Planet Homeless: Governance arrangements in Amsterdam, Copenhagen and Glasgow

Homelessness is a complex policy issue that all local governments face. But, at the same time, local authorities often have very little influence on the causes of homelessness, such as de-institutionalization, drug addiction, and release from detention or evictions. Seen in a European context, Northern European local governments all face similar policy challenges, such as addressing rough sleeping and promoting the flow of people out of temporary shelters.
From a public administration perspective, homelessness can be identified as both a ‘wicked’ and a persistent social and policy issue. Interventionist governments tend to implement policies that might end up creating new problems – hospitalization being an example of this– and at the moment there is little known about the effectiveness of public administration policies that aim to address homelessness. From this study I hope to gain more insight into exactly which elements of administrative and political approaches, or which governance arrangements, are most effective in this respect.
To this end, this research has examined developments in three northern European countries (Denmark, the United Kingdom and the Netherlands) over the past ten years.. Homelessness has made its way onto the political agenda because, in a time of relative prosperity, it is no longer considered an acceptable phenomenon and has therefore become an issue that has had to be addressed. It is true that, as various financial and moral strategies have been combined, possible new opportunities have arisen to tackle the problem of homelessness. In this research, I ask the question of whether differences in the administrative-political approach to homelessness between Northern European metropolises impact upon the quality of the level of the facilities offered, as well as the related social results. Does a difference within the governance arrangement that is applied in relation to homelessness lead to a difference in the quality of facilities for homeless people and a difference in the social results?

To answer this question, I first examined, through a literature study, exactly which elements of governance arrangements to combat homelessness currently exist and exactly which aspects of these elements must be studied in order to explain which outcomes (in terms of the quality of services offered) and which societal effects different arrangements may have.
There are three elements of a governance arrangement that are crucial to this: policy, structure and management. Within the ‘policy’ element, I distinguish three aspects: policy goals, policy instruments and the basic assumptions (moral and empirical) underpinning the policy. The ‘structure’ element concerns the level at which resources and responsibilities that offer possible solutions to homelessness are allocated. In addition, the network structure is an important aspect of the structure element. Lastly, under the ‘management’ element are grouped relations that local government within the network structure has with non-profit organisations and private organisations. Flexible or more distant relationship between politics and the civil service within local government itself are also subject to examination. In addition, the dominant role of local officials and the value at the local level attached to certain conceptions of accountability can also be distinguished as relevant aspects of management.
The effectiveness of a governance arrangement is assessed on the basis of the quality of the services offered and the related societal effects. The quality of the services offered can be measured by using a number of performance indicators for public mental health care (PMHC). These performance indicators measure the integrated nature of the services offered, the degree to which this also includes psychiatric services, and the supply of temporary or permanent housing. Efficiency is also one of the criteria on the basis of which the quality of the services offered is defined.
The societal effects of the governance arrangement are assessed on the basis of indicators related to the number of homeless people (who reside in the facilities provided) in the city and the number of people sleeping outside. In addition, a picture of the societal effects is obtained by recording the opinions and reactions of the broader public.

In the context of this study on the subject of homelessness, three hypotheses on the impact of a governance arrangement on the quality of the services offered and the societal results have been put forward. The first hypothesis proposes that specific variations in the way in which a governance arrangement is organised impact upon the efficacy of the arrangement in terms of integrated service coverage and customised services. The second hypothesis states that networks that are more heterogenic are more effective in the field of housing. Heterogenic networks are more capable of preventing homelessness than homogenic networks. The third hypothesis argues that in a more centralised structure there is a greater effectiveness in terms of efficiency.

These three hypotheses and the other findings of the literature study are outlined in the following theoretical model.

Figure 1 Theoretical relation between the governance arrangements applied at local level in respect of homelessness and their efficacy






















In order to empirically investigate the three hypotheses, using the results of the literature study, I sought as much variation in the political-administrative contexts of northern European cities as possible. I decided to study northern European cities because they are comparable in terms of prosperity and in the last decade they have demonstrated activity in the setting of policies to address homelessness. In Northern Europe, different administrative traditions can be distinguished. These can be summarized under the headings of Anglo-Saxon (more centralist structure and corporatist network), Scandinavian (more decentralized structure and pluralist network) and continental (a mixture of the administrative traditions of the first two).
On the basis of this trichotomy, Glasgow, Copenhagen and Amsterdam were chosen as cities where it would be possible to analyse the organisation and effectiveness of the governance arrangement. The local authorities of these cities were willing to cooperate with the investigation. In each city, the relevant policy documents were studied and confidential semi-structured interviews among ten relevant stakeholders (policy makers, politicians, executive institutions, clients) have been carried out. The results of the three case studies provide a well-documented picture of the different aspects of the theoretical framework outlined above..

In Copenhagen, the policy appears to be highly problematic. Efforts focus on the groups of homeless people with the most complex needs. However, the only instrumentation tools offered to these persons are temporary emergency solutions. Copenhagen's main instrument is Housing First. This appears to be only partially available for the toughest group of homeless people. The assumptions underlying the policy appear to exhibit a mentality of permissiveness. The homeless person is seen as a free bird, an independent individual, who is better off not being rushed.
The governance structure fits the metaphor of an archipelago, which is not addressed in the policy and impacts on the quality of the services offered and the societal results. There is only a partly decentralized structure. The structure is characterised by central control, achieved in part as a result of funding a network of longer-term social relief providers who offer their services in the city of Copenhagen. There is a close interrelationship between society and politics. The civil service is given the task here of responding flexibly to this close interrelationship. This may explain the incident-focused practice, as well as proposals made by politicians directly reflected in the policy.
This governance arrangement has questionable outputs: on the street, homeless persons with psychiatric problems (those who are the responsibility of the local region) cannot be helped because there is a clear lack of solutions to address the problems faced by these individuals. The societal effect of the governance arrangement is that both the number of homeless people who make use of shelter facilities and the number of homeless people who reside on the street have increased during the intensified homeless policy.

Glasgow’s policy focuses on the strength and self-responsibility of homeless persons themselves. Because of the significant social challenges that this town is facing and the central obligation to offer statutory homeless people a new permanent home, many efforts are being made to prevent homelessness. Persons who report themselves as homeless, and who it is felt could benefit from the supply of services other than just shelter, are offered alternative tools and the city also seeks other solutions, such as the repairing of family relationships through mediation. To prevent evictions, active negotiation takes place with landlords considering such action. Homelessness policy is part of the housing policy. The involvement of housing agencies in these instruments is significant.
Both a clear state central structure and a centralizing trend are visible here. At the time of the research (all of 2013 until the beginning of 2014), the possibility of more Scottish independence would have actually produced less autonomy for Scottish cities. Also, there is generally little public confidence in the solutions offered by government: to achieve solutions to societal issues politically elected people tend to take up higher public positions. In this context, political accountability is much valued and impacts on agreements with providers of care, as in the case of adverse incidents. Accountability is an important element in this management culture.
Glasgow has proved quite successful in achieving the policy goals that the city has set itself. As compared to the other two cases, the city reached the highest outflow of homeless people into permanent housing. Also, despite the varying support for the policy from local residents and in public opinion, the city managed to improve the social effects. The results of the local policy are closely monitored and subject to national benchmarking.

Finally, in Amsterdam, the policy objectives are aimed at prevention, rehabilitation and flow through and out of shelters. However, the instruments of the person-centred approach in this municipality appear to still be mainly focused on helping the people who are most in need or display troublesome behaviour. The recently extended capacity