feet
or
0.0167
miles
wide
(
88
÷
5,280
feet)
)
.
One
mile
of
roadway
therefore
disturbs
0.0167
square
miles,
or
10.67
acres
(
0.0167
x
640
acres/
square
mile)
.

4­
22
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
include
projects
that
do
not
closely
match
the
model
project
characteristics.
Since
highway
and
road
funding
can
fluctuate
from
year
to
year,
EPA
estimated
the
average
miles
and
average
cost
over
the
period
1995­
2000.
Table
4­
5
shows
these
data,
with
all
dollar
values
expressed
in
1997
dollars.
12
Once
all
dollar
amounts
were
expressed
in
constant
year
dollars,
EPA
summed
the
number
of
miles,
federal
funds,
and
total
costs
across
the
two
functional
classifications
and
four
improvement
types
to
generate
an
overall
estimate
of
total
cost
and
miles
affected.
The
total
cost
was
then
divided
by
the
miles
affected
to
generate
a
weighted
average
cost
per
mile
over
all
relevant
improvement
types
and
functional
classifications.
Table
4­
5
shows
the
weighted
average
cost
is
$
5.4
million
per
mile.

Some
caveats
should
be
noted
about
the
data
from
the
 
Highway
Statistics
 
series,
,
and
as
used
in
EPA
 
s
impact
model.
First,
the
dollar
amounts
used
represent
obligated
funds,
rather
than
actual
finished
project
cost.
Therefore,
the
final
project
cost
(
as
well
as
the
actual
payment
to
private
sector
contractors
carrying
out
the
work)
may
be
different
than
the
costs
reported
here
(
Benedict
2002)
.
13
Second,
the
costs
reported
in
Table
FA­
10
of
 
Highway
Statistics
 
are
for
multi­
­
year
projects
(
Benedict
2002)
.
This
does
not
present
a
serious
problem
for
the
analysis
because
the
costs
provide
consistent
estimates
of
project­

level
costs
and
affected
miles
with
which
to
calculate
a
project­
level
cost
per
mile.
The
fact
that
project
completion
may
span
multiple
years
is
not
particularly
relevant
for
this
analysis.
These
caveats
aside,

this
is
the
most
complete
and
well­
documented
set
of
data
available
on
the
cost
for
highway
construction
projects
nationwide.
14
The
results
of
this
analysis
are
presented
in
Sections
5.2
and
5.4.

12
Values
were
converted
to
1997
equivalents
using
data
from
Table
PT­
1
of
the
Highway
Statistics
publication,
 
Price
Trends
for
Federal­
Aid
Highway
Construction
 
(
(
FHWA,
2001a)
.

13
Actual
costs
may
be
higher
due
to
unforeseen
construction
problems.
However,
to
the
extent
this
occurs,
it
will
lessen
the
impacts
of
the
proposed
rule
as
modeled.
Higher
costs
per
mile
will
decrease
the
average
number
of
miles
constructed
per
year.
Fewer
miles
constructed
results
in
fewer
acres
disturbed,
and
therefore
lower
compliance
costs.

14
EPA
previously
has
used
an
estimate
of
$
24.61
million
per
mile
as
an
estimate
for
highway
project
cost
(
weighted
rural
and
urban
average;
FHWA
2001b)
.
This
figure,
from
the
FHWA
Office
of
Program
Administration,
may
reflect
many
improvement
types
and
other
costs
that
EPA
determined
should
not
be
included
in
this
analysis.
It
also
contains
significant
costs
for
land
acquisition,
engineering,
design,
and
other
work
that
would
not
be
paid
to
the
contractor
for
actual
construction.

4­
23
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
5.
Obligation
of
Federal­
Aid
Highway
Funds
for
Selected
Highway
Improvements
and
Functional
Classifications
­
1995
to
2000.
(
Thousands
of
1997
dollars)

Type
of
Improvement
Urban
Total
Interstate
Other
Freeways
and
Expressways
New
Construction
Number
Of
Miles
Federal
Funds
Total
Cost
Cost
per
Mile
175
$
1,231,171
$
1,393,799
$
7,984
277
$
1,226,600
$
1,584,583
$
5,714
452
2,457,771
2,978,382
$
6,591
Relocation
Number
Of
Miles
Federal
Funds
Total
Cost
Cost
per
Mile
17
$
243,936
$
272,084
$
16,062
46
$
213,422
$
270,509
$
5,924
63
457,358
542,593
$
8,668
Reconstruction­
Added
Capacity
Number
Of
Miles
Federal
Funds
Total
Cost
Cost
per
Mile
536
$
2,206,338
$
2,680,896
$
5,001
331
$
1,330,439
$
1,674,158
$
5,062
867
3,536,778
4,355,055
$
5,024
Major
Widening
Number
Of
Miles
Federal
Funds
Total
Cost
Cost
per
Mile
307
$
1,086,999
$
1,273,760
$
4,152
192
$
800,507
$
1,041,609
$
5,429
499
1,887,507
2,315,369
$
4,643
Total
Number
Of
Miles
Federal
Funds
Total
Cost
Cost
per
Mile
1,034
$
4,768,445
$
5,620,539
$
5,434
846
$
3,570,968
$
4,570,860
$
5,406
1,880
8,339,413
10,191,398
$
5,421
Source:
Based
on
FHWA
1996­
2001,
Highway
Statistics
1995­
2000
,
Table
FA­
10.

4­
24
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.3
IMPACTS
ON
MODEL
ESTABLISHMENTS
In
this
section
EPA
presents
the
methodology
used
to
analyze
the
establishment­
level
impacts
of
the
proposed
rule.
Section
4.3.1
outlines
the
impact
analysis
for
a
model
establishment
undertaking
a
model
project.
Section
4.3.2
generalizes
and
extends
this
model
establishment
analysis
to
estimate
the
industry­
wide
closure
impacts
and
employment
losses
due
to
the
proposed
regulatory
options.
Finally,

Section
4.3.3
analyzes
whether
the
proposed
rule
could
present
a
barrier
preventing
new
firms
from
entering
a
market,
thereby
protecting
existing
firms
from
competition.

4.3.1
Model
Establishment
Analysis
This
section
presents
the
inputs
to
the
model
establishment
analysis,
discusses
the
development
of
balance
sheet
and
income
statement
information,
and
develops
the
methodology
for
assessing
potential
regulatory
impacts
in
terms
of
changes
in
model
establishment
financial
ratios.

4.3.1.1
Inputs
to
the
Model
Establishment
Analysis
EPA
began
by
identifying
data
to
characterize
the
typical
financial
conditions
of
model
businesses
in
the
construction
and
development
industry.
This
data
is
used
to
develop
a
financial
model
of
the
firm,
and
to
analyze
the
impacts
of
the
regulatory
options
on
firm
financial
conditions.
The
sections
below
present
the
methodology
used
to
analyze
financial
impacts
on
a
model
firm,
and
then
extend
the
methodology
to
project
facility
closures
and
employment
losses.

The
Bureau
of
the
Census
recently
published
a
profile
of
the
residential
homebuilding
industry
that
allows
analysts
and
others
to
examine
data
in
ways
that
were
not
previously
available
(
Rappaport
and
Cole
2000)
.
In
particular,
the
study
presents
data
by
size
of
builder,
where
the
builder
 
s
size
is
defined
in
terms
of
the
number
of
housing
units
completed
(
previously
such
breakdowns
were
available
only
on
the
basis
of
employment
size
or
revenue
size)
.
EPA
used
this
profile
to
develop
financial
snapshots
of
typical
residential
home
builders.

4­
25
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
From
the
profile,
EPA
determined
the
average
value
of
construction
work
(
revenues)
completed
by
builders
of
various
sizes
based
on
the
number
of
housing
units
started
in
1997.
EPA
combined
the
average
construction
revenue
data
for
such
builders
with
more
detailed
financial
data
on
the
homebuilding
industry
from
Dun
and
Bradstreet
(
2000)
(
D&
B)
.
The
D&
B
data
was
then
scaled
to
the
size
of
the
builder
in
the
Census
profile,
using
the
ratio
of
revenues
to
total
assets.

4.3.1.2
Balance
Sheet
and
Income
Statement
for
Model
Establishment
Table
4­
6
presents
the
balance
sheet
and
income
statement
for
a
model
firm
in
the
single­
family
residential
construction
sector.
EPA
constructed
the
model
firm
financial
statement
using
D&
B
 
s
1999
­

2000
Industry
Norms
and
Key
Business
Ratios,
and
the
Census
special
report
on
the
homebuilding
industry.
The
basic
approach
was
to
calculate
the
ratio
of
key
components
of
the
balance
sheet
and
income
statement
to
net
sales,
and
then
scale
the
value
of
these
components
to
the
size
of
the
model
firm.

The
model
firm
financials
shown
in
Table
4­
6
are
based
on
a
firm
with
$
1.99
million
in
revenues,
which
is
the
average
for
homebuilders
in
the
10
to
24
home
per
year
size
class
(
one
of
the
size
classes
defined
in
the
Census
report)
.

For
the
single­
family
and
multifamily
residential
construction
sectors,
EPA
constructed
a
series
of
model
facilities,
one
for
each
housing
unit
starts
class.
A
financial
statement
for
each
model
firm
was
generated
from
these
revenue
estimates
using
the
method
discussed
above
and
illustrated
in
Table
4­
6.

The
Census
special
study
covers
the
single­
family
and
multifamily
construction
sectors,
but
does
not
cover
the
commercial
and
industrial
building
construction
sectors.
To
construct
model
facilities
for
these
sectors,
EPA
used
1997
Census
of
Construction
data
which
is
available
by
employment
size
class.
First,

EPA
determined
the
employment
class
in
each
sector
corresponding
to
the
median
sized
firm
in
terms
of
revenues.
This
employment
class
became
the
basis
for
a
single
model
facility
for
each
sector.
For
both
the
commercial
and
industrial
sectors,
median
revenues
were
in
the
50
to
99
employee
class.
Within
that
employment
class,
EPA
then
calculated
revenues,
employment,
and
costs
per
establishment
in
order
to
further
characterize
the
model
facility.

4­
26
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
For
the
four
construction
sectors
analyzed,
EPA
used
D&
B
 
s
 
typical
 
establishment
balance
sheet
data
from
the
following
four­
digit
SIC
industries:
15
 
Single­
family
residential
construction:
SIC
1531
 
Multifamily
residential
construction:
SIC
1522
 
Manufacturing
and
industrial
building
construction:
SIC
1541
 
Commercial
and
institutional
building
construction:
SIC
1542
 
Highway
and
street
construction:
SIC
1611
For
the
model
establishment
presented
in
Table
4­
6,
revenues
were
determined
from
Census
data.
All
other
components
are
determined
by
the
percentages
taken
from
the
D&
B
 
typical
 
balance
sheet
for
SIC
1531.
The
ratio
of
revenues
(
net
sales)
to
total
assets
is
used
to
determine
total
assets
(
and
therefore
total
liabilities)
;
the
dollar
value
of
the
remaining
components
are
derived
using
the
percentages
in
the
right
hand
column.

15
Although
most
of
the
data
used
in
this
economic
analysis
is
reported
on
an
NAICS
basis,
the
most
recent
D&
B
report
still
uses
the
SIC
system
for
reporting
purposes.
EPA
believes
the
SIC­
based
data
from
D&
B
can
be
applied
to
the
corresponding
NAICS
industries,
since
there
is
a
high
degree
of
overlap
in
the
industry
definitions.

4­
27
1
2
3
4
5
6
7
8
9
­
­
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
6.
Model
Single­
Family
Residential
Construction
Firm
Financial
Data
Sources:
D&
B
2000;
Census
2000c;
CCH
1999.
Line
Item
Dollars
Percent
Assets
Cash
$
163,390
11.9%

Accounts
Receivable
$
122,199
8.9%

Notes
Receivable
$
9,611
0.7%

Inventory
$
417,399
30.4%

Other
Current
$
303,438
22.1%

Total
Current
Assets
$
1,016,037
74.0%

Fixed
Assets
$
216,938
15.8%
Other
Non­
current
$
140,049
10.2%

Total
Assets
$
1,373,023
100.0%

Liabilities
10
Accounts
Payable
$
112,588
8.2%

11
Bank
Loans
$
23,341
1.7%

12
Notes
Payable
$
201,834
14.7%

13
Other
Current
$
391,312
28.5%

14
Total
Current
Liabilities
$
729,075
53.1%

15
Other
Long
Term
$
162,017
11.8%

16
Deferred
Credits
$
10,984
0.8%

17
Net
Worth
$
470,947
34.3%
18
Total
Liabilities
&
Net
Worth
$
1,373,023
100.0%

Operating
Income
19
Net
Sales
$
1,987,009
100.0%

20
Gross
Profit
$
453,038
22.8%

21
Net
Profit
After
Tax
$
23,844
1.2%

22
Working
Capital
$
286,962
4.3.1.3
Methodology
for
Analysis
of
Regulatory
Impacts
on
Model
Establishment
For
each
model
firm,
EPA
examined
the
economic
impacts
of
each
regulatory
option
on
four
different
financial
ratios:
(
1)
Gross
Profit,
(
2)
Current,
(
3)
Debt
to
Equity,
and
(
4)
Return
on
Net
Worth.

Industry
publications
cite
these
financial
ratios
as
particularly
relevant
to
the
construction
industry
4­
28
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
(
Kone,
2000;
Benshoof,
2001)
.
Two
of
the
ratios
examined
are
based
on
operating
income
(
gross
profit,

and
return
on
net
worth)
,
and
two
are
based
on
the
balance
sheet
statement
(
current,
and
debt
to
equity)
.

Based
on
literature
reviews,
industry
focus
group
input,
and
econometric
evidence,
EPA
believes
the
level
of
CPT
to
customers
to
be
high
in
the
construction
industry.
Complete,
or
100
percent
CPT
implies
zero
direct
impacts
on
the
construction
industry.
Complete
CPT
in
the
residential
sector,
for
example,
essentially
results
in
all
compliance
costs
being
capitalized
into
the
cost
of
the
house,
which
is
then
assumed
to
be
paid
for
over
30
years
as
part
of
the
homebuyer
 
s
mortgage.
In
this
analysis,
EPA
has
taken
a
conservative
approach
that
results
in
a
 
worst­
case
 
scenario,
,
and
is
based
on
the
opposite
extreme
 
zero
CPT.
.
That
is,
EPA
assumed
all
compliance
costs
are
borne
by
the
developer­
builder.

EPA
also
examined
more
realistic
scenarios
incorporating
the
effects
of
partial
CPT
on
the
builder.
EPA
used
a
market
model
approach
to
estimate
CPT
(
i.
e.
,
the
ratio
of
the
increase
in
market
price
to
incremental
compliance
costs)
for
each
of
the
four
construction
sectors
analyzed.
EPA
 
s
estimates
of
CPT
range
from
a
low
of
85
percent
for
the
manufacturing
and
industrial
building
sector
to
a
high
of
92
percent
for
the
multifamily
residential
housing
sector.
Assuming
positive
CPT,
builders
incur
compliance
costs
multiplied
by
one
minus
the
CPT
percentage;
the
remaining
costs
are
passed
through
to
customers
in
the
form
of
higher
prices.
16
Thus,
for
each
compliance
cost
estimate,
EPA
examines
impacts
two
ways:
first
assuming
zero
CPT,
second,
assuming
positive
CPT.

EPA
assumes
that
compliance
costs
affect
each
model
firm
 
s
balance
sheet
in
the
following
manner.
Construction
costs
are
typically
financed
with
a
short
term
construction
loan.
The
value
of
the
loan
tends
to
run
about
80
percent
of
the
value
of
the
project,
with
the
developer
providing
the
remainder
of
the
capital.
The
simplified
balance
sheet
presented
in
Table
4­
7
illustrates
how
a
construction
loan
equal
to
$
Q
affects
the
construction
firm
 
s
balance
sheet
if
the
lending
institution
requires
the
builder
to
finance
20
percent
of
the
cost
of
the
loan.

16
Assume,
for
example,
that
the
market
analysis
shows
that
housing
prices
increase
by
$
0.80
of
every
dollar
in
increased
construction
costs
per
unit
built,
then
CPT
is
80
percent.
If
the
proposed
regulation
adds
$
200
in
construction
costs
per
house,
the
builder
incurs
impacts
from
$
40
in
increased
costs
not
offset
by
increased
revenues
[
(
1
­
0.8)
*
$
200
]
,
while
the
house
buyer
pays
an
additional
$
160
(
0.8*
$
200)
for
the
house.

