The programming of decisions for disaster and emergency response: A Luhmannian approach

Abstract

Spanish

French

Organizations managing disasters face a paradox. They need to build stable, reliable structures that are flexible enough to allow adaptation to such unexpected events. Much planning for concrete disaster response operations involves scenarios. From a Luhmannian perspective, this approach is characteristic of a form of ‘if-then’ conditional programming. Extant research on emergencies and disaster management, however, has remained silent about other than scenario-based planning. This article draws on sociological decision theory to highlight alternative forms of planning for disasters. It presents the possibilities to build stable structures for disaster management by making use of conditional programmes that rely on space instead of scenarios, and by making use of what Luhmann calls ‘programme nesting’. It illustrates this argument with a case study of emergency management in a large German city at the origin of this new planning method.

Keywords

Decision premises decision programmes disaster preparedness disaster studies Niklas Luhmann organizational sociology scenario planning sociological decision theory

Introduction

Emergency and disaster response organizations face the challenge of establishing structures that are both stable and permit unexpected incidents to be dealt with (Apelt, 2014: 74–76; Lindell, 2013: 806; Weick and Sutcliffe, 2007: 23–27). The focus in preparing concrete responses to emergencies and disasters has hitherto been mainly on scenario-based planning (see e.g. Alexander, 2000: 89; 2005: 163–169; Carter, 1992: 162–164, 233ff.; Federal Emergency Management Agency, 2010; Harrald and Mazzuchi, 1993; Mezey, 2011: 123; Walker et al., 2011: 165). Such scenarios seek to anticipate all possible incidents (thus rendering them expectable) so that chains of appropriate measures for managing emergencies can be prepared. However, the scenario approach has faced criticism for quite some time (see Kolen and Helsloot, 2014: 612; Oloruntoba, 2013: 1674; Perry and Lindell, 2003: 342ff.; Wright and Goodwin, 2009: 813–819). Nevertheless, alternatives for planning concrete measures have so far attracted little attention.

We consequently follow Michael Lindell’s (2013: 812ff.) appeal to relate so-called ‘disaster studies’ more strongly to social science theory (similarly Tierney, 2007). Specifically, we draw on Niklas Luhmann’s (2003: 45) sociological decision theory and the concept of decision premises to investigate how to prepare for the unexpected in responding to emergencies with other types of planning. From the viewpoint of Luhmannian decision theory, scenario-based disaster planning is a variety of ‘if-then’ or conditional programming (Luhmann, 2006: 271). On the basis of an empirical study, we correspondingly show that there are other possibilities of conditional programming for emergency and disaster planning than the scenario approach. We take as our starting point the municipal Emergency Response Plan for the Düsseldorf old town, which provides the programmatic basis for joint responses to emergencies by the fire and emergency department, police and other response services. This plan focuses essentially on geographical space as the basis for joint planning to deploy resources appropriately in meeting particular emergencies. This focus permits a specific mode of operationalizing interagency programming over and beyond scenarios. We identify two key forms of programming: concretizing ‘if-then’ as ‘where-then’, and programming specific means to be deployed for indeterminate purposes (dealing with actual incidents rather than anticipating potential incidents). The former is a specific way of dealing with the programme stimulus identification problem (see March and Simon, 1993: 167ff.); the latter can be described as ‘programme nesting’, a theoretical notion that has attracted little attention to date (Luhmann, 1964: 11).

Although this possibility of programming resources for organizational purposes (Luhmann, 2006: 265) has long been known as an adequate means of handling the relationship between stability and flexibility (Luhmann, 1964; 1966: 39), noteworthy empirical applications have been lacking to date. This article offers the first more far-reaching empirical examination of programme nesting, a concept that has hitherto been treated only theoretically. We come to the conclusion that the conditional programming of means for so-called purposive programmes can be possible and useful without the ends themselves having to be specified.

The usual scenario-based planning paradigm and the role of space

The programming of decisions

In addressing the theory of planning for emergencies and disasters, we draw on the concept of decision premises in organizations (Kühl, 2013: 97–110; Luhmann, 2006: 222–329; Simon, 1997: 23ff.). From this perspective, planning means deciding on the premises for future decisions (Luhmann, 2006: 230). Decision premises are indicators of the correctness of prospective decisions (Simon, 1997: 308). Premises that can be explicitly set include communication channels (also hierarchies); personnel recruitment and placement; and programmes (Luhmann, 2003: 45–47). These premises do not determine decisions but they do influence them strongly. Although the concept of decision premises is used primarily to describe organizational phenomena, both hierarchies (see Ahrne and Brunsson, 2011) and programmes (see Luhmann, 1995: 203ff.) are to be found in other contexts. It is thus possible to establish joint (operational) hierarchies and decision programmes at the interorganizational level, too.

If we wish to establish how concrete responses to disasters can be planned, the focus is on programmes. Even though communication channels and personnel can also be planned in advance for disasters (and this is also done), what measures are or can be actually taken can be decided only through programmes. Whereas in the event of emergencies, decisions about personnel, communication channels and hierarchies determine who is responsible and in what positions, who communicates with whom, who is authorized to issue instructions and to whom, programmes can provide for concrete measures and relevant objectives. Two types of programme can be distinguished: conditional programmes and purposive programmes (Luhmann, 2003: 45). Conditional programmes take an ‘if-then’ form (Seidl and Becker, 2006: 28). They determine that if a particular event occurs, certain measures will then be taken. Such programmes are thus input-oriented, since only correct identification of the programme stimulus (input) leads to the planned measures being carried out (output).

Purposive programmes, by contrast, are output-oriented because they define fixed goals or purposes (as outputs) that are supposed to be attained. Although the choice of means to attain the ends can typically be limited, it is basically left to the competent decision maker (Kühl, 2013: 104). By their nature, purposive programmes cannot be used at the concrete level to prepare for emergencies. The specific purpose of intervention (e.g. to extinguish a fire) arises only when an incident occurs.

In the literature, purposive and conditional programmes have largely been addressed separately. Among the few exceptions are a number of works in German (e.g. Kühl, 2001: 215; Winter, 1998: 78), which, however, are concerned only with replacing conditional by purposive programmes or converting the former into the latter, and not with how the two types of programme interact. Nevertheless, empirical evidence shows that mixed purposive and conditional programmes do occur. One variant takes the form of ‘programme nesting’ (Luhmann, 1964: 11). This refers to the use of conditional programmes as a means within the framework of purposive programmes.

Nesting conditional and purposive programmes gives greater elasticity: although a set of conditional programmes initially reduces complexity, their use as a means within purposive programmes allows new complexity in pursuing ends (Luhmann, 1964: 11). Luhmann had very early discussed the possibility of nesting programmes (Luhmann, 1964) and had repeatedly returned to the issue (Luhmann, 1966, 2006). However, it attracted little further attention (a rare exception being Rommelspacher, 2011: 151–154). We shall be coming back to this.

Scenario-based planning as conditional programming and the neglected possibility of nesting programmes

In the existing literature, emergency planning has usually taken the form of conditional programming, often on a large-scale scenario basis (see e.g. Alexander, 2000: 89; 2005: 163–169; Carter, 1992: 162–164, 233ff.; Federal Emergency Management Agency, 2010; Harrald and Mazzuchi, 1993; Luhmann, 2006: 271; Mezey, 2011: 123; Walker et al., 2011: 165). What we call scenario planning is the attempt to anticipate many or all potential emergency scenarios and to plan in detail the measures that would have to be taken in each case.

Scenario planning is used in many fields (e.g. Lambert and Sadra, 2005; Ten Brinke et al., 2010; Wallace and de Balogh, 1985). The US Federal Emergency Management Agency (2010), for instance, makes extensive use of it. We have access to a similar (confidential) planning document for Germany relating to critical infrastructures. Such conditional programme planning relies essentially on all the necessary programme activators (e.g. incidents) being correctly foreseen and appropriate responses being planned, and, in an emergency, on the competent authorities being able to judge the situation correctly to avoid setting off the wrong chains of measures. The hope is that extremely detailed and far-sighted planning will prepare for all, even unforeseeable eventualities. The results are substantial documents, often hundreds of pages long, setting out copious conditional programmes in which anticipated incidents activate a chain of predefined measures for addressing all contingencies that could arise in the course of an incident and measures.

Although there are many useful applications for this sort of ‘deterministic planning’ (Kolen and Helsloot, 2014: 612), for example dealing with unexploded ordnance, this approach has been criticized for a number of reasons (see Kolen and Helsloot, 2014: 612; Lindell and Perry, 1992: 37ff.; Oloruntoba, 2013: 1674; Perry and Lindell, 2003: 342ff.; Wright and Goodwin, 2009: 813–819). Thus Frosdick (1997) points out that it is impossible to anticipate all events and contingencies. Perry and Lindell (2003: 343) stress that it can be difficult for decision makers in a given operation to identify the predominant scenario and appropriate emergency plan, which in organization theory corresponds to the well-known problem of identifying the correct stimulus for the set ‘if-then’ chains (see March and Simon, 1993: 167ff.). Kolen and Helsloot (2014: 628–631) show that such deterministic planning cannot take account of the actual uncertainties facing decision makers on the incident ground, and that reasonable alternative options in emergencies are disregarded.

However, it appears to be difficult to develop alternatives in specific planning for measures to handle unpredictable events. At any rate, alternative proposals for emergency planning have hitherto been rare. Some researchers merely suggest greater optimization of existing scenario-based procedures (see e.g. Altay and Green, 2006; Oloruntoba, 2013; Wright and Goodwin, 2009). Perry and Lindell (2003: 342), by contrast, limit themselves to more general ‘principles of response’. Kolen and Helsloot (2014: 631) suggest taking better account in planning of uncertainties as such, for instance by immediately interpreting emergency situations as crises. Furthermore, the importance of developing a safety-oriented culture is often emphasized as being more relevant than planning (Alexander et al., 2009; Weick and Sutcliffe, 2007).

The role of geographical space in emergency planning

Since in our case geographical space plays a key role as a planning category, we need to consider it briefly. In this context, space has tended to play a secondary role in the literature. Space is a category, for example, in identifying significant geographical risk zones such as potential flood hazard areas (see Mishra et al., 2012). There are also strategies for preparing evacuation plans (see Perry et al., 1981). Moreover, the development of detailed emergency maps for informational purposes in emergencies (see Alexander, 2002) and the further development of geographic information systems (see Montoya, 2003) are of some importance. In all, space plays an important but usually only auxiliary role in emergency and disaster planning.

What makes a ‘good’ plan?

On the basis of the existing literature, general criteria for what constitutes a good emergency and disaster response plan can be identified. Critics of scenario-based planning stress that a plan should not be too detailed. This increases the risk of the so-called ‘paper plan syndrome’ (Auf der Heide, 1989: 33) because although exhaustive written plans give an impression of good preparation, they prove difficult to consult or can prove simply impracticable in an emergency. Plans should therefore be regularly used and reproduced on the ground, because as decision premises they are produced, reproduced and adjusted only in the course of actually making decisions (see March and Simon, 1993: 161; Seidl and Becker, 2006: 27). Instead of determining actions for concrete responses, plans should leave room for flexibility (Perry and Lindell, 2003: 342ff.). Emergency planning and emergency management should therefore be kept as separate as possible (Quarantelli, 2008: 324). Prospective responses should aim not only at speed but also be appropriate with respect to measures and resource utilization. To this end, it is sometimes necessary not to take immediate large-scale action in emergencies but first of all to inspect the situation and then decide whether further measures need to be taken (Perry and Lindell, 2003: 342).

Such planning should be centralized at the local authority level (Alexander, 2005: 161) and be interorganizational (Perry and Lindell, 2003: 343). The agencies involved should reach agreement and coordinate on this basis, and take cognisance of their mutual actions. If possible, interagency communication channels and operational hierarchies should be clearly defined (Moynihan, 2009). It is also important that planning be a continuous process (Quarantelli, 2008: 321) with constant coordination and ongoing development. For interagency coordination, it is important that there be a single plan for the organizations involved, which is clearly and unambiguously worded and compatible with other existing plans (e.g. in the individual agencies) (Alexander, 2005: 162).

The setting: The case of Düsseldorf

In what follows, we examine the Old Town Emergency Response Plan of the Düsseldorf municipality. The Düsseldorf municipality has general operative responsibility for non-police emergency management in the city and is therefore in charge of fire and rescue services and technical assistance of all sorts. This assistance is entrusted to an autonomous department of the city administration responsible for non-police emergency management – the fire and emergency department (FED). With a population of over 600,000, Düsseldorf has a FED counting some 1000 professional and 300 volunteer personnel.

Although the FED forms the core of non-police emergency management, it depends on interagency cooperation in both minor and major emergencies. For current emergencies, we counted some 80 organizations at the official level alone that cooperate with the Düsseldorf FED and often with one another (see Berthod et al., 2015). Some of these organizations also cooperate in planning for potential emergency and disaster responses. There are various emergency plans that are supposed to be drawn up in interagency cooperation, for instance to deal with unexploded ordnance or airport incidents. One such plan is the Emergency Response Plan under study.

It addresses the Düsseldorf ‘old town’, which still constitutes the centre of the city. It is criss-crossed by narrow historical lanes and has a high concentration of eating establishments (some 300). There are often a large number of visitors, attending major events like the annual carnival or who gather there on normal weekends. Dense development in the old town in combination with the often dense crowds make access to the area very difficult for fire and other response services. After problems had arisen in interagency cooperation during incidents in the old town, the FED, the police and the municipal office for public order decided in May 2011 to develop a joint response plan for the area. In collaboration with the traffic management office and the construction supervision authority, these three organizations consequently elaborated the Old Town Emergency Response Plan, which has been in force since 2013.

Methodology

The present approach was developed in the context of a qualitative field study on cooperation between public authorities and organizations in the safety and security fields in Düsseldorf. Given the primary responsibility of the FED for non-police emergency management, this was our point of departure. Over a period of 14 months in 2014 and 2015, we conducted our enquiry in the form of a snowball study. Setting out from our gatekeeper, the FED, we gained access to other organizations, which enabled us to access others in turn. We engaged in repeated periods of unstructured participant observation, accompanying incident command and coordination bodies during major events and disasters, including week-long participation in everyday emergency operations. We also conducted both partly structured and spontaneous informal interviews, and gained access to many documents such as operation reports, operational plans and minutes of meetings.

Beginning with a number of partly structured interviews on the various levels and characteristics of urban emergency and disaster management (Berthod et al., 2015), we were able to conduct first participant observation with two researchers at the carnival procession on Shrove Monday. This is the biggest regular event in the city, attracting a crowd of over 500,000. We accompanied the public authorities and organizations in charge of safety and security as they coordinated their activities during the event. Since it took place in the Düsseldorf old town, we came across the Old Town Emergency Response Plan for the first time. In the course of the event, we observed that the agencies involved acted essentially on the basis of this plan.

Although we had formed no a priori hypotheses before collecting our data, we had been inspired first from a public administration perspective by the existing literature on disaster studies and second, by general organizational sociology theory. It seemed to us that the Old Town Emergency Response Plan deviated from usual theoretical assumptions. We consequently focused part of our enquiry on the plan and its theoretical peculiarities. Our fundamental strategy was guided by what Robert Yin (2009: 130) has described as ‘relying on theoretical propositions’. Our point of departure was Niklas Luhmann’s decision theory. We then applied the analytical technique of ‘explanation building’ (Yin, 2009: 141). On the basis of our first interviews and observations, we formulated a number of initial assumptions about the theoretical particularities of the Old Town Emergency Response Plan, concentrating at this stage on describing the anomalous nature of the plan. We also directed our attention during the subsequent field study to obtaining more knowledge about the response plan. We were interested above all in observing how it was actually put into practice. We were participant observers both in everyday operations and at major events. In a number of rounds we collected new data that allowed us to revise our initial assumptions.

Overall, we drew on statements by 22 of many interviewees from 16 partly structured interviews and nine ethnographic interviews relating to the Old Town Emergency Response Plan. Ten of our interviewees were from the Düsseldorf FED, four from the police, three from the Red Cross, three from the German St John Ambulance and two from other organizations. Then there were the minutes and memoranda from our observations, as well as various documents. The interviews varied widely in nature. Sometimes they were comprehensive and lasted for a number of hours, sometimes they were brief ethnographic interviews conducted during observation (even in the course of operations). In keeping with our methodological procedure, there was no standardized questionnaire; we adapted questions to our constantly revised theoretical assumptions. The same is true of our observations, which became more and more focused as analysis proceeded. Important focal topics in interviews included the experience of response personnel with implementation of the plan on the incident ground; the background and history of the plan; interagency cooperation in operations on the basis of the plan; problems and difficulties in using it. Our analysis having reached a certain degree of saturation, we conducted five final interviews mainly to validate our findings (to some extent also in relation to recent operations).

The Old Town Emergency Response Plan

The Old Town Emergency Response Plan provides the basis for safety and security-related activities in the case of emergencies, disasters and major public events undertaken by each organization within the old town, notably by the FED, the police, the traffic management office and the office for public order, as well as organizers of major public events. The plan applies throughout the year and around the clock, both in everyday operations and for every type of event, and serves as a basis for the activities of other response agencies. The situation in the old town is constantly monitored and classified by the responsible police station. Situation classes A to C indicate the density of crowds in the various areas of the old town. Classes 1 to 4 indicate whether an event or event-like occasion is taking place and whether or not a ‘concrete response situation’ has already developed or not. On this depends whether the organizers have to be involved in making decisions. What sort of emergency situation has arisen, however, is irrelevant:

It doesn’t matter whether it’s a heart attack or an explosion, things go consistently by the same standard. (interviewee 1, FED)

The default class is 1-A unless the responsible police station reports that a higher level situation has been registered. The response situation class in the old town determines the activation of predefined measures, e.g. the FED, police and office of public order contacting other agencies or exchanging liaison officers. At this point in time, the response plan prescribes contacts between agencies and the point at which these predetermined communication channels are to be put in operation. Moreover, space-related measures are usually defined, notably installing road blocks, cancelling taxi stands and establishing standardized access/egress routes for the emergency services, as well as evacuation measures.

Depending on the classification of the situation and, especially on where an incident has taken place, the plan defines the access routes to be used by emergency services and routes for evacuation should the need arise. Access/egress routes are coupled directly to the location of an incident – to the exact number of the building:

In the Andreasstrasse, the entire street is reached via access C … Access to Bolkerstrasse is subdivided. From 1 to 25 access C is used, and from 30 to 57 access D. (interviewee 24, FED)

The core of the concept is a site plan (Figure 1) that brings together a range of information on a map of the old town such as access/egress routes (‘Zufahrt’ A, B, C, etc.) and potential cordon points (small white/red circles). The map also shows marshalling areas for the fire and emergency service (red and partly red circles), police (green triangles) and public order office (blue squares), as well as various sites for first-aid posts (partly blue circles). Patient transfer points are also shown (yellow and partly yellow circles). They come into operation when a class C situation is reached. In this case, rescue service units no longer bring vehicles into the relevant areas of the old town, because they would be heavily obstructed. Rescue personnel then proceed on foot, transferring patients to ambulances at the fixed transfer points. A leading member of the FED (interviewee 71) described a typical instance:

The police call and say the crowds in the old town are so dense we can’t get through. According to the Old Town Emergency Response Plan we have to do something [Class C]. I would then pass on the information and say we also have difficulty getting through with our vehicles and even on foot. Emergency medical service agencies must set up a post outside so that we pass people on there [first-aid posts]. We get them out of there on stretchers and then they’re treated in emergency vehicles [patient transfer points].

Figure 1.

Düsseldorf old town site plan for emergency response.

Moreover, at the C level reconnaissance is required in the event of more serious incidents to assess the situation. The plan also provides for keepout areas (red areas in Figure 1) where none of the agencies on the ground are permitted to park their vehicles (unless, of course, the incident has taken place directly in or in the immediate vicinity of such areas) in order to give emergency service personnel room to act.

Implementation of the plan

It took time for the Old Town Emergency Response Plan to be implemented, and naturally it works only if we keep to it. Starting with the fire and rescue services, the police, public order office, as well as medical services engaged for events. So if you elaborate such expensive, complex plans and response concepts, everyone necessarily has to know them and abide by them. Such a plan has to live. (interviewee 27, FED)

During our enquiry, we paid special attention to the actual implementation of the plan, to whether and how it was used. In the field we observed, among other things, that the site plan (Figure 1) forming part of the Emergency Response Plan was displayed in many places by various agencies. St. John Ambulance, at least, had even integrated the plan in their geographic information system, where the site map was ‘superimposed on and scaled to fit’ the ground plan of the area (interviewee 65, St. John Ambulance). Furthermore, we observed the use of the plan in actual operations on more than one occasion.

Our interviewees also confirmed that implementation of the response plan was well advanced and was beginning to prove successful. For instance, a leading member of the Red Cross had this to say:

What definitely works much better is where a rescue resource [e.g. an ambulance] is brought in from, … or where a crew park their vehicle and then perhaps cover the last few yards to the scene on foot with their equipment. So everyone knows very precisely that they can’t simply park their vehicles in the red zone. Such things work every day when the old town is crowded; it has to be said, it really works on a practical everyday basis. (interviewee 67)

Although members of various organizations also told us about specific problems in implementing the plan and continuing coordination difficulties (for example, one agency did what it had agreed to do only after repeated demands), the people involved are working on eliminating these difficulties. A leading police officer (interviewee 54), for instance, stressed that at smaller events such as the Christmas market, event-related elements of the plan would in future also be used to drill personnel in implementing it. Overall, the plan seems to have become well established. According to the statistics, the FED, for example, have apparently reduced emergency response times by several minutes since introduction of the plan. And the response plan has already provided a model for similar concepts, for example, to deal with airport incidents.

Geographical space as a key planning dimension

Geographical space is the crux of the Old Town Emergency Response Plan. The particular characteristics of the old centre of the city required the plan to be limited to a defined area. The old town is a venue for events and social life; together with the physical properties of the area and the wealth of eating establishments it houses this made it a suitable entity for planning purposes. From the outset of the planning process, geographical space was thus the basis for the interagency elaboration of the emergency response plan. A police officer described how things began:

That was the background: what areas should not be occupied by the police, as well, what evacuation routes we needed from the old town; that was the basis. And the second step was added, namely where transfer points were to be sited for emergency services. (interviewee 66, police)

In drawing up the plan, the organizations involved defined their spatial requirements for maximum personnel deployment, for instance in the form of marshalling areas. To avoid encumbering the scene of operations with too many personnel, they first assemble in the marshalling areas before actually being deployed.

This geographical focus is accordingly reflected in the final response plan. Space is the yardstick in the plan for the continual assessment of the situation in the old town. Furthermore, if an incident occurs, space is the decisive factor for the measures to be taken. In contrast to scenario-based response plans, in which various scenarios such as ‘bomb alert’ or ‘cellar fire’ are predefined, the Old Town Emergency Response Plan concentrates primarily on where an incident occurs and not on what it is. It is initially unimportant whether a bomb alert, a cellar fire, or something else is involved.

