Political Budget Cycles and the Civil Service:
Evidence from Highway Spending in US States
David Bostashvili
Gergely Ujhelyi
Amazon.com
Department of Economics
[email protected]
University of Houston
[email protected]
January 28, 2017
Abstract
Civil service protections are thought to stabilize government by ensuring continuity
in bureaucracies around elections. We show that these rules also stabilize by dampening
politicians’incentive to create budget cycles. Civil service reforms undertaken by US
states in the second half of the 20th century provide an opportunity to study political
budget cycles in highway spending with and without a civil service system. We …nd that
without civil service highway spending is 12% higher in election years and 9% higher
in the year before an election. By contrast, under civil service highway spending is
essentially smooth over the electoral cycle.
For useful comments on an earlier version of this paper, we thank Sutirtha Bagchi, Steven Bednar,
Aimee Chin, Michael Conlin, Steven Craig, William Hankins, Elaine Liu, Vikram Maheshri, Bent Sorensen,
Michael Ting, Andrew Zuppann, and participants at the 14th Missouri Economics Conference and meetings
of the Public Choice Society, APET, National Tax Association and SEA. The views expressed here are the
authors’and do not represent those of Amazon.com or its a¢ liates.
1
Introduction
One of the main arguments for giving bureaucrats civil service protections such as tenure
is that they ensure the stability of government around elections. Without these protections
the practice of political patronage is believed to result in large turnover in bureaucracies as a
newly elected administration replaces state employees with its trusted allies and supporters.
In this paper we investigate another way in which a civil service system can create stability
in government activity: by reducing the political budget cycle in government expenditures.
As shown by a large literature, politicians have an incentive to adjust …scal policies during
the course of an election cycle. For example, they have an incentive to spend on projects
favored by voters in election years while delaying less popular expenditures until after the
election. Such incentives can give rise to political budget cycles: cycles in expenditures (or
other …scal categories) within a politician’s term in o¢ ce. The literature emphasizes that
political budget cycles are not a general phenomenon but rather depend on institutional
factors such as the age of democracy, electoral rules, or political transparency (Persson and
Tabellini, 2003; Brender and Drazen, 2005; Shi and Svensson, 2006; Alt and Lassen, 2006).
In this paper we propose that bureaucratic institutions should also be expected to shape
political budget cycles.
Political decision makers have less incentive to time spending decisions to the electoral
cycle if they have less control over how the money is spent. For example, a governor may have
an incentive to increase highway spending before the election if new construction projects
can freely be targeted to politically important constituencies. But if the location of projects
is determined to a large extent by career bureaucrats who enjoy civil service protections and
are guided by technical rather than political considerations, the governor’s ability to target
projects may be limited. In this case, increasing highway spending before the election may
not be worthwhile. Thus, a civil service system can dampen the political budget cycle.
To test this idea, we study highway spending by US state governments in the second
half of the 20th century. Fueled by the federal Highway Trust Fund and the Interstate
Highway program, highway construction and maintenance was one of the biggest areas of
state government activity in this period.1 This is an area requiring technical expertise, but
also one that can be highly politically lucrative if a politician can make sure that a new
road segment is built at the right place at the right time. To study highway spending with
and without civil service, we take advantage of recently collected data on the timing of civil
service reform in US state governments.
1
In spite of its magnitude, studies of the highway program are scarce. A notable exception is Knight
(2002) who studies the political economy of federal highway transfers in the period 1983-1997.
2
Throughout the 20th century US state governments changed their bureaucratic organization from political patronage to civil service (the “merit system”), introducing merit-based
recruitment and other protections from politics. While the reforms were similar, they occurred at di¤erent times in di¤erent states and we use them as a natural experiment to study
the impact of civil service on the budget cycles in highway expenditures. We study both
the introduction of the statewide merit system and, using newly collected information, the
introduction of merit systems speci…cally in states’highway departments.
Our …ndings provide strong evidence for political budget cycles in highway spending
that are conditional on bureaucratic organization. Under patronage, we …nd that highway
spending is 9% higher in the year before an election and 12% higher in the election year
as compared to the …rst year immediately following the election. By contrast we …nd no
evidence of political budget cycles under civil service. Under civil service, highway spending
is essentially smooth over the electoral cycle. These …ndings survive a variety of robustness
checks, including di¤erent samples and estimation methods.
These results generalize the idea that a civil service system can ensure the “stability”and
“continuity”of government activity in election times. While it is natural that, compared to
patronage, civil service rules create stability within the bureaucracy, our …ndings show that
this stability can extend to other areas of government as well - in particular, to the policy
choices of election-minded politicians.
2
Related literature
Our paper brings together two distinct and large literatures: one on bureaucracies and the
other on political budget cycles.
With respect to bureaucracies, the broad question we address is: What is the impact of
civil service rules? Several theoretical studies relate to this question at least indirectly by
studying society’s incentives to delegate decisions to independent, expert bureaucrats (e.g.,
Maskin and Tirole, 2004; Alesina and Tabellini, 2007, 2008) or politicians’incentives to do the
same (Epstein and O’Halloran, 1999; Gailmard and Patty, 2007; Fox and Jordan, 2011; Ting,
2002, 2012). Ujhelyi (2014a) asks about the welfare e¤ects of civil service rules when they
a¤ect the interaction of politicians and bureaucrats. The empirical literature on civil service
rules is much smaller as the di¢ culty of obtaining comparable data on institutional reforms
in a large number of jurisdictions often makes identi…cation challenging. Early studies of the
economic impact of the civil service include Rauch (1995) (infrastructure investment in US
cities), Rauch and Evans (2000) (cross-country growth), and Krause et al. (2006) (accuracy
of state revenue forecasts). To improve identi…cation, the recent literature has focused on
3
civil service reforms in US states, …nding that it is more di¢ cult for politicians to get reelected
under civil service than under patronage (Folke et al., 2011), and that politicians circumvent
state bureaucracies that are under civil service by using intergovernmental transfers (Ujhelyi,
2014b). The present paper continues the agenda of seeking to understand the impact of civil
service rules on politicians’ behavior by asking how the civil service shapes government
spending over the electoral cycle.
With respect to political budget cycles, the broad question we address is: What institutional factors shape the cycles?2 The literature has found signi…cant di¤erences in budget
cycles between developed and developing countries and between “new”and “old”democracies (Brender and Drazen, 2005; Shi and Svensson, 2006). Recent studies have investigated
some of the institutional di¤erences that could account for this. In particular, studies have
found that cycles are more likely to be present under less budgetary transparency, less media freedom, and more corruption (Akhmedov and Zhuravskaya, 2004; Shi and Svensson,
2006; Alt and Lassen, 2006). Persson and Tabellini (2003) …nd more pronounced cycles
in majoritarian and presidential systems, and across US states Rose (2006) …nds that the
cycles are weakened by strong balanced-budget requirements. The present paper continues
the agenda of looking for institutional di¤erences in political budget cycles and proposes
that bureaucratic institutions may be a relevant source of heterogeneity. Since the strength
of civil service protections di¤ers between countries, this could provide a complementary
explanation for the observed di¤erence in the budget cycle of developed and developing
countries.
Apart from a focus on a new source of institutional heterogeneity, our empirical setting
has several advantages for identi…cation relative to some of the earlier studies. First, we are
studying heterogeneity in actual institutions rather than in an index of perceived institutions.
Second, in our period of study bureaucratic institutions change within state over time, so
we are not simply comparing budget cycles in the cross section. Third, we have the usual
bene…ts of within-country data that, compared to cross-country analyses, hold …xed a variety
of institutional and economic factors. Particularly important in our case is the fact that
elections are held at the same exogenously …xed dates.
We follow the recommendation of the recent literature and look for political budget
cycles in speci…c budget items rather than broad categories like total spending or de…cits
(Drazen and Eslava, 2010). Highway spending decisions in US states provide an economically
important area to look for cycles, and, given the extensive involvement of bureaucrats in
highway projects, this is an ideal setting to study the impact of bureaucratic institutions on
2
See Drazen (2001) and Eslava (2011) for surveys of the political budget cycle literature.
4
these cycles.3
3
3.1
Background
Highway spending by state governments
“Highway expenditures” cover a broad range of expenditures, including the “construction,
maintenance, and operation of highways, streets, and related structures, bridges, tunnels,
ferries, street lighting and snow and ice removal.”4 In monetary terms, highways were one
of the main areas of government activity in US states for most of the 20th century. In the
1950s and 60s in the average US state highway expenditures accounted for 25-35% of all state
government spending. This share declined over time but remained above 10% throughout
the 1970s and 80s (see Figure 1).
0
0
.05
500
million $
1000
share of direct expenditures
.1
.15
.2
.25
.3
.35
1500
Figure 1: Direct highway expenditures by US state governments
1950
1960
1970
year
1980
1990
2000
Notes: Direct highway spending as a share of total direct spending (solid line) and in 2009 dollars (dashed
line) in the average US state.
Highway expenditures are a politically important spending category (Knight, 2002).
Apart from their potential long-term e¤ects on economic development, these projects are
often highly visible and a¤ect the everyday life of a large number of political constituents.
3
Evidence of political budget cycles in capital projects such as road construction has previously been
found in Israel (Brender, 2003), India (Khemani, 2004), and Colombia (Drazen and Eslava, 2010) among
others.
4
US Census Bureau, Census of Governments, http://www.census.gov/govs/state/de…nitions.html.
5
Several features of the US system of highway …nance suggest that these expenditures can
be particularly valuable political tool for state governments. The construction, ownership,
and maintenance of highways is the responsibility of state governments. This principle was
codi…ed in the Federal-Aid Road Act of 1916 and the states have actively resisted attempts
by the federal government that they perceived as infringing on this responsibility.5 State
governments are responsible for deciding which projects are undertaken, where they are
located, and who is hired to work on them.
While highway projects are the states’ responsibility, funding for these projects comes
mainly from the federal government who reimburses and in some cases advances most of
the costs. The 1956 Federal-Aid Highway Act establishing the Highway Trust Fund for the
development of the interstate highway system set the federal funding share at 90 percent.
Since then, the federal share has varied across projects but has typically remained above 75
percent. Over time, the range of projects qualifying for federal funding has expanded significantly, encompassing not just construction and maintenance of the roads themselves, but
also public transportation projects (e.g., bus lanes or replacing unwanted highway segments
with rail systems), highway beauti…cation and safety projects (including landscaping), parking lots, bridges, parkland preservation, the acquisition of rights-of-way, relocation assistance
to those a¤ected by construction, and the purchase of ferry boats (see CBO, 1978).
Thus highway expenditures are politically attractive to state governments not only because they can bene…t speci…c constituents and are highly visible, but also because they are
funded largely from external sources. There is extensive anecdotal evidence on the in‡uence
of politics on highway projects. One common phenomenon is project location being guided
by political considerations. Others include various forms of corruption documented in several
congressional reports investigating the escalating costs of the Interstate Highway Program
(see Dilger, 1989). Observers have noted that patronage states were particularly prone to
ine¢ ciencies:
“One very expensive by-product of partisanship in highway practice is the passing-over by a
new highway administration of roads already planned and designed by its predecessor. Also, in
spoils states, the administrator may retain favored consultants to design projects which are not
constructed; he may change priority ratings after preliminary engineering is partly or completely
done. The practice sometimes even involves acquisition of right of way for projects never completed.
In certain cases, parts of fully-engineered projects are constructed and the remaining technical work
discarded.” (Martin, 1959, p169)
5
For example, when the federal government made e¤orts to impose investment standards for highway
projects, language had to be inserted in a 1973 bill to reassure the states that the provision of federal funds
“shall in no way infringe on the sovereign rights of the States to determine which projects shall be federally
…nanced.” (CBO, 1978, p56).
6
In this paper, we document one particular feature of the politics of state highway spending: the political budget cycles in highway expenditures.
3.2
Civil service reform
Throughout the 20th century US states changed the systems of personnel management in
their bureaucracies, replacing patronage (or the “spoils system”) with civil service (or the
“merit system”). Under patronage, public employees from top cabinet positions down to
maintenance personnel were bound to their political patrons, were expected to provide political services such as contributions to campaign funds, and were likely to be replaced when
the administration was voted out of o¢ ce. Under a merit system, hiring was based on open
competitive examinations, political services were prohibited, and the emphasis was placed
on …lling bureaucracies with people with technical expertise and committed to a career in
the civil service. While the patronage system ensured that bureaucrats would be loyal to
the politicians in o¢ ce, the goal of the merit system was to make them independent.
Modeled after the federal Pendleton Act of 1883, the statewide merit systems adopted
by the states had a comparable set of provisions. They all contained provisions for meritbased recruitment and protection from politics, and provided for a bipartisan Civil Service
Commission or similar body to supervise implementation and enforce the rules. Although
all states except Texas eventually adopted a statewide merit system, adoption of the reforms
was slow: 22 states adopted the merit system after 1950 (see Ujhelyi (2014b) for details).
What was the cause of civil service reform in the states? The empirical evidence suggests
that politicians themselves had little to gain from giving up patronage: following the reforms
there was a reduction in the number of public employees who could potentially provide
political support (Ujhelyi, 2014b) and politicians forced to give up patronage had trouble
getting reelected (Folke et al., 2011). In the more descriptive literature, while there are
speci…c instances where an incumbent governor may have used merit protections to “lock
in” his loyal supporters, historians and public administration scholars generally agree that
the reform movement was bottom-up, fueled by the good government movement rooted
in the Progressive Era. The main driver of reform appears to have been pressure from
various citizen groups and the voters themselves (National Research Council, 1952; Tolchin
and Tolchin, 1971; Mosher, 1982; Ingraham, 1995). In several instances the transition to a
statewide merit system was initiated by a referendum and codi…ed in the state’s constitution.
In recent years a number of states followed the federal government’s lead and engaged
in a new wave of bureaucratic reforms. This process, which started at the federal level with
the 1978 Civil Service Reform Act and at the state level with a 1996 reform in Georgia
7
has, in many respects, gone in the opposite direction than the earlier reforms, weakening
civil service protections and aiming to increase bureaucrats’responsiveness to managers and
policy makers. A number of states are currently engaged in reforms along these lines, which
makes understanding the impact of merit system protections of current policy relevance.6
3.3
Highway department merit systems
While most states did not establish a centralized statewide merit system until the second
half of the 20th century, every state had speci…c departments with their own merit systems
before then.7 To check whether this was the case for highway departments, we collected new
data on the timing of merit system adoptions in these departments. Speci…cally, for each
state that did not have a statewide merit system in 1960, we checked whether its highway
department adopted its own merit system before the statewide merit system was introduced.
