ON DEFINING FRONTLINE WORK AND COMMAND

Considerable attention has been devoted in organisation sciences and public administration literature to the functioning of frontline organisations such as social welfare organisations, police and youth care (e.g., Lipsky, Reference Lipsky1980). Frontline organisations share several defining characteristics (e.g., Maynard-Moody & Musheno, Reference Maynard-Moody and Musheno2003; Henderson & Pandey, Reference Henderson and Pandey2013). First, in frontline organisations there are frequent face-to-face interactions between frontline workers and clients. Second, primary tasks are often carried out outside the organisation, that is, distant from the head office. Third, primary tasks are carried out in an ever-changing environment and are difficult to plan in advance. The organisation sciences and public administration literature has been primarily concerned with the degree to which frontline workers implement policy in practice and the way frontline organisations attempt to influence policy implementation by using ‘distant’ management tools, such as performance indicators and training.

This paper is about a barely explored topic in public administration research, that is frontline command. In most frontline organisations, frontline workers solve problems in an environment relatively free of managerial supervision (Henderson & Pandey, Reference Henderson and Pandey2013). However, some frontline organisations assume that direct command of frontline workers is necessary for the effectiveness of frontline work and have appointed frontline commanders to meet this presumed need. The direct supervision of frontline workers by frontline commanders is referred to in this paper as frontline command. Frontline command can be defined as making decisions in the frontline and ensuring that frontline workers carry out these decisions accordingly. Frontline command under high levels of time pressure, such as in emergency situations, is also known as (incident) command and control. Fire services but also some departments of the police (e.g., major criminal investigation) have frontline commanders in place to control frontline work. But also private organisations such as oil drilling companies and petrochemical organisations have frontline commanders at work to ensure proper frontline work.

Frontline commanders typically have to deal with critical, ill-structured and complex problems. They are essentially complex problems for which no straightforward and stable definition can be given. Frontline commanders must therefore actively construct these problems. This is what Karl Weick (Reference Weick1988) meant by enacted sensemaking: the nature and scope of a problem becomes evident only after efforts are underway to tackle it. Table 1 shows the most important characteristics of ill-structured, complex problems.

Table 1 Characteristics of ill-structured, complex problems (compiled from Rittel & Webber, Reference Rittel and Webber1973; Simon, Reference Simon1973; Zsambok & Klein, Reference Zsambok and Klein1997; Funke, Reference Funke2001)

An example of an ill-structured and complex problem in firefighting practices is a rapidly-spreading fire in a warehouse containing hazardous substances. Initially, the precise location and size of the fire is often unclear, also whether there are people still inside and the extent to which the hazardous substances are involved. This information is necessary in order to establish a satisfactory tactic. In addition, it is hard to predict precisely how the fire will develop as well as the consequences for the vicinity. Further, there are conflicting interests: on the one hand the fire causes a column of smoke that will affect the vicinity, on the other hand, large amounts of polluted water used for extinguishing the fire will have a significant impact on the environment.

In addition, frontline commanders often have to solve critical problems in environments rich in sensory stimuli (e.g., Flin, Reference Flin1996). In regard to firefighters, for example, these could be the colour and density of the smoke, the flow, the temperature, the magnitude of the fireline but also noises, which are audible and may indicate the imminent collapse of a structure. In addition to the many sensory perceptions that frontline workers must process in order to gain insight into the problem, there are at least as many additional sensory stimuli with which frontline workers are involuntarily and often unconsciously confronted. For example, the distracting noise produced by the fire engine’s pump, onlookers in the vicinity of a major accident, a large amount of radio communication,and extreme weather conditions.

In the following section we discuss the way in which frontline workers make decisions and their implications for frontline commanders. We then look in greater depth at frontline command and present a general model comprising five steps to illustrate the constructs of frontline command. In each step we present recommendations provided by psychological research and management sciences designed to improve frontline command. Most of this research can be classified as NDM research.

THE NEED FOR FRONTLINE COMMAND

In the 1980s and 1990s a lot of research was carried out into the way in which frontline workers make decisions in their own working environment. This research was known as Naturalistic Decision Making, abbreviated to NDM (Zsambok & Klein, Reference Zsambok and Klein1997; Lipshitz, Klein, Orasanu, & Salas, Reference Lipshitz, Klein, Orasanu and Salas2001; Lipshitz, Klein, & Carroll, Reference Lipshitz, Klein and Carroll2006). The principal finding from NDM studies was that frontline workers in the majority of cases make satisfactory decisions (though not always optimal!) on the basis of their experience. This principal finding was encompassed by Gary Klein and colleagues in a theory named ‘Recognition-Primed Decision Making’ (RPD) (Klein, Calderwood, & Clinton-Cirocco, Reference Klein, Calderwood and Clinton-Cirocco1986; Klein, Reference Klein1989). According to this theory, frontline workers possess the ability, on the basis of a number of indicators, to recognise a new situation and subsequently to choose an approach which in a similar situation in the past worked satisfactorily. This ability is also known as intuition. Nobel prize winner Herbert Simon (Reference Simon1992: 155) described intuition as follows: ‘The situation has provided a cue: this cue has given the expert access to information stored in memory, and the information provides the answer. Intuition is nothing more and nothing less than recognition’. RPD is rooted in empirical research into (American) firefighting operations but was also used successfully to describe decision making among doctors, pilots, chess players and other professionals (Klein, Reference Klein1993, Reference Klein1998, Reference Klein2008).