4­
29
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
The
loan
reduces
current
assets
by
the
amount
of
capital
the
builder
is
required
to
pay
but
increases
noncurrent
assets
by
the
total
value
of
the
project;
long
term
debt
is
increased
by
the
amount
of
the
loan
(
0.80Q)
.
The
baseline
balance
sheet
financial
ratios
for
the
model
firm
will
be
calculated
on
the
basis
of
the
center
column,
while
the
post­
regulatory
financial
ratios
will
be
calculated
on
the
basis
of
the
right
hand
column.
The
value
of
Q
was
set
equal
to
the
incremental
capital
compliance
costs
of
the
proposed
rule.
EPA
used
the
same
framework
for
all
four
sectors
analyzed.

Table
4­
7.
Impact
of
Compliance
Costs
on
Developer­
Builder
 
s
Balance
Sheet
Line
item
Baseline
Post
Loan
Current
Assets
$
A
$
A
­
.
20Q
Noncurrent
assets
$
B
$
B
+
Q
Total
Assets
$
A
+
$
B
$
A
+
$
B
+
.
80Q
Current
Liabilities
$
D
$
D
Long
Term
Debt
$
E
$
E
+
.
80Q
Net
Worth
$
F
$
F
Debt
plus
Equity
$
D
+
$
E
+
$
F
$
D
+
$
E
+
$
F
+
.
80Q
Note:
Q
equals
incremental
compliance
costs.

4.3.1.4
Analysis
of
Financial
Ratios
for
Model
Establishment
Few
financial
ratios
have
clearly
defined
critical
values
that
indicate
whether
a
firm
is
performing
well
or
poorly.
Furthermore,
analysts
often
find
that
a
firm
can
perform
well
in
one
financial
category
(
debt
management,
for
example)
,
yet
poorly
in
another
(
perhaps
rate
of
return)
.
Lacking
such
hard
and
fast
rules
for
interpreting
financial
ratios,
analysts
tend
to
emphasize
trends
over
time,
comparisons
among
competitors,
or
comparisons
between
industries,
rather
than
a
single
critical
value
for
any
particular
ratio.
The
sections
below
briefly
describe
the
four
ratios
examined
for
this
analysis.

4­
30
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Gross
Profit
Ratio
The
gross
profit
ratio
measures
the
ratio
of
pretax
operating
profit
to
revenues:

gross
profit
ratio
 
gross
profit
 
(
net
sales
&
operating
costs)
net
sales
net
sales
Gross
profits
are
line
item
20
on
the
model
firm
balance
sheet
and
income
statement
(
Table
4­
6)

while
net
sales
are
line
item
19.
This
ratio
measures
the
decline
in
pretax
operating
income
relative
to
the
firm
 
s
volume
of
business.
Under
the
worst­
case
scenario
(
zero
CPT)
,
the
post
compliance
gross
profit
ratio
for
the
model
firm
would
be:

gross
profit
ratio
 
(
net
sales
&
operating
costs
&
pre
&
tax
compliance
costs)
net
sales
An
increase
in
compliance
costs
decreases
the
value
of
the
gross
profit
ratio;
the
firm
is
relatively
worse
off.

Return
on
Net
Worth
Return
on
net
worth
measures
the
rate
of
return
from
the
firm
relative
to
the
owner
 
s
investment:

return
on
net
worth
 
net
profit
after
tax
net
worth
Net
profit
after
tax
is
line
item
21
on
the
model
firm
balance
sheet
and
income
statement
(
Table
4­
6)
while
net
worth
is
line
item
17.
Should
the
rate
of
return
on
this
line
of
business
fall
too
much,
then
investors
have
better
opportunities
for
their
capital;
they
would
start
investing
their
capital
in
other
industries
instead
of
construction,
and
the
construction
industry
would
contract.
Compliance
costs
reduce
net
profit,
and
therefore
reduce
return
on
net
worth:

4­
31
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
return
on
net
worth
 
(
net
profit
after
tax
&
post
&
tax
compliance
costs)
net
worth
EPA
multiplied
compliance
costs
by
one
minus
the
effective
tax
rate
to
estimate
post­
tax
compliance
costs.
To
determine
the
effective
tax
rate,
EPA
assumed
taxable
income
was
equal
to
gross
profit
(
line
item
20
on
Table
4­
6)
;
EPA
used
Federal
corporate
tax
rates
plus
the
average
state
corporate
tax
rate
(
6.6
percent)
for
the
specified
level
of
taxable
income.
Note
that
return
on
net
worth
is
a
much
more
sensitive
ratio
than
the
other
ratios
considered
above
because
it
is
calculated
on
a
post­
tax
basis.

As
can
be
observed
in
line
item
21,
post­
tax
profits
are
a
much
smaller
percent
of
net
sales
than
gross
profit.

Current
Ratio
The
current
ratio
is
defined
as:

current
ratio
 
current
assets
current
liabilities
Current
assets
are
line
item
6
on
the
model
firm
balance
sheet
and
income
statement
(
Table
4­
6)

while
current
liabilities
are
line
item
14.
The
current
ratio
is
a
liquidity
ratio
that
measures
the
availability
of
cash
and
near
cash
assets
to
meet
short­
term
obligations.
Clearly
if
current
liabilities
exceed
current
assets
(
i.
e.
,
the
current
ratio
is
less
than
one)
,
the
firm
cannot
meet
all
its
short­
term
financial
obligations.
Although
the
current
ratio
has
a
well
defined
critical
threshold,
detrimental
financial
impacts
can
occur
before
the
ratio
falls
below
one.
Again,
using
EPA
 
s
conservative
worst­
case
assumption
to
estimate
the
impact
of
the
proposed
rule
on
the
model
firms
 
s
finances,
the
post­
regulatory
current
ratio
is:

current
ratio
 
(
current
assets
&
.
20
×
pre
&
tax
compliance
costs)
current
liabilities
An
increase
in
compliance
costs
decreases
the
value
of
the
current
ratio;
the
firm
is
relatively
worse
off.

4­
32
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Debt
Management
The
debt
to
equity
ratio
is
a
ratio
that
measures
how
much
a
firm
 
s
financing
has
been
borrowed
from
creditors:

total
debt
debt
to
equity
ratio
 
owner
equity
Total
debt
is
the
sum
of
line
items
14
(
current
liabilities)
,
15
(
other
long
term
liabilities)
,
and
16
(
deferred
credits)
on
the
model
firm
balance
sheet
and
income
statement
(
Table
4­
6)
,
while
owner
equity
is
line
item
17
(
net
worth)
.
The
debt
to
equity
ratio
presents
amount
of
capital
borrowed
relative
to
that
supplied
by
the
owners.
If,
for
example,
the
debt
to
equity
ratio
is
1.9,
then
$
1.90
has
been
borrowed
for
every
$
1
of
capital
provided
by
the
owners.
If
the
debt
to
equity
ratio
becomes
too
high,
creditors
would
be
reluctant
to
lend
further
capital
unless
the
owners
provide
more
equity.
Incremental
compliance
costs
mean
that
the
builder
would
increase
long
term
debt
by
the
amount
of
the
loan
(
0.80
×
capital
cost)
)
.
Thus
the
post
compliance
debt
to
equity
ratio
is
calculated
as:

debt
to
equity
ratio
 
(
total
debt
%
0.
80
×
pre
&
tax
compliance
costs)
net
worth
An
increase
in
compliance
costs
increases
the
value
of
the
debt
to
equity
ratio
and
the
firm
is
relatively
worse
off.

4.3.1.5
Compliance
Cost
Inputs
into
Financial
Ratio
Analysis
EPA
estimated
engineering
compliance
costs
based
on
project
size,
climatic,
geographical,
and
other
characteristics.
To
project
economic
impacts
using
these
costs,
EPA
determined
the
costs
incurred
by
each
model
establishment,
then
converted
these
compliance
costs
to
costs
per
establishment
based
on
the
following
formula:

costs
per
establishment
 
(
costs
per
acre)
×
(
acres
per
start)
×
(
starts
per
establishment)

4­
33
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
EPA
estimated
average
compliance
costs
per
acre
based
on
project
size.
These
are
a
weighted
average
of
engineering
costs
by
environmental
region
(
see
section
4.4.3
for
details
of
the
weighted
average
of
compliance
costs
per
acre
calculation,
and
section
4.6.2.
for
discussion
of
regional
characteristics
and
compliance
costs)
.

For
the
single­
family
residential,
commercial,
and
manufacturing
construction
sectors,
the
estimated
number
of
units
started
per
establishment
is
essentially
identical
to
the
number
of
buildings
started.
For
the
multifamily
residential
construction
sector,
however,
Census
reports
the
number
of
units
started,
but
each
building
contains
a
number
of
units.
EPA
therefore
estimated
the
average
number
of
units
per
building
to
convert
units
started
to
buildings
started.

Using
data
from
1999
and
2000,
EPA
examined
the
number
of
units
built
in
various
building
classes
(
e.
g.
,
35,500
units
in
buildings
containing
2
to
4
units,
48,000
units
in
buildings
containing
5
to
9
units)
to
construct
a
weighted
average
(
U.
S.
Census
Bureau
2000b)
.
Assuming
the
midpoint
of
each
building
class
interval
represents
the
average
number
of
units
per
building
in
each
class
(
e.
g.
,
apartment
buildings
in
the
2
to
4
units
per
building
class
contain
an
average
of
3
apartments
per
building)
,
EPA
divided
total
units
per
class
by
the
midpoint
of
the
class
to
estimate
the
number
of
multi­
unit
buildings
in
each
class.
EPA
then
calculated
a
weighted
average
of
units
per
building
using
the
class
midpoints
weighted
by
the
estimated
number
of
buildings
constructed
in
each
class.
Using
this
approach,
EPA
estimated
an
overall
average
of
10.8
units
per
multi­
family
residential
building
nationwide.

EPA
used
a
variety
of
sources
to
estimate
average
acres
per
start.
For
single­
family
residential
construction,
EPA
based
its
estimate
of
acres
per
start
on
the
median
lot
size
from
the
Census
report
Characteristics
of
New
Housing
(
U.
S.
Census
Bureau,
2000a)
.
For
multifamily
residential,
commercial,

and
industrial
sectors,
EPA
combined
data
on
the
typical
 
building
 
footprint
from
R.
.
S.
Means
(
2000)

with
the
ratio
of
building
footprint
to
site
size
from
the
Center
for
Watershed
Protection
(
CWP,
2001)
to
estimate
average
acres
per
start.

For
the
model
highway
and
street
construction
contractor,
EPA
used
data
from
Dun
&

Bradstreet,
the
1997
Census
of
Construction
,
and
the
1995­
2000
editions
of
the
Federal
Highway
Administration
 
s
(
FHWA
 
s)
Highway
Statistics
publication.
EPA
used
1997
Census
data
to
construct
a
model
highway
and
street
construction
establishment
based
on
median
revenues
for
establishments
in
4­
34
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
NAICS
234110.
Using
the
same
methodology
EPA
developed
distributions
of
financial
ratios
for
Dun
&

Bradstreet
data
for
SIC
1611
(
highway
and
street
construction)
.
To
estimate
the
number
of
acres
disturbed,
and
hence,
total
establishment
compliance
costs,
ERG
estimated
miles
of
highway
constructed
per
year
by
dividing
model
establishment
revenues
by
the
estimated
cost
per
mile
constructed,
$
5.4
million,
which
was
derived
in
Table
4­
5.
17
4.3.2
Extension
of
Model
Facility
Analysis
to
Project
Industry
Closures
EPA
extended
the
model
facility
framework
described
here
to
project
closures
and
employment
losses
resulting
from
the
proposed
regulation.
The
primary
analysis,
based
upon
analysis
of
financial
ratios,
is
presented
in
Section
4.3.2.1.
EPA
also
conducted
a
sensitivity
analysis,
comparing
the
results
of
the
primary
analysis
to
an
estimate
of
closures
and
employment
losses
using
an
alternative
approach
based
on
cashflow
changes.
This
alternative
approach
is
outlined
in
Section
4.3.2.2.
The
results
of
the
primary
analysis
are
in
Section
5.5,
while
the
sensitivity
analysis
is
presented
in
Appendix
5B.
Before
explaining
these
methodologies,
however,
EPA
first
presents
information
on
how
the
number
of
affected
establishments
and
employees
was
determined
for
use
in
this
analysis.

4.3.2.1
Estimation
of
Affected
Establishments
and
Employment
The
proposed
rule
contains
three
regulatory
options,
each
of
which
would
apply
to
sites
of
varying
sizes.
Option
1
applies
to
sites
of
one
acre
or
larger,
Option
2
applies
to
sites
of
five
acres
or
larger,
and
Option
3
(
no
regulation
option)
applies
to
all
sites.
To
accurately
reflect
the
number
of
entities
affected
under
each
option,
EPA
has
adjusted
the
closure
and
employment
loss
methodology
to
account
for
the
number
of
establishments
affected
under
each
option.
This
section
describes
the
process
used
to
make
these
adjustments.

EPA
again
used
data
from
the
Census
special
study
of
the
home
building
industry
(
Rappaport
and
Cole,
2000)
to
obtain
the
number
of
establishments
by
housing
unit
starts
class.
EPA
concluded
that
17
As
described
in
Section
4.2.7,
EPA
estimated
that
one
mile
of
highway
will
disturb
10.67
acres
of
land.

4­
35
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
this
data
provided
the
best
source
for
estimating
the
number
of
establishments
and
employees
potentially
affected
under
each
option.
Using
the
estimated
density
of
2.67
single­
family
housing
units
per
acre
(
see
Table
4­
2)
,
establishments
starting
between
one
and
four
single­
family
housing
units
per
year
were
excluded
under
Option
1
because
they
are
unlikely
to
disturb
more
than
one
acre
on
a
given
project.
18
Establishments
in
both
the
1­
4
and
5­
9
housing
unit
starts
per
year
categories
were
similarly
excluded
under
Option
2,
since
the
maximum
number
of
housing
units,
nine,
equates
to
only
3.3
acres.
19
This
makes
it
unlikely
many
builders
in
these
size
classes
disturb
more
than
five
acres
on
an
individual
project
basis.
The
Census
report
estimates
that
50,661
single­
family
builders
start
between
one
and
four
housing
units
per
year,
while
another
12,708
builders
start
between
five
and
nine
units
per
year.
EPA
further
concluded
that
1,904
multifamily
builders
starting
between
two
and
nine
multifamily
units
per
year
are
unlikely
to
disturb
more
than
five
acres
on
a
given
project,
and
excluded
these
from
the
universe
of
establishments
potentially
affected
under
Option
2.
Affected
employment
is
determined
in
the
same
manner
as
affected
establishments.
The
Census
study
reports
the
number
of
employees
in
each
housing
unit
start
category,
and
this
number
is
subtracted
as
above
under
each
option.

The
adjustments
above
were
made
for
the
residential
construction
industries
only.
There
are
two
reasons
for
this:
(
1)
the
Census
special
study
only
covers
single­
family
and
multifamily
residential
construction
establishments;
and
(
2)
EPA
believes
that
commercial
and
industrial
building
establishments
are
overall
more
likely
to
disturb
five
acres
or
more
during
the
course
of
each
project.
Therefore,
no
adjustments
are
made
to
the
nonresidential
building
establishment
and
employment
counts.

Table
4­
8
shows
the
establishment
count
adjustment
for
each
option,
while
Table
4­
9
shows
the
adjustment
to
employment.

18
Using
the
density
of
2.67
units
per
acre,
four
housing
units
per
year
equates
to
a
maximum
of
1.5
acres.
This
makes
it
unlikely
a
large
percentage
of
establishments
in
the
1­
4
housing
units
per
year
category
disturb
more
than
one
acre
at
a
time
on
a
regular
basis.

19
Again,
this
would
be
the
maximum
land
area
disturbed
in
a
year.
The
maximum
disturbed
on
an
individual
project
could
be
even
less.

4­
36
­
­

­
­
­
­

­
­
­
­

­
­

­
­
­
­

­
­
­
­
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
8.
Number
of
Establishments
in
the
Construction
and
Development
Industry
Adjusted
for
Regulatory
Option
Coverage
Industry
Baseline
[
a
]
Option
1
Option
2
Adjustment
for
1
acre
exclusion
Adjusted
Number
Adjustment
for
5
acre
exclusion
Adjusted
Number
Single­
family
84,731
(
50,661)
34,070
(
12,708)
21,362
Multifamily
4,603
4,603
(
1,904)
2,699
Commercial
39,810
39,810
39,810
Industrial
7,742
7,742
7,742
Potentially
affected
establishments
136,886
86,225
71,613
[
a
]
Previously
adjusted
for
remodeling
establishments
and
land
development
establishments.
.
See
Section
2.3.5
for
discussion
of
this
adjustment.
Figures
may
not
add
to
totals
due
to
rounding.
Source:
Rappaport
and
Cole
(
2000)
and
EPA
estimates.