To do this, we relied on several sources, including contemporary news reports, government
documents, and a 1952 study by the Highway Research Board of the National Research
Council (NRC, 1952).8
As it turns out, in our period of study only 5 states had their highway departments
introduce a merit system before the statewide merit system was established: Arizona, Idaho,
Texas, South Carolina, and Washington (see Table 1).9 For all other states that did not have
a statewide merit system by 1960, their highway department came under civil service when
the statewide merit system was eventually established.
Given our focus on highway expenditures, how should we think about situations when
a state’s highway department adopted its own merit system prior to the statewide system?
One can argue that, because highway expenditures go through the bureaucrats in state
highway departments, what is relevant is whether this department is under a merit system
at the time when spending occurs. The highway department’s merit system is the most
likely to represent a constraint for politicians’ability to in‡uence where and how projects
are undertaken, who is hired to work on them, etc.
6
Because these second wave reforms are much more heterogenous than the …rst wave of reforms, studying
them directly is left for future research.
7
One example are departments administering funds under the Social Security Act (such as public health
and social welfare). In 1939-1940 federal legislation mandated the introduction of merit systems for these
departments in all states as a condition for funding. No such requirement was adopted for any other
department, including highway departments.
8
See the Appendix for details.
9
The case of Texas is unusual: although it never had a statewide merit system, since 1940 its Highway
Department has operated a de facto merit system, with all key positions occupied by career engineers
promoted through the ranks based on merit. The department’s independence from state politics is reinforced
by a 1946 amendment to the State Constitution establishing a dedicated fund that e¤ectively removes
legislative control over the department’s budget (see Gri¢ n, 1974; Morehead, 1984).
8
Table 1: Timing of statewide and highway department-speci…c civil service reforms in the
sample
State
Statewide merit
system introduced
West Virginia
1989
Mississippi
1977
Montana
1976
North Dakota
1975
South Dakota
1973
Arkansas
1969
South Carolina
1969
Arizona
1968
Delaware
1968
Florida
1967
Idaho
1967
Iowa
1967
Pennsylvania
1963
Utah
1963
New Mexico
1961
Washington
1961
Kentucky
1960
Texas
-
Prior highway department
merit system introduced
1950
1957
1951
1955
1940
Notes: A missing year in the second column indicates that the state highway department
…rst came under a merit system when the statewide civil service was established. States not
listed introduced a statewide merit system before 1960. Texas never had a statewide merit
system. Sources: Dates of statewide merit system adoptions are from Ujhelyi (2014b). See
the Appendix for the sources of the highway department-speci…c information.
On the other hand, one can also argue that a merit system speci…c to a particular
department is qualitatively di¤erent from a statewide civil service system. For example,
enforcement of the department-speci…c merit system may not be as vigorous when state
government as a whole is still under patronage (National Research Council, 1952).10 Even
if a department-speci…c merit system functions ‡awlessly, it may not create the same constraints for politicians as a civil service system covering most bureaucrats does. For example,
highway construction projects can involve other departments besides highways (e.g., agriculture/forestry, health, etc.). If some of these bureaucrats enjoy civil service protections while
others do not, a politician could still be able to in‡uence, e.g., the location of the project by
exerting pressure on some of the decision-makers.
10
Similarly, while the introduction of statewide civil service can re‡ect voter pressure for good government,
a department-speci…c merit system may serve other purposes, such as “locking in” employees loyal to the
current administration.
9
In the main analysis we do not take a stance on this issue and simply show that we get
similar results using either the statewide or the department-speci…c civil service reforms. We
brie‡y explore potential di¤erences in the impact of the two reforms in section 6.3.
4
Data
Our period of study is 1960-1995. The starting date re‡ects two considerations: data availability (in particular for the citizen ideology measure discussed below), and the establishment
of the federal Highway Trust fund in 1956. The latter not only gave a boost to highway construction throughout the US, it also lowered state governments’cost share to 10% on most
projects. This likely made it easier for state politicians to use highway spending as a political vehicle and we expect state government behavior to di¤er before and after the 1956
act. The end date of the study period also re‡ects two considerations: a need to have as
long a panel as possible to avoid a bias in …xed e¤ects regressions with lagged dependent
variables (see below), and making sure that we are comparing similar institutional reforms.
As discussed in section 3.2, a 1996 reform in Georgia began a second wave of civil service
reforms at the state level, and including these would be di¢ cult due to the heterogeneity in
their provisions.11
Our main outcome of interest, per capita real highway expenditures by state governments,
comes from the US Census Bureau’s Census of Governments. We restrict attention to direct
expenditures (expenditures made directly by the state government as opposed to transfers
to local governments) which account for 85% of state government’s highway expenditures in
the average state.12
Our main independent variables are the merit system indicators discussed above and
indicators for the gubernatorial cycle in the state (election year / election year minus one
/ election year minus two, with the post-election year serving as the omitted category).
Focusing on the election of the chief executive (here, the governor) is standard and re‡ects
the idea that he is the actor with the most in‡uence (including in some cases formal veto
power) over government spending decisions. In our study period, two states held governor’s
elections every two years, while several states moved from a two to a four-year cycle. Because
politicians’ behavior in a two-year cycle is likely to be fundamentally di¤erent from their
behavior in a four-year cycle, we restrict attention to four-year cycles. This gives us a total
of 359 election cycles in 44 states.13 As we show below, our results are robust to restricting
11
As we show below, the results are robust to considering shorter sample periods.
See Ujhelyi (2014b) for a detailed analysis of the impact of civil service reform on intergovernmental
expenditures.
13
Excluded are New Hampshire and Vermont which have two-year cycles, and Rhode Island which switched
12
10
attention to those states that had a four-year cycle for the entire sample period.
As control variables, we use characteristics common in the literature on institutions and
policy outcomes (Besley and Case, 2003). In particular, we control for government resources
such as the tax base, measured by state real per capita income and its squared, as well as for
demographic variables - population size and its squared, and the fractions of state population
that are school-aged (5–17) and elderly (over 65) - to capture the demand for government
services. We also control for political characteristics that might be correlated with political
cycles, the introduction of the merit system, and expenditures. We include a dummy for
Republican control of both houses of the state legislature, a dummy for Democratic control
of both houses, as well as an indicator for the governor’s party a¢ liation. We also include
the Berry et al. (1998) measure of voter ideology, which creates an index of voter liberalism
by using the ideology rating of congressional candidates and their vote shares. Finally, we
include the percentage of urban population which is likely to a¤ect highway construction
and which also has been suggested as a potential correlate of civil service reform (Ruhil and
Camoes, 2003).
With an exercise like ours, the timing of the variables matters. Governments report
expenditures by …scal year, which typically run from July 1 of the previous year to June 30
of the given year (e.g., …scal year 1970 ran from July 1, 1969 to June 30, 1970). Elections are
held in November, and we therefore match (for example) the 1970 election year with …scal
year 1970.14 Thus, we match election cycle indicators to expenditures based on the …scal
year. However, because the …scal year 1970 budget was set in 1969 and spending from this
budget began in 1969, all other independent variables are lagged by 1 period. For example,
a merit system adopted in calendar year 1969 is matched to expenditures in …scal year 1970
as it is unlikely to a¤ect expenditures made in …scal year 1969. This also ensures that our
independent variables can reasonably be considered predetermined in the regressions below.
Table 2 reports the de…nitions and summary statistics of all the dependent and independent variables used in the empirical analysis. The data sources are given in Appendix
A.
to a four-year cycle in 1994. As is standard in the literature, we also exclude Alaska and Hawaii which are
considered …scal outliers, and Nebraska, which has a nonpartisan legislature.
