Legislative Apportionment and Fiscal Policy in the American States
| Author | H. George Frederickson, Yong Hyo Cho |
| DOI | 10.1177/106591297402700101 |
| Published date | 01 March 1974 |
| Date | 01 March 1974 |
| Subject Matter | Articles |
5
LEGISLATIVE
APPORTIONMENT
AND
FISCAL
POLICY
IN
THE
AMERICAN
STATES
H.
GEORGE
FREDERICKSON,
Indiana
University
and
YONG
HYO
CHO,
University
of
Akron
ELIBERATE
changes
in
major
public
policy
are
seldom
dramatic,
tending
D
to
come
instead
in
minor
increments.
Only
in
retrospect
can
these
incre-
JL-~
ments
be
totaled
in
a
way
that
enables
whole
epochs
of
policy
change
to
be
clearly
identified.
The
transition
from
a
patronage
to
a
merit
public
service
was
surely
epochal,
but
the
policy
was
changed
gradually
by
minor
increments.
And
so
it
has
been
in
the
widening
of
the
franchise,
the
regulation
of
industry
and
com-
merce,
the
expansion
of
government’s
responsibility
for
the
poor,
municipal
reform,
and
other
major
epochs
in
American
politics.
The
incremental
character
of
policy
change
has
always
provided
a
challenge
to
the
political
scientist.
As
he
develops
hypotheses
asserting
relationships
between
policy
and
policy
results,
he
has
great
difficulty
identifying
and
measuring
specific
policy
changes
and
tracing
them
to
particular
changes
in
policy
results.
As
a
question
of
policy,
the
apportionment
of
state
legislative
bodies
has
been
an
exception
to
these
generalizations
for
several
reasons.
First,
apportionment
is
rather
easy
to
measure.
Second,
there
is
no
dearth
of
hypothesized
relationships
about
the
&dquo;quality&dquo;
of
the
apportionment
of
state
legislative
bodies
and
the
policy
behavior
of
those
bodies.
Third,
the
change
from
general
malapportionment
to
relative
equity
occurred,
in
policy-making
terms,
during
a
very
short
span
of
time.
Between
the
Supreme
Court’s
decision
in
1962
and
1968
all
significant
malappor-
tionment
in
state
legislatures
had
virtually
disappeared.
Fourth,
because
reappor-
tionment
was
more
dramatic
than
most
policy
change,
it
is
presumed
to
be
easier
to
trace
apportionment
equalization
to
changes
in
policy
outcome.
For
these
rea-
sons,
almost
all
studies
of
policy
results
and
their
correlates
include
a
measure
of
apportionment
as
a
potential
predictor.
The
results
have
not
been
impressive.
The
three
best
known
measures
of
appor-
tionment
(Schubert-Press,
Dauer-Kelsay,
David-Eisenberg)
have
not
been
robust
determinants
of
either
expenditure
or
non-expenditure
policy
outcomes.
Thomas
Dye’s
well-known
analysis
of
policy
outcomes
in
the
American
states
concludes
that
legislative
apportionment
is
unimportant
as
an
indicator
of
policy
behavior.
All
of
his
political
measures
are
similarly
unimpressive
as
predictors:
&dquo;...
the
political
choices
of
malapportioned
legislatures
are
not
noticeably
different
from
policy
NOTE:
The
authors
owe
heavy
thanks
to
John
Bachellor,
Don
Pond,
Robert
Firestine,
Laura
and
Janet
Spaulding,
Clare
Donaher,
Dorothy
Smith,
Ann
Dolliver,
Ronald
Weber,
and
William
Shaffer,
and
Jack
Rossotti.
The
research
reported
here
was
supported
by
the
State
Legislatures
Project
of
the
National
Municipal
League.
This
is
a
revision
of
a
paper
presented
at
the
1970
Conference
of
the
American
Political
Science
Association
in
Los
Angeles,
California.
6
choices
of
well-apportioned
legislatures
The
works
of
Ira
Sharkansky,2
Richard
Hofferbert,3
and
Brian
Fry
and
Richard
Winters4
all
reach
virtually
the
same
conclusion.
