Public Organization Review

, Volume 9, Issue 3, pp 213–234

Emergent Management Strategies in a Public Agency: A Case Study of Alternative Fuel Vehicles


DOI: 10.1007/s11115-009-0081-z

Cite this article as:
Jacob, B., Welch, E. & Simms, T. Public Organ Rev (2009) 9: 213. doi:10.1007/s11115-009-0081-z


The ability of public organizations to invest in emerging technologies is dependent upon the degree to which they can effectively manage the risks of being a lead-user in a political environment. However, little is known about the dimensions and implications of the different forms of risk faced by innovative public organizations as well as the strategies employed to manage them. This paper addresses these issues by studying how one public agency implements a program of replacing its transportation fleet with alternative fuel vehicles (AFVs).


Coping Organizational learning Emergent strategies Alternative fuel vehicles Public sector management 


Research on the adoption and integration of emerging technologies in the public sector has focused primarily on information technology and electronic government (Moon 2002; Brown and Brudney 2004; Welch and Pandey 2007). Serving as portals for citizens to access government services, these technologies are extremely visible. However, other technologies have been introduced in less visible areas of public agencies such as: transportation, waste water treatment, and recycling. These ‘quiet’ advances are important to understand for a variety of reasons. For example, in some cases public sector organizations are the first places that new technologies are introduced making them influential for the ongoing end-user innovation process (von Hipple 2005). Also, these new technologies force public agencies to address problems and issues that arise from related risks (Deyle 1994). Thus, studying the implementation process for new innovations, offers insights to broader questions of risk management in the public sector.

Organizations that adopt new technologies face a host of uncertainties regarding the technology’s effectiveness, reliability and costs. In the public sector, these uncertainties are exacerbated by the additional constraint of public accountability (Mulgan 2000). Hence, the ability of public organizations to invest in emerging technologies is fundamentally dependent upon the degree to which it can effectively manage the risks of being a lead-user in a political environment. That said, little is known about the dimensions and implications of the different forms of risk faced by innovative public organizations as well as the strategies employed to manage them. This paper addresses these issues by studying how one public agency implements a program of replacing its transportation fleet with alternative fuel vehicles (AFVs).

The extant literature describes two viable strategies for managing uncertainty in public organizations: planned and emergent. Planned strategies are, typically, developed prior to the adoption of the innovation and are the result of a rational analytic process. Implementation is articulated in detailed plans that leave managers with little room for discretionary or opportunistic behavior (Golden 1990; Deyle 1994). In contrast, emergent strategies are characterized by experiential learning (Mintzberg and Waters 1985). The goals and objectives of a particular program are clearly established during policy adoption, but implementation is left deliberately vague. Emergent strategies arise from public managers ‘coping,’ (Chia and Holt 2006; Armistead and Clark 1994) ‘improvising’ (Weick 1998) or ‘groping along’ (Behn 1988). Through a process of trial-and-error, public managers learn which strategies and actions are more, or less likely, to lead to a policy’s successful implementation (Behn 1988; Golden 1990; Sanger and Levin 1992). This allows public officials to “evaluate and refine their initial program after trying it out, rather than employing extensive analysis and a set of alternatives before launching an initiative” (Deyle 1994, 459). However, the literature is unclear about how the lessons gleaned from trial-and-error are translated and communicated within the organization and subsequently influences managerial decisions. Additionally, while a growing body of evidence suggests that emergent strategies are more successful for the implementation of innovative policies and programs (Behn 1988; Golden 1990; Sanger and Levin 1992), these studies are based upon innovations that have been implemented in relatively ‘low-risk’ environments, which makes generalization to situations of high uncertainty difficult (Deyle 1994). In short, the literature offers few insights with respect to how emergent strategies can be successfully applied to public organizations at the cutting edge of technology.

We address the above shortcoming in two ways. First we develop a model of emergent strategies that explicitly accounts for ‘learning’ and its relationship to the uncertainty and risk faced by innovative public organizations that must adapt and innovate around emerging technologies. Second, we test this model by studying an application of emergent strategies in one public agency—the Forest Preserve District of DuPage County (FPDDC), Illinois—as it implements a program of replacing approximately one hundred vehicles from traditional gasoline vehicles to Alternative Fuel Vehicles (AFVs).

AFV fleet conversion presents a compelling example to study emergent management strategies of innovations in public organizations. First, because conversions are voluntary there are no established protocols or regulations that the organization is mandated to follow. Second, the market for AFV technology remains unsettled and support from various state and federal grants (e.g. Federal Congestion Mitigation and Air Quality, State of Illinois Biodiesel rebate, and Illinois Alternate Fuels Rebate) is unpredictable. Government funding of AFV programs can change as political priorities shift. Third, fleet services agencies are accountable to political bodies and ultimately to citizens for AFV related decisions and outcomes. The ‘trials’ and ‘errors’ that characterize the implementation process for emerging technology implementation must be justified to a broad set of stakeholders that have different demands and may be less patient with delays in outcomes due to a process of learning (Deyle 1994). Thus, studying the details of an emergent strategy in the context of AFV fleet conversion will provide novel insights with respect to how new technologies can be successfully adopted in public organizations.

The paper begins by first presenting a description of the FPDDC. It then discusses the relevant literature, propositions and methods. The findings section describes deviations from the initial strategic plan set forth at the outset of the FPDDC’s AFV program, provides an explanation for these changes in terms of trial-and-error learning, and points to the emergent dimensions of the AFV strategy. A concluding section outlines the implications for theory and practice.

Forest preserve district of DuPage County

The FPDDC has a traditional commissioner form of government with seven members elected to staggered four-year terms. It owns and manages more than 24,000 acres of property and has an annual budget of more than $175 million. Prior to 1999, each of the twelve departments in the FPDDC was responsible for the budgeting and purchasing of their own vehicles while the Fleet Services Department provided only routine service and maintenance of department vehicles.