14
This follows the standard way of matching in the literature which ensures that most spending matched
to a given year occurred before an election was held in that year (e.g., Brender and Drazen, 2005).
11
12
3
De…nition
Mean Std. Dev. Min
Max
Real per capita direct expenditures on highways
320.74
163.36
94.02 1380.73
Real per capita direct expenditures
2164.82
761.88
631.10 4844.17
Real per capita direct expenditures on regular highways 308.69
164.88
79.55 1380.73
Real per capita expenditures on toll highways (all such
12.05
21.32
0.00
229.35
expenditures are direct expenditures)
1 if statewide merit system is in place
0.91
0.29
0
1
Indicators for the election year and the number of years
before the next election
Log (state population in 1000)
8.09
0.98
5.67
10.36
Fraction of population aged 5-17
0.22
0.03
0.15
0.31
Fraction of population aged > 65
0.11
0.02
0.04
0.19
Annual income per capita ($1000)
24.52
5.81
8.90
43.95
Fraction of urban population
0.67
0.14
0.36
0.93
1 if Democratic party has a majority in both houses of
0.60
0.49
0
1
the state legislature
1 if Republican party has a majority in both houses of
0.20
0.40
0
1
the state legislature
1 if governor is a Democrat
0.62
0.49
0
1
Measure of citizen ideology (liberalism)
0.45
0.17
0.01
0.94
Notes: All monetary values are in real 2009 dollars. N = 1387. See the Appendix for data sources.
Governor’s party
Citizen ideology
Rep. control
Population
Kids
Aged
Income
Urban
Dem. control
Merit
Ele0 , Ele 1 , Ele 2 , Ele
Variable
Highway expenditures
Direct expenditures
Regular highway expenditures
Toll highway expenditures
Table 2: Variable de…nitions and summary statistics
5
Speci…cation
Our speci…cation follows the standard approach to estimating political budget cycles in the
literature. To test for the possibility that these cycles di¤er under civil service and patronage,
we include the M erit variable and its interaction with the electoral cycle. Speci…cally, we
estimate
yst =
0
X
(
Elest +
Elest
M eritst ) + M eritst
(1)
= 2
+ ys;t
1
+ X0st +
s
+
t
+ "st ;
where yst is per capita highway expenditures in state s in year t. The indicators Elest capture
the election cycle, with Elest taking a value of one years from the next election (Elest3 , or
the post-election year, is the omitted category). The variable M eritst takes the value of one
if a statewide merit system is in place in year t. Control variables include lagged highway
expenditures ys;t 1 , the various time-varying state characteristics X0st described above, and
state and year …xed e¤ects. The Robustness section contains alternative speci…cations,
including one where all the state characteristics X0st are also interacted with the merit system
indicator.
The coe¢ cients of particular interest in Equation (1) are the ’s and ’s. The coe¢ cients
capture the presence of political budget cycles under patronage (M eritst = 0), while the
coe¢ cients
measure the di¤erence in this cycle under civil service. For example, 0 > 0
would indicate the presence of an election year budget cycle under patronage and 0 < 0
would indicate that this is dampened by the merit system.
Our empirical setting o¤ers several advantages for identi…cation. First, unlike some of
the institutions studied in the previous literature, bureaucratic organization changed over
time within states during our period of study. Thus, we are identifying the and parameters both by comparing the budget cycle across civil service and patronage states, and
by comparing changes in the budget cycle when a state switches from patronage to civil
service. Second, we avoid the serious identi…cation concerns that arise if election dates are
endogenous. Elections in our setting are always held on the same date, …xed exogenously.
As is well-known, simple …xed e¤ects estimates of (1) give biased results when the number
of periods is small due to the presence of the lagged dependent variable. Because most papers
in the political budget cycle literature study short panels of 10-20 periods, they often use
variants of the di¤erence GMM estimation methods proposed by Holtz-Eakin et al. (1988)
and Arellano and Bond (1991) to overcome this bias (e.g., Shi and Svensson, 2006; Drazen
and Eslava, 2010). Our panel, which contains 35 years for most states, is closer to a length for
13
which the standard …xed e¤ects estimates is typically viewed as appropriate. For example,
Judson and Owen (1999) recommend using the standard …xed e¤ects speci…cation for panels
longer than 30 periods.15 Studies of political budget cycles using standard …xed e¤ects
estimation include Persson and Tabellini (2003) and Brender and Drazen (2005), who study
panels of length 38 and 41, respectively. In the main analysis we follow these authors and
use standard …xed e¤ects estimation. We then show that our results are unchanged using
the more involved Arellano-Bond type strategies.
6
6.1
Results
Main result
Table 3 reports the results of estimating speci…cations in which the dependent variable is
real per capita direct expenditures on highways (in 2009 dollars). The …rst column is a
benchmark speci…cation that tests for the presence of a cycle in the average state, without
di¤erentiating between civil service and patronage. The results do not show any evidence of
a political budget cycle: highway spending in any year of the political cycle is statistically
indistinguishable from the …rst year.
In the second column, we interact the election cycle indicators with Merit to allow for
the possibility that political budget cycles di¤er under patronage and civil service. Once this
institutional heterogeneity is accounted for, our estimate of the political budget cycle under
patronage becomes large and statistically signi…cant. This estimated political budget cycle
is illustrated on panel A of Figure 2. The point estimates indicate that per capita highway
spending in election years is $38.41 higher than in the year after the election. In addition,
highway spending in the year before an election is also larger, by $29.81 per capita. Finding
electoral e¤ects on highway spending in the pre-election year is not surprising given that
highway construction projects can take a long time to complete. Having a project completed
in an election year may require expenditures in the previous year. Compared to average
spending in a post-election year, these …gures represent increases of 9% (pre-election year)
and 12% (election year), respectively.
Column (2) of Table 3 shows that the estimated budget cycle is only present under
patronage and disappears under civil service. For election years and pre-election years, the
interactions of the political cycle indicators with Merit are statistically signi…cant, and,
compared to the estimates under patronage, have similar magnitudes but the opposite signs.
15
Using the Arellano-Bond type methods in long panels is not without costs. The large number of potential
instruments creates di¢ culties for identi…cation and model selection (see Roodman (2009a) for a detailed
discussion).
14
Table 3: Political cycles in highway spending and the merit system
(1)
6.51
(4.01)
1.05
(4.22)
-6.12
(4.03)
0
Ele
Ele
1
Ele
2
Ele0
Merit
Ele
1
Merit
Ele
2
Merit
Merit
Merit system:
R2
N
0.68
1,387
(2)
38.41***
(9.40)
29.81**
(12.42)
7.35
(14.77)
-35.95***
(9.60)
-32.94**
(12.72)
-15.34
(14.55)
17.57*
(8.79)
Statewide
0.68
1,387
(3)
40.37***
(12.96)
39.43**
(15.95)
13.46
(18.91)
-37.27***
(13.50)
-42.67**
(16.27)
-21.78
(18.75)
18.87*
(11.12)
Department-speci…c
0.68
1,387
Notes: The dependent variable is real per capita highway expenditures. Regressions control for state and year …xed e¤ects, lagged highway expenditures, log state
population and its square, real per capita income and its square, the fraction of population aged 5-17 and the fraction aged 65 and over, Dem. control, Rep. control,
the governor’s party, urbanization and citizen ideology. Robust standard errors clustered by state in parentheses. ***, **, * denote signi…cance at 1, 5, and 10 percent,
respectively.
