Sign up

Public versus private delivery of municipal solid waste services: the case of North Carolina.
Abstract:
This paper examines the effects of different institutional arrangements and characteristics on cost savings, efficiency gains, and productivity of delivering municipal solid waste services. A cost function approach is employed, and North Carolina municipal data for three years (1997, 2001, and 2003) are used for the analysis. Empirical findings indicate that there is no significant difference in cost savings between public delivery and private contractor delivery of solid waste services, a finding similar to those of other recent studies. There are three possible reasons for this. First, the threat of competition and contracting out might have led to cost savings in the cases of public delivery. Second, there might be a lack of competition because a few large private contractors have been able to win follow-on contracts over the years. Third, there might be substantial transaction costs arising as the result of private contracting. (JEL H40, H83, Q53)

Article Type:
Report
Subject:
Public contracts (Management)
Refuse collection (Contracts)
Author:
Bae, Suho
Pub Date:
07/01/2010
Publication:
Name: Contemporary Economic Policy Publisher: Western Economic Association International Audience: Academic; Trade Format: Magazine/Journal Subject: Business; Economics Copyright: COPYRIGHT 2010 Western Economic Association International ISSN: 1074-3529
Issue:
Date: July, 2010 Source Volume: 28 Source Issue: 3
Topic:
Event Code: 200 Management dynamics; 610 Contracts & orders received; 490 Contracts & orders let Computer Subject: Contract agreement; Company business management; Government contract
Product:
Product Code: 4953020 Municipal Solid Waste; 4953100 Solid Waste Collection; 4953300 Solid Waste Disposal; 9106420 Solid Waste Programs; 9106424 Refuse Disposal Programs NAICS Code: 562111 Solid Waste Collection; 56221 Waste Treatment and Disposal; 92411 Administration of Air and Water Resource and Solid Waste Management Programs SIC Code: 4953 Refuse systems
Geographic:
Geographic Scope: United States Geographic Code: 1USA United States
Accession Number:
231418927
Full Text:
I. INTRODUCTION

For a number of years government organizations have attempted to enhance the effectiveness and efficiency of public service delivery in various ways, including privatization of municipal solid waste collection and disposal. The contracting out of services to the private sector has increased, although many municipalities still deliver services directly.

Several studies have found that cost minimization is a major consideration for service delivery choices in waste and recycling services (Bel and Miralles, 2003; Walls, Macauley, and Anderson, 2005). According to Bel and Miralles (2003), compared with political factors such as political party affiliation and ideology, service cost is the most important factor when municipalities decide whether to contract out municipal solid waste services to private companies or deliver services themselves. It has been generally understood that the contracting out of municipal solid waste services can yield significant cost savings. Contrary to this general understanding, however, recent evidence shows that there is no relationship between contractual arrangements and cost savings (e.g., Bel and Costas, 2006; Callan and Thomas, 2001; Dijk-graaf and Gradus, 2003, 2007).

The purpose of this paper is to examine the effects of different institutional arrangements and service characteristics--including private contractors versus public delivery--on cost savings, efficiency, and productivity of municipal solid waste services. To do so, this paper employs a cost function approach and uses municipal data from North Carolina covering three years (1997, 2001, and 2003).

This paper contributes to the current literature in several ways. As some previous studies have done (e.g., Szymanski, 1996; Szymanski and Wilkins, 1993), this paper uses data from different years (i.e., 1997, 2001, and 2003). This approach allows us to control for intertemporal factors by using yearly dummies. In particular, this approach is useful to control for contractual dynamics and the problem of holdup in the case of private contracting. After successfully winning initial contracts in the first tendering process, private contractors are able to keep winning the contracts and raise the price of solid waste service delivery year after year. In this paper, solid waste costs include collection, recycling, and disposal of commercial and industrial nonhazardous solid waste, as well as residential waste; in most studies, solid waste costs include only collection of residential waste. Although this approach makes it difficult to directly compare empirical findings in this paper with other studies, it allows for a more complete view of overall cost savings, efficiency, and productivity gains. In addition, this paper employs a more complex and detailed model in examining the effects of different institutional arrangements and characteristics on cost savings, efficiency gains, and productivity of delivering municipal solid waste services; it introduces more variables to provide more precise estimations and control for region- and municipality-specific factors.

Section II discusses different institutional arrangements and service characteristics in municipal solid waste service delivery and provides an overview of current literature. In Section III, the translog cost function for empirical estimation is developed and discussed, along with the cost share equation, elasticity of substitution, and productivity. Section IV explains variables and data sources employed for empirical analysis and presents descriptive results. Section V presents empirical findings, and Section VI summarizes important findings and discusses study limitations.

II. PUBLIC VERSUS PRIVATE DELIVERY OF MUNICIPAL SOLID WASTE SERVICES

Generally, public domestic services are delivered using one of the three broad institutional arrangements: (1) (1) public delivery, (2) privatization, and (3) contracting out. According to Domberger and Jensen (1997), privatization means the transfer of ownership of physical assets from the public to private sector. But privatization does not necessarily promote competition in service delivery. (2) Contracting out means the organizations that offer the lowest prices in the bidding process are chosen to deliver particular services for the duration of the contract term. The distinction is that contracting out is "competition for the market as opposed to competition in it" (Domberger and Jensen 1997). According to Vickers and Yarrow (1991), however, contracting out is also a kind of privatization. Contracting out allows the contractor to appropriate any financial surplus gained through service delivery, meaning that the profit is transferred to the contractor. The transfer of the profit is central to ownership (Vickers and Yarrow, 1991).

Most public domestic services, including municipal solid waste services, have natural monopoly characteristics, implying that the cost of the service delivery decreases when only one service provider delivers the service within a region. But through a competitive bidding process, services can be delivered at the lowest prices (Demsetz, 1968). Private contractors may have stronger incentives to save on service delivery costs and realize technical efficiencies than public entities (Bel and Miralles, 2003). Therefore, contracting out may be able to deliver public services with natural monopoly characteristics at the lowest cost.

However, there are at least three major issues with contracting out. First, the marketplace transaction costs might be substantially higher than the benefits; transaction costs arising from contracting out include the "writing of specifications and contracts, and evaluating tenders, and negotiating the final contract with the winning tenderer" (Domberger and Jensen, 1997). Additionally, according to Batley (1996, p. 749), contracting out requires new governmental roles, such as "setting broad frameworks of policy and standards, drawing up and monitoring contracts, regulating contractors and monopolists, coordinating, financing and supporting producers, and supporting consumers with information and finance." These new roles incur costs as well. Second, contracting out may not be complete because it is very difficult to write everything that is likely to happen for the duration of the contract term and at every stage of service delivery into the contract (Bel and Miralles, 2003; Domberger and Jensen, 1997). Third, contracting out may result in low quality of services as private contractors may have a stronger incentive to reduce service delivery costs than to improve or even maintain quality of service (Blank, 2000; Bloomfield. 2006).

Several studies have found that contracting out leads to lower solid waste collection costs and creates a higher level of productivity than delivery by local governments (e.g., Batley, 1996; Domberger, Meadowcroft, and Thompson. 1986: McDavid, 1985; Post. Broekema. and Obirih-Qpareh, 2003: Reeves and Barrow. 2000; Szymanski and Wilkins, 1993); (3) estimated cost savings vary from 20% to 49%. Some studies fail to find significant differences between public units that won contracts in the competitive tendering process and private firms (Domberger, Meadowcroft, and Thompson. 1986; Szymanski and Wilkins. 1993). According to Ohlsson (2003). public and private firms adopt different production technologies, and private firms pay more for capital stocks than public firms. Using data on Swedish municipalities and firms, Ohlsson found that direct public service delivery costs are 6% lower than those of private contractors.

On the other hand, recent evidence shows that there is no significant difference in cost savings between private contractors and public delivery (e.g.. Bel and Costas, 2006; Callan and Thomas, 2001: Dijkgraaf and Gradus, 2007). There are several possible explanations for this. First, if a considerable threat of competition and contracting out exists in delivering municipal solid waste services, public managers are more concerned about cost savings, efficiency, and productivity gains, which may improve the performance of public delivery and result in no cost difference between public delivery and private contractors. For example, in areas where services were delivered by public entities that were adjacent to areas where services were outsourced to private companies, economic performance of public delivery improved considerably (Hodge, 1998). Second, private contracting itself may not be sufficient for performance improvements if there is no competition or weak competition (Dijkgraaf and Gradus, 2007; Vickers and Yarrow, 1991). In North Carolina, only a few large private contractors, including Waste Management. Waste Industries, and BFI Waste Service, collect and dispose of residential solid waste. In addition, these private contractors have been able to win follow-on contracts over the years, which may imply weak competition in the residential solid waste market in North Carolina. Third, transaction costs arising from contracting out could be substantially large and exceed the benefits arising from contracting out (Post, Broekema, and Obirih-Opareh, 2003). (4)

As previously mentioned, this paper accounts for the collection, recycling, and disposal of municipal solid waste. Here, municipal solid waste includes not only residential solid waste but also commercial and industrial nonhazardous solid waste. Therefore, as this paper considers broader aspects of the delivery of municipal solid waste services, our findings may not be directly comparable to the existing literature, which focuses on the cost-saving effects of contracting out service delivery of residential solid waste collection. To do empirical estimation, this paper accounts for more institutional and service characteristics related to collecting, recycling, and disposing municipal solid waste.

III. EMPIRICAL MODEL OF MUNICIPAL SOLID WASTE SERVICE COSTS

Based on the hedonic cost function approach of Feigenbaum and Tecples (1983) and Schmit and Boisvert (1996), the production function for solid waste collection, recycling, and disposal in a local jurisdiction can be written as

Q(Y; [Z.sub.1], [Z.sub.2], ..., [Z.sub.l]) = f(L, K, E), (1)

where Q(*) is an index of output in a local jurisdiction, Y is the total amount of solid waste collected, recycled, and disposed of in the local jurisdiction, and [Z.sub.1], [Z.sub.2], ..., [Z.sub.l] are the variables representing institutional and service attributes associated with collection, recycling, and disposal of Y. Thus, output of solid waste (presented as Q) reflects the annual collection, recycling, and disposal of solid waste (measured in tons per year and presented as Y), as well as associated institutional and service characteristics. L is labor input, K is capital input, and E is energy input.

According to duality theory, (5) an indirect cost function can be derived from the production function in equation (1) as

(2) C = C(Q(Y; [Z.sub.1], [Z.sub.2], ..., [Z.sub.l]); [P.sub.w], [U.sub.k], [P.sub.e]; [F.sub.1], [F.sub.2], ..., [F.sub.m]),

where solid waste cost per collection at a residential site (C) is a function of exogenously determined input prices ([P.sub.w], [U.sub.k], [P.sub.e]), output (Q), and fixed factors ([F.sub.1], [F.sub.2], ..., [F.sub.m]). [P.sub.w] is the hourly wage of a solid waste worker, [U.sub.k] is the user cost of private capital, and [P.sub.e] is the price of energy per unit.

This paper utilizes a hybrid translog cost function for empirical estimation, which can be approximated by the second-order Taylor expansion series (Cave, Christensen, and Trethe-way, 1980). It allows for interaction terms and quadratic functional forms among explanatory variables. A translog cost function approach has been employed in many previous studies to analyze performance of public service delivery (e.g., Bruggink, 1982; Estache and Rossi, 2002; Feigenbaum and Teeples, 1983; Schmit and Boisvert, 1996).

From the cost function in Equation (2), the following translog cost function is used for empirical estimation:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where E represents a dummy variable for whether a local jurisdiction has its own municipal electric system (as data on price of electricity are not available). Yearly and regional dummies are employed to control for unexplained variations across years and regions in North Carolina. It is assumed that In Q = ln Y + ln g ([Z.sub.1], [Z.sub.2], ..., [Z.sub.l]) and ln g(*) = [SIGMA][[beta].sub.t](ln [Z.sub.l]) (Feigenbaum and Teeples, 1983; Schmit and Boisvert, 1996).

In the translog cost function in Equation (3). to ensure that the production is convex and to satisfy the cost-minimization problem, we need two assumptions (Mas-Colell, Whinston, and Green. 1995, pp. 139-143): the symmetry requirement, which is [beta].sub.ij] = [[beta].sub.ji] for i [not equal to] j; and the homogeneity of degree one, which is as follows: (6)

(4) [SIGMA][[beta].sub.i] = 1; and [SIGMA][[beta].sub.ij] = [SIGMA][[mu].sub.yi] = 0 for i, j = L, K.

Utilizing Shephard's lemma, two demand functions in the shares of total solid waste collection, recycling, and disposal costs can be derived from the translog cost function in Equation (3). Because the cost shares sum to unity, however, error terms are correlated across the cost share equations and the covariance matrix is singular. In other words, any share equation can be expressed as a linear combination of the other equation. To avoid this problem of singularity, it is necessary to delete one share equation. In this paper, the share equation of private capital is deleted. (7) The share equation of labor via Shephard's lemma is derived from the function in Equation (3). as follows:

[S.sub.L] = ([[[partial derivative] ln C]/[[partial derivative] ln [P.sub.w]]]) = [[beta].sub.w] + [[beta].sub.ww] ln [P.sub.w] + [[mu].sub.yw] ln Y + [[beta].sub.wk] ln [U.sub.k], (5)

where [S.sub.L] is the cost share of labor. In the translog cost function and the share equation of labor, the restrictions on the slope coefficients are imposed across the equations (Cameron and Trivedi, 2005, p. 210). In other words, each slope coefficient in Equation (5) has the same slope coefficient in Equation (3). For example, [[mu].sub.yw]S are identical to each other in Equations (3) and (5).

The translog cost function and the share equation of labor are estimated as a system, using Zellner's seemingly unrelated regression (SUR) method, to increase efficiency (Greene, 2003, pp. 614-635; Zellner, 1962). This method is equivalent to the equation-by-equation ordinary least squares method (Cameron and Trivedi, 2005, pp. 209-210).

A. Elasticities of Substitution

Own-price and cross-price elasticities of input demand and Morishima elasticities of substitution are estimated using parameters from the translog cost function in Equation (3). Own-price and cross-price elasticities of input demand are written as

(6) [[member of].sub.ii] = [[[beta].sub.ii]/[S.sub.i]] + [S.sub.i] - 1 and [[member of].sub.ij] = [[[beta].sub.ij]/[S.sub.i]] + [S.sub.j](i [not equal to] j) for i, j = L, K,

where [[member of].sub.ij] > 0 implies that labor and private capital are substitutes, and [[member of].sub.ij] < 0 implies that they are complements, [[member of].sub.ii] represents the change in the demand for input i with an increase (or decrease) in input i"s price, and [[member of].sub.ij] tells us the change in the demand for input i with an increase (or decrease) in input j's price.

Morishima elasticities of substitution can be expressed from own-price and cross-price elasticities of input demand in Equation (6):

(7) [[theta].sub.ij] = [[member of].sub.ij] - [[member of].sub.jj](i [not equal to] j) for i, j = L, K,

where [[theta].sub.ij] measures percent change in the factor ratio of input i to input j when the price of input j increases by 1%, under the assumption that the price of the other input is held constant (Nguyen and Streitwieser. 1999). (8)

B. Total Factor Productivity Level and Efficiency Gains

To compare the productivity level between public delivery and private contractors, the value of residual in each municipality is employed, because it is the logarithmic deviation of its total factor productivity (TFP) level from the average productivity of all municipalities (Bae. 2009; Lichtenberg and Siegel, 1987; Martin. McHugh, and Johnson. 1991). As this paper employs a cost function approach for empirical estimation, a smaller value of residual represents a better TFP level.

The value of residual also measures relative efficiency gain or loss compared to the predicted value of the estimated translog cost function. In other words, as a cost functional approach is employed for empirical estimation, a residual value below zero represents efficiency gain, and a value above zero, efficiency loss (Reeves and Barrow, 2000).

IV. DATA SOURCES, VARIABLES, AND DESCRIPTIVE RESULTS

A. Data Sources and Variables

This paper analyzes what factors and characteristics, including institutional arrangements and service characteristics, influence performance of municipal solid waste services. It utilizes North Carolina municipal data, which are available for three years (1997, 2001. and 2003) from the solid waste survey results conducted by the North Carolina League of Municipalities (NCLM). (9) Municipal finance data, including total solid waste expenditures and solid waste salaries, were obtained from North Carolina County and Municipal Financial Information (North Carolina Department of Stale Treasurer Web site). Median household income data were obtained from Census 2000 Demographic Profiles (Census Bureau Web site), data on form of municipal government from Forms of Government and Methods of Election in North Carolina Cities (University of North Carolina School of Government Web site), and data on whether a municipality is part of a metropolitan statistical area (MSA) from Business and Economic Statistics (Rand California Statistics Web site). (10)

In North Carolina, all municipalities (i.e., cities, towns, and villages) are incorporated and do not have home rule." But North Carolina municipal governments are delegated powers that are equivalent to those that counterparts in home rule states enjoy (Bluestein, 2006). North Carolina municipalities provide crucial public services to citizens, including providing water, sewer, solid waste, police, street, recreation, land use planning, and fire protection services (Fleer, 1994, pp. 198-216; NCLM Web site). They enjoy considerable autonomy in deciding how and under what institutional mechanisms those services are delivered to citizens. In North Carolina, municipal solid waste services are either delivered by municipalities or contracted out to private companies. Several large private service providers, such as Waste Management, Waste Industries, and BFI Waste Service, collect and dispose of municipal solid waste under contracts with the municipalities.

The existing literature extensively uses municipal data within one state in the United States, as done in this paper (e.g., Callan and Thomas, 2001; Collins and Downes, 1977; Hirsch, 1965; Pier, Vernon, and Wicks, 1974; sec Bel and Warner, 2008 for details). This approach allows for estimating more precisely the effects of institutional arrangements and service characteristics on cost savings, efficiency gains, and productivity of delivering municipal solid waste services, and also controls for state-level legislative and regulatory factors in empirical estimation, including environmental regulations and regulations related to municipal solid waste services (Vickers and Yarrow, 1991). (12) However, region- and municipality-specific factors may still be important in explaining solid waste service costs. To reduce this problem, this paper employs regional dummies and municipality-specific variables such as form of municipal government, median household income, and urban/rural location (explained below).

Solid waste cost per collection at a residential site (13) is employed as the dependent variable in empirical estimation, which allows for a more parsimonious specification and more precise estimations by directly accounting for number of residential sites served and frequency of collection. This value was obtained from total costs of solid waste collection, recycling, and disposal divided by the total number of collections in a year (which was constructed from the frequency of residential solid waste collection and the number of residential sites served).

Total amounts of municipal solid waste collected, recycled, and disposed of (presented as Y) are employed to represent the demand for municipal solid waste services. As previously mentioned, municipal solid waste includes not only residential solid waste but also commercial and industrial nonhazardous solid waste. Hourly wage per solid waste worker was obtained from solid waste salaries divided by the number of total hours worked, which was constructed as [total work days (except Saturdays, Sundays, and official federal holidays) multiplied by the number of full-time equivalent employees multiplied by 8 hr per day]. (14) The ratio of capital outlays to total expenditures in each fiscal year in a municipality is employed as a proxy for user cost of capital, under the assumption that user cost of private capital increases as the municipality spends more money on capital outlays. Cost share of labor was obtained from expenditures on solid waste salaries divided by total solid waste expenditures.

Several variables are also employed to represent service attributes associated with solid waste collection, recycling, and disposal, including institutional arrangements of service delivery. The first variable concerns who collects the residential solid waste: the municipality or a private contractor. This is coded as 1 if the municipality collects residential solid waste and coded as 0 if a private contractor collects it. As mentioned in Section II, empirical evidence in the existing literature shows mixed results on the relationship between contractual arrangements and cost savings. The second variable is whether the municipality implements a recycling and reuse program: if the municipality implements a recycling and reuse program, it is coded as 1; otherwise, it is coded as 0.

The third variable is whether a sanitary landfill facility is operated by the municipality, private company, county, or regional entity. When a sanitary landfill is run by either the municipality or under cooperation between the municipality and another entity (i.e., private company, county, or regional entity), it is coded as 1; otherwise, it is coded as 0. When the municipality collects residential solid waste, implements a recycling and reuse program, and operates a sanitary landfill facility, it may be able to reduce costs of solid waste collection, recycling, and disposal through economies of scope. Economies of scope are realized through "joint use of inputs, shared marketing programs, or common program administration'' (Callan and Thomas, 2001). Thus, in the presence of economies of scope, costs of solid waste services are expected to be lower.

The fourth variable is whether a municipality delivers special services or concessions for a downtown business district (DBD). Under municipal delivery of special services or concessions for DBDs, costs of solid waste services are expected to be higher because municipal delivery of those services requires additional resources. If a municipality delivers special services or concessions for DBDs, it is coded as 1; otherwise, it is coded as 0. The fifth variable is the distance to a landfill site. With an increase in round-trip miles to the landfill, costs of solid waste services are expected to increase.

The sixth variable is whether the municipality has its own electric system. When the municipality has its own system, costs of solid waste services are expected to be higher, because the electric system also produces solid waste during electricity generation, transmission, and delivery. The seventh variable is the percentage of yard waste composted and mulched out of total yard waste. The more yard waste is composted and mulched, the greater the solid waste service costs are expected to be because that type of waste requires additional equipment and personnel. The eighth variable is the municipality's average ratio of debt to total revenues over a 5-yr period, from 1997 to 2001; this value is employed as a proxy for fiscal stringency. Large ratios represent weak fiscal stringency.

The ninth variable is the municipal government structure. There are two main forms of municipal government in North Carolina: council-manager and mayor-council. The council-manager form is more popular among municipalities in North Carolina than the mayor-council form. Professional managers in the council-manager form may be able to bring leadership and expertise to municipalities and attract other career professionals to local administrations, including professionals in delivery of municipal solid waste services (Fleer, 1994, pp. 199-212). Thus, costs of solid waste services are likely to be lower for the municipalities that have the council-manager form than for the municipalities that have the mayor-council form. But empirical evidence shows mixed results, and recent literature reports that the distinction between the two forms of municipal government is blurring, partly because many mayor-council municipalities increasingly hire professional managers and administrators (Frederickson, Johnson, and Wood, 2004, pp. 52-67; Jung, 2006). The tenth variable is the median household income in the municipality; this is employed to represent the socioeconomic characteristics of the municipality. Median household income is expected to be positively related to solid waste costs because high-income municipalities tend to spend more money on municipal solid waste services. The eleventh variable considers whether the municipality is part of an urbanized area; this may be important to control factors such as the size of the labor market. If the municipality is located in an urban area, known as an MSA, it is coded as 1; otherwise, it is coded as 0. These last three variables--form of municipal government, median household income, and urban/rural location--are employed to represent municipality-specific factors and characteristics that might explain municipal solid waste costs.

Two yearly dummies (2001 and 2003) are included to control for unexplained variations across years. In particular, this approach can control for contractual dynamics and the problem of holdup in the case of private contracting. After successfully winning initial contracts in the first tendering process, private contractors may be able to keep winning contracts in the tendering process and raise the price of solid waste service delivery year after year. The year 1997 is used as the base year.

Six regional dummies are employed to control for unexplained variations across regions in North Carolina. As shown in Figure 1, North Carolina is organized into seven economic development regions: Advantage West, Charlotte, Piedmont Triad, Research Triangle, Northeast, Eastern, and Southeast. Local governments located in a region are similar in such terms as natural endowments, level of development, and economic/industrial structure. The Advantage West region is used as the base category for the other six regions.

[FIGURE 1 OMITTED]

B. Descriptive Results

Table 1 reports descriptive results for municipal solid waste services in North Carolina. Municipalities in North Carolina are major players in the delivery of residential solid waste collection services. The average cost of solid waste collection, recycling, and disposal per collection at a residential site in municipalities using public delivery ($3.06) is lower than the average cost in municipalities using private contractors ($3.78), whereas the average tons of municipal solid waste being collected, recycled, and disposed of in municipalities using public delivery is more than twice as large as municipalities using private contractors.

Hourly wage per solid waste worker in municipalities using public delivery ($12.61) is higher than in municipalities using private contractors ($10.45), which is also consistent with previous findings (e.g., Post, Broekema, and Obirih-Opareh, 2003). (15) In contrast, user cost of capital is slightly higher in municipalities using private contractors than in municipalities using public delivery: the average ratio of capital outlays to total expenditures in a municipality, used as a proxy for user cost of capital, is 0.164 for public delivery, whereas it is 0.184 for private contractors.

More municipalities using public delivery implement recycling and reuse programs (69%) than those using private contractors (44%). More municipalities using public delivery deliver special services or concessions to DBDs (42%) than municipalities using private contracting arrangements (27%). Also, 17% of municipalities using public delivery operate sanitary landfill facilities, while 19% of municipalities using private contractors operate them.

Private contractors transport solid waste to sanitary landfill facilities farther (28 mil) than municipalities using public delivery (25 mil). More municipalities using public delivery own their own electric systems (31%) than those using private contractors (26 %). In municipalities using public delivery, more yard waste is composted and mulched (63%) than in municipalities using private contractors (53%).

More municipalities using public delivery (87%) have the council-manager form of government than those using private contractors (65%). Municipalities that directly deliver residential solid waste collection services are fiscally less stringent than those that contract those services out to private companies. Median household income in municipalities using private contractors ($41,534) is higher than that in municipalities using public delivery ($37,765). More municipalities using public delivery (57%) are part of MSAs than municipalities using private contractors (49%).

In municipalities using public delivery, solid waste services are more labor-intensive, whereas they are more capital-intensive in those using private contractors. As shown in Table 1, the average share of labor cost is 0.46 in municipalities using public delivery, whereas it is 0.25 in municipalities using private contractors.

V. EMPIRICAL RESULTS

Table 2 reports empirical results for municipal solid waste services obtained from the estimation of three translog cost functions, along with their share equations for labor, using Zellner's SUR estimation method. One advantage of a translog cost function approach is that it allows for incorporation of other variables of interest while keeping the theoretical integrity of the model (Caves, Christensen. and Trethe-way, 1980). (16) As mentioned above, this paper includes several variables of interest about service attributes associated with solid waste collection, recycling, and disposal, including institutional arrangements of service delivery, local characteristics, and fixed factors.

The results, as shown in Table 2. illustrate that all the first-order and second-order coefficients of input prices and total amounts of solid waste collected, recycled, and disposed of (except the first-order coefficient of total amounts of solid waste) are significant and show the expected signs. [R.sup.2]'s for the cost functions are 0.005 to 0.014. while [R.sup.2]'s for the share equations for labor are 0.044 to 0.045 across the three models.

Several variables are employed to represent service attributes associated with solid waste collection, recycling, and disposal. Institutional arrangements in the delivery of residential solid waste collection services have no significant effect on solid waste cost per collection at a residential site. This is consistent with recent evidence that demonstrates no significant difference in cost savings between public delivery and private contractors (e.g., Bel and Costas, 2006: Callan and Thomas. 2001: Dijkgraaf and Gradus, 2007).

Municipal implementation of recycling and reuse programs has no effect on cost savings. Solid waste cost per collection at a residential site is not significantly different under municipal operation of a sanitary landfill facility than under operation by a county/regional entity or private company. Also of interest is whether economies of scope exist when municipalities deliver residential solid waste collection services, implement recycling and reuse programs, and operate sanitary landfill facilities. As shown in Model 3 in Table 2, however, the interaction term is not significant, implying the nonexistence of economies of scope.

When there is no municipal ownership of the electric system, average solid waste cost per collection at a residential site is significantly lower, ranging from 21% to 22%, than when there is a municipal ownership, probably because electric systems produce solid waste during electricity generation, transmission, and delivery. (17) Fiscal stringency, measured as the average ratio of debts to total revenues from 1997 to 2001, is negatively associated with cost savings, probably because large municipal debt may cause great pressure on cost savings in the delivery of public programs and services, including municipal solid waste services. In municipalities with a mayor-council form of government, average solid waste cost per collection at a residential site is significantly lower, ranging from 30% to 32%, than municipalities with a council-manager form.

A. Elasticities of Substitution

The whole sample in this paper is divided into two subsamples: municipalities that directly deliver residential solid waste collection services and municipalities that outsource those services to private contractors. Table 3 reports own-price and cross-price elasticities of input demand, and Morishima elasticities of substitution (18) for three samples. These were calculated from estimated parameters in Model 2 in Table 2 and average cost shares of labor in Table 1. As there are two input factors, as shown in Table 3, in absolute terms, the own-price elasticity of labor is identical to the cross-price elasticity of labor demand with respect to capital, and the own-price elasticity of capital is identical to the cross-price elasticity of capital demand with respect to labor. As shown in Table 3, two Morishima elasticities of substitution are identical to each other.

All the own-price elasticities in the three samples have the theoretically correct negative signs and are inelastic for increase in their own prices. In the whole sample, along with a 10% increase in hourly wage per solid waste worker, the demand for solid waste workers decreased by 3.8%. A 10% increase in user cost of capital led to a decrease in the demand for private capital stock by 2.8%. But there are some large differences in own-price elasticities between the two subsamples. The own-price elasticity of labor is -0.36 in the subsample with municipalities using public delivery and -0.42 in the sub-sample with municipalities using private contractors. The own-price elasticity of capital is -0.30 in the former subsample, and -0.14 in the latter.

In the three samples, cross-price elasticities of input demand are positive but not large, implying that solid waste labor force and private capital are weak substitutes. In the whole sample, a 10% increase in hourly wage per worker increased the demand for private capital stock by 2.8%, whereas the demand for solid waste workers increased by 3.8%, along with a 10% increase in user cost of capital. But there are some large differences in cross-price elasticities between the two subsamples. The cross-price elasticity of capital demand with respect to labor is 0.30 in the subsample with municipalities using public delivery, whereas it is 0.14 in the subsample with municipalities using private contractors. The cross-price elasticity of labor demand with respect to private capital is 0.36 in the former subsample, whereas it is 0.42 in the latter.

All the Morishima elasticities of substitution are positive but smaller than unity, implying that labor and capital are weak Morishima substitutes. Municipalities using public delivery have slightly higher Morishima substitutability than municipalities using private contractors. Along with a 10% increase in hourly wage per worker (user cost of capital), the ratio of capital to solid waste workers (the ratio of solid waste workers to capital) increased by 6.6% in the former subsample but only by 5.6% in the latter.

B. Total Factor Productivity Level and Efficiency Gains

Table 4 shows productivity and efficiency gains (19) either with public delivery or private contractors. As this paper uses a cost function approach, a smaller value of residual represents a better TFP level (Bae, 2009; Lichtenberg and Siegel, 1987; Martin, McHugh, and Johnson, 1991). These residuals were obtained from the estimated translog cost function in Model 2 in Table 2. As shown in Table 4, the average TFP level with public delivery is higher than with private contractors.

The residuals also measure relative efficiency gains or losses compared to the prediction of the estimated translog cost function in Model 2 in Table 2. In other words, the residual value below zero represents efficiency gain, and above zero, efficiency loss (Reeves and Barrow, 2000). As shown in Table 4, in the whole sample, 45.6% of municipalities (i.e., 115 out of 252 municipalities) achieved efficiency gains: 48.8% of municipalities using public delivery (i.e., 102 out of 209 municipalities) achieved efficiency gains, whereas 30.2% of municipalities using private contractors achieved efficiency gains (i.e., 13 out of 43 municipalities). Figure 2 also shows relative efficiency gains or losses under the two institutional arrangements. More residuals with public delivery are below zero than those with private contractors, implying that municipalities using public delivery exhibited more efficiency gains than municipalities using private contractors.

[FIGURE 2 OMITTED]

VI. CONCLUSIONS

Municipalities deliver collection, recycling, and disposal of municipal solid waste services either through public delivery or private contractors. This paper has examined the effects of different institutional arrangements and characteristics on cost savings, efficiency gains, and productivity of delivering solid waste services. To do so, we have employed a cost function approach and have used North Carolina municipal data for the years 1997, 2001, and 2003.

The empirical findings indicate that there is no significant difference in cost savings between public delivery and private contractors, as other recent studies have shown as well (e.g., Bel and Costas, 2006; Callan and Thomas, 2001; Dijkgraaf and Gradus, 2007). Municipal ownership of an electric system and the council-manager form of municipal government are positively associated with high solid waste costs. Fiscal stringency, measured as the ratio of debts to total revenues, is negatively associated with cost savings, probably because high debts result in great pressure for cost savings. Solid waste labor force and private capital are weak substitutes, although there is some difference in substitutability under the two institutional arrangements. Municipalities using public delivery of residential solid waste collection services realize more efficiency gains than those using private contractors; they also achieve higher productivity levels using public delivery than using private contractors.

This paper has several limitations. First, it has analyzed cost savings effects of different institutional arrangements for delivery of solid waste services, but it has not looked at the effects on the quality of those services, mainly due to the unavailability of data. Private contractors may lead to lower quality solid waste services because they may have a stronger incentive to reduce costs rather than maintain or improve service quality. Further research needs to be done to examine the effects of different institutional arrangements on quality and availability of municipal solid waste services. Second, this study was not able to analyze why there was no difference in the costs of municipal solid waste services between public delivery and private contractors, mainly due to data unavailability. There are several possible factors that could lead to this outcome: (1) considerable threats of competition and contracting out, (2) lack of competition, and/or (3) transaction costs. Thus, future research should examine which of these factors, if any, are the most important in explaining why there was no difference in the costs of solid waste services between public delivery and private contractors.

REFERENCES

Bae. S. "The Responses of Manufacturing Businesses to Geographical Differences in Electricity Prices.'" Annals of Regional Science. 43, 2009, 453-72.

Batley, R. "'Public-Private Relationships and Performance in Service Provision." Urban Studies, 33, 1996, 723-51.

Bel, G., and A. Costas. "Do Public Sector Reforms Get Rusty? Local Privatization in Spain." Journal of Policy Reform. 9(1), 2006, 1-24.

Bel, G., and A. Miralles. "Factors Influencing the Privatization of Urban Solid Waste Collection in Spain." Urban Studies. 40, 2003, 1323-34.

Bel, G., and M. Warner. "Does Privatization of Solid Waste and Water Services Reduce Costs? A Review of Empirical Studies." Resources. Conservation and Recycling. 52, 2008, 1337-48.

Blackorby, C. and R. R. Russell. "Will the Real Elasticity of Substitution Please Stand Up? (A Comparison of the Allen/Uzawa and Morishima Elasticities.)" American Economic Review. 79, 1989, 882-8.

Blank. R. M. "When Can Public Policy Makers Rely on Private Markets? The Effective Provision of Social Services." Economic Journal. 110, 2000, 34-49.

Bloomfield. P. "The Challenging Business of Long-Term Public-Private Partnerships: Reflections on Local Experience." Public Administration Review. 66, 2006, 400-11.

Bluestein, F. S. "Do North Carolina Local Governments Need Home Rule?" Popular Government. Fall 2006, 15-24.

Bruggink, T. H. "Public Versus Regulated Private Enterprises in the Municipal Water Industry: A Comparison of Operating Costs." Quarterly Review of Economics and Business, 22, 1982, 111-25.

Callan, S. J., and J. M. Thomas. "Economies of Scale and Scope: A Cost Analysis of Municipal Solid Waste Services." Land Economics, 77(4), 2001, 548-60.

Cameron, A. C, and P. K. Trivedi. Microeconometrics: Methods and Applications. New York: Cambridge University Press. 2005.

Caves, D. W., L. R. Christensen. and M W. Tretheway. "Flexible Cost Functions for Mulliproduct Firms." Review of Economics and Statistics. 62, 1980, 477-81.

Census Bureau. Census 2000 Demographic Profiles. Accessed June 10. 2008. http://censtats.census.gov/pub/Profiles.shtml.

Collins, J. N., and B. T. Dowries. "The Effect of Size on Provision of Public Services: The Case of Solid Waste Collection in Smaller Cities." Urban Affairs Quarterly, 12(3), 1977, 333-47.

Connaughton, J. E., and R. A. Madsen. "The Economic Impacts of the North Carolina Motorsports Industry." Economic Development Quarterly, 21(2), 2007, 185-97.

Demsetz. H. "Why Regulate Utilities?" Journal of Law and Economics. 9, 1968, 55-65.

Dijkgraaf, E., and R. H. J. M. Gradus. "Cost Savings of Contracting Out Refuse Collection." Empirica, 30, 2003, 149-61.

Dijkgraaf, E., and R. H. J. M. Gradus. "Collusion in the Dutch Waste Collection Market." Local Government Studies, 33(4), 2007, 573-88.

Domberger, S., and P. Jensen. "Contracting Out by the Public Sector: Theory. Evidence, Prospects." Oxford Review of Economic Policy, 13(4), 1997, 67-78.

Domberger, S., S. Meadoweroft, and D. Thompson. "Competitive Tendering and Efficiency: The Case of Refuse Collection." Fiscal Studies, 7(4), 1986, 69-87.

Dubin, J. A., and P. Navarro. "How Markets for Impure Public Goods Organize: The Case of Household Refuse Collection." Journal of Law, Economics, and Organization. 4(2), 1988, 217-41.

Estache, A., and M. A. Rossi. "How Different Is the Efficiency of Public and Private Water Companies in Asia?" World Bank Economic Review. 16, 2002, 139-48.

Feigenbaum, S., and R. Teeples. "Public Versus Private Water Delivery: A Hedonic Cost Approach." Review of Economics and Statistics. 65, 1983, 672-78.

Fleer, J. D. "Grassroots Governments: The Many Who Govern." in North Carolina Government and Politics, edited by J. D. Fleer. Lincoln, NE: University of Nebraska Press. 1994, 197-216.

Frederickson, H. G., G. A. Johnson, and C. H. Wood. The Adopted City: Institutional Dynamics and Structural Change. Armonk, NY: M.E. Sharpe, 2004.

Greene, W. II. Econometric Analysis. Fifth Edition, Upper Saddle River, NJ: Prentice Hall, 2003.

Hirsch, W. Z. "Cost Functions of An Urban Government Service: Refuse Collection." Review of Economics and Statistics, 47(1), 1965, 87-92.

Hodge, G. "Contracting Public Sector Services: A Meta-Analytic Perspective of the Internal Evidence." Australian Journal of Public Administration. 57(4), 1998, 98-110.

Jung, C. "Forms of Government and Spending on Common Municipal Functions: A Longitudinal Approach." International Review of Administrative Sciences. 72(3), 2006, 363-76.

Lichtenberg, F. R., and D. Siegel. "Productivity and Changes in Ownership of Manufacturing Plants." Brookings Papers on Economic Activity, 3, 1987, 643-83 (special issue on Microeconomics).

Martin, S. A., R. McHugh. and S. R. Johnson. The Influence of Location on Productivity: Manufacturing Technology in Rural and Urban Areas. Center for Economic Studies. Census Bureau. Discussion Paper, CES 91, 1991, 10.

Mas-Colell, A., M. D. Whinston, and J. R. Green. Micro-economic Theory. New York: Oxford University Press. 1995.

McDavid, J. "The Canadian Experience with Privatizing Residential Solid Waste Collection Services.'* Public Administration Review, 45, 1985, 602-8.

Nguyen, S. V., and M. L. Streitwieser. "Factor Substitution in U.S. Manufacturing: Does Plant Size Matter?" Small Business Economics. 12, 1999, 41-57.

North Carolina Department of State Treasurer. North Carolina County and Municipal Financial Information. Raleigh, NC. Accessed November 10, 2005, http://www.treasurer.state.nc.us/lgc/units/unitlist's.htm. 2005.

North Carolina League of Municipalities. What Are We Doing With Garbage? Raleigh, NC: North Carolina League of Municipalities. 1997.

North Carolina League of Municipalities. What Are We Doing with Garbage? Raleigh. NC: North Carolina League of Municipalities, 2001.

North Carolina League of Municipalities. What Are We Doing with Garbage? Raleigh, NC: North Carolina League of Municipalities. 2003.

North Carolina League of Municipalities. About Cities and Towns. Raleigh. NC. Accessed June 10, 2008. http://www.nclm.org/about%20cities%20and%201owns/aboul.htm.2008.

Ohlsson, H. "Ownership and Production Costs: Choosing between Public Production and Contracting-Out in the Case of Swedish Refuse Collection." Fiscal Studies, 24, 2003, 451-76.

Pier, W. J., R. B. Vernon and J. H. Wicks. "An Empirical Comparison of Government and Private Production Efficiency." National Tax Journal. 27(4), 1974, 653-56.

Post, J., J. Broekema, and N. Obirih-Opareh. "Trial and Error in Privatization: Experiences in Urban Solid Waste Collection in Accra (Ghana) and Hyderabad (India)." Urban Studies. 40, 2003, 835-52.

Rand California Statistics. MSAs in the State NC. Business and Economic Statistics. Accessed June 15. 2008. http://ca.rand.org/stats/economics/MSA/NC.html.

Reeves, E., and M. Barrow. "The Impact of Contracting Out on the Costs of Refuse Collection Services: The Case of Ireland." Economic and Social Review. 31(2), 2000, 129-50.

Schmit, T. M., and R. N. Boisvert. A Hedonic Approach to Estimating Operation and Maintenance Costs for New York Municipal Water Systems. Department of Agricultural. Resource, and Managerial Economics. Cornell University. Working Paper. 96-12. 1996.

Szymanski, S. "The Impact of Compulsory Competitive Tendering on Refuse Collection Services." Fiscal Studies. 17(3), 1996, 1-19.

Szymanski, S., and Wilkins. S. "Cheap Rubbish? Competitive Tendering and Contracting Out in Refuse Collection-1981-1988." Fiscal Studies, 14(3), 1993, 109-30.

University of North Carolina School of Government. Forms of Government and Methods of Election in North Carolina Cities. Chapel Hill. NC: The University of North Carolina. Accessed July 12, 2008. http://www.sog.unc.edu/pubs/FOG/index.php.2008.

Vickers, J., and G. Yarrow. "Economic Perspectives on Privatization." Journal of Economic Perspectives. 5(2), 1991, 111-32.

Walls, M., M. Macauley, and S. Anderson. "Private Markets, Contracts, and Government Provision: What Explains the Organization of Local Waste and Recycling Markets?" Urban Affairs Review. 40(5), 2005, 590-613.

Zellner, A. "An Efficient Method of Estimating Seemingly Unrelated Regressions and Tests for Aggregation Bias." Journal of the American Statistical Association. 57, 1962, 348-68.

ABBREVIATIONS

SUR: Seemingly Unrelated Regression

TFP: Total Factor Productivity

NCLM: North Carolina League of Municipalities

MSA: Metropolitan Statistical Area

DBD: Downtown Business District

(1.) A referee made good suggestions about the distinction between "provision" and "delivery" (or "production") in municipal solid waste services. According to Bel and Warner (2008) and Dubin and Navarro (1988), public provision means that a municipality is responsible for the existence and provision of municipal solid waste services. In public provision, service delivery is made by either municipalities or private companies. Private provision means that a private company takes responsibility for the existence and provision of municipal solid waste services without government intervention.

(2.) After privatization, it is still possible that private companies operate in noncompetitive markets and thus enjoy considerable market power (Bel and Miralles, 2003).

(3.) According to Post. Broekema, and Obirih-Opareh (2003). private contractors reduce costs of solid waste collection through lowering wages and using old and unsophisticated equipment.

(4.) In case studies on Accra in Ghana and Hyderabad in India. Post. Broekema, and Obirih-Opareh (2003) show-that transaction costs arising from contracting out could be substantial. Batley (1996) also points out that the occurrence of transaction costs arising from contracting out is not generally accounted for in discussions of net effects.

(5.) Let K be a non-empty closed set, and let [[mu].sub.k](*) be its support function. According to the duality theorem, if K is strictly convex, then the minimizing vector is unique at [bar.p]. and the support function [[mu].sub.k](*) is concave and differentiable at [bar.p]. If K is convex hut not strictly convex, then the function [[mu].sub.k](*) at [bar.p] is still equal to the minimizing set. which is not unique but multivalued (Mas-Colell. Whinston, and Green, 1995, pp. 66-67).

(6.) The author estimated the unrestricted model without any implicit assumptions and conducted tests on the assumptions. In general, the data support these assumptions. These are available upon request.

(7.) When either the share equation of private capital or the share equation of labor is deleted, there is no difference in empirical results.

(8.) The Allen-Uzawa partial elasticity of substitution is a popular measure in analyzing substitutability among input factors. It is written as: [[sigma].sub.ii] = [[c.sub.ii]/[S.sub.i]] and [[sigma].sub.ij] = [[c.sub.ij]/[S.sub.i]] (i [not equal to] j) for i. j = L. K. As Blackorby and Russell (1989) point out. however, the Morishima elasticity of substitution is a better measure than the Allen elasticity of substitution in measuring substitution relationships among inputs, especially because the Morishima elasticity (1) allows natural asymmetry of substitution relationships among input factors. (2) measures case of substitution among input factors, and (3) provides complete information on factor shares.

(9.) Two hundred and thirty of the 400 cities and towns (57.5%). 179 of the 413 cities and towns (43.3%). and 217 of the 413 cities and towns (52.5%) in North Carolina responded to the 1997. 2001. and 2003 surveys, respectively (NCLM, 1997. 2001. and 2003). Missing observations arc excluded, so 252 observations are used for empirical analysis (84 municipalities in 1997, 78 in 2001, and 90 in 2003).

(10.) All continuous variables measured in dollars are reported in constant 2003 dollars.

(11.) An incorporated municipality means that the General Assembly (or in a few cases, a former state agency, the Municipal Board of Control) has granted a charter authorizing the establishment of a municipal corporation, along with the powers, authority, and responsibilities of the municipal government (Fleer, 1994, pp. 198-216; NCLM Web site).

(12.) As a referee suggests, single-country studies in countries like the Netherlands and Sweden arc similar to single-state studies using municipal data in one state in the United States because of the uniformity of regulations (e.g., Dijkgraaf and Gradus, 2003, 2007; Ohlsson, 2003; see Bel and Warner. 2008. for more details).

(13.) According to Bel and Warner (2008), among the 18 empirical studies on solid waste collection that they reviewed, 9 studies employed different measures of average cost as dependent variable, while 8 studies and 1 study employed total costs and output, respectively. Different measures of average cost include cost per ton collected (Hirsch, 1965; Ohlsson, 2003), cost per collection unit and cost per employee (Szymanski and Wilkins, 1993), cost per household (Szymanski, 1996), and cost per yard of garbage collected (Dubin and Navarro. 1988).

(14.) In constructing hourly wage per solid waste worker, this paper assumes that solid waste workers have the same average length of working hours with either public entities or private contractors. But, private contractors may hire fewer employees who work longer hours than employees at public entities. In that case, this assumption may bias the empirical results.

(15.) Hourly wage per solid waste worker ($12.85) was 12% higher in municipalities that are part of MSAs than that ($11.48) in municipalities that are not.

(16.) In addition, as homogeneity in input prices and symmetry of the second-order terms are imposed, the estimated cost functions, along with their share equations. satisfy all regularity conditions of a theoretically valid cost model.

(17.) The estimated coefficient is interpreted as the cost of municipal solid waste services per collection at a residential site under municipal ownership of an electric system as a proportion of cost per collection at a residential site not under municipal ownership, given by [e.sup.-[beta]] (Reeves and Barrow, 2000). For example. [beta] = 0.231 in Model 2. thus [e.sup.-0.231] = 0.79, indicating 21% lower not under municipal ownership than under municipal ownership.

(18.) Although institutional arrangements in the delivery of residential solid waste collection services have no significant effect on cost savings, the comparisons in elasticities of substitution across the two comparison groups (i.e., municipalities using public delivery and those using private contractors) may provide some interesting patterns on the substitutability between two inputs (i.e., labor and private capital).

(19.) Although there is no significant difference in cost savings between public delivery and private contractors, the comparisons in efficiency gains and productivity across the two comparison groups may provide some interesting patterns on efficiency gains and productivity between the two institutional arrangements as well.

* The author would like to thank Michael I. Luger, Robert H. Wilson, and anonymous referees for their helpful comments and suggestions on earlier drafts.

SUHO BAE *

Bae: Assistant Professor. Graduate School of Governance. Sungkyunkwan University. 53, Myeongnyun-dong 3-ga, Jongno-gu, Seoul 110-745, Republic Korea. Phone 82-2-760-0443. Fax 82-2-766-8856, E-mail: baes@skku.edu

doi:10.1111/j.1465-7287.2009.00180.x
TABLE 1
Descriptive Results

Variable                           All Sample    Public     Private
                                      Mean      Delivery   Contractor
                                   (Standard      Mean        Mean
                                   Deviation)  (Standard   (Standard
                                               Deviation)  Deviation)

Average cost of solid waste per       3.179       3.056       3.779
collection (C)                       (2,931)     (1.648)     (6.119)

Total amounts of solid waste         26,748      29,919      11,336
collected, recycled, and disposed   (64,210)    (69.824)    (14.174)
(tons) (Y)

Hourly wage ([P.sub.w])              12.242      12.611      10.449
                                     (6.489)     (5.596)     (9.623)

User cost of capital ([U.sub.k])      0.168       0.164       0.184
                                     (0.122)     (0.121)     (0.124)

Municipal implementation of           0.644       0.686       0.440
recycling program                    (0.475)     (0.463)     (0.486)

Municipal operation of sanitary       0.174       0.170       0.194
landfill facility                    (0.376)     (0.374)     (0.392)

Municipal delivery of special         0.396       0.421       0.274
services to DBDs                     (0.488)     (0.495)     (0.439)

Round-trip to landfill facility      25.076      24.565      27.562
(mil)                               (31.303)    (32.034)    (27.682)

Municipal ownership of electric       0.302       0.311       0.256
system                               (0.460)     (0.464)     (0.441)

Yard waste composted and mulched     61,722      63.416      53.488
(%)                                 (44.427)    (44.021)    (45.988)

Council-manager government form       0.833       0.871       0.651
                                     (0.373)     (0.336)     (0.482)

Fiscal stringency                     0.053       0.056       0.040
                                     (0.053)     (0.054)     (0.043)

Household income                     38.408      37.765      41.534
                                    (11.405)    (10.689)    (14.125)

MSA                                   0.556       0.569       0.488
                                     (0.498)     (0.496)     (0.506)

Cost share of labor                   0.423       0.457       0.253
                                     (0.273)     (0.238)     (0.360)

N                                      252         209         43


TABLE 2
Empirical Results: Translog Cost Function

                              Model 1      Model 2      Model 3

DV : InC                      Estimate     Estimate     Estimate
                             (Standard    (Standard    (Standard
                               Error)       Error)       Error)

Constant                      3.316        3.163        3.430
                             (2.310)      (2.405)      (2.401)

In Y                         -0.227       -0.233       -0.269
                             (0.166)      (0.166)      (0.167)

In [P.sub.w]                  0.370        0.368        0.370
                             (0.094) ***  (0.095) ***  (0.095) ***

In [U.sub.k]                  0.630        0.632        0.630
                             (0.094) ***  (0.095) ***  (0.095) ***

1/2[(In Y).sup.2]             0.044        0.044        0.050
                             (0.019) **   (0.019) **   (0.019) ***

1/2[(In [P.sub.w]).sup.2]     0.083        0.084        0.083
                             (0.016) ***  (0.016) ***  (0.016) ***

1/2[(Inc [U.sub.k]).sup.2]    0.083        0.084        0.083
                             (0.016) ***  (0.016) ***  (0.016) ***

(In Y)(In [P.sub.w])         -0.033       -0.033       -0.033
                             (0.009) ***  (0.009) ***  (0.009) ***

(In Y)(In [U.sub.k])          0.033        0.033        0.033
                             (0.009) ***  (0.009)***   (0.009)***

(In [P.sub.w])(In            -0.083       -0.084       -0.083
[U.sub.k])                   (0.016) ***  (0.016) ***  (0.016) ***

Municipal delivery of         0.138        0.114        0.133
residential solid waste      (0.135)      (0.143)      (0.143)
services

Municipal implementation of   0.119        0.122        0.180
recycling program            (0.108)      (0.110)      (0.116)

Municipal operation of       -0.098       -0.078        0.123
sanitary landfill facility   (0.131)      (0.139)      (0.191)

(Municipal delivery of                                 -0.421
residential solid waste                                (0.275)
services) x (municipal
implementation of recycling
program) x (municipal
operation of landfill
facility)

Municipal delivery of         0.104        0.092        0.102
special services to DBDs     (0.108)      (0.110)      (0.109)

In(round-trip miles to       -0.033       -0.035       -0.038
landfills)                   (0.036)      (0.037)      (0.037)

Municipal ownership of        0.255        0.231        0.242
electric system              (0.112) **   (0.120) *    (0.119) **

Yard wastes composted and     0.0002       0.0001       0.0001
mulched                      (0.001)      (0.001)      (0.001)

Fiscal stringency            -2.203       -1.967       -2.046
                             (1.001) **   (1.057) *    (1.054) *

In (household income)        -0.160       -0.145       -0.165
                             (0.208)      (0.217)      (0.216)

MSA                          -0.073       -0.074       -0.082
                             (0.112)      (0.123)      (0.122)

Council-manager government    0.359        0.384        0.376
form                         (0.151) **   (0.154) **   (0.153) **

2001                                       0.094        0.084
                                          (0.120)      (0.120)

2003                                       0.048        0.028
                                          (0.116)      (0.116)

Northeastern region                        0.117        0.107
                                          (0.214)      (0.213)

Eastern region                             0.090        0.065
                                          (0.193)      (0.193)

Research triangle region                  -0.088       -0.082
                                          (0.193)      (0.192)

Southeastern region                       -0.122       -0.141
                                          (0.185)      (0.185)

Piedmont triad region                      0.030        0.050
                                          (0.181)      (0.180)

Charlotte region                           0.047        0.046
                                          (0.182)      (0.181)

N                                252          252          252

[R.sup.2] (In C)              0.0048       0.0057       0.0141

[R.sup.2] (labor)             0.0451       0.0441       0.0448

Note: *. **, and *** significant at the 10%, 5%, and 1% levels,
respectively.


TABLE 3
Elasticities of Substitution

             All Sample       Public Delivery   Private Contractor

           Price  Morishima  Price   Morishima  Price   Morishima

Labor,    -0.380             -0.360             -0.417
Labor

Capital,  -0.278             -0.303             -0.142
Capital

Capital,   0.278    0.657     0.303    0.663     0.142    0.558
Labor

Labor,     0.380    0.657     0.360    0.663     0.417    0.558
Capital

Note: Elasticities calculated at sample means.


TABLE 4
Total Factor Productivity Level and Efficiency Gains

                  All Sample   Public   Private Contractor
                              Delivery

TFP level            0.0314   -0.0100          0.2325
Efficiency gains    45.63%    48.80%          30.23%
Gale Copyright:
Copyright 2010 Gale, Cengage Learning. All rights reserved.