As part of a re-organization in 1999, all vehicle-related responsibilities were centralized within the Fleet Services Department. Subsequent to this change, the department assumed responsibility for the implementation of the AFV program. The re-structured Fleet Services Department was relatively small, only employing about eight mechanics at any given time, but operating with a high degree of autonomy with respect to the larger political environment. The Fleet Services Department of the FPDDC is managed by a fleet manager and a shop foreman. Together they are responsible for day-to-day operations of the department, including the implementation of the AFV program.

While recognizing that there are many possible ways in which we could examine AFV management strategies at FPDDC Fleet Services, we have chosen to focus on their purchasing decisions. Unlike other managerial decisions, such as hiring, which only occur intermittently, vehicle purchase decisions provide a consistent and observable reference from which to explore emergent managerial behavior. Each year, DuPage County allocates approximately $1 million to the Fleet Services Department for capital replacement, most of which is used for vehicle purchases. Financing for AFV purchases must come from these funds as well as from State and Federal grants and rebates.1

For the purchase of gasoline vehicles, the Fleet Services Department currently employs a, fundamentally, rational and analytic decision-making process for determining the number and type of vehicles to purchase. The process can be characterized by the following steps:
  1. 1.

    An in-house vehicle database is queried for those vehicles that are more than 7 years old or have greater than 50–70,000 miles. Of the 180 in-road vehicles the department maintains, approximately 50 vehicles usually meet these criteria.

  2. 2.

    The vehicles meeting these criteria are examined by the shop foreman, who has intimate knowledge of each vehicle. Based on his personal assessment the list is shortened to approximately 30 or 40 vehicles.

  3. 3.

    Each vehicle is evaluated and scored individually. The score results in the vehicle meeting one of three categories: keep, re-life2, or replace.

  4. 4.

    As a final check, the department head, assistant and shop foreman meet to finalize the decisions on which vehicles to replace. This results in a list of about 10 vehicles that the department will have to replace in that particular year.


This decision process provides reliable information on the replacement of gasoline vehicles with a similar gasoline vehicle. However, it breaks down when decisions must be made about replacing a gasoline vehicle with an AFV.

In 2001, the FPDDC voluntarily adopted a policy to convert its entire fleet, of over one-hundred vehicles, to AFVs within 10 years. The initial conversion plan called for equal distribution of ten AFV purchases per year throughout the ten-year period. It also expected that each vehicle would use 100-percent alternative fuel upon conversion. Originally, Fleet Services expected that the resulting fleet would comprise approximately 60-percent Liquid Petroleum Gas (LPGAS) vehicles, 30-percent Compressed Natural Gas (CNGAS) vehicles and 10-percent 85-percent Ethanol (E-85) vehicles.3 Given the clear objectives laid out in the initial plan, the traditional four step rational plan would only need to consider the type of AFV that should be purchased. However, the decision to replace gasoline vehicles with AFVs was complicated by the unpredictability of both the AFV private market and the related public funding. The final purchase decision often had to be ‘improvised’ based upon unforeseen constraints existing in the external market.

By embracing a new technology, the FPDDC could no longer rely on a rational analytic process for the ‘routine decision’ to replace its vehicles. Rather, it had to make ad hoc decisions to cope with external market uncertainties, while still meeting programmatic objectives. In doing so, it placed additional burdens on its internal technical capacities. Different types of AFVs employ different technologies, many of which were not fully understood. The organization needed to learn how to make adaptations to all types of new AFV technologies. This scenario suggests that the FPDDC is an early adopter organization faced with substantial uncertainties. The next two sections describe the risks of operating as a lead-user public organization and develops a model of organizational coping that relies on trial-and-error learning.

Dimensions of risk in a lead user environment

The Fleet Services Department can be aptly characterized as a lead user organization. Lead users are “…ahead of the majority of users in their populations with respect to an important market trend, and they expect to gain relatively high benefits from a solution to the needs they have encountered” (Von Hipple 2005, pg. 4). Typically, such needs cannot be filled by existing products and lead users are forced to develop solutions for themselves. Pursuing program objectives as a lead user in an uncertain environment creates a context replete with risks that must be understood and managed for the program to be successful (March 1988). As they relate to new technologies in a public organization—such as AFVs—we identify three types of risk—external, internal and political.

In the case of AFVs, the private market is characterized by unsettled technology trajectories and market preferences. These factors make predicting availability, cost and quality of AFV technology a difficult task. Compounding these market uncertainties, AFV programs are often supported by various state and federal government grants and rebate programs (e.g. Federal Congestion Mitigation and Air Quality, State of Illinois Biodiesel rebate, and Illinois Alternate Fuels Rebate).4 Political shifts and budget constraints often make these supports unreliable. Taken together, uncertainties in the private AFV market and public funding shape the external dimensions of risk faced by agency managers.

As with most new technologies, explicit and reliable knowledge about how to maintain, repair or adapt the new technology is often not available. The primary goals of the FPDDC conversion program were to reduce emissions and operational costs of the fleet. To accomplish the goals, the department understood that it would need to replace gasoline fuel vehicles with AFVs that utilized LPGAS, CNGAS or E-85. LPGAS and CNGAS would require technical modifications unfamiliar to the Fleet Services Department. The internal uncertainty that resulted from insufficient ‘know-how’ about technical diagnostics, repairs and AFV conversions of LPGAS and CNGAS vehicles represents a second dimension of risk.

These two dimensions of risk—internal uncertainty derived from limited technical expertise and external uncertainty resulting from changes in the private market for AFV technology and public funding sources—imply that the answers to several key managerial questions are complex and not readily discernable. For example, the development of local fueling and repair infrastructure must be considered at the same time that the agency is faced with changing fuel prices and technology preferences of the market. Similarly, understanding whether it is better to purchase pre-converted vehicles from the manufacturer or undertake post-purchase conversion is linked to the level and development of ‘in house’ technical capacity of fleet personnel.

Finally, purchase decisions must also be made accountable to political bodies and ultimately to citizens. Accountability demands often require development of rules and procedures that constrain decisions and increase political oversight (Deyle 1994). Such oversight represents a third important dimension of risk faced by AFV managers, as bureaucratic constraints imposed from outside the organization can limit the innovativeness of public organizations (Moon and Bretschneider 2002). Hence, fleet conversion presents a context in which a lead user organization must sustain a high degree of openness to the environment and a tolerance for technical uncertainty while demonstrating traditional, stability-oriented values typically demanded of public organizations.

For lead user public organizations, risk manifests itself as: 1) internal technical uncertainty 2) external market & support uncertainty, and 3) the uncertain political context. The literature on policy innovations identifies some of these uncertainties (see for example, Deyle 1994; Roberts 1992); however it says little about how they interact. As discussed below (and depicted in Fig. 1), the three uncertainties interact such that the management of any one, in isolation, is unlikely to mitigate the overall risk.
Fig. 1

Model of emergence due to coping through learning

Internal uncertainty—characterized by the presence or absence of technical knowledge—is potentially the most manageable dimension of risk faced by fleet agencies. Internal uncertainty can be reduced by consciously developing in-house expertise, through trial-and-error learning and, to the extent that it is available, formal technical training. By reducing internal uncertainty, AFV managers are also able to control other types of risk. For example, innovations such as technical modifications to vehicles or conversion technology by the organization may feed back into the market, thereby affecting future development of technology in ways that conform to the needs of the organization (Von Hipple 2005). Further, the more capable the agency is at addressing technical issues, the more likely they will be able to efficiently achieve promised outcomes; thus, limiting the need for political oversight.

Limiting internal uncertainty does not negate all the risks faced by an AFV manager. For example, while some innovations can feed back into the external markets, others are a function of the market moving in response to other users of the technology. Thus, even if fleet service agencies innovate and develop strong in-house expertise, AFV managers are still likely to face high levels of uncertainty in technology markets. Managers also face another external risk over which they have almost no control: the availability of government program support and funding. Much like the private markets, public support for different forms of AFV technology may shift in ways that negate the gains made by in-house expertise. Additionally, while innovation may serve to buffer excessive political oversight, oversight occurs for a wide variety of rationales not associated with service quality. As a result, AFV managers must simultaneously foster innovation as a means of limiting oversight while actively seeking to minimize excessive levels of oversight that inhibit innovation.

In sum, we expect that a public AFV program will face risks similar to other lead user organizations, such as limited internal know-how and shifting technology trajectories in the private market, but with the added constraints of a public accountability. Interactions among the different types of uncertainty add complexity. One likely response is simply to ‘cope’ with problems as they arise while ‘learning’ from failures and successes until the program and related technology matures.

Mitigating risk: a coping framework

The literature on fleet replacement considers purchase decisions to be an outcome of minimizing the costs associated with operation and maintenance. Our interest in studying the management strategies of AFVs was, in part, motivated by mounting evidence that, while cost minimization may be a valid long-term goal, such ‘optimal’ solutions seem unrealistic in a lead user public organization. Instead, the environment of AFV implementation dictates that AFV management strategies will develop over time based upon the cumulative knowledge that an agency gains from its efforts to effectively address internal, external and political uncertainties. Standardized management strategies should emerge as the agency develops ‘in-house’ expertise that decreases the agency’s internal uncertainty while providing a degree of flexibility to cope with external uncertainties (Cyert and March 1963) within a context of political oversight. This section presents a model of the emergence of management strategies for public organizations that are lead users of technology. It considers the case of AFV purchase decisions that are made through a process of trial-and-error learning. This model and discussion motivate four propositions.

Although an AFV program may begin with a clear set of goals, funding sources and operational protocols, unanticipated opportunities and constraints arise, forcing fleet service managers to make decisions ‘on-the-fly.’ They have to take actions to overcome practical difficulties—they will have to cope (Chia and Holt 2006; Armistead and Clark 1994).

Coping (i.e. dealing with practical difficulties) takes into account a range of daily decision-making devices, such as strategic thinking that incorporates longer term organizational objectives (Graetz 2002), intuition based on outcomes from past experience (Khatri and Ng 2000), or improvisation (Weick 1998). For example, given the choice to purchase one of two different AFVs, neither of which satisfies previously established standards or program goals, the AFV manager might: 1) select the vehicle with some thought as to how the program’s goals will change in the long run, 2) make the purchase based on the vehicle that most closely reflects past purchase decisions, or 3) select an entirely new course of action. Thus, an essential component of coping is the ability to draw on a range of decision-making resources that may be appropriate for the particular situation.

In coping, organizations learn (Geenhuizen and Nijkamp 2003). For example, although an agency may have the in-house expertise to efficiently service and maintain a particular type of AFV, the private market may have shifted and either the vehicles for which the agency has developed an expertise are no longer available or the funding to support the purchase may be unavailable. In this context, managers can: 1) ignore the uncertainty, stick with the initial plan and hope for the best, 2) do nothing until the uncertainty resolves itself, 3) identify and specify key uncertainties and attempt to work around them, 4) reduce the uncertainty perhaps through research or shifting it onto someone else, and 5) treat uncertainty as an opportunity for crafting creative policies (Geenhuizen and Nijkamp 2003). Each of these approaches represents different coping mechanisms. Excepting the first two ‘do nothing’ alternatives, these mechanisms suggest that AFV managers may be able to mitigate uncertainty through a process of organizational learning. We propose a model (Fig. 1) in which AFV managers cope with uncertainty by relying on a process of organizational learning.

Organizations—both public and private—evolve and adapt. Two key frameworks have been put forward in the literature to explain how organizations learn—complex adaptive systems and organizational evolution. The adaptive systems approach posits that organizations interact directly with the external environment and constantly strive toward some form of stasis. Thus, adaptive behavior requires learning through direct interactions with the environment (Holland 1995). By contrast, the evolutionary approach suggests that organizations encode learned inferences from history into routines that guide behavior, thereby enabling learning to be passed to future generations of employees (March 1991). Encoding inferences into organizational memory (March 1988) can take many forms including formal rulebooks, tacit knowledge, and culture maintained through systems of socialization and control. These frameworks allow us to move beyond the stereotypical characterization of public agencies as inert, unimaginative and excessively bureaucratic toward a perspective of public agencies as open, flexible and capable of learning.