Under civil service, the impact of election years is only $2.46 per capita (= 38:41 35:95) and
the impact of the pre-election year is $-3.13 (= 29:81 32:94), neither of which is statistically
di¤erent from 0.16 This is illustrated on panel B of Figure 2. Introduction of the civil service
appears to dampen the political budget cycle in highway expenditures.
In column (3), we use the highway department-speci…c merit system rather than the
statewide merit system as our civil service indicator. As can be seen, the results are very
similar. We explore the implications of having only a department-speci…c merit system
without statewide civil service in section 6.3 below.
6.2
Robustness
In this section we present several robustness checks on our main result.
16
In the second year of the cycle we observe a small but statistically signi…cant $7.99 decline compared to
the post-election year.
15
Figure 2: Political cycles in highway spending under civil service and patronage
Notes: The …gure shows the political budget cycles in highway expenditures (in real 2009 dollars per
capita) implied by the estimates in Table 3, column (2) under patronage (Panel A) and civil service (Panel
B). In each case the base category is the post-election year, normalized to 0. The 95 percent con…dence
interval is shown by the grey lines.
6.2.1
Estimates on a balanced panel
Because we restrict attention to 4-year gubernatorial terms, the set of states in the sample
changes over time. In particular, states move from 2 to 4 year terms throughout our sample
period. Does the changing set of states a¤ect our results? In Column (1) of Table 4 we
present regressions on a balanced panel of states, using only the 32 states that had 4-year
terms throughout the sample period.17 Our results are very similar to those obtained earlier,
indicating that the changing set of states does not a¤ect our …ndings.
6.2.2
Di¤erent time period
Our main estimates covered the period 1960-1995. There are two reasons to wonder whether
the estimates are robust to considering a shorter period. First, the nature of highway spending changed over time: while the focus in the earlier period was on construction, later
projects fell increasingly under maintenance. Dilger (1989) suggests that 1983 was a turning
point in this respect (see also Knight, 2002). Second, the approach to civil service reform
changed over time (see section 3.2). After an emphasis on merit system protections and
bureaucratic independence during the …rst wave of civil service reforms, the second wave
17
Two states, Florida and Illinois, had 4-year terms throughout the sample period but moved gubernatorial
elections from presidential election years to midterm election years (in 1966 and 1978, respectively). In both
cases this resulted in one 2-year term for the governor in o¢ ce at the time of the change and we exclude
these 2-year terms from the sample for those states.
16
Table 4: Political cycles in highway spending and the merit system, robustness
Dep. Var.:
Balanced panel Before 1983
(1)
(2)
Ele0
39.78***
42.72***
(12.65)
(11.87)
1
Ele
27.69**
30.59**
(11.78)
(12.83)
2
Ele
16.19
6.71
(17.60)
(13.70)
Ele0 Merit
-38.29***
-34.27***
(12.62)
(11.60)
Ele 1 Merit
-32.19**
-34.26**
(11.96)
(13.40)
2
Ele
Merit
-24.75
-14.11
(17.31)
(14.67)
Merit
20.48**
15.32
(9.76)
(10.14)
R2
0.72
0.68
N
1,110
862
With full interactions
(3)
30.79***
(10.74)
26.24**
(12.35)
-5.09
(16.16)
-28.27**
(11.56)
-28.89**
(12.89)
-2.15
(16.47)
11.88
(9.22)
0.93
1,387
Notes: The dependent variable is real per capita highway expenditures. Column (1) restricts attention to the 32 states with 4-year gubernatorial terms throughout the sample period. Column (2)
restricts attention to 1960-1983. Column (3) includes interactions of the election cycle indicators
with all the control variables. Here reported coe¢ cients on Ele0 , Ele 1 , and Ele 2 are the estimated
marginal e¤ects of these indicators when the value of each control variable is …xed at its sample
mean. All regressions control for state and year …xed e¤ects, lagged highway expenditures, log state
population and its square, real per capita income and its square, the fraction of population aged 5-17
and the fraction aged 65 and over, Dem. control, Rep. control, the governor’s party, urbanization
and citizen ideology. Robust standard errors clustered by state in parentheses. ***, **, * denote
signi…cance at 1, 5, and 10 percent, respectively.
of reforms emphasized accountability to managers and bureaucratic responsiveness. This
led to a weakening of civil service protections. While at the state level the second wave of
reforms did not start until a 1996 reform in Georgia, policy changes at the federal level came
earlier, with the 1978 Civil Service Reform Act. Thus it may be that the operation of state
bureaucracies in the latter part of the sample (and in particular the 1989 introduction of the
merit system in West Virginia) is less comparable than in earlier years.
As a robustness check, we repeat our regressions shortening the sample period by a third,
to 1960-1983. Column (2) of Table 4 shows that the …ndings for this period are very similar
to those obtained earlier.
17
6.2.3
Controlling for heterogeneity in the political budget cycle
We check whether the di¤erent cycles observed in patronage versus civil service states could
be due to some (observed) factor correlated with civil service reform. To do this, we control
for heterogeneity of the political budget cycles across all control variables in X by interacting
each of these variables with the election cycle indicators. The results are in Column (3) of
Table 4. Reported coe¢ cients on the election cycle indicators are the estimated marginal
e¤ects of these indicators when the value of each control variable is …xed at its sample mean.
As can be seen, the results are similar to those reported earlier. Highway spending still
increases in election and pre-election years under patronage but not under civil service. The
estimated impact of Merit on the political budget cycle reported earlier was not an artifact
of failing to account for heterogeneity across other observable characteristics of the states.
6.2.4
Di¤erence GMM estimates for dynamic panel data
As discussed in section 5, our panel is long enough that any bias in the …xed e¤ects estimates
due to the presence of the lagged independent variable is likely to be minimal. By contrast,
in a panel this long the Arellano-Bond type methods designed to address the bias can be
problematic due to the proliferation of instruments.
Nevertheless, to check the robustness of our …ndings, we performed various versions of
the di¤erence GMM estimation. In each case, one …rst di¤erences equation (1) to eliminate
the state …xed e¤ects:
yst =
3
X
(
Elest +
Elest
Civilst ) +
civilst
(2)
=0
+
ys;t
1
+
X 0st +
t
+
"st :
Then, observing that ys;t 1 and "st are necessarily correlated, higher lags of yst are used
as instruments for ys;t 1 .
Our results from this exercise are in Table 5. Column (1) follows the Anderson and Hsiao
(1982) approach and uses lags t 2 and t 3 as “standard” instruments. The coe¢ cient
estimates on the variables of interest are similar to those obtained above, but the overidenti…cation test fails, indicating that some of the instruments may not be exogenous. A test
of serial correlation indicates that the presence of 2nd order autocorrelation is just rejected
at 10%, suggesting that using the t 2 lag as an instrument may not be appropriate. The
next column uses lags t 3 and t 4 as instruments, resulting in better model performance
(and similar coe¢ cient estimates). In column (3) we use the same lags but treat each year
as a separate equation following the Arellano and Bond (1991) method. As can be seen, our
18
results are very similar. Finally, in column (4) we include further lags (up to t 10) and
again …nd that the coe¢ cients of interest change very little.
In sum, our …ndings reported above appear robust to the use of di¤erent estimation
methods.
19
20
a
With a large number of instruments Hansen’s J-statistic is too weak to be meaningful (p = 1.00).
