Maximizing Efficiency in State Infrastructure Finance: The Role of Competition,Citizen Monitoring Capacity,and Institutions

Abstract

In the United States, despite record levels of public infrastructure spending, evidence on rising traffic congestion and deteriorating infrastructure condition raises questions about the efficiency of government infrastructure spending. This research aims at empirically assessing and explaining the relative efficiency in producing public highway infrastructure outcomes among American states. To achieve this purpose, a semiparametric analysis—the two-stage double bootstrap data envelopment analysis method—is applied to examine how highway infrastructure efficiency scores can be estimated and explained by a number of exogenous variables among 47 American states from 1995 to 2009. This study finds that there is a large efficiency variation among states in terms of producing quality highway infrastructure services. Furthermore, interstate competition, jurisdiction size, fiscal capacity, and political and fiscal institutions are the key factors influencing the efficiency performance of state highway infrastructure systems.

Keywords

efficiency infrastructure finance highways states

Introduction

The search for efficiency improvements in the public sector is a topic of great interest in the academic field of public administration and among practical professionals (e.g., Andrews & Entwistle, 2014; Lindlbauer, Winter, & Schreyögg, 2016; Pérez-López, Prior, & Zafra-Gómez, 2015; Ruggiero, Duncombe, & Miner, 1995). The movement of New Public Management swept around the world in the 1990s with the slogan of creating a government that “works better and costs less.” The recent Great Recession of 2008-2010 had a significant adverse impact on government revenues and has drawn renewed attention to improving government efficiency and reducing the costs of public service delivery (Pérez-López et al., 2015).

To date, scholars have shown interest in estimating both the efficiency of specific services of the public sector—public libraries (e.g., Hemmeter, 2006), public schools (e.g., Ruggiero et al., 1995), public hospitals (e.g., Lindlbauer et al., 2016), public utilities (e.g., Haug, 2008)—and the overall efficiency of the public sector in providing multiple public services among municipalities (e.g., Borge, Falch, & Tovmo, 2008; Geys, Heinemann, & Kalb, 2010), states (e.g., Pang, Tafti, & Krishnan, 2014), and countries (e.g., Afonso, Schuknecht, & Tanzi, 2005, 2010). Although research measuring efficiency and modeling bureaucratic behavior has burgeoned in recent years, an empirical analysis of the drivers of public infrastructure efficiency has received less attention.

This study attempts to further the research on public service efficiency by exploring the determinants of efficiency in one of the understudied areas of the public sector—the U.S. state highway infrastructure. There are three reasons for studying efficiency in state highway infrastructure. First, highways are one of the key public services provided by U.S. state governments. All fifty state governments play a crucial role in managing and financing state highway infrastructure services. Second, state highway infrastructure systems are the largest component of state capital assets (Bartle & Chen, 2014; Chen, 2016). In 2015, total state spending on highways was over US$120 billion. Third, the highway transportation infrastructure sector is notorious for low efficiency, spending waste, and corruption (e.g., Liu & Mikesell, 2014; Winston, 2013). Despite record levels of public infrastructure investment, evidence on rising traffic congestion and deteriorating physical condition raises questions about the efficiency of government transportation infrastructure spending (Chen, 2016; Winston, 2013). The latest 2017 Report Card for America’s Infrastructure continued to give the nation’s critical infrastructure an overall poor grade of D+ (American Society of Civil Engineers [ASCE], 2017). According to ASCE (2017), one out of every five miles of highway pavement is in poor condition, one in nine of the nation’s bridges is rated as structurally deficient, and more than two out of every five miles of America’s urban interstates are congested. Given the significant amount of government infrastructure investment and the widely expressed concern about the declining quality of the American public infrastructure systems, research about the efficiency of infrastructure finance is especially timely and important.

To fill the research gap, this study aims at empirically assessing and explaining the relative efficiency performance in producing highway infrastructure outcomes among American states. We draw on theories of public sector efficiency from the public choice literature and empirically test the link between theories and measures of state highway infrastructure efficiency. A semiparametric analysis—the two-stage double bootstrap data envelopment analysis (DEA) method—is applied to examine how highway infrastructure efficiency scores can be estimated and explained by a number of exogenous variables among 47 American states from 1995 to 2009. The results reveal that there is a large efficiency variation among states in terms of producing quality highway infrastructure services. Furthermore, interstate competition, jurisdiction size, fiscal capacity, and political and fiscal institutions are the key factors influencing the efficiency performance of state highway infrastructure.

The article proceeds as follows. The next section reviews existing theories of public sector efficiency and develops testable hypotheses. The subsequent section presents a detailed explanation of data and methodology, followed by the discussion of empirical findings. The final section concludes with a summary of findings and discussion of policy implications.

Theoretical Background and Hypothesis Development

Little research on infrastructure efficiency has been done. Public choice theory can be used to guide such inquires. In the following section, we draw on major theories of public sector efficiency from the public choice literature (Breton & Wintrobe, 1975; Leibenstein, 1966, 1978; Migué, Belanger, & Niskanen, 1974; Moene, 1986; Niskanen, 1971, 1975; Ostrom, 1972; Wyckoff, 1990) and derive testable hypotheses potentially related to the degree of efficiency in state infrastructure finance.

Competition

The introduction of competition in the provision of public services offers incentives for improving government efficiency by influencing “citizen’s willingness to pay for public services or their willingness to stay in the jurisdiction” (Grosskopf, Hayes, Taylor, & Weber, 2001, p. 454). Furthermore, competition increases external pressure on public sector managers and pushes them to be more efficient in providing public services (e.g., Borge et al., 2008; Ruggiero et al., 1995). Empirically, at the local level, Grossman, Mavros, and Wassmer (1999) assert that competition within the central city’s metropolitan area (measured by the number of cities in a central city’s metropolitan statistical area) significantly contributes to better public sector efficiency. Grosskopf et al. (2001) find that public school competition (measured by student enrollments in public school districts) exerts a significant and positive effect on education efficiency.

At the state level, there is growing empirical evidence about interstate competition in fiscal policy to attract new industries and firms (e.g., Bruce, Carroll, Deskins, & Rork, 2007; Case, Rosen, & Hines, 1993; Kenyon, 1991). In an early study, Gray (1976) explores the dynamics policy process of state politics and contends that the model of competitive threats explains the variations in both state welfare and education expenditures. Public highway infrastructure systems have a great visibility and often play a critical role in supporting state and local economic development and business growth (Macmanus, 2004). Therefore, infrastructure becomes a key area of state and local fiscal competition. Empirically, Case et al. (1993), Bruce et al. (2007), and Witko and Newmark (2009) present the evidence that states engage in highway expenditure competition with other states (race to the top) to attract new economic activities. Despite the paucity of empirical studies on state-level competition and public sector efficiency, there is some general literature discussing the potential efficiency-enhancing effects of interstate competition. For instance, Kenyon (1991) argues that competition at the state and local levels promotes responsiveness of governments to their citizens and improves the efficiency of public service provision. O’Connell and Yusuf (2011) show that residents use neighboring states to provide anchor values. This offers an example of how residents use information about other states to pressure their state to be efficient or comparable. Based on the previous discussion, the study makes the first hypothesis:

Hypothesis 1: States facing stronger interstate competition in the provision of highways are positively associated with highway infrastructure efficiency.

Size of Jurisdiction

Niskanen’s model of bureaucratic behavior suggests that the larger the sizes of jurisdictions, the less efficient the governments. This is because the greater monopolistic power associated with a larger jurisdiction decreases the incentives of both voters and legislators to monitor governments (Niskanen, 1971, 1975). Similarly, Ostrom (1972) argues that smaller organizations are easier to monitor and more responsive to their voters. She further asserts that an increasing government size is often associated with “decreased responsibility of local officials and decreased participation by citizens” (Ostrom, 1972, p. 487) and therefore affects public sector efficiency negatively. However, empirical evidence is mixed. Empirically, it has been found that larger jurisdiction sizes are positively associated with local police department inefficiency (e.g., Hayes, Razzolini, & Ross, 1998) and state government inefficiency (Pang et al., 2014). In contrast, De Borger, Kerstens, Moesen, and Vanneste (1994) find that Belgian municipalities with larger population sizes operate more efficiently due to the presence of economies of scale. In view of these considerations, it is not possible to establish a priori the sign of the relationship between jurisdiction size and public sector efficiency. Therefore, the hypothesis is as follows:

Hypothesis 2: The sizes of state-administered highway systems are related to state highway infrastructure efficiency.

Fiscal Capacity

The relationship between a jurisdiction’s fiscal capacity and its efficiency in achieving better highway infrastructure outcomes with given resources is complex. Note that our efficiency measure essentially captures state highway technical efficiency, that is, whether to use given resources to increase outcomes at an optimum. On one hand, states with greater fiscal capacity might have more capable administrators and employees and thus would be able to manage state highway infrastructure systems more efficiently. From this perspective, stronger fiscal capacity indicates better highway infrastructure efficiency. However, on the other hand, stronger fiscal capacity may increase the on-the-job leisure of politicians and public managers (De Borger et al., 1994), foster featherbedding of politicians and public managers (Wyckoff, 1990), and allow governments to provide public services in an inefficient form because public officials have fewer incentives to provide services as efficiently as possible (Balaguer-Coll & Prior, 2009; Borge et al., 2008; Duncombe, Miner, & Ruggiero, 1997). Empirically, the majority of previous research suggests that jurisdictions with greater fiscal capacity tend to be less efficient in the provision of public services (Duncombe & Yinger, 1997; Eom & Rubenstein, 2006; Grosskopf et al., 2001; Pang et al., 2014). In sum, whether state fiscal capacity has a positive or negative impact on state highway efficiency is an empirical question. Based on the above discussions, it is not possible to establish a priori the sign of the relationship between fiscal capacity and public sector efficiency. Therefore, the next hypothesis is expressed as follows:

Hypothesis 3: Fiscal capacity is related to state highway infrastructure efficiency.

Political Environment

Public choice theorists have discussed the effects of various political environments on public sector efficiency (e.g., Breton & Wintrobe, 1975; Moene, 1986). Concerning political ideology, the political economy literature suggests that the socialist parties prefer a larger public sector than nonsocialists. Moreover, they have strong ties to public sector unions. In general, public sector unions, as special interest groups, tend to support the increase of wages and generally oppose performance incentive schemes (Borge et al., 2008; Falch, 2001). This may result in lower efficiency in the provision of public services. Empirically, Borge et al. (2008) and Kalb (2010) find that municipalities governed by left-wing parties have lower efficiency. In the United States, Democratic members usually prefer a larger government size and have strong ties to public sector unions. Therefore, the larger the influence of democratic ideology, the less efficient state highway infrastructure. According to these discussions, the study offers the following hypotheses:

Hypothesis 4a: States having larger ratios of Democratic members in the State House are negatively associated with highway infrastructure efficiency.

Hypothesis 4b: States having larger ratios of Democratic members in the State Senate are negatively associated with highway infrastructure efficiency.

Hypothesis 4c: States having Democratic governors are negatively associated with highway infrastructure efficiency.

Hypothesis 4d: States having a larger share of government employee union memberships are negatively associated with highway infrastructure efficiency.

Political leadership and strength affect public sector efficiency. Theoretically, strong political leadership may be more likely to resist pressure to accommodate low efficiency with inflated budgets (Geys et al., 2010; Kalb, 2010) and have more power in “internal bargains with public sector unions about the implementation of incentive schemes and other means of increasing performance” (Borge et al., 2008, p. 483). Empirically, majority governments are found to achieve higher levels of efficiency (Borge et al., 2008). In the United States, states with divided governments usually adjust slowly to exogenous environmental changes and are subject to weak political leadership (Alt & Lowry, 1994; Poterba, 1994). So, the next hypothesis is specified:

Hypothesis 4e: States with divided governments are negatively associated with highway infrastructure efficiency.

Fiscal Institutions

In the Leviathan model, the public choice theorists argue that government officials are self-interested actors who maximizing discretionary budget slack for private, rather than public, gains (Niskanen, 1971, 1975). Fiscal institutions serve as ex-ante rules that limit the policy choices of government officials, and they can bind leviathan because they prescribe what politicians can and can not do (e.g., Chan & Mestelman, 1988; Moene, 1986). A hard budget constraint will lower budgetary slack because the agency faces a fixed budget and cannot increase its budget by reducing inefficiency (Borge et al., 2008). In the United States, tax and expenditure limitations (TELs) on state and local governments have been passed with the presumption that they will limit the growth of government and raise government efficiency (e.g., Stallmann, 2007). Similarly, most state governments have constitutional or statutory requirements to balance the budgets, known as balanced budget requirements (BBRs). Empirical research on fiscal rules shows that BBRs are an effective tool for fiscal restraint (constraining expenditure) at the state level (e.g., Smith & Hou, 2013), which may have the potential to push the efficiency improvement of government policy. In view of these considerations, the next hypothesis is proposed:

Hypothesis 5: States having stringent fiscal and budgetary institutions are positively associated with highway efficiency.

Method and Data

Model Specification

The hypotheses are tested in the following econometric model:

\begin{array}{l} H i g h w a y E f f i c i e n c y_{i t} = β_{0} + β_{1} {(C o m p e t i t i o n)}_{i t - 1} + β_{2} {(J u r i s d i c t i o n S i z e)}_{i t} + \\ β_{3} {(F i s c a l C a p a c i t y)}_{i t} + β_{4} {(P o l i t i c a l E n v i r o n m e n t)}_{i t} + β_{5} {(F i s c a l I n s t i t u t i o n s)}_{i t} \\ + β_{6} {(C o n t r o l s)}_{i t} + γ θ_{i} + ψ ω_{t} + ε_{i t} . \end{array}

where the dependent variable of HighwayEfficiency_it is the state highway infrastructure efficiency score in state i at time t. Competition_it is the variable of measuring interstate highway competition. JurisdictionSize_it is a variable measuring the size of state-administered highway system. FiscalCapacity_it is a variable to capture state fiscal capacity. PoliticalEnvironment_it is a vector of state political variables. FiscalInstitutions_it is a set of state fiscal institutional variables. Controls_it is a series of control variables capturing state variations in residents’ characteristics, environmental conditions, and highway production costs. θ_i and ω_t are dummy variables to control for both state and year-invariant unobserved heterogeneity. ε_it means errors.

Dependent Variable—State Highway Efficiency Index

The dependent variable of interest, the efficiency score of state highway infrastructure, is not directly observed in our data. We construct this variable by choosing appropriate highway input and outcome variables to analyze highway efficiency scores in each state.

State highway inputs

Following Deller and Halstead (1994) and Kalb (2010), three monetary highway input variables are included. The first variable is real per capita state-administered highway capital spending (HwyCapitalExp). Highway capital outlay consists of expenditures associated with highway improvements, which include acquisition of right of way, preliminary and construction engineering, and new highway construction and reconstruction. Highway capital outlay aims at expanding road capacity and easing traffic congestion. The second is real per capita state-administered highway maintenance spending (HwyMaintExp). Highway maintenance outlay consists of expenditures associated with system preservation, which includes routine and regular expenditures required to keep highways and bridges in usable condition (U.S. Department of Transportation, 2010). Current highway outcomes not only rely on ongoing highway investment but also depend on the past level of highway capital stock (e.g., Winston & Langer, 2006). The third is real per capita state-administered highway capital stock. Following previous studies (Chen, 2017; Winston & Langer, 2006), this study develops a highway capital stock measure (HwyCapitalStock) by using the cumulative levels of past highway investments with depreciation.¹

State highway outcomes

We prefer to use highway outcomes rather than highway outputs. This goes beyond the simple idea of efficiency (ratios of outputs to financial inputs). Actually, we take the quality of public highway infrastructure into account.² Drawing from recent literature (Chen, Kriz, & Wang, 2016; Heckman, 2015; Neshkova & Guo, 2013), three highway outcome variables are utilized. The first variable is state road quality, which is defined as the percent of acceptable roads in state-administered highway systems (GoodRoads). Roads are in acceptable quality when their International Roughness Index (IRI) scores are less than or equal to 170 (Chen, 2014; Chen et al., 2016; Neshkova & Guo, 2013). The second is state bridge condition (GoodBridges), which refers to the percent of state-owned bridges that are not structurally or functionally deficient (Chen, 2014; Chen et al., 2016; Heckman, 2015). The third is state highway traffic congestion, which is measured by the percent of state-administered urban road miles that are congested (RoadCongestion). To derive a statewide measure, this research uses the traffic Volume/Service Flow (V/SF) ratio, which measures the actual flow of traffic relative to a theoretical maximum amount of road carrying capacity. Roads are congested when their traffic V/SF ratios exceed 0.80 (Chen, 2014; Chen et al., 2016). Table 1 gives a summary of descriptive statistics for the inputs and outcomes of state highway infrastructure.

Table 1.

Summary of State Highway Infrastructure Inputs and Outcomes.

	n	M	SD	Minimum	Maximum
Highway inputs
HwyCapitalExp	705	194.54	40.37	65.61	580.76
HwyMaintExp	705	34.99	19.05	15.01	194.54
HwyCapitalStock	705	2,295.30	560.71	709.10	7,890.50
Highway outcomes
GoodRoads%	705	82	11	38	99
GoodBridges%	705	72	9	38	90
RoadCongestion%	705	10	5	0	26

State highway efficiency index

Following the mainstream of public education efficiency studies (e.g., Duncombe & Yinger, 1997; Eom & Lee, 2014), we define highway technical efficiency as the ability to produce the maximum possible highway outcomes given a set of highway inputs. When highway inputs (e.g., labor, capital, materials) are replaced by monetary inputs (highway spending), highway technical efficiency is equal to highway spending efficiency. This means, a state is said to be inefficient if it spends more on highways than other states with the same highway outcomes and the same highway costs. The degree of inefficiency is measured by the extent of this excess highway spending. We apply the nonparametric DEA to calculate state highway infrastructure efficiency scores for every state in each year.

Key Independent Variables

The study adopts the variable of HwyCompetition to measure the competition in providing state highway infrastructure from neighboring states. This variable is defined as the average highway efficiency score in neighboring states in the previous year (1-year lag). Given the great visibility of state highway infrastructure systems and the critical role of transportation investment in supporting economic development (Bruce et al., 2007; Witko & Newmark, 2009), it is expected that neighboring states with better highway efficiency performance will push one state to be more efficient in its own provision of highway infrastructure. To measure the size of jurisdictions, the study uses the variable of the log of state-administered highway lane miles (LnHwySize). The sign of the variable of LnHwySize is ambiguous. We use the log of state real median household income (LnMedianIncome) to measure state fiscal capacity. A higher level of median household income largely determines the fiscal capacity of states (e.g., De Borger & Kerstens, 1996; De Borger et al., 1994; Kalb, 2010). Its sign is not a priori clear; however, we expect fiscal capacity would affect state highway efficiency.

There are five variables to capture state political institutions. They are the ratio of Democratic members in the State House (DemoHouse), the ratio of Democratic members in the State Senate (DemoSenate), a dummy variable if a Democratic governor (DemoGovernor), the percent of state government employees who have union memberships (PublicSectorUnion), and a dummy variable if the governor and majority of both chambers of the legislature are in different parties (DividedGov). It is expected that Democratic ideology, public sector union memberships, and divided governments negatively affect state highway efficiency. Fiscal institution variables in this study include a dummy variable of whether states enacted a constitutional or statutory TELs, a dummy variable of whether the governor must submit a balanced budget (SubmitBalancedBudget), a dummy variable of whether state legislature must pass a balanced budget (PassBalancedBudget), a dummy variable of whether the governor must sign a balanced budget (SignBalancedBudget), and a dummy variable of whether state governments may carry over year-end deficit (DeficitCarry). It is expected that the existence of these TELs and BBRs is associated with better highway infrastructure efficiency.

Control Variables

Characteristics of residents

We use the variables of the log of licensed drivers per 1,000 population (LnDriverPop) and the log of total registered vehicles per capita (LnVehicle) to capture state road users’ characteristics. Most Americans use public roads on a daily basis and have an opinion on the quality of highway transportation services (Neshkova & Guo, 2013). In addition, road users pay a significant amount of motor fuels tax and motor vehicle registration fees to fund state highway infrastructure systems. As taxpayers, they concern about the efficient provision of highway infrastructure service (Macmanus, 2004). This argument is similar to homeowners. In the United States, local property taxes are the majority revenue source to fund public schools. As property taxpayers, homeowners are generally willing to monitor school district officials in the provision of public education services. In the same vein, we expect that states with larger amounts of licensed drivers and vehicle owners are associated with better highway infrastructure efficiency due to the potential monitoring efforts from highway users. We also use the variable of the percent of state population with a bachelor’s degree or higher to control for state education level (CollegeEducation). It might be likely that college-educated citizens have more capacity and willingness to monitor state officials to efficiently provide public infrastructure (Duncombe & Yinger, 1997; Mueller, 1989).

Environmental and cost factors

The study adds a set of control variables to capture state variations in the highway production costs and environmental conditions. We use the variable of the log of state mean hourly road construction wage (LnHwyWage) to measure highway labor cost. To capture land prices for new road construction, the variable of the log of state farmland value per acre (LnLandPrice) is used. Highway production is influenced by environmental conditions including population density, urbanization, temperature, precipitation, traffic flow, and road damage of heavy trucks (Egilmez & McAvoy, 2013). It is therefore necessary to control for these factors. LnPopDensity is measured as the log of population per land square miles. Urbanization is operationalized as the percent of state population living in urban areas (Urban). Precipitation index (PDSI) is measured by the annual average Palmer Drought Severity Index (PDSI), which ranges from −10 to +10 with negative and positive signs indicating dry and wet, respectively. Temperature is measured as average winter temperature based on the months of December, January, and February (Temperature). LnVMT is measured as the log of annual vehicle miles traveled. TruckVMT is defined as the heavy truck share of annual VMT. LnVMT and TruckVMT are lagged one year to avoid endogeneity. We also control for the variables of state trade and economic activities. These variables include state share of total U.S. exports (StateExports), state truck industry employment (TruckEmployment), and state wholesales distribution industry employments (WholesalesEmployment). It should be noted that we excluded the variable of state imports because it is very highly correlated with the variable of state exports. Because these economic activities rely on efficient and well-maintained highway infrastructure systems, we expect that these variables have a positive sign on state highway infrastructure efficiency. Table 2 offers a descriptive summary for all variables included in the second stage.

Table 2.

Summary of Variables Used in the Second-Stage Regression.

Variable	n	M	SD	Minimum	Maximum
Competition
HwyCompetition	705	0.8901	0.057	0.8611	0.9768
Fiscal capacity
LnMedianIncome	705	2.18	0.19	1.79	2.9
Jurisdiction size
LnHwySize	705	2.06	0.21	1.67	2.45
Political institutions
DemoGovernor	705	0.58	0.49	0	1
DemoHouse	705	0.48	0.16	0.11	1
DemoSenate	705	0.48	0.15	0.06	1
DividedGov	705	0.35	0.48	0	1
PublicSectorUnion (%)	705	32.9	17.8	5.2	73.1
Fiscal institutions
Tax expenditure limitations	705	0.51	0.50	0	1
SubmitBalancedBudget	705	0.83	0.38	0	1
PassBalancedBudget	705	0.73	0.44	0	1
SignBalancedBudget	705	0.19	0.39	0	1
DeficitCarry	705	0.87	0.33	0	1
Characteristics of residents
LnDriverPop	705	2.95	0.03	2.86	3.05
LnVehicles	705	11.81	0.8	9.41	13.41
CollegeEducation (%)	705	25	5	12	38
Economic condition
Unemployment (%)	705	5	2	2	13
StateExports (%)	705	2.13	3.01	0.54	17
LnTruckEmployment	705	7.73	2.89	2.55	16.82
LnWholesalesEmployment	705	21.69	4.25	13.1	35
Highway production cost variables
LnHwyWage	705	3.25	0.4	2.29	4.33
LnLandPrice	705	1.12	0.1	0.75	1.4
Environmental factors
LnPopDensity	705	1.96	0.56	0.7	3.08
Urban (%)	705	33.39	10.58	7	62.6
Precipitation	705	0.49	2.15	−6.11	7.19
Temperature	705	0.67	0.16	0.3	0.99
LnVMT	705	0.11	0.04	0.01	0.28
TruckVMT (%)	705	40.1	7	38	42.6

Data and Method

The units of analysis in this research are 47 contiguous states (excluding Hawaii, Alaska, and Nebraska). Alaska and Hawaii are excluded because these two states are outliers in state highway capital expenditures and federal highway grants (Bruce et al., 2007).³ Following previous studies (e.g., Witko & Newmark, 2009), Nebraska was dropped because of its unique unicameral and nonpartisan legislature. Due to the consideration of data availability, the time period of panel data analysis is from 1995 to 2009.⁴ The definitions of all variables and their data sources are shown in Supplementary Appendix A.

The study uses a two-stage double bootstrap DEA method. In the first stage, the study estimates a multiple-input and -outcome DEA model with the 15 years of data jointly. DEA is a nonparametric efficiency modeling approach that analyzes input-to-output ratios for decision-making units (DMUs), which in this case are state-administered highway systems (Charnes, Cooper, & Rhodes, 1978).⁵ DEA identifies the most efficient state and sets the remaining states relative to this benchmark, resulting in efficiency scores. The most efficient state is assigned an efficiency score of 1, whereas all other states have lower efficiency scores less than 1 (meaning inefficiency). For example, an efficiency score of 0.9 would indicate that states would need to reduce inputs by 10% to be fully efficient in an input-oriented model. Because of given highway outcomes, the objective is to reduce the use of resources. The input orientation has been acknowledged to be more appropriate because states have more control over inputs (highway investment) than outcomes. In addition, the DEA method with variable returns to scale (VRS) is assumed because states display different sizes of their highway infrastructure systems.

After estimating the DEA model, the bootstrapping procedure developed by Simar and Wilson (2007) is used to calculate bias-corrected DEA efficiency scores. These corrected efficiency scores then become the dependent variable in the second-stage truncated regression, which examines the determinants of state highway infrastructure efficiency.⁶ A truncated regression means that the efficiency score is truncated at the value of 1, and there are no efficiency scores greater than 1 included in the second-stage regression model. The double bootstrap procedure is estimated using Algorithm 2 of Simar and Wilson (2007), with 500 and 1,000 replications for bias correction and for confidence intervals, respectively.⁷

Estimation Results

The First-Stage DEA Results: Estimating State Highway Infrastructure Efficiency Scores

In the first stage, the study uses the VRS output-oriented DEA model to derive state highway infrastructure efficiency index. The calculated DEA efficiency scores are bounded between 0 and 1, with 1 being more efficient. Table 3 presents the DEA results of original and bias-corrected highway infrastructure efficiency scores. The mean value of the original (uncorrected) efficiency score for the entire period is 0.954. In contrast, the mean value of the bias-corrected efficiency score is 0.859. This implies that the bias-corrected highway efficiency scores are on average, lower than the uncorrected efficiency estimates, inferring that the uncorrected efficiency estimates are upward biased. Without correcting for bias, the estimated results would have indicated that states were performing more efficiently than they actually were in the provision of state highway infrastructure.

Table 3.

Original and Bias-Corrected Efficiency Scores of State Highway Infrastructure.

Year	n	Original DEA estimates		Bias-corrected DEA estimates		Number of efficient states	% of efficient states
Year	n	M	SD	M	SD	Number of efficient states	% of efficient states
1995	47	0.949	0.047	0.854	0.052	13	27.66
1996	47	0.951	0.044	0.856	0.048	13	27.66
1997	47	0.944	0.052	0.850	0.057	12	25.53
1998	47	0.952	0.052	0.857	0.055	17	36.17
1999	47	0.942	0.053	0.848	0.058	13	27.66
2000	47	0.950	0.052	0.855	0.059	16	34.04
2001	47	0.953	0.045	0.858	0.050	15	31.91
2002	47	0.938	0.051	0.844	0.055	8	17.02
2003	47	0.951	0.046	0.855	0.051	10	21.28
2004	47	0.961	0.045	0.864	0.052	18	38.30
2005	47	0.959	0.043	0.863	0.047	17	36.17
2006	47	0.965	0.042	0.869	0.046	18	38.30
2007	47	0.966	0.042	0.870	0.049	17	36.17
2008	47	0.967	0.038	0.871	0.042	19	40.43
2009	47	0.970	0.036	0.873	0.040	20	42.55
All	705	0.954	0.046	0.859	0.050	15	32.06

Note. Estimation of the bias-corrected efficiency score is based on the first stage of Algorithm 2 of Simar and Wilson (2007), modified for the left and right boundaries of input-oriented efficiency scores. Estimation is obtained using the statistical software R—with installing the latest version of “rDEA” package (May 8, 2016. Version 1.2-4). DEA = data envelopment analysis.

The bias-corrected efficiency score of 0.859 indicates that on average, state governments reach 85.9% of the efficiency of their best practice peers in terms of providing quality highway infrastructure outcomes. It also suggests that state highway infrastructure systems could, therefore, have reduced all highway monetary inputs by 14.1%, while maintaining the same level of highway outcomes if they had been fully efficient. During the period of 1995 to 2009, Arizona, Maine, and Wyoming (the best performers) are consistently ranked as the most efficient states (their efficiency scores remain consistently at 1) from 1995 to 2009. Conversely, New Jersey is the worst performer with the lowest average highway efficiency score (0.7602) during the same period. Supplementary Appendix B presents a summary of the mean values of state highway efficiency scores. The wide variation of state highway infrastructure efficiency scores highlights the need to further explore the factors driving state highway efficiency differentials, which will be addressed in the next section.

The Second-Stage Bootstrapped Truncated Regression: Determinants of State Highway Infrastructure Efficiency

The main focus of this study is to use the two-stage semiparametric approach with double bootstrap procedures suggested by Simar and Wilson (2007) to make robust estimations about the effects of exogenous variables on public sector efficiency. The dependent variable in the second-stage truncated regression is the state highway infrastructure efficiency scores developed in the first stage. Therefore, a positive (negative) coefficient indicates a positive (negative) marginal effect on highway infrastructure efficiency. Table 4 summarizes the bootstrapped truncated regression results.⁸

Table 4.

Main Results of the Second-Stage Bootstrapped Truncated Regression.

Variables	Expected sign	Bias-corrected coefficient	Confidence intervals
Variables	Expected sign	Bias-corrected coefficient	95% Bootstrap lower bound	95% Bootstrap upper bound
Competition
HwyCompetition	+	0.090**	0.072	0.125
Jurisdiction size
LnHwySize	±	−0.026***	−0.019	−0.037
Fiscal capacity
LnMedianIncome	±	−0.141***	−0.132	−0.153
Political institutions
DemoGovernor	−	0.005**	0.003	0.006
DemoHouse	−	−0.015	−0.008	−0.024
DemoSenate	−	−0.007	−0.005	−0.014
DividedGov	−	−0.016**	−0.01	−0.021
PublicSectorUnion	−	−0.006**	−0.004	−0.007
Fiscal institutions
TELs enacted dummy	+	0.001**	0.0007	0.0011
SubmitBalancedBudget	+	0.006**	0.004	0.012
PassBalancedBudget	+	0.005**	0.003	0.009
SignBalancedBudget	+	0.004	0.002	0.007
DeficitCarry	−	−0.01	−0.007	−0.011
Characteristics of residents
LnVehicle	+	0.009**	0.006	0.012
LnDriverPop	+	0.052	0.039	0.061
CollegeEducation	+	0.191***	0.165	0.287
Economic conditions
Unemployment	±	0.036***	0.027	0.047
StateExports	+	−0.525	0.418	0.585
LnTruckEmployment	+	0.061**	0.0598	0.0662
LnWholesalesEmployment	+	−0.027	−0.022	−0.0038
Highway production costs
LnLandPrice	−	−0.042*	−0.032	−0.055
LnHwyWage	−	−0.066**	−0.054	−0.078
Environment factors
LnPopDensity	±	−0.041**	−0.033	−0.052
Temperature	+	0.000	0.000	0.000
Precipitation	−	−0.001	−0.0006	−0.0018
Urban	±	0.029	0.022	0.035
TruckVMT	−	−0.046	−0.03	−0.056
LnVMT	−	0.053	0.039	0.067
Control for years	Yes
Control for states	Yes
Log Likelihood	1,644.28
N	705

Note. Year dummy variables are included in the model but omitted for the table. Double-bootstrap truncated estimation Algorithm 2 (n = 1,000). TELs = tax and expenditure limitations.

p < .10. **p < .05. ***p < .010.

While competition is generally viewed as one way of improving public sector efficiency, a direct analysis of this effect in the public infrastructure sector is rare. As expected, interstate highway competition has a positive and statistically significant sign. This finding is in line with the educational efficiency literature (e.g., Grosskopf et al., 2001), and supports the hypothesis about the positive efficiency gains of interstate competition in the provision of state highway infrastructure. This may be because state residents use information about other states to pressure their state to be efficient or comparable in the provision of state highway infrastructure (O’Connell & Yusuf, 2011), and competition promotes responsiveness of governments to their citizens and improves the efficiency of public service provision (Kenyon, 1991)

Regarding jurisdiction size, the coefficient of the variable of state-administered highway size is negative and significant, indicating that the adverse effect of larger jurisdiction sizes (decreased citizen monitoring activities) outweighs the potential efficiency benefit of economies of scale. Concerning fiscal capacity, the variable of state median household income shows a relatively larger and negative sign. This finding is in line with the previous literature of state and local public sector efficiency (e.g., De Borger & Kerstens, 1996; Pang et al., 2014). It suggests that higher income levels increase the fiscal capacity and therefore may foster potential inefficiency because public officials have fewer incentives to provide services as efficiently as possible (Balaguer-Coll & Prior, 2009; Borge et al., 2008; Duncombe et al., 1997; Kalb, 2010).

Turing to fiscal institution variables, all of them show an expected sign but have varying statistical significances. We find that the existence of state tax expenditure limits is significantly associated with better highway infrastructure efficiency. As for BBRs, two dummy variables of the governor must submit a balanced budget and state legislature must pass a balanced budget are positive and statistically significant. These findings are in accordance with previous local government studies of Borge et al. (2008) and Falch (2001), and support the hypothesis that stringent budgetary constraints reduce public sector inefficiency. Taking a look at political institution variables, contrary to the expectation, the variable of a Democratic governor has a positive and statistically significant sign. It is hard to explain the contrasting result. Democratic governors often associated with preferences for higher spending and a larger government size, but they may not necessarily imply lower efficiency (Geys et al., 2010). Another possible explanation is that the country’s infrastructure is in bad repair and that willingness to spend more on roads and highways leads to a more efficient system.⁹ Conversely, the two variables of the ratios of Democratic members in State Houses and Senates have a negative but statistically insignificant sign. It is interesting to find that the sign in the variable of divided governments is negative and statistically significant. This finding supports the hypothesis that divided state governments achieve lower highway efficiency due to the weaker political leadership (Geys et al., 2010; Kalb, 2010). Finally, as expected, the variable of public sector union is negative and significant, inferring that the larger the percent of state government employees who are union memberships, the worse state highway efficiency (Borge et al., 2008).

Among our control variables of capturing residents’ characteristics, we find that an increasing number of total registered vehicles per capita is accompanied by a rising state highway efficiency score. This implies that vehicle owners use roads on a daily basis and may be more likely to monitor state officials to efficiently provide highway infrastructure. Similarly, the variable of the percentage of state population with a bachelor’s degree or higher has an expected positive and significant sign. This could be due to the fact that highly educated people have more capacity and willingness to monitor state officials in the efficient provision of public services. Regarding other control variables, two highway cost variables of state farmland value per acre and state mean hourly road construction wage have a negative and statistically significant coefficient. This indicates that states facing larger highway production costs must spend more to achieve the same highway outcomes in comparison with states with lower highway production costs. In addition, several highway environmental variables are statistically significant. State population density shows a negative sign. While a high population density points to cost advantages due to the regional concentration of public services (De Borger & Kerstens, 1996), it can also induce a greater amount of road travel, and therefore results in higher costs to maintain road quality and reduce traffic congestion. Unemployment has a positive coefficient. This could be due to the fact that higher unemployment rates may reduce the demand for road travel and therefore raises state highway efficiency. In addition, we find that the variable of truck industry employment has a positive coefficient, indicating that states with more people working in the truck industry are associated with better highway efficiency.

Robustness Checks

The study used various procedures to check the robustness of our results.¹⁰ Table 5 in Supplementary Appendix D presents the results of robustness checks. In Model 1, we added the state of Nebraska into the sample and re-estimated the data with dropping the two variables of Democratic House and Democratic Senate. In Model 2, we changed the orientation of the DEA model and used the output-oriented DEA model. The DEA estimation may be sensitive to outliers. Model 3 excluded state highway infrastructure efficiency scores, which are under the bottom 10% and above 90%. In Model 4, we changed the number of bootstrap replications performed in the second stage, with 1,000 and 2,000 replications for bias correction and for the confidence intervals, respectively. Finally, following the strategy of Linna (1998) and Lindlbauer et al. (2016), instead of computing efficiency scores for the pooled data, Model 5 calculated state highway efficiency scores by running the DEA model separately for each cross-sectional data from 1995 to 2009. As shown in Table 5, when these five robustness checks are included, our key variables of competition, jurisdiction size, fiscal capacity, and political and fiscal institutions remain significant with the same sign though the significance levels in political and fiscal institutions variables became weaker in Model 3 (due to a reduced sample size).

Conclusion and Policy Implications

The measurement of public sector efficiency has received increasing attention in recent years. While there are numerous studies on the measurement and explanation of public sector efficiency, particularly in the public education sector, the estimation and analysis of efficiency in the public infrastructure sector has attracted far less attention in this literature. Therefore, the purpose of this research is to add to the existing literature by estimating and studying the determinants of public highway infrastructure efficiency for a broad panel of U.S. state governments. A semiparametric approach (the two-stage double bootstrap DEA method) is applied to estimate and examine how highway efficiency scores can be explained by a number of exogenous variables among 47 American states from 1995 to 2009.

The key results of the analysis can be synthesized as follows. There is a large efficiency variation over time in state highway infrastructure during the period of 1995 to 2009. On average, the mean bias-corrected state highway efficiency score is 0.859 (less than 1), indicating that on average, state governments achieve nearly 86% of the efficiency level among the best performers and the overall state governments could reach the full efficiency level by the reduction of 14% of the three monetary highway inputs. Concerning the key highway infrastructure efficiency determinants, interstate competition, democratic governor, the existence of constitutional or statutory state tax expenditure limits, the governor must submit a balanced budget, and state legislature must pass a balanced budget influence state highway efficiency positively. Conversely, state median household income, state-administered highway size, divided government, and public sector unions negatively affect state highway efficiency.

This study contributes to the literature of public sector efficiency in four important ways. First, efficiency studies on U.S. state governments are rare except Pang et al. (2014). In fact, there is a lack of studies estimating both the service-specific efficiency and the overall government efficiency in providing multiple public services. To the best of our understanding, this study represents the first academic effort to empirically estimate and explain efficiency differences in American state infrastructure sector. Second, in this study, public choice theory is extended to explain government inefficiency in the provision of public infrastructure. Consistent with theoretical and empirical literature on public sector efficiency (e.g., Grosskopf et al., 2001; Grossman et al., 1999), competition, jurisdiction size, fiscal capacity, and political and fiscal institutions are the key efficiency drivers in the research context of state highway infrastructure sector. Third, our empirical analysis is based on a long panel of data of American state governments rather than cross-sectional data used in most previous studies (e.g., Duncombe et al., 1997; Grosskopf et al., 2001; Grossman et al., 1999; Ruggiero et al., 1995). This improves the credibility of our research findings. Fourth, this research uses recent advances in the econometric efficiency analysis—a two-stage double bootstrap DEA (see Note 10). This new approach is able to correct for efficiency estimation bias arising in the traditional two-step DEA/Tobit approach, which is predominantly used to examine the determinants of public sector efficiency (e.g., Eom & Rubenstein, 2006; Ruggiero et al., 1995).

This research has important policy implications for current discussions about the challenge of U.S. infrastructure crisis. First, there is a general consensus that U.S. transportation infrastructure has become badly deteriorated primarily due to inadequate outlays for maintenance and the underfunding of new investment needs. However, a large increase in transportation infrastructure spending is not enough to resolve the infrastructure crisis. Given our research findings, it is hard to favor a large increase in current transportation funding level without taking the efficiency of public provision of infrastructure into consideration. Second, this study suggests that government agencies can seek ways to improve the efficient provision of public infrastructure. The effective institutional setup (hard budget constraints) is one way to reduce public service inefficiency. In addition, productivity clauses and performance incentives could be incorporated into the design of institutional control to minimize potential efficiency loss. Third, this study suggests that government agencies can become more efficient in the provision of public services by seeking greater citizen monitoring efforts. On one hand, it is suggested that state governments may need to offer quality information about state infrastructure conditions and make such information more transparent and accessible to citizens. On the other hand, state governments are suggested to find ways to encourage citizen monitoring via multiple methods of citizen involvement (public hearings, citizen advisory boards, open forums) and participatory budgeting (Ebdon & Franklin, 2004).

The study has a number of limitations that will hopefully inspire more research in this area. First, our three highway inputs are measured in monetary terms (expenditure measures). Due to data availability, alterative highways inputs are not utilized. There is a need for more research to improve these measurements, especially about the construction of capital stock. Second, our efficiency findings are measure specific because we only include three key highway outcome measures. Scholars are encouraged to incorporate other valuable highway outcomes. Third, as opposed to the use of a nonparametric DEA method, future studies can use the alternative panel stochastic frontier method to estimate and explain state highway efficiency differences. However, despite the above limitations, our research sheds new lights on the measurement and determinants of public infrastructure efficiency.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Can Chen, PhD, is an assistant professor at the Department of Public Administration, Steven J. Green School of International and Public Affairs of Florida International University. His research interests are infrastructure finance, capital budgeting, and infrastructure management.

References

Afonso

Schuknecht

Tanzi

(2005). Public sector efficiency: An international comparison. Public Choice, 123, 321-347.

Afonso

Schuknecht

Tanzi

(2010). Public sector efficiency: Evidence for new EU member states and emerging markets. Applied Economics, 42, 2147-2164.

Alt

J. E.

Lowry

R. C.

(1994). Divided government, fiscal institutions, and budget deficits: Evidence from the states. American Political Science Review, 88, 811-828.

American Society of Civil Engineers. (2017). ASCE’s 2017 report card for America’s infrastructure. Available from https://www.infrastructurereportcard.org/americas-grades/

Andrews

Entwistle

(2014). Public service efficiency: Reframing the debate. Abingdon, UK: Routledge.

Balaguer-Coll

M. T.

Prior

(2009). Short-and long-term evaluation of efficiency and quality. An application to Spanish municipalities. Applied Economics, 41, 2991-3002.

Bartle

J. R.

Chen

(2014). Future issues in state transportation finance. In Rubin

M. M.

Willoughby

K. G.

(Eds.), Sustaining the states: The fiscal viability of American state governments (pp. 211-234). New York, NY: CRC Press.

Borge

Falch

Tovmo

(2008). Public sector efficiency: The roles of political and budgetary institutions, fiscal capacity, and democratic participation. Public Choice, 136, 475-495.

Bradford

D. F.

Malt

R. A.

Oates

W. E.

(1969). The rising cost of local public services: Some evidence and reflections. National Tax Journal, 22(2), 185-202.

10.

Breton

Wintrobe

(1975). The equilibrium size of a budget-maximizing bureau: A note on Niskanen’s theory of bureaucracy. Journal of Political Economy, 83, 195-207.

11.

Bruce

Carroll

D. A.

Deskins

J. A.

Rork

J. C.

(2007). Road to ruin? A spatial analysis of state highway spending. Public Budgeting & Finance, 27(4), 66-85.

12.

Case

A. C.

Rosen

H. S.

Hines

J. R.

(1993). Budget spillovers and fiscal policy interdependence: Evidence from the states. Journal of Public Economics, 52, 285-307.

13.

Chan

K. S.

Mestelman

(1988). Institutions, efficiency and the strategic behavior of sponsors and bureaus. Journal of Public Economics, 37, 91-102.

14.

Charnes

Cooper

W. W.

Rhodes

(1978). Measuring the efficiency of decision making units. European Journal of Operational Research, 2, 429-444.

15.

Chen

(2014). Measuring state highway sustainability: Taking the fiscal dimension into account. Public Works Management & Policy, 19, 255-276.

16.

Chen

(2016). Effects of fiscal stress on state highway infrastructure finance: A composite index approach. Municipal Finance Journal, 37(2), 1-28.

17.

Chen

(2017). Does money matter for infrastructure outcomes? The effects of public infrastructure finance on state infrastructure quality. Journal of Public Budgeting, Accounting & Financial Management, 29, 375-408.

18.

Chen

Kriz

K. A.

Wang

(2016). How does the health of transportation infrastructure affect state credit ratings? An empirical analysis. Public Finance Review, 44, 660-680.

19.

Daraio

Simar

(2007). Advanced Robust and nonparametric methods in efficiency analysis: Methodology and applications. Berlin, Germany: Springer Science.

20.

De Borger

Kerstens

. (1996). Cost efficiency of Belgian local governments: A comparative analysis of FDH, DEA, and econometric approaches. Regional Science & Urban Economics, 26, 145-170.

21.

De Borger

Kerstens

Moesen

Vanneste

. (1994). Explaining differences in productive efficiency: An application to Belgian municipalities. Public Choice, 80, 339-358.

22.

Deller

S. C.

Halstead

J. M.

(1994). Efficiency in the production of rural road services: The case of New England towns. Land Economics, 70, 247-259.

23.

Duncombe

Miner

Ruggiero

(1997). Empirical evaluation of bureaucratic models of inefficiency. Public Choice, 93, 1-18.

24.

Duncombe

Yinger

(1997). Why is it so hard to help central city schools? Journal of Policy Analysis and Management, 16, 85-113.

25.

Ebdon

Franklin

A. L.

(2004). Searching for a role for citizens in the budget process. Public Budgeting & Finance, 24(1), 32-49.

26.

Egilmez

McAvoy

(2013). Benchmarking road safety of US states: A DEA-based Malmquist productivity index approach. Accident Analysis & Prevention, 53, 55-64.

27.

Eom

T. H.

Rubenstein

(2006). Do state-funded property tax exemptions increase local government inefficiency? An analysis of New York State’s STAR program. Public Budgeting & Finance, 26(1), 66-87.

28.

Eom

T. H.

Lee

S. H.

(2014). A longitudinal analysis of impacts of court-mandated education finance reform on school district efficiency. Journal of Public Budgeting, Accounting & Financial Management, 26(1), 1-49.

29.

Falch

(2001). Collective bargaining in the public sector and the role of budget determination. European Journal of Political Economy, 17, 75-99.

30.

Geys

Heinemann

Kalb

(2010). Voter involvement, fiscal autonomy and public sector efficiency: Evidence from German municipalities. European Journal of Political Economy, 26, 265-278.

31.

Gray

(1976). Models of comparative state politics: A comparison of cross-sectional and time series analyses. American Journal of Political Science, 20, 235-256.

32.

Grosskopf

Hayes

K. J.

Taylor

L. L.

Weber

W. L.

(2001). On the determinants of school district efficiency: Competition and monitoring. Journal of Urban Economics, 49, 453-478.

33.

Grossman

P. J.

Mavros

Wassmer

R. W.

(1999). Public sector technical inefficiency in large U.S. cities. Journal of Urban Economics, 46, 278-299.

34.

Haug

(2008). Effects of democratic control on the efficiency of local public enterprises: Empirical evidence for water suppliers in eastern Germany. Public Finance and Management, 8, 1-19.

35.

Hayes

K. J.

Razzolini

Ross

L. B.

(1998). Bureaucratic choice and non-optimal provision of public goods: Theory and evidence. Public Choice, 94, 1-20.

36.

Heckman

A. C.

(2015). The search for better government performance: Understanding the impact of management quality and resources on transportation infrastructure outcomes in the American states. Public Management Review, 17, 187-208.

37.

Hemmeter

J. A.

(2006). Estimating public library efficiency using stochastic frontiers. Public Finance Review, 34, 328-348.

38.

Kalb

(2010). Public sector efficiency: Applications to local governments in Germany. Berlin, Germany: Springer Science.

39.

Kenyon

D. A.

(1991). Competition among states and local governments: Efficiency and equity in American federalism. Washington, DC: The Urban Institute.

40.

Leibenstein

(1966). Allocative efficiency vs. “X-efficiency.” American Economic Review, 56, 392-415.

41.

Leibenstein

(1978). On the basic proposition of X-efficiency theory. American Economic Review, 68, 328-332.

42.

Lindlbauer

Winter

Schreyögg

(2016). Antecedents and consequences of corporatization: An empirical analysis of German public hospitals. Journal of Public Administration Research and Theory, 26, 309-326.

43.

Linna

(1998). Measuring hospital cost efficiency with panel data models. Health Economics, 7, 415-427.

44.

Liu

Mikesell

J. L.

(2014). The impact of public officials’ corruption on the size and allocation of U.S. state spending. Public Administration Review, 74, 346-359.

45.

Macmanus

S. A.

(2004). “Bricks and mortar” politics: How infrastructure decisions defeat incumbents. Public Budgeting & Finance, 24(1), 96-112.

46.

Migué

Belanger

Niskanen

W. A.

(1974). Toward a general theory of managerial discretion. Public Choice, 17, 24-43.

47.

Moene

K. O.

(1986). Types of bureaucratic interaction. Journal of Public Economics, 29, 333-345.

48.

Mueller

D. C.

(1989). Public choice II. Cambridge, UK: Cambridge University Press.

49.

Neshkova

M. I.

Guo

(2013). Public participation and organizational performance: Evidence from state agencies. Journal of Public Administration Research and Theory, 22, 267-288.

50.

Niskanen

W. A.

(1971). Bureaucracy and representative government. Chicago, IL: Aldine–Atherton.

51.

Niskanen

W. A.

(1975). Bureaucrats and politicians. Journal of Law and Economics, 18, 617-643.

52.

O’Connell

Yusuf

J. E.

(2011). An information model to present proposals to increase taxes: Two examples in the context of American values. International Journal of Public Administration, 34, 180-189.

53.

Ostrom

(1972). Metropolitan reform: Propositions derived from two traditions. Social Science Quarterly, 53, 474-493.

54.

Pang

M. S.

Tafti

Krishnan

M. S.

(2014). Information technology and administrative efficiency in U.S. state governments: A stochastic frontier approach. MIS Quarterly, 38, 1079-1101.

55.

Pérez-López

Prior

Zafra-Gómez

J. L.

(2015). Rethinking new public management delivery forms and efficiency: Long-term effects in Spanish local government. Journal of Public Administration Research and Theory, 25, 1157-1183.

56.

Poterba

J. M.

(1994). State responses to fiscal crises: The effects of budgetary institutions and politics. Journal of Political Economy, 102, 799-821.

57.

Ruggiero

Duncombe

Miner

(1995). On the measurement and causes of technical inefficiency in local public services: With an application to public education. Journal of Public Administration Research and Theory, 5, 403-428.

58.

Simar

Wilson

P. W.

(2007). Estimation and inference in two-stage, semi-parametric models of production processes. Journal of Econometric, 136, 31-64.

59.

Smith

D. L.

Hou

(2013). Balanced budget requirements and state spending: A long–panel study. Public Budgeting & Finance, 33, 1-18.

60.

Stallmann

J. I.

(2007). Impacts of tax & expenditure limits on local governments: Lessons from Colorado and Missouri. Journal of Regional Analysis and Policy, 37(1), 62-65.

61.

U.S. Bureau of Economic Analysis (2004). BEA Depreciation Rate. Retrieved from https://www.bea.gov/national/FA2004/Tablecandtext.pdf

62.

U.S. Department of Transportation. (2010). 2008 Status of the Nation’s Highways, Bridges, and Transit: Conditions and Performance. Washington, DC. Retrieved from https://www.fhwa.dot.gov/policy/2008cpr/

63.

Winston

(2013). On the performance of the U.S. transportation system: Caution ahead. Journal of Economic Literature, 51, 773-824.

64.

Winston

Langer

(2006). The effect of government highway spending on road users’ congestion costs. Journal of Urban Economics, 60, 463-483.

65.

Witko

Newmark

A. J.

(2009). The strange disappearance of investment in human and physical capital in the United States. Journal of Public Administration Research and Theory, 20, 215-232.

66.

Wyckoff

P. G.

(1990). Bureaucracy, inefficiency, and time. Public Choice, 67, 169-179.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB