By Charles
Sampson
Introduction
A main
concern of municipal governments[1]in the
United States is their capacity to finance, predictably and reasonably,
services and programs for their residents.
Scholars have questioned whether that capacity is impacted by the
structure of administrative authority in various municipal governments. The existence of the notion of whether the
structure of administrative authority has an independent effect on policy
(expenditure decisions) is characterized by many titles and occupies much space
in the literature. This research intends
to observe data and make an empirical assessment of whether two forms of
municipal administrative authority are distinguished by their expenditure
patterns.
Municipal
governments spend their money on a variety of services including education,
transportation, health, welfare, safety and housing. The larger municipalities spend a greater
percentage of their budget on education services and health, hospitals, and
public welfare. Smaller municipalities spend a greater percentage on
transportation and environment, housing and waste management (National League
of Cities, 2006). Municipal governments
differ along the line of their size, election method and administrative
authority. Administrative (and
political) arrangements are left to the respective states that (in turn) grant
municipal status vis-à-vis charters.
This state of affairs holds largely due to the fact that the
Constitution of the United States makes no mention of local government.
Administrative authority is determined by the form of government granted in a
charter and there are five distinct forms: council manager
(58%)
mayor-council (38%); commission (3%); town meeting and representative town
meeting (<1 span="" style="mso-spacerun: yes;"> (Renner, 1988; Boynton
and DeSantis, 1990). The 2006 National
Municipal Form of Government Survey conducted by ICMA indicates that the vast
majority of American cities have either the Council-Manager or the
Mayor-Council form of government.
Introduction
A main
concern of municipal governments[1]in the
United States is their capacity to finance, predictably and reasonably,
services and programs for their residents.
Scholars have questioned whether that capacity is impacted by the
structure of administrative authority in various municipal governments. The existence of the notion of whether the
structure of administrative authority has an independent effect on policy
(expenditure decisions) is characterized by many titles and occupies much space
in the literature. This research intends
to observe data and make an empirical assessment of whether two forms of
municipal administrative authority are distinguished by their expenditure
patterns.
Municipal
governments spend their money on a variety of services including education,
transportation, health, welfare, safety and housing. The larger municipalities spend a greater
percentage of their budget on education services and health, hospitals, and
public welfare. Smaller municipalities spend a greater percentage on
transportation and environment, housing and waste management (National League
of Cities, 2006). Municipal governments
differ along the line of their size, election method and administrative
authority. Administrative (and
political) arrangements are left to the respective states that (in turn) grant
municipal status vis-à-vis charters.
This state of affairs holds largely due to the fact that the
Constitution of the United States makes no mention of local government.
Administrative authority is determined by the form of government granted in a
charter and there are five distinct forms: council manager
(58%)
mayor-council (38%); commission (3%); town meeting and representative town
meeting (<1 span="" style="mso-spacerun: yes;"> (Renner, 1988; Boynton
and DeSantis, 1990). The 2006 National
Municipal Form of Government Survey conducted by ICMA indicates that the vast
majority of American cities have either the Council-Manager or the
Mayor-Council form of government.
Complementary and competing theories about municipal expenditures and form of government
Social scientists have studied
municipal expenditures and governance from several theoretical perspectives,
and as one might expect there is more than one school of thought about whether
and/or how municipal expenditure patterns can be explained by form of government
(FOG). The impact of continuously
changing forms of American local governments is partially explained in social
science theory generally described as institutionalism or new institutionalism
(Rogers; 1959 March and Olsen 1989; DiMaggio and Powell 1991; Lynn 1998; and
Frederickson, Johnson and Wood, 2004).
There is no single new
institutional approach; instead there are a number of variants such as
historical institutionalism, sociological institutionalism, and rational choice
institutionalism (Hall and Taylor, 1996; Peters, 1999). Peters (1999) outlines four features of
institutions that help to demarcate institutions from other entities First,
institutions are ‘in some way a structural feature of society’ (Peters,1999,
p18). These can be formal (organizations) or informal (networks) both of which
work to shape relations between actors;
Second, institutions entail an element of stability; Third, institutions, impact behavior; Fourth,
there is ‘some sense of shared values and meanings´ (Peters, 1999: 18) among
the members of an institution.
Institutionalism reflects are ‘the sets of rules that guide and
constrain actors´ behavior’ (Lowndes, 2001: 1958). Scharpf (1997:38) sees
institutions as ‘systems of rules that structure the courses of actions that a
set of actors may choose.
When
Iapply the institutional lens to American local government, Isee structural
characteristics, i.e., administrative arrangements or forms of city government
that represent collective institutions.
A form of government represents a particular institutional nuance and
these administrative arrangements may be connected to a municipal government’s
capacity to finance, predictably and reasonably, services and programs for
their residents.
Basic
Municipal forms of government and their significance
Municipal
governments differ along the line of their size, election method, and their
administrative and political arrangements.
Twice in each decade the International City Management Association
(ICMA) surveys municipal governments in the United States. These surveys reveal
that there are five common organizational arrangements (forms) for municipal
governments: 1) mayor-council, 2) council-manager, 3) commission, 4) town
meeting and, 5) representative town meeting.
The legislative bodies are the result of charters, ordinances or other
actions of their respective state governments that, on their face, facilitate
political constitutions for the cities. Observations of the ICMA surveys for
the period between 1981 and 2006 reveals that the most frequently cited forms
of government identified by respondent cities are the council-manager and the
mayor-council forms. These two forms are
used by 84% of the cities which participate in that survey .
The
basic council-manager form is a structural arrangement for a non-political
executive as the administrative centerpiece of a municipal government. It parallels the organization of the business
corporation complete with voters as stockholders who elect a board of directors
(council) and a mayor as chair of that board.