Although RPD is often an effective decision making strategy, in certain cases it may lead to unsatisfactory decisions. Two specific scenarios can be described:

• ∙ First, recognition may hinder the ability of judgment. A frontline worker can think that he or she is dealing with a prototypical situation but overlook certain (contradictory) indicators (e.g., Orasanu & Martin, Reference Orasanu and Martin1998). The fatal fire at De Punt mentioned earlier, is one such example. The first firefighters to arrive at the scene believed, on the basis of various indicators, that they were dealing with a prototypical fire but were overwhelmed when the fire suddenly spread; in retrospect there had been signs indicating the chance of spread, which had been overlooked. In addition, the fireground can be of such magnitude that crews concentrating on a particular area are unable to oversee the entirety. In such instances, decision making on the basis of the situation in a single area is understandable but hardly sensible given the extent of the fire.

• ∙ Second, there may be a lack of recognition because the decision maker does not have the right experience and/or the (learning) environment does not provide reliable feedback. If the environment does not provide any timely or accurate feedback, it will be impossible for the decision maker to gain reliable insight in the causality between his or her actions and their consequences (Kahneman & Klein, Reference Kahneman and Klein2009). Or as astutely stated by Hogarth (Reference Hogarth2003: 17): ‘You cannot learn from feedback you do not receive and some feedback may simply act to increase confidence in erroneous beliefs’. An illustrative example in the Netherlands is the use of so-called flashover containers. In these containers a flashover (i.e., a deflagration of fuel rich smoke) can be experienced by firefighters in a controlled environment. The training aims at learning the indications that a flashover is about to happen. The result, however, is that firefighters are less alarmed by the flashover because they feel that, as in the container, just getting down on your knees will get you out of trouble. In reality a flashover will usually mean that part of the ceiling collapses trapping firefighters or worse. In addition, the frontline worker may be confronted with a situation of which he or she has no previous experience. In such cases most frontline workers tend to make a decision which has worked well in the past and which they can adapt according to the result. And thus the frontline worker learns as he goes, though sometimes with less than desirable results.

In such cases, frontline command is desired as frontline workers should in no way be expected to be able to protect themselves from making mistakes. A further core finding from NDM studies is that frontline workers are susceptible to biases which affect their ability to assess the effectiveness of their decision making, such as the confirmation bias or the tendency to confirm a previously formed hypothesis by (a) interpreting all gathered information as such, (b) subsequently search only for signs that validate that hypothesis and (c) discount negative evidence (Ask & Granhag, Reference Ask and Granhag2005). A further known bias is that of the sunk cost bias, or the tendency to want to complete a task that is nearing the final phase regardless of circumstances (Meij, Reference Meij2004). This bias can have serious consequences in firefighting operations. A known phenomenon in interior firefighting, for example, is that crews attacking a fire source are reluctant to withdraw because they have made a great effort to reach the fire source, even when a fire chief no longer deems it safe to continue an interior attack.

FADCM: THE FIVE STEPS OF FRONTLINE COMMAND

Various models have been developed within military and aviation sciences that describe steps for successful decision making. Prototypical examples are OODA (Observe, Orient, Decide and Act) (Brehmer, Reference Brehmer2000, Reference Brehmer2005), DOODA (Dynamic OODA) (Brehmer Reference Brehmer2005, Reference Brehmer2007; Jensen, Reference Jensen2009), FORDEC (Facts, Options, Risks/Benefits, Decide, Execute and Check) (e.g., Li, Li, Harris, & Hsu, Reference Li, Li, Harris and Hsu2014). These models are quite similar to one another. First, the models are rooted in cybernetics (Ashby, Reference Ashby1956; Morgan, Reference Morgan1982). Cybernetic models are always goal-directed and cyclical. This means that on the basis of acquired information, goals are established and the model is continually applied until the goal, with or without interim adjustments, is achieved. Second, these models have in common that they both distinguish roughly the same four steps: gather more information, analyse the problem, make decisions and check if the desired result has been achieved. Third, these models are developed to aid decision making of individual decision makers who often have to execute the decisions by themselves (such as military pilots).

