American Police Unions: A Hindrance or Help to Innovation?

Abstract

In response to the crises of the 1960s, police departments in the United States embarked upon great change in their policies, operations, and programs. This literature, however, has largely neglected the duality of the role of police unions in the adoption of new programs and practices. Learning about the key actors in the innovations process, however, is crucial to understanding how organizational change emerges. This article builds upon previous research to understand the effects of police unions on the adoption of innovation. Findings indicate that police unions may actually promote rather than hinder the implementation of new innovation.

Keywords:

• police
• unions
• community policing
• innovation

Notes

¹ For an in-depth discussion of the development of the labor movement in American policing, please see Burpo (1971) or Walker (1977).

² The year 2003 is the most recent available LEMAS data set. LEMAS data of 2007 have been collected but not yet released.

³ The use of LEMAS data does limit some of our understanding of the nuances of innovation. For example, there are different levels of sophistication in terms of technology (Sever et al., 2008). Similarly, community policing can involve a multitude of strategies and activities that can be implemented differently in departments depending upon community needs. LEMAS data, however, offer a glimpse into police activity across a wide range of police agencies.

⁴ As noted early, more than 90% of police agencies serving populations larger than 25,000 report adopting COP (Hickman & Reeves, 2001).

⁵ Specifically, as suggested by Maguire (1997) vertical differentiation is measured as follows:Vertical differentiation (Height) = (Salary of Chief – Salary of Entry Officer)/Salary of Entry Officer.

⁶ This investigation tried to control for this by including the age of the agency in the analysis (King, 2009). This approach, however, ignores fluctuations in agency practices that can only be captured by multiple data points.