بررسی ارتباط بین رهبری یکپارچه و عملکرد بخش دولتی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|8758||2010||16 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : The Leadership Quarterly, Volume 21, Issue 2, April 2010, Pages 308–323
This study develops the concept of integrated leadership in the public sector. Integrated leadership is conceived as the combination of five leadership roles that are performed collectively by employees and managers at different levels of the hierarchy. The leadership roles are task-, relations-, change-, diversity-, and integrity-oriented leadership. Using data from the Federal Human Capital Survey and Program Assessment Rating Tool, we analyze the relationship between integrated leadership and federal program performance. The findings from the empirical analysis indicate that integrated leadership has a positive and sizeable effect on the performance of federal sub-agencies. The study concludes with a discussion of the implications of the findings and limitations of the study.
نتیجه گیری انگلیسی
The present study sought to synthesize leadership and public administration research to develop and measure the concept of integrated leadership in the public sector. Integrated leadership is conceived of as the combined efforts of organizational members across multiple levels of the hierarchy to perform five leadership roles: task-, relations-, change-, diversity-, and integrity-oriented leadership. The findings show that integrated leadership in the public sector matters when it comes to improving organizational performance. Integrated leadership is positively correlated with performance in the federal government as measured using PART results scores. Moreover, the size of leadership's effect on performance is meaningful, but the estimate of size is sensitive to the method used to compute it. In terms of substantive effect on the PART results score, a one standard deviation increase in integrated leadership results in almost a 21 point increase in the dependent variable. This would appear to be a sizeable effect and one that far exceeds previous estimates reported in the public management literature (Fernandez, 2005 and Meier & O'Toole, 2002) using the same estimation method. Adopting another approach more commonly used by leadership researchers, we find that integrated leadership accounts for about 4% of the variance in performance, an impact of smaller size. Although modest, a leadership effect of this latter size is not trivial, especially in light of the scale of federal operations. Public management scholars should be mindful of how using different approaches to estimating size effects can produce divergent results. The approach taken by leadership scholars that involves estimating the percent of the variance in performance accounted for by leadership variables offers the advantage of allowing researchers to compare size effects across studies using different measures of performance. This advantage might prove to be a critical one in the public sector, where we find a dearth of performance measures available for comparing different types of public organizations. While the advent of shared leadership as a legitimate approach to the study of leadership is a recent phenomenon, organization theorists for several decades have demonstrated the salient role played by leaders at multiple organizational levels. Small groups research has shown that leadership at low levels of the hierarchy is a key determinant of work team effectiveness (Guzzo & Dickson, 1996 and Kerr & Tindale, 2004). Other researchers have found that middle managers serve a critical role in organizations as architects and champions for organizational change (Dutton, Ashford, O'Neill, & Lawrence, 2001). This study of shared leadership in the federal bureaucracy should encourage public management scholars to adopt a broader view of leadership, one that recognizes the contributions to organizational effectiveness made by leaders at multiple levels. The continued treatment of leadership as a role played by a top executive runs the risk of failing to capture the full range of leadership behavior and efforts underway in public organizations. With this point in mind, we caution the reader about the study's limited ability to accurately capture leadership behavior at all levels of the federal sub-agencies in our sample. The majority of survey items used to measure integrated leadership referred to the behavior of team leaders, supervisors and managers; only a few asked directly about employees' perceptions of the agency's senior leadership. The findings might very well underestimate the contributions of senior executives to organizational performance. Other limitations to this study warrant discussion. While there are five leadership roles embedded in the concept of integrated leadership, it is important to note that the data prevented direct measurement and analysis of collaborative leadership, a leadership role that is becoming increasingly important in the governance arena. Collaborative management and leadership entails a variety of inter-organizational activities, including shaping policy issues and public priorities; mobilizing resources; facilitating coordination and mutual adjustment; resolving conflict; building trust; and measuring performance (Adams and McGuire, 2003 and Bingham and O’Leary, 2008Thomson & Perry, 2006). Networked arrangements in which multiple actors are involved in co-producing a public service without any one of them being in charge places a premium on this type of leadership (Crosby & Bryson, 2005). Although the integrated leadership construct taps into dimensions such as diversity- and integrity-oriented leadership that are relevant for collaborative leadership, additional data gathering and analysis is needed to explore the relationship between collaborative leadership and the five leadership roles examined in this study. The approach taken to defining and measuring integrated leadership might also raise questions regarding convergent and discriminant validity. As the results of the higher-order confirmatory factor analysis (CFA) indicate, the five-dimensional model of integrated leadership exhibits a better fit with the data than a one-dimensional model. All five leadership roles are positively correlated with one underlying construct, which we label integrated leadership, and yet, they are sufficiently different to suggest they capture distinct dimensions of leadership. However, most of the goodness-of-fit measures in the higher-order CFA are near the borderline between good and poor model fit, indicating room for improvement. In particular, there is reason to believe that a six-dimensional model of integrative leadership that includes a leadership style aimed at managing collaboratively across organizational boundaries would offer a better model fit than our five-dimensional model of integrated leadership. Moreover, the five dimensions representing different leadership roles exhibit fairly high bivariate correlations, ranging from r = 0.61 (change- and diversity-oriented leadership) to r = 0.77 (task- and relations-oriented leadership). These correlations suggest the need for additional research on the discriminant validity of the integrated leadership construct and more general issues of appropriate model specifications for the construct (Jarvis, MacKenzie, & Podsakoff, 2003). In fact, we believe future research about integrated and shared leadership should place a high priority on measurement issues because of the centrality of measurement for further development of shared and integrated leadership research. The findings are strengthened by the fact that the dependent and independent variables are measured using different data sources, helping to minimize common source bias. Another thorny methodological issue that arises in this study is simultaneity bias. Attribution theory of leadership posits that followers make positive attributions of leaders for a variety of reasons, including when the organization performs well (Awamleh & Gardner, 1999 and Lord & Maher, 1991). This raises the specter of a simultaneous relationship between leadership and performance when using employee attitudes to measure leadership behavior and traits: good leadership improves performance, but higher levels of performance cause followers to make positive attributions of leaders. We tried to minimize simultaneity bias by using independent and dependent variables measured at different points in time. That is, roughly half of the PART results scores were assigned by OMB in 2007, while the data used to measure integrated leadership was collected by OPM in 2006. Had the data been available to derive all the PART results scores in 2007, we would have concluded more confidently that the relationship between integrated leadership and PART results was a causal one rather than a mere statistical correlation. As such, the results should be interpreted with some caution. Finally, the data sources and the sample of organizations on which the empirical analysis is based suggest that the findings are generalizable across the federal bureaucracy. Whether or not a leadership style that incorporates the same five roles has a positive influence on performance in smaller state and local agencies, and what the size of such an effect might be, are empirical questions that remain to be answered.