Bel, Fageda, and Warner conduct a meta-regression analysis to empirically test if there is systematic support for the claim that private provision of water distribution and solid waste collection services achieves lower overall costs than public provision.

The authors introduce a diverse range of existing theoretical and empirical literature spanning the fields of public choice, property rights, transaction costs, and industrial organization. Their review of public choice and property rights literature examines the differences between public ownership (typically monopolistic) and private ownership (typically competitive), focusing on the varying incentives between these two market structures. Generally, public ownership is characterized by incentives such as political interests and public support rather than profit. Thus, the benefits from innovations and cost reduction are often marginalized because they do not accrue directly to politicians, often leading to excess supply and/or inefficiency. Conversely, private ownership is expected to provide incentives for innovation and cost reduction (albeit often with lower quality of service) since the existence of private property rights induces profit maximization behavior. However, these incentives are largely dependent on the existence of a competitive market structure which is not always the case in public service provision.

Transaction costs and industrial organization literature focus on market structure conditions and the principal-agent problem. According to transaction costs literature, the choice to shift to private production is determined by estimated transaction costs, costs of contracting out, and monitoring costs. Clearly, each of these costs vary with the type of service and prevalent market structure conditions. Industrial organization literature emphasizes differences in aligning managerial actions and ownership objectives between private and public ownership. However, the extent of this alignment, once again, may be determined by market structure conditions and contractual frameworks.

As a whole, there are mixed conclusions on whether private ownership results in reduced costs and recent papers indicate that there are essentially no differences in costs between private and public provision. This is primarily due to a lack of competitive market structures, high transaction and contracting costs, and a general improvement in efficiency of publicly provided services.

Following the literature review, the authors describe their meta-regression analysis, which uses data collected from other empirical studies as the sample set. In this case, the set consists of 27 empirical studies which compare the costs of public and private production while controlling for comparable variables. There are two major goals of the meta-regression analysis: (i) studying the impacts of sample size, time period, geographical area, service characteristics, type of panel data and functional form of regression equations used in each empirical study, and (ii) analyzing whether cost impacts of privatization are prone to publication biases. The inherent advantage of such an analysis is its ability to study the impacts of several variables more robustly using a wide range of individual empirical studies.

From the regression results, the authors conclude that there is no clear evidence of significant cost differences between private and public provision across the range of studies (high dispersion around the mean t-statistic of -1.49). Moreover, results from certain variables suggest that studies conducted in recent years are less likely to find differences in public and private production costs. Additionally, lower costs under private production were more likely for solid waste collection than for water distribution. This could be due to higher scale economies naturally present in solid waste collection as well as the lack of competition and high transaction costs common to water distribution. For example, higher transaction and quality-monitoring costs in water distribution services may erode potential cost savings from private provision. Finally, it seems that there may be evidence of occasional publication bias; papers that have statistically significant results that establish cost savings are more likely to be published.

In conclusion, privatization, in itself, may not result in systematic cost reductions. More emphasis may be laid upon factors such as the kind of service, market structure dynamics, transaction and monitoring costs, as well as the policy environment and regulatory framework. These factors may be more critical than the privatization process on its own, which only changes ownership status and does not inherently imply cost reductions.