Lead users, such as the FPDDC, do not have the benefit of established protocols and market proven products. They need to develop new knowledge through an internal process of trial-and-error. Such experiential processes are personal, rooted in action, based in a specific context, and result in concrete know-how (Polyani 1967). Difficulties and costs associated with the communication of tacit knowledge are often referred to as the ‘stickiness’ of information’ (Nonaka and Takeuchi 1995). Thus, a lead user will need to effectively convert tacit (sticky) knowledge to more explicit (less sticky) forms of knowledge (Polyani 1967), through appropriate communication channels. In the case of AFVs, the expectation is that an agency’s ability to effectively address technical issues, such as routine diagnostics and repairs is increased through trial-and-error learning. Tacit knowledge is communicated internally among shop-floor workers and fed back into the process of trial-and-error learning until the innovation or solution to the particular problem becomes explicit. The agency evolves from one in which little was known about AFV technology to one that is able to efficiently address technical issues even as new members join the organization. It also evolves toward a more coherent strategy that includes standardized procedures and decision criteria that emerge from coping. The development of explicit knowledge serves two purposes: it lowers internal uncertainty and allows the agency to effectively interact with its external environment.

Faced with uncertainties arising from 1) obsolete technology, 2) changing fuel and vehicle prices, 3) the entry and exit of new firms, 4) and variability in public funding, the organization can inform private firms and public agencies of important internal innovations and subsequent funding needs. Moreover, while communication from the AFV agency to external entities may help reduce external uncertainty, there is also the opportunity for the private firms and funding agencies to communicate their priorities back to the agency. This external feedback loop helps to focus the direction and priorities of the internal learning process and enables the organization to evolve with and adapt to its environment.

By reducing internal and external uncertainties through the process of innovation, the agency is more likely to effectively meet end-user needs in an efficient and effective manner. For example, by generating high internal technical knowledge through trial-and-error learning, the agency will reduce the likelihood that end-users experience excessive down-time while their cars are being repaired. Also, by shaping the external market, the agency may increase the consistency of the type and quality of the replacement vehicles that they purchase. Satisfying end-users helps the agency reduce the potential for political oversight, a key factor for the innovation process (Moon and Bretschneider 2002).

Opportunism proposition

The model anticipates that public managers will rely on a process of organizational learning to cope with the dynamic and fluid characteristics of public AFV programs. As a lead user, the organization faces risks that are both internally and externally derived. Internally, managers aim to develop in-house expertise on technical issues related to vehicle conversion, repair and maintenance to reduce uncertainty about technologies that are new and not well tested. Externally, managers try to understand and anticipate changes in the overall AFV landscape, including the availability of future funding sources, sources of AFV technology, and related infrastructure. However, due to high levels of technical complexity and slow learning processes, managers are unable to make decisions that strategically address long-term goals of the organization. Instead, managers make decisions about fleet management in the short-term that can be broadly justified in terms of long-term goals, but are more reactive. In lead user organizations where external technology environments are uncertain and internal technical capacity is limited, managers are likely to react to opportunities as they arise. This suggests a positive relationship between the level of risk and opportunistic decision making of managers.
  1. P1.

    The higher the internal and external risks, the more likely vehicle purchase decisions will be opportunistic as opposed to strategic.


Oversight-performance propositions

The FPDDC Fleet Services agency is directly accountable to the users and indirectly accountable to the citizens who are served by the various agencies in the District. Given that the Forest Preserve has established an umbrella policy that commits the District to a ten-year conversion program, Fleet Services must ensure that the conversion process stays on schedule and that the fleet operates at a normal service level. Failure to maintain the conversion schedule or to maintain normal operation levels may result in greater political oversight, external interference in the operations of the agency, and ultimately in the delay or collapse of the replacement program.

Prior research on accountability finds that the mechanism of accountability is related to (1) performance expectations and (2) the nature of the core task. Performance expectations are divided into input, process, output and outcome, while the nature of core work ranges from routine to non-routine (Romzek and Johnston 2005; Romzek and Dubnick 1994). When performance expectations are primarily input oriented and when tasks are routine, hierarchical accountability mechanisms are generally preferred: rules and standard operating procedures ensure accountable actions. When tasks are less routine and performance expectations related to near-term outputs that can be clearly recognized and assessed, political oversight is often the preferred accountability mechanism. However, in public organizations when tasks are non-routine and performance expectations are longer-term outcome oriented, professional qualifications serve to ensure accountability.

Fleet operations usually entail moderately routine activities and outputs that are more clearly assessed: reliable vehicles. However, fleet agencies are also professional organizations; a substantial level of training is required before an individual can be hired as a mechanic. Also, the conversion process to untested AFV technology necessitates trial-and-error activities and decision processes that are best defined as non-routine. Dual expectations in which clients demand reliable vehicles and the program demands a continuous conversion process means that the agency must manage both day-to-day service outputs and longer-term program outcome expectations. It is well recognized that a high level of external political oversight constrains the innovativeness of public organizations (Moon and Bretschneider 2002; Welch et al. 2004). Therefore, to reduce the potential of political oversight, the Fleet Services agency must integrate new AFV technology sufficiently quickly and seamlessly such that it satisfies the service quality expectations of its end users. If reliability and performance of the alternative fuel vehicles falls, end users could call for greater political oversight of the agency. Likewise, a public lead user that successfully resists political oversight would also be more likely to successfully fulfill the output expectations of its end users and the outcome demands of the program.
  1. P2.

    Political oversight will increase as the performance of the organization decreases (on schedule conversion of reliably performing AFVs).

  2. P3.

    The performance of the organization (on schedule conversion of reliably performing AFVs) will decrease as political oversight increases.


Knowledge proposition

Finally, over time, the repetition of procedures, internal communication about problems and solutions, and similar communications with private firms, can transform tacit knowledge into explicit knowledge making information more reliable and accessible. At least three benefits result. First, the agency will have developed the ‘know how’ to address technical issues as they arise, which in turn minimizes internal uncertainty. Second, by transferring this knowledge to the market, the agency directs and focuses the efforts of private firms, in essence shaping and stabilizing the market. Third, the agency can communicate the value and support of existing external funding sources and any need for different funding opportunities. This suggests that as knowledge becomes more explicit, a more coherent managerial strategy (in this case regarding the purchase of AFVs) should emerge. Underlying this statement is the realization that if technical knowledge fails to become explicit, the organization runs the risk of either making uninformed decisions or becoming overly dependent on the knowledge of a few key employees and the agency’s. The final proposition then is based on the development of explicit knowledge.
  1. P4.

