Police and Delinquency Intervention/Prevention Programs

Public Services?

TlieGseof Community-Oriented Polidi^ Gerasimos A.

Gianakis, University of Central Florida G. John Davis, III, St. Petersburg Police Department Wbat are heal law enforcement organizations adopting under tbe banner of community-oriented policing? Does implementation of community-oriented policing entail changes to the traditional structure of local law enforcement agencies?

To what extent are implementation and structural changes being supported by changes in operational policies and administrative systems?

The authors address these issues through a survey of local law enforce- ment agencies in Elorida. They find that community-oriented policing manifests itself in a variety of forms, but they all seem to center on chan^ng the officer rather tban the organization. Even when structural change occurs, the impact on existing policies and systems is minimal The authors question tbe long range prospects for this innovation.

If the sound bites of the reinventing govern- ment movement are ever to attain the status of sound scripture, they must first be operational- ized in terms of specific services. These slogans call for a market-oriented, customer-driven gov- ernment, owned by empowered communities and featuring decentralized services focused on preventing rather than on curing. These new initiatives are delivered by public organizations employing participative, team-oriented manage- ment systems and results-oriented evaluative criteria (Osborne and Gaebler, 1992). Although community policing builds on previous innova- tions in policing, such as problem-oriented policing (Goldstein, 1990; 1979) and team policing (Schwartz and Clarren, 1977; Sherman et al., 1973) that predate the formal reinventing government movement, community policing has become recognized as the law enforcement manifestation of that movement (Turner and Wiatrowski, 1995). Like outcome-oriented, mission driven government, community polic- ing has become a global phenomenon (Bayley, 1994a; Weatheritt, 1991) that refiects reinven- tion's market values and outcome-oriented management precepts.

In his capacity as chairman of the Task Force on Government Accomplishment and Account- ability of the American Society for Public Administration, Raymond T. Olsen reviewed nearly 40 case studies of outcome management strategies in the public sector, and he concluded that the core management systems of the orga- nizations pursuing these strategies were general- ly not being realigned to accommodate them (Olsen, 1997); that is, these management sys- tems continue to reinforce a process of control rather than supporting a results-driven orienta- tion.

Public Administration Review • November/December 1998, Vol. 58, No. 6 485 This article examines the implementation of the com- munity policing model by local law enforcement agencies in Florida. It focuses on the extent to which adoption has affected existing policies, procedures, and organiza- tional structures. Local law enforcement service delivery systems have typically been housed in hierarchical bureaucracies featuring command-and-control manage- ment systems. Such systems would appear to be antithet- ical to the spirit and substance of community policing.

Community policing has suffered from conceptual confusion in both research and practice (Roberg and Kuykendall, 1993; Wycoff, 1991; Greene and Taylor, 1991).

Community policing has been implemented in a wide variety of ways, manifesting diflFerences in person- nel, organizational structures, deployment schemes, patrol modes, operational functions, geographical scopes, and degree of involvement of citizens and coordination with external agencies (Bayley, 1994a; 1994b; Sadd and Grinc, 1994). However, despite this confusion, commu- nity policing is widely accepted by politicians and police professionals as an innovative way to deliver police ser- vices (Eck and Rosenbaum, 1994). As early as 1984, 143 police agencies surveyed nationally reported utilizing community policing (Trojanowicz and Harden, 1985).

Forty-two percent of all police departments serving pop- ulations of over 50,000 recently reported that they had adopted some form of community policing (Trojanowicz, 1994).

For Bayley (1994a) the basic elements of community policing are: consultation with community groups regarding their security needs; command devolution so that those closest to the community can determine how to best respond to those needs; mobilization of agencies other than the police to assist in addressing those needs; and remedying the conditions that generate crime and insecurity through focused problem solving. He and oth- ers (Moore, 1994; Eck and Rosenbaum, 1994) are com- fortable with the ambiguity that surrounds the concept, but WycofF(1991) suggests that it may now be a barrier to effective communication about police practices. Walk- er (1992) notes that community policing may be victim- ized by its popularity and rapid expansion inasmuch as proper planning and implementation appear to be lack- ing. Moore calls for "a survey of the field of policing to determine to what extent community policing is moving from rhetorical to operational levels throughout the country" (298). Like the reinventing government move- ment in general, the question of whether the community policing model represents a lasting innovation is very much an open one.

Issues in Community Policing Community policing calls for a new breed of police officer (Trojanowicz and Bucqueroux, 1990) operational- izing a new professionalism based on democratic values such as participation and openness, rather than on tech- In order to successfully implement their community policing programs, most researchers contend that police organizations must adopt an "organic" organizational structure.

nological values rooted in substantive expertise (Skolnick and Fyfe, 1995). The community policing officer nego- tiates and designs policing for particular areas (Bayley, 1994a), and these policing strategies are directed to proactive prevention rather than reactive detection (Tro- janowicz and Bucqueroux, 1990). One issue that emerges here is whether these attitudinal and behavioral changes at the police officer level will be supported by structural changes in the police organization. Conven- tional wisdom in organization theory holds that an orga- nization's structure should be designed to optimize the functioning of its operational technology. However, it is possible that the "911" emergency response function will continue to drive the structure ofthe police organization, and the community policing model will be forced to fmd its place within the hierarchical military model that has traditionally housed this reactive function. The mainte- nance of a dual proactive/reactive patrol capacity would certainly strain the resources of most agencies.

In order to successfully implement their community policing programs, most researchers contend that police organizations must adopt an "organic" organizational structure, a participatory manj^ement style, new reward structures, new training programs and selection criteria, and new control systems (Roberg, 1994; Kelling and Moore, 1991). Skolnick and Fyfe (1995) identify a decentralized command structure as an essential element of community policing. For Trojanowicz and Bucqueroux (1990), community policing entails the formal enrich- ment or enlargement of the job of patrol officer, and increasing the autonomy of the patrol officer calls for the enlargement of citizen participation as oversight to pre- vent potential abuses. Hence, the community policing model makes a host of demands on the hierarchical, mili- tary model, which has been largely closed to public par- ticipation. This may be the reason that in 1994, Moore could report that "in practice, no department has yet fully implemented community policing as an overall phi- losophy" (290).

Following Bittners (1970) classic analysis ofthe police function, Klockars (1991) suggests that community policing "is best understood as the latest in a fairly long tradition of circumlocutions whose purpose is to conceal, mystify, and legitimate police distribution of nonnego- tiably coercive force" (240). These circumlocutions are necessary because the police function is inherently offen- sive but society must nevertheless reconcile itself to its necessity. Bittner described the first three of these circum- 486 Public Administration Review • November/December 1998, Vol. 58, No.

6 * be implemented?'—the concept is already extremely pop- A M partnership and true co-production of police ular with policy makers—but rather 'how should it be LLiiimplemented?'" (1994, 125).

However, the model is ld b l fU d i ill b p services would entail some "dvilianimon" of police clearly plagued by conceptual confUsion, and it will be il : implemented in an organizational structure that may be positions and an enhanced public oversight of police antithetical to its basic precepts.

Issues in the Organization of , , ,. . . , . u Law Enforcement locutions—the legalization, the militarization, and the professionalization of the police function. For Klockars, The local police organization is treated in the literature community policing is an attempt to clothe coercive force as an example of what Meyer and Rowan (1983) called in the attractive weave of community, cooperation, and the "institutionalized organization." These are organiza- crime prevention; but "police can no more create com- tions employing ambiguous technologies in uncertain munity or solve problems of urban anomie than they can environments that produce outcomes that are difficult to be legalized into agents of the court or depoliticized into appraise. Hence, issues of appropriate technology and pure professionals" (1991, 257). Other researchers ques- structural adaptation to environmental change cannot be tion the very concept of "community" (Crank, 1994). resolved on the basis of efficiency. In response to this Community policing foctises on maintaining order, and uncertainty, "internal participants and external con- the political community defines public order; however, stituents alike call for institutionalized rules that promote the necessary political consensus may prove to be elusive, trust and confidence in outputs and buffer organizations particularly in the areas that are most afflicted with "dis- from failure" (Meyer and Rowan, 1983, 36). However, order" (Mastrofski, 1991). Buerger (1994) suggests that institutional rules inevitably come into conflict with community mobilization may also be difficult. He whatever efficiency criteria may exist, because "the gener- believes that the community policing model does not alized rules of the institutional environment are often seek to "empower" communities but rather to enhance inappropriate to specific situations" (Meyer and Rowan, the police response to crime through intelligence gather- 1983, 37).

ing. A full partnership and true co-production of police Meyer and Rowan (1983) identify four possible orga- services would entail some "civilianization" of police nizational responses to the dynamic tension created by positions and an enhanced public oversight of police the need to conform to "ceremonial" rules in order to activities. maintain legitimacy and the need to resolve the uncer- Formal evaluations of some manifestations of commu- tainties regarding day to day operations precipitated by nity policing have been undertaken (Rosenbaum, 1994; that conformance. First, the organization can resist the Greene and Mastrofski, 1991). These studies have yield- incorporation of ceremonial rules, but it may then be ed small positive effects or contradictory results, and they unable to establish its legitimacy based on efficiency cri- have been piqued by methodological problems, indud- teria. Second, it can maintain conformity by cutting off ing weak implementations of the program (Bayley, environmental relations, but this would be tantamount to 1994b; Greene and Taylor, 1991). The selection of crite- acknowledging that the operational "myths" embodied in ria for evaluating these programs highlights some of the the institution's rules don't actually work. Third, the issues regarding the ambiguity of the community policing organization can acknowledge that its structure does not concept. Moore (1994) suggests that the reduction of the manifest an optimal response to the demands of its oper- fear of victimization and increases in feelings of security ational environment, but this would surely compromise have independent value apart from any actual reduction its claims to legitimacy. Fourth, the organization can in crime rates. For Moore, community mobilization and promise reforms in the future, but this may also weaken the enhancement of community control are also valued the case for current legitimacy. Alternatively, the organi- ends in themselves. Shaw (1993) contends that commu- zation can decouple its formal structures from its opera- nity policing efforts should be evaluated on the extent to tional activities. "The assumption that formal structures which they are successful in restructuring police organiza- are working is buffered from the inconsistencies and tions so that they may continue to be adaptive and inno- anomalies involved in technical activities" (Meyer and vative, and enhance the creative potential of their person- Rowan, 1983, 39). Rather than focusing on the coordi- nel. Greene and Taylor (1991) suggest more technical nation and control of operational activities on the basis of evaluation criteria that still reflect the spirit of communi- efficiency criteria, management becomes a largely cere- ty policing, such as the number of crimes committed by monial fiinction focusing on elaborate displays of confi- persons from outside the community. dence, satisfaction, and good faith in the face of ambigu- Despite these real and potential problems, Wilkinson ous goals.

and Rosenbaum contend that "the ftindamental question In the police literature, the elements of the institution- about community policing in the 1990s is not 'should it alized organization are expressed as the "dual realities" of Reinventing or Repackaging Public Service?

The Case of Community-Oriented Policing 487 Methodology The community policing issues and concepts in police organization described above guided the development of a questionnaire to examine the two issues in this study. First, what are Florida law enforcement agencies implement- ing under the rubric of community policing, and second how is it affecting the existing formal structure? The ques- tionnaire was field tested with sworn managerial personnel from a police department and a sheriff's office. The questionnaire was mailed to the chief executive officers of the 116 local law enforcement agencies in the state with at least 50 sworn personnel. Bayley (1994b) has suggested that most small police departments employ some variant of community policing by defatilt, in that the structure of small agencies is simpler and their policies and processes less formally defined. The initial mailing and a follow-up letter yielded a response rate of 77.6% (N = 90). Of the police departments 54 of 76 responded; of the sheriff's offices 27 of 40 responded. Eight respondents did not iden- tify their agency. The chief executive officer was asked to complete the questionnaire from his or her personal per- spective.

All but one of the 90 chief executives indicated that their agencies employed a variant of community policing— striking testimony to the elasticity of this "one size fits all" model. Equally striking was the median date of adop- tion—a fairly recent 1993. These results suggest that law enforcement executives have gradually accepted the "fact. ..that community policing is the only form of policing available for anyone who seeks to improve police oper- ations, management, or relations with the public" (Eck and Rosenbaum, 1994, 4), and who can be against that?

Any ambitions to identify determinants of adoption were quashed by the virtually unanimous implementation of some variant of community policing by this sample.

The respondents were also asked to identify: (a) the specific community policing variant adopted (ranging from the adoption of a community policing philosophy, to the implementation of special community policing units, to organization-wide implementations entailing geographic decentralization); (b) the effects that adoption had had on selected policies and procedures; (c) the elements of the adopted variant; (d) the operational activities of each agen- cy's patrol personnel; (e) the relative strength of resistance to adoption from various potential sources; (f) the rela- tive severity of various potential operational problems; (g) the goals of the community policing variant; (h) demo- graphic and timeline information; and (i) perceptions of new training needs, and identification of agencies that had been influential in the community policing movement. Specifics regarding these dimensions are displayed in the tables that summarize the results.

This research explores the potential for differences in the above dimensions between police departments and sherifFs offices, large and small agencies, accredited and nonaccredited agencies, as well as differences associated with selected demographic variables. Sheriffs are usually elected officials in Florida, and they may be more amenable to closer ties with the community, or the electorate, than appointed police chiefs. One of the thrusts of community policing is to endow large agencies with some of the benefits of small agencies, and this would require greater structural change in the former. The law enforcement accreditation process prescribes the adoption of spe- cific managerial processes and organizational mechanisms. Accredited agencies may be committed to bureaucratic appurtenances that may constrain their capacity to implement the community policing model. The respondents were also divided into two groups based on the organizational pervasiveness of the adopted variation of community policing (maximal implementers and minimal implementers, as described below), and differences between these groups are also examined. The goal of this research is to determine what is being implemented and by whom, and to identify impacts on organizational structure and operational policies. Perceived problems associated with imple- mentation are also examined in order to gain additional insights into the nature of operational impacts.

policing that precipitate the "bifurcation of authority" that characterizes the typical police organization (Brown, 1981), and the dichotomy between "management cops and street cops" that yields the "two cultures of policing" (Reuss-Ianni and Ianni, 1983; Reuss-Ianni, 1983). In his research, Caiden observed that "the top and bottom of the police department were two different worlds — There was no central control, only an insistence on authority and a pervading sense of powerlessness" (1977, 15).

The strategies of the institutionalized organization are clearly manifested in the community policing move- ment: the promise of future reform; the display of good faith and confidence; the promulgation of ambiguous goals and the implementation of uncertain technologies; the avoidance of evaluation on the basis of efficiency cri- teria in favor of "good faith" criteria; and the idea that the new model is simply a reaction to the failure of the reac- tive model, which severed ties with the community and endangered the legitimacy of the organization.

Does the community policing model represent an attempt to couple operational activities and the environ- ment in which they occur more closely with formal orga- nizational structures in order to enhance the technical efficiency of those activities, or is community policing simply a strategy to buttress the legitimacy of the existing formal structures while leaving them intact? The contin- Public Administration Review • November/December 1998, Vol. 58, No.

6 ued commitment of the police establishment to evalua- tion through bureaucratic displays is evinced by another recent innovation, the police accreditation movement.

Alternatively, if the organization's formal structures are effectively buffered from the operational environment, could it be feasible to implement community policing as simply an operational technology buffered from the for- mal organization? The requirements for participation, openness, empowerment, and decentralization, as well as the lessons learned from the failure of the team policing innovation in the 1970s—strongly resisted by middle managers facing marginalization (Roberg, 1994; Sherman et al., 1973)—^suggest not.

Results Table 1 lists alternative community policing models in order of increasing organizational pervasiveness, and it depicts the percentage of the responding law enforcement agencies employing each model. The adoption rates of the three types of agencies examined are also displayed.

In each case, the most common variant employed was restructuring patrol operations with decentralized substa- tions—a fairly pervasive implementation featuring geo- graphic decentralization and impacts on the patrol func- tion as a whole. The second most common model adopted, except for police agencies, was the addition of specialized sworn patrol units to the traditional patrol function. The percentage of agencies indicating that they had adopted community policing only as a department philosophy is rather small given the fact that community policing is often described as a philosophy rather than a technology. The type of model adopted was not associat- ed with agency type, agency size (measured by number of sworn personnel), or whether the agency was accredited.

Analysis also indicated that there were no statistically sig- nificant relationships among the three agency variables, except that accreditation tended to be more common among large agencies (Cramer's V = .23; significance = .05).

For the remainder of the analysis, the respondents were divided into two groups based on their level of adoption: those adopting variants featuring "philosophy only" or "specialized units" (minimal implementation), and those entailing some level of "restructuring" (maxi- mal implementation). A separate analysis conducted on this variable and a dichotomized version of the date of adoption variable yielded no significant differences in level of implementation associated with date of adoption, so that the possibility that minimal implementation was simply a transitional stage was discounted.

The relative importance of the goals of the community policing model adopted is depicted in Table 2. Goals concerning community relations (closer community rela- tions, problem solving, and public demands for closer relationships) are consistently ranked highest in impor- tance. Goals concerning crime fighting, crime informa- tion gathering, and patrol efficiency follow. Goals reflect- ing the internal structures of the agencies (chain of command, empowering patrol officers, decentralization, internal communications, supervision, culture) are per- ceived as less important. However, the internal structure goals are rated significantly higher in importance in agen- cies manifesting the maximal implementation variants of the community policing model than in those employing minimal implementations; of course, the extent of struc- tural changes was one of the criteria for group member- ship.

Maximal implementers were also more likely to emphasize the development of patrol officer job skills and the reduction of controls on sworn personnel than mini- mal implementers, although the latter relationship was not a significant one. Reducing police misconduct is also a significantly more important goal for maximal imple- menters. Smaller agencies appear to be more concerned with misconduct than larger ones, but this difference is not statistically significant. Police departments, accredit- ed agencies and maximal implementers also appear to be more amenable to changing their missions than their respective counterparts, but these rather large observed differences were not statistically significant. Perhaps most significant here is that all of the goals recorded averages well into the somewhat important range, with the excep- tion of shorten the chain of command. Decentralize operational decisions, loosening needless controls, and providing better supervision were the only other goals to score less than 3.50. Although maximal implementers rated the internal structure goals higher in importance than the minimal implementers, these were still not among their highest rated goals.

Table 1 Community Oriented Policing Model Adopted By ^ency Type (Percentages) COP Model Patrol philosophy only Department philosophy only Administrative unit only Specialized civilian patrol units Specialized sworn patrol units Restructuring of entire patrol operations to COP Restructuring of entire department to COP Restructuring of patrol operations with decentralized substations Restructuring of department with decentralized substations All Agencies (N = 89) 0 3.6 2.4 2.4 24.3 4.7 18.5 1 33.5 10.5 Police (N = 54) 0 3.0 3.0 0 21.9 4.9 23.7 36.9 6.8 Sheriff's (N = 27) 0 4.9 0 4.9 28.0 4.9 8.8 31.9 16.5 Sworn < 147 (N = 43) 0 3.3 3.3 3.3 24.7 3.3 22.3 31.9 8.0 Sworn > 147 (N = 42) 0 4.4 0 0 26.4 6.9 14.2 36.1 11.8 Accredited (N = 35) 0 0 0 0 19.2 7.8 10.7 42.1 19.2 Nonaccr.

{N = 52) 0 5.3 3.3 3.3 27.3 3.3 23.3 29.3 5.3 Reinventing or Repackaging Public Service? The Case of Community-Oriented Policing 489 Table 2 Importance of COP Goals By Agency Type (Means)' Goals Shorten chain of command Empower patrol personnel Closer relations with community Decentralize operational decisions Community problem solving Reducing police misconduct Reducing crime rates Crime information from community Identifying service preferences Increasing patrol efficiency Reducing calls for service Protecting businesses effectively Changing department's mission Developing officer job skills Changing department's culture Public demands for closer relations Increasing resources Better communications within dept.

Loosening needless controls Providing better supervision 1.

1 snot a goal; 2 = not an important goal; 3 ° somewhat important goal; 4 = important goal; 5 = very important goal * T-test p< .05 " T-test/>< .01 All Agencies (N = 89) 2.46 3.83 4.73 3.43 4.63 3.50 4.36 4.06 4.05 4.14 3.87 4.05 3.94 3.85 3.68 4.40 3.81 3.91 3.35 3.44 Police (N=54) 2.43 3.94 4.77 3.42 4.67 3.51 4.31 3.92 3.96 4.04 3.82 3.98 4.25 3.82 3.69 4.49 3.73 3.84 3.25 3.43 Sheriffs (N=27) 2.71 3.68 4.65 3.42 4.58 3.38 4.46 4.38* 4.08 4.32 3.84 4.23 3.38 3.80 3.52 4.23 3.81 3.81 3.50 3.48 Swom <147 (N = 43) 2.23 3.78 4.73 3.23 4.59 3.80 4.37 4.05 4.03 4.18 3.75 3.93 3.44 3.85 3.68 4.43 3.85 4.05 3.13 3.48 Sworn >147 (N = 42) 2.71 3.90 4.80 3.64 4.76 3.23 4.46 4.10 4.05 4.18 3.98 4.24 3.54 3.88 3.68 4.41 3.76 3.73 3.54 3.48 Accredited (N = 35) 2.81 4.00 4.71 3.65 4.74 3.50 4.53 4.21 4.18 4.21 4.09 4.24 4.94 3.88 3.79 4.48 3.73 4.00 3.48 3.39 Nonaccr.

(N = 52) 2.27 3.73 4.74 3.31 4.54 3.49 4.26 3.96 3.96 4.10 3.74 3.90 3.24 3.82 3.60 4.34 3.88 3.84 3.26 3.47 Minimal (N= 25) 2.86 3.32 4.70 2.86 4.57 2.86 4.43 4.00 3.91 3.95 3.83 3.91 2.82 3.18 2.96 4.13 3.65 3.43 2.96 2.91 Maximal (N = 64) 2.94** 4.00** 4.74 3.63** 4.65 3.72* 4.34 4.08 4.10 4.20 3.87 4.10 4.33 4.08** 3.95" 4.49* 3.87 4.08** 3.48 3.61* The changes in policies and procedures associated with the adoption community policing are displayed in Table 3.

Responses in the "other" category usually referred to information collected elsewhere in the questionnaire.

One respondent made reference to a "citizens' police academy." The overall frequency with which policies were changed or added was low, particularly in light of the range of goals recorded above. Changes in internal aflairs procedures and citizen complaint procedures were rela- tively rare, although sheriff's offices reported changes in the latter much more frequently than police departments (this difference was not statistically significant). Shift assignments were affected in a majority of the cases, and dispatch policies and shift rotation procedures were affected in a third of the cases. These changes indicate significant impacts on manpower allocation and deploy- ment schemes. However, the most frequently reported change was in the area of training programs (81.8 per- cent), the only policy area other than shift assignments changed or added by a majority of all agencies. For mini- mal implementers, the area of training programs was the only one cited by a majority of the agencies. This may be Table 3 Policies Changed or Added Due to COP Adoption Policy Area Patrol shift assignments Patrol shift rotation Patrol officer evaluation criteria Requirements for promotion Sworn employment selection criteria Training programs Internal affairs procedures Patrol report writing Patrol dispatch Citizen complaint procedures Patrol overtime Tuition reimbursement programs Job rotation programs Off-duty assignments Other All Agencies (N = 89) 60.2 33.3 30.7 12.5 21.6 81.8 5.7 15.9 33.3 19.3 9.1 8.0 12.5 6.8 9.1 *Cramer'sV/>< .05 "Cramer's V/>< .01 Police (N = 54) 63.0 34.0 33.3 16.7 27.8 81.5 9.3 11.1 35.8 14.8 9.3 5.6 14.8 5.6 11.1 by Agency Type (Percentages) Sheriffs (N = 27) 61.5 34.6 30.8 3.8 3.8 84.6 0 26.9 26.9 30.8 7.7 15.4 7.7 7.7 7.7 Sworn < 147 {N = 43) 58.1 30.2 34.9 16.3 30.7 79.1 4.7 9.3 25.6 11.6 4.7 2.3 11.6 9.3 7.0 Sworn > 147 (N = 42) 61.0 35.0 26.8 7.3 9.8 (.26)* 82.9 4.9 19.5 37.5 24.4 14.6 9.8 9.8 0 (221* 12.2 Accredited (N = 35) 57.1 34.3 37.1 8.6 11.4 88.6 5.7 22.9 41.2 20.0 8.6 11.4 14.3 8.6 8.6 Nonaccr.