Table
4­
9.
Employment
in
the
Construction
and
Development
Industry
Adjusted
for
Regulatory
Option
Coverage
Industry
Baseline
[
a
]
Option
1
Option
2
Adjustment
for
1
acre
exclusion
Adjusted
Number
Adjustment
for
5
acre
exclusion
Adjusted
Number
Single­
family
340,874
(
128,940)
211,933
(
41,940)
169,993
Multifamily
35,160
35,160
(
6,064)
29,096
Commercial
549,567
549,567
549,567
Industrial
148,861
148,861
148,861
Potentially
affected
employees
1,074,462
945,521
897,517
[
a
]
Previously
adjusted
for
remodeling
establishments
and
land
development
establishments.
.
See
Section
2.3.5
for
discussion
of
this
adjustment.
Figures
may
not
add
to
totals
due
to
rounding.
Source:
Rappaport
and
Cole
(
2000)
and
EPA
estimates.

4­
37
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.3.2.2
Closure
and
Employment
Impacts
Based
on
Financial
Ratio
Analysis
To
assess
the
impacts
on
firm
closures,
EPA
first
selected
a
criterion
for
determining
when
a
facility
is
considered
 
impacted
 
by
the
proposed
rule.
.
As
discussed
above,
financial
ratios
rarely
have
well­
defined
thresholds
that
correlate
with
financial
health
or
distress.
On
previous
effluent
guidelines
(
e.
g.
,
MP&
M)
,
EPA
has
defined
the
critical
value
for
financial
stress
as
that
value
of
a
financial
ratio
that
defines
the
poorest
performing
25
percent
of
firms
(
i.
e.
,
the
lowest
quartile)
.
EPA
then
assumes
that
a
facility
is
financially
stressed
if
its
pre­
regulatory
financial
ratio
lies
above
the
lowest
quartile
for
that
ratio,
but
its
post­
regulatory
ratio
falls
in
that
lowest
quartile
range.
20
To
estimate
the
number
of
establishments
in
each
industry
that
would
be
financially
distressed
by
the
proposed
regulation,
EPA
first
approximated
a
cumulative
distribution
function
for
each
financial
ratio
based
on
D&
B
data.
Figure
4­
1
illustrates
the
current
ratio
cumulative
distribution
function
for
SIC
1531,
used
to
analyze
single­
family
residential
construction.
The
baseline
curve
represents
the
pre­

regulatory
cumulative
distribution
function.
This
curve
indicates
that
25
percent
of
establishments
have
a
current
ratio
below
1.1
(
1.1
thus
becoming
the
critical
value
for
determining
financial
distress)
,
25
percent
of
establishments
have
a
current
ratio
greater
than
1.1
but
less
1.4
(
the
median)
,
25
percent
have
a
current
ratio
greater
than
1.4
but
less
than
2.9,
and
25
percent
have
a
current
ratio
greater
than
2.9.
21
20
For
example,
according
to
D&
B,
25
percent
of
establishments
in
SIC
1531
have
a
current
ratio
less
than
1.1,
and
75
percent
have
a
current
ratio
greater
than
1.1.
If
an
establishment
 
s
pre­
regulatory
current
ratio
is
greater
than
1.1,
but
its
post­
regulatory
current
ratio
is
less
than
1.1,
EPA
would
classify
the
firm
as
financially
distressed.

21
The
minimum
and
maximum
values
for
the
current
ratio
are
not
provided
by
D&
B.
For
completeness
EPA
selected
 
reasonable
 
values
to
represent
the
end
points
of
the
curve.
.
This
has
no
effect
on
the
analysis
because
the
lowest
and
highest
ranges
are
not
used
in
the
analysis.

4­
38
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
EPA
then
calculates
the
post­
regulatory
current
ratio
for
the
quartile
values.
This
shifts
the
cumulative
distribution
function
for
the
current
ratio
to
the
left.
Using
the
post­
regulatory
curve
in
this
example,
approximately
40
percent
of
establishments
now
have
current
ratios
less
than
or
equal
to
the
critical
value
of
1.1.
Thus,
about
15
percent
of
establishments
in
this
sector
incur
incremental
financial
Figure
4­
1
Pre
­
and
Post­
regulatory
Cumulative
Distribution
Function
for
Current
Ratio
SIC
1531:
Ope
rative
Builde
rs
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
Current
Ratio
Probability
Baseline
Postregulatory
Critical
Value
distress
due
to
compliance
costs
(
i.
e.
,
40
percent
below
1.1
on
the
post
regulatory
curve
minus
25
percent
below
1.1
in
the
baseline)
.

4­
39
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
A
firm
that
exhibits
or
experiences
financial
distress
under
a
single
measure
of
financial
performance
would
not
necessarily
always
shut
down.
Therefore,
EPA
constructed
similar
cumulative
distribution
functions
for
the
debt
to
equity,
and
return
on
net
worth
ratios,
then
estimated
the
probability
of
incremental
financial
distress
under
each
measure.
22
To
assess
the
economic
achievability
of
the
proposed
rule,
EPA
assumes
that
the
probability
of
establishment
closure
due
to
incremental
compliance
costs
is
equal
to
the
average
probability
of
incremental
financial
distress
under
each
of
the
three
financial
ratios:
current,
debt
to
equity,
and
return
on
net
worth.
Multiplying
this
probability
by
the
number
of
establishments
in
the
sector,
EPA
obtains
an
estimate
of
the
number
of
establishments
projected
to
close
due
to
the
proposed
regulation.
Intuitively,
EPA
is
making
an
implicit
assumption
that
establishments
incurring
financial
distress
under
one
ratio
are
also
incurring
distress
under
the
other
two
ratios.
If
an
establishment
is
distressed
under
multiple
measures
of
financial
health,
it
is
highly
likely
the
establishment
will
close.
23
Employment
losses
are
estimated
by
multiplying
the
number
of
establishments
projected
to
close
by
the
average
number
of
employees
per
establishment.

Finally,
to
project
sector­
wide
impacts,
EPA
aggregated
closure
and
employment
impacts
over
all
combinations
of
model
establishments
and
project
sizes
examined.
Thus,
closures
for
a
single
sector
are
calculated
as
a
weighted
average
where
the
weights
are
determined
by:
(
1)
the
relative
frequency
of
establishments
represented
by
each
model
in
the
sector,
and
(
2)
the
relative
frequency
of
a
particular
project
size
among
all
projects
performed
by
the
sector.
EPA
also
adjusted
the
universe
of
affected
establishments
to
reflect
the
regulatory
coverage
of
each
option.
Thus,
for
Option
1
(
which
applies
to
sites
of
one
acre
or
greater)
EPA
excluded
establishments
in
the
1­
4
housing
starts
category
on
the
assumption
that
few
of
these
small
builders
are
likely
to
disturb
more
than
one
acre
per
project.
Similarly,

where
Option
2
would
apply
to
sites
of
five
acres
or
more,
EPA
excluded
establishments
in
both
the
1­
4
22
D&
B
does
not
provide
quartile
values
for
the
gross
profit
ratio.

23
A
strict
interpretation
of
this
implicit
assumption
would
result
in
EPA
always
selecting
the
smallest
probability
of
incremental
financial
distress
from
among
the
three
measure.
However,
EPA
determined
this
was
not
analytically
desirable
because
the
results
would
always
be
determined
by
the
least
sensitive
measure
of
distress.
Therefore,
EPA
selected
an
average
of
the
three
probabilities
to
measure
closure
rates.
Note
that
in
reality,
establishments
may
incur
distress
under
one
ratio,
but
not
under
another,
thus
being
less
likely
to
close.
It
is
possible
that
the
set
of
establishments
incurring
distress
under
the
current
ratio,
for
example,
is
completely
separate
from
the
set
of
establishments
incurring
distress
under
the
debt
to
equity
ratio.
However,
EPA
has
no
information
on
which
to
base
an
estimate
of
such
joint
probabilities.
Assuming
the
sets
of
establishments
incurring
distress
are
identical
results
in
a
more
conservative
estimate
of
closures.

4­
40
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
and
5­
9
housing
starts
class.
Assuming
the
national
average
of
density
of
2.67
houses
per
acre
(
see
Table
4­
2)
a
five
acre
site
would
support
an
average
of
13.3
housing
units.

4.3.2.3
Closure
and
Employment
Impacts
Based
on
Cashflow
Analysis
As
a
check
on
the
financial
ratio­
based
approach
to
projecting
establishment
closure
impacts,

EPA
developed
a
cashflow
model
and
constructed
a
statistical
distribution
of
establishments
around
each
representative
model.
This
allowed
EPA
to
estimate
the
probability
that
establishments
would
have
insufficient
cashflow
to
afford
the
estimated
compliance
costs.

Modern
financial
theory
states
that
an
investment
should
not
be
undertaken
if
cashflow
is
expected
to
be
negative
after
the
investment
is
undertaken
(
Brealy
and
Myers,
1996;
Brigham
and
Gapenski,
1997)
.
In
the
context
of
this
proposed
rule,
if
compliance
costs
exceed
cashflow,
then
post­

regulatory
cashflow
would
be
negative.
Under
these
circumstances
EPA
projects
that
the
establishment
would
close;
EPA
has
used
this
standard
for
projecting
establishment
closures
for
a
number
of
past
effluent
guidelines
(
e.
g.
,
Transportation
Equipment
Cleaning,
Industrial
Laundries,
Iron
and
Steel)
.

Basing
the
cashflow
analysis
on
the
model
facilities
only
means
that
all
establishments
represented
by
a
particular
model
would
be
projected
to
remain
open
if
the
model
establishment
earns
cashflow
exceeding
compliance
costs,
and
all
would
close
if
the
model
establishment
 
s
cashflow
is
less
than
estimated
compliance
costs.
In
reality,
the
model
establishment
represents
a
family
of
establishments,
some
with
greater
cashflow
than
the
model,
some
with
less
cashflow
than
the
model.

Thus,
there
is
some
probability
that
establishments
would
close
due
to
compliance
costs
even
if
the
model
establishment
 
s
cashflow
exceeds
compliance
costs.
By
developing
a
probability
distribution
for
each
model
establishment
 
s
cashflow
with
known
mean
and
variance,
EPA
can
estimate
this
probability.

Multiplying
the
probability
that
compliance
costs
exceed
cashflow
(
i.
e.
,
that
post­
regulatory
cashflow
is
negative)
by
the
number
of
establishments
represented
by
the
model,
EPA
obtains
the
projected
number
of
closures
for
that
option.
To
develop
the
cashflow
distribution,
EPA
first
estimated
the
mean
and
variance
of
cashflow
associated
with
each
model
establishment.
EPA
based
its
estimate
of
mean
cashflow
on
the
1997
Census
of
Construction.
EPA
calculated
average
revenues,
payroll,
material
costs,

and
work
subcontracted
out
to
others
within
each
model
class
(
starts
class
for
single
and
multifamily
residential,
employment
class
for
commercial
and
manufacturing
sectors)
by
dividing
each
Census
value
4­
41
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
by
the
number
of
establishments
in
the
class.
EPA
then
estimated
taxable
income
per
model
establishment
as:
revenues
minus
payroll,
material
costs,
and
work
subcontracted
out
to
others.

Adjusting
taxable
income
for
taxes
and
interest
payments
results
in
estimated
model
establishment
cashflow.
EPA
applied
Federal
corporate
tax
rates,
plus
the
average
state
corporate
tax
rate
to
establishment
income.
EPA
assumed
interest
payments
comprise
25
percent
of
taxable
income.

EPA
based
its
estimate
of
the
variance
of
each
model
establishment
 
s
cashflow
distribution
on
the
U.
S.
Small
Business
Administration
 
s
 
births
and
deaths
 
database,
,
a
special
tabulation
prepared
for
SBA
by
Census
(
SBA
1999)
.
EPA
calculated
the
ratio
of
establishment
closures
to
total
establishments
for
the
1989
to
1998
time
period
at
the
four­
digit
SIC
level
from
this
database.
24
Assuming
these
establishments
were
closing
because
their
cashflow
was
less
than
zero,
EPA
used
the
model
mean
and
the
assumption
of
a
normal
distribution
to
estimate
the
variance
for
the
distribution
that
would
result
in
a
probability
of
zero
cashflow
(
or
less)
equal
to
the
closure
rate
estimated
from
the
births
and
deaths
database.

With
estimated
mean,
variance,
and
assumed
distribution
of
cashflow
for
each
model
establishment,
it
is
a
straightforward
exercise
to
estimate
the
probability
of
closure
due
to
the
proposed
rule.
Figure
4­
2
illustrates
how
this
analysis
was
conducted.
The
 
estimated
normal
 
curve
represents
the
distribution
of
a
model
establishment
with
mean
cashflow
of
$
1
million,
and
a
variance
set
so
that
the
probability
of
cashflow
less
than
zero
is
about
17
percent
(
as
determined
from
SBA
 
s
 
births
and
deaths
 
database)
)
.
The
critical
value
is
equal
to
estimated
compliance
costs
 
in
this
example
set
equal
to
$
400,000.
25
Figure
4­
2
shows
that
based
on
this
distribution,
about
27
percent
of
establishments
earn
cashflow
less
than
estimated
compliance
costs.
However,
17
percent
of
establishments
had
negative
cashflow
prior
to
incurring
the
compliance
costs
(
i.
e.
,
the
 
baseline
closures
 
)
.
Therefore,
about
10
percent
of
establishments
in
this
example
would
be
projected
to
close
due
to
the
regulation
(
e.
g.
,
27
24
Note
that
the
level
of
detail
in
the
database
was
sufficient
to
allow
EPA
to
estimate
separately
the
closure
rates
for
small
and
large
business
establishments.

25
This
large
estimated
compliance
cost
was
selected
only
for
the
purposes
of
making
the
figure
clear
and
does
not
reflect
actual
anticipated
compliance
costs.

4­
42
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Figure
4­
2
Baseline
Distribution
Function
with
Bounds
for
Facility
Cashflow
0.00
0.25
0.50
0.75
1.00
$
0
$
1,000
$
2,000
$
3,000
$
4,000
Cashflow
(
x
$
1,000)
Probability
Estimated
Normal
Lower
Bound
Upper
Bound
Critical
Value
percent
with
cashflow
less
than
compliance
costs
minus
the
17
percent
with
cashflow
less
than
zero)
.
If
150
establishments
are
in
this
model
class,
and
the
average
employment
per
establishment
is
20
workers
in
this
class,
than
EPA
would
project
15
establishments
would
close
and
300
employees
would
lose
their
jobs
due
to
the
regulation.

Because
of
the
uncertainties
inherent
in
estimating
cashflow
and
variance
for
this
analysis,
EPA
estimated
a
range
of
closure
and
employment
impacts.
EPA
created
upper
and
lower
bounds
to
its
4­
43
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
estimated
cashflow
distribution
by
multiplying
the
distribution
 
s
variance
by
plus/
minus
25
percent.
This
creates
the
bands
observed
around
the
estimated
normal
distribution
in
Figure
4­
2.
Therefore,
although
the
methodology
follows
the
logic
outlined
above,
EPA
reports
an
upper
and
lower
bound
for
projected
closures
based
on
bands
around
the
actual
estimated
variance
of
cashflow.
The
results
of
this
analysis
are
presented
in
Appendix
5A.

4.3.3
Analysis
of
Barriers
to
Entry
Barriers
to
entry
are
typically
assumed
to
occur
if
the
cost
of
complying
with
a
regulation
substantially
increases
the
firm
start­
up
costs.
For
example,
if
a
rulemaking
required
that
all
facilities
invest
substantially
in
a
wastewater
treatment
system,
then
an
entrepreneur
might
be
discouraged
from
starting
an
enterprise.
The
increased
capital
cost
serves
as
a
barrier
to
new
entry
to
the
industry.

The
situation
in
the
construction
industry
is
somewhat
different
than
that
outlined
above.
In
terms
of
the
capital
expense
needed
to
start
a
firm,
the
proposed
rule
has
little
direct
impact.
The
proposed
rule
does
not
require
a
firm
to
purchase
and
install
any
capital
equipment,
and
thus
the
level
of
capital
expenditures
required
to
start
up
a
firm
are
not
directly
affected
by
the
proposed
rule.

Landis
(
1986;
see
section
2.4.1.4.2
for
details)
identifies
two
significant
classes
of
barrier
to
entry
specific
to
the
construction
industry
that
are
not
related
to
capital
equipment:
(
1)
entry
costs
to
participate
in
a
given
market
(
e.
g.
,
local
development
fees
or
abnormally
high
land
costs)
,
and
(
2)
input
cost
differentials
(
e.
g.
,
the
new
entrant
must
pay
a
higher
price
for
inputs
than
existing
firms)
.
These
barriers
to
entry,
however,
also
appear
unaffected
by
the
proposed
rule.
To
the
extent
that
either
of
these
barriers
already
exist
in
any
given
market,
they
would
not
be
differentially
impacted
by
the
proposed
rule.

