Report | Economic Growth

Privatization and the Market Role of Local Government

Briefing Paper #112

Download PDF

Privatization and the market role of local government

Small growth in contracting underscores dominance of service provision by public employees

by Mildred Warner and Amir Hefetz

In the early 1990s, Chautauqua County, N.Y. was pulling out of an economic slump. Keen to maintain services while reducing taxes and building government reserves, the county executive aggressively restructured government services using a mix of strategies – sometimes privatizing services, but at other times bringing previously privatized work back in house. The county also cooperated with other local governments to achieve economies of scale, and introduced performance-based incentives to promote efficiency and investment in new technologies within county departments. By 1997 the county funds were up, the tax rate had fallen, and services had expanded. In the end, the county’s officials decided that, barring a few exceptions, neither the private nor public sector offered an inherent advantage, so they found a middle path between those who stress the advantages of privatization and those who advocate for the direct provision of services by government. (For more details on Chautauqua County’s experiences, see the case study in Appendix C.)

Successful local governments such as Chautauqua County’s recognize privatization as one of a number of available tools for restructuring their operations. Other such tools include inter-municipal cooperation, performance-based incentives, and cooperative labor-management relations. All of these options can help governments enhance economic and social efficiency. While privatization represents the most widely pursued alternative to public service delivery, local governments use a wide set of restructuring strategies in order to address a variety of market conditions and a broad range of social concerns.

Local governments often find themselves to be the sole buyer of a given type of service, and, as such, must give careful attention to their role in the market. When it comes to interacting with the marketplace, governments can play many roles, acting as regulator, provider, and contract manager. Sometimes governments even create a market simply by their need to deliver a certain kind of service. In these cases governments must ensure competition, acceptable costs, adequate service quality, and stability in service provision.

Our analysis of national longitudinal survey data from the International City/County Management Association (ICMA) for 1982 to 1997 shows that:

  • Although local governments employ a broad array of restructuring alternatives, service delivery by public employees remains the dominant form of provision, hovering around an average of 60% when all services are taken into account.
  • Despite the increasing political acceptance of private forms of governmental service delivery, privatization to for-profit and non-profit firms has risen only slightly, from 22% in 1982 to 24% in 1997, an increase that has not been steady.
  • The next most common alternative form of service delivery-inter-municipal cooperation, which achieves similar economies of scale but keeps work in the public sector-also has risen slightly, from 12% in 1982 to 15% in 1997.
  • Ninety percent of all public service restructuring takes one of three forms-privatization to for-profit companies, privatization to non-profits, and inter-municipal cooperation.

Data for this study are based on ICMA surveys of alternative service delivery arrangements of local governments. ICMA surveys have been conducted every five years since 1982 and include all U.S. counties with populations over 25,000 and all cities with populations over 10,000. The ICMA surveys provide a very comprehensive view of the complex mix of services offered by local governments and the array of delivery alternatives employed (see Appendix Table A-2). Conducted in 1982, 1988, 1992, and 1997, the surveys measure the form of service delivery for 64 different services in seven broad areas: public works and transportation, public utilities, public safety, health and human services, parks and recreation, culture and art, and support functions. Forms of contracting out measured by the ICMA surveys include: contracts with other governments (inter-municipal cooperation); contracts with for-profit or non-profit providers; and use of subsidies, franchises, or volunteers. The surveys also measure local government managers’ reasons for considering alternative forms of service delivery. The stability in survey design and large sample size (over 1,500 municipalities) make the ICMA surveys an excellent source of data for comparing governmental restructuring over time.

The persistence of public service delivery
Why does traditional provision of services by government employees persist as the dominant form of service delivery? Past theorizing offers two possible answers: government failure, or quasi-market failure. Government failure may explain why privatization doesn’t occur even when it might be appropriate. Such failure may take the form of bureaucratic concerns over loss of control or power (Niskanen 1971). Some studies also suggest that traditional forms of public service provision may persist because of civil service employees’ opposition to privatization (McGuire et al. 1987; Savas 2000). Support for this view in our findings, however, is limited.

In theory, privatization attempts to address government failure by separating service provision from service delivery, potentially giving the citizen/consumer more choice. When local governments outsource service delivery, the government remains responsible for the finance and provision of the service, but production and delivery are carried out by private firms or non-profit organizations. These contracts, though, are not always forged in truly competitive markets, a situation that can lead to incorrect service specification-where cost savings come at the expense of service quality-or to a loss of community values (e.g., when privacy is sacrificed in order to contract out tax collection) (Lowery 1998).

Local governments employ several strategies to minimize such problems. They may resort to different levels of contracting in different service areas. They may mix public and private provision in the same service area. Or they may reclaim responsibility for service provision in areas that were formerly contracted out. The availability of these complex restructuring strategies may explain why there has been limited growth in overall privatization trends.

Our research challenges traditional views on service restructuring as a black-and-white choice between only public or private provision, and we provide both a theoretical and empirical basis for understanding the complexity of local government restructuring behavior. Using national statistics, we address the full range of restructuring alternatives and the stability of those choices. We complement the data on trends with examples from the case of Chautauqua County, N.Y. to illustrate the complexity of local government restructuring.

Furthermore, our research shows that ideology does not dictate local service delivery decisions. Rather, as other research has also shown, pragmatic local governments explore a range of options and recognize the need to structure their market for the sake of economic efficiency, service quality, and accountability to community values (Warner and Hebdon 2001).

Why did privatization increase so little?
The privatization movement stemmed from a recognition that the political choices about service provision can be separated from the actual mechanics of service delivery (Boyne 1996; Oakerson 1987). Privatization is advocated as a solution to problems of governmen
t failure. Government monopoly of production, bureaucratic unresponsiveness to citizen needs, and budget-maximizing public officials are typical examples of the kinds of government failure that privatization claims to remedy (Eggers and O’Leary 1995; Savas 2000). The benefits of both public-sector engagement and market discipline can be achieved through quasi-markets that separate provision from production through contracts.

But contracts and quasi-markets also create failures of their own. Lowery (1998) discusses three types of quasi-market failure: failure of market formation, preference error, and preference substitution. The first of these, failure of market formation, results from a lack of competition, which is often due to the limited number of suppliers that can actually provide many public services. If privatization merely substitutes a private monopoly for a public one, then savings will likely disappear after the initial contract. Another kind of market failure is caused by lack of full information and high transaction costs in contract specification and monitoring. Difficulty in contract specification and monitoring can result in service delivery that does not match public preferences. A third form of market failure results when citizen preferences are not in line with collective preference as expressed through government provision. When decisions regarding policy provision are separated from the delivery of the service, “citizen consumers” may exercise choice based on factors such as race or income rather than efficiency or quality, the latter of which government providers intend to be the motivators of citizen choice.

Is the lack of growth in privatization due to government failure?
Theory would suggest that the same processes that create government inefficiency, bureaucratic unresponsiveness, government monopoly production, and principal agent problems (Eggers and O’Leary 1995; Savas 2000) also cause governmental reluctance to privatize. However, published research and the ICMA data provide limited support for this view.

Contracting is not new. One reason why there may be so little growth in privatization is that contracting out is not a new phenomenon for local governments. Many services were contracted out in the 19th century and only brought in house as part of the “good government” movement launched to stem problems of corruption (Adler 1999). Henig (1989) argues that contracting by local governments has been a long-standing practice, and that naming these activities “privatization” was part of an effort by public-choice proponents to show empirical support for their propositions. Thus, the trend shown in Figure 1 is flat because the baseline is 1982, when contracting out was already common.

 Figure 1

Fiscal stress was declining. The fact that fiscal stress was declining over much of this period may explain some of the lack of growth in privatization. Pressure to restructure local government service delivery is driven, in part, by concerns with fiscal stress and the notion that private providers, who are subject to the rigors of market competition, will be more efficient than government. So if privatization offers efficiency gains, we might expect to see higher levels of privatization during recessionary periods (e.g., 1982) or times of local government fiscal stress (e.g., 1992). Indeed, the up and down pattern of the privatization trends does appear to mirror the overall economy’s cycles. Beginning in 1988, ICMA tracked local government concerns with fiscal stress, tax pressure, and cost control. However, with the exception of fiscal stress, which rose in 1992, the percentage of governments reporting these factors as important in their decision to restructure shows a steady decline over the period (see Table 1).

Table 1: Factors affecting local government restructuring 

Opposition to privatization is stable or falling. The lack of growth in privatization also can’t be explained by a strong opposition to the practice. Some argue that privatization reflects a desire for a “smaller” government. While this seems to be the general public’s view of privatization, internal opposition to government contracting would be expected from line employees, department heads, and elected officials. But with the exception of line employees, who have the most to lose from privatization, the ICMA data show opposition from department heads and elected officials is declining or stable. Although the percentage of government managers who cite a “change in political climate emphasizing a decreased role for government” as a factor in their decision to restructure has almost doubled since 1988, it still accounts for less than a fifth of all respondents. This may reflect greater public trust in local government and a recognition of the essential nature of the services it provides (Conlan 1998).

Managerial capacity is rising. Lack of managerial capacity does not appear to explain the limited increase in privatization. The restructuring that must occur in privatization requires experience and leadership on the part of local government officials (Bartle et al. 1996; Boyne 1998). It also requires capacity and willingness to take risks (Warner and Hebdon 2001). Almost 70% of all the responding governments report that they are studying the feasibility of private delivery, and over 95% report that they currently contract out for at least one service. The number of governments reporting that they lacked staff expertise in contract management dropped by half from 1988 to 1997 to less than 6%. The percentage allowing competitive bidding between public employees and private providers increased by 50%, and there was a similar increase in governments offering programs to minimize the effect on displaced public employees. However, lack of monitoring and low levels of competitive bidding suggest room for managerial improvement.

Another indicator of rising managerial capacity is the increased use of internal competition – performance-based incentives, merit-based pay and bonuses, performance-based budgeting – to encourage departments and work teams to identify opportunities to increase efficiency (Osborne and Gaebler 1992). These efforts are often characterized by organizational flexibility and increased labor-management cooperation (Appelbaum and Batt 1994; U.S. Department of Labor 1996; Martin 1999). When departments are allowed to reinvest savings into improved technology, further efficiency gains result. These internal forms of restructuring address two of the most serious causes of public-sector inefficiency – outdated technology, and rigid organizational management systems. The ICMA surveys of alternative service delivery do not assess internal restructuring. However, a special 1993 survey of total quality management in local government found 55% of responding cities reported using these principles (West et al. 1994). A 1998 ICMA survey of government managers’ support for principles of “reinventing government” (e.g., entrepreneurial government, consumer orientation, internal competition) also showed wide acceptance of these principles (Feldman 1999). Thus, limited growth in privatization may be due to enhanced governmental capacity to achieve service delivery efficiencies internally.

Is the limited growth in contracting due to quasi-market failure?
Public goods are those goods and services for which externalities are high and exclusion of free riders is difficult. Private contracts may help government address important capital, technology, and management constraints. But governments still retain their responsibility for ensuring quality service and addressing the externalitie
s that cause goods and services to be public in the first place (e.g., the public health benefits of frequent garbage collection). Thus, careful attention must be given to the way in which governments structure these contract markets. The sources of quasi-market failure outlined by Lowery (1998) – failure of a market to form, high costs of contracting, and potential loss of community values – are not merely theoretical but are reflected in the ICMA data.

Contracting is a difficult process. What makes service contracting such a difficult process? For services that are easily specified and measured and for which there are many alternate suppliers, privatization should be easy and may yield cost savings (Starr 1987). However, a competitive market of alternative suppliers may not exist in every community. Thin markets of alternative providers can be especially important at the local level (Hirsh 1995) and among rural governments where the benefits of privatization may be short lived (Kodrzyski 1994).

Many local government services are complex in nature, making quality harder to monitor. Even in the private sector the decision to make or buy is a complex one. Just like private firms, governments must carefully assess things like market structure, transaction costs, and the difficulty of contract specification and monitoring before making a decision to contract out (Nelson 1997). For both the private sector and government, timeliness, dependability, and quality often are more important than cost efficiency (Bartle et al. 1996; Pouder 1996). In a study of local government restructuring in New York State, quality concerns were highest among those governments that contracted out more because they no longer held internal control over service delivery (Warner and Hebdon 2001).

In the decision to privatize, governments often fail to include the costs of contracting and monitoring, which can average over 20% of total project costs (Prager 1994; Pack 1989). To overcome difficulties in specifying services and monitoring quality, many local governments develop ongoing, close relations with their contractors. This form of relational contracting may ensure quality, but it also erodes the distance between government and contractor and may lead to rising costs or corruption (Sclar 2000).

The ICMA surveys do not address monitoring costs, but they do ask governments if they monitor service-delivery costs and quality. Although over 95% of governments report some level of contracting, less than half report any monitoring activity at all. Cost was the most commonly monitored item, but it was monitored by only 40% of responding governments. Similarly, contract compliance with service delivery standards was monitored by only 38% of governments. Citizen satisfaction was monitored by less than a third of responding governments. Only 11% of governments reported keeping the service-complaint mechanism in house (see Table 1).

Efficiency gains are hard to secure. Clear evidence of the increased efficiency of private contracts does not exist. Case studies showing increased efficiency under privatization (Savas 2000; Eggers and O’Leary 1995) are counterbalanced by cases showing increased costs (Sclar 2000). Boyne (1998), in his analysis of a range of studies looking at the efficiency of private contracts, finds no clear support for greater efficiency. Other research suggests that the benefits from privatization are overstated (Stein 1990; Ferris 1996). Lack of competition, high costs of contracting and monitoring, and erosion in service quality are commonly cited reasons (Kodrzycki 1994; Prager 1994).

The labor-intensive nature of most public services means that most of the potential savings from privatization are likely to come from wages and benefits or reductions in service quality (Hebdon 1995). Concern about losses in wages and benefits causes many local governments to require that contractors maintain similar wage and benefit levels after privatization (Martin 1999). In these cases, cost savings must come from increased process efficiency or decreased service quality. Because the complex nature of many local government services is hard to specify in a contract, the potential for erosion in service quality is high. This was demonstrated by Sclar (1997) in the case of Albany Fleet maintenance, a case in which costs rose and service quality declined as a result of privatization. Other cases have resulted in improvements, but the key seems to rest in the nature of the market and contract specification (Osborne and Gaebler 1992; Sclar 2000).

Concern about loss of public values. Loss of direct control over service delivery means care must be given to delineate the full range of social benefits included in a service (Lowery 1998). Given the difficulties in contract specification, government contractors must ensure valuable social benefits are not lost once service delivery is contracted (Starr 1987). Some benefits of public provision extend beyond the service itself and include legal rights of redress, freedom of expression, and public access (Sullivan 1987; Moe 1987). In some cases, governmental responsibility remains despite the fact that private parties control service delivery. This enduring responsibility and liability may cause governments to be cautious in relinquishing control over service delivery.

When service provision is separated from delivery, and the “consumer/citizen’s” decision is swayed by factors other than collective well-being, preference substitution can occur (Lowery 1998). One only has to look at the segregated nature of metropolitan areas to understand that “consumer/citizen” decisions about housing and education are often based on race and income rather than efficiency or service quality (Altshuler et al. 1999). Public services provide the means by which local governments build community (Marmolo 1998; Frug 1998), so if contracting causes these broader community benefits to be lost, then an important public good is also sacrificed. Potential efficiency gains through contracting must be weighed against the community-building mission of local government.

Understanding government restructuring behavior
Local governments recognize the potential for both government failure and quasi-market failure, but the evidence suggests that the limited growth in privatization thus far is not primarily due to government failure. Governments must balance concerns over efficiency with concerns over service quality and community values. Further analysis of the ICMA restructuring data shows an active market-structuring role on the part of local governments. These governments employ a range of restructuring options that vary according to service and local market conditions; they mix public and private provision for the same service, and recognize the dynamic nature of local service delivery by shifting among different delivery alternatives over time.

Governments recognize that their role in structuring the market is critical to ensure competition and true efficiency gains from privatization. As the primary funder of the contracted service, local government is in a monopsonist position and must take care to identify the potential market of alternative suppliers and structure contracts to ensure that quality is maintained and efficiency gains secured. This is why we often see mixed public/private provision where part of the service is provided in-house and part by private contract. A local government plays many roles in the market, including that of regulator, contractor, and direct service provider. Depending on the service in question, a local government selectively chooses which role to play, with an eye to ensuring service quality, dependability, efficiency, and community values.