Once the first space-related measures have brought the first response crews onto the scene, further assessment and decisions about what concrete measures are to be taken to handle the specific type of incident are left to the command personnel. For major incidents, response operations are basically under the command of the municipal FED (in the case of non-police emergency management) or the police (in the case of emergencies related to criminal activity). Where operations are being directed by the FED, all participating agencies are subordinated to the department or assist it within the framework of their own responsibilities. For example, the office of public order and the police carry out appropriate measures such as installing cordons, evacuation and directing traffic. Similarly, in the case of mass casualty incidents, emergency medical service agencies such as the Red Cross or St. John Ambulance answer to the FED.

The FED naturally has standard operating rules (see e.g. Südmersen and Cimolino, 2014), special decision-making rules (see e.g. Graeger et al., 2009) and established procedures. As far as content is concerned, however, these rules are unspecific and/or modular in structure. Standard operating rules determine, for example, how response vehicles are to be manned, what radio channels are to be used, and how tasks are to be allocated when fighting fires in buildings. The rules for command decision making lay down how decisions are to be made but not which ones. By their nature, these procedures can be combined on a modular basis and are not the subject of the interagency response plan.

In combination with the Old Town Emergency Response Plan, these rules allow operations to be scaled up or down and continuously adapted to events. While incidents can rapidly change – for example if a cellar fire develops into a shopping mall conflagration with many injured – those in charge can constantly adapt their operational decisions without being dependent on any initial definition of the incident. The plan works independently of any scenario:

The whole thing is really independent of scenarios. Whether we have a police operation, say, an unexploded bomb or the threat of an attack or whatever … or there’s a fire and the FED have to get to the building and, let’s say, push their way through crowds and make room or whatever, this fits in with all scenarios, it is independent of any scenario. … And it fits in with whatever happens. Even if a plane crashed in the old town, we could start our work in coordination through the Old Town Response Plan without having to think about it. … It was important for us to work independently of a scenario. Otherwise, we tie ourselves down too much in advance, and then someone says ‘it’s not that sort of situation, that’s not what it’s about’. (interviewee 77, police)

Possibilities for conditional programming in the response plan

At first glance, the importance of geographical space for emergency planning is surprising. As we have seen, space has hitherto played more of a secondary role as a planning category in the literature. In this respect, the category space has so far never been taken as the key dimension of planning. In this regard, the plan under study presents an interesting case. Using space as the basis for planning clearly offers programming possibilities that have hitherto been neglected in previous research.

Although the Old Town Emergency Response Plan also contains many conditional programming elements, it differs from the usual scenario-based concepts. There are thus no set initial scenarios to activate large-scale chains of measures. But the plan contains a large set of parallel and independently activatable conditional programmes. We identify two categories of modular programming with different functions and addressing different problems. The one mainly concerns the initial stage of incident response. They are ‘if-then’ programmes taking the form of ‘where-then’. The other category offers resources for purposive programming, so that we have programme nesting. We now go into the two in greater detail.

‘If-then’ as ‘where-then’

Programmes that come into play in the initial phase of response operations are an important component of the Düsseldorf Old Town Emergency Response Plan. They standardize access/egress routes and organize the precautionary manning of cordon control points. This programming compensates for the typical absence of command decision makers on the ground when an incident occurs. The situation is different with large-scale events such as concerts or festivals. Commanders are then regularly already on the scene monitoring such events and can decide themselves on entry and exit routes for emergency response services.

The measures to be taken for everyday incidents in event-free periods depend primarily on the location of the incident and not on the type of incident involved. The typical ‘if-then’ conditional programme thus takes the form of a ‘where-then’ programme (i.e. if something happens in a specific place, a measure is activated). This operationalization of programming addresses the problem of stimulus identification in activating programmes to allow predefined measures to be initiated at all (March and Simon, 1993: 168–171). In the present case, the stimulus has typically to be identified by a dispatcher on receipt of an emergency call. Information about the location of an incident – possibly from the caller’s mobile GPS coordinates – generally makes this identification easier than the description of the incident given by the caller.¹ As a support, large numbers are displayed on lampposts along the Rhine embankment (see Figure 1, indicated by the rectangular white boxes) to enable callers to determine their precise location.

Programme nesting

On the basis of the initial ‘where-then’ programmes, temporary and provisional measures are activated until response commanders arrive on the scene. Some aspects of planning therefore concentrate on making the necessary resources (personnel and equipment) available on the incident ground as usefully and as fast as possible. In concrete terms, it is not primarily a matter of what measures the agencies involved ought to take in an emergency but how they ensure that the necessary means are made available as rapidly as possible so that further (initially flexible) measures can be taken at all and then adapted to the changing circumstances of the operation.

Another part of the Old Town Emergency Response Plan accordingly deals with communication between agencies (communication channels). Thus, from certain classes of situation onwards, the response agencies are required to exchange liaison officers and agency response commanders to coordinate their activities. These – also conditional programme – elements ensure interagency coordination on the scene. Moreover, the Old Town Emergency Response Plan provides for many potential measures to be activated by the responsible decision makers should the need arise. Apart from a detailed evacuation plan based on predefined access/egress routes, there are, for example, a concept for cordoning off certain streets, information sheets and announcements for visitors to the old town, plans for moving taxi stands and clearing restaurant and cafe terraces. The Response Plan sets no criteria for activating such measures but merely stipulates which agencies have the right to do so. For example, the police or the office of public order are allowed to install cordons in accordance with the set access control plan. If an incident has already occurred, for instance, the police, who are always on the scene owing to their location in the old town, could decide to clear streets in accordance with the evacuation plan while fire and rescue services move in by the predefined access routes.

As we have seen, the Response Plan also provides for setting up, accessing and operating first-aid stations, transfer points and marshalling areas. These elements of the plan can also be interpreted as conditional programmes. Although many relate strongly to geographical space, their activation is decided by response commanders on the incident ground. This means a shift in the fundamental logic of planning. Whereas the scenario-based plans described are basically input-oriented, in the present case we are dealing rather with an end/means schema. But the specific purposes of the operation (cellar fire, carousel accident, gas leak, etc.) are left open in the planning.