As
the
model
establishment
analysis
shows,
the
proposed
rule
might
increase
borrowing
to
finance
building
projects.
This
could
affect
a
potential
industry
entrant
indirectly
in
that
it
may
need
marginally
more
start­
up
capital
in
order
to
obtain
the
somewhat
larger
short­
term
construction
loan
to
undertake
a
project.
Once
again,
however,
the
new
entrant
would
still
face
essentially
the
same
requirements
that
existing
firms
face
to
secure
a
loan.
Thus,
new
entrants
should
not
be
differentially
4­
44
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
affected
by
the
proposed
rule
in
such
a
way
that
they
would
be
unable
to
compete
effectively
with
existing
firms.

To
examine
the
potential
for
barriers
to
entry,
EPA
calculated
the
ratio
of
estimated
compliance
costs
to
each
model
establishment
 
s
current
assets
and
total
assets.
If
these
ratios
are
small,
then
EPA
concludes
that
the
proposed
rule
would
have
little
effect
on
the
ability
of
a
new
entrant
to
find
financing
for
a
project.
Note
that
in
this
analysis
EPA
compares
total
compliance
costs
to
assets.
This
step
probably
overestimates
impacts.
It
is
more
likely
that
a
new
entrant
would
need
to
provide
only
20
percent
of
the
incremental
compliance
costs
and
would
obtain
the
remaining
80
percent
from
conventional
construction
loan
financing
sources
(
see
Section
4.3.1.3)
 
as
would
an
existing
firm.
.

4.4
NATIONAL
COMPLIANCE
COSTS
As
noted
above,
EPA
developed
engineering
costs
for
four
categories
of
land
use
(
single­
family
residential,
multifamily
residential,
commercial,
and
industrial)
and
six
project
size
categories
(
1,
3,
7.5,

25,
70,
and
200
acres)
.
Estimates
of
the
national
costs
of
the
effluent
guidelines
regulations
are
obtained
by
multiplying
the
per­
acre
costs
developed
for
each
land
use
and
size
class
combination
by
the
number
of
acres
of
each
type
estimated
to
be
developed
each
year;
taking
into
account
the
applicability
of
each
option
in
terms
of
site
size.

Estimates
of
the
number
of
acres
developed
nationally
per
year
are
available
from
the
U.
S.

Department
of
Agriculture
 
s
(
USDA
 
s)
National
Resources
Inventory
(
NRI)
.
This
source
does
not,

however,
identify
the
type
of
development
or
subsequent
nature
of
the
land
use,
nor
the
distribution
of
acreage
by
site
size.
The
following
sections
describe
the
NRI
estimates
and
EPA
 
s
approach
to
distributing
the
developed
acreage
by
type
of
development
and
site
size.

4.4.1
National
Estimates
of
Disturbed
Acreage
The
NRI,
a
program
of
the
USDA
 
s
Natural
Resources
Conservation
Service,
is
designed
to
track
changes
in
land
cover
and
land
use
over
time.
The
inventory,
conducted
every
five
years,
covers
all
non­

4­
45
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Federal
land
in
the
United
States
(
75
percent
of
the
U.
S.
total)
.
The
program
captures
land
use
data
from
some
800,000
statistically
selected
locations.
From
1992
to
1997,
an
average
of
2.24
million
acres
per
year
was
converted
from
nondeveloped
to
developed
status
(
USDA,
2000)
.
Table
4­
10
shows
the
allocation
of
this
converted
land
area
by
type
of
land
or
land
cover.
As
seen,
land
previously
classified
as
forest
land
accounted
for
41.9
percent
of
the
total,
while
land
previously
classified
as
cropland
accounted
for
25.6
percent
and
land
previously
classified
as
pastureland
accounted
for
17.4
percent.
No
further
breakdown
by
type
of
converted
land
use
is
available.

EPA
assumes
that
some
of
the
2.24
million
acres
converted
from
an
undeveloped
to
developed
state
each
year
would
be
exempt
from
the
requirements
of
the
proposed
rule
due
to
small­
site
or
low­
soil­

loss­
potential
waivers.
Based
on
the
engineering
analysis
of
sites
likely
to
be
eligible
for
such
waivers,

EPA
has
reduced
the
acreage
subject
to
active
construction
controls
to
2.18
million
acres
(
U.
S.
EPA,

2002)
.

In
the
following
section
EPA
develops
estimates
of
the
distribution
of
this
acreage
by
type
of
development
and
by
project
size.
EPA
also
estimates
the
amount
of
acreage
potentially
excluded
from
coverage
under
the
site
size
exclusions
specified
for
Option
1
and
Option
2
(
i.
e.
,
below
one
and
below
5
acres,
respectively)
.
With
the
resulting
estimates
of
acreage
distributed
by
project
type
and
size
class,

EPA
can
then
apply
the
appropriate
per­
acre
engineering
costs
to
obtain
estimates
of
national
costs.

4.4.2
Distribution
of
Acreage
by
Project
Type
To
allocate
the
NRI
acreage,
EPA
has
estimated
the
distribution
of
acres
developed
by
type
of
development
in
the
following
way.
In
the
first
step,
EPA
multiplied
the
number
of
building
permits
issued
annually
by
estimates
of
the
average
site
size.
Thus
for
single­
family
residential
construction,

EPA
multiplied
the
number
of
new
single­
family
homes
authorized
by
building
permit
by
the
average
lot
size
for
new
single­
family
construction.
Estimates
for
other
types
of
construction
are
based
on
extrapolations
from
the
Census
permit
data
and
EPA
estimates
of
average
project
size.
In
the
second
step,
EPA
adjusts
the
estimates
of
acres
converted
to
reconcile
any
differences
between
the
total
number
of
acres
accounted
for
using
this
approach
and
the
total
acres
developed
estimated
by
the
NRI.
Finally,

4­
46
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
EPA
allocates
the
total
by
type
of
construction,
site
size,
and
region
and
adjusts
each
regional
value
to
an
integer
to
ensure
that
only
whole
sites
are
considered.

Table
4­
10.
Acres
Converted
from
Undeveloped
to
Developed
State
a
(
1992­
1997)

Type
of
land
Acres
Converted
to
Development
1992­
1997
(
000)
annual
average
Percent
contribution
by
type
of
land
Cropland
574.8
25.6%

Conservation
Reserve
Program
land
1.5
0.1%

Pastureland
391.2
17.4%

Rangeland
245.9
11.0%

Forest
land
939
41.9%

Other
rural
land
89.1
4.0%

Water
areas
and
federal
land
1.8
0.1%

Total
2,243.4
100.0%

a
NRI
defines
developed
land
as
a
combination
of
the
following
land
cover/
use
categories
large
urban
and
built­
up
areas
,
small
built­
up
areas
,
and
rural
transportation
land
.
These
are
defined
as
follows:
 
Large
urban
and
built­
up
areas
.
A
land
cover/
use
category
composed
of
developed
tracts
of
at
least
10
acres
 
meeting
the
definition
of
urban
and
built­
up
areas.
b
 
Small
built­
up
areas.
A
land
cover/
use
category
consisting
of
developed
land
units
of
0.25
to
10
acres,
which
meet
the
definition
of
urban
and
built­
up
areas.
b
 
Rural
transportation
land.
A
land
cover/
use
category
which
consists
of
all
highways,
roads,
railroads
and
associated
right­
of­
ways
outside
urban
and
built­
up
areas;
also
includes
private
roads
to
farmsteads
or
ranch
headquarters,
logging
roads,
and
other
private
roads
(
field
lanes
are
not
included)
.
b
Urban
and
built
up
areas
are
in
turn
defined
as:
 
Urban
and
built­
up
areas.
A
land
cover/
use
category
consisting
of
residential,
industrial,
commercial,
and
institutional
land;
construction
sites;
public
administrative
sites;
railroad
yards;
cemeteries;
airports;
golf
courses;
sanitary
landfills;
sewage
treatment
plants;
water
control
structures
and
spillways;
other
land
used
for
such
purposes;
small
parks
(
less
than
10
acres)
within
urban
and
built­
up
areas;
and
highways,
railroads
,
and
other
transportation
facilities
if
they
are
surrounded
by
urban
areas.
Also
included
are
tracts
of
less
than
10
acres
that
do
not
meet
the
above
definition
but
are
completely
surrounded
by
Urban
and
built­
up
land.
Two
size
categories
are
recognized
in
the
NRI:
areas
of
0.25
acre
to
10
acres,
and
areas
of
at
least
10
acres.

Source:
USDA,
2000.

Single­
family
residential
Census
data
indicate
that
in
recent
years
the
number
of
new
single­
family
housing
units
authorized
has
averaged
just
over
1.0
million
units
per
year
(
see
Table
4­
11)
.
As
seen
in
Table
4­
12,
the
average
lot
size
for
new
single­
family
housing
units
is
13,553
square
feet,
or
0.31
acres
(
1
acre
=
43,560
4­
47
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
square
feet)
.
Using
the
average
lot
size,
however,
would
underestimate
the
total
acreage
converted
for
single­
family
residential
projects
because
this
acreage
does
not
include
common
areas
of
developments
that
are
not
counted
as
part
of
the
owner
 
s
lot
 
streets,
sidewalks,
parking
areas,
storm
water
management
structures,
and
open
spaces.

To
account
for
this,
EPA
examined
data
obtained
from
a
survey
of
municipalities
conducted
in
support
of
the
Phase
II
NPDES
storm
water
rule
(
EPA,
1999)
.
This
survey
identified
14
communities
that
consistently
collected
project
type
and
size
data
as
part
of
their
construction
permitting
programs.
26
EPA
 
s
review
of
permitting
data
from
these
communities
covered
855
single­
family
developments
encompassing
18,134
housing
units.
The
combined
area
of
these
developments
was
11,460
acres.
This
means
that
each
housing
unit
accounted
for
0.63
acres
(
11,460
acres
÷
18,134
units
=
=
0.63
acres
per
unit)
.
This
estimate,
essentially
double
the
average
lot
size,
appears
to
more
than
account
for
the
common
areas
and
undeveloped
areas
in
a
typical
single­
family
residential
development.
For
this
reason,

EPA
averaged
the
Census
estimate
of
the
national
average
lot
size
(
0.31
acres)
and
the
Phase
II
NPDES
storm
water
estimate
of
0.63
acres
per
unit
to
arrive
at
an
estimate
of
0.47
acres
per
unit.
This
number
was
multiplied
by
the
average
number
of
single­
family
housing
units
authorized
by
building
permit,
1.04
million,
to
arrive
at
an
estimate
of
490,231
acres
(
see
Table
4­
15)
.

Table
4­
11.
New
Single­
Family
and
Multifamily
Housing
Units
Authorized,
1995­
1997
Year
All
Housing
Units
Single­
Family
Housing
Units
Multifamily
Housing
Units
1995
1,332,549
997,268
335,281
1996
1,425,616
1,069,472
356,144
1997
1,441,136
1,062,396
378,740
1995­
1997
avg
1,399,767
1,043,045
356,722
Source:
Census
2000b.
Series
C40
New
Privately
Owned
Housing
Units
Authorized.

26
The
communities
were:
Austin,
TX;
Baltimore
County,
MD;
Cary,
NC;
Ft.
Collins,
CO;
Lacey,
WA;
Loudoun
County,
VA;
New
Britain,
CT;
Olympia,
WA;
Prince
George
 
s
County,
MD;
Raleigh,
NC;
South
Bend,
IN;
Tallahassee,
FL;
Tuscon,
AZ;
and
Waukesha,
WI.

4­
48
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
12.
Average
and
Median
Lot
Size
for
New
Single­
Family
Housing
Units
Sold,
1995­
1997
Year
Average
Lot
Size
(
Square
Feet)
Median
Lot
Size
(
Square
Feet)

1995
13,290
9,000
1996
13,705
9,100
1997
13,665
9,375
1995­
1997
avg
13,553
9,158
Source:
Census
2000a.
Series
C25
Characteristics
of
New
Housing:

Multifamily
Residential
For
residential
construction
other
than
single­
family
housing,
EPA
divided
the
average
number
of
units
authorized
over
1995­
1997
(
356,722,
from
Table
4­
11)
by
the
average
number
of
units
per
new
multifamily
building.
The
average
number
of
units
per
building
was
obtained
by
examining
the
distribution
of
units
by
unit
size
class
in
Census
data
(
U.
S.
Census
Bureau,
2000b)
.
This
report
shows
the
number
of
units
by
building
size
class
(
2
to
4
units,
5
to
9
units,
10
to
19
units,
20
or
more
units)
.
27
EPA
estimated
the
number
of
buildings
in
each
size
class
(
using
data
for
1999
and
2000)
by
dividing
the
number
of
units
in
each
class
by
the
average
number
of
units.
The
total
number
of
units
were
then
divided
into
the
estimated
number
of
buildings
to
arrive
at
the
average
number
of
units
across
all
building
size
classes.
When
this
was
done,
the
average
number
of
units
was
estimated
to
be
10.8.

EPA
next
examined
data
on
the
average
site
size
for
multifamily
residential
developments.
The
Center
for
Watershed
Protection
reports
estimates
from
one
survey
in
which
the
footprint
for
multifamily
buildings
occupied
an
average
of
15.6
percent
of
the
total
site
(
CWP,
2001)
.
EPA
assumed
that
the
average­
sized
multifamily
building
(
10.8
units)
would
have
two
floors
and
that
each
unit
occupies
the
national
average
of
1,095
square
feet
(
NAHB,
2002)
.
The
total
square
footage
accounted
for
by
living
space
is
thus
11,826
square
feet.
Multiplying
by
a
factor
of
1.2
to
account
for
common
areas
and
other
non­
living
space
(
utility
rooms,
hallways,
stairways)
,
and
dividing
by
2
to
reflect
the
assumption
of
a
2­

27
The
average
number
of
units
was
derived
using
data
for
1999
and
2000,
since
data
for
prior
years
was
not
available
at
this
level
of
building
size
detail.

4­
49
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
story
structure,
EPA
obtained
a
typical
building
footprint
of
7,096
square
feet
(
11,826
x
1.2
÷
2
=
=
7,096)
.

Combining
this
with
the
CWP
estimate
of
the
building
footprint
share
of
total
site
size
(
15.6
percent)
,
the
average
site
size
was
estimated
to
be
42,485
square
feet
(
7,096
÷
0.156
=
=
45,487)
,
or
just
over
one
acre
(
1.04
acres)
.

EPA
compared
the
average
site
size
obtained
using
this
approach
with
data
from
the
14
community
study
referenced
above.
That
study
 
s
review
of
permitting
data
identified
286
multifamily
developments
covering
a
total
of
3,476
acres.
The
average
site
size,
12.1
acres,
is
considerably
higher
than
that
obtained
above.
EPA
has
no
indication
that
the
permits
reviewed
in
these
communities
are
for
projects
of
a
larger
than
average
size.
For
purposes
of
this
analysis,
EPA
has
taken
the
midpoint
of
the
estimates,
6.5
acres,
as
the
average
size
of
multifamily
projects.
This
number
was
multiplied
by
the
average
number
of
multifamily
housing
developments
authorized
by
building
permit,
35,672,
to
arrive
at
an
estimate
of
231,868
acres
(
see
Table
4­
15)
.

Nonresidential
construction
EPA
lacked
current
data
on
the
number
of
nonresidential
construction
and
development
projects
authorized
annually
because
the
Census
Bureau
ceased
collecting
data
on
the
number
of
permits
issued
for
such
projects
in
1995.
EPA
therefore
used
regression
analysis
to
forecast
the
number
of
nonresidential
building
permits
issued
in
1997,
based
on
the
historical
relationship
between
residential
and
nonresidential
construction
activity
(
see
Section
4.5.3)
.
Using
this
approach,
EPA
estimates
that
a
total
of
426,024
nonresidential
permits
were
issued
in
1997.
These
represent
a
variety
of
project
types,

including
commercial
and
industrial,
institutional,
recreational,
as
well
as
nonresidential,
nonbuilding
projects
such
as
parks
and
road
and
highway
projects.

EPA
first
combined
a
number
of
nonresidential
project
types
into
a
larger
 
commercial
 
category,
which
included
hotels
and
motels,
retail
and
office
projects,
and
religious,
public
works,
and
educational
projects.
28
EPA
 
s
reasoning
for
including
the
latter
categories
under
the
commercial
category
28
The
commercial
category
included:
hotels/
motels,
amusement,
religious,
parking
garages,
service
stations,
hospitals,
offices,
public
works,
educational,
stores,
other
nonresidential
buildings.