(N = 52) 64.7 34.0 25.5 15.7 27.5 78.4 5.9 11.8 27.5 17.6 9.8 3.9 11.8 5.9 9.8 Minimal {N = 25) 45.8 20.8 16.7 0 0 66.7 0 4.2 16.7 12.5 8.3 0 4.2 4.2 8.3 Maximal {N = 64) 65.1 38.7 34.9 17.5 (24)* 28.6 (.32)** 87.3 (.24)* 7.9 20.6 40.3 (23)* 22.2 9.5 11.1 15.9 7.9 9.5 490 Public Administration Review • November/December 1998, Vol. 58, No. 6 Table 4 Pattol Officer Duties by Agency Type (Percentages) Duties Conduct follow-ups to calls Investi^te felonies Propose solutions to community problems Write beat reports Develop strategic plans for community problems Attend read-ofFs Regularly address citizen groups Random patrol in beat or zone Business checks Liaison with crime analysis Forward citizen complaints Crime prevention programs Referrals to social service agencies Organize crime watch groups Set patrol goals and objectives All Agendes (N = 89) 61.4 46.6 85.2 51.1 60.2 62.1 71.6 83.0 83.0 43.7 53.4 68.2 70.5 54.5 39.8 * Cramer's V />< .05 ** Cramer's V />< .01 Police (N = 54) 53.7 48.1 85.2 48.1 64.8 54.7 70.4 83.3 81.5 43.4 51.9 63.0 66.7 50.0 40.7 1 Sheriff's (N = 27) 73.1 42.3 84.6 50.0 53.8 73.1 73.1 88.5 84.6 46.2 61.5 73.1 76.9 57.7 38.5 Sworn <147 (N = 43) 51.2 37.2 83.7 53.5 53.5 61.9 65.1 81.4 81.4 40.5 41.9 69.8 62.8 58.1 32.6 Sworn >147 (N = 42) 68.3 53.7 87.8 48.8 68.3 61.0 78.0 87.8 87.8 46.3 70.7 68.3 80.5 51.2 43.9 Accredited (N = 35) 65.7 51.4 88.6 54.3 57.1 65.7 82.9 , 94.3 82.9 54.3 54.3 71.4 80.0 57.1 37.1 Nonaccr.

(N = 52) 56.9 43.1 82.4 47.1 60.8 58.0 62.7 78^4 82!4 36.0 51.0 64.7 62.7 52.9 39.2 Minimal (N = 25) 50.0 45.8 87.5 62.5 62.5 54.2 83.3 79.2 79.2 50.0 70.8 87.5 79.2 66.7 33.3 Maximal (N = 64) 66.7 47.6 84.1 47.6 58.7 66.1 68.3 84.1 84.1 40.3 47.6 61.9 ( 2S^** 68.3 50.8 41.3 evidence that the law enforcement establishment believes the community policing model simply requires changing the police ofFicer. This emphasis on the development of a new breed of police officer is also supported by the fre- quency of changes in selection criteria (21.6 percent; sig- nificantly greater in smaller agencies than in larger), and patrol officer evaluation criteria (30.7 percent). A sepa- rate analysis of those respondents who had reported a change in training policies indicated that they were more likely to also report changes in evaluation criteria than those reporting no changes in training policies (34.7 per- cent to 12.5 percent); these were also more likely to report changes in promotion criteria (13.9 percent to 6.3 percent) and selection criteria (25 percent to 6.3 per- cent).

Given the low frequency with which changes were made in other policy areas, this may also reflect a "patrol officer change strategy" at work.

The personnel strategy is supported by the fact that the maximal implementers recorded significantly greater impacts than the minimal implementers in the areas of training, selection criteria, and promotion criteria, and a larger if not a significantly different impact in patrol eval- uation criteria. This indicates that structural decentral- ization entails an even greater emphasis on the realign- ment of human resource management systems. The goal of empowering patrol personnel was also associated with adoption of the maximal model, and this goal is clearly related to structural decentralization and obviously calls for adjustments in human resource management systems.

However, it is clear that the human resource systems of police organizations are major targets of their community policing efforts. The law enforcement establishment appears to be relying on the personnel change strategy to implement community policing, and when structural changes are made, this reliance necessarily increases.

The responsibilities and duties of patrol officers are summarized in Table 4. Proposing solutions to commu- nity problems and conducting random patrol and busi- ness checks emerge as the most common duties. In Table 2, community relations goals and crime fighting goals emerged as the highest rated. This manifests an elemen- tal problem in the community policing model—the suc- cessful merging ofthe reactive, 911 emergency response mode, and the proactive, community problem-solving mode. Accredited agencies are less likely to abandon the random patrol function than unaccredited agencies. This relationship does not extend to large versus small agen- cies, even though size and accreditation are positively related. Despite the support recorded by the goals of decentralizing decision making, solving community problems, and empowering patrol officers, relatively few agencies give patrol officers responsibility for setting patrol goals and objectives. This frequency increases by 50 percent, however, when this duty is expressed as "developing strategic plans based on community prob- lems or issues." The chief executives apparently prefer less formal expressions of patrol officer empowerment.

The wide range of duties undertaken by patrol officers is also clearly evident in these data. Twelve of the 15 cate- gories recorded majorities. However, the most striking finding is that patrol officer duties do not vary signifi- cantly with level of implementation, particularly in light of the fact that minimal implementation models often do not directly involve patrol officers. Curiously, patrol offi- cers in minimal implementation agencies are significantly more likely to engage in crime prevention activities, which is a nontraditional, proactive policing function often associated with community oriented policing.

Reinventing or Repackaging Public Service!

The Case of Community-Oriented Policing 491 Table 5 Community Policing Program Elements 1 All Agencies Element (N = 89) Mobilization of citizen groups Citizen review function within dept.

Citizen review function within jurisdiction Internal affairs unit within dept.

Enhanced ordinance enforcement Enhanced truancy enforcement Coordination with other city and county agencies Coordination with state social service agencies Coordination with local schools Regular meetings with citizen groups Decentralization of investigative function to substations Liaison with formal youth services unit Measurement of patrol outcomes Police program evaluation capacity Enhanced codes enforcement Golf cart patrols Horseback patrols Walking beats Bicycle patrols Resource centers/storefronts Citizen satisfaction surveys Community newsletter * Cramer's V /)< .05 ** Cramer's V/i 76.1 17.0 9.1 46.6 n.i 68.2 88.6 81.8 95.5 86.4 21.8 64.8 35.2 20.7 76.1 9.1 19.3 46.6 92.0 56.8 72.7 40.9 <.O1 by Agency Type (Percentages) Police (N = 54) 72.2 18.5 11.1 48.1 77.8 70.4 88.9 77.8 94.4 87.0 9.4 59.3 42,6 20.8 83.3 11.1 18.5 50.0 94.4 64.8 66.7 42.6 SherifFs (N = 27) 88.5 19.2 7.7 53.8 65.4 65.4 92.3 92.6 (.39)" 96.2 88.5 50.0 (.47)" 76.9 23.1 19.2 65.4 (.33)' 7.7 26.9 38.5 88.5 46.2 84.6 38.5 Sworn <147 (N = 43) 72.1 9.3 2.3 32.6 69.8 65.1 79.1 72.1 93.0 79.1 9.5 53.5 27.9 7.1 76.7 4.7 7.0 37.2 93.0 48.8 65.1 34.9 Sworn >147 (N = 42) 82.9 24.4 17.1 63.4 (.31)" 78.0 70.7 97.6 (.29)" 90.2 (.23)' 97.6 95.4 (24)' 34.1 (.30)" 73.2 41.5 29.3 (29)** 78.0 12.2 34.1 (.34)" 61.0 (.24)* 92.7 63,4 80.5 46.3 Accredited (N = 35) 77.1 17.1 11.4 60.0 85.7 80.0 94.3 82.9 97.1 97.1 34.3 74.3 48.6 28.6 82.9 14.3 20.0 51.4 100 60.0 85.7 37.1 Nonaccr.

(N = 52) 74.5 17.6 7.8 35.3 (.24)* 62.7 (.25)' 60.8 84.3 80.4 94.1 78.4 (.27)* 12.0 (.27)* 56.9 25.5 (.24)* 14.0 70.6 5.9 17.6 41.2 86.3 (.25)* 52.9 64.7 (.23)' 43.1 Minimal (N = 25) 79.2 16.7 8.3 29.2 79.2 62.5 91.7 79.2 95.8 91.7 20.8 58.3 12.5 12.5 83.3 0 33.3 45.8 100 54.2 58.3 41.7 Maximal (N = 64) 74.6 15.9 9.5 54.0 (.22)* 69.8 71.4 87.3 82.5 95.2 85.7 22.6 66.7 44.4 (.30)** 24.2 73.0 12.7 14.3 (.22)* 47.6 88.9 57.1 77.8 41.3 The elements of the adopted community policing models are summarized in Table 5. Bicycle patrols appear to be a component of virtually all community ori- ented policing programs. This element alone would limit the definition of community policing to the adoption of a new technology that can support community policing efForts, directed patrol strategies, or the traditional ran- dom patrol function. Channels for coordinating activi- ties with other government agencies, citizen groups, and other divisions within the agency also appear to be widespread, particularly coordination with local schools.

Police functions not traditionally related to crime fight- ing, such as ordinance enforcement, code enforcement, and truancy enforcement, also appear to be integral parts of a majority ofthe programs. Formal structures for citi- zen review of police activities and the implementation of an internal investigative function appear to be less widespread. Larger agencies are more likely to have both than smaller agencies, and accredited agencies are more likely to have an internal affairs division. The vast major- ity of agencies schedule regular meetings with citizen groups, particularly accredited and larger agencies.

Almost three quarters of the agencies survey their citizens regarding their satisfaction with police services, and 40 percent publish a newsletter. Clearly, these methods of securing citizen input are preferred over formal citizen review boards. The capacity to measure patrol outcomes and evaluate police programs also appears to be relatively rare.

Larger agencies are much more likely to evaluate than smaller agencies, and accredited agencies and those with maximal implementations are more likely to moni- tor outcomes.

Data on resistance to adoption from tbe point of view of tbe chief executive are summarized in Table 6. Over- all, law enforcement unions and patrol officers are ranked as tbe sources of strongest resistance, followed closely by first-line supervisors. However, tbese averages still fall on the supportive side of tbe scale. Perceived supervisor resistance exceeds patrol officer resistance in large agen- cies and accredited agencies, as well as in maximal imple- mentation agencies. In tbe latter case, tbis may be evi- dence of at least tbe perception of real structural change 492 Public Administration Review • November/December 1998. Vol. 58, No. 6 Table 6 Resistance to COP Adoption By Agency Type (Means)' Sources of Resistance Patrol officers Investigators Elected officials First line supervisors Unions Middle managers Business community Citizen groups Top managers Minority community Civilian personnel 1.

1 = very supportive; All Agencies (N = 89) 2.65 2.51 1.62 2.63 2.71 2.31 1.55 1.41 1.52 1.72 2.32 2 = supportive; 3 = neutral; Police (N = 54) 2.71 2.62 1.62 2.68 2.70 2.36 1.56 1.48 1.55 1.78 2.35 Sheriffs (N = 27) 2.44 2.24 1.67 2.44 2.74 2.25 1.48 1.24 1.48 1.63 2.24 4 = not supportive; 5 Sworn <147 (N = 43) 2.71 2.48 1.55 2.49 2.69 2.16 1.52 1.45 1.51 1.85 2.29 = very resistant Sworn > 147 (N = 42) 2.56 2.62 1.68 111 2.75 2.53 1.54 1.36 1.56 1.61 2.38 Accredited (N = 35) 2.73 2.61 1.59 2.82 2.75 2.58 1.48 1.24 1.59 1.75 2.18 Nonaccr.

(N = 52) 2.62 2.46 1.64 2.48 2.69 2.16 1.62 1.54 1.50 1.69 2.45 Minimal (N = 25) 2.83 2.54 1.67 2.63 2.90 2.38 1.54 1.46 1.63 2.04 2.39 Maximal (N = 64) 2.58 2.50 1.59 2.64 2.62 2.28 1.57 1.40 1.49 1.61 2.30 tbat tbreatens tbe role of middle management. However, the difference in ranking is due to tbe reduction of resis- tance on tbe part of patrol officers in agencies witb maxi- mal implementations ratber tban a large increase in tbe level of supervisor resistance in tbose agencies. Overall, in fact, tbe various agency types manifested a similar pat- tern, and analysis identified no statistically significant relationships between any of agency variables and per- ceived resistance in any of tbe identified groups. Curi- ously, tbe chief executives of maximal implementation agencies tended to report lower levels of perceived resis- tance than those from minimal implementation agencies, but, once again, these differences were not significant.

Tbe resistance of patrol officers can be understood in light of the wide ranging duties they are asked to assume under most community policing models, as well as the fact that most programs appear to focus on remaking the patrol officer.

The data regarding perceptions of operational prob- lems associated with adoption are displayed in Table 7.

Funding and service demands emerge as the highest ranked problem areas overall, but service quality does not appear to be suffering as a result. Smaller agencies and police departments reported significantly greater prob- lems with unreal expectations associated with adoption than did their counterparts, but tbese were not rated as severe. Problems with managerial styles were also consis- tently reported, although these, too, were not ranked as severe problems. Funding problems, paperwork prob- lems, and problems witb patrol supervision were more serious in maximal implementation agencies than in min- imal implementation agencies, but, once again, these dif- ferences were not statistically significant. Planning capac- ity was among the lowest rated problems areas, altbough in Table 5 program evaluation capacity and measurement of patrol outcomes were not widespread. One could speculate that if real structural change were occurring, planning capacity would be strained. It would also appear tbat law enforcement executives bave little interest in the application of structured analysis to their pro- grams.

None of the three agency type variables emerged as a Table 7 Operational Problems Associated with COP Adoption By Agency Type (Means)' Operational Problem Areas Funding Community relations Patrol accountability Conflicting goals Managerial styles Equipment Training Unclear mission Patrol supervision Unreal expectations Sworn personnel Service demands Service quality Paperwork load All .^encies (N = 89) 2.38 1.21 1.77 2.02 2.31 1.78 1.91 2.02 2.06 1.85 1.91 2.35 1.52 1.93 Interdepartmental communication 2.15 Planning capacity 1.85 Police (N = 54) 2.28 1.18 1.75 2.06 2.43 1.72 1.94 2.13 2.04 1.98 1.94 2.37 1.52 1.90 2.17 1.94 Sheriffs (N = 27) 2.67 1.20 1.75 1.88 2.08 1.96 1.84 1.72 1.96 1.52* 1.60 2.20 1.48 1.84 2.12 1.72 1.1 = not a problem; 2 = slight problem; 3 = problem; 4 = severe problem; * T-test /)<.O5 Sworn <147 (N = 43) 2.23 1.27 1.71 2.14 2.42 1.67 1.88 2.23 2.09 2.07 2.02 2.26 1.52 1.98 2.21 1.91 Sworn > 147 (N = 42) 2.50 1.18 1.81 1.95 2.23 1.92 1.92 1.85 2.03 1.64* 1.82 2.46 1.51 1.87 2.15 1.82 Accredited (N = 35) 2.58 1.27 1.88 2.12 2.47 1.71 1.88 2.06 2.24 1.85 2.03 2.55 1.61 2.00 2.24 1.88 5 = severe and continuing problem Nonaccr.

