Governments use public–private partnerships (PPPs) as their agents to finance, design, build, maintain and operate their public infrastructure. Despite wide use, many PPPs have produced poor outcomes, including large transaction costs, renegotiations and bankruptcies. Society delegates the authority to build and operate public infrastructure to governments, which must then choose the means of provision. The alternatives are either government-financed design-build contracting, followed by government operation and maintenance—traditional procurement (TP)—or a PPP. We examine this choice using principal–agent and political economy theories. We evaluate the performance of PPPs versus TP against the normative goal of social welfare (economic efficiency). As well, in a review of the empirical literature through 2022, we find no convincing evidence that PPPs provide superior social welfare, nor evidence that many projects been evaluated on this basis. Governments’ continued preference for PPPs in many cases is best explained by political goals and political economy theory. A review of recent empirical evidence supports the view that political economy variables contribute to PPP adoption.

政府利用公私合作伙伴关系（PPP）作为其融资、设计、建设、维护和运营其公共基础设施的代理人。尽管广泛使用，许多 PPP 却产生了不良结果，包括巨额交易成本、重新谈判和破产。社会将建设和运营公共基础设施的权力委托给政府，然后政府必须选择提供方式。替代方案是政府资助的设计建造承包，然后是政府运营和维护（传统采购（TP））或 PPP。我们使用委托代理和政治经济学理论来研究这种选择。我们根据社会福利（经济效率）的规范目标来评估 PPP 与 TP 的绩效。同样，在对截至 2022 年的实证文献的回顾中，我们没有发现令人信服的证据表明 PPP 提供了优越的社会福利，也没有证据表明许多项目是在此基础上进行评估的。在许多情况下，政府对 PPP 的持续偏好可以通过政治目标和政治经济学理论得到最好的解释。对最近经验证据的回顾支持了政治经济变量有助于 PPP 采用的观点。