4­
50
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
is
based
on
engineering
judgment
that
storm
water
management
practices
would
be
similar
across
each
project
type.
The
total
estimated
number
of
commercial
permits
in
1997
was
254,566
(
59.7
percent
of
the
nonresidential
total)
.
EPA
retained
the
industrial
category,
which
totaled
12,140
permits
(
2.8
percent)
,
separately.
Storm
water
management
practices
for
such
sites
generally
differ
from
those
for
commercial
or
residential
sites.
The
residual,
159,318
permits
(
37.4
percent)
,
are
nonbuilding,

nonresidential
projects
that
include
parks,
bridges,
roads,
and
highways.
EPA
accounts
for
these
projects
in
the
steps
described
further
below.

For
the
commercial
and
industrial
categories,
EPA
reviewed
the
project
size
data
collected
from
the
14­
community
study
referenced
earlier
(
EPA,
1999)
.
This
study
identified
817
commercial
sites
occupying
5,514
acres
and
115
industrial
sites
occupying
689
acres.
The
average
site
size
is
6.75
and
5.99
acres,
respectively.

EPA
also
reviewed
estimates
from
CWP
(
2001)
on
the
average
percent
of
commercial
and
industrial
sites
taken
up
by
the
building
footprint.
These
percentages,
19.1
and
19.6
respectively,
were
multiplied
across
the
model
project
site
sizes
of
1,
3,
7.5,
25,
70,
and
200
acres
to
estimate
the
size
of
building
on
each
site,
assuming
single­
story
buildings
in
each
case.
These
estimates
are
shown
in
Table
4­
13.

Table
4­
13.
Average
Building
Square
Footage
Project
Size
(
Acres)
Commercial
Industrial
1
8,320
8,555
3
24,960
25,666
7.5
62,400
64,164
25
207,999
213,880
70
582,397
598,863
200
1,663,992
1,711,037
Estimates
were
obtained
by
multiplying
the
site
size
in
square
feet
by
the
percentage
of
the
site
estimated
to
be
occupied
by
the
building
footprint,
based
on
data
from
CWP
(
2001)
.
Source:
EPA
estimates.

4­
51
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
As
seen
in
the
table,
the
average
building
size
corresponding
to
the
6­
to
7­
acre
sites
estimated
from
the
14­
community
study
are
in
the
60,000
square
feet
range.
EPA
next
examined
R.
S.
Means
(
2000)
,
which
provides
cost
data
for
 
typical
 
commercial
and
industrial
buildings.
.
As
part
of
the
cost
data,
Means
identifies
the
typical
range
of
building
sizes
based
on
a
database
of
actual
projects.
Table
4­

13
shows
the
typical
size
and
size
range
for
a
variety
of
building
types
that
would
fall
into
either
the
commercial
or
industrial
categories.
While
some
of
the
building
types
correspond
with
the
estimated
average
of
60,000
square
feet,
these
appear
high
for
other
categories,
such
as
low­
rise
office
and
supermarkets,
warehouses,
and
elementary
schools.
EPA
believes
generally
that
there
are
more
small
projects
than
large
ones.
As
a
result,
EPA
inferred
that
this
approach
would
suggest
an
average
building
size
of
25,000
square
feet,
which
implies
an
average
site
size
of
3
acres,
based
on
Table
4­
14.

Table
4­
14.
Typical
Building
Sizes
and
Size
Ranges
by
Type
of
Building
Building
Category/
Type
Typical
Size
(
Gross
Square
Feet)
Typical
Range
(
Gross
Square
Feet)

Low
High
Commercial
­
Supermarkets
20,000
12,000
30,000
Commercial
­
Department
Store
90,000
44,000
122,000
Commercial
­
Low­
Rise
Office
8,600
4,700
19,000
Commercial
­
Mid­
Rise
Office
52,000
31,300
83,100
Commercial
­
Elementary
a
41,000
24,500
55,000
Industrial
­
Warehouse
25,000
8,000
72,000
a
For
purposes
of
this
analysis
EPA
combines
a
number
of
building
types,
including
educational,
under
the
commercial
category.
Source:
R.
S.
Means
(
2000)
.

To
reconcile
the
estimates
obtained
from
the
two
approaches,
EPA
has
taken
the
midpoint
of
the
estimates.
For
commercial
development,
EPA
assumes
an
average
site
size
of
4.87
acres
(
the
average
of
6.75
and
3.0
acres)
and
for
industrial
development
EPA
assumes
an
average
site
size
of
4.50
acres
(
the
average
of
5.99
and
3.0
acres)
.

The
resulting
average
project
sizes
were
then
multiplied
by
the
estimated
number
of
commercial
and
industrial
permits
to
obtain
an
estimate
of
the
total
acreage
developed
for
these
project
categories.

4­
52
­
­
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
15
shows
the
results
of
this
 
bottom­
up
 
approach
to
estimating
the
number
of
acres
of
land
developed.
The
overall
estimate
of
the
amount
of
land
developed
is
2.01
million
acres
per
year.

Residential
single­
family
development
accounts
for
24.4
percent
of
the
total,
multifamily
development
for
11.5
percent
of
the
total,
commercial
for
61.4
percent,
and
industrial
for
2.7
percent.

Table
4­
15.
National
Estimates
of
Land
Area
Developed
Per
Year,
Based
on
Building
Permit
Data
Type
of
Construction
Permits
Average
Site
Size
a
Acres
Disturbed
Number
Pct.
of
Total
Number
Pct.
of
total
Residential
Single­
family
1,043,045
77.5%
0.47
490,231
24.4%

Multifamily
35,672
2.7%
6.5
231,868
11.5%

Nonresidential
Commercial
b
254,566
18.9%
4.9
1,234,645
61.4%

Industrial
12,140
0.9%
4.5
54,630
2.7%

Total
1,345,423
100.0%
2,011,374
100.0%

a
For
single­
family
residential,
this
is
the
average
of
the
average
lot
size
for
new
construction
in
1999
(
Census
1999)
and
the
average
obtained
in
EPA
(
1999)
.
For
all
other
categories,
the
site
sizes
are
EPA
assumptions
based
on
representative
project
profiles
contained
in
R.
S.
Means
(
2000)
and
the
14­
community
survey
conducted
in
support
of
the
Phase
II
NPDES
storm
water
rule
(
EPA
1999)
.
See
also
Tables
4­
7
and
4­
8.
b
A
number
of
project
types
were
grouped
together
to
form
the
 
commercial
 
category,
,
including:
hotels/
motels,
amusement,
religious,
parking
garages,
service
stations,
hospitals,
offices,
public
works,
educational,
stores,
other
nonresidential
buildings.

The
estimate
of
total
acreage
developed,
2.01
million
acres,
can
be
compared
with
the
estimate
provided
by
the
NRI.
From
Table
4­
10,
NRI
estimates
that
a
total
of
2.24
million
acres
are
converted
from
undeveloped
to
developed
status
each
year.
As
noted
above,
some
acreage
would
not
be
covered
by
the
proposed
rule
or
site
size
limitations
due
to
waivers.
The
estimated
acreage
subject
to
the
proposed
rule
2.18
million
acres.
29
EPA
considers
the
estimate
of
2.01
million
acres
(
Table
4­
15)
to
be
close
to
the
estimates
obtained
from
NRI.
Areas
not
accounted
for
in
EPA
 
s
estimates
include
those
converted
as
a
result
of
29
This
is
technically
the
acreage
covered
under
Option
1,
which
affects
sites
of
one
acre
or
more
in
size.
Estimates
of
the
acreage
covered
under
Option
2,
which
affects
sites
of
five
acres
or
more,
are
made
in
Section
4.4.4.

4­
53
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
road,
highway,
bridge,
park,
monument,
and
other
nonbuilding
construction
projects.
30
EPA
has
not
developed
engineering
costs
applicable
to
these
types
of
projects,
but
assumes
that
the
builders
and
developers
of
these
areas
would
face
compliance
costs,
i.
e.
,
the
acres
should
not
be
excluded
from
the
analysis.
For
the
purpose
of
developing
national
compliance
costs,
therefore,
EPA
has
allocated
the
entire
NRI
acreage,
adjusted
for
waivers,
according
to
the
distribution
shown
in
the
final
column
of
Table
4­
16.
31
Table
4­
16.
National
Estimates
of
Land
Area
Disturbed
Based
on
National
Resources
Inventory
Totals
Type
of
Construction
Acres
Based
on
Permits
Data
Adjusted
NRI
Acreage
b
Number
a
Pct.
of
Total
Residential
Single­
family
490,231
24.4%
533,878
Multifamily
231,868
11.5%
252,182
Nonresidential
Commercial
c
1,234,645
61.4%
1,332,476
Industrial
54,630
2.7%
57,523
Total
2,011,374
100.0%
2,176,058
a
From
Table
4­
15.
b
This
column
distributes
the
total
acreage
estimated
in
NRI
to
be
converted
on
an
annual
basis
(
adjusted
for
waivers)
according
to
the
distribution
by
type
of
development
estimated
through
analysis
of
permits
data.
See
also
Tables
4­
11
through
4­
14.
c
A
number
of
project
types
were
grouped
together
to
form
the
 
commercial
 
category,
,
including:
hotels/
motels,
amusement,
religious,
parking
garages,
service
stations,
hospitals,
offices,
public
works,
educational,
stores,
other
nonresidential
buildings.

4.4.3
Distribution
of
Acreage
by
Project
Size
The
next
step
in
the
national
compliance
cost
analysis
is
to
allocate
the
number
of
acres
in
each
of
the
four
land
use
categories
according
to
project
size.
The
project
size
distribution
is
based
on
the
survey
of
municipalities
conducted
in
support
of
the
Phase
II
NPDES
storm
water
rule
(
EPA,
1999)
.

This
survey
identified
14
communities
that
consistently
collect
project
type
and
size
data
as
part
of
their
30
As
noted
above,
EPA
estimates
there
are
approximately
159,000
such
projects
permitted
each
year.

31
This
distribution
implies
that
the
acres
not
accounted
for
from
the
NRI
(
see
Table
4­
10)
will
be
costed
at
the
weighted
average
cost
across
the
single­
family
residential,
multifamily
residential,
commercial,
and
industrial
categories.

4­
54
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
construction
permitting
programs.
Table
4­
17
shows
the
distribution
by
project
size
for
each
land
use
category.

Following
allocation
to
project
size
class,
EPA
also
allocated
the
acreage
to
one
of
19
eco
­

regions,
based
on
geographical
information
system
(
GIS)
modeling.
Non­
linearity
of
installation
costs
made
it
inaccurate
to
consider
partial
sites.
So,
these
totals
were
adjusted
to
ensure
that
only
whole
sites
would
be
considered
for
each
category
of
type,
site
size,
and
region.
Further
detail
on
this
step
in
the
analysis
can
be
found
in
the
Development
Document
(
U.
S.
EPA,
2002)
.

The
final
step
in
the
national
compliance
cost
analysis
is
to
multiply
the
number
of
acres
in
each
eco­
region,
size
class,
and
land
use
category
by
the
applicable
cost
per
acre.
These
costs
are
shown
in
Chapter
Five.

4.4.4
Estimates
of
Acreage
Covered
by
Option
2
Table
4­
16
above
shows
the
distribution
of
acreage
affected
under
Option
1
of
the
proposed
rule,

which
would
apply
to
sites
of
one
acre
or
larger.
The
additional
acreage
excluded
under
the
site
size
limitations
of
Option
2
(
five
acres)
was
obtained
by
estimating
the
acreage
in
sites
above
one
acre
and
below
five
acres
in
size.
The
3­
acre
size
class
represents
projects
greater
than
1
acre
and
less
than
5
acres.
This
category
was
subtracted
from
the
matrix
of
acreage
by
region,
type,
and
size
class
as
allocated
by
the
GIS.
As
shown
in
Table
4­
17,
the
14­
community
study
(
EPA,
1999)
found
that
6.0
percent
of
acreage
developed
for
single­
family
housing
was
assigned
to
sites
in
the
3­
acre
size
class.

EPA
estimated
that,
after
rounding,
roughly
6.1
percent
of
acreage
converted
to
single­
family
housing
units
would
be
excluded
under
Option
2.
EPA
made
similar
estimates
of
the
acreage
converted
to
multi­

family,
commercial,
and
industrial
uses
that
would
be
excluded
under
Option
2.
Table
4­
18
shows
the
distribution
of
acreage
affected
under
Option
2
of
the
proposed
rule.

4­
55
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
17.
Distribution
of
Permits
by
Site
Size
Site
Size
(
Acres)
No.
of
Permits
Acres
by
Size
Pct.
Acres
by
Size
Single­
Family
Residential
1
266
266
2.3%

3
228
684
6.0%

7.5
138
1,035
9.0%

25
175
4,375
38.2%

70
30
2,100
18.3%

200
15
3,000
26.2%

Total
852
11,460
100.0%

Multifamily
Residential
1
43
43
1.2%

3
100
300
8.6%

7.5
61
458
13.2%

25
71
1,775
51.1%

70
10
700
20.1%

200
1
200
5.8%

Total
286
3,476
100.0%

Commercial
1
266
266
4.8%

3
356
1,068
19.4%

7.5
86
645
11.7%

25
91
2,275
41.3%

70
16
1,260
22.9%

200
0
0
0.0%

Total
815
5,514
100.0%

Industrial
1
39
39
5.7%

3
55
165
23.9%

7.5
10
75
10.9%

25
8
200
29.0%

70
3
210
30.5%

200
0
0
0.0%

Total
115
689
100.0%

4­
56
­
­
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Table
4­
17.
Distribution
of
Permits
by
Site
Size
Site
Size
(
Acres)
No.
of
Permits
Acres
by
Size
Pct.
Acres
by
Size
Total
1
614
614
2.9%

3
739
2,217
10.5%

7.5
295
2,213
10.5%

25
345
8,625
40.8%

70
59
4,270
20.2%

200
16
3,200
15.1%

Total
2,068
21,139
100.0%

Based
on
permitting
data
from
the
following
municipalities
or
counties:
Austin,
TX;
Baltimore
County,
MD;
Cary,
NC;
Ft.
Collins,
CO;
Lacey,
WA;
Loudoun
County,
VA;
New
Britain,
CT;
Olympia,
WA;
Prince
George
 
s
County,
MD;
Raleigh,
NC;
South
Bend,
IN;
Tallahassee,
FL;
Tuscon,
AZ;
and
Waukesha,
WI
(
EPA,
1999)
.
Source:
EPA
estimates.

Table
4­
18.
Estimates
of
Acreage
Affected
Under
Proposed
Rule
Option
2
Type
of
Construction
Acreage
Affected
Under
Option
1
a
Percent
Excluded
Under
Option
2
b
Acreage
Affected
Under
Option
2
Residential
Single­
family
533,878
6.1%
501,100
Multifamily
252,182
8.8%
229,958
Nonresidential
Commercial
c
1,332,476
20.4%
1,061,108
Industrial
57,523
25.7%
42,733
Total
2,176,058
1,834,898
a
From
Table
4­
15.
b
Based
on
analysis
of
site
size
distributions
found
in
EPA
(
1999)
.
c
A
number
of
project
types
were
grouped
together
to
form
the
 
commercial
 
category,
,
including:
hotels/
motels,
amusement,
religious,
parking
garages,
service
stations,
hospitals,
offices,
public
works,
educational,
stores,
other
nonresidential
buildings.
Source:
EPA
estimates.

4­
57
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.4.5
Operation
and
Maintenance
Costs
For
any
incremental
ESC
requirements
triggered
under
Option
2,
EPA
estimated
the
percentage
of
capital
costs
of
each
technology
that
would
be
required
annually
to
operate
and
maintain
the
facilities.

Those
facilities
with
a
limited
useful
life
were
assigned
percentages
sufficient
to
replace
them
at
the
appropriate
time.
These
were
converted
to
costs
per
acre
for
each
option.
The
O&
M
costs
are
assumed
to
be
incurred
for
a
one­
year
period
during
the
active
phase
of
construction.

4.5
IMPACTS
ON
THE
NATIONAL
HOUSING
MARKET
4.5.1
Description
of
National
Housing
Market
Model
EPA
takes
three
complementary
approaches
to
estimating
the
market
impacts
of
the
proposed
rule.
Two
treat
the
nation
as
a
single
market;
the
third
treats
each
city
as
a
distinct
housing
market.
The
first
approach
assumes
all
of
the
costs
of
compliance
with
the
regulation
are
passed
through
to
the
home
buyer.
If
the
home
is
more
costly,
fewer
households
would
be
able
to
qualify
for
a
mortgage
to
purchase
it.
This
change
in
market
size
is
an
indicator
of
the
impact
of
the
proposed
regulation.
In
the
second
approach,
the
costs
of
compliance
shift
the
national
housing
supply
curve
in
a
linear
partial
equilibrium
model.
A
portion
of
the
increased
costs
raises
the
price
of
new
housing
while
the
balance
is
absorbed
by
the
builder.
Higher
prices
and
lower
quantities
change
the
welfare
of
participants
in
the
housing
market.