    As knowledge becomes more explicit with respect to internal and external environments, AFV purchase decisions will become more standardized.


Methodology and data

As noted by Whittington (1996), “[t]o understand strategizing better, we will need close observation of strategists as they work their ways through their strategy-making routines” (p. 734). Thus, given the objective of this study—to better understand the dimensions of emergent strategies—we have employed a case-study methodology. We have relied primarily on qualitative data collected through an open-interview process thereby, trading off concerns about external validity for the opportunity to explore richer detail of a qualitative study. Although we rely heavily on findings from interviews, we also draw on two other sister studies: a cost effectiveness analysis (Haller et al. 2007) and an end user adoption study (Johns et al. 2009) of AFV conversions at the FPDDC. These additional studies allow us to ‘triangulate’ our findings using different data sources and methodologies (Yin 1984). That is, to the degree that these studies—based on different data sources and analyses—support our findings suggest we can be more confident in generalizing beyond our particular case.5

Seven interviews in total were administered to the department manager, assistant manager, shop foreman, and department mechanics.6 The objectives of these interviews were to: 1) ‘drill down’ to the fundamental criteria in the day-to-day decisions surrounding the management of the AFV program, particularly relating to vehicle purchase decisions, 2) explore the process of learning within the organization and 3) understand the internal and external communication channels with respect to the creation and institutionalization of new knowledge.

The interviews took the form of guided conversations (McNeil 1990), where the interviewer could deviate from the instrument to follow particularly interesting and valuable points. It is also worth noting that while we were primarily interested in understanding the emergence of a single dimension of AFV management, namely, purchase decisions, we developed the instrument around four management issues: 1) technical knowledge, 2) training, 3) hiring and 4) purchasing. This allowed us to consider purchasing decisions within the broader managerial context.

The propositions to be examined in the study are developed around three constructs—purchase decisions, internal innovation, and political oversight. First, purchase decisions are expected to vary (between opportunism and standardization) in relation to the stage of the AFV program and the dimensions of risk. In the early stages of an AFV program where uncertainty in each of the three dimensions is relatively high, we expect purchase decisions to be opportunistic. As the AFV program matures, knowledge of the external and internal environment will increase reducing uncertainty and making purchase decisions more standardized.

To understand how purchase decisions varied over time, respondents were asked to identify the last AFV purchased. This was followed by a series of probes such as: “why did you purchase this vehicle?” and “how involved were you in the decision to purchase this particular vehicle?” Responses to these questions provided a detailed picture of the existing AFV purchase criteria and process. With this picture in hand, interviewees were asked to explain how the current process had changed over the past 5 years (i.e. since the beginning of the AFV program). Responses that suggested the criteria for purchase was either ‘availability’ or, was based on some ‘pre-existing goals and objectives’ helped to distinguish between ‘opportunistic’ and ‘standardized’ purchases.

The second construct, internal innovation is represented as the degree to which shop-floor workers were able to diagnose and address new problems for which there was no standard operating protocol. To understand how internal innovation varied over the last 5 years of the FPDDC’s AFV program, interviewees were asked to identify the last mechanical or technical AFV problem they faced and explain how that problem was addressed. This was followed with an inquiry into how the problem was diagnosed and how the diagnosis and response had changed over the past 5 years. To capture changes in levels of internal knowledge, we asked if the problems had increased or decreased and whether they were finding more or fewer new problems that needed to be addressed. However, the key to our model is that ‘performance improving knowledge’ is developed through a process of trial-and-error learning. Thus, if new knowledge is derived from some other process, such as formal training or, performance is improved by some exogenous factor such as increased funding or improvements in AFV technology, then our model would be an incorrect depiction of how public organizations manage the risks associated with new technologies. Accordingly, because we wanted to attribute improvements in diagnostic abilities innovations occurring through trial-and-error learning we asked interviewees if the changes they identified were a result of factors such as: 1) internal training, 2) improved AFV technology or 3) more funding to buy better vehicles. To the degree that responses to these factors were negative, would suggest evidence in favor or our model.

The third construct we considered was political oversight. That is, we sought to understand how intrusive the accountability process was with respect to the AFV program. Accordingly, we considered oversight in terms of the number of required reporting devices by asking interviewees to describe the last communication they had with the County Board. Responses to this question were probed to understand if the communication was formal and scheduled, and how this relationship had changed over the past 5 years. This line of inquiry allowed us to identify if the accountability process had become more or less stringent since the inception of the program.

Underlying each of our propositions was the idea that AFV managers must find ways to limit the three dimensions of risk. A key challenge we faced in developing the interview instrument was the operationalization of these risk dimensions (internal, external and oversight). First, having characterized internal uncertainty as limitations in technical knowledge we drew on responses to questions regarding the shop-floor workers ability to diagnose and address new problem. Thus, if respondents indicated that they were unable to effectively diagnose any AFV related problems or that they were facing new problems that they had yet to encounter, internal uncertainty could be characterized as relatively high.

Second, to understand the two factors that characterize the external uncertainty faced by the FPDDC a series of probes asked respondents about changes in the private AFV market and changes in available funding sources. For example, to understand the perceived dynamics of private market component of external uncertainty we asked “Are there more or fewer AFVs available for purchase?” and “Why did you think these changes occurred?” Similarly, to shed light on the factors that influence external uncertainty we asked probative questions such as: “Have funding sources changed?’ and “What if you don’t have enough funding to meet your needs?” Responses to these and other questions allowed us to consider the degree to which the external environment had shifted and how these factors have affected AFV purchase decisions.