Only
one
piece
of
research
finds
apportionment
to
be
a
significant
indicator
of
policy
outcomes.
Alan
Pulsipher
and
James
Weatherby
examined
expenditure
variance
in
the
states
(using
an
aggregate
dependent
variable
which
included
local
with
state
spending)
for
functions
such
as
police,
highways,
and
parks
and
recrea-
tion.
They
concluded
that
malapportionment
is
generally
statistically
significant
as
a
predictor
of
spending
variance.6
This
finding
prompted
an
intriguing
exchange
of
letters
between
Pulsipher
and
Weatherby
and
Dye
and
Dyson.
Dye
and
Dyson
correctly
point
out
that
Pulsipher
and
Weatherby
test
only
for
the
possible
effect
of
malapportionment
(and
party
competition)
without
testing
for
the
possible
effect
of
other
social,
economic,
or
political
indices;
that
they
do
not
give
their
R2
(the
amount
of
variance
explained ) ;
and
that
the
&dquo;contribution
of
these
political
varia-
bles
are
not
as
great
as
the
contribution
of
socioeconomic
variables.&dquo;
In
response
Pulsipher
and
Weatherby
argue
that
Dye
and
Dyson’s
model
contains
a
methodo-
logical
non
sequitur
&dquo;namely,
that
if
economic
and
demographic
factors
are
more
important
than
political
factors,
than
political
factors
are
unimportant.&dquo;
6
It
is
with
some
temerity
that
we
enter
this
cross
fire.
But,
for
several
reasons,
we
here
present
yet
another
attempt
to
assess
the
impact
of
legislative
apportion-
ment
on
policy
outcomes.
First,
we
are
impressed
by
the
starkly
different
conclu-
sions
reached
by
those
who
research
the
apportionment
question.
Political
scientists
studying
state
legislative
and
political
behavior
in
an
institutional
and
non-quanti-
tative
way,
as
well
as
experienced
state
politicians,
almost
universally
conclude
that
apportionment
is
of
fundamental
importance
in
determining
&dquo;who
gets
what,
when,
and
how.&dquo;
So-called
policy
analysts,
with
the
exception
of
Pulsipher
and
Weather-
by,
come
to
the
opposite
conclusion.
Are
these
conclusions
reconcilable?
Second,
the
apportionment
&dquo;rules
of
the
game&dquo;
changed
very
quickly,
providing
a
unique
opportunity
to
test
the
relationship
between
new
policy
makers
and
consequent
policy.
Third
previous
research
on
apportionment
exhibits
the
weaknesses
listed
below.
a.
Policy
study
has
not
been
longitudinal.
Most
studies
are
based
on
the
analy-
sis
of
a
single
year
and
in
almost
all
cases
a
year
prior
to
1962
(the
year
of
Baker
v.
Carr)
and
an
attempt
to
assess
the
relationship
between
malapportionment
and
1
"Malapportionment
and
Public
Policy
in
the
States,"
Journal
of
Politics,
27
(August
1965),
599.
See
also
his
Politics,
Economics,
and
the
Public:
Policy
Outcomes
in
the
American
States
(Chicago:
Rand
McNally,
1966),
pp.
270-81.
2
Spending
in
the
American
States
(Chicago:
Rand-McNally,
1968).
Sharkansky
does
con-
tend,
however,
that
his
findings
that
previous
legislative
spending
behavior
is
an
impor-
tant
determinant
of
subsequent
behavior
is
evidence
of
the
salience
of
political
influence
in
money
decisions.
3
"The
Relations
Between
Public
Policy
and
Some
Structural
and
Environmental
Variables
in
the
American
States,"
American
Political
Science
Review,
60
(June
1966),
508-28.
4
Brian
R.
Fry
and
Richard
F.
Winters,
"The
Politics
of
Redistribution,"
American
Political
Science
Review,
64
(June
1970),
508-23.
