探索促进更高程度共同生产的协同因素组合效应

在 20 世纪 70 年代后期，Elinor Ostrom 及其同事提出了公共部门的共同生产概念，强调公民如何与服务提供者共同参与服务的交付，尤其是在执法领域。自此以后，共同生产的理解不断扩展，涵盖了志愿者、非政府组织和社区组织等多方利益相关者。现代共同生产定义为服务提供者（包括公共、私人或第三部门）与用户或志愿者在整个服务交付生命周期中（从委托到评估）进行合作的过程。共同生产突出了共同责任，模糊了传统的角色分工，有别于更广泛的公民参与概念。

共同生产与协同治理密切相关，后者强调多方利益相关者的参与、跨部门合作和以共识为基础的决策。Ansell 和 Gash（2008）以及 Emerson 等人（2012）提出的协同治理框架中，激励、领导力和制度设计等要素对于成功的共同生产也至关重要。公平、信任和参与等核心价值观支撑着这两种方法，加强了共同生产和协同治理在提升公共价值、利用多元资源促进社会利益方面的联系。

关于共同生产的研究指出了公民参与、领导力、组织文化和沟通动态等因素的重要影响。然而，现有研究通常孤立地对待这些要素，忽略了它们的协同作用。这一不足因案例研究和定性方法的主导地位而加剧，从而限制了研究成果的普适性。尽管一些定量分析已逐步出现，但这些研究很少考虑多因素的联合效应，显示出更综合的研究方法需求。

本研究填补了这一空白，通过考察协同因素如何互动以增强共同生产效果。研究采用模糊集定性比较分析法（fsQCA），基于协同治理案例数据库中的共同生产案例，识别出协同因素的有效组合。研究结果提供了关于共同生产成功机制的深入理解，具有理论和实践意义，并为未来研究提供了方向。

In the late 1970s, Elinor Ostrom and her colleagues introduced the concept of co-production in the public sector, highlighting how citizens can actively participate in service delivery alongside providers, particularly in law enforcement. Since then, the understanding of co-production has expanded, encompassing a wider range of stakeholders, including volunteers, NGOs, and community organizations. Modern definitions describe co-production as a process in which providers (including public, private, or third sectors) and users or volunteers cooperate throughout all stages of the service delivery lifecycle—from commissioning to assessment. Co-production emphasizes shared responsibility and blurs traditional role distinctions, distinguishing it from broader concepts of citizen participation.

Co-production is closely related to collaborative governance, which emphasizes multi-stakeholder engagement, cross-sector collaboration, and consensus-based decision-making. Frameworks of collaborative governance proposed by Ansell and Gash (2008) and Emerson et al. (2012) identify key elements like incentives, leadership, and institutional design, which are also crucial for successful co-production. Core values such as equity, trust, and engagement underpin both approaches, reinforcing the alignment between co-production and collaborative governance in enhancing public value and leveraging diverse resources for societal benefit.

Research on co-production has highlighted the significant influence of factors such as citizen engagement, leadership, organizational culture, and communication dynamics. However, existing studies often examine these elements in isolation, overlooking their potential synergistic effects. This limitation is further amplified by the predominance of case studies and qualitative methodologies, which restrict the generalizability of findings. While some quantitative analyses have begun to emerge, they seldom consider the combined effects of multiple factors, underscoring the need for a more integrative approach.

This study addresses this gap by examining how collaborative factors interact to enhance co-production outcomes. Using fuzzy set qualitative comparative analysis (fsQCA) based on co-production cases from the Collaborative Governance Case Database, the study identifies effective combinations of collaborative factors. The findings provide an in-depth understanding of mechanisms driving successful co-production, with implications for both theory and practice, and suggest directions for future research.

本研究使用定性比较分析（QCA）方法，旨在探索协同要素在共同生产中的组合效应。QCA方法非常适合揭示复杂的因果路径，尤其在共同生产涉及多种相互关联的因素时表现尤为出色。通过辨别必要条件和充分条件，QCA能够识别不同因素组合如何促成共同生产，并提供比单一或对比性案例设计更广泛的结论。与大样本回归研究相比，fsQCA更擅长揭示多条因果路径，从而为理解多因素交互对共同生产的影响提供了更全面的视角。

为适应数据的特性，本研究采用模糊集QCA（fsQCA）方法，因为该方法能够处理来自李克特量表的连续数据，避免二元分类可能导致的信息丢失。fsQCA的独特优势在于其允许更细致的集合隶属度，捕捉社会现象中的细微差别。这种灵活性有助于数据细节的保留，更精确地反映理论构念，尤其是在合作和共同生产的背景下。FsQCA的数据校准能力也增强了分析的有效性，更贴切地体现了非二元的理论概念和实证情境。