Yes, it’s like a modular toolkit with the transfer points marked in. I don’t have to use all of them; maybe I don’t have to use any. But if I don’t have a better idea, then I say ‘I’ll take transfer point 27’, I only have to radio them the info, then everyone knows at once. That means the Response Plan also includes this diagram, it has all the tools. What ones I use is up to me. (interviewee 1, FED)

As we have seen, this form of resource programming through conditional programmes (programme nesting) is not untypical (see Luhmann, 2006: 265); however, it is striking that in the present case concrete purposes are not programmed at all – indeed, they cannot be programmed. The particular incident and the consequent specific operational purpose are left open in the planning, which concentrates on making resources available for indeterminate purposes.

Shift to leadership decisions – the growing importance of personnel

In the event of an incident, the programmatic nature of the plan then shifts the onus for decisions essentially onto the leadership. This increases the complexity facing incident commanders.

One big problem is you can evacuate only when you have clearly understood what the situation is. If you simply evacuate because an alarm has been given, you clear x streets and cafe terraces – you know how it looks in the old town. And the people just decamp without paying [in restaurants, pubs, etc.]. This means you cause great economic losses and have a hard time explaining things to the proprietors on your own. (interviewee 9, FED)

In scenario-based plans, the onus for decisions is determined above all by identification of the programme stimulus and consequent measures and procedures. Success depends essentially on the knowledge resources of the agencies involved, on how well elaborated and complete programme sequencing is, on far-reaching anticipation and correct identification of the initial cause of the incident (and the manageable definition of the stimulus in the programme itself). In the present case, by contrast, command personnel have constantly to make independent decisions on the basis of ongoing situation assessment. Complexity reduction is left to the personal qualities of the responsible commander. The change in the form of programme thus ultimately reduces the importance of programmes for decision making and increases the importance of the decision premise (command) personnel and their assessment of the situation. Dependence on well trained and chosen command personnel increases.

Functions of programming

Flexibility in operation

Having discussed the organization-theoretical particularities of the plan, we turn now to its positive functions. The yardsticks mentioned for a good emergency response plan provide the background. Whereas in scenario-based planning the emergency situation has to be assessed as fast as possible in order to activate the correct (predefined) measures provided for under the programme, means-oriented planning permits (within the context of managing the specific incident) the situation to be first examined and assessed. Only within the framework of continuous situation assessment do command personnel make independent decisions on what measures are to be taken. This programme nesting thus provides a high level of planability together with an elasticity within this planning that is difficult to achieve with the scenario approach. This means that the Old Town Emergency Response Plan, unlike conditionally programmed plans, can also keep emergency planning separate from emergency management.

Conditional programme resources are modular in structure, so that, for example, taking a specific transfer point into operation does not trigger further programmes. This allows operations to be scaled up, if, for instance, what initially appears to be an everyday incident develops into a full-scale disaster. Whereas scenario-based plans can start with their disaster scenarios only when an actual incident has reached a climax, in the Old Town Emergency Response Plan modular conditional programmes can be combined and scaled up.

We can scale up everything from a heart attack to an explosion without having to tell anyone at any stage that things have to be done differently because we’re into the Old Town Emergency Response Plan; that is the case from the start, and not as with some response plans that delay the daily process: everything is really a continuous transition, and the individual firefighter who is perhaps called in when an explosion occurs won’t notice any difference. (interviewee 1, FED)

Developing a common language

Given its form, the Emergency Response Plan presented constitutes a common language for all the agencies involved. The key components of the plan are thus named and marked on a common map, the site plan (Figure 1). It is used on the ground by all agencies. Since these agencies had participated in developing the plan, there can be no danger of semantic misunderstandings. This appears to facilitate implementation within individual organizations, since they can translate the elements of the plan into their own structures and even develop them without argument. For example, the police use the Response Plan transfer points for internal prisoner transfers. There is hence a common plan that is clearly and understandably formulated and which is compatible with the internal plans of the individual agency.

In the present case, all the organizations involved accordingly know who is talking to whom and what is meant when specific points are named:

We have a map, we have a grid reference, it’s the same for everyone, and that makes things easier from the outset in such an operation. There are no orientation problems any more, there are no coordination problems. … a preparatory operational landscape has been superimposed on the old town that is the same for everyone. And where everyone is talking about the same thing. If, for instance, someone says ‘I need marshalling area number so and so’, everyone knows: ‘Aha that’s for the FED, for the police, they access from there, the route has to be kept clear’, and so on. (interviewee 76, police)

This brings us to another aspect of the function of this common language. Because, for instance, marshalling areas, transfer points and first-aid stations are already provided for as conditional programme resources, incident commanders have to decide only which of the numbered points are to be used. The agencies involved therefore do not need to reach agreement on what measures are to be taken and where, or on what they are to be called. Much of the otherwise necessary interagency coordination of what personnel are to be deployed and where, etc. is therefore superfluous. Basal coordination requirements between agencies are consequently reduced.

In the present case we are dealing with programmes, so that they can also be combined with other types of structure. We have pointed to the importance of establishing clear communication channels and the appropriate operational hierarchies. The Old Town Emergency Response Plan provides for both. Whereas in Germany operational hierarchies are regulated above all by law – with concomitant, synchronized status ordering of fire service, emergency medical service, technical support (THW) and auxiliary agencies – communication channels are additionally set by the Response Plan. What is more, certain communication tools are kept in reserve in the form of programmes activatable where required. For example, a predefined coordination group can be activated in which liaison officers organize interagency coordination.

Facilitating the integration of external organizations

Establishing a common language enables agencies and personnel from elsewhere to be immediately integrated. Such external resources are needed relatively frequently. In the past, this aspect has led to coordination problems, for example when police from other parts of the state are called in.

These external units now need only to be supplied with the relevant site plan (Figure 1), which provides them with all the information they need to take prompt action when instructed without the need for further coordination. One firefighter explained (interviewee 9):

Anyone [with no prior information] only needs to look at the map to know what’s what. … Even if I haven’t used it for two years, I can still find my way around. … [The police] also have external personnel and also speak the same language. Go to 17, then he goes to 17. It used to be: go to the snack stand. But where? Snack stand. Don’t know exactly.

