How to balance between embedding and deviation?
It frequently occurs that organisations deviate from their original policies when carrying them out, because of, for example, new information and insights, demands of stakeholders or problems arisen during further implementation. Based on case studies at four housing associations in the Netherlands, this paper will deal with the difficulties on daily decisions that are made during the implementation process. The paper will present a conceptual model in which the relation between embedding and deviation is shown. Then the paper will go into a number of measures that can be taken by housing association to assure that policies are embedded. After that, the paper will shift focus to deviations of policy. I will present a conceptual model that makes housing associations more conscious of their actions and helps them to balance between embedding and deviation.
Introduction: the implementation problem
When a policy has been set, the implementation starts. Housing associations, like other organisations, want their policies to be implemented in practice. However, they also know that sometimes it could be wise to deviate from their policy. For example due to new information and insights, the demands of stakeholders or just because too many problems are expected during further implementation. Based on case studies at four housing associations in the Netherlands, this paper will go into the difficulties on daily decisions that are made during the implementation process. The paper will present a conceptual model in which the relation between embedding and deviation is shown. Then the paper will go into a number of measures that can be taken by housing association to assure that policies are embedded. After that, the paper will shift focus to deviations of policy. I will present a conceptual model that makes housing associations more conscious of their actions and helps them to balance between embedding and deviation.
The starting point of this paper is the problem that the policies of organizations in practice are not always realized. A major problem with implementation is that it is often thought of being easy. This is evident from the statement “if the policy has been established, it is only a matter of executing”, used in different versions by many employees of various organizations. The word ‘only’ suggests that establishing a policy requires more attention than the implementation. It is questionable whether implementation is really that easy. The formulation of Ekkers (2002:115) suggests not. He writes: when a policy is adopted, it should be implemented, at least that is usually intended. Otherwise it was not established at all. If we define implementation of housing management strategies – in line with the more general definition of Maarse (1998:102) – as any activity carried out in order to realize concrete measures from the policy in a physical sense. Between establishing a measure like ‘demolition of this property and new construction at this location’ in the strategic housing management plan and the completion of the new homes many efforts have to be made by internal and external staff. As Day and Wensley (1983:86) emphasize, such process is more complicated than suggested by many process models.
If we focus on housing associations, the problem of embedding and implementing policies has gained importance. The housing associations are private organisations with a social task. Since 1995, they are no longer financially dependent on the government. An important consequence of this change of regulations is that housing associations more often can (and should) decide on measures regarding their housing stock. Since 1995, therefore a growing number of housing associations have developed a strategic housing management plan, in which they indicate how they want to steer their housing stock. Strategic housing management fills the gap that arose through the abolition of state control to the housing associations. In the first years after the change of regulations, housing associations have paid much attention to policy making. The process model of Van den Broeke (1998) is both a reflection and a cause of this trend. According to various studies, strategic housing management became one of the main policy themes of housing associations at the beginning of this century (Heeger and Nieboer 2003).
At this moment, the strategic housing management plans have been used by most housing associations for a while already, making a longitudinal study possible. To see what (f)actors are important in this implementation, at four housing associations a number of projects have been followed in time. The investigated housing associations differ in size and scope. There are two housing associations which operate in small villages and rural areas (Groenveld Wonen, 2000 dwellings and Leyakkers, 6500 dwellings). Two other housing associations own property in large cities (Bo-Ex, 8000 dwellings and Vestia Rotterdam Noord, 7000 dwellings). At these housing associations a total of 15 implementation projects have been followed. The projects vary in size from the removal of a partitioning wall, to a large-scale neighbourhood approach. Other projects include renovation, demolition and new construction and the buying and selling of buildings and locations.
The detailed description of the cases (Dankert 2008) has shown that many projects are changed during implementation, implicating that the relationship between the established policies and projects is rather weak. All projects investigated showed at least some changes during the course of implementation. In one third of all cases the outcomes of a projects deviates to such extent that this also affects the original overarching policy objectives. The conclusion that the realized measures do not always comply with the original policy is not surprising. Others have shown that this is often the case in policy implementation (e.g. Pressman and Wildavsky 1984; and Flyvbjerg 1998).
Housing associations make strategic plans for their housing stock. Based on a number of overarching objectives, dwellings are designated for letting, renewal, sale or demolition. Such policies are embedded in the organization by translation of the policy to other places in the organization. These other places include the annual budget, annual plans, information in the primary databases, project plans and agreements with third parties. Via these routes policies find their way within the organization. During implementation processes minor and major daily decisions that affect the performance are made. It is at these moments that, whether or not aware, the choices are made between complying to the original policy or to depart from it.
A lot of variables are important when it comes to embedding a policy. Literature on implementation is mainly on the possibilities to meet the proposed policy during the implementation (see for example overview papers of O’Toole 1986 and 2000; Matland 1995). Research on implementation of policy tends to find a lot of variables that influence the implementation. In one of his review studies, O’Toole (1986) counted over 300 different variables. This huge amount of variables does not only show the difficulties of implementation practices, but it also should have consequences for research on implementation issues. As Matland (1995:146) has correctly argued, a literature with over 300 variables needs more structure rather than more variables. In this paper I shall only highlight the variables that meet two requirements. First these variables should pop up from the cases, thus being relevant to housing associations. Secondly, in this paper I will only take variables into account that can be influenced directly by housing associations themselves.
In literature on housing management, the active implementation of such policy is also known as the ‘embedding’ of policy in the organization (Eskinasi 2008, Van Os 2007). In essence, embedding a policy is largely about the establishing of connections. In this case between the strategic housing management plan and for example a computer system or a budget. Because a housing association cannot implement on its own, also links with other stakeholders have to be established. Officials, aldermen, building materials, contractors, financial resources, architects, and others contribute to implementation.
Even if policy is firmly embedded, housing associations sometimes depart from previous policies. These deviations are made sometimes knowingly and sometimes unknowingly. If a housing association is aware of this policy deviation, it is mostly justified on the basis of new questions from stakeholders or blockages of the original plan. In the cases, deviate from the policy is not always perceived as negative. Many respondents talk about deviations from the policy in terms of ‘taking opportunities’, ‘flexibility’ and ‘taking into account new insights.’ However, housing associations also believe that policy implementation is important. In the implementation literature the issue of departure from policy has not been discussed a lot. From the perspective of science of law Terpstra and Havinga (2001) introduce different types of implementation. The type ‘professional policy implementation’ has been described by them as follows:
“If considered necessary, professionals may decide to deviate from the formal rules or to create (their own) new rules. In contrast to the bureaucratic style, in the professional style a correct application of formal rules is not viewed as the main task. More weight is given to the achievement of goals such as environmental protection, adequate social assistance, or fairness in decisions.” (Terpstra en Havinga 2001:105)
Another perspective is that of policy evaluation. Among scientists from this field there is a debate on how the results of policy must be evaluated. On the one hand, there is a school which advocates the conformity of results to the original policy. On the other hand, there is a growing group of authors who claim that the results of the policy in the first place must be assessed on their performance (Faludi and Korthals Altes, 1994).
From the cited authors above – Terpstra and Havinga, and Faludi and Korthals Altes – we can conclude that sometimes deviation from policy can be a deliberate choice rather that something to be avoided at all times. This conclusion also arises from the cases. The following examples also make this clear. In target group selection for a new build apartment complex by Groenveld Wonen, the housing association chose at the last moment to let the apartments to young people, while in the original policy housing for the elderly is strongly addressed, and the homes are also suitable for this end. It appeared that these stakeholders had been called for this deviation. An example of deliberate policy deviation at a smaller scale is (at Bo-Ex) the example that a partitioning wall was not removed (although this was part of the policy) in some cases, because the new tenant had requested so. In both examples, the solutions seem to be legitimate at first sight, despite the policy deviation.
The aim of this paper is to introduce a conceptual model on embedding and deviation of policy, and to show how this model could be worked out into a model that serves as a handle to housing associations in their daily work. From the introduction above two opposing movements can be seen. On the one hand, daily decisions aim at implementing the strategic housing management plan. On the other hand, daily decisions may deviate from the policy. Embedding policy is needed for implementation. At the same time, flexibility is needed in order to deviate from policies when that is necessary. The relation between embedding and deviation of policy as described above has been summarized in the following conceptual model.
The conceptual model is heavily based on the due process model that has been developed by Bruno Latour (2004) and Michel Callon (1986). There are different ways to steer the decisions on embedding and deviation. In the following I want to introduce the due process model that possibly could serve as a handle to housing associations when they are implementing their policies. The model of Latour (2004) and Callon (1986) is primarily a model which states that proposals, ideas, plans and the like, that arise every day should not to be accepted or rejected as soon as they pop up. Instead, they have to subject to a careful process before either being accepted or rejected. The word ‘due’ in the name of the model is to be understood as ‘carefully’. This meaning has been taken from the American legal system where the term due process is the guarantee of a careful process that protects the rights of suspects (Hyman, 2005).
Although other models could also serve as a handle to housing associations, there are tree arguments in favour of introducing the due process model in this paragraph. In the first place, following the due process model will increase the number of deviations that are made deliberately above deviations that are left unnoticed. This will prevent random deviation from well-considered policies. Secondly, the use of the due process model may contribute to the legitimacy of the housing association. This is important because the legitimacy of housing associations is under discussion. According to Brandsen and Helderman (2004), there is even talk of a “lingering legitimacy issue.” Through the due process model stakeholders will be able to have a say on (possible) deviation of the policies of housing associations. An important feature of the due process model is that not only stakeholders can have their say. Through contractors and controllers for example, engineering and financial aspects of proposals will also be present in the discussion. The third argument for the introduction of the due process model emerges from the interests of the housing association to smoothen the implementation process. Various researches show that companies are dependent from many external (f)actors in the implementation of their plans (O’Toole, 1986). Applying the due process model enables organizations to take better account of the wishes and interests of stakeholders involved in the measures that have to be implemented. This allows them to smoothen the implementation process.
Embedding is – in the words of Latour (2004) – about institutionalisation. One of the four general rules of Latour is about the institution of agreements. An agreement concerning how a project and the existing structures can ‘live together’ can be seen as a starting point. However, institution is about much more than only putting a signature under an agreement on paper. The building that is targeted at should also change in a physical sense. Only then a plan has become real. From the moment that a plan from the housing management strategy has been put into practice, it belongs to the existing structures. As a result, it is important at this last stage to close the discussion. The legitimacy of the plan should not be questioned anymore. Tasks that have been agreed on in terms of hierarchy should be carried out. As Latour (2004:109) puts it, agreements have to be institutionalized. In the form of a general rule: “Once the [measure from the housing management plan] have been instituted, you shall no longer question their legitimate presence at the heart of collective life.” For housing associations this means that after the definitive decisions with stakeholders about how to go on with a plan, this outcome should not be questioned over and over again. The implication of this is that al agreements that have been made during the phase of hierarchy have to be fulfilled.
Although embedding is also about closing the discussion about alternatives, deviation from policy is still possible. In order to prevent random deviations, deviations are to be due processed. At this point the other three general rules of Latour come into the foreground. Problematization is the key word of the first general rule of Latour. He formulates it as: “You shall not simplify the number of [options] to be taken into account in the discussion.” (Latour 2004:109). For Latour, this rule is about the need to give a new candidate for existence some space to introduce themselves. A new option on how to develop the housing stock should neither be implemented nor neglected too soon. Problematization is able to make sure that institution or exclusion will not happen too soon. The second general rule deals with consultation about the characteristics of the new option (Latour, 2004). The discussion is now no longer about the question whether the option (that now has evolved into a plan) at stake is legitimate at all, but how it can be filled in. All those who are involved in the implementation process should explicate their vision on the implementation. In the case of housing associations that are planning to build or renovate a building thus constructors, architects, local politicians, tenants and other stakeholders should be consulted. Consultation is merely about explicating the different viewpoints. Discussion about how different viewpoints can live together is not important yet. Latour (2004:109) formulates the second general rule as follows: “You shall make clear that the number of voices that participate in the articulation of [the option/plan at stake] is not arbitrarily short-circuited.” Hierarchy is the third general rule of the due process model. Hierarchy is about fitting the new option/plan into existing structures. Sometimes it is obligatory that the new option/plan can ‘live together’ with the existing structures. This is for example the case when we face legislation as part of the existing structures. Sometimes fitting in existing structures is not obligatory, but strongly preferred by one of the participants (for example the maintenance of an existing organization structure). By fitting in the option/plan its essence can change. However, at the same time, the existing structures also change. Hierarchy comes with a lot of discussion with the spokesmen of different parts of the existing structures. During the discussions it will be examined how the new plan fits in with the existing structures. Latour (2004:109) formulates this again in the form of a general rule: “You shall discuss the compatibility of new [options/plans] with [the existing structures], in such a way as to maintain them all in the same common world that will give them their legitimate place.” For the professionals at a housing association this means that they should be flexible enough to change (details of) a project in such a way that it fits in with the existing structures. At the same time it should be possible to change the existing structures in order to make a plan feasible. When the hierarchy has been set and decided on, the deviation from the original policy has been set. However, because it has still to be realized in practice. Therefore, after hierarchy one has to go back to embedding.
There are several measures taken by the investigated housing associations to ensure that policies are implemented. In this paragraph I will go into these measures. In the first place it is important that the policy is regarded as input for the annual budget. In addition, in some cases annual plans of activity are made. In practice it seems that such annual plans and budgets are used by employees, whereas policy documents are not used in daily practice. The policy should therefore be incorporated in the annual plans and budgets in order to be executed. From the cases there are two examples of this. At Bo-Ex and Vestia Rotterdam Noord the authors of the housing management plan are involved in the drafting of the annual budget, to ensure that measures from the housing management plan are translated into the budget. At Vestia Rotterdam Noord it has been standard for some time that the update of the housing management plan was linked to the drafting of the annual budget. This seems to be important because the annual budget can only be based on the housing management plan if it is
It is also important that the policy is taking into account in each stage of renovation and building projects. Some parts of the policy can also be processed in the databases that are used in daily processes. In this way, through the automation, at the right moment a signal is issued as parts of the policy apply. For housing associations this is useful, for example, for measures to be implemented just before a dwelling is going to be issued to a new tenant. At Bo-Ex an example of this can be found in the measure to change the floor plan of apartments in some buildings. This measure has to be executed every time a tenant terminates its contract. To remind the employees working on this a label is captured in the primary automation system, on the page where they find also other relevant information for this work process.
For the external embedding, the policy should be input for performance agreements, contracts and other agreements with third parties. This is not unidirectional. The agreements with third parties in turn also determine the scope of the policy of the housing association. In order to embed its focus on the elderly, Groenveld Wonen has set a cooperation agreement with a local care agency. By Bo-Ex and Leyakkers, performance agreements with the municipality were agreed to set the right circumstances for realising new dwellings in the future. Agreements were made about land prices and numbers of dwellings that the housing association can build every year.
In addition to formal arrangements with third parties as a stakeholder housing associations can have their say on the policies of third parties. Also through this approach, the cooperation may be easier. An example of this is that two housing associations from this study have worked together with the local council in preparing a local housing policy.
Furthermore, monitoring of progress is important. Monitoring can reveal potential deviations, and put them on the agenda. The best example of this is the use of a balanced scorecard (BSC) at Leyakkers. Since 2001, from the strategic housing management the aim is to let 40% of all dwellings allocated each year to elderly people (older than 60), and to also let 40% of the dwellings to people under 30 years. During the first years after this aim was set, it was only embedded in regular meetings with the municipalities about the new building of dwellings. However, only after this aim became part of the BSC in 2004 the results of the monitoring led to further embedding. From the BSC it became clear that both aims were not (always) realized. By labelling more existing dwellings especially for these target groups, Leyakkers was able to get closer to the 40% it set in 2001.
Regarding the staff, the right circumstances for employees to implement the policy have to be created. This can be done by giving sufficient time for their part of the implementation and by explicitly assessing their contribution to the policy implementation. Many large and small decisions that employees make every day are not captured within any of the above measures. It is thus important that the staff knows the main goals of the policy, so they are always able to incorporate the spirit of the policy into their actions. Internal marketing of the policy through presentations and popular versions of documents can make a contribution to this.
In recent literature on implementation there is more attention to the fact that implementation is an ongoing policy process, and therefore needs to be flexible (e.g. Boonstra 2004). Several authors make proposals for a flexible embedding of policies. This should ensure that the objectives are achieved as far as possible, albeit through flexible ways. The consequence is that the recommendations for embedding arising from the cases cannot be considered as an exhaustive list. In practice it may be that new and / or other means should be explored in order to embed the policy at stake.
Deviation from policy
At the housing associations in this study different examples can be found about how deviations from the original policy can be approached. From the cases it becomes clear that deviations are not always made conscious. The conceptual model presented in this paper should be able to help housing associations becoming more aware of the balance between embedding and deviation. Below I will present examples from the cases about problematization, consultation and hierarchy where elements of this model have been used.
Problematization of a new issue to be done in order to establish a possibility to deviate from a policy at all. At this time it is especially important not to come up with ready-made solution too soon. In practice it appears to be difficult not to proceed with the introduction of solutions right from the start. In the investigated cases examples of problematization are rare. The best example of problematization is the ‘yes, if’ adage of Leyakkers. When stakeholders need the assistance or approval from Leyakkers on a certain issue, Leyakkers always responds with ‘yes, if’ to their question. In addition, tenants have many choices regarding the maintenance of their homes. The ‘yes, if’ attitude, new questions and issues are never embraced or excluded directly, but there is room for discussion about the issue at stake. On smaller scale, there are also examples of factors that lead to problematization. At Groenveld Wonen, figures on the characteristics of the building and its popularity among people looking for a house became an instrument for problematization. The figures showed that the houses, originally designed for the elderly, had to let to young people increasingly because the apartments were not suitable for elderly anymore. When the corporation further analyzed this problem, it became clear that from demographic analysis could be inferred that the number of elderly would increase. Another example is a director who, albeit after a long discussion, asked aloud what exactly the real problem was in the discussion on daylighting. Asking aloud led to the problematization of the situation.
In the phase of consultation the only task is to gather as much different viewpoints towards the newly raised issue as possible. In practice we see that in this phase it is difficult to find out who and what will be affected by the new issue at stake. Often consultation is not done explicitly, but in a more implicit way. The architect makes its drawings of the floor plan. In that way he makes very explicit what he wants with the projects. In all cases the builders of course did financial prognoses very early in the process. In that way they made explicit how much they would like to get from the project in financial terms. However in other cases views were not made explicit at all. This caused trouble in later phases of the projects. This was for example the case at one project of Bo-Ex. Bo-Ex already decided in favour of demolition above renovation for a building. However, protests from the tenants followed. Only after going back into the consultation phase, and relooking at the renovation and demolition alternative, the tenants agreed to demolition. If the consultation had been done before taking decisions (the phase of hierarchy), the protests from the tenants could have been avoided.
All new issues are going through the phases of problematization and consultation. Only in the phase of hierarchy it is decided whether or not a new issue is going to be embedded or not. During this phase the housing association should try, together with others involved, to see how the statements from the phase of consultation can be brought together into one single solution. In the larger projects of the housing association involved in this study this is done by lots of talking between parties. The project champions at the housing associations talk about projects do different stakeholders, take their comments back to the architect, go to talk to the stakeholders again with an updated version etcetera. In the case one project of Groenveld Wonen it became clear that this can be a very time consuming process, especially when the project does not make any progress while at the same time regulations are changing. Although the aim of this phase is to bring different viewpoints from the phase of consultation together, hierarchy is also the phase were sometimes has to be decided to not embed a new issue. When a decision has been made, the focus should shift back to embedding this decision. When the new issue has to be adapted the techniques for embedding, which have been discussed elsewhere in this paper, come into the foreground again.
The examples from the cases show that in general there is a tendency to go to the phase of hierarchy too soon. Housing associations are good at making decisions. However, they are not always aware of new issues being a deviation from the original policy (perplexity). Consultation is not always done properly, which can cause problems later on in the process.
Conclusion: balancing between embedding and deliberate policy deviation
Policy is made to implement. The internal and external embedding of policy is necessary to meet this aim. Without embedding a policy no one will look after the policy anymore. Because after the adoption of a new policy new issues may emerge, and because unforeseen events may appear, it is also important to have the opportunity to occasionally deviate from the policy. By respondents in this study, deliberate policy deviation often is seen as positive. In the literature on policy implementation, however, deviation is usually seen as negative, and is focused on improving the embedding of policies (e.g. Pressman and Wildavsky, 1984). This is a limited approach. Housing associations implement not for implementation. Much more important is that (the outcome of) the implementation process is legitimized.
The cases show that housing associations legitimize their performance through the implementation of policies. Sometimes policies are embedded, and sometimes they deviate from their policies. Employees at the housing associations are not always aware of conflicts between embedding and deviation. However, arbitrary deviation from policy is to be avoided. The conceptual model presented in this paper can be used to deliberately determine whether or not the housing association should deviate from its policy in certain circumstances. The elements of the conceptual model presented in this paper have been put into practice already more or less by the housing associations in this study. Further research should reveal to what extend this is the case at a larger scale. Another point for further research is to make the model itself more concrete in order to make it more suitable for housing associations.
Boonstra, J.J. (2004) (Ed.) Dynamics of Organizational Change and Learning. Wiley Publishers, Chichester.
Brandsen, T. en J.K. Helderman (2004). Volkshuisvesting. In: H. Dijstelbloem, P.L. Meurs en E.K. Schrijvers (red.). Maatschappelijke dienstverlening. Een onderzoek naar vijf sectoren. WRR-Verkenning 6. Pp. 65-132, Amsterdam University Press, Amsterdam.
Callon, M. (1986) Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St Brieuc Bay. In J. Law (Ed.), Power, action and belief: A new sociology of knowledge. London: Routledge & Kegan Paul.
Dankert, R. (2008). Verbindingen in de implementatie black box. Uitvoering van strategisch voorraadbeleid en vastgoedsturing door woningcorporaties. Delft: OTB.
Day, George S. en Robin Wensley (1983). Marketing Theory with a Strategic Orientation. In: Journal of Marketing. Volume 47 (herfst 1983), pp. 79-89.
Ekkers, Paul (2002). Van volkshuisvesting naar woonbeleid. Reeks Planologie, deel 4. Sdu Uitgevers, Den Haag.
Eskinasi, Martijn (2008). Corporaties & vastgoedsturing. Tweede druk. Nestas communicatie, Almere.
Faludi, Andreas en Willem Korthals Altes (1994). Evaluating communicative planning: a revised design for performance research, European Planning Studies, 2, pp. 403– 418.
Flyvbjerg, Bent (1998). Rationality and Power: Democracy in Practice. University of Chicago Press, Chicago/London.
Heeger, Henk en Nico Nieboer (2003). Van bouwtechniek naar woontechniek en woonmilieu. Trends en ontwikkelingen in het voorraadbeleid. In: Aedes-Magazine. Nr 13/14, pp. 56-59.
Hyman, A. T. (2005) The little word ‘due’. In: Akron Law Review 38 (1): 1-51.
Latour, Bruno (2004). Politics of Nature. How to bring the sciences into democracy? Harvard University Press, London.
Maarse, J.A.M. (1998). De uitvoering van overheidsbeleid. In: Hoogerwerf, A. en M. Herweijer (eds.). Overheidsbeleid. Een inleiding in de beleidswetenschap. Samson, Alphen aan de Rijn.
Matland, R. E. (1995) Synthesizing the implementation literature: the ambiguity-conflict model of policy implementation. In Journal of public administration research and theory 5 (2): 145-174.
O’Toole, L. Jr. (1986) Policy recommendations for multi-actor implementation: an assessment of the field. In: Journal of public policy 6 (2): 181-210.
O’Toole, L. Jr. (2000) Research on Policy Implementation: Assessment and Prospects. In: Journal of Public Administration Research and Theory 10 (2): 263-288.
Pressman en Wildavsky (1984). Implementation. How great expectations in Washington are dashed in Oakland. University of California Press, Berkely/Los Angeles/London.
Terpstra, J. en T. Havinga (2001). Implementation Between Tradition and Management: Structuration and Styles of Implementation. In: Law & Policy 23 (1): 95-116.
Van den Broeke, Rob (1998). Strategisch voorraadbeleid van woningcorporaties: informatievoorziening en instrumenten. Delft University Press, Delft.
Van Os, Peter (2007). Mensen, stenen, geld. Het beleidsproces bij woningcorporaties. RIGO, Amsterdam.