The board and chair appoint a chief administrative officer (manager) who
is expected to foster administrative competence and efficiency.
The parallel model for political
authority is the mayor-council form which is intended to highlight
representative government at the municipal level. In the strong-mayor form of this
government, the mayor is given almost total administrative authority, with the
power to appoint and dismiss department heads without council approval. Likewise,
the mayor prepares and administers the budget, although that budget often must
be approved by the city council. In some strong-mayor governments, the mayor
will appoint a chief administrative officer (CAO) who will supervise department
heads, prepare the budget, and coordinate departments. This CAO is responsible
only to the mayor. In the weak-mayor
form of this government the
council possesses both legislative and executive authority. The council
may appoint officials and must approve of mayoral nominations. The council also
exercises primary control over the municipal budget. As indicated by the ICMA
data, the commission form and the town hall meeting forms have virtually
disappeared.
The fact that cities are continuously changing
frameworks for governance is arguably a response to battles between competing
municipal-based forces in a democratic society, but more important are the
outcomes of these redesign processes.
Frederickson, et al (2004) imply that continuous change in the
governance framework erases some of the distinguishing features between basic
structures and subsequently results in structures that increasingly resemble
each other. Their contribution is
referred to as the “adapted cities framework” which posits the notion that
demands for mayoral leadership, political responsiveness and administrative
efficiency is conciliated and best captured as the “Type III city.” This form attempts to capture the best of
both the elements of administrative efficiency and political responsiveness.
The majority of the literature supports that reform
governments, i.e., council manager (CM) are more business-like than all
others. Public choice theories suggest
the expected effects of government form and of other variables on municipal
expenditures. Booms (1966) is credited
with demonstrating that city managers are more likely than mayor-council forms
of government to minimize the cost of municipal services. That work was one of
the first in urban public finance literature to investigate the efficiency of Council
Manager Cities as measured by spending level differences. It hypothesized that the form of city
government has as important effect on the levels of public expenditures for
municipal functions and that city managers are more likely than elected mayors
to minimize the cost of producing municipal services; DeSantis and Renner
(1994) report that structural characteristics of local governments are
important and that CM types spend less than MC types. Lineberry and Fowler (1967) are credited with
being the scholars to systematically investigate the direct and indirect
effects of reformed city structures.
Their study purports that CM forms plus at large elections tend to lower
tax and spend at lower levels than MC forms.
They conclude that political structure a significant intervening
variable that affects the relationship between the socio-economic
characteristics of municipalities and their policy outputs. Specifically MC municipalities tend to be
more responsive to the demographic characteristics of their constituents
compared to CM forms. They argue that CM
governments make public policy less responsive to the demands arising out of
social conflicts in the population. Lyons (1978) supports Lineberry and Fowler’s
findings that reformed cities react more readily to pressure from groups that
want lower taxes and less spending.
Feiock, Jeong, and Kim (2003) report that CM forms allow administrators
and elected officials to more easily resist opportunistic behaviors. A substantial segment of the literature
counters the claim of CM superiority or argues that FOG makes for no difference
relative to policy decisions. Clark
(1968) stated that municipal reform types are correlated with the highest
levels of aggregate spending; Liebert
(1974) suggests that if the functional inclusiveness of municipalities is
controlled, the relationship between structure and policy outputs disappears
altogether. A number of studies argue
that there is no difference in FOG as it relates to municipal
expenditures: Deno and Mehay (1987);
Farnham (1987); Hayes and Chang (1990); Morgan and Pelissero (1980); and Reid
(1991).
Most of the related literature treats the
administrative authority issue as exogenous; however, Sass (1991) and Fahey
(1998) argue that structure is endogenous.
Contributions by McDonald (2005) complement both Sass and Fahey by
arguing that if a structure of local government produces inefficient choices
the structure would eventually change.
When
in Doubt, Follow the Money
This
study attempts to “follow the money.” By observing municipal expenditure data,
we empirically assess whether and/or how the most frequently used forms of
municipal administrative authority are connected to expenditure patterns. The units of analysis are the two forms of
organizational arrangements most frequently found in American municipalities
(cite ICMA): the council-manager and the mayor-council) forms.
Data
Description
This
data set consists of numerical and qualitative variables from 126 cities, each
with population’s greater than100K, with a mayor-council or council-manager
form of government (FOG) during the period 1967 – 2007, a forty year time
span. The data captured in this study
were collected by the Census of Governments which conducts an enumeration twice
in each decade (during the second and seventh years). The data in this study are for October of the
year preceding the Census e.g., 2006, 2001, and so forth and reflect the
spending activity of the observed officials.
Thus, there are t=9 time-indexed measurements for each variable for each
city. Each time point is also described
by a dichotomous variable representing the minority status of the mayor for
that given year.
Dependent
variables
Scholars of urban politics tend to classify spending into three
basic categories allocational, developmental and redistributive spending
(Peterson, 1981; Stein, 1990). There is no universally accepted model for
placing specific spending subcategories into these three larger spending
areas.
Allocational policy is defined as spending on a range of basic
city services that can be considered housekeeping services. Allocational spending funds
developmentally and distributionally neutral housekeeping obligations. These
expenditures function with less than perfect efficiency and spillovers of
social benefits and social costs cannot always be identified. Moreover, many allocational functions involve
sub functions for which the amount of inter-community spillover varies. It is difficult to define an area of benefit
consumption for certain services; many of the beneficiaries (especially the
indirect beneficiaries) cannot even be identified, and, therefore, the area to
which benefits extend cannot be circumscribed.