Notes: The dependent variable is real per capita highway expenditures. GMM estimates on …rst di¤erences using lags of the dependent
variable as Anderson-Hsiao or Arellano-Bond style instruments. Reported model diagnostics are the p-values from Hansen’s J-statistic and
from Arellano-Bond’s autocorrelation tests. Columns (3-4) and the autocorrelation tests use xtabond2 in Stata (Roodman, 2009b). Regressions
control for state and year …xed e¤ects, lagged highway expenditures, log state population and its square, real per capita income and its square,
the fraction of population aged 5-17 and the fraction aged 65 and over, Dem. control, Rep. control, the governor’s party, urbanization and
citizen ideology. Robust standard errors in parentheses. ***, **, * denote signi…cance at 1, 5, and 10 percent, respectively.
Anderson-Hsiao Anderson-Hsiao Arellano-Bond Arellano-Bond
(1)
(2)
(3)
(3)
Ele0
37.96**
34.38***
34.73***
38.18***
(16.90)
(12.81)
(11.48)
(10.25)
1
Ele
31.24
22.37*
27.09**
29.23**
(21.15)
(13.08)
(12.96)
(12.72)
2
Ele
7.45
4.89
6.23
8.00
(16.26)
(10.76)
(13.32)
(14.05)
Ele0 Merit
-34.43**
-32.17**
-32.45***
-35.56***
(16.27)
(12.71)
(11.55)
(10.07)
Ele 1 Merit
-35.17*
-26.13**
-31.26**
-32.78**
(20.15)
(12.97)
(13.13)
(12.78)
2
Ele
Merit
-15.15
-11.20
-14.02
-15.48
(16.85)
(11.26)
(12.98)
(13.69)
Merit
27.82
26.56
22.07
10.72
(30.27)
(24.68)
(22.36)
(18.60)
Lags used as instruments
2, 3
3, 4
3, 4
3-10
a
a
Overidenti…cation test p-value
0.08
0.43
Test of 1st order serial correlation
0.21
0.03
0.00
0.00
Test of 2nd order serial correlation
0.10
0.06
0.04
0.05
Test of 3rd order serial correlation
0.79
0.89
0.94
0.85
Test of 4th order serial correlation
0.74
0.75
0.73
Number of instruments
52
52
115
280
N
1307
1275
1339
1339
Estimation method
Table 5: Political cycles in highway spending and the merit system, GMM estimates
6.3
6.3.1
Further results
Department-speci…c vs. statewide merit system
How do patterns of highway spending vary under a merit system speci…c to the highway
department vs. a merit system with statewide coverage? We can explore this question
because we have periods in our data with department-speci…c merit system but no statewide
civil service (see Table 1). However, since we only have a limited number of these periods (38
state-year observations in four states18 ), these results should be taken merely as suggestive.
Figure 3 shows the results of estimating equation (1) with both the statewide and the
department-speci…c merit variables (and their respective interactions with the election cycle).19 Panel A is for patronage, panel B for a department-speci…c merit system only, and
panel C for statewide civil service. Panels A and C are similar to Figure 2 and show the
dampening of the budget cycle under a statewide merit system compared to patronage. Interestingly, the pattern in Panel B is somewhere between the two, with no pre-election year
increase in spending, but still a spike in spending in election years. This may suggest that
a department-speci…c merit system without a statewide merit system dampens the political
budget cycle somewhat, but still leaves opportunities for increased spending, especially closer
to the election.20 These interpretations are subject to the caveat above regarding the small
number of observations in our data that are used to identify the patterns in Panel B.
18
In Table 1 although a …fth state, Arizona, had a department-speci…c merit system before statewide civil
service was introduced, its governors were serving 2-year terms and is therefore excluded from the sample
for that period.
19
The parameter estimates are given in the Appendix.
20
It is possible that politically motivated spending in the pre-election year di¤ers from election year
spending: for example, construction projects may have to be initiated sooner than maintenance work in
order to yield electoral bene…ts. Figure 3 may suggest that, on its own, a department-speci…c merit system
may constrain the former more than a latter. Exploring this further would be an interesting topic for future
research.
21
22
Notes: The …gure shows the political budget cycles in highway expenditures (in real 2009 dollars per capita) implied by estimates of Equation (1)
that include both the statewide and the department-speci…c merit variables, and their respective interactions with the electoral cycle indicators. In
each panel the base category is the post-election year, normalized to 0. The 95 percent con…dence interval is shown by the grey lines.
Figure 3: Political cycles in highway spending under statewide or highway department-speci…c civil service, or patronage
6.3.2
Toll vs. regular highways
At the most general level, political budget cycles refer to any changes in …scal categories
correlated with the electoral cycle. Budget cycles can arise from politicians’ and voters’
focus on expenditures or revenues (or both). For example, if cycles re‡ect politicians’desire
to please voters and voters are “…scal conservatives”(Peltzman, 1992), politicians’incentive
may be to increase revenues and lower de…cits before elections.
Our results above provide evidence of a focus on a particular type of expenditure in
the context of US state politics. However, this interpretation may need to be quali…ed due
to the presence of toll highways. Toll highways create revenue for state governments, and
spending on these highways could in principle be motivated by a desire to increase revenues.
If the political budget cycle was driven by spending on toll highways, this could indicate that
incumbent politicians are in fact motivated by revenues rather than expenditures. Because
the Census of Governments reports spending on toll and non-toll highways separately, we
can check for this by estimating separate regressions for the two categories. The results are
in Table 6, and they clearly indicate that the budget cycles under patronage arise in non-toll
highway expenditures. Spending on toll highways shows no cycles under either patronage or
civil service, indicating that a desire to increase revenues is unlikely to drive the cycle.
23
Table 6: Political cycles in toll vs. non-toll highway spending and the merit system
Dep. Var.:
Toll highways
(1)
Ele0
6.64
(7.60)
1
Ele
6.17
(5.27)
Ele 2
2.53
(2.24)
Ele0 Merit
-4.71
(7.69)
Ele 1 Merit
-5.65
(5.16)
2
Ele
Merit
-2.09
(2.14)
Merit
3.86
(2.70)
R2
0.39
N
1,387
Non-toll highways
(2)
31.55***
(7.16)
23.10**
(11.16)
4.27
(13.96)
-31.16***
(8.06)
-26.87**
(12.07)
-12.83
(13.94)
13.46*
(7.45)
0.68
1,387
Notes: The dependent variable is real per capita highway expenditures on toll and nontoll highways, respectively. All regressions control for state and year …xed e¤ects, the lagged
dependent variable, log state population and its square, real per capita income and its square,
the fraction of population aged 5-17 and the fraction aged 65 and over, Dem. control, Rep.
control, the governor’s party, urbanization and citizen ideology. Robust standard errors
clustered by state in parentheses. ***, **, * denote signi…cance at 1, 5, and 10 percent,
respectively.
6.3.3
Non-highway spending
Highway spending is a natural category in which to look for political budget cycles due
to its well-documented political importance. Are the political budget cycles in highway
expenditures under patronage large enough to show up in total government spending? Do
cycles appear in other spending categories as well? Table 7 provides some answers to these
questions. Column (1) indicates the presence of cycles in total expenditures under patronage
(but not under civil service). Column (2) further shows that this is entirely driven by highway
spending: restricting attention to non-highway expenditures shows no cycles.