The
third
approach
estimates
a
linear
partial
equilibrium
model,
like
the
national
model,
for
215
metropolitan
statistical
areas
(
MSAs)
based
on
local
measures
of
residential
construction
activity.
This
approach
measures
changes
in
affordability
in
terms
of
the
Housing
Opportunity
Index
(
HOI)
,
a
well
publicized
measure
of
housing
availability.
The
following
sections
explain
each
model
in
detail.

4.5.1.1
Complete
Cost
Pass
Through
and
Housing
Affordability
Landis
 
(
(
1986)
and
Luger
and
Temkin
 
s
(
2000)
surveys
suggest
that
all
of
the
additional
costs
of
compliance
with
new
storm
water
regulations
would
be
passed
through
to
new
home
buyers
in
the
form
of
higher
prices
for
a
unit
of
a
given
quality.
The
quantity
of
new
housing
built
would
not
change
4­
58
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
because
demand
is
driven
by
demographics
more
than
marginal
price
considerations,
i.
e.
,
demand
is
inelastic,
and
competition
in
supply
is
limited
because
of
oligopolistic
markets
in
many
areas
and
infinitely
elastic
supply
in
others.
An
increase
in
the
price
of
a
home
increases
the
income
necessary
to
qualify
for
a
home
mortgage
to
purchase
the
home,
and
so
reduces
the
number
of
households
able
to
afford
it.
One
measure
of
the
impact
of
the
regulation
is
the
change
in
the
number
of
households
that
can
afford
the
new
home.

EPA
developed
its
market
model
parameters
from
the
previously
described
model
projects,

Census
data,
and
the
housing
economics
literature.
Simple
assumptions
about
expected
proportionate
profit
margins,
borrowing,
and
contingencies
discussed
in
Section
4.2
indicate
that
added
incremental
compliance
costs
are
multiplied
by
a
factor
of
1.5
to
1.8
in
the
final
consumer
price.
Luger
and
Temkin
(
2000)
report
a
compliance
cost
multiplier
of
2
to
6
times
actual
compliance
costs.
The
higher
multiplier
may
reflect
a
tight
housing
market
in
high
growth
regions.
The
median
house
price,
from
the
industry
profile,
is
taken
as
the
baseline
price.
The
median
price,
P0
,
with
the
additional
compliance
costs,
C,

multiplied
by
a
factor
for
added
time
and
borrowing,
m,
equals
the
new
price,
PN
,
which
is
the
starting
point
for
calculating
the
effect
of
the
proposed
regulation
on
affordability,
welfare
measures,
and
other
market
model
results:

P
N
 
P
0
%
mC
(
1)

where:

P
N
 
New
Price
with
ESC
Compliance
Costs
P
0
 
Median
New
Home
Price
m
 
Cost
Multiplier
C
 
ESC
Compliance
Costs
The
monthly
payment
for
principal,
interest,
taxes,
and
insurance
(
PITI)
for
the
new
home
is
based
on
the
new
price:

4­
59
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
FP
N
(
r
PI
 
12
)

(
2)

12
)
&
360
1
&
(
1
%
r
P
N
T
 
t
(
3)
1,
000
P
N
I
 
s
(
4)
1,
000
PITI
 
PI
%
T
%
I
(
5)

where:

PI
 
Monthly
Principal
and
Interest
F
 
Proportion
of
New
Home
Cost
that
is
financed
r
 
Annual
Mortgage
Interest
Rate
T
 
Monthly
Tax
Payment
t
 
Monthly
Tax
Rate
per
Thousand
Dollars
Value
I
 
Monthly
Insurance
Premium
s
 
Monthly
Insurance
Rate
per
Thousand
Dollars
Value
PITI
 
Principal
,
Interest
,
Taxes
,
and
Insurance
Fannie
Mae
guidelines
limit
borrowers
 
PITI
payments
to
no
more
than
28
percent
of
their
gross
income.
.

The
value
for
F,
0.774,
and
r,
0.0752,
the
mortgage
terms,
are
national
averages
for
the
typical
30­
year
fixed
rate,
private
mortgage
in
the
base
period
(
FHFB,
2001)
.
Values
for
t,
$
1/
$
1,000
value,
and
s,

$
0.25/
$
1,000
value,
are
from
a
recent
study
of
housing
affordability
(
Savage,
1999)
.
The
gross
income
necessary
to
qualify
for
the
mortgage
at
the
new
price,
under
this
criterion,
Y,
is
given
by:

4­
60
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Y
 
12
PITI
(
6)
0.
28
Table
4­
19
illustrates
the
calculations
using
Option
2
costs.
In
Chapter
Five,
EPA
uses
this
approach
to
estimate
the
number
of
households
priced
out
of
the
new
housing
market
as
a
result
of
each
regulatory
option
or
combination
of
options.

Table
4­
19.
Change
in
Housing
Affordability
 
Sample
Calculation
Source:
EPA
estimates.
Data
element
Baseline
Option
2
Average
per
lot
cost
difference
from
baseline
Difference
in
cost
per
lot
times
multiplier
$
0
$
0
$
111
$
201
Home
price
$
288,397
$
288,598
Principal
and
interest
Real
estate
taxes
Homeowner'
s
insurance
Total
principal,
interest,
taxes,
and
insurance
$
1,564
$
288
$
72
$
1,924
$
1,565
$
289
$
72
$
1,926
Income
necessary
to
qualify
for
mortgage
Change
in
income
necessary
Number
of
households
shifted
(
thousands)

Percent
change
in
number
of
qualified
households
$
82,472
$
0
0
0.0%
$
82,529
$
58
­
29
­
0.15%

The
change
in
the
number
of
households
who
qualify
for
a
mortgage
to
finance
the
baseline
home
price
but
cannot
afford
the
home
with
the
added
compliance
costs
is
imputed
from
Census
Bureau
statistics
of
household
income.
The
Census
Bureau,
Current
Population
Survey,
reports
the
money
income
of
households
in
21
income
classes
from
zero
to
over
$
100,000
(
U.
S.
Census
Bureau
2000d)
.

Table
4­
20
shows
the
Census
distribution.
Each
income
class,
except
the
top
one,
spans
$
5,000
in
annual
income.
If
households
are
evenly
distributed
within
each
class,
then
a
change
of
$
1,000
from
the
baseline
income
necessary
to
qualify
to
the
new
income
necessary
excludes
one
fifth
of
the
members
of
the
income
class
from
qualifying
for
the
new
mortgage
level.
Since
the
incremental
costs
of
compliance
4­
61
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
are
relatively
small,
the
new
price
usually
falls
within
the
same
income
class
as
the
baseline
price
and
the
number
of
households
per
$
1,000
change
in
price
is
adequate
to
find
the
change
in
number
of
qualifying
households.
If
the
qualifying
income
for
the
baseline
price
is
in
a
different
income
class
than
the
qualifying
income
for
the
new
price,
the
number
of
households
per
$
1,000
change
in
price
in
each
class
is
calculated
and
the
number
of
households
disqualified
calculated
in
parts.

Table
4­
20.
Household
Information
for
Imputing
Changes
in
Ownership
Possibilities
a
Calculated
from
proportion
of
owner­
occupied
to
total
housing
units
multiplied
by
number
of
households
in
income
class.
Source:
Household
Income:
U.
S.
Census
Bureau,
Current
Population
Reports,
P60­
209,
Money
Income
in
the
United
States:
1999,
U.
S.
GPO:
Washington,
2000;
Housing:
U.
S.
Census
Bureau,
American
Housing
Survey
for
the
United
States:
1999,
Table
2­
12
Income
Characteristics
of
Occupied
Units,
http:
/
/
www.
census.
gov/
hhes/
www/
housing/
ahs/
ahs99/
tab212.
html
Current
Population
Survey
American
Housing
Survey
Annual
Household
Income
(
$
1,000)
Households
(
1,000)
Households
That
Own
Home
a
(
1,000)
Percent
Owned
for
Income
Class
Total
Housing
Units
(
1,000)
Owner­
Occupied
Units
(
1,000)

<
5
5­
9
10­
14
15­
19
20­
24
25­
29
30­
34
35­
39
40­
44
45­
49
50­
54
55­
59
60­
64
65­
69
70­
74
75­
79
80­
84
85­
89
90­
94
95­
99
100>
Total
3,010
6,646
7,661
7,482
7,238
6,890
6,381
6,016
5,565
4,958
4,789
4,064
4,112
3,380
2,927
2,903
2,526
2,023
1,736
1,568
12,832
104,707
1,456
3,051
3,906
3,935
3,946
4,000
3,891
3,794
3,875
3,452
3,674
3,118
3,360
2,762
2,392
2,372
2,227
1,784
1,531
1,383
11,674
70,071
48.4%
45.9%
51.0%
52.6%
54.5%
58.1%
61.0%
63.1%
69.6%
69.6%
76.7%
76.7%
81.7%
81.7%
81.7%
81.7%
88.2%
88.2%
88.2%
88.2%
91.0%
66.9%
5,839
6,728
7,780
7,037
7,369
6,867
7,469
5,951
9,778
 
8,184
 
11,985
 
 
 
6,548
 
 
 
11,267
102,802
2,824
3,089
3,967
3,701
4,017
3,987
4,555
3,753
6,808
 
6,278
 
9,793
 
 
 
5,774
 
 
 
10,250
68,796
The
proportion
of
households
in
the
marginal
income
class
that
already
own
their
home
indicates
the
size
of
the
market
possibly
affected.
According
to
the
Census
Bureau
 
s
American
Housing
Survey,
in
4­
62
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
1999,
48.4
percent
of
households
with
less
than
$
5,000
income
owned
their
own
home
while
91
percent
of
those
with
income
over
$
100,000
annually
own
their
home.
Overall,
66.9
percent
of
households
own
their
home.
32
The
rate
of
home
ownership
for
the
larger
income
classes
from
the
housing
survey
was
applied
to
all
of
the
income
classes
of
the
population
survey
within
the
same
range
(
indicated
by
the
ditto
marks
in
Table
4­
20)
.
The
total
number
of
households
with
income
greater
than
that
required
to
qualify
for
the
baseline
home
is
the
total
number
of
households
that
could
afford
the
baseline
home.
Since
this
is
the
group
that
may
be
in
the
market
for
a
new
home,
substantial
changes
in
the
proportion
of
this
group
that
can
afford
it
may
represent
large
changes
in
the
size
of
the
market
for
new
homes
attributable
to
the
construction
and
development
regulation.

4.5.1.2
National
Partial
Equilibrium
Modeling
Another
approach
to
evaluating
the
impact
of
the
proposed
regulation
on
housing
markets
is
based
on
a
household
production
function
partial
equilibrium
model.
Empirical
studies
find
a
highly
elastic
supply
and
a
somewhat
inelastic
demand
for
new
housing
(
DiPasquale,
1999)
.
These
estimated
elasticities
and
the
assumption
that
compliance
costs
of
new
environmental
regulations
result
in
only
marginal
changes
in
prices
and
quantities
allow
the
market
to
be
modeled
with
a
simple
linear
partial
equilibrium
market
model
similar
to
the
ones
used
in
other
recent
EPA
regulations
(
U.
S.
EPA,
2001b)
.

The
modeling
situation
is
similar
to
that
used
by
Montgomery
(
1996)
to
forecast
wood
product
demand.
The
linear
partial
equilibrium
model
can
be
viewed
as
a
reduced
form
of
a
more
complex
structural
model.
We
can
assume,
for
example,
that
all
of
the
instrumental
variables
are
the
same
in
both
the
baseline
and
alternatives,
i.
e.
,
the
regulation
does
not
change
U.
S.
population
growth,
carpenters
 
wages,
wood
product
prices,
and
so
forth.
Montgomery
 
s
(
1996)
modeling
equation
(
equation
12
in
the
paper)
is
simply
a
linear
supply
curve
and
equations
6,
8,
and
10
reduce
to
a
linear
demand
curve.
The
simpler
model
will
provide
the
same
results
as
the
more
complex
structural
model
given
small
marginal
changes
in
costs
and
unchanging
long
run
assumptions.

32
The
American
Housing
Survey
uses
fewer
income
groups
than
the
Current
Population
Survey.

4­
63
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
National
statistics
of
residential
housing
starts
from
the
Census
of
Construction
establish
the
baseline
quantity
for
the
model.
The
baseline
price
is
the
median
new
home
price
derived
from
the
project
cost
model.
This
combination
is
the
baseline
market
equilibrium
where
supply
equals
demand.

To
indicate
highly
elastic
supply,
EPA
assumes
a
price
elasticity
of
supply
of
4.0.
DiPasquale
(
1999)

cites
studies
with
estimates
for
new
housing
supply
elasticity
from
0.5
to
infinity
but
the
majority
of
the
long
run
estimates
are
in
the
3
to
13
range.
Housing
demand
elasticity
is
equally
controversial.
EPA
assumes
a
price
elasticity
of
demand
of
­
0.7
to
indicate
a
somewhat
inelastic
demand
function.

Sensitivity
tests
of
these
assumptions
are
shown
in
Appendix
5B.

Given
a
baseline
equilibrium
point
(
P0
,
Q0
curve.
)
and
these
elasticities,
EPA
identified
a
linear
supply
Q
 
"
%
$
P
(
7)

Where:

Q
 
Number
of
residential
building
permits
issued
P
 
Price
of
new
home
"
 
Intercept
calibrated
from
baseline
equilibrium
 
Q
0
&
$
P
0
$
 
Coefficient
on
price
 
E
s
×
Q
0
P
0
E
s
 
Supply
elasticity
of
new
homes
>
0
A
linear
demand
curve
was
derived
similarly.

4­
64
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Q
 
F
%
(
P
(
8)

Where:

F
 
Intercept
calibrated
from
baseline
equilibrium
 
Q
0
&
(
P
0
(
 
Coefficient
on
price
 
E
d
×
Q
0
P
0
E
d
 
Demand
elasticity
of
new
homes
<
0
EPA
assumes
the
baseline
condition
is
in
equilibrium
so
these
two
equations
are
equal.
The
increased
costs
of
compliance
raise
builders
 
costs
and
shift
the
supply
curve
upward
to
the
left.
.
The
change
in
prices
and
quantities
depends
on
the
relative
slopes
of
the
supply
and
demand
curves.
EPA
chose
to
model
the
increased
costs
as
a
slope­
preserving
change
in
the
supply
curve
intercept,
"
,
rather
than
an
elasticity­
preserving
change
in
slope.
The
new
intercept
is
calculated
as:

"
 
Shocked
intercept
s
(
9)
 
Q
0
&
$
(
P
0
%
ESC)

where
ESC
is
the
per
unit
costs
of
compliance
with
the
proposed
regulation.
The
new
price
is
given
by:

"
S
&
F
P
N
 
(
&
$
(
10)

Equilibrium
prices
and
quantities
are
then
recalculated
using
the
new
price
and
shocked
intercept.

Unlike
the
complete
cost
pass
through
method
described
above,
some
of
the
costs
of
compliance
in
the
partial
equilibrium
model
may
be
absorbed
by
the
builder.
The
proportions
flowing
to
consumers
4­
65
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
and
builders
depends
on
the
relative
elasticities
of
supply
and
demand.
The
literature
suggests
cost
pass
through
rates
are
very
high
in
this
industry
(
DiPasquale,
1999)
.
With
the
supply
and
demand
elasticities
selected
as
representative
of
the
literature,
Es
=
4
and
Ed
=
­
0.7,
the
cost
pass
through
is
85
percent.
Thus,

the
industry
absorbs
15
percent
of
the
costs
of
compliance
and
passes
the
remainder
on
to
home
buyers
as
a
price
increase.

The
partial
equilibrium
model
has
a
number
of
implications
for
the
welfare
of
society.
When
the
supply
curve
shifts
following
introduction
of
incremental
compliance
costs,
consumers
lose
some
of
their
benefits
from
the
product
in
absorbing
those
compliance
costs.
This
results
in
a
loss
of
consumer
surplus.
How
the
consumer
surplus
is
lost
is
irrelevant
from
a
welfare
economics
perspective.

Consumers
may
choose
cheaper
options
in
the
construction
of
their
new
homes
such
as
lower
quality
carpets
or
cabinets.
They
may
accept
less
expensive,
smaller
homes.
Or,
they
may
just
pay
the
higher
price
and
forego
other
spending.
In
any
case,
the
home
would
provide
less
utility
than
it
might
have
without
the
ESC
costs.
Different
choices
would
affect
which
industries
feel
the
impact
in
the
regional
economy.
Changes
in
housing
options
would
impact
builders
and
suppliers.
Decreased
overall
spending
would
impact
a
wide
range
of
consumer
goods
industries.
For
simplicity,
EPA
assumed
that
consumers
would
reduce
other
spending
in
response
to
the
price
change.
The
reduction
in
home
sales
volume
and
consumer
spending
in
other
sectors
reduces
employment
in
construction
and
all
other
parts
of
the
economy.
Indirect
effects
of
the
regulation
on
the
whole
economy
are
estimated
using
Regional
Input­

Output
Modeling
System
(
RIMS)
multipliers
published
by
the
U.
S.
Department
of
Commerce.
The
multiplier
analysis
indicates
the
ultimate
changes
in
gross
domestic
output
and
employment
attributable
to
the
new
regulation.