In this study, allocational spending includes: 1) education, 2)
education other than capital, 3) fire protection, 4) police protection, 5)
sewerage, 6) electric utilities, 7) gas utilities and 8) general
expenditures. This categorization of
spending is defined by the Technical Documentation Manual and roughly parallels core divisions in the urban
political arena
By
contrast, developmental expenditures include what Molotch (1976) refers to as
preconditions of growth they include expenditures as defined by the Technical
Documentation Manual, e.g., 1) water
utilities fund the ‘operation, maintenance,
and construction of public water supply systems’; 2) regular highways include expenditures for the ‘maintenance,
operation, repair, and construction of non-toll highways, streets, roads,
alleys, sidewalks, bridges, and related structures’; 3) parks and recreation funds the ‘provision and support of
recreational and cultural-scientific facilities’; 4) sewerage expenditures fund the ‘provision, maintenance, and
operation of sanitary and storm sewer systems and sewage disposal and treatment
facilities’; 5) central staff services
funds ‘government-wide executive, administrative, and staff service agencies’;
6) general public buildings
covers spending for ‘construction, equipping, maintenance, and operation of
general public buildings not related to specific functions or agencies’; 7) parking facilities funds the
‘provision, construction, maintenance, and operation of local government public
parking facilities operated on a commercial basis’; and 8) health services other funds the
‘provision of services for the conservation and improvement of public
health.’ These expenditures have
historically supported business development and sustained cities, and they
continue to be the ‘significant dynamic’ of contemporary local political
economy and are critical to public resources and local issue agendas. Our
listing is intended to complement Tiebout (1956), and Logan and Molotch (1987)
arguments that cities will spend funds to attract businesses and tax-paying residents.
Finally,
redistributive policies are those that tend to target and benefit
less-advantaged residents. In this study, they are operationalized to include
health, hospitals other than capital, welfare, and limited public housing.
Hajnal and Trounstine (2010) state that a
range of different kinds of surveys all find that poor minority
residents are especially concerned about redistribution and social services,
whereas Whites and the middle class are especially concerned about attracting
businesses and other aspects of development, reducing taxes and improving their
quality of life through better parks and recreation and easier transportation
(Clark & Ferguson, 1983; Lovrich, 1984; Welch, Sigelman, Bledsoe, &
Combs, 2001). Thus, the three
dependent variables in this study are three types of municipal expenditures,
i.e., allocational, developmental, and redistributive.
Independent Variables/Covariates
There are two independent variables
used in this analysis -- race and form of government. For mayoral race/ethnicity, we use a binary
variable coded as 1 for minority mayor and 0 for
non-minority (white) mayor. Sources for
the race variable include the Census of Governments, and International
City Management Association database. A second covariate in our data is the form
of government (FOG), with only two possible outcomes, mayor council (coded as
1) and city manager (coded as 2). These two (of five
organizational arrangements) account for >85% of all American municipal
forms of government according to International City Management Association
(ICMA, 2006). It should be noted that
FOG is the only covariate that is time-invariant for all 8 time periods. The other independent variable, Race, is a
time varying covariate.
The
Hypothesis
Much of
the public management literature which treats municipal government structure
argue that mayor council (MC) forms of government are inclined to favor
political responsiveness as a behavior that is more important than
administrative efficiency. Thus, I
develop an argument for two related hypothesis along the three forms of
spending that cities engage 1) allocational; 2) developmental and 3) redistributive:
H1 Mayor
councilgovernments will spend moreper capitain allocational spending, than CM
because this type of spending covers the “housekeeping” functions of city
government such as police and fire operations; garbage collection, etc.,
likewise;
H2 CM
(council-manager) governmentswill invest more, per capita, in developmental
spending, than MC cities, since CM governments are inclined toward
administrative efficiency and favor business-like activity which attract
industries that are likely to generate employment while attracting
comparatively higher income residents.
Thus, spending on transportation infrastructure, parks and recreation,
tax support for entrepreneurial development, i.e., tax increment districts, and
activities which attract businesses that provide jobs and attract high income
residents will distinguish that set of behaviors from officials in MC
operations.; and
H3 MC
government’s will devote more per capita to
redistributive activities, than CM governments, although neither of the municipal
operations will favor redistributive spending.
Municipal spending on housing.health-care and welfare for the
populations that are more financially dependent
is likely to be found in an institutional setting which is organized to
be more politically responsive.
Method of Statistical Analysis: The Unconditional
Linear Growth Model
To begin, Ipresent the growth model
which includes time as a single predictor and allows for random slope and
intercept terms. This random coefficient
regression model can be used to fit city-specific linear models for the
expenditure data for each of the three dependent variables, developmental,
allocational and redistributive.
Diagnostic checks establish the goodness of fit for the model so that
additional analysis to ascertain the effect, if any, of the independent
variables can follow.
Let
denote the expenditure value for city j and time
point i. The two-level model is given
by:
Level 1:
, where
, and
Level 2:
,
,
where
The
parameters
and
can be interpreted as the average (across all
cities) intercept and slope across all 125 cities. Graphs of our data indicate nonzero
intercepts and slopes, so it is expected that these parameters are
significantly different from zero. If
the variance components
and
are statistically significant, this verifies that
modeling the intercept and slope as random was a good strategy. For each city, predictions at each time
point can be plotted along with the original data to investigate how well the
linear model fits.
Assessment of Form of Government
Testing the effect of form of
government (FOG) is straightforward.
Since the form of government stays constant for each of the eight time
points for each city, it is examined by including it as a level-2
predictor. The model from above becomes:
Level 1:
, where
, and
Level 2:
,
, where
Rejection
of the hypothesis
implies that form of government has an effect
on the intercepts; significance of
implies
a relationship between form of government and the slope.
