公共管理者的教育类型与优先提供核心服务和沟通之间的关联

公共管理中的一个重要问题是公共组织应该优先还是避免聘用专业经理（即专业背景与其所监管的工作相匹配的经理）。研究表明，专业经理在公共组织中与领域特定绩效更高相关，理论表明他们的领导方式与非专业经理不同。然而，没有实证证据支持后一种说法。为了解决这个问题，我们调查了专业经理是否比非专业经理更重视核心服务提供并更好地沟通。通过对员工在学校、养老、日托和就业环境中的领导者行为报告进行多层次回归，我们发现专业经理在除学校以外的所有环境中都更重视核心服务提供，在养老和就业环境中沟通更好，但在其他环境中则不然。这些发现与理论预期一致，从而强调了教育背景在塑造管理行为方面的重要性。

An important question in public management is whether public organizations should prefer or avoid specialist managers (i.e., managers whose professional backgrounds match the work they oversee). Research has shown that specialist managers are associated with better field-specific performance in public organizations, and theory suggests that they lead differently than nonspecialist managers. However, there is no empirical evidence to support this latter claim. To remedy this, we investigate whether specialist managers prioritize core service provision more and communicate better than nonspecialist managers. Through multilevel regressions of employee reports of leaders' behaviors in the contexts of schools, eldercare, daycare, and employment, we find that specialist managers prioritize core service provision more in all contexts but schools and communicate better in eldercare and employment but not elsewhere. These findings are consistent with the theoretical expectations and thus emphasize the importance of educational background in shaping managerial behavior.

引言

    公共管理中的一个重要问题是公共组织是否应该偏爱或避免聘用专业经理（即专业背景与其所监管的工作相匹配的经理）。公众对这个问题的辩论往往是主观的，专业经理的支持者和反对者都提到了聘请他们担任经理的非凡优势或劣势。然而，经验证据较少，管理资格和适合性问题迄今为止研究不足。文献中的三个缺点阻碍了对专业经理的彻底了解。

    第一个缺点是缺乏关于专业经理行为的经验证据。第二个缺点涉及管理水平。最后一个缺点与迄今为止研究的部门有关。到目前为止，我们不知道证据是否扩展到中低专业化程度的公共服务提供领域中数量更多的经理。

    我们试图分两步缓解这些缺点。首先，我们调查了专业经理和非专业经理之间的行为差异，以弥补他们领导方式缺乏的实证证据。其次，我们调查了一线经理，并分析了福利国家的四项关键公共服务：学校、养老、日托和就业。我们利用丹麦某市所有经理和员工的横断面数据集来调查我们的假设。我们发现，第一个假设得到了广泛支持，即专业经理比非专业经理更重视核心服务提供。我们发现，第二个假设得到了部分支持，即专业经理比非专业经理沟通得更好（我们只发现在养老和就业方面有统计学显著影响）。总之，我们的研究通过扩大研究成果、服务提供领域和分析的领导水平方面的研究范围，为公共行政研究中专业经理的文献做出了贡献。

    An important question in public management is whether public organizations should prefer or avoid specialist managers (i.e., managers whose professional backgrounds match the work they oversee). The public debate on the matter is often opinionated, and proponents and opponents of specialist managers refer to extraordinary advantages or dis-advantages of having them as managers. However, the empirical evidence is more modest and questions of managerial qualifications and fit are understudied so far. Three shortcomings in the literature impede a thorough understanding of specialist managers.

    The first shortcoming is a lack of empirical evidence on specialist managers' behaviors. The second shortcoming concerns the level of management. The final shortcoming relates to the sectors researched so far. So far, we do not know whether the evidence extends to the much more numerous managers in public service provision areas of middle-to-low degrees of professionalization.

    We attempt to alleviate these shortcomings in two steps. First, we investigate differences in behaviors between specialist and nonspecialist managers to remedy the lacking empirical evidence on their approach to leadership. Second, we investigate frontline managers, and we analyze four key public services in the welfare state: schools, eldercare, daycare, and employment. We investigate our hypotheses with a cross-sectional dataset of all managers and employees in a Danish municipality. We find broad support for our first hypothesis that specialist managers prioritize core service provision more than nonspecialist managers. We find partial support for our second hypothesis that specialist managers communicate better than nonspecialist managers (we only find statistically significant effects in eldercare and employment). In sum, our study contributes to the literature on specialist managers in public administration research by broadening the scope of research in terms of outcomes, service provision areas, and leadership levels analyzed.  