The findings of our study provided mixed support for our particular propositions. However, we found evidence supporting our more general claim for the emergence of an AFV management strategy due to a coping/learning mechanism. More specifically, we find evidence of 1) opportunistic purchase decisions at the early stages of the AFV program, 2) the need and ability of the organization to engage in trial-and-error forms of learning, and 3) the emergence of standardized forms of knowledge as the AFV program matures. This section outlines the findings for each of the propositions.

Opportunism proposition

This proposition anticipates a positive relationship between ‘risks’ and opportunistic AFV purchases: the higher the risks the more likely purchase decisions will be opportunistic. Positive support for this proposition requires evidence of the presence of high levels of uncertainty, purchases that could be characterized as opportunistic, and linkage between the two. Support for these was found not only in our interview responses but also in the cost effectiveness analysis published by Haller, et al. (2007).

Interviews provided clear evidence that during the early stages of the AFV program, AFV technology was not well understood and protocols for diagnosis and repairs were non-existent. For example, the early propane vehicles purchased would often “run lean” causing the engine to back fire and “blow up” its air-box.7 The manual provided with the original conversion kit failed to address any troubleshooting measures for this problem. Instead, it was up to the maintenance team to diagnose the problem, try different solutions, and enter into lengthy communications with the supplier, who often did not understand the origins of the problem. Beyond this particular issue, respondents generally characterized the early stages of the AFV program as a time in which internal technical knowledge was quite low.

“The original system was almost a guessing game…”

“I basically had to learn [about AFV fuel systems] on my own….a lot of those were trial-and-error.”

“There is nothing else [training] offered around…you have to learn on your own…”

“[Person X] has spent a lot of time figuring out how to deal with these small problems…”

Moreover, because the organization is relatively small, communication is typically informal, particularly with respect to AFVs. Solutions to new problems are often communicated by word of mouth and developed by one or two key mechanics. Thus, as evidenced in the following quotes, while the mechanics indicated that they “have got all the bugs [worked out],” the knowledge is still tacit in nature because: 1) no AFV repair protocol has been written and, 2) Fleet Services relies extensively on the two AFV ‘experts.’

“there is no program in place yet for training anyone else…”

“That is why we are inefficient, because we got a lot of propane vehicles over there that need to come over here [for repair] and they sit here for a while because…[Person X] is busy….and will leave something sitting out there…”

These findings support the fundamental underpinning of our model: that lead user organizations must not only develop knowledge but they must also make the knowledge explicit before a mature strategy can emerge. Even though AFV technology is becoming increasingly well understood by certain employees in the organization, it remains tacit with respect to the organization as a whole. Further, these findings suggest that, even at the mid-point of a ten-year program, the FPDDC’s AFV program, with respect to ‘learning,’ is still in its infancy.

The external dimension of risk was also evident in the early stages of the AFV program. Respondents identified significant and rapid changes in the private market, which altered the original program goals.

“…it [AFV technology] changes year to year by the time you get into something there is something new.”

“When we got into this, Ford, GM and Dodge were making vehicles, so you had a lot of choices…now you got a pick-up truck and that is it [so our purchase decisions are constrained by the market].”

The cost effectiveness analysis (Haller et al. 2007) also showed that many of the external funding sources, such as the Federal Congestion Mitigation and Air Quality (CMAQ) grant and the Illinois EPA Biodiesel rebate, did not come to fruition “due to factors that lie outside of the Forest Preserve’s control” (Haller et al. 2007, p. 228). The FPDDC suffered from the unexpected failure of the AFV emission reduction credits market. This market was designed to allow non-mandated fleets that implement AFV replacement plans to receive emission reduction credits for each AFV that they purchased. “Emission credits were estimated to hold a certain market value and, as an early adopter, the Forest Preserve would have been able to sell their credits to other fleets that found it more cost effective to buy credits rather than implement an AFV conversion program” (Haller et al. 2007, p. 228). However, this market has not developed (Winebrake and Farrell 1997). To date, the Forest Preserve has only been able to take advantage of the Alternative Fuel Vehicle Rebate Program administered through the Illinois EPA.8 The total amount rebated to the Forest Preserve, as of 2005, was $169,842, approximately one-fourth of the anticipated amount (Haller et al. 2007). Thus, the external environment clearly remains in flux, providing further evidence that the AFV purchasing strategy is likely to remain characterized by coping.

In sum, the interviews and cost-effectiveness analysis provided evidence of high internal and external dimensions of uncertainty. However, the inability of the FPDDC to cope with internal and external uncertainties should not be overstated. To some degree, the FPDDC was able to limit some of these through its ‘local’ environment. That is, it was able to use its relationships with local companies to cope with technical and logistical issues. As the quote below indicates the FPDDC has been able to customize its own market with respect to its conversions.

We were buying the vehicles ready to convert but, now we [were] having problems finding someone one who could convert them…our local distributor, made arrangements… that they would drive one of my trucks to Arizona. They would do the conversion at no cost to us. Have it run through the actual EPA test then they would transfer the truck back to me once it was certified. Then I would run the truck and anytime I decided I didn’t like the system they would remove it at no charge to me.

As will be shown in the next section, the FPDDC has also mitigated much of its risk by adequately supporting end-users. However, despite their influence over local or regional suppliers of AFVs, larger automobile companies (e.g. Ford and GM) are still the primary drivers of the AFV market. Thus, while individual public AFV organizations are unlikely to direct the AFV market as a whole, their ability to communicate to agencies within their sub-markets (i.e. local or regional) may lead to important ‘custom-made’ solutions. Although, the FPDDC has been able to mitigate some of the risks it faced the organization is still in the early stages of learning and coping, as internal knowledge has yet to become explicit and the external environment has yet to stabilize.

As we proposed, purchases at the early stages of the FPDDC’s AFV program could be aptly characterized as opportunistic. This opportunism is well reflected in the following interview responses.

“The decision is often based on what we can get….”

“Why did we get them? It was a deal that we couldn’t pass up for the technology…we got some of our natural gas vehicles because they were sitting someplace …. And we were made a deal that we couldn’t pass up.”

“The new vehicles we buy are decisions that are based on what we need and what we can get.”