Assessment of Minority Mayor
For assessing the effect of race,
Iuse two methods. The first approach is
to examine the residuals from the fitted growth curves, where the model does
not include race as a covariate. If race
is not affecting expenditures, then the set of residuals corresponding to the
minority mayors should have the same zero-mean distribution as the set of
residuals for white majors. This method
is popular in litigation statistics, (Gastwirth, 1975, 1984, 1988, 2000;
Gastwirth and Wang, 1987) in which a multiple regression model is used to
predict salary, and then the distributions of female residuals is compared to
that of males.
For comparing these two
distributions, I use both parametric methods (which assume normality of the
underlying densities) and nonparametric, distribution-free methods. The parametric method is the two sample
t-test, which compares the true means.
For the nonparametric methods, Icompare the medians using the Wilcoxon
test. Finally, a distribution-free test
of equality of the two distributions is carried out using the
Kolmogorov-Smirnoff procedure.
As an alternative, Ican include race
as a time-varying random effect. In this
approach, each city is modeled as a function of both time and race. The predicted values no longer appear linear
because the time points corresponding to a minority mayor are adjusted for the
estimated race effect. The model
becomes:
Level
1:
, where
,
Level
2:
,
,
where
Significance of
implies a race effect. If the variance component
is statistically significant, then the race effect
varies across cities.
Results and Discussion
Table1, below,provides descriptive
statistics while controlling for inflation by using for per capita spending for
the three types of municipal expenditures.
The data demonstrate that mayor-council governments spend more than
city-manager governments in all three areas of spending. Moreover, in the
instances where there is a minority mayor associated with the FOG, spending in
each of the three categories is greater in comparison with the white mayor in
each FOG.
Descriptive
Statistics for the Three Types of Governmental Expenditures (Per Capita)
Variables
|
N
|
Mean
|
StdDev
|
|
Developmental
|
||||
FOG
|
Mayor Council (MC)
|
450
|
66.699
|
90.597
|
City Manager (CM)
|
569
|
58.130
|
70.759
|
|
Race
|
White
|
959
|
55.122
|
72.780
|
Minority
|
141
|
99.456
|
104.280
|
|
Allocational
|
||||
FOG
|
Mayor Council (MC)
|
458
|
321.933
|
528.207
|
City Manager (CM)
|
569
|
261.213
|
376.370
|
|
Race
|
White
|
967
|
258.280
|
423.745
|
Minority
|
141
|
484.794
|
565.255
|
|
Redistributive
|
||||
FOG
|
Mayor Council (MC)
|
456
|
29.316
|
63.409
|
City Manager (CM)
|
556
|
17.580
|
36.869
|
|
Race
|
White
|
957
|
18.661
|
42.903
|
Minority
|
135
|
48.087
|
77.963
|
Note: The
measurement unit is dollar per 1000 people. N=1027
Developmental Expenditures:
The log transformed development data
shows the need for a variance-stabilizing transformation, so the log transform
was used.
Table 2.Results of the Linear Growth Model for
Developmental Expenditure
Parameter
|
Linear
Growth Model
|
||||
Fixed Effects
|
Estimate
(Standard Err)
|
||||
Initial
status,
|
Intercept
|
1.2638*** (0.0440)
|
|||
Rate
of Change,
|
Slope
|
0.5017*** (0.0065)
|
|||
Variance Components
|
|||||
Level
1
|
Within
person
|
0.1982*** (0.0096)
|
|||
Level
2
|
In
initial status
|
0.1613*** (0.3199)
|
|||
In
Rate of Change
|
0.0018**
(0.0062)
|
||||
Covariance
|
0.0001
(0.0035)
|
||||
Note: * p<.05, ** p<.01, *** p<.001
In this linear
growth model, the parameters contain fixed effects and random effects.
is the fixed effect parameter that can be interpreted as the average (across all cities) intercept when TIME=0 (year of 1967), and is the fixed effect parameter that can be interpreted as the average slope across cities. Significance of the variance components indicates that the cities have different slopes and intercepts.
is the fixed effect parameter that can be interpreted as the average (across all cities) intercept when TIME=0 (year of 1967), and is the fixed effect parameter that can be interpreted as the average slope across cities. Significance of the variance components indicates that the cities have different slopes and intercepts.
Results
of the Conditional Linear Growth Model (for developmental expenditures, below)
show fixed effects impact for time and FOG.
Table 3.Results of the Conditional Linear Growth
Model for Developmental Expenditure
Parameter
|
Linear
Growth Model
|
||||
Fixed Effects
|
Estimate
(Standard Err)
|
||||
Initial
status,
|
Intercept
|
1.3501*** (0.1518)
|
|||
Rate
of Change,
|
Time
|
0.5173*** (0.0227)
|
|||
FOG
|
-0.0450 (0.0928)
|
||||
Time*FOG
|
-0.0103 (0.0139)
|
||||
Variance Components
|
|||||
Level
1
|
Within
person
|
0.1973*** (0.0099)
|
|||
Level
2
|
Intercept
|
0.1658*** (0.0326)
|
|||
Time
|
0.0020**
(0.0007)
|
||||
Covariance
|
0.0003
(0.0037)
|
||||
Fixed Effects
|
|||||
Initial
status,
|
Intercept
|
1.2571*** (0.0439)
|
|||
Rate
of Change,
|
Time
|
0.4975*** (0.0065)
|
|||
Race
|
0.1807**
(0.0576)
|
||||
Variance Components
|
|||||
Level
1
|
Within
person
|
0.1952*** (0.0010)
|
|||
Level
2
|
Intercept
|
0.1612*** (0.0307)
|
|||
Time
|
0.0017**
(0.0007)
|
||||
Race
|
0.0271
(0.0376)
|
||||
Covariance
|
-0.0003 (0.0035)
|
||||
0.0171
(0.0260)
|
|||||
-0.0018 (0.0049)
|
|||||
Note: * p<.05, ** p<.01, *** p<.001;Time:
0, 1, 2, 3, 4, 5, 6, 7, 8; Race: 1=Minority, 0=White; FOG: 1= Mayor Council, 2=
City Manager
Form
of Government
Results
of the Linear Growth Model for Developmental expenditure, including form of
government as a covariate are given below.