24
Table 7: Total expenditures and non-highway expenditures
Dep. Var.:
Total expenditures
(1)
Ele0
27.35*
(14.95)
1
Ele
44.86**
(19.55)
2
Ele
2.91
(21.23)
Ele0 Merit
-24.25
(16.51)
Ele 1 Merit
-60.17**
(24.96)
2
Ele
Merit
-26.70
(23.41)
Merit
76.68***
(27.84)
2
R
0.98
N
1,387
Non-highway expenditures
(2)
-12.01
(12.24)
11.54
(12.14)
-6.03
(14.66)
11.92
(14.25)
-24.47
(16.79)
-9.70
(14.68)
50.09**
(21.17)
0.98
1,387
Notes: The dependent variable is real per capita total direct expenditures in column (1) and
real per capita non-highway direct expenditures in column (2). Regressions control for state
and year …xed e¤ects, lagged highway expenditures, log state population and its square, real
per capita income and its square, the fraction of population aged 5-17 and the fraction aged 65
and over, Dem. control, Rep. control, the governor’s party, urbanization and citizen ideology.
Robust standard errors clustered by state in parentheses. ***, **, * denote signi…cance at 1,
5, and 10 percent, respectively.
7
Conclusion
Bureaucratic institutions matter for policy and the behavior of politicians. In this paper we
found that civil service protections can stabilize government activity over time by dampening
the political budget cycle. In particular, we found signi…cant budget cycles in the highway
expenditures of US state governments under patronage but no cycles under civil service.
These …ndings may suggest a possible explanation for some of the cross-country di¤erences observed in previous studies: political budget cycles may be more prevalent in political
systems characterized by patronage but less likely to occur under civil service. While the
potential of civil service to stabilize the bureaucracy has long been recognized, our results
suggest that this institution may also have a “multiplier” e¤ect by stabilizing the policies
chosen by election-minded politicians.
25
References
[1] Akhmedov, A., and E. Zhuravskaya (2004): “Opportunistic Political Cycles: Test in a
Young Democracy Setting,”Quarterly Journal of Economics 119(4), 1301-1338.
[2] Alesina, A., and G. Tabellini (2007): “Bureaucrats or Politicians? Part I: A single
policy task,”American Economic Review 97(1), 169-179.
[3] Alesina, A., and G. Tabellini (2008): “Bureaucrats or politicians? Part II: Multiple
policy tasks,”Journal of Public Economics 92, 426–447.
[4] Alt, J. E., and D.D. Lassen (2006): “Transparency, Political Polarization, and Political
Budget Cycles in OECD Countries,”American Journal of Political Science 50(3), 530550.
[5] Anderson, T. W., and C. Hsiao (1982): “Formulation and estimation of dynamic models
using panel data,”Journal of Econometrics 18, 47–82.
[6] Arellano, M., and S. Bond (1991): “Some Tests of Speci…cation for Panel Data: Monte
Carlo Evidence and an Application to Employment Equations,” Review of Economic
Studies 58(2), 277-297.
[7] Berry, W. D., E. J. Ringquist, R. C. Fording, and R. L. Hanson (1998): “Measuring
Citizen and Government Ideology in the American States, 1960-93,”American Journal
of Political Science 42, 327-48.
[8] Besley, T., and A. Case (2003): “Political institutions and policy choices: Evidence
from the United States,”Journal of Economic Literature 41, 7-73.
[9] Brender, A., and A. Drazen (2005): “Political Budget Cycles in New versus Established
Democracies,”Journal of Monetary Economics, 52(7), 1271-1295.
[10] Congressional Budget O¢ ce (1978): Highway Assistance Programs: A Historical Perspective, Washington, DC: Congressional Budget O¢ ce.
[11] Dilger, R.J. (1989): National intergovernmental programs, Englewood Cli¤s, NJ: Prentice Hall.
[12] Drazen, A. (2001): “The Political Business Cycle after 25 Years,”In B.S. Bernanke and
K.Rogo¤ (Eds.): NBER Macroeconomics Annual 2000, 75-117, Cambridge, MA: MIT
Press.
26
[13] Drazen, A., and M. Eslava (2010): “Electoral Manipulation via Voter-Friendly Spending: Theory and Evidence,”Journal of Development Economics 92(1), 39-52.
[14] Epstein, D., and S. O’Halloran (1999): Delegating Powers. A Transaction Cost Politics Approach to Policy Making under Separate Powers, Cambridge, UK: Cambridge
University Press.
[15] Eslava, M. (2011): “The Political Economy of Fiscal De…cits: A Survey,” Journal of
Economic Surveys 25(4), 645-673.
[16] Folke, O., S. Hirano, and J. M. Snyder, Jr. (2011): “Patronage and Elections in US
States,”American Political Science Review 105(3), 567-585.
[17] Fox, J., and S.V. Jordan (2011): “Delegation and Accountability,” Journal of Politics
73(3), 831–844.
[18] Gailmard, S., and J.W. Patty (2007): “Slackers and Zealots: Civil Service, Policy
Discretion, and Bureaucratic Expertise,” American Journal of Political Science 51(4),
873-889.
[19] Holtz-Eakin, D., W. Newey, and H. S. Rosen (1988): “Estimating vector autoregressions
with panel data,”Econometrica 56, 1371–1395.
[20] Ingraham, P.W. (1995): The Foundation of Merit: Public Service in American Democracy, Baltimore, MD: Johns Hopkins University Press.
[21] Judson, R. A., and A. L. Owen (1999): “Estimating Dynamic Panel Data Models: A
Guide for Macroeconomists,”Economics Letters 65(1): 9-15.
[22] Khemani, S. (2004): “Political Cycles in a Developing Economy: E¤ect of Elections in
the Indian States,”Journal of Development Economics 73(1): 125-154.
[23] Knight, B. (2002): “Endogenous federal grants and crowd-out of state government
spending: theory and evidence from the federal highway aid program,”American Economic Review 92(1), 71–92.
[24] Krause, G. A., D. E. Lewis, and J. W. Douglas (2006): “Political appointments, civil
service systems, and bureaucratic competence: Organizational balancing and executive
branch revenue forecasts in the American states,”American Journal of Political Science
50(3), 770-787.
27
[25] Maskin, E., and J. Tirole (2004): “The Politician and the Judge: Accountability in
Government,”American Economic Review 94(4), 1034-1054.
[26] Mosher, F. (1982): Democracy and the public service, 2nd ed., New York, NY: Oxford
University Press.
[27] National Research Council (1952): Merit System Provisions in State Highway Employment, Washington, DC: Highway Research Board, NRC.
[28] Peltzman, S. (1992): “Voters as Fiscal Conservatives,”Quarterly Journal of Economics
107(2), 327-361.
[29] Persson, T., and G. Tabellini (2003): The Economic E¤ects of Constitutions, Cambridge, MA: MIT Press.
[30] Rauch, J. E. (1995): “Bureaucracy, infrastructure, and economic growth: Evidence from
US cities during the Progressive era,”American Economic Review 85(4), 968-979.
[31] Rauch, J.E., and P.B. Evans (2000): “Bureaucratic structure and bureaucratic performance in less developed countries,”Journal of Public Economics 75, 49-71.
[32] Rogo¤, K. (1990): “Equilibrium political budget cycles,” American Economic Review
80, 21–36.
[33] Roodman, D. (2009a): “A Note on the Theme of Too Many Instruments,” Oxford
Bulletin of Economics and Statistics 71(1): 135-158.
[34] Roodman, D. (2009b): “How to do xtabond2: An introduction to di¤erence and system
GMM in Stata,”The Stata Journal 9(1) 86–136.
[35] Rose, S. (2006): “Do Fiscal Rules Dampen the Political Business Cycle?”Public Choice
128(3-4), 407-431.