本研究的数据源自协同治理案例数据库（CGCD），该数据库涵盖了44个来自全球不同部门的协同治理实例，包括环境、安全、社会服务、健康、基础设施、教育和文化等领域。CGCD是由全球学者和实践者共同开发的开源资源，采用61项问卷标准化编码，并提供关于一般案例信息、主要特征、起始条件、制度设计、领导力、协同过程、问责和协同成果等八个维度的详细定性描述。这些维度与既有的治理框架保持一致，为协同治理的分析提供了高质量和全面的数据支持。尽管CGCD中的案例并不一定代表所有协同治理实例，但其涵盖了协同治理的已知和未知面向，并由专家精心筛选为信息价值最高的案例，即便可能存在部分信息缺口，也保证了案例的代表性。

为聚焦具有深度用户和公民参与的共同生产，本研究对CGCD中的案例进行了两步筛选。首先，筛选出涉及公民或公民团体的案例（基于问题#23），确保分析的案例具备共同生产的潜力。接着，根据公民在协作中的影响力（基于问题#52），将5分李克特量表中的3分设定为最低标准，确保所选案例具备实质性公民参与。这一筛选标准符合共同生产概念，保证了用户在服务设计和交付中具有影响力，从而避免了未真正体现共同生产核心特征的案例。最终筛选出了18个符合条件的共同生产案例，具体描述如表1所示。

This study employs Qualitative Comparative Analysis (QCA) to explore the combinatorial effects of collaborative factors on co-production. QCA is particularly well-suited for uncovering complex causal pathways, especially when co-production involves multiple interconnected factors. By identifying necessary and sufficient conditions, QCA can discern how different configurations of factors contribute to co-production and provides broader conclusions than single or comparative case designs. Compared to large-sample regression studies, fsQCA is better equipped to reveal multiple causal pathways, offering a more comprehensive perspective on the effects of multi-factor interactions on co-production.

To align with the nature of the data, this study employs fuzzy set QCA (fsQCA), as it can handle continuous data from Likert scales, avoiding information loss that may result from dichotomization. FsQCA’s unique advantage lies in its ability to accommodate nuanced set membership, capturing subtle variations often present in social phenomena. This flexibility helps preserve data detail and provides a more precise reflection of theoretical constructs, particularly in the context of collaboration and co-production. Additionally, fsQCA’s calibration capabilities enhance analytical validity by allowing for a nuanced representation of theoretical concepts and empirical scenarios that are not strictly binary.

The data comes from the Collaborative Governance Case Database (CGCD), which includes 44 collaborative governance instances from various global sectors, such as environment, security, social services, health, infrastructure, education, and culture. CGCD is an open-access resource developed by scholars and practitioners worldwide, coded through a standardized survey of 61 questions and providing detailed qualitative descriptions across eight dimensions: general case information, main case characteristics, starting conditions, institutional design, leadership, collaborative processes, accountability, and collaborative outcomes. These dimensions align with established governance frameworks, providing high-quality, comprehensive data for collaborative governance analysis. While CGCD cases do not necessarily represent all collaborative governance instances, they encompass both known and unknown aspects of governance, selected by experts as the most informative cases available, ensuring representativeness despite potential information gaps.

To focus specifically on co-production with deep user and citizen participation, this study conducted a two-step selection of cases from the CGCD. First, cases involving citizens or citizen groups were selected (based on Question #23) to ensure potential for co-production analysis. Second, to capture substantial citizen participation, a minimum threshold of 3 on a 5-point Likert scale was set for the degree of citizen influence in collaboration (based on Question #52). This threshold aligns with the conceptual understanding of co-production, ensuring that users exert meaningful influence in the design and delivery of services, thus avoiding cases that do not fully embody the core characteristics of co-production. This rigorous selection process resulted in 18 eligible co-production cases, as detailed in Table 1.