(N = 52) 2.24 1.18 1.73 1.98 2.24 1.84 1.96 2.02 1.94 1.86 1.84 2.28 1.48 1.92 2.10 1.84 Minimal (N = 25) 2.08 1.13 1.78 1.96 2.38 1.79 1.88 2.13 1.87 1.83 1.79 2.33 1.58 1.75 2.04 1.79 Maximal (N = 64) 2.48 1.25 1.78 2.05 2.30 1.79 1.92 2.00 2.13 1.85 1.97 2.35 1.48 2.00 2.20 1.87 Reinventing or Repackaging Public Service!

The Case of Community-Oriented Policing 493 major determinant of any of tbe seven program areas examined. Data on the race and sex of the chief execu- tive officer did not manifest the variance required for meaningful analysis. The executive's years of education, years in office, years with the agency, and years in law enforcement were not significantly related to level of implementation. The executives were asked to identify what they perceived to be the top five training needs associated with community policing. Fifty-five identified one or more needs. It was often difficult to group idiosyncratic titles of training modules, but the most fre- quent areas identified were: problem-solving techniques (28 citations); cultural diversity (19); interpersonal com- munication (14); community relations/mobilizing com- munity involvement (10); "basic" community oriented policing (8). These reflect the primary goals of commu- nity engagement and problem solving cited above, as well as the idea that community policing simply entails the retraining of the patrol officer.

Discussion Who is adopting community policing in Florida?

Everyone! What are they adopting? The range of goals identified by the chief executives of the agencies would lead one to believe they are adopting something for everyone. No single, definitive model of community policing emerged from this research. The most common model adopted by local law enforcement agencies in Florida entailed the restructuring of the patrol function and the geographic decentralization of patrol operations to substations. The second most frequently adopted vari- ant was the simple addition of specialized community policing units to the existing patrol function. The model adopted was not related to agency size, although the research literature Implies that community policing would entail greater structural and procedural changes for larger agencies. Police agencies and sheriff's offices exhib- ited similar adoption patterns, despite the fact that the chief executives of the latter are elected. We had specu- lated that elected officials might be more amenable to closer relationships with the community, but sheriff's offices may have been more decentralized to begin with, and hence, would not have adopted the more structurally pervasive models at a greater rate than police depart- ments. Neither were adoption patterns associated with the accreditation status of the agency. The accreditation movement focuses on the professionalism of the law enforcement organization, and we had posited that a commitment to bureaucratic processes and administrative systems could potentially confiict with the values of police officer empowerment and devolution of com- mand.

Our results, then, confirmed the ambiguity and elas- ticity of the community policing concept, which are well documented in the research literature and often presented as strengths of the concept. Some theorists have charac- terized community policing as a "philosophy" of policing with no attendant structural or procedural elements.

Others hold that structural decentralization and the adoption of organic organizational forms are essential ele- ments of the concept. We presented the respondents with nine models of community policing, each of which entailed varying degrees of structural change and geo- graphic decentralization. We had asked them to indicate which model most accurately described the approach of their agencies, and the bi-modal distribution described above emerged. In an effort to identify the actual impacts of community policing on the policies, proce- dures, and operations of the adopting agencies, we divid- ed our sample into two groups based on the organization- al pervasiveness of the model adopted. Minimal implementers merely adopted a philosophy or added spe- cialized units to their existing service delivery structure; maximal implementers reported some geographic decen- tralization of operations, and/or restructuring of their overall patrol operations or the operations of the entire agency. If real organizational change is associated with the implementation of community policing, maximal implementers should report more extensive organization- al impacts than minimal implementers.

Once again, implementation group membership was not associated with any of the three agency type variables.

Maximal implementers rated the community policing program goals related to the internal structure of the organization as more important than the minimal imple- menters. However, this difference in the relative impor- tance of goals was not reflected in differences in impacts on existing policies and procedures, or on patrol officers duties, or on the substantive elements of the community policing programs. In fact, both groups rated all of the listed goals as important, and the maximal implementers rated all but one higher than the minimal implementers (the five internal structure goals were among only nine rated higher that were statistically significant). The inci- dence of impacts was low for most policies and proce- dures in both groups. Training programs emerged as the most frequently affected policy area (the only one record- ing a majority among the minimal implementers, and one of two recording a majority among the maximal implementers). Both groups appeared to be adding com- munity policing duties to traditional patrol duties. For example, a vast majority of patrol officers in both groups continued to provide random patrol in addition to proposing solutions to community problems.

How is it that agencies implementing community policing models entailing varying degrees of structural change can manifest so few differences in program sub- stance, patrol duties, and impacts on operational policies and procedures? It may be due to the fact that the major- ity of these agencies are seeking to implement communi- ty policing by changing the officer rather than the organi- zation. The maximal implementers reported dramatically more frequent impacts on training programs, promotion- 494 Public Administration Review • November/December 1998, Vol. 58, No.

6 al criteria, and selection criteria than the minimal imple- menters, and we found evidence that both groups are relying on a patrol officer change strategy to implement their versions of community policing. There were also some indications that some respondents may have exag- gerated the extent of structural change associated with their models. In two cases observed in closer detail, the degree of reported decentralization may have been over- stated: one that reported decentralization to substations had actually implemented storefront operations that did not house command responsibilities. Another that reported decentralization had actually shortened a very long chain-of-command at the very top when two assis- tant chiefs and three deputy chiefs became three assistant chiefs, followed by the majors, lieutenants, and sergeants.

Alternatively, following Meyer and Rowan (1983), decen- tralization may be viewed as an attempt to further buffer the formal organization from the potential environmental threats associated with community policing as police offi- cers take ownership of community problems. Field oper- ations and the formal organization are so loosely coupled that geographical decentralization can occur without dis- rupting the formal policies and bureaucratic systems that define the latter. We have suggested that the failure of past innovations indicates that such an arrangement does not bode well for the long-range prospects of community policing.

The potential for real geographic decentralization and command devolution may also be limited by resource constraints. Maximal implementers reported more prob- lems with funding than minimal implementers, as well as more frequent problems with supervision. This may be due less to resistance from supervisors who fear being made superfluous than to the perceived unmet need to provide greater levels of on site supervision with increased decentralization. However, perceived resistance from supervisors did exceed resistance from officers among maximal implementers. Although this result was not sta- tistically significant, it may indicate that supervisors short-circuit the implementation of supportive policies and procedures because they oppose decentralization and command devolution. In any case, except for changes in training programs, the models featuring restructuring and/or geographic decentralization were not accompa- nied by any widespread realignment of existing opera- tional policies and procedures.

Although this research examined only a limited num- ber of possible factors determining the adoption of com- munity policing, the initial conclusion here is that imple- mentation occurs in the context of local history and political culture, and these can be highly idiosyncratic.

This view is reflected in the comments of an early respon- dent to the survey who did not complete the question- naire, but rather appended a two page letter explaining that his agency had been operating in a community policing mode for quite some time. Their approach, however, reflected the agency's ongoing response to the Maximal ^ implementers reported more problems with funding than minimal implementers, as well as more frequent problems with supervision.

particular needs of the citizens of the jurisdiction rather than the conscious adoption of any of the options listed in the questionnaire. For this respondent, community policing represented the capacity to respond to those needs.

This view reflects Shaw's (1995) contention that community policing programs should be judged on their capacity to produce organizational forms and processes that provide for continual innovation in the operation of law enforcement service delivery systems—that is, forms that harness the creative potential of the agency's internal constituency to respond to the enunciated needs of its external constituency.

From this perspective, one would expect to observe some structural decentralization, the empowerment of patrol officers to use discretion in service delivery, and an openness to information from the environment.