4.5.1.3
Regional
Partial
Equilibrium
Modeling
and
the
Housing
Opportunity
Index
Each
of
the
approaches
described
above
treats
housing
as
a
national
market
with
the
same
demand
elasticities
applying
across
the
country.
In
reality,
however,
market
conditions
can
vary
widely
from
region
to
region,
state
to
state,
and
city
to
city.
Markets
vary
both
in
the
level
of
activity
and
the
structure
of
the
industry.
Costs
of
compliance
would
undoubtedly
be
easier
to
pass
through
to
consumers
in
a
hot
housing
market
than
in
a
depressed
market.
EPA
 
s
third
modeling
approach
captures
such
4­
66
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
regional
variation
by
setting
up
a
partial
equilibrium
model
for
each
Metropolitan
Statistical
Area
(
MSA)

and
using
statistics
of
the
level
of
activity
in
the
MSA
to
select
the
parameters
of
the
model.

The
Census
Bureau
collects
information
about
housing
starts
as
well
as
the
size
of
the
existing
housing
stock
at
the
MSA
level.
EPA
infers
that
where
housing
built
during
the
1990s
represents
a
large
proportion
of
the
total
current
housing
stock,
the
new
housing
market
is
active
and
demand
would
be
expected
to
be
less
elastic
than
in
areas
with
slower
growth.
As
discussed
above,
the
long
run
supply
of
new
housing
is
assumed
to
be
quite
elastic
overall.
These
facts
provide
the
basis
for
selecting
elasticities
to
represent
housing
markets
at
the
MSA
level.

EPA
developed
separate
partial
equilibrium
models
for
each
MSA.
Like
the
national
models
described
above,
EPA
used
building
permit
and
median
new
home
price
data
to
establish
the
baseline
equilibrium
point
for
each
MSA.
Demand
elasticities
were
estimated
based
on
the
ratio
of
new
housing
units
authorized
to
housing
stock
over
the
period
1990
to
1996
(
Census,
1998)
.
EPA
mapped
regions
where
this
ratio
is
lowest
to
the
most
elastic
estimates
of
demand
found
in
the
literature
and
those
where
the
ratio
is
highest
to
the
least
elastic
demand
elasticity
estimates.
EPA
believes
this
approach
captures
the
relative
differences
in
demand
elasticity
between
active
and
depressed
housing
markets
around
the
country.

Each
MSA
model
is
shocked
with
the
estimated
compliance
costs
for
the
median
new
home
in
the
region.
The
model
then
estimates
changes
in
prices,
quantities,
and
welfare
measures
for
each
MSA.

As
there
are
more
than
200
MSAs,
it
is
not
practical
to
report
all
of
the
individual
results.
Instead,
all
of
the
MSAs
in
a
Census
division
are
averaged
together
to
give
a
sense
of
the
effect
of
compliance
costs
on
each
region
of
the
nation.

Affordability
is
a
significant
concern
for
some
stakeholders.
The
National
Association
of
Home
Builders
(
NAHB)
publishes
the
Housing
Opportunity
Index
(
HOI)
for
180
MSAs.
HOI
measures
the
proportion
of
the
housing
stock
a
family
with
the
median
income
can
afford.
NAHB
compares
the
median
family
income
to
the
actual
distribution
of
home
prices
in
the
MSA.
EPA
does
not
have
access
to
such
detailed
price
information.
Instead,
EPA
assumes
home
prices
are
normally
distributed
about
the
median
with
standard
deviation
of
1.
Thus,
our
rough
HOI
(
RHOI)
is
the
cumulative
probability
of
homes
with
prices
less
than
the
maximum
PITI
that
the
median
income
can
afford.

4­
67
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Algebraically:

30
(
Median
Income
×
0.
28)
e
rt
dt
RHOI
 
Z
(
1,
1)
(
m
0
)
(
11)
Median
Sales
Price
The
numerator
represents
the
present
value
of
the
maximum
PITI
payment
that
the
median
income
can
afford
at
the
prevailing
mortgage
rate,
r,
over
a
typical
30­
year
fixed
rate
loan.
The
denominator
is
simply
the
median
sales
price.
When
this
ratio
is
equal
to
one,
the
median
income
family
can
afford
the
median
sales
price
home
or,
equivalently,
half
the
families
can
afford
the
median
sales
price
home.
The
normal
cumulative
density
function
with
mean
of
one
and
variance
of
one,
is
represented
by
Z(
1,1)
(
@
)
.

Thus,
if
the
median
income
family
can
afford
more
than
the
median
sales
price
home,
the
ratio
will
be
greater
than
one,
and
the
Z(
1,1)
(
@
)
function
will
indicate
the
proportion
of
homes
the
family
can
afford.

For
MSAs
with
HOIs
reported
by
NAHB,
EPA
adjusts
the
variance
of
the
normal
curve
so
that
RHOI
yields
the
NAHB
baseline
HOI
index
(
NAHBHOI)
.
The
variance
scaling
factor
is:

Z
&
1
(
0,
1)
(
RHOI
)
V
 
*
Z
&
1
*
(
12)

(
0,
1)
(
NAHBHOI
)

where
Z(
0,1)
­
1
is
the
inverse
of
the
standard
normal
cumulative
distribution.
Changing
the
variance
of
Z(
1,1)

from
one
to
V
causes
RHOI
to
equal
NAHBHOI
at
the
observed
median
family
income.
In
those
MSAs
where
NAHB
does
not
calculate
HOI,
unadjusted
RHOI
is
reported.
33
To
assess
the
impact
of
the
regulation,
the
adjusted
HOI
is
calculated
with
the
new
sales
price
from
the
market
model.
The
percent
change
in
adjusted
HOI
is
an
indicator
of
the
added
stress
of
compliance
costs
on
the
housing
market.

Like
the
full
pass
through
model
discussed
above,
the
MSA
HOI
model
shows
how
changes
in
costs
affect
home
buyers.
This
approach
has
the
advantage
of
recognizing
local
market
differences
and
33
In
13
MSAs,
the
distribution
of
home
prices
is
so
different
from
normal
that
RHOI
cannot
approximate
NAHBHOI
with
the
variance
adjustment.
These
MSAs
were
deleted
from
the
results.

4­
68
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
applying
them
within
the
model.
Average
HOI
among
MSAs
in
Census
divisions
before
and
after
compliance
costs
are
reported
in
Chapter
Five.

4.5.2
Inputs
to
the
National
Housing
Market
Model
The
analysis
uses
the
average
price
of
the
model
home
worked
out
in
Section
4.2,
$
284,632,
as
a
starting
point.
Buyers
in
2000
financed
an
average
of
77.4
percent
of
the
home
purchase
price
at
an
interest
rate
of
7.52
percent
(
FHFB,
2001)
.
EPA
assumes
a
30­
year
conventional
fixed
rate
mortgage
for
ease
of
calculation.
EPA
also
assumes
a
monthly
real
estate
tax
rate
of
$
1
per
$
1,000
of
home
value
and
insurance
payment
of
$
0.25
per
$
1,000
of
home
value
(
Savage
1999)
.
These
assumptions
are
applied
to
the
revised
home
price
to
derive
an
estimate
of
the
monthly
principal,
interest,
taxes,
and
insurance
(
PITI)

payment
generally
required
to
purchase
a
new
home.

In
Chapter
Five
EPA
uses
this
approach
to
estimate
the
number
of
households
priced
out
of
the
new
housing
market
as
a
result
of
each
regulatory
option.

4.5.3
Multifamily
and
Non­
Residential
Construction
Market
Models
EPA
developed
three
market
models
of
the
multifamily
and
non­
residential
construction
industry.

All
three
are
similar
to
the
residential
regional
partial
equilibrium
model.
They
treat
each
state
as
a
separate
market
with
adjusted
demand
elasticities.
Each
model
produces
estimates
of
changes
in
prices,

quantities,
and
welfare
measures.

The
commercial
market
is
highly
disaggregated
into
regional
markets.
Office
rents
for
similar
buildings
(
Class
A
space)
range
from
$
17/
square
foot/
year
in
Wichita
to
more
than
$
60/
square
foot/
year
in
San
Francisco
(
Grubb
&
Ellis
2001)
.
This
disparity
shows
that
arbitrage
among
markets
is
not
possible
and
space
in
each
area
should
be
considered
a
different
commodity.
Many
real
estate
companies
maintain
data
on
conditions
in
regional
markets.
Typically,
activity
in
the
market
is
measured
in
terms
of
the
vacancy
rate
and
asking
rents.
EPA
developed
a
market
model
for
office
space
similar
to
the
regional
partial
equilibrium
models
developed
for
residential
construction
to
indicate
the
effects
on
commercial
construction.

4­
69
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
The
Census
Bureau
discontinued
collection
of
non­
residential
building
permit
information
in
1994.
To
estimate
non­
residential
building
permits
issued
in
later
years,
EPA
regressed
non­
residential
building
permits
on
residential
building
permits,
the
value
of
non­
residential
buildings
put
in
place
(
VPIP)
,

and
a
time
trend.
Since
the
relationship
among
these
variables
differs
from
state­
to­
state,
regressions
were
estimated
at
the
state
level.
Three
different
regressions
were
estimated.
Several
states
showed
a
distinct
shift
in
building
permits
data
when
the
Census
sample
changed
from
17,000
permit­
granting
jurisdictions
to
19,000
jurisdictions
in
1983.
In
states
where
this
difference
was
apparent,
only
observations
after
1983
were
used
in
the
final
projection.
In
addition
some
states
had
strong
trends
which
were
correlated
with
residential
building
permits.
Since
this
multicollinearity
reduced
the
influence
of
residential
building
permit
data
in
later
projection
years,
a
regression
was
also
estimated
without
the
trend
variable.
The
three
regressions
are:

 
1980­
1994
data;

 
1983­
1994
data;
and
 
1980­
1994
data
estimated
without
the
trend
variable.

Each
regression
was
also
estimated
using
only
data
through
1993
to
test
their
ability
to
forecast
the
next
year
outside
of
the
sample,
i.
e.
,
1994.
The
regression
which
gave
the
best
out
of
sample
projection
to
1994
and/
or
had
the
highest
correlation
coefficient
for
the
state
was
selected
to
be
used
for
that
state
 
s
projection.
Thus,
each
state
projection
uses
the
model
that
best
predicts
its
pattern
of
non­
residential
development.
EPA
allocates
the
number
of
non­
residential
building
permits
estimated
for
each
state
to
commercial,
industrial,
and
other
projects
based
on
the
number
of
permits
issued
for
each
type
of
project
in
the
1994
building
permit
data.
The
commercial
category
is
a
catch­
all
which
includes
public
buildings,

hotels,
amusements,
and
educational
buildings,
in
addition
to
office
and
retail
buildings.
EPA
implicitly
assumes
that
these
projects
would
employ
best
management
practices
that
are
similar
to
those
required
for
office
or
retail
space.
A
separate
category
for
industrial
projects
and
a
third
category
for
non­
building
permits
are
also
allocated
from
the
1994
data.

In
the
partial
equilibrium
model,
the
quantity
of
construction
in
each
category
is
measured
by
the
number
of
building
permits
issued.
Rental
rates,
in
dollars
per
square
foot
per
year,
are
closely
watched
indicators
of
demand
for
commercial
space
and
serve
as
our
price.
Rents
and
activity
reports
for
35
retail
space
markets
around
the
country
from
a
recent
real
estate
marketing
firm
report
(
Grubb
and
Ellis,
2001)

4­
70
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
provide
the
baseline
information
for
the
market
model.
Like
the
ratio
of
new
building
permits
to
housing
stock
in
the
residential
model,
EPA
used
the
activity
reports
to
create
a
scale
of
demand
intensity
which
was
then
used
to
map
to
each
market
an
appropriate
demand
elasticity
from
a
range
of
possible
market
elasticities.

Demand
for
office
and
retail
space
is
relatively
insensitive
to
small
changes
in
price.
Since
non­

residential
construction
activity
tends
to
be
driven
by
interest
rates,
job
growth,
and
locational
factors
rather
than
building
costs,
cost
pass
through
is
very
high.
Huffman
(
1988)
,
for
example,
found
that
impact
fees
were
largely
passed
on
to
end
users
in
the
long
run.
EPA
therefore
applies
a
range
of
elasticities
from
­
0.01
to
­
0.80
to
represent
relatively
inelastic
demand
for
commercial
space.
In
regions
with
many
vacancies,
lessees
can
be
more
sensitive
to
price
so
a
more
elastic
demand
curve
is
used.
In
regions
with
tight
markets,
lessees
have
fewer
options
and
generally
have
little
choice
but
to
pay
the
asking
price,
so
demand
is
less
elastic.
Builders
can
pass
on
a
higher
proportion
of
their
costs
in
tight
markets
than
in
soft
markets.
Even
in
the
softest
market,
however,
83
percent
of
costs
are
passed
through
to
consumers
with
these
assumptions.

The
number
of
non­
residential
building
permits
was
projected
at
the
state
level
while
the
Grubb
and
Ellis
commercial
data
is
from
35
selected
cities.
Since
there
is
insufficient
building
permit
data
to
model
each
city,
EPA
models
each
state
as
a
separate
market
with
the
average
rent
and
activity
rate
for
the
cities
within
the
state
representing
the
state
market.
The
assumption
is
reasonable
where
state
office
and
retail
markets
are
concentrated
in
one
city,
or
one
city
is
representative
of
general
statewide
market
conditions.
The
assumption
is
less
defensible
in
large
states
with
many
population
centers,
since
market
conditions
may
vary
from
city
to
city
within
such
states.
Almost
half
of
the
states
were
not
represented
by
cities
in
the
Grubb
and
Ellis
data.
For
these
states,
the
average
rent
and
activity
values
for
cities
within
the
Census
division
containing
the
state
were
used
to
indicate
state
market
conditions.

The
industrial
space
market
model
is
similar
to
the
commercial
model.
It
uses
the
vacancy
rate
for
industrial
space
as
an
indicator
of
market
activity
and
the
rental
rate
for
warehouse
space
as
the
price.

Industrial
space
users
are
considerably
more
mobile
and
price
sensitive
than
commercial
or
residential
space
consumers
so
demand
for
industrial
space
is
more
elastic.
The
range
used
in
this
analysis
is
­
0.2
to
­
1.5.

4­
71
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
The
multifamily
housing
market
model
uses
the
same
format
as
the
non­
residential
models.
The
activity
measure
is
the
proportion
of
the
housing
stock
built
in
the
1990
to
1996
time
period.
Separate
price
series
or
rental
rates
for
multifamily
housing
are
not
reported
so
the
single­
family
housing
prices
were
taken
as
a
near
substitute.
EPA
assumed
that
elasticities
of
demand
are
also
similar
to
those
for
single­
family
housing.

The
multifamily
and
non­
residential
models
apply
equations
7
through
10
above
to
estimate
supply
and
demand
curves.
Compliance
costs
are
converted
to
the
same
units
as
the
rental
rates,
given
the
model
project.
The
increase
in
cost
shifts
the
supply
curve
to
the
left
and
upward.
Market
results
may
be
reported
in
terms
of
changes
in
rents
and
building
permits,
as
well
as
changes
in
consumer
and
producer
surplus,
and
can
be
converted
to
changes
in
indirect
employment
using
the
RIMS
II
multiplier.

4.6
NET
ECONOMIC
IMPACTS
Environmental
regulations,
while
imposing
costs
on
the
regulated
industry,
may
also
provide
a
stimulus
to
firms
that
make
or
install
environmental
controls,
or
provide
other
services
related
to
reglatory
compliance.
The
output
and
jobs
created
by
new
spending
in
the
environmental
industry
offsets,
to
some
extent,
the
loss
of
output
in
the
affected
industry.
In
the
case
of
C&
D,
the
same
firms
that
now
do
much
of
the
site
preparation
work
would
also
be
charged
with
implementing
ESCs,
and
likely,
conducting
ESC
certification
and
inspection.
Contractors
would
be
hired
to
build
sedimentation
ponds,
improve
grades,
and
construct
any
incremental
ESCs
triggered
by
the
proposed
regulation.
Thus,

while
the
regulation
is
costly
in
one
sense,
much
of
that
cost
flows
directly
back
into
the
industry,

stimulating
more
activity,
output,
and
employment.