假设

假设H1. 专业经理比非专业经理更重视提供核心服务。

假设H2. 专业经理比非专业经理沟通得更好。

Hypothesis H1. Specialist managers prioritize core service provision more than nonspecialist managers.

Hypothesis H2. Specialist managers communicate better than nonspecialist managers.

研究设计

    我们利用了丹麦一个大型市镇的数据。丹麦是一个小型一元论国家，拥有庞大的福利国家和多层级政府体系，包括国家、地区和市级。丹麦市镇是调查专业和非专业管理人员的一个很好的案例，因为它使我们能够调查不同专业化程度的分支机构一线管理人员的领导力，而这在文献中是缺失的。在本研究中，我们调查了埃斯比约市的专业和非专业管理人员，该市由一个大城市埃斯比约、几个城镇和小村庄组成。埃斯比约在规模、经济、税收、人口统计和政治气候方面与大多数丹麦市镇相当。

    我们的样本包括 2020 年秋季丹麦埃斯比约市的所有管理人员和员工，并基于三个独立来源：管理人员和员工就业岗位的登记数据、关于员工对其管理人员行为和管理人员正规教育的看法的领导力评估调查，以及包括管理人员和员工性别、年龄和任职年限的工作环境调查。我们排除了除一线经理以外的经理以及无法汇总为字段的领域的一线经理（n = 189 名经理）。因此，我们还排除了与其余一线经理无关的员工（n = 2928 名员工）。这产生了一个嵌套在 135 名经理中的 1996 名员工的数据集。我们关注的是该市最大和最中心的公共服务提供领域：学校、日托、老年人护理和就业。

    我们用两个各有五项的指数来衡量员工对其经理优先考虑核心服务提供和沟通的看法。在经理层面，我们纳入了经理的年龄、性别和任职年限，即在当前职位上工作的年限。此外，我们纳入了经理的控制范围，定义为直接向经理汇报的员工人数，以及经理完成的最高管理培训水平（如果有）。最后，我们纳入了衡量经理是否暂时替代管理职位的指标。在员工层面，我们纳入了衡量员工年龄、性别和任职年限的指标，同样定义为在当前职位上工作的年限。除经理完成的最高管理培训水平外，所有背景特征均来自埃斯比约市的登记处，该信息是通过埃斯比约市的内部调查获得的。

    我们的方法分为两个步骤。首先，我们使用多级模型来解决我们数据的嵌套结构。我们在回归模型中控制了经理和员工的任职年限、年龄和性别以及经理的顶级教育、控制范围和就业状况。其次，我们使用逆概率加权来平衡专业和非专业经理之间不均衡的员工人口结构。我们使用连续独立员工级别变量年龄和任期的组均值中心化。我们还使用了 Mundlak 设备，它是员工级别变量的组集合。所有模型均使用 R 4.1.1 和 lme4 1.1-27.1 的完全最大似然估计法，用随机截距估计。我们分两步研究了多层模型的稳健性。首先，我们在没有逆概率加权的情况下估计了模型。其次，我们通过估计每个模型的最佳无偏线性预测因子来仔细检查模型，以确保经理级别的未解释方差不会混淆主要分析中的模式。

    We draw on data from a large Danish municipality. Denmark is a small unitarian country with a large welfare state and a multi-tier government system with a national, regional, and municipal tier. A Danish municipality is a good case for investigating specialist and nonspecialist managers, because it allows us to investigate the leadership of frontline managers across branches with varying levels of specialization, which is missing in the literature. In this study, we investigate specialist and nonspecialist managers employed in Esbjerg municipality, which consists of a single large city, Esbjerg, a few towns, and small rural villages. Esbjerg compares to most Danish municipalities in size, economy, taxation, demographics, and political climate.