[36] Ruhil, A.V.S., and P.J. Camoes (2003): “What Lies Beneath: The Political Roots of
State Merit Systems,” Journal of Public Administration Research and Theory 13(1),
27-42.
[37] Shi, M., and J. Svensson (2006): “Political Budget Cycles: Do They Di¤er across
Countries and Why?”Journal of Public Economics 90(8-9), 1367-1389.
[38] Ting, M.M. (2002): “A Theory of Jurisdictional Assignments in Bureaucracies,”American Journal of Political Science 46(2), 364-378.
28
[39] Ting, M.M. (2012): “Legislatures, Bureaucracies, and Distributive Spending,” The
American Political Science Review 106(2), 367-385.
[40] Tolchin, M., and S. Tolchin (1971): To the Victor... Political Patronage from the Clubhouse to the White House, New York, NY: Random House.
[41] Ujhelyi, G. (2014a): “Civil Service Reform,”Journal of Public Economics 118, 15-25.
[42] Ujhelyi, G. (2014b): “Civil Service Rules and Policy Choices: Evidence from US State
Governments,”American Economic Journal: Economic Policy 6(2), 338-380.
29
A
A.1
Appendix
Additional tables
Table 8: Political cycles under statewide or department-speci…c merit systems
Ele0
Ele
1
Ele
2
Ele0
40.33***
(12.93)
39.42**
(15.90)
14.06
(18.97)
-32.57***
(7.62)
-13.49
(11.44)
0.82
(7.27)
17.92*
(9.84)
-5.40
(15.51)
-29.53
(19.67)
-23.30
(19.90)
1.84
(14.18)
0.69
1,387
Merit
Ele
1
Merit
Ele
2
Merit
Merit
Ele0
HwyMerit
Ele
1
HwyMerit
Ele
2
HwyMerit
HwyMerit
R2
N
Notes: The dependent variable is real per capita highway expenditures.
HwyMerit = 1 if a department-speci…c merit system is in place. The regression controls for state and year …xed e¤ects, lagged highway expenditures,
log state population and its square, real per capita income and its square,
the fraction of population aged 5-17 and the fraction aged 65 and over, Dem.
control, Rep. control, the governor’s party, urbanization and citizen ideology.
Robust standard errors clustered by state in parentheses. ***, **, * denote
signi…cance at 1, 5, and 10 percent, respectively.
A.2
Data sources
Statewide merit systems
See Ujhelyi (2014b) and the sources reported there.
30
Highway department merit systems
Arizona. “Arizona Highway Changes Proposed,” Prescott Evening Courier, Jan 19, 1955,
p4. “State Highway Examination Dates Are Set,”Prescott Evening Courier, Jul 9, 1957, p8.
Arkansas. “Suggestion To Revamp Road System Heard,” Northwest Arkansas Times, Feb
16, 1952, p1. “Legislature,”The Courier News, Feb 28, 1957, p8.
Delaware. “Stalling on Merit System, du Pont Says of Democrats,”The News Journal, Oct
11, 1961, p10.
Florida. Whitney, J.C. (2008): Florida expressways and the public works career of Congressman William C. Cramer, Graduate Theses and Dissertations, http://scholarcommons.usf.edu/etd/563.
Idaho. Fifty Years of Professional Engineering in Idaho, 1960, Boise, ID: Idaho Society of
Professional Engineers. 100th Anniversary, 2010, Boise, ID: Idaho Society of Professional
Engineers.
Iowa. “Murray: Want Possible Ho¤a-Roads Tieup?”Ames Daily Tribune, Oct 21, 1958, p1.
“Ending Road Works Spoils,”The Des Moines Register, Sep 22, 1967, p4.
Kentucky. “Combs Plans Merit System,”Kentucky New Era, Apr 10, 1959, p17.
Mississippi. “Highway Problems Aired,” The Delta Democrat-Times, Jul 13, 1973, p16.
“Expanded Merit System Passed,”Clarion-Ledger, Feb 11, 1977, 12C.
Montana. “Commission Will Speed Up State Highway Program,” The Independent Record,
Dec 18, 1962, p1.
New Mexico. “State Highway Department Has A Record Turnover,” Albuquerque Journal,
Nov 18, 1951, p6.
Oklahoma. Odell, W.H. (1950): The patronage system in Oklahoma. Norman, OK: Transcript Co.
Pennsylvania. Martin, J.W. (1959): “Administrative Dangers in the Enlarged Highway
Program,”Public Administration Review 19(3), 164-172.
Texas. Smith, Gri¢ n Jr. (1974): “The Highway Establishment and How it Grew and Grew
and Grew,” Texas Monthly 2(4), April, p76-93. http://www.texasmonthly.com/issue/april1974. Morehead, R. (1984): Dewitt C. Greer, King of the Highway Builders, Austin, TX:
Eakin Press.
Utah. “Merit System Triggers Organizational Dispute,” The Deseret News, Dec 24, 1962,
pB-1.
South Carolina. “Thurmond Hails Road Department Merit System,” The Index-Journal,
Sep 23, 1950, p5. NRC (1952).
Washington. A History of Personnel Systems for Washington State, 1989, Olympia, WA:
Washington State Department of Personnel.
West Virginia. “Wright Asks All Highway Workers in U.S. Put Under Merit System,” The
31
Raleigh Register, Jul 25, 1962, p2. “Jay Names Civil Service Commission,” The Raleigh
Register, Jun 28, 1977, p6.
Consumer Price Index
U.S. Department of Labor: Bureau of Labor Statistics, http://www.bls.gov. Consumer Price
Index for All Urban Consumers, not seasonally adjusted. Annual value obtained by averaging
across months. 2009 = 100.
State expenditures
US Census Bureau, State Government Finances Publication Historical Data Base, state
government variables. Direct Expenditure, Regular Hwy-Direct Exp, Total Hwy-Direct Exp,
Total Hwy-Total Exp, Toll Hwy-Total Exp.
Income and population
Bureau of Economic Analysis: Regional Economic Accounts, http://www.bea.gov/regional/spi/.
State Annual personal income. Population …gures reported in this source are midyear estimates of the Census Bureau.
Aged and kids
US Census Bureau. The post-1970 data was compiled by List, J.A., and D.M. Sturm (2006):
“How Elections Matter: Theory and Evidence from Environmental Policy,”Quarterly Journal of Economics 121(4), 1249-1281. The pre-1970 was entered from Population Projection
(P25) Reports. Year 1969 linearly interpolated.
Percent urban
US Census Bureau. Urban and Rural Population 1900-1990, released 1995, available at
http://www.census.gov/population/censusdata/urpop0090.txt. Years between censuses were
linearly interpolated.
Party control and governor’s party
Burnham, W. Dean, “Partisan Division of American State Governments, 1834-1985,” Conducted by Massachusetts Institute of Technology, ICPSR ed. Ann Arbor, MI: Interuniversity
Consortium for Political and Social Research [producer and distributor], 1986. All variables
merged so that they re‡ect party composition for the given year (for election years, party
composition re‡ects the pre-election situation). Before 1975, this requires shifting the variables forward by 1 year. Governor’s party: corrections as listed in Ujhelyi (2014b).
Years 1985-1996 from Council of State Governments: Book of the States, various volumes.
Citizen ideology
32
Berry et al. (1998). This index uses ideological ratings of congressional candidates by the
Americans for Democratic Action and the AFL/CIO’s Committee on Political Education
and their vote shares to estimate the ideological composition of electoral districts; these are
then aggregated to form a statewide measure of citizens’ideology (degree of liberalism, on
a scale 0-100).·
33