These instructions are generally given – at least in the case of non-police emergency management – by local commanders, since incident command is always in the community affected. Local commanders have been trained to handle the (entire) Old Town Emergency Response Plan and use it regularly.

Limitations and conclusion

There are some limitations to the approach under study. Our investigation is a single case study looking at the possibilities of programming that have been achieved by focusing on geographical space as a planning category. However, in our view this planning procedure is suitable only for specific areas and not for all fields of emergency planning. Although space must always be regarded as a strategic resource in emergency responses, we see the form of space-based emergency planning we have presented rather as a possibility for achieving both robust and flexible planning in very complex social hazard zones. For such areas – e.g. dense inner cities, airports, railway stations – there is a complex connection between large crowds, difficult local conditions and differing organizational responsibilities. It seems to us, however, that planning whole cities or even rural areas down to the last detail is neither useful nor practicable. Incidents in ‘normal’ areas presumably require a more general planning approach with established interagency communication channels, general coordination and mutual training. We should also recall that scenario-based planning can be useful in many fields where specific and probable risk scenarios are involved. This is particularly the case, for instance, in dealing with unexploded ordnance or with flooding or evacuation planning. We therefore see the planning approach presented not as a general alternative but as supplementing the tools offered by other planning approaches.

Our study shows that the application of sociological theory can bring new insights in emergency and disaster research. We have thus been able to show what programming possibilities there are for disaster planning over and beyond the usual deterministic conditional programme planning. On the one hand, there are possibilities for stimulus setting for conditional programme planning components that are not based on the definition of scenarios but, as in the present case, operate with location definition. This enables ‘if-then’ programming in the form of ‘where-then’ programming. On the other hand, we establish that means for purposive programmes can be specified through conditional programmes, even though the relevant specific purposes are not defined. The two possibilities observed permit ex ante decisions while retaining or increasing flexibility. They offer stable planning that can at the same time be used for many unexpected events. Both ‘where-then’ programming and space-based programme nesting allow responses to be reliably defined while ensuring openness for the actual purposes of intervention.

Our study thus also provides two general contributions to decision research. For the first time we have empirically examined the theoretical figure of programme nesting. We show that within the framework of programme nesting, conditional programmes can be set as resources for purposive programmes without having to specify the purposes. Through conditional programmes, means can thus be programmed without defined ends, with a shift to nested programmes only when concrete purposes (e.g. operational goals) are set.

We assume that there are other possibilities for programming in emergency and disaster response planning that we have not been able to demonstrate with our single case study. With reference to Lindell (2013), we see further scope for research in which sociological theory can prove useful. Moreover, we see the concept of decision programming as only one of many possibilities for applying sociological theory to disaster studies. Future research could, for example, usefully look into others of the decision premises mentioned and into societal context conditions.

Footnotes

Acknowledgements

We gratefully acknowledge the support of the numerous individuals and organizations we were able to observe and interview in the course of our investigation, and especially Thomas Hußmann with the Düsseldorf fire and emergency department. We also want to thank Jörg Sydow for his encouraging feedback and guidance in the project, Maja Apelt and the participants of the research colloquium on organizational sociology at the University of Potsdam in September 2014, as well as Stefan Kühl and the participants of the research colloquium on organizational sociology at Bielefeld University in January 2015. Last but not least we wish to thank Svenja Hammer for extensive feedback on prior versions, Rhodes Barrett for his help with translating our original manuscript, as well as the editor Eloísa Martín and the anonymous reviewers for their constructive suggestions.

Funding

This research received funding from Deutsche Forschungsgemeinschaft (grant SY 32/6-1).

Notes

Author biographies

Michael Grothe-Hammer, MA, is a research associate at the Department of Management of the Freie Universität Berlin and at the Institute of Social Sciences at Helmut Schmidt University Hamburg. He studied sociology at Bielefeld University. His current research interests include organizational sociology, qualitative research methods, sociological systems theory, high reliability theory and sociology of disasters. His work has appeared in the Journal of Contingencies and Crisis Management, Organizational Research Methods as well as Systems Research and Behavioral Science.

Olivier Berthod is a lecturer at the Department of Management of the Freie Universität Berlin, where he also received his Doctorate. He has held teaching appointments at the University of Cologne, the Toulouse School of Management, the ESCP Europe and the German Aerospace Research Centre among other institutions. His research focuses on networks, governance and leadership in the face of uncertainty. His work has appeared in various outlets and edited volumes, and reports on issues as diverse as the management of UNESCO World Heritage sites, team dynamics in cockpits, the organization of food safety, and the prevention of crises and catastrophes in large urban areas.

References

Ahrne

Brunsson

(2011) Organization outside organizations: The significance of partial organization. Organization 18(1): 83–104.

Alexander

(2000) Scenario methodology for teaching principles of emergency management. Disaster Prevention and Management 9(2): 89–97.

Alexander

(2002) Principles of Emergency Planning and Management. Oxford: Oxford University Press.

Alexander

(2005) Towards the development of a standard in emergency planning. Disaster Prevention and Management 14(2): 158–175.

Alexander

Bramati

Simonetta

(2009) Emergency preparedness training and education in Lombardy Region, Italy: Survey of supply and demand. Natural Hazards Review 10(3): 77–83.

Altay

Green Walter

III (2006) OR/MS research in disaster operations management. European Journal of Operational Research 175(1): 475–493.

Apelt

(2014) Organisationen des Notfalls und der Rettung: Eine Einführung aus organisationssoziologischer Perspektive. In: Jenki

Ellebrecht

Kaufmann

(eds) Organisationen und Experten des Notfalls: Zum Wandel von Technik und Kultur bei Feuerwehr und Rettungsdiensten. Münster: Lit Verlag, pp. 69–84.