    Our sample includes all managers and employees in Esbjerg Municipality in Denmark in fall 2020 and is based on three separate sources: register data on managers and employees' employment positions, a leadership evaluation survey of employees' perceptions of their managers' behaviors and managers' formal education, and a work environment survey including managers and employees' gender, age, and tenure. We excluded managers other than frontline managers and frontline managers in areas that could not be aggregated as fields (n = 189 managers). Accordingly, we also excluded employees who were not related to the remaining frontline managers (n = 2928 employees). This resulted in a dataset of 1996 employees nested in 135 managers. We focused on the largest and most central public service provision areas in the municipality: schools, daycare, eldercare, and employment.

    We measured employees' perception of their managers' prioritization of core service provision and communication with two indices of five items each. At the manager level, we included the manager's age, gender, and tenure, that is, number of years employed in the current position. Moreover, we included the manager's span of control defined as the number of employees reporting directly to the manager, and the manager's highest level of completed managerial training, if any. Finally, we included a measure for whether the manager is temporarily substituting a managerial position. At the employee level, we included measures for the employee's age, gender, and tenure, similarly defined as the number of years employed in the current position. All background characteristics were obtained from registries in Esbjerg municipality except the manager's highest level of completed managerial training, which was obtained through an internal survey in Esbjerg municipality.

    Our approach had two steps. First, we used multilevel models to address the nested structure of our data . We controlled for managers and employees' tenure, age, and gender as well as managers' top-off education, span of control, and employment status in the regression models. Second, we used inverse probability weighting to balance out uneven employee demographics across specialist and nonspecialist managers. We used group mean centering of the continuous independent employee-level variables age and tenure. We also used Mundlak devices, which are group aggregates of employee-level variables. All models were estimated with random intercept using full maximum likelihood estimation using R 4.1.1 and lme4 1.1-27.1. We investigated the robustness of the multilevel models in two steps. First, we estimated the models without inverse probability weighting. Second, we scrutinized the models by estimating the best unbiased linear predictors for each model to ensure that unexplained variance at the manager level is not confounding the pattern in the main analysis.

结果

    在所有情况下，优先考虑核心服务提供的系数均为正值，这与我们的假设一致。在老年护理（β = 0.51，p = 0.062）、日托（β = 0.33，p = 0.099）和就业（β = 0.75，p = 0.067）方面，这种关联具有统计学显著性，为 10%，但在学校（β = 0.09，p = 0.743）方面则不具有统计学显著性。因此，我们发现员工认为专业经理在日托、老年护理和就业方面更优先考虑核心服务提供，但在学校方面则并非如此。

    在所有情况下，沟通的系数均为正值，这与我们的假设一致。这种关联在养老护理 (β = 0.70, p = 0.018) 和就业 (β = 0.88, p = 0.035) 方面具有统计学显著性，但在日托 (β = 0.11, p = 0.480) 和学校 (β = 0.10, p = 0.742) 方面则不显著。统计学显著性影响范围从 0.7 到 0.9，接近 7 分制的满量表点。因此，我们发现员工认为专业经理在养老护理和就业方面比非专业经理沟通更好，但在学校和日托方面则不然。

    我们的敏感性分析支持了我们的主要发现。首先，主要分析中的模式对于省略逆概率加权的权重是稳健的。其次，每个模型的最佳无偏线性预测因子的估计值大多与零重叠。这意味着我们的模型几乎完美地解释了管理者层面的差异，从而降低了管理者层面无法解释的差异混淆主要分析模式的可能性。因此，我们将结果的稳健性解释为我们的结果不仅仅是方法论的产物的证据。

    The coefficients for prioritizing core service provision are positive across all contexts, which is consistent with our hypothesis. The association is statistically significant at 10% in eldercare (β = 0.51, p = 0.062), daycare (β = 0.33, p = 0.099) and employment (β = 0.75, p = 0.067), but not statistically significant in schools (β = 0.09, p = 0.743). Thus, we find that employees perceive specialist managers as prioritizing core service provision more in the context of daycare, eldercare, and employment, but not in schools.

    The coefficients for communication are positive across all contexts, which is consistent with our hypothesis. The association is statistically significant in eldercare (β = 0.70, p = 0.018) and employment (β = 0.88, p = 0.035), but not in daycare (β = 0.11, p = 0.480) and schools (β = 0.10, p = 0.742). The statistically significant effects range from 0.7 to 0.9, which is close to a full-scale point on the 7-point scale. Thus, we find that employees perceive that specialist managers communicate better than nonspecialist managers in eldercare and employment but not in schools and daycare.