The results show that developmental expenditures are not affected by the
form of government.
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
3.0491
(0.1519) ***
|
||
Rate
of Change,
|
Time
|
0.2821
(0.02270) ***
|
||
FOG
|
-0.04879
(0.09288)
|
|||
Time*FOG
|
-0.00659
(0.001388)
|
|||
Variance Components
|
||||
Level
1
|
Within
person
|
0.1369
(0.006880)***
|
||
Level
2
|
Intercept
|
0.1891
(0.003253)***
|
||
Time
|
0.0003096
(0.00073)***
|
|||
Covariance
|
-0.00382
(0.003666)
|
|||
The Minority Mayor
The
residuals for each city at each time point from the linear growth modelare
usedto further explore the expenditure pattern between white and non-white
mayors. Positive residuals indicate a mayor spent more on the average
development expenditure than predicted; whereas negative residuals indicate a
mayor spent less on the average development expenditure than predicted.
The
parametric t test was performed on these residuals to compare the average
between white and non-white mayors. The results show that there was moderate
statistical significance between the averages, (t= -1.85, p=0.0642) with the mean for non-white larger than that for white.
For the
nonparametric test, we find significance of the race effect on developmental
spending. Wilcoxon test was performed on
these residuals to compare the median residuals between white and non-white
mayors. The results show that there was a statistical significance between the
median residuals for white and non-white mayors (W=84276, Z=1.889, p=0.0294). Kolmogorov-Smirnov test was performed on
these residuals to examine the distribution of residuals for white and
non-white mayor. The results show that there was statistical significance
between the residuals’ distributions for white and non-white mayors (D=0.128 p=0.0352).
Method
2: Minority Mayor as a Level-1 Covariate
Results of the
Linear Growth Model for Developmental expenditure
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
2.9535
(0.04378) ***
|
||
Rate
of Change,
|
Time
|
0.2651
(0.006610) ***
|
||
Race
|
0.1203
(0.04882) *
|
|||
Variance Components
|
Estimate
(Standard Err)
|
|||
Level
1
|
Within
person
|
0.1355
(0.006733) ***
|
||
Level
2
|
Intercept
|
0.1825
(0.03036) ***
|
||
Time
|
0.002896
(0.00069) ***
|
|||
Race
|
0.02073
(0.02536)
|
|||
Covariance
|
-0.00445
(0.003467)
|
|||
0.02674
(0.02226)
|
||||
-0.00372
(0.004029)
|
||||
Time: 0, 1, 2,
3, 4, 5, 6, 7, 8; Race: 1=Black, 0=white
Again, this is
more evidence of the effect of race, since the parameter for race is
significant at the .05 level.
Allocational
Expenditures
Results of the
Linear Growth Model for Allocational expenditure
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
4.0577
(0.03771) ***
|
||
Rate
of Change,
|
Slope
|
0.3379
(0.00551) ***
|
||
Variance Components
|
Estimate
(Standard Err)
|
|||
Level
1
|
Within
person
|
0.07087
(0.003415) ***
|
||
Level
2
|
In
initial status
|
0.1486
(0.02277) ***
|
||
In
Rate of Change
|
0.002558
(0.00049) ***
|
|||
Covariance
|
-0.00319
(0.002442)
|
|||
Form of Government
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
4.2481(0.1288)***
|
||
Rate
of Change,
|
Time
|
0.3580
(0.01846)***
|
||
FOG
|
-0.1189
(0.07897)
|
|||
Time*FOG
|
-0.01315
(0.01132)
|
|||
Variance Components
|
||||
Level
1
|
Within
person
|
0.07067
(0.003537)***
|
||
Level
2
|
Intercept
|
0.1495
(0.02387)***
|
||
Time
|
0.002435
(0.000495)***
|
|||
Covariance
|
-0.00343
(0.002520)
|
|||
The results show
that allocational expenditures are not affected by the form of government.
Method 1 for Assessing Minority Mayor
The
residuals for each city at each time point from the linear growth model are
usedto further explore the expenditure pattern between white and non-white
mayors. Positive residuals indicate a mayor spent more on the average
development expenditure than predicted; whereas negative residuals indicate a
mayor spent less on the average development expenditure than predicted.
T test
was performed on these residuals to compare the average residuals between white
and non-white mayors. The results show that there was no statistical
significance between the average residuals for white and non-white mayors (t=
-0.53, p=0.59).
The
Wilcoxon test was performed on these residuals to compare the median residuals
between white and non-white mayors. The results show that there was no
statistical significance between the median residuals for white and non-white
mayors (W=80517, Z=0.6569, p=0.2556).
Kolmogorov-Smirnov test was performed on these residuals to examine the
distribution of residuals for white and non-white mayor. The results show that
there was no statistical significance between the residuals’ distributions for
white and non-white mayors (D=0.0532, p=0.8764).
Results
of the Linear Growth Model for Allocational expenditure with a level-1 race
covariate confirm the previous results of insignificance of race on
allocational spending.