Auf der Heide

(1989) Disaster Response: Principles of Preparation and Coordination. Toronto: CV Mosby Company.

Berthod

Grothe-Hammer

Sydow

(2015) Some characteristics of high-reliability networks. Journal of Contingencies and Crisis Management 23(1): 24–28.

10.

Carter

(1992) Disaster Management: A Disaster Manager’s Handbook. Manila: Asian Development Bank.

11.

Federal Emergency Management Agency (2010) Comprehensive Preparedness Guide 101: Developing and Maintaining Emergency Operations Plans, 2010. Washington, DC: Federal Emergency Management Agency, US Department of Homeland Security.

12.

Frosdick

(1997) The techniques of risk analysis are insufficient in themselves. Disaster Prevention and Management 6(3): 165–177.

13.

Graeger

Cimolino

Vries

Südmersen

(2009) Einsatz- und Abschnittsleitung: Das Einsatz-Führungs-System. Heidelberg: ecomed.

14.

Harrald

Mazzuchi

(1993) Planning for success: A scenario-based approach to contingency planning using expert judgment. Journal of Contingencies and Crisis Management 1(4): 189–198.

15.

Kolen

Helsloot

(2014) Decision-making and evacuation planning for flood risk management in the Netherlands. Disasters 38(3): 610–635.

16.

Kühl

(2001) Über das erfolgreiche Scheitern von Gruppenarbeitsprojekten: Rezentralisierung und Rehierarchisierung in Vorreiterunternehmen der Dezentralisierung. Zeitschrift für Soziologie 30(3): 199–222.

17.

Kühl

(2013) Organizations: A Systems Approach. Farnham: Gower.

18.

Lambert

Sarda

(2005) Terrorism scenario identification by superposition of infrastructure networks. Journal of Infrastructure Systems 11(4): 211–220.

19.

Lindell

(2013) Disaster studies. Current Sociology 61(5–6): 797–825.

20.

Lindell

Perry

(1992) Behavioral Foundations of Community Emergency Planning. Washington, DC: Hemisphere.

21.

Luhmann

(1964) Lob der Routine. Verwaltungsarchiv 55: 1–33.

22.

Luhmann

(1966) Recht und Automation in der öffentlichen Verwaltung: Eine verwaltungswissenschaftliche Untersuchung. Berlin: Duncker & Humblot.

23.

Luhmann

(1995) Social Systems. Stanford, CA: Stanford University Press.

24.

Luhmann

(2003) Organization. In: Bakken

Hernes

(eds) Autopoietic Organization Theory: Drawing on Niklas Luhmann’s Social Systems Perspective. Copenhagen: Copenhagen Business School Press, pp. 31–52.

25.

Luhmann

(2006) Organisation und Entscheidung. Wiesbaden: VS Verlag für Sozialwissenschaften.

26.

March

Simon

(1993) Organizations. Cambridge: Blackwell.

27.

Mezey

(2011) Recent methods of homeland security and safety planning. In: Mezey

Strunz

(eds) Führung von Einsatzkräften. Frankfurt am Main: Peter Lang, pp. 121–138.

28.

Mishra

Fuloria

Bisht

(2012) Enhancing disaster management by mapping disaster proneness and preparedness. Disasters 36(3): 382–397.

29.

Montoya

(2003) Geo-data acquisition through mobile GIS and digital video: An urban disaster management perspective. Environmental Modelling and Software 18(10): 869–876.

30.

Moynihan

(2009) The network governance of crisis response: Case studies of incident command systems. Journal of Public Administration Research and Theory 19(4): 895–915.

31.

Oloruntoba

(2013) Plans never go according to plan: An empirical analysis of challenges to plans during the 2009 Victoria bushfires. Technological Forecasting and Social Change 80(9): 1674–1702.

32.

Perry

Lindell

(2003) Preparedness for emergency response: Guidelines for the emergency planning process. Disasters 27(4): 336–350.

33.

Perry

Lindell

Greene

(1981) Evacuation Planning in Emergency Management. Lexington, MA: Lexington Books.

34.

Quarantelli

(2008) Ten research derived principles of disaster planning. In: Boin

(ed.) Crisis Management: Volume 1. Los Angeles: Sage, pp. 321–324.

35.

Rommelspacher

(2011) Automatisierung von Führungsentscheidungen. Wiesbaden: Vieweg +Teubner.

36.

Seidl

Becker

(2006) Organizations as distinction generating and processing systems: Niklas Luhmann’s contribution to organization studies. Organization 13(1): 9–35.

37.

Simon

(1997) Administrative Behavior: A Study of Decision-making Processes in Administrative Organizations. New York: Free Press.

38.

Südmersen

Cimolino

(2014) Brandbekämpfung im Innenangriff. Heidelberg: Hüthig Jehle Rehm.

39.

Ten Brinke

WBM

Kolen

Dollee

. (2010) Contingency planning for large-scale floods in the Netherlands. Journal of Contingencies and Crisis Management 18(1): 55–69.

40.

Tierney

(2007) From the margins to the mainstream? Disaster research at the crossroads. Annual Review of Sociology 33(1): 503–525.

41.

Walker

Giddings

Armstrong

(2011) Training and learning for crisis management using a virtual simulation/gaming environment. Cognition, Technology and Work 13(3): 163–173.

42.

Wallace

de Balogh

(1985) Decision support systems for disaster management. Public Administration Review 45: 134–146.

43.

Weick

Sutcliffe

(2007) Managing the Unexpected: Resilient Performance in an Age of Uncertainty. San Francisco: Jossey-Bass.

44.

Winter

(1998) Politikum Polizei: Macht und Funktion der Polizei in der Bundesrepublik Deutschland. Münster: Lit Verlag.

45.

Wright

Goodwin

(2009) Decision making and planning under low levels of predictability: Enhancing the scenario method. International Journal of Forecasting 25(4): 813–825.

46.

Yin

(2009) Case Study Research: Design and Methods. Thousand Oaks, CA: Sage.