Two points are worth noting. First, while purchases in the early stages of the AFV program can be characterized as opportunistic, they seemed primarily to be a result of external uncertainties as opposed to internal uncertainties. Thus, while we proposed that greater overall risk should lead to opportunistic behavior, it seems that the presence of a particular dimension of risk—external risks—are more likely to lead to opportunistic behavior.9 Second, it seems that even at the mid-point of the program at least some of the vehicles purchased remain a result of opportunistic behavior. For example, recent purchases of two hybrids AFVs were clearly a function of opportunism. One interviewee stated:

“…we just bought two hybrids because they were available and a hell of a price.”

Opportunistic behavior is not particularly surprising as FPDDC Fleet Services was only in its fifth year at the time of the interviews. Additionally, as one of only a few local agencies with an aggressive and voluntary replacement program, its lead user position may require opportunistic purchasing even after internal knowledge is more explicit. In sum, the interviews (and cost-effectiveness analysis of Haller et al. 2007) suggest that the AFV program can be characterized by high levels of uncertainty and that this uncertainty has lead to greater levels of opportunistic behavior.

Oversight-performance proposition

The oversight-performance proposition suggests a recursive relationship between oversight and innovation. On the one hand, less political oversight will help foster the innovative behavior required at the early stages of an AFV program, while, on the other hand, trial-and-error learning that results in program delay or in unreliable vehicles, typical of innovative processes, will increase oversight. This proposition requires an understanding of the degree to which political bodies observed the AFV program and the degree to which their (in)actions affected the ability of the FPDDC to effectively replace traditional gasoline vehicles with AFVs.

Overall, our findings show that there has been almost no variation in the level of political oversight over the first 5 years of the program. We believe that the evidence indicates that the agency has been able to successfully adapt and maintain new technology such that service levels have not been adversely affected. For example, shop floor mechanics seem confident that technical knowledge has improved to the point where they have “got all the bugs” and now “know where to look right away” when diagnosing and repairing AFVs.

The interviews also show that continued trust between the political powers and the AFV manager, established prior to the adoption of the AFV program, has been a key component in its current ability to maintain autonomy. The following quotes indicate the level of decision-making autonomy that they enjoy.

“…as far as the scrutiny…[X] used to have walk in with a stacks of reports on the thing that was being replaced….that stopped”

The extent, to which trial-and-error learning has gone on with few repercussions at the FPDDC, such as described in the example on blown up air boxes, validates both the ability of the agency to swiftly convert trial-and-error learning to routine outputs as well as the maintenance of trust in the professionalism of the organization. Respondents described initial efforts to adopt new AFV technology as “a guessing game.” However, there are no indications of increased political oversight. Simultaneously, the autonomy of the agency to make its own purchasing decisions and to experiment with technology adaptation appears to be strong and unchallenged. Had external political oversight increased, it is possible that the department’s ability and willingness to engage in the ‘guessing game’ would have decreased.

The second part of this proposition relates to performance. That is, how well has the agency been able to provide reliable vehicle replacements to the various County departments? In the face of limited oversight, as described above, our expectation is that the agency’s performance, regardless of technical limitations should be relatively high. The interview evidence suggests this is the case. That is, trial-and-error learning resulted in viable solutions for end users. When discussing propane vehicle conversions, one interviewee offered the following statements:

“ has [become] better. In the beginning almost all of [the end users] were against it, ‘cause it just wasn’t their common get in start it up and go.... [But] for the most part they are all for it now, so long as the vehicle is reliable and does what it needs to do, which at the beginning was a bit of a problem.”

“[The propane conversions] have gotten better because [our mechanics] have spent a lot of time working on these problems.”

Other findings from survey research conducted on the FPDDC end-users also found that end-users of AFVs in the FPDDC “rated the maintenance of the vehicles highly” (Johns et al. 2006). In sum, although trust with political oversight agencies was developed prior to the inception of the AFV program, the agency has, at the very least, been able to maintain this trust in the face of internal uncertainties. In such an environment, our model suggests that the agency should be able to perform relatively well. Our findings provide support for this proposition.

Knowledge proposition

The final proposition we explored considers the relationship between internal and external forms of knowledge and the standardization of purchase decisions. This proposition suggests that in the latter stages of an AFV program, when technical issues and the nature of the external environment are well understood, purchase decisions are more likely to be made on some established and standardized set of criteria. That is, we would expect the emergence of a more routinized purchasing strategy. This proposition requires evidence of explicit, as opposed to tacit, knowledge of the internal and external environments, which can be related to a standardized set of criteria for AFV purchase decisions.

First, as has been shown in the previous two sets of propositions, internal knowledge has gone from a “guessing game” to one in which technical problems are well understood. Responses in our interviews provided further evidence of this new knowledge becoming, at least to some degree, explicit in the operation of the department.10

“One of my jobs was to teach those guys how to look at different stuff on the propane and [other AFV] fuel systems…”

“If it is a problem that I found, that would take someone a long time to find I write it down on the work ticket. I am also pretty good at communicating [these problems] with [Person X].”

The department has also developed and maintains an extensive vehicle database that the each vehicle was in the shop for repair, the nature of the repair, costs of the repair and fuel usage. While this database is used in the part of the purchase process, as described in the case description, it is still not actively employed as a device for documenting new types of problems and solutions. Rather informal communication and notes on the work tickets is the preferred approach.

“If I can’t figure something out, I’ll just get on the phone to [X]…it is easier than tracking down the notes on the work ticket...”

However, as evidenced by the previous two propositions, purchase decisions are still very much a function of what is made available in the market (e.g. hybrid vehicles) and knowledge remains primarily tacit. This has several implications. First, knowledge of the external market is not well understood. The department is still not able to anticipate the availability of particular types of technologies or vehicle makes. Moreover, as evidenced by the quote below, the agency is still subject to shifts in external funding.

“There is currently no sunset date [on State Funding that supported the program], but if they do decide to establish one we might find ourselves in trouble.”

Thus, while it seems that the internal dimension of risk is starting to become more explicit external dimensions of risk remains high enough to prevent the emergence of a purchasing strategy.