本研究使用 fsQCA 3.0 软件分析数据，以识别共同生产的必要条件和充分条件。该软件通过真值表和布尔简化算法系统地分析条件（协同要素）与结果（共同生产）之间的关系，并计算一致性和覆盖率得分，以评估不同条件组合在解释结果中的相关性。通过生成多种解决方案，fsQCA 提供了因果关系的全面视角，揭示了实现共同生产的多种路径。

首先，为了确定必要条件，分析了在所有实现共同生产的案例中始终出现的条件。根据 Hossain 等人（2022）以及 Rihoux 和 Ragin（2009）的建议，研究设定了 0.9 的一致性和 0.5 的覆盖率作为基准。结果显示，激励是实现共同生产的潜在必要条件，表明它在实现结果时具有重要支撑作用。

在充分条件分析中，研究识别了能够促成高水平共同生产的不同协同要素组合。通过 fsQCA 软件创建的真值表中，共列出四个条件的所有可能组合。在 18 个案例中，fsQCA 识别出九个组合，其中四个组合与高水平的共同生产相关。研究将原始一致性阈值设为 0.75，并采用中间解决方案，以平衡复杂性和可解释性。结果显示，实现高共同生产的三条路径具有较高的覆盖率（0.890）和一致性（0.799），支持了模型的稳健性。

具体而言，第一条路径（管理型共同生产）结合了制度设计、激励和领导力，突出了领导力和制度结构在协作过程中的作用。在挪威 Svelvik 市（案例 C17）中，任务委员会促进了公民与政府间的协作，以提升公共服务质量。然而，该案例也揭示了部分政治家对参与的犹豫，表明在管理型共同生产中，领导者应构建共享的愿景，以确保沟通顺畅并充分实现协作潜力。

第二条路径（领导型共同生产）依赖于领导力、激励和协同过程，显示了领导力和协作机制（如对话）在推动公共服务中的重要性。美国的 Blackfoot Challenge（案例 C2）是该路径的典型案例，领导者通过有效整合多元利益相关者，共同应对复杂环境问题，体现了强有力的领导与协作机制在实现高水平共同生产中的作用。

第三条路径（自组织共同生产）展示了在缺乏集中式领导的情况下，通过制度设计、激励和协同过程的协同作用也能实现高水平的共同生产。米兰的 Area C 拥堵收费计划（案例 C9）为自组织协作的典型，广泛的公民参与与对地方管理的信任推动了协作进程，彰显了公民在推动共同生产中的重要作用。

This study analyzed the data using fsQCA 3.0 software to identify the necessary and sufficient conditions for co-production. This software systematically examines the relationships between conditions (collaborative factors) and outcomes (co-production) by using truth tables and Boolean minimization algorithms. It calculates consistency and coverage scores to assess the relevance of different configurations of conditions in explaining the outcome. By generating multiple solutions, fsQCA provides a comprehensive perspective on causal relationships, revealing various pathways to achieve co-production.

To determine necessary conditions, the analysis focused on factors that consistently appear in all cases where co-production is observed. Following recommendations by Hossain et al. (2022) and Rihoux and Ragin (2009), the study set benchmarks of 0.9 for consistency and 0.5 for coverage. Results indicated that incentives are a potential necessary condition for achieving co-production, suggesting that they play an important supportive role in producing the outcome.

In the analysis of sufficient conditions, the study identified different combinations of collaborative factors that contribute to high levels of co-production. In the truth table created by fsQCA software, all possible configurations of the four conditions were listed. Among the 18 cases, fsQCA identified nine combinations, four of which were associated with high co-production. The study set the raw consistency threshold at 0.75 and adopted the intermediate solution to balance complexity and interpretability. Results showed that three pathways to high co-production displayed notable coverage (0.890) and consistency (0.799), supporting the robustness of the model.

Specifically, the first pathway, termed “managed co-production,” combines institutional design, incentives, and leadership, highlighting the role of leadership and institutional structure in the collaborative process. In the municipality of Svelvik, Norway (Case C17), task committees promoted collaboration between citizens and the government to improve public service quality. However, this case also revealed some reluctance among politicians to participate, indicating that leaders in managed co-production should foster a shared vision to ensure smooth communication and fully realize collaborative potential.

The second pathway, “led co-production,” relies on leadership, incentives, and the collaborative process, demonstrating the importance of leadership and collaborative mechanisms (such as dialogue) in advancing public service. The Blackfoot Challenge in the United States (Case C2) exemplifies this pathway, where leaders effectively integrated diverse stakeholders to address complex environmental issues, underscoring the role of strong leadership and collaborative mechanisms in achieving high co-production.

The third pathway, “self-organized co-production,” shows that institutional design, incentives, and collaborative processes can also lead to high co-production even in the absence of centralized leadership. Milan’s Area C congestion charge initiative (Case C9) exemplifies self-organized collaboration, where broad citizen participation and trust in local governance drove the collaboration, highlighting the critical role of citizen involvement in fostering co-production.