We have combined the steps of these models to develop a model particularly aimed at frontline command and thus includes a step in which the decision is formulated into an order. We used the following acronym, FADCM: Factfinding (gathering of information), Analysis (problem analysis), Decision making (making a decision), Communication (issuing the order), Monitoring (monitoring the units’ performance of the ordered tasks). FADCM is based on assumptions that differ fundamentally from those underlying the problem-solving model (Orientation, Evaluation, Control) developed by Robert Bales and Fred Strodtbeck, which is currently often used in the training of frontline commanders in the Netherlands. That model assumes that decision makers adhere to a rational process of choice making in which, on the basis of all the information they have gathered, they formulate a series of options, which they subsequently consider one by one in order to arrive at the best decision. Nobel Prize winner Herbert A. Simon showed long ago, however, that this is not a ‘natural’ method of decision making for people (Simon, Reference Simon1992). Specifically in regard to frontline work Weick (Reference Weick1995), for example, stated that it is impossible for people to form a complete and accurate impression of reality and to retain it. Moreover, Weick stated (Reference Weick1995) that it is impossible for frontline commanders to focus on all individual aspects of complex problems. In contrast to the model still used frequently in the Netherlands, the FADCM model thus attempts to take account of the way in which people make decisions in their daily working lives. To this end, the FADCM model incorporates insights derived from NDM studies. The model illustrates step-by-step the capabilities and limitations of frontline commanders and how they can apply their knowledge in practice. As far as we know, insights from NDM research have not previously been combined in a single integral modal for frontline command.

We shall now discuss, step-by-step through the FADCM model, core insights from psychological research and the subsequent practical implications for frontline commanders. Practical implications are in fact recommendations that frontline commanders within firefighting services and elsewhere can apply directly to work practices. Each step is illustrated by means of an example from firefighting practices. First, a remark about our method. In regard to the recommendations based on the literature, which we give per FADCM step, we know they work but we do not know to what extent. Further empirical research is needed to qualify the effectiveness of these recommendations for different frontline organisations in varying circumstances.

CONCLUSION

We have presented, for each FADCM step, insights obtained from psychological research and organisation sciences, and described the significance of these insights for frontline commanders leading frontline workers. Overall, the FADCM model illustrates clearly that as a result of human factors and practical limitations, frontline workers can only to a very limited degree be managed and controlled by frontline commanders (cf. McLennan et al., Reference McLennan, Holgate, Omodei and Wearing2006). The implication is that the performance of frontline work is largely determined by the degree to which frontline workers are enabled by their organisation to self-manage their tasks (Groenendaal et al., Reference Groenendaal, Helsloot and Scholtens2013). To enhance self-organisation, frontline organisations should promote education, training and gaining experience in different task environments.

Frontline commanders can still, however, play an essential part by making decisions, which compensate for the shortcomings of frontline workers as mentioned in this paper. Frontline commanders should remember this and at each incident ask themselves where they can or should ‘make a difference’. In summary, we present the following concrete suggestions:

• ∙ maintain, in the first instance, spatial and mental distance;

• ∙ from a distance, look for indications that action is required;

• ∙ determine the broad outline of any indication you find to avoid the pitfall of micromanagement;

• ∙ trust your intuition to determine at which moment a critical decision is required;

• ∙ if you feel the need to act, ask yourself if this is the decision which will make a difference. Realise that you can make only a few decisions per incident because each order you issue requires careful communication and monitoring;

• ∙ before making the decision, take time to think about the consequences of that decision. One consequence could be that for a considerable time you would be prevented from having a clear view on the incident resulting in the necessity for frontline command backup;

• ∙ translate the decision into an order and communicate this to the frontline workers. Explain why you have made the decision and ask an open question in order to check if the decision has been understood correctly;

• ∙ monitor the execution of the issued order and adjust if necessary. In the event you cannot be physically present during the execution of an order, ask an open question to determine precisely what the frontline workers have done. New indications may manifest which require a further decision;

• ∙ monitoring is a specific task that requires concentration so it is unwise, as frontline commander, to take on additional tasks until execution of the critical decision is complete.

These key points have been used for some years in training fire chiefs to improve their frontline command skills. Crisislab is currently running a research programme into the effectiveness of these courses. Thus far, the results are looking positive: indeed, frontline commanders themselves feel they are more able to make ‘a difference’. Moreover, frontline workers are very happy with the FADCM model adopted by their commanders whereby they are more autonomous in decision making. Some frontline workers now finally appreciate that frontline commanders are not merely ‘extras’, but that they provide an important contribution to incident management.

In conclusion, we would like to emphasise that the insights from the studies we have compiled are not relevant only to firefighting services. In more general terms, we believe that every incident manager can benefit from the points of consideration we have listed, such as ‘consider the consequences carefully before making a decision’, ‘limit the work load’ and ‘monitor carefully the execution’. It is important that managers who do not personally operate at the frontline (e.g., tactical or strategic decision makers) understand the limitations inherent to manageability of the frontline and the command capacity of frontline commanders. They should take into consideration the practicality of their decisions and seek advice from experts with frontline experience. Finally, our article provides pointers for those in organisations who are involved in preparation for incidents and crises. First, our article demonstrates clearly that the essence of effective crisis management lies in preparation. During crises, frontline workers, logically, do what they have learned. The necessity for frontline command can be limited to an extent if frontline workers are well-trained. Second, our article provides guidelines for the development and assessment of crisis management exercises in which the key points described earlier can be used as an assessment framework.