By longitudinally linking the ICMA surveys, we are able to compare the stability of contracts and in-house provision. Some instability reflects changes in service demands, new technologies, and capital-investment ne
eds that benefit from privatization or partnership with the private sector. But part of the instability in contracts also reflects dissatisfaction with contracted services and the need to bring work back in-house to maintain quality. For services on which public health and safety depend, quality and dependability are often more important than efficiency. The flat privatization trends are not due to lack of governmental effort. Rather, they reflect concerns with quasi-market failure. Pragmatic local governments are cautious innovators. (See Appendix C.)

Governments structure markets to ensure competition
Ensuring efficiency and quality in public service delivery requires attention to both public and private markets. Competition can promote efficiency, and increased attention is being given to external forms of competition. To make effective use of external providers, governments must give special attention to the way they structure contracts in local supplier markets. Competitive bidding has become quite common among local governments and is used to ensure contestability in contract markets (Martin 1999; Osborne and Plastrick 1997). Many governments compare internal cost estimates with external bids to create benchmarks for efficient costs. The ICMA surveys show that 16.5% of governments report allowing competitive bidding (Table 1.) This benchmarking process often results in mixed provision of services whereby part of the service or service area may be delivered by a private contractor, and part by public employees. This redundancy in service delivery not only promotes cost efficiency, it also creates competition in quality and ensures service dependability in case of contract failure (Miranda and Lerner 1995). Contestability has been widely used in England and Australia and may be achieved through inter-departmental competition as well. The lack of efficiency gains from privatization is attributed, in part, to contestable markets among public bureaucracies (Stein 1990).

There may be a progression from providing services totally in-house, to mixed public/private provision, and ultimately to total contracting out. The ICMA data show that, in 1982, governments were more likely to use mixed provision, contracting out a part of the service while keeping the remainder in house. Mixed provision accounted for 47% of all restructuring in 1982, while total contracting out accounted for 53%. However, after 16 years of experience with privatization, the ICMA surveys still show a third of all restructuring was in the form of mixed public/private provision in 1997. Of the roughly 40% of services that are restructured, services that are totally contracted out accounted for 70% of the restructuring, while mixed public/private provision accounted for 30% in 1997. Thus, while the level of contracting has risen, the level of mixed public/private provision remains significant.

Not surprisingly, mixed provision is most common in service areas where contracting out is high. It is low only in public utilities, where returns to economies of scale (through natural monopolies) would make redundancy too expensive. Even for the areas in which total contracting out has risen, mixed provision has remained stable. Many governments realize they must remain players in the market to ensure competition, service quality, and accountability to citizens.

The persistence of mixed public/private provision confirms Miranda and Lerner’s (1995) notion of “benchmarking,” a process by which governments intentionally use both public employees and contractors in the same service area to ensure quality and competitiveness. In order to guide provision, governments must have a hand in service delivery, and studies show that this redundancy does not increase costs but rather helps ensure service quality and fair bidding (Miranda and Lerner 1995). Chautauqua County Executive Andrew Goodell describes it this way:

The secret is competition. Whether it is a public or private monopoly, it is still a monopoly and will charge monopoly rents. Local governments must create competition where none exists-even if that means competition between the public and private sectors.

Local government restructuring is a complex process
Given the broad mandate of local government, it is not surprising that a wide range of service delivery alternatives are used. While the most common form of service delivery is public provision with public employees entirely, privatization and inter-municipal cooperation are the most common alternatives.

In the ICMA data, public services are usually provided through the use of public employees except for public works, health and human services, and culture and arts. There has always been a high level of non-profit provision in health and human services and culture and arts – a reflection of the evolution of our social welfare state from non-profit providers. Inter-municipal cooperation and non-profits traditionally have accounted for close to 70% of all service provision in these two areas. The level of contracting in public utilities also has been high traditionally (around 50%), and it has remained at that level. Where the dramatic growth in contracting out has occurred is in public works/transportation, public safety, and support functions – areas where services are more easily specified and where back-office functions (e.g., dispatching, purchasing, etc.) benefit from economies of scale.

Privatization and cooperation are the most popular forms of contracting. Privatization to for-profit and non-profit firms and inter-municipal cooperation are the most common alternatives, together accounting for almost 90% of all restructuring cases. Franchises, subsidies, vouchers, and volunteers account for the remaining 10%. For-profit privatization is the most common alternative and rose from 16% to 19% of all service provision from 1982 to 1997. For-profit privatization is most common in support functions, public works, public utilities, and public safety (Table 2). These services are more easily specified, are more likely to have competitive supplier markets, and, in the case of support functions and public safety, include indirect services (e.g., dispatching and back-office processes) not directly visible to the citizen. This makes for-profit privatization viable.