4.6.1
Welfare
Effects
In
terms
of
the
welfare
effects
discussed
in
Section
4.5.1,
both
the
consumer
and
producer
surpluses
are
converted
to
costs
of
production.
Consumer
surplus
represents
income
that
would
have
been
used
by
consumers
to
purchase
other
products
or
for
enjoyment.
Producer
surplus
would
have
flowed
to
the
owners
of
the
firm
and
probably
to
consumption
or
investment
in
other
industries.
Both
4­
72
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
quantities
thus
flow
out
of
the
construction
sector.
Only
to
the
extent
that
the
compliance
costs
that
would
be
absorbed
are
greater
than
the
sum
of
production
loss
plus
the
consumer
surplus
lost
would
the
regulation
result
in
a
net
increase
in
activity
in
the
construction
sector.

Both
the
loss
and
the
gain
in
employment
are
estimated
by
applying
RIMS
II
multipliers
to
the
changes
in
output
derived
from
the
market
models.
Construction
activity
generates
approximately
37.8
jobs
per
million
dollars
of
output
while
general
consumer
spending
generates
only
27.3
jobs
per
million
dollars
of
spending.
Shifting
spending
from
consumers
to
construction
would
increase
overall
employment.
As
some
readers
may
be
interested
in
both
the
losses
and
gains
in
construction
employment,
both
aspects
are
shown
in
Chapter
Five,
as
well
as
the
loss
in
employment
from
lost
consumer
spending.

4.6.2
Regional
Impacts
For
this
analysis,
EPA
examines
the
potential
impacts
to
specific
regions
by
assessing
whether
the
proposed
C&
D
regulations
could
have
community
or
regional
level
impacts.
Such
impacts
could
alter
the
competitive
position
of
the
C&
D
industry
across
the
nation
or
lead
to
growth
or
reductions
in
C&
D
activity
(
in­
or
out­
migration)
in
different
regions
and
communities.
Traditionally,
the
distribution
of
C&
D
establishments
has
echoed
the
general
regional
distribution
of
U.
S.
population,
with
some
parts
of
the
industry
responding
to
short
or
long
term
shifts
in
population
distribution.

EPA
does
not
expect
that
the
proposed
C&
D
regulations
would
have
a
significant
impact
on
where
construction
and
development
takes
place,
or
the
regional
distribution
of
construction
and
development
activity.
On
the
one
hand,
regulatory
costs
would
be
lower
in
regions
with
lower
rainfall
and
reduced
soil
erodibility.
This
would
tend
to
favor
projects
being
developed
in
such
regions.
At
the
same
time,
however,
a
project
located
in
a
low
rainfall
region
would
rarely
be
a
perfect
substitute
for
the
same
project
in
a
high
rainfall
region.
So
many
factors
go
into
a
locational
decision
that
few
homeowners,
companies,
or
industrial
firms
are
likely
to
make
their
decision
on
where
to
build
based
solely
upon
the
relative
costs
of
storm
water
controls.
Thus,
EPA
does
not
expect
the
proposed
C&
D
regulations
to
significantly
influence
the
prevailing
pattern
of
construction
and
development
activity.

4­
73
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
EPA
 
s
market
model
accounts
for
regional
market
influences
by
creating
state
and
MSA
level
partial
equilibrium
models
for
each
sector.
These
models
are
used
to
quantify
the
regional
impacts
in
terms
of
output
and
employment.
Like
the
national
employment
effects,
state
employment
changes
are
calculated
using
RIMS
II
multipliers.
Regional
multipliers
were
not
available
for
this
analysis
so
the
national
multipliers
were
used.
The
results,
therefore,
overstate
the
employment
impacts
within
the
region
but
indicate
the
effect
of
changes
within
the
region
on
the
nation
as
a
whole.
Tables
summarizing
state
impacts
are
included
in
Chapter
Five.

4.6.3
International
Trade
As
part
of
its
economic
analysis,
EPA
has
evaluated
the
potential
for
changes
in
U.
S.
trade
(
imports,
exports)
of
construction
and
development
related
goods
and
services.
A
significant
component
of
the
U.
S.
construction
and
development
industry
operates
internationally,
and
in
addition
numerous
foreign
firms
operate
in
the
U.
S.
EPA
judged,
however,
that
the
potential
for
U.
S.
construction
and
development
firms
to
be
differentially
affected
by
the
proposed
rule
is
negligible.
The
proposed
rule
would
be
implemented
at
the
project
level,
not
the
firm
level,
and
would
affect
only
projects
within
the
U.
S.
All
firms
undertaking
such
projects,
domestic
or
foreign,
would
be
subject
to
the
proposed
rules.

U.
S.
firms
doing
business
outside
the
U.
S.
would
not
be
differentially
impacted
compared
to
foreign
firms,
nor
would
foreign
firms
doing
business
in
the
U.
S.

The
proposed
rule
may
stimulate
or
depress
demand
for
some
construction­
related
goods.
To
the
extent
that
the
proposed
rule
acts
to
depress
the
overall
construction
market,
demand
for
conventional
construction­
related
products
may
decline.
This
decline
may
be
offset
by
purchase
of
goods
and
services
related
to
storm
water
management.
Overall,
EPA
does
not
anticipate
that
any
shifts
in
demand
for
such
goods
and
services
resulting
from
the
proposed
regulation
would
have
significant
implications
for
U.
S.

and
foreign
trade.

4­
74
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.7
GOVERNMENT
IMPACTS
4.7.1
Administrative
Costs
EPA
has
analyzed
the
administrative
costs
to
governments
associated
with
the
proposed
rule.

EPA
assumes
that
the
majority
of
construction­
related
regulatory
costs
would
be
associated
with
processing
general
permits.
As
noted
previously,
EPA
assumes
that
the
majority
of
NPDES
Phase
I
and
Phase
II
NPDES
storm
water
permit
programs
are
fully
implemented,
and
that
any
new
regulatory
requirements
would
be
superimposed
upon
these
programs.

Under
Option
1,
EPA
assumes
that
no
incremental
costs
would
be
imposed
on
governmental
units.
Under
Option
2,
EPA
estimates
that
each
state
would
incur
costs
to
revise
existing
regulations
to
reflect
the
shift
of
regulatory
coverage
from
Part
122
to
Part
450.
Based
on
the
assumption
that
all
states
would
change
their
storm
water
programs
to
include
certification
of
sedimentation
basins
and
other
aspects
of
the
proposed
rule,
EPA
estimated
the
costs
of
establishing
such
a
program.
The
costs
are
based
on
assumptions
about
the
number
of
labor
hours
states
would
allocate
to
amending
such
programs,

and
the
applicable
labor
rate.
Further
details
on
these
assumptions
and
costs
can
be
found
in
the
Development
Document
(
EPA,
2002)
.

4.7.2
Compliance
Costs
EPA
estimates
that
government
entities
(
federal,
state,
and
local)
commission
as
much
as
one
quarter
of
the
total
value
of
construction
work
completed
in
the
U.
S.
each
year.
As
final
owner
of
a
substantial
amount
of
the
industry
output,
governments
would
bear
some
of
the
compliance
costs
associated
with
the
proposed
rule,
assuming
these
costs
are
passed
on
from
developers
and
builders.
In
Chapter
Five,
Section
5.8,
EPA
allocates
the
government
share
of
compliance
costs
based
on
the
government
share
of
industry
output.
Further
details
about
government
costs
can
also
be
found
in
Chapter
Ten.

4­
75
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
4.8
REFERENCES
Benshoof,
M.
2001.
An
Inside
Look
at
Builders
 
Books.
.
Housing
Economics.
National
Association
of
Home
Builders,
May.

Brealy,
Richard
A.
,
and
Stewart
C.
Myers.
1996.
Principles
of
Corporate
Finance
(
5
th
ed.
)
.
New
York:
The
McGraw­
Hill
Companies,
Inc.

Brigham,
Eugene
F.
,
and
Louis
C.
Gapenski.
1997.
Financial
Management:
Theory
and
Practice
(
8
th
ed.
)
.
Fort
Worth:
The
Dryden
Press.
pp.
428­
431.

Brown,
Dirk
S.
G.
2002.
User
Fee­
Based
Financing
in
the
2000s.
Stormwater.
3(
1)
:
50­
54.

Busco,
Dana,
and
Greg
Lindsey.
2001.
Designing
Stormwater
User
Fees:
Issues
&
Options.
Stormwater.
2(
7)
:
42­
44
CCH.
1999.
2000
U.
S.
Master
Tax
Guide.
Chicago:
CCH
Incorporated.

CWP.
2001.
Impervious
Cover
and
Land
Use
in
the
Chesapeake
Bay
Watershed.
Ellicott
City,
MD:
Center
for
Watershed
Protection,
January.
Additional
data
table,
 
Chesapeake
Bay
Watershed
Impervious
Cover
Results
by
Land
Use
Polygon,
 
received
via
a
facsimile
from
Tetra
Tech,
,
Inc.
,
September
20,
2001.

Dun
&
Bradstreet.
2000.
1999­
2000
Industry
Norms
and
Key
Business
Ratios.

ERG.
2001.
Distribution
of
Acreage
Disturbed
Estimates.
Memo
from
ERG,
Inc.
to
the
U.
S.
Environmental
Protection
Agency,
October
25,
2001.

ERG.
1999.
Real
Estate
Development
Financing.
Memo
from
ERG,
Inc.
to
the
U.
S.
Environmental
Protection
Agency,
December
28,
1999.

Fannie
Mae.
2001.
Glossary:
Qualifying
Guidelines.
Available
at:
http:
/
/
www.
homepath.
com/
cgi­
bin/
WebObjects­
4/
HomePathWOF.
woa/
8/
wa/
Glossary?
topic=
Glo
ssary&
title=
Qualifying%
20Guidelines&
wosid=
IL1000D5300pz500E&
oid=
3684.
Accessed
on:
July
18,
2001.

FASU.
1997.
Florida
Association
of
Stormwater
Utilities:
1997
Stormwater
Utilities
Survey.
Available
at:
http:
/
/
www.
fasu.
org/
publications/
surveys/
.
Accessed
on
January
31,
2002.

FHFB
(
Federal
Housing
Finance
Board)
.
2001.
Monthly
Interest
Rate
Survey
(
MIRS)
Periodic
Summary
Tables.
Available
at:
http:
/
/
www.
fhfb.
gov/
MIRS/
mirs.
htm.
Accessed
on:

FHWA.
2001.
Federal
Highway
Administration.
Typical
Interstate
System
Cost
per
Mile.
Fax
from
C.
Duran,
FHWA
Office
of
Program
Administration,
to
D.
Metivier,
ERG,
Inc.
September
19,
2001.

4­
76
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
Grubb
&
Ellis.
2001.
Office
Market
Trends:
A
Survey
of
the
Nation
 
s
Office
Markets.
Summer.
Available
at:
www.
grubb­
ellis.
com.
Accessed
on
September
11,
2001.

Hirsch,
Albert
A.
1994.
Residential
Construction
from
a
Long­
Run
Perspective.
Survey
of
Current
Business.
June:
30­
41.

Keller,
Brant
D.
2001.
Buddy,
Can
You
Spare
a
Dime?
Stormwater.
2(
2)
:
38­
42.

Kone,
D.
L.
2000.
Land
Development
9
th
ed.
Home
Builder
Press
of
the
National
Association
of
Home
Builders.
Washington,
DC.

Luger,
M.
I.
and
K.
Temkin.
2000.
Red
Tape
and
Housing
Costs.
New
Brunswick,
New
Jersey:
CUPR
Press.

NAHB
(
National
Association
of
Home
Builders)
.
2002.
Characteristics
of
New
Multifamily
Buildings
1987­
1999.
Available
at:
http:
/
/
www.
nahb.
com/
multifamily/
characteristics.
htm.
Accessed
on
May
29,
2001.

NAHB
2001a.
Housing
at
the
Millennium:
Facts,
Figures,
and
Trends.
Washington,
D.
C.
:
The
National
Association
of
Home
Builders.
Available
at:
http:
/
/
www.
nahb.
com/
housing_
issues/
facts.
htm.

NAHB.
2001b.
Building
a
Balance:
Cost
Breakdown
of
A
Single
family
Home.
Available
at:
http:
/
/
www.
nahb.
com/
housing_
issues/
balance_
2.
htm.

Rappaport,
B.
A.
,
and
T.
A.
Cole.
2000.
1997
Economic
Census
 
Construction
Sector
Special
Study:
Housing
Starts
Statistics
 
A
Profile
of
the
Homebuilding
Industry.
U.
S.
Census
Bureau,
July.

Ross,
D.
and
S.
Thorpe.
1992.
 
Impact
Fees:
Practical
Guide
for
Calculation
and
Implementation.
 
Available
at:
http:
/
/
www.
revenuecost.
com/
imp_
fees.
html.
Accessed
on
July
11,
2001.

R.
S.
Means.
2001.
Heavy
Construction
Cost
Data
15
th
Annual
Edition.
Kingston,
Massachusetts:
R.
S.
Means
Co.

R.
S.
Means.
2000.
Building
Construction
Cost
Data
58
th
Annual
Edition.
Kingston,
Massachusetts:
R.
S.
Means
Co.

Savage,
H.
A.
1999.
Who
Could
Afford
to
Buy
a
House
in
1995?
Washington:
U.
S.
Census
Bureau.
Supplemental
material
is
available
at:
http:
/
/
www.
census.
gov/
hhes/
www/
housing/
hsgaffrd/
afford95/
aff95src.
html.

Sierra
Club.
2000.
Sprawl
Costs
Us
All:
How
Your
Taxes
Fuel
Suburban
Sprawl.
Available
at:
http:
/
/
www.
sierraclub.
org.
Accessed
on
September
20,
2000.

Tetra
Tech,
Inc.
2001.
SWV2.
xls.
Microsoft
Excel
Spreadsheet.
Received
October
9,
2001.

ULI.
2000.
Urban
Land
Institute.
Market
Profiles
2000:
North
America.

4­
77
Economic
Analysis
of
Construction
and
Development
Proposed
Effluent
Guidelines
May
2002
U.
S.
Census
Bureau.
2000a.
Current
Construction
Report
C25:
Characteristics
of
New
Housing,
1999.
Issued
July
2000.

U.
S.
Census
Bureau.
2000b.
Current
Construction
Report
C40:
New
Privately
Owned
Housing
Units
Authorized,
1999.
Issued
July
2000.

U.
S.
Census
Bureau.
2000c.
1997
Economic
Census:
Construction:
Subject
Series.
January.
2000.

U.
S.
Census
Bureau.
2000d.
Current
Population
Reports,
P60­
209,
Money
Income
in
the
United
States:
1999.
Washington,
D.
C.
:
U.
S.
GPO.

U.
S.
Census
Bureau.
1999.
American
Housing
Survey
for
the
United
States:
1999.
Available
at:
http:
/
/
www.
census.
gov/
hhes/
www/
housing/
ahs/
ahs99/
tab212.
html.

U.
S.
Census
Bureau.
1998.
State
and
Metropolitan
Area
Data
Book
1997­
98
(
5
th
Edition)
.
Washington,
D.
C.
:
U.
S.
GPO.

U.
S.
Department
of
Agriculture.
2000.
1997
Natural
Resources
Inventory
Summary
Report.
Table
8.
Changes
in
land
cover/
use
between
1992
and
1997.
http:
/
/
www.
nhq.
nrcs.
usda.
gov/
NRI/
1997/
summary_
report/
original/
table8.
html
U.
S.
EPA.
2002.
Development
Document
for
the
Effluent
Guidelines
for
the
Construction
and
Development
Point
Source
Category.
EPA­
821­
R­
02­
007.

U.
S.
EPA.
2001a.
Estimation
of
Capital
Costs
for
Technology
Options.
Draft
dated
July
12,
2001.

U.
S.
EPA.
2001b.
Economic
Analysis
of
the
Proposed
Revisions
to
the
National
Pollutant
Discharge
Elimination
System
Regulation
and
the
Effluent
Guidelines
for
Concentrated
Animal
Feeding
Operations.
EPA­
821­
R­
01­
001.
January.

U.
S.
EPA
2001c.
Summary
of
Focus
Group
Meetings
with
the
National
Association
of
Home
Builders.
Chicago,
IL.
March
13.

U.
S.
EPA
2001d.
Summary
of
Focus
Group
Meetings
with
the
National
Association
of
Home
Builders.
Dallas,
TX.
March
20.

U.
S.
EPA.
1999.
Economic
Analysis
of
the
Final
Phase
II
Storm
Water
Rule.
Office
of
Wastewater
Management.

U.
S.
Small
Business
Administration,
1999.
Employer
Firm
Births
and
Deaths
by
Employment
Size
of
Firm,
1989­
1998.
Available
at:
http:
/
/
www.
sba.
gov/
advo/
stats/
dyn_
b_
d8998.
pdf.

Wright,
1996.
Paul
H.
Wright.
Highway
Engineering,
6
th
edition.
New
York:
John
Wiley
&
Sons.
1996.

4­
78