    Our sensitivity analyses support our main findings. First, the pattern in the main analysis is robust to omitting the weights from inverse probability weighting. Second, the estimates of the best unbiased linear predictors for each model mostly overlap zero. This implies that our models almost perfectly account for the variance at the manager level, which reduces the chance that unexplained variance at the manager level confounds the pattern in the main analysis. Thus, we interpret the robustness of the results as evidence that our results are not merely methodological artifacts.

讨论

    据我们所知，我们的研究首次提供了有关专业和非专业管理者行为的实证证据，并为推进专业管理者领导理论奠定了重要的实证基础。我们的研究结果表明，根据员工的说法，专业管理者比非专业管理者更重视核心服务提供，沟通也更清晰，这两种行为被认为有利于特定领域的绩效。然而，我们的研究结果揭示了一些有趣的变化。虽然优先考虑核心服务提供的效果存在于除学校以外的所有环境中，即日托、养老和就业，但沟通的效果只存在于养老和就业中。

    考虑到这一点，我们的研究结果仍然有几个有趣的理论含义。首先，我们为 Goodall 和 Bäker 的理论论点提供了实证支持，即专业管理者比非专业管理者更重视组织中的核心服务。其次，我们的研究结果支持 Mumford 等人的理论论点，即专业管理者比非专业管理者沟通更清晰。第三，我们的研究结果支持了理论论点，即专业经理的优势也适用于一线经理，因为我们表明专业经理的领导方式与一线的非专业经理不同。第四，在所研究的行业中，区分专业经理和非专业经理在低到中等专业化水平上是有用的。最后，我们想强调一下我们的研究结果中一个令人震惊的细微差别：在小学的背景下，没有统计学上显著的关联。

    我们的研究有一些局限性，这为进一步的研究开辟了道路。首先，正如方法部分所指出的，我们的研究设计是横断面的，这限制了因果结论的范围。其次，我们的研究设计不允许我们研究潜在的理论机制。第三，优先考虑核心服务提供和沟通是重要的结果，但其他领导行为也可能影响员工或组织绩效。最后，尽管我们有一个相当全面的数据集，但在调查的每个公共服务提供领域中，非专业人员的比例相当低。

    To our knowledge, our study is the first to present empirical evidence on specialist and nonspecialist managers' behavior and to establish an important empirical foothold to advance the theory on how specialist managers lead. Our findings indicate that specialist managers, according to the employees, prioritize core service provision more and communicate clearer than nonspecialist managers, two behaviors that are theorized to be conducive to field-specific performance. However, our findings reveal some interesting variation. While the effect for prioritizing core service provision is found in all contexts but schools, that is, daycare, eldercare and employment, the effect for communication is only found in eldercare and employment.

    Keeping this in mind, our findings still have several interesting theoretical implications. First, we provide empirical support for Goodall and Bäker's theoretical argument that specialist managers prioritize the core service in an organization more than nonspecialist managers. Second, our findings support Mumford et al.'s theoretical argument that specialist managers communicate more clearly than nonspecialist managers. Third, our results support the theoretical arguments that the advantages of specialist managers also apply to frontline managers, because we show that specialist managers lead differently than nonspecialist managers at the frontline. Fourth, the distinction between specialist and nonspecialist managers proved useful at the low to medium levels of specialization in the sectors examined. Finally, we want to highlight a thrilling nuance to our findings: the absence of a statistically significant association in the context of primary schools.

    Our study has some limitations, which open up avenues for further research. First, as flagged in the methods section, our research design is cross-sectional, which limits the scope for causal conclusions. Second, our research design does not allow us to examine the underpinning theoretical mechanisms. Third, prioritizing core service provision and communication are important outcomes, but other leadership behaviors may also affect employees or organizational performance. Finally, even though, we have a quite comprehensive data set, the proportion of nonspecialists in each public service provision area under investigation is quite low.