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
4.0524
(0.03770) ***
|
||
Rate
of Change,
|
Time
|
0.3382
(0.005735)***
|
||
Race
|
0.04099
(0.04159)
|
|||
Variance Components
|
||||
Level
1
|
Within
person
|
0.06828
(0.003394) ***
|
||
Level
2
|
Intercept
|
0.1489
(0.02273) ***
|
||
Time
|
0.002795
(0.000537)***
|
|||
Race
|
0.04164
(0.02113) *
|
|||
Covariance
|
-0.00356
(0.002531)
|
|||
0.004420
(0.01752)
|
||||
-0.00528
(0.002955)
|
||||
Redistributive
Expenditures
Results of the
Linear Growth Model for Redistributive expenditure
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
0.7204
(0.1428) ***
|
||
Rate
of Change,
|
Slope
|
0.4167
(0.01956)***
|
||
Variance Components
|
||||
Level
1
|
Within
person
|
0.5510
(0.02750)***
|
||
Level
2
|
In
initial status
|
2.2866(0.3223)***
|
||
In
Rate of Change
|
0.03647(0.006191)***
|
|||
Covariance
|
-0.2391(0.04048)***
|
|||
Form
of Government:
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
2.0844
(0.3161)***
|
||
Rate
of Change,
|
Time
|
0.3144
(0.04444) ***
|
||
FOG
|
-0.5305
(0.1943) **
|
|||
Time*FOG
|
0.01468
(0.02731)
|
|||
Variance Components
|
||||
Level
1
|
Within
person
|
0.2981
(0.01511) ***
|
||
Level
2
|
Intercept
|
0.9629
(0.1437) ***
|
||
Time
|
0.01587
(0.002901) ***
|
|||
Covariance
|
-0.7206
(0.01684) ***
|
|||
The
results of the Linear Growth Model for Redistributive expenditure are different
that the previous expenditure categories.
Here, form of government is a statistically significant factor for
redistributive (p-value < .01.) The
negative estimated coefficient for
is .-531, indicating that the mayor council form
spends more on redistributive than city manager.
Assessing Minority Status: Analysis of
the Residuals
Residual
analysis can be used to explore the expenditure pattern between white and non-white
mayors. Positive residuals indicate a mayor spent more on the average
development expenditure than predicted; whereas negative residuals indicate a
mayor spent less on the average development expenditure than predicted. We carried out the parametric two sample
t-test to compare the means of the residuals between white and non-white
mayors. The results show that there was moderate statistical significance
between the average residuals for white and non-white mayors (t= -1.89,p=0.0605). The Figure below shows that the average for
minority mayors is larger than for white mayors.
Next I summarize
the nonparametric methods. The Wilcoxon
test was performed on these residuals to compare the median residuals between
white and non-white mayors, and the results show that there was a statistical
significance between the median residuals for white and non-white mayors
(W=82915, Z=2.6636, p=0.0039). The nonparametric Kolmogorov-Smirnov
two-sample test for the residuals’ distribution shows that there was a statistical
significance between the residuals’ distributions for white and non-white
mayors (D=0.1705 p=0.0021, ) with the
distribution for minority mayors having larger percentiles than that for white
mayors.
Method
2: Minority Mayor as a Level-1 Covariate
Results of the
Linear Growth Model for Redistributive expenditure
Parameter
|
Linear
Growth Model
|
|||
Fixed Effects
|
Estimate
(Standard Err)
|
|||
Initial
status,
|
Intercept
|
1.2417
(0.09511) ***
|
||
Rate
of Change,
|
Time
|
0.3322
(0.01314) ***
|
||
Race
|
0.1687
(0.06748) *
|
|||
Variance Components
|
||||
Level
1
|
Within
person
|
0.2949
(0.01444) ***
|
||
Level
2
|
Intercept
|
1.0108
(0.1438) ***
|
||
Time
|
0.01578
(0.002805) ***
|
|||
Race
|
0
|
|||
Covariance
|
-0.0729
(0.01647) ***
|
|||
0.02171
(0.06921)
|
||||
-0.00841
(0.01182)
|
||||
In this linear
growth model, the parameter
is the difference between Black and white mayor in
the log Redistributive expenditure, and is significant at the .05 level. The variance component
is not statistically different from zero, indicating
that the moderate race effect is consistent across cities.
Summary and Conclusions
Municipal officials, i.e., mayors
and councils shape their expenditure decisions by defining the situations, determiningtheirroles,
assessing the appropriateness of different expenditures and responding to the
situations in a manner that meets their challenge within the constraints of
their financial wherewithal. The appropriateness
of the different spending decisions is discussed below.
With regard to developmental
spending the analysis indicated that
such expenditures are not affected by the form of government.Developmental
spending policy has the goal of enhancing the local fiscal base and of
subsequently attracting high-end taxpayers.
It is a policy that virtually every municipality would pursue; a sort of
pump priming activity that is vital to attracting businesses and jobs.Such an
expenditure approach is connected to fiscal capacity theory which argues that
local jurisdictions must fight for wealthier residents and businesses and
simultaneously balance a competitive tax-rate that is low enough to attract
wealthier residents and businesses with the provision of goods and services.
City officials buy into a strict division of labor between
governmental and private sector entities in control and manipulation of
resources (Imbroscio, 1998). Such
spending decisions emphasize the interdependence of governmental and non-governmental
forces in meeting economic and social challenges. Therefore, it is not surprising that we find
no sufficient statistical evidence suggesting that form of government affects
developmental spending.