5
"Malapportionment,
Party
Competition,
and
the
Functional
Distribution
of
Government
Expenditures,"
American
Political
Science
Review,
62
(December
1968),
1207-19.
6
"Communications,"
American
Political
Science
Review,
63
(June
1969), 529.
7
policy
variance.
If
reapportionment
is
presumed
to
change
policy,
then
the
subject
should
be
approached
from
the
perspective
of
longitudinal
analysis.
b.
The
aggregation
of
jurisdictions.
Because
the
Bureau
of
the
Census
some-
times
reports
state
and
local
government
spending
together,
several
analysts
have
attempted
to
correlate
apportionment
with such
aggregated
outcome
indices.
If
it
is
the
states
that
are
apportioned,
how
can
we
expect
the
policy
decisions
of
cities,
counties,
special
districts,
and
the
like
to
have
any
sensible
correlation
with
state
reapportionment?
If
reapportionment
changes
the
state
policy
makers,
we
might
logically
expect
changes
in
state
policy.
c.
There
has
been
an
overemphasis
on
the
level
of
total
state
spending
and
its
determinants
and
an
underemphasis
on
spending
for
specific
functions.
How
large
the
pie
is
and
what
determines
its
size
is
an
interesting
question.
An
equally
inter-
esting
question
is
what
functions
of
government
get
what
share
of
that
pie
and
what
determines
the
size
of
shares.
d.
Too
few
inventive
variables
of
the
political
variety,
and
particularly
of
the
apportionment
variety,
have
been
utilized.
Similarly,
too
few
change
variables
of
this
sort
have
been
organized.
e.
Because
of
a
preoccupation
with
the
spending
of
jurisdictions
and
attempts
to
measure
the
impact
of
this
spending,
too
little
attention
has
been
given
to
the
redistribution
of
resources
in
the
political
system.
(Fry
and
Winters
are
exceptions
to
this
and
a
substantial
improvement
on
previous
work.)
Fourth,
the
models
used
by
policy
analysts
have
probed
only
that
aspect
of
legisla-
tive
apportionment
having
to
do
with
district
population
equity.
Those
familiar
with
the
processes
of
legislative
apportionment
and
reapportionment
regard
the
phenomenon
as
far
too
complex
to
be
comprehended
by
an
index
which
accounts
only
for
equity.
The
processes
of
apportionment
involve
the
protection
of
incum-
bents
by
gerrymandering
and
other
techniques;
the
protection
of
committee
chair-
manships,
particularly
on
those
committees
that control
state
aid
to
localities;
the
relative
strength
of the
party
in
power
at
the
time
of
apportionment;
and
whether
the
party
in
power
is
that of
the
incumbent
governor.
To
fully
comprehend
legis-
lative
apportionment
and
its
relationship
to
policy
consequence,
the
analyst
must
recognize
the
complexity
of
the
subject
and
invent
measures
to
account
for
that
complexity.
Our
hypotheses,
derived
primarily
from
the
contentions
of
experienced
politi-
cians
and
non-quantitative
analysts
of
state
legislative
behavior,
are
designed
to
accommodate
the
complexity
of
the
subject,
as
well
as
to
put
into
quantitative
form
as
many
of
the
&dquo;common
sense&dquo;
observations
of
these
analysts
as
possible.
First,
we
hypothesize
that
a
legislative
apportionment
scheme
determines
to
a
significant
extent
who
holds
the
purse
strings ;
therefore,
the
characteristics
of
the
constituen-
cies
(such
as
rural-agricultural
versus
urban-industrial)
do
influence
both
the
size
of
the
state
purse
and
the
manner
in
which
the
money
in
that
purse
is
distributed.
Second,
it
seems
reasonable
to
predict
that
it
will
be
easier
to
discern
the
influ-
ence
of
reapportionment
in
1967
and
1969 than
to
discern
the
influence
of
mal-
apportionment
in
1962.
Malapportioned
legislatures
have
maintained
their
inequi-
ties
for
many
years,
during
which
the
effect
of
malapportionment
was
hidden
under
To continue reading
Request your trial