结论和实际意义

    在本研究中，我们调查了专业经理是否比非专业经理沟通更好，更重视核心服务提供。这些主题对于公共管理研究很重要，因为它们被理论化为专业经理与特定领域绩效正相关关系的基础。我们发现，在日托、养老和就业方面，专业经理比非专业经理更重视核心服务提供，但在学校方面并非如此。其次，我们发现专业经理在养老和就业方面沟通更好，但在学校和日托方面则不然。我们的研究结果以三种方式为专业经理和非专业经理的文献做出了贡献。首先，我们提供了第一个实证证据，支持专业经理在优先考虑核心服务和沟通方面的行为与非专业经理不同这一命题。因此，我们迈出了绘制专业经理行为与特定领域绩效之间相关性基础的第一步。其次，我们表明专业经理的优势并不局限于高层管理的公共经理这一小众群体。第三，我们将研究的部门范围扩大到公共部门中一些最大的服务领域。我们的结果表明，专业和非专业管理者之间的区别对其他级别的公共管理（一线管理者）以及专业化程度低到中等的服务领域具有相当大的解释力。

    我们的研究结果对公共管理者和教授公共管理的专业人士也具有实际意义。公共部门的高层管理人员在启动招聘流程时应该意识到这些优势（以及非专业人士的劣势）。他们可以留出时间更彻底地检查申请人的核心服务提供方法以及他们在沟通中处理专业思维导图的技能。对于非专业一线管理者来说，这并不意味着他们就一定不合格。我们的研究结果对公共管理教育也有影响。许多教育项目都侧重于弥合研究、教学和实践之间的差距。我们的研究结果强调了将通才技能与实践导向（专业）知识相结合的重要性，并通过使用以公共管理者实践经验为基础的教学技巧，积极地将现实生活经验引入教育学习空间。通过这种方式，公共管理课程可以帮助公共管理人员在现实生活中培养和应用所需的技能。因此，尽管本文的结果表明专业经理人比非专业经理人具有优势，但我们的结果也指出了非专业经理人可以指导和发展其管理工作的可能领域。

    In this study, we investigated whether specialist managers communicate better and prioritize core service provision more than nonspecialist managers. These topics are important for public administration research because they are theorized to underpin specialist managers' positive correlation with field-specific performance. We found that specialist managers prioritize core service provision more than nonspecialist managers in daycare, eldercare, and employment but not in schools. Second, we found that specialist managers communicate better in eldercare and employment but not in schools and daycare. Our findings contribute to the literature on specialist and nonspecialist managers in three ways. First, we provide the first empirical evidence supporting the proposition that specialist managers behave differently than nonspecialist managers with regard to prioritizing core services and communication. Thus, we take the first step in charting the underpinnings of the correlation between specialist managers' behavior and field-specific performance. Second, we show that the advantages of specialist managers are not confined to the small niche of public managers at the top level of management. Third, we extend the scope of examined sectors to some of the largest service areas in the public sector. Our results show that the distinction between specialist and nonspecialist managers has substantial explanatory power for other levels of public management (frontline managers) and in service areas with low to middle levels of specialization.

    Our results also have practical implications both for public managers and for professionals teaching public administration. Top-level managers in the public sector should be aware of these advantages (and disadvantages for nonspecialists) when they initiate hiring processes. They could set aside time to more thoroughly examine the applicants' approaches to core service provision and their skills to address professional mental maps in their communication. For nonspecialist frontline managers, this does not imply that they are automatically unqualified. Our results also have implications for public administration educations. Many educational programs have focused on bridging the gap between research, teaching, and practice. Our results emphasize the importance of combining generalist skills with practice-oriented (professional) knowledge and drawing real-life experiences actively into the educational learning spaces by using pedagogical techniques that are grounded in the public managers hands-on experiences. In this way, public administration programs may help public managers develop and apply the needed skills in real life professional settings. Thus, even though, the results in this article indicate that specialist managers have advantages compared to nonspecialist managers, our results also point out possible areas where nonspecialist managers may direct and develop their management efforts.

《公共行政》微信公众号刚刚上线不久，我们期待来自各界的宝贵意见，您可将您的建议发送给后台。如果您对我们推送的文章感兴趣，也可在下方留言评论，与其他读者一起交流。

”