Researchers have identified decentralization of the com- mand structure as an essential element of community policing (Skolnick and Fyfe, 1995; Bayley, 1994a; Kelling and Moore, 1991). Two-thirds of our respondents char- acterized their models as entailing some restructuring and/or decentralization, but we have indicated that the extent may have been exaggerated and restructuring is not accompanied by the realignment of operational poli- cies and procedures. In regard to the empowerment issue, patrol officers were often charged with responsibil- ity for developing solutions to community problems, but less often allowed to set patrol goals and objectives.

Patrol officers may be empowered to engage the commu- nity, but the organization will continue to engage its per- sonnel through existing bureaucratic mechanisms. The overall approach seems to entail changing the officer before changing operational policies, administrative pro- cedures, and command hierarchy.

Empowerment and openness must proceed hand in hand in order to ensure accountability (Trojanowicz and Bucqueroux, 1990). Some respondents reported concern regarding potential police misconduct, but very few appear to be amenable to the idea of formal citizen over- sight. Many programs feature regular meetings with citi- zen groups, mobilization of citizen groups, citizen satis- faction surveys, and community newsletters. In about half the cases, patrol officers are responsible for forward- ing citizen complaints, but existing citizen complaint procedures were affected by only about 20 percent of the community policing programs. Restructuring and decen- tralization were widely reported, and these imply the devolution of discretionary decision making to the police officer; however, if police officer empowerment is occur- ring, law enforcement executives seem reluctant to open Reinventing or Repackaging Public Service? The Case of Community-Oriented Policing 495 their agencies to the community oversight necessary to provide for accountability. Patrol accountability emerged as one of the lowest rated problem areas for both the minimal and maximal implementers.

The responding agencies are eager, however, to coor- dinate their efforts to address community problems with other relevant public agencies. Bayley (1994a) observed that this coordination function can serve to empower the local law enforcement agency as a "one-stop-shopping" mega-store for public services. Efforts to empower patrol officers are also limited by the fact that these, and addi- tional duties associated with community policing, are apparently being added to their traditional 911 emergen- cy response and crime fighting activities. The collection of information from the community appears to reflect the need for intelligence to support crime fighting, as suggested by Buerger (1994), as much as it does the desire to identify community problems and service pref- erences. The reduction of crime rates and the collection of crime information emerged as among the most highly rated goals of the community policing programs. Ran- dom patrol activities and business checks on the one hand, and proposing solutions to community problems on the other, were the three most frequently cited duties of patrol officers. This may explain why the greatest level of resistance to the adoption of community policing has come from sworn personnel and their unions.

It does not appear that the essential elements of com- munity policing as described in the research literature are finding their way into the Florida programs. Walker (1992) contends that community policing programs have been plagued by poor planning and weak implementa- tion. The measurement of patrol outcomes and a formal program evaluation capacity were not elements of the Florida programs, although both were more prevalent in agencies that were maximal implementers. At the same time, planning capacity was among the lowest rated problem areas. Apparently, these agencies have neither the capacity nor the willingness to evaluate the relative success of their programs. Is the general public being asked, once again, to accept community policing as a good faith effort of the police establishment to reform policing? This good faith effort is evidenced by the fact that all of the goals of the community policing programs are rated as at least somewhat important with the excep- tion of shortening the chain of command.

If community policing is all things, then what is it? Bicycle patrols emerged from this study as a universal technology of community policing. Bicycles may simply be a more effi- cient way to deliver traditional patrol service in down- town areas and public housing projects, and bicycle patrols also facilitate intelligence gathering to support crime fighting. As a symbol of community policing, bicycle patrols characterize it as a simple repackaging of traditional policing in new technological clothes.

Some of these findings may be attributed to the rela- tive newness of the majority of the programs; over half had been operational for less than four years. Policy changes may not have been completed, resistance may not have crystalized, and problems may not have had time to make themselves manifest. However, analysis yielded no relationships between date of adoption and level of implementation. Although this was not an explicit goal ofthe research, little promise of evolutionary reform was identified. The relative newness ofthe major- ity of the programs may also be evidence that in some cases the community policing model was adopted simply to take advantage of the $30 billion crime bill passed by the Congress in 1994, which virtually mandated the development of a community policing plan by local law enforcement agencies. Funding problems was the highest ranked operational problem associated with adoption, and increasing resources was ranked higher as a goal than providing better supervision, lessening controls on sworn officers, changing department's culture, decentralization of operational decision making, and shortening the chain of command, and almost as high as empowering patrol personnel, and developing patrol officer job skills. It is unlikely that federal funds to add new officers can sustain the momentum of the community policing movement, and any cosmetic changes made by local law enforcement agencies in pursuit of these funds will fade away (leaving additional fiscal burdens on the local governments that must fully fund the new positions after three years). No funding source can simply mandate the "metamorphosis of management systems" necessary to support innovation in the delivery of public services called for by Olsen (1997).

Conclusion Community policing is a lot of things, and all the agencies polled here are adopting at least one of them.

No agency, however, has affected existing operational policies and administrative procedures across the board, and even those professing structural change and geo- graphic decentralization seem to focus on changing the officer rather than the organization. The substance of the typical community policing program in Florida seems to center on changing the attitudes and behaviors of the patrol officer in order that he or she can more effectively engage the public and collect information on community problems and service preferences, as well as criminal intelligence, and steer citizens to public agencies offering relevant services. This is accompanied by some geo- graphic decentralization of the patrol function, but there is less evidence of the devolution of decision making authority. Thus, the organizational impacts of communi- ty policing have been minimal. Personnel policies have been changed in order to develop and support new atti- tudes and behaviors on the part of patrol officers. This is necessary in order to make the officers more amenable to assuming additional duties, as well as more competent to carry them out. The public is polled, but kept at arm's 496 Public Administration Review • November/December 1998, Vol. 58, No. 6 length from the formal organization. Hence, the changes in operational policies and administrative systems, the devolution of decision making and authority, and the engaging of the community that would signal reinvention were not observed. What was observed was a repackaging of policing in terms of the remaking of the patrol officer and additions to the basic patrol service package.

The results of this research support the characteriza- tion of the local law enforcement agency as one of Meyer and Rowan's (1983) institutionalized organizations. We witnessed the implementation of an ambiguous technolo- gy through an organizational structure better suited to the operationalization of known cause and effect relation- ships.

The formal structure of the organization is decou- pled from the realities of day to day operations, and this makes it possible to change the technologies at the bot- tom without disturbing the top. There is also the promise of future reform and the highlighting of good faith efforts, but the community is kept at bay despite whatever good faith efforts it may manifest. Community policing emerges as a ritual in the ceremony that is police management—a ceremony apparently designed to shield the organization from the formal evaluation of its tech- nologies in order to protect the organization's flow of resources.

Klockars (1991) contends that the promise of commu- nity policing may be ephemeral given the social anomie that characterizes many urban communities; the police are helpless to serve communities that do not really exist.

But the penetration of these urban areas by a better bureaucracy wielding new bureaucratic tools—particular- ly a bureaucracy housing the coercive power of the state—can only serve to exacerbate social alienation. The community policing movement is reaching out to the public through newsletters and surveys, as well as through the efforts of community oriented patrol offi- cers with little formal discretion to meet its enunciated needs.

The power of the empowered patrol officer is constrained by the maintenance of accountability to an extended chain of command and its administrative sys- tems.

A real community policing would call for citizens engaging their police agencies politically, rather than the agencies engaging the public bureaucratically. This approach would also provide oversight of officers who were truly empowered to meet community needs. Cer- tainly, none of this was observed in this study.

The conclusion here is that the general tenets of the reinvention movement may prove difficult to opera- tionalize in specific substantive service areas. The required operational and structural changes cannot sim- ply be mandated through legislation like the 1994 anti- crime bill, or the Government Performance and Results Act of 1993.

Innovators must consider the political con- text, reform history, and technological issues in each ser- vice, and obstacles associated with each of these areas must be identified and acknowledged before they can be overcome. Of course, reinvention does not occur over night. Klockars (1991) recounts a police practitioner commenting that criticizing community policing is akin to criticizing the song selections of a singing dog—the important point is that the dog is singing! But this canine has sung before, and the captious might at least be some- what mollified by a chorus of structured analysis and on- going evaluation. This is a legitimate and necessary role for professional bureaucracy, but it is one that law enforcement organizations apparently avoid.