Summary of findings

The findings from our study provide evidence of continued opportunism despite relative success at limiting the various dimensions of risk. First, with respect to limiting internal uncertainty, the FPDDC has clearly developed sufficient internal expertise for addressing technical AFV issues. However, while well developed in some respects, this knowledge remains tacit as the diagnostics and protocols are exclusively within the purview of two key employees, which has been shown to lead to internal inefficiency when addressing several AFV issues at once. Second, external uncertainty remains relatively high as the private market is extremely fluid and the availability of AFV technology from one time period to the next is still very unpredictable. However, the agency has been able to mitigate these uncertainties, to a limited extent, by customizing their relationship with local AFV suppliers with respect to providing conversion kits that meet EPA standards, and by sharing the lessons they learned from internal process of trial-and-error with private sector technology providers. Finally, the FPDDC has very successfully managed the level of political oversight it faces. Although the high levels of trust exhibited by this agency seem to have been developed prior to the AFV program, our findings suggest that, at the very least, this trust has been maintained by the agency’s ability to successfully meet end-user needs.


The Forest Preserve District of DuPage County, Illinois, has undertaken an innovative and aggressive program of AFV fleet replacement. As a lead user of technology in the public sector, they faced risks from internal technical uncertainties, external market and funding uncertainties and uncertainties regarding political oversight. We have shown that the county used emergent strategies based upon trial-and-error learning to cope with purchasing, maintenance and conversion problems as they arose. The interview evidence also suggest that external risks, in particular instability in funding sources and technology availability, place extremely high constraints on a small public organization’s ability to develop and maintain a coherent management strategy. Beyond these findings, this paper points to some broader conclusions that are relevant for other fleet services agencies as well as other public lead user organizations.

First, despite the fact that the FPDDC Fleet Services is in its sixth year of the conversion program, the process of converting of tacit knowledge to explicit knowledge is slow. Only two of the eight full time mechanics are fully trained in the range of conversion technologies even though over fifty percent of the vehicles are AFVs. While some internal efforts to train more mechanics have begun, few if any written materials have been developed. Tacit knowledge on AFVs is still informally rather than formally communicated. These delays are probably due to both the ongoing level of technical uncertainty and the limited resources and time of the organization to develop explicit materials. For small agencies, allocation of resources toward formal training and development of materials may result in the reduction of resources dedicated to maintaining service quality, a key component in the reduction of political oversight risk. Additionally, the inherent stickiness of information combined with resource limitations of small public agencies means that tacit knowledge will only be slowly converted to explicit knowledge. Policy makers who set up programs and managers who administer them must make available and apply resources to ensure that the conversion process occurs. Attention to this process should help encourage the lead user agency to move from emergent to more planned strategies of implementation.

Second, despite the fact that FPDDC has purchased over 50 vehicles and is in its sixth year, the purchasing process remains opportunistic. The addition of one step in the four-step rational analytic method for determining vehicle turnover has bumped a routine vehicle purchasing decision process into a consistently ad hoc one. External uncertainties will probably continue to force the agency to purchase vehicles in a relatively opportunistic way. How soon the agency can return to a more routine analytic method is unclear. However, the lessons for public managers and policy makers are clear: conversion of a fleet of vehicles to AFV is complex. Organizational change that faces substantial internal and external uncertainty will require the organization to dramatically change the framework for its decision making and operations for a long period of time.

To succeed, the political environment must remain stable. Policy makers need to recognize that trial-and-error patterns of learning will include setbacks and delays. This means that oversight and evaluation activities must recognize and account for these issues. Even as the public continues to turn to performance-based approaches to management, technology intensive programs that are subject to highly fluid external environments may require alternative evaluation approaches. If strategies are in fact emergent, as our study suggests, then evaluative measures will need to be more flexible than typical performance based evaluation methods allow. Additionally, fleet managers must recognize that the task of conversion, for example, is both technical and political. Continuous efforts to both maintain service outputs and attain program goals are critical.


Anticipated federal and state level grants, rebates, emission reduction credits programs and, reductions in fuel use and maintenance, the initial strategy developed by the FPDDC expected the AFV conversion effort to actually net a savings of approximately $150,000 over ten years.


This describes a vehicle that needs significant repairs but, the cost of these is less than the purchase of a new vehicle.


Observations of the FPDDC AFV program at the half-way point of the program paints a picture at odds with this initial plan. For example, 1) to date the department as replaced 102 vehicles, which is more than 50 percent of its 180 on-road vehicles, however, the department anticipates “falling behind” this aggressive replacement schedule, 2) alternative fuel has lagged behind expected use by approximately 10 percent, 3) the net outlay per AFV is 15 percent higher than expected, and 4) the amount of grants and rebates has fallen short of expectations.


In the case of the FPDDC, it was anticipated that these grants would be used to offset the costs of infrastructure development, vehicle purchases and fuel purchases.


Haller et al.’s cost effectiveness analysis explicitly quantifies the difference between the initial strategic plan and outcomes at the mid-point of the AFV program, which is critical for considering the opportunism proposition (P1). Johns et al.’s end user adoption study helped us understand the relationship between internal innovations and the implementation of the AFV program.


Important baseline information with respect to key actors, AFV technology, and program goals and objectives was collected through a series of baseline interviews in 2005. However, the main empirical study was conducted over a three-month period in 2006.


Which costs approximately $800 per air box.


This program provides a cash rebate for 80-percent of the incremental extra cost of an alternative fuel vehicle with a maximum rebate of $4,000 per vehicle. Given that the Forest Preserve is a public agency exempt from property tax, the credit is actually taken by the auto dealer and passed through as a discount on vehicle price to the purchaser.


Which, however, is not to say that technical knowledge is not an important factor in the purchasing decisions. AFV managers made clear that they are less inclined to purchase AFV’s in which shop floor mechanics have found recurring problems or are more difficult to work with.


Although, as will be described below, technical knowledge in the FPDDC is still not truly explicit, as most of the technical know how is shared between only two key mechanics.


Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.School of Politics and EconomicsClaremont Graduate UniversityClaremontUSA
  2. 2.University of IllinoisChicagoUSA
  3. 3.University of ChicagoChicagoUSA

Personalised recommendations