Table 2: Trends in local government restructuring by service area 

Inter-municipal cooperation is the second most common restructuring alternative. This sort of cooperation accounts for about 15% of service provision. Cooperation is highest in health and human services and culture and arts. These are complex service areas where the social benefit extends beyond the direct-service recipient. Cooperation also is high in the same areas where privatization is high (e.g., public works, public safety, and parks and recreation) because of the benefits of economies of scale. However, governments choosing cooperation over privatization enjoy these scale economies without resorting to private sector provision.

Privatization to the non-profit sector is the third significant restructuring alternative, but it averages only 5% of service provision. Non-profit privatization is most common in health and human services and culture and arts. Non-profits traditionally have provided these services – in fact many of these services began to receive public contracts only after the need overwhelmed non-profit providers and public attitudes shifted toward including these services as public benefits. Although non-profit provision as a service-delivery alternative has not risen, new partnerships between government and non-profit providers create opportunities for improvements in service quality and efficiency.

The level of contracting is falling. Although the level of contracting, when averaged across all places and all
services, is rising slightly (as seen in Figure 1), both the percentage of governments contracting out and the percentage using high levels of contracting is dropping. The increase is due to a few governments that do very high levels of contracting and an increase in contracting in public works and support functions, as shown in Table 2.

Most governments have at least one case of contracting out. In 1992, almost equal numbers of governments used for-profit privatization (91%) and cooperation (88%) for at least one service. By 1997 the percentage of governments contracting out actually dropped. The percentage of governments with at least one case of privatization fell to 85%, and cooperation dropped to 73%. Non-profit privatization was less common and dropped most dramatically from 65% of all governments in 1992 to 52% of all governments in 1997 (Table 3).

Table 3: Level of restructuring by local government 

If a government chooses to restructure service delivery, it is likely to do so in several services. In fact, the majority of governments reported at least one case of privatization (for-profit and non-profit) and of inter-municipal cooperation. However, the level of restructuring by place declined from 1992 to 1997. The percentage of governments restructuring three or more services dropped from 74% in 1992 to 67% for for-profit privatization and to 53% for cooperation by 1997. Non-profit privatization showed a similar drop from 33% of responding governments having three or more cases in 1992 to 25% in 1997.

Contracting out is more likely in suburbs. Public choice theory suggests that suburban areas would be the most favored in the market for public goods and services (Boyne 1996). Service demands will be more homogeneous (given a more homogeneous population), and residential choice (mobility) is higher. Figure 2 shows both privatization to for-profit firms and inter-municipal cooperation are highest among suburban governments. As expected, public provision is highest among metropolitan and independent rural governments. Rural independent governments have the lowest privatization levels (probably due in part to lack of a competitive market of private suppliers), and metropolitan governments have a lower level of inter-municipal cooperation, in part because they are already large enough to enjoy economies of scale (Warner and Hefetz 2001).

Figure 2: Restructuring by metro status, 1997

Cooperation is highest among both suburbs and independent (rural) areas, where smaller sizes make cooperation necessary to gain scale. Metropolitan places where service areas are already large may realize only limited economy-of-scale gains through increased cooperation. These trends may also be, in part, a reflection of the forces driving suburbanization in the first place. The desire to escape the true social costs of provision of urban services may encourage the isolation of high-need, high-cost urban populations (Frug 1998).

Use of alternative forms of service delivery also varies along income lines. Public provision of all services is most common among areas with lower average incomes, while privatization is more likely as community income levels rise (Warner and Hefetz 2001). If cost savings from privatization were large, we might expect lower-income areas to privatize more. Instead, these data suggest there is something in the nature of supplier markets and transactions costs that makes privatization a less compelling option for low-income areas. Cooperation, by contrast, appears to be income-neutral in its effects. Markets create competitive landscapes. It appears that, by bringing markets into public-service delivery, we see a competitive landscape emerging that favors richer, suburban locales.