On the other hand, it does appear that developmental spending
is different between minority and white governed cities. The MGM is challenged to emphasize
developmental spending because there is an understanding of the linkage between
urban government and the business sector. Because MGM cities have a larger
concentration of disadvantaged citizens and they also have the goal of
attracting more business, they will spend more on developmental efforts. As city governments adjust their efforts, it
is not unreasonable to find that expenditure decisions are also affected.
There
was insufficient statistical evidence to show that allocational spending is
affected by the form of government and minority status of mayor. Allocational spending by municipal
governments reduce the likelihood of catastrophic conflagrations, wholesale
violations of personal property, community epidemics and the use of public
spaces as dumps and junkyards.Thus, it is not surprising that the majority of
empirical studies, including this one, have found little difference in the
levels of allocational policies by white and black mayors (Keller (1978);
Nelson (1978). Moreover, our finding
complements Helen Thompson (2001) who examined policy priorities of black
mayors in the United States and subsequently developed a typology of black mayoral
agendas and a profile of their terms. Her study concluded that while there is a
high level of congruence in agenda priorities among the country’s black mayors,
the policies do not appear to be different from those of non-black mayors. Our
research here uses modern statistical methods and complements these findings.
Perhaps the most intriguing finding is that exogenous factors
such as mayoral race and institutional factors such as a form of government are
statistically significant for redistributive spending behavior.This observation
invites a number of noteworthy linkages to previous works. First, I believe my
contribution sheds light on Peterson’s redistributive construct and elucidates
the theory of need responsiveness which suggests that redistributive spending
will increase as the number of needy residents rise.
Likewise, Schneider’s (1987) community homogeneity hypothesis
was conducted at a time when there were not many instances of minority governed
municipalities. The community
homogeneity hypothesis rationalized that, within fiscally heterogeneous
communities, we can expect to see lower redistributive spending because the
nature of redistributive spending is more obvious and thereby prompts vocal
resistance from higher income residents.
It also was conducted when there was a dearth of MGM’s. Since minority
governed cities are more homogeneous, than are majority governed
municipalities, our observation of higher redistributive spending is consistent
with Schneider.
Emergence of MGM’s began gradually after Friesma’s (1969)
characterization of the minority mayor-ship as a “hollow prize”.This “hollow
prize” was a constellation of impoverished central cities with dependent
populations that the new black mayor inherited; it is now demonstrable that
this rising class of new black mayors inherited a set of conditions which would
not easily allow spending policies designed to discourage concentrations of
those dependent populations because they were already in residence.This study
uncovers an acute irony: the“ hollow prize”has become emblematic of unintended
challenges of PL 89-110: herculean leadership challenges with limited resources
for this new set of leaders. It also
convinces us that institutionalism does not account for existence of the hollow
prize, but it does distinguish its presence.
References
Adrian, Charles and Charles Press (1968)
Governing Urban America New York, McGraw-Hill
Advisory Commission on Intergovernmental
Relations, (1963) ‘Performance of Urban Functions: Local and Areawide,
(Washington, DC: U.S. Government Printing Office)
Beckman, Norman (1966) ‘How Metropolitan
Are Federal and State Policies?’ Public Administration Review 96-106.
Booms, B. H.(1966). ‘City Governmental
Form and Public Expenditure Levels’.National Tax Journal 19: 187-199.
Boynton, Robert P., and Victor S.
DeSantis (1990), “Form and Adaptation: A Study of the Formal and Informal
Functions of Mayors, Managers, and Chief Administrative Officers” Baseline
Data Report, 22 (1) Washington, DC:
ICMA
Bureau, U.S.C.(1992)Census of Governments.
Washington, DC, U.S. Department of Commerce
_____________(2000) Statistical
Abstract of the United States.Washington, DC, U.S. Department of Commerce.
Clark, Terry N., ed (1968) Community
Structure and Decision-Making: Comparative Analysis. (San Francisco:
Chandler Publishing Company)
Deno, K.T., and Stephen L. Mehay (1987).
“Municipal Management Structure and Fiscal Performance: Do City Managers Make a
Difference’.” Southern Economic Journal 53: 627-642
Derthick, Martha (1972) New Towns In-Town
(Washington, DC: Urban Institute).
DiMaggio, Paul J. and Walter W. Powell
1983 The Iron Cage Revisited: Institutional Isomorphism and Collective
Rationality in Organizational Fields. American
Sociological Review 48:147-160.
DeSantis, Victor and Tari Renner
(1994).“The Impact of Political Structure in American Counties.”Public
Administration Review54: 291-295
_________________________ (2002).City
Government Structures: An attempt at clarification. In H.G. Frederickson and J. Nalbandian (Eds.)
The future of local government administration: The Hansell symposium (pp
71-80). Washington, DC: International City /County Management association
Dye, Thomas, and John A. Garcia
(1978).“Structure, Function, and Policy in American Cities’.Urban Affairs
Quarterly14:103-122.
Eleazar, Daniel (1990) The American
Mosaic (Boulder, CO: Westview Press.
Fahey, Colleen (1998). The Choice of
Local Government Structure in Massachusetts: A Historical Public Choice
Perspective.” 79:2 Social Science Quarterly 443-444
Farnham, P.G. (1987). “Form of
Government and Median Voter.”Social Science Quarterly68: 569-582.
___________ (1990) “The Impact of
Citizen Influence on Local Government Expenditure’ Public Choice64: 201-212.
Feiock, Jong, and Kim (2003) “Credible
Commitment and Council-Manager Government: Implications for Policy Instrument
Choices.” Public Administration Review63(5): 616-625.
Fleischmann, A., Gary P. Green, and Tsz
Man Kwong (1992) “What’s a City to Do: Explaining the Difference in Local
Economic Development Policies.” Western Political Quarterly 45(3)
Frederickson, H. George, Gary A.