• • • Gerasimos A. Gianakis is an assistant professor in the Public Administration Department at the University of Central Florida, where he teaches courses in organization- al theory, budgeting, and financial management. His research on the organization of policing has appeared in the Journal of Criminal Justice, the American Journal of Criminal Justice, and the American Review of Public Administration. He was formerly a management analyst with the St. Petersburg, Florida, Police Department and a budget manager with St. Petersburg's Budget Department.

G. John Davis III is the chief of police for the City of St. Petersburg, Florida. He holds a doctoral degree in criminology from the Florida State University, and he is an adjunct professor with the Department of Criminal Justice at the University of South Florida, where he teach- es courses in statistical applications in criminal justice and police management.

References Bayley, David H. (1994a). "International Differences in Com- munity Policing." In Dennis P. Rosenbaum, ed.. The Chal- lenge of Community Policing. Thousand Oaks, CA: Sage, 278-281.

(1994b). Police for tbe Euture. Oxford: Oxford Uni- versity Press.

Bittner, Egon (1970). The Eunctions of Police in Modem Soci- ety. Washington, DC: National Institutes of Mental Health.

Brown, Michael K. (1981). Working the Street: Police Discretion and the Dilemmas of Reform. New York: Russell Sage Publi- cations.

Buerger, Michael E. (1994). "The Limits of Community." In Dennis P. Rosenbaum, ed.. The Challenge of Community Policing. Thousand Oaks, CA: Sage, 270-273.

Caiden, Gerald E. (1977). Police Revitalization. Lexington, MA: Lexington Books.

Crank, John P. (1994). "Watchmen and Community: Myth and Institutionalization in Policing." Law and Society Review28(2): 325-351.

Eck, John E., and Dennis P. Rosenbaum (1994). "The New Police Order: Effectiveness, Equity, and Efficiency in Com- munity Policing." In Dennis P. Rosenbaum, ed., Tbe Chal- lenge of Community Policing. Thousand Oaks, CA: Sage, 3- 26.

Reinventing or Repackaging Public Service?

The Case of Community-Oriented Polidng 497 Coldstein, Herman (1990).

Problem-Oriented Policing.

New York: McGraw-Hill.

(1977).

"Improving Policing: A Problem-Oriented Approach." Crime and Delinquency 25 (April): 236-258.

Creene, Jack R, and Ralph B. Taylor (1991). "Community- Based Policing and Foot Patrol: Issues of Theory and Evalu- ation." In Jack R. Creene and Stephen D. Mastrofski, eds..

Community Policing: Rhetoric or Reality. New York:

Praeger, 195-223.

Creene, Jack R., and Stephen D. Mastrofski, eds. (1991).

Community Policing:

Rhetoric or Reality.

New York:

Praeger.

Kelling, Ceorge L, and Mark H. Moore (1991). "From Politi- cal Reform to Community: The Evolving Strategy of Police." In Jack R. Creene and Stephen D. Mastrofski, eds..

Community Policing:

Rhetoric or Reality.

New York:

Praeger, 3-25.

Klockars, Carl B. (1991). "The Rhetoric of Community Polic- ing." In Jack R. Creene and Stephen D. Mastrofski, eds..

Community Policing:

Rhetoric or Reality.

New York:

Praeger, 239- 258.

Mastrofski, Stephen D. (1991). "Community Policing as Reform: A Cautionary Tale." In Jack R. Greene and Stephen D. Mastrofski, eds..

Community Policing:

Rhetoric or Reality.

New York: Praeger, 47-67.

Meyer, John W., and Brian Rowan (1983). "Institutionalized Organizations: Formal Structure as Myth and Ceremony." In John W. Meyer and Richard Scott, eds..

Organizational Environments:

Ritual and Rationality.

Beverly Hills, CA:

Sage, 21-44.

Moore, Mark H. (1994). "Research Synthesis and Policy Implications." In Dennis P. Rosenbaum, ed., Tbe Challenge of Community Policing.

Thousand Oaks, CA: Sage, 285- 299.

Olsen, Raymond T. (1997). Outcomes Management Requires Realignment." PA Times 20(2):

10-11.

Osborne, David, and Ted Gaebler (1992).

Reinventing Govem- ment.

New York: Addison-Wesley.

Reuss-Ianni, Elizabeth, and R A. J. Ianni (1983). "Street Cops and Management Cops: The Two Cultures of Policing." In Maurice Punch, ed.. Control in the Police Organization.

Cambridge, MA: MIT Press, 251-274.

Reuss-Ianni, Elizabeth (1983). The Two Cultures of Policing.

New Brunswick, NJ: Transaction Books.

Roberg, Roy R. (1994). "Can Today's Police Organization Effectively Implement Community Policing?" In Dennis P.

Rosenbaum, ed.. The Challenge of Community Policing.

Thousand Oaks, CA: Sage, 249-257.

Roberg, Roy R., and Jack Kuykendall (1993).

Police and Soci- ety.

Belmont, CA: Wadsworth.

Rosenbaum, Dennis P., ed. (1994). The Challenge of Commu- nity Policing.

Thousand Oaks, CA: Sage.

Sadd, Susan, and Randolph Grinc (1994). "Innovative Neigh- borhood Oriented Policing: An Evaluation of Community Policing Programs in Eight Cities." In Dennis P. Rosen- baum, ed..

The Challenge of Community Policing.

Thousand Oaks, CA: Sage, 27-52.

Schwartz, A., and S. N. Clarren (1977). The Cincinnati Team Policing Experiment:

A Summary Report.

Washington, DC:

The Urban Institute and Police Foundation.

Shaw, J. Wiford (1993). "Community Policing to Take Guns off the Street." Behavioral Science and the Law 11 (Autumn): 361-374.

Sherman, Lawrence W., Catherine H. Milton, and Thomas V.

Kelley (1973).

Team Policing:

Seven Case Studies.

Washing- ton, DC: Police Foundation.

Skolnick, Jerome H., and James J. Fyfe (1995). "Community Policing Would Prevent Police Brutality." In Paul A. Win- ters, ed..

Policing the Police.

New York: Greenhaven Press, 45- 55.

Trojanowicz, Robert C. (1994).

Community Policing:

A Survey of Police Departments in the United States.

Lansing, MI:

National Center for Community Policing, Michigan State University.

Trojanowicz, Robert C, and Bonnie Bucqueroux (1990).

Community Policing:

A Contemporary Perspective.

Cincin- nati, OH: Anderson.

Trojanowicz, Robert C, and H. A. Harden (1985). The Status of Contemporary Community Policing.

East Lansing, MI:

Michigan State University.

Turner, Robyne, and Michael B. Wiatrowski (1995). "Com- munity Policing and Community Innovation: The New 'Institutionalism' in American Government." In Peter C.

Kratcoski and Duane Dukes, eds.. Issues in Community Policing Cincinnati, OH: Anderson, 261- 270.

Walker, Samuel (1992). The Police in America: An Introduction.

2nd ed. New York:

McGraw- Hill.

Weatheritt, Mollie (1991). "Community Policing: Rhetoric or Reality?" In Jack R. Greene and Stephen D. Mastrofski, eds..

Community Policing:

Rhetoric or Reality.

New York:

Praeger, 153-175.

Wilkinson, Deanna L., and Dennis P. Rosenbaum (1994).

"The Effects of Organizational Structure on Community Policing: A Comparison of Two Cities." In Dennis P.

Rosenbaum, ed.. The Challenge of Community Policing.

Thousand Oaks, CA: Sage, 110- 126.

Wycoff, Mary Ann (1991). "The Benefits of Community Policing: Evidence and Conjecture." In Jack R. Greene and Stephen D. Mastrofski, eds..

Community Policing:

Rhetoric or Reality.

New York: Praeger, 103-120.

498 Public Adnunistration Review • November/December 1998, Vol.

58, No.

6