Restructuring is a dynamic process
Government service delivery is responsive to changes in citizen demand, technology, and local market conditions. New services, service shedding, new contracting out, and the contracting back in of previously privatized services are all captured in the dynamic nature of the restructuring process. While stability in service delivery may reflect a lack of willingness to innovate on the part of government, the instability in contracts may reflect a lack of dependability of private providers. There is a high cost to switching providers, and one important feature of government service is dependability and stability. Interestingly, the stability of alternative forms of service provision has received little attention in the literature (Lopez de Silanes et al. 1997 is an exception). Warner and Hebdon were the first to measure specifically reverse privatization (contracting back in house of services previously privatized) in a survey of local governments in New York State. Township and county officials in New York described privatization as just one side of the coin; they also were willing to bring previously contracted work back in house (Warner 2000; Warner and Hebdon 2001). In the New York survey, contracting back in accounted for 8% of all restructuring cases. However, New York, with its exceptionally high levels of public-sector unionization, may have higher-than-normal levels of contracting back in.

To test whether the New York results were similar to local governments across the nation, we paired ICMA survey responses across survey years to track shifts in forms of service delivery for each responding government. Roughly 40% of responding governments were the same in each pair of years, from 1982 to 1988 (749), 1988 to 1992 (638), and 1992 to 1997 (628). (See Appendix B for a fuller explanation of our methodology for pairing survey responses.)

Almost all governments contract out and contract back in. Table 4 shows the percentage of governments engaged in new contracting, both out and in, for the three paired time periods: 1982-88, 1988-92, and 1992-97. Almost all governments engaged in new contracting out (96% in 1992-97 period). New contracting out includes any movement toward more contracting (see Appendix Table B-1). The percentage of governments engaging in at least three cases of new contracting out rose from 74% in the first period to around 85% for the later two time periods. On average across all governments, 7.9 services were newly contracted out between 1992 and 1997. That the level of new contracting out could be so high when the overall level of contracting is relatively stable suggests that new contracts are balanced against contracts being brought back in house.

Table 4: Level of new contracting-out and contracting back in by local governments 

Indeed, most governments that contract out also contract back in. From 1992 to 1997 88% of governments had contracted back in at least one service, and 65% had contracted back in more than three services. On average across all places, five services were contracted back in from 1992 to 1997 (more than half the level of new contracting out reported above). Although the level and percentage of governments contracting back in dropped over the three time periods, these data show that dissatisfaction with contracting out is a reality and causes some governments to bring services back in house.

Provision entirely via public employees is the most stable. To determine which forms of service delivery were the most stable, we compared public, cooperative, and for-profit service provision during the most recent period for which data are available (1992-97). Figures 3 and 4 clearly show that delivery of services entirely by public employees is the most common and the most stable fo
rm of provision – more than two-and-a-half times as stable as either alternative. On average, over half of all services provided by public employees were still provided that way in 1997. By contrast, only one-fifth of services provided by cooperation or privatization in 1992 were still provided that way in 1997.

Figure 3: Stability of in-house service provision, 1992-97 


Figure 4: Stability of service delivery alternatives, 1992-97 


Changes in service delivery may be due to the addition of new services, service shedding, shifts toward alternative forms of delivery, or contracting back in services that were previously contracted out. There are important differences in the movement patterns for each alternative form of service delivery. While public provision is the most stable form of service delivery, the largest movements within public provision (22.6%) are toward some alternative (either cooperation or privatization), which is a reflection of the restructuring process. Service shedding from public provision is quite low (6.2%). Within the privatization alternative, the largest movements are toward privatization (29.0%). However, movements from privatization toward other forms (about 23% for either cooperation or public provision) and service shedding from for-profit privatization (16.1%) are also quite high. This may explain why the overall level of privatization has barely risen despite high rates of new privatization.

When looking specifically at cooperation, there is a balance between movement toward cooperation (15.9%), movement away from cooperation (18.8%), and use of cooperation for the provision of new services (13.5%). But cooperation has the highest levels of service shedding (33.0%) of any of the three forms of service provision. Some of the service shedding may reflect new regional forms of provision, with the responsibility for a service assumed by another government or a special district or authority, thus relieving the local municipality of the responsibility. The ICMA data, however, do not distinguish among types of inter-municipal cooperation.

We can see from these figures that both service shedding and new service provision are highest in the privatization and cooperation alternatives. Part of the instability of these alternatives is due to their flexibility in enabling governments to meet changes in citizen demand for services or changes in local market conditions.

Why are contracts unstable? Stability in form of service provision is important. The significance of contracting back in what had once been a privatized service has been given little attention in the theoretical or empirical literature. More research is needed into the nature of contracting and the reasons for instability in alternative service delivery. Absent a new national survey directly measuring the trends in contracting back in, we cannot know why it occurs nor distinguish competitive bidding from failed contracts. At our request, the National Association of Counties added three questions about the stability of private contracts to its spring 2000 survey of governmental operations. Of the 144 county respondents, 14% (20) reported bringing a previously contracted service back in house during the last three years. Not surprisingly, the highest levels of contracting back in were for services where contracting out was also high: data processing, garbage collection and disposal, building maintenance, and emergency medical service. Respondents indicated factors important in their decision to bring service delivery back in house. Half reported problems with the service provider and economic concerns. A quarter reported problems with contracting and monitoring, and another quarter listed management and labor issues. Only 10% of respondents listed political concerns as a factor (NACO 2000).

Additional evidence from 22 case studies of contracting back in (drawn from the ICMA sample) shows that most governments contract back in because of dissatisfaction with contracts (Ballard and Warner 2000). Difficulty and costs of monitoring, problems with service quality, cost and lack of flexibility in contracting, and failure to realize anticipated savings were commonly cited as reasons for contracting back in. Not unlike the contracting decision in the private sector (to make or to buy), public officials find the need for flexibility, responsiveness to changing citizen concerns, and service quality to be crucial determinants (Nelson 1997). Even simple services that are logical candidates for market-based provision may have important public attributes that are lost when provided via contracts. Decisions to bring work back in house are also based on increased efficiency in public sector provision and are often accompanied by cooperation between labor and management. Thus, internal process improvements provide an important and viable alternative to privatization, providing similar efficiency gains with fewer downside risks (Ballard and Warner 2000).

This research provides evidence that the limited growth in privatization over the past 16 years is not due to government failure. Local government service-delivery restructuring is a complex process that requires keen attention to the relative benefits of both internal production and market-based contracting. The complex mix of restructuring alternatives and new evidence on the stability of service delivery alternatives illustrate the active role of local governments in structuring markets to ensure economic and social efficiency. Far from a retreat of government, increased use of market-based provision requires an expansion of government into the administration of markets both public and private (Blanchard et al. 1998). Pragmatic local officials recognize the complexity of the restructuring process and the importance of ensuring service quality, efficiency, and the preservation of community values. Given the difficulties inherent in contracting and monitoring, it is not surprising that the majority of local government services are still provided by public employees. These trends suggest that the scope for privatization may be more limited than some advocates believe.

While the overall trends show a slight drop in in-house public provision, and slight increases in privatization and inter-municipal cooperation, they mask the significant movement between these categories over the 16-year period under study (1982-97). Most governments that contract out also contract back in. The major restructuring alternatives – privatization and cooperation – are both shown to be highly unstable. This reflects the flexible way contracts are used by local governments to ensure economic and social efficiency. Governments can meet citizen demands for new services first through contracts and later by developing in-house delivery capacity. This is an example of market success. However, instability in contracts also reflects inadequate monitoring and government dissatisfaction with the contracting process. This reflects market failure. For the majority of services, the most stable mode of delivery continues to be provision of services by public employees.

In the end, the evidence suggests that local governments recognize the need for flexibility in determining how services are delivered. Privatization would seem to be an important, but limited, tool. To be effective, privatization requires that local governments increase their role in structuring the market, which they accomplish by giving special attention to contract specification, monitoring, and the nature of competition in supplier markets. For many local government services, conditions for efficient market provision are not present, hence public provision by public e
mployees remains the preferred service delivery option. Even when privatization is possible, governments often must remain providers in the market to ensure competitive contracts and continued cost savings. Osborne and Gaebler, in their popular book, Reinventing Government (1992), admonished local government to “steer rather than row.” Our research shows that local governments do what any good captain would, they steer and row. To adequately steer the service delivery process and ensure efficient market functioning, governments maintain direct involvement in service delivery.

–December 2001

Mildred Warner is an assistant professor and Amir Hefetz is a research specialist in the Department of City and Regional Planning at Cornell University.

See related work on Public Investment | Economic Growth

See more work by Amir Hefetz and Mildred Warner