Johnson, and Curtis Wood, ‘The Changing Structure of American Cities: A Study
of the Diffusion of Innovation’, Public Administration Review64 (3), pp320 – 330.
Hannah, Susan B. (2001). ‘Writing Home
Rule Charters in Michigan: Current Practices in Constitution-Making’ in J.R.
Baker, ed. The Lanaham Readings in State and Local Government,
Baltimore, MD, Lanaham Publishers Inc.
Hansell, S.B., (1998, April 23-25). Form
and function in Michigan local government.
Paper presented at the Midwest Political Science Association. Chicago
___________ (1999) Reforming the reform:
Variations on mayor-council form of government, part 2. Public Management, 81
(1): 27-28.
Hayes, K., and S. Chong (1990).“The
Relative Efficiency of City Manager and Mayor-Council Forms of Government.”Southern
Economic Journal57: 167-177.
Helsing, Jeffrey N., (2000) Johnson’s
War/ Johnson’s Great Society: The Guns and Butter Trap (Westport, Conn:
Praeger).
Howard, Ebenezer (1965) Garden Cities
of To-morrow (Cambridge, MA: MIT Press)
Liebert, R., (1974). “Functions and
Expectations: A Re-analysis of Recent Research.” Social Science Quarterly54: 765-783.
Lineberry, R. L. and E.P. Ferguson
(1967).“Reformism and Public Policies in American Cities.”American Political
Science Review61 (September):
701-717.
Lyons, W. (1978) “Reform and Response in
American Cities: Structures and Policy Reconsidered.’” Social Science
Quarterly15: 118-132.
Lynn,
Laurence E., Jr. 1998 The New Public Management: How to Transform a Theme Into
A Legacy. Public Administration
Review, 58:3 (May/June): 231-237.
March, James and J.P. Olsen (1989) Rediscovering
Institutions (New York: NY) the Free Press
Marcuse, Herbert (2001) Towards a
Critical Theory of Society (New York: Rutledge)
Maser, Steven M. (1985) ‘Demographic
Factors Affecting Constitutional Decisions: The Case of Municipal Charters’. Public
Choice 47 (2): 121-162
McDonald, Lynn (2005). Government
Structure and Local Public Expenditures
Miller, Gary (1985). ‘Progressive Reform
and Induced Institutional Preferences’.Public Choice 47(2) 163-181
__________ 2000. ‘Above Politics: Credible Commitment
and Efficiency in the Design of Public Agencies. Journal of Public
Administration Research and Theory 10 (2) 289-327.
Morgan, D. R., and John P. Pelissero
(1980). “Urban Policy: Does Political Structure Matter?” American Political
Science Review 74: 999-1006.
National League of Cities (2006) ‘About
Cities: Municipal Expenditures’ http://www.nlc.org/about_cities
_101/179.cfm?TYPE=popup
Ostrom,
V., C. M. Tiebout, and R. Warren.(1961) "The Organization of Government in
Metropolitan Areas: A Theoretical Inquiry." The American Political Science
Review 55 (4):831-42.38
Ostrom, V., Robert Bish, and
ElinorOstrom (1988).Local Government in the United States.San Francisco,
CA, Institute for Contemporary Studies.
Osterlind, Steven J., and Barbara
Tabachnick (2001) SPSS for Windows Workbook to Accompany Using Multivariate
Statistics, 4th ed., Allyn and Bacon
Rehfus, John and Eric Stow, Eds. (1975)
‘A Symposium: The Governance of New Towns’ Public Administration Review
v 35 (May/June).
Renner, Tari (1988) “Elected Executives:
Authority and Responsibility”, Baseline Data Report V 20 (3) Washington,
DC: International City Management Association
Reid, G. J., (1991) “tests of
Institutional Versus Non-Institutional Models of Local Public Expenditure
Determination.” Public Choice70:
315-333.
Rogers , J. F. 1959 Institutionalism versus
Anti-Institutionalism in American Education. Educational Theory, 9: 169–173.
doi: 10.1111/j.1741-5446.1959.tb01263.
Sass, Tim R. (1991). “The Choice of
Municipal Government Structure and Public Expenditures”, 71Public Choice 71-87
Svara, James H. (1990) Official
leadership in the city: Patterns of conflict and cooperation. New York: Oxford
University Press.
_____________ (1994) Facilitative
Leadership in Local Government: Lessons from Successful Mayors and Chairpersons
(San Francisco: Jossey-Bass) 1sted.
____________ (1987) “Mayoral Leadership
in Council-manager Cities: Preconditions versus Preconceptions.” Journal of
Politics49: 207-222.
Tiebout, Charles (1956) “A Pure Theory
of Local Expenditures.”Journal of Political Economy 64: 416-424.
1997 Census of Governments (1999)
State and Local Finances v 4 (Washington, DC: U.S. Department of Commerce.
[1]
As defined for census statistics on governments, municipal governments
are political subdivisions within which a municipal corporation has been
established to provide general local government for a specific population
concentration in a defined area. This includes all active governmental units
officially designated as cities, boroughs, (except in Alaska), towns, (except
in the six New England States, Minnesota, New York, and Wisconsin), and villages.
In Alaska, the term ‘borough’ corresponds to units classed as county
governments. In New England, Minnesota,
New York, and Wisconsin, the term ‘town’ refers to an area subdivision which
may be legally termed a municipal corporation and have a similar governmental
organization but has no necessary relationship to a concentration of
population, and thus corresponds to the term ‘township in other states.
Tidak ada komentar:
Posting Komentar
For yout correction, write your comment in here. Thank you.
(Tulislah komentar anda di sini untuk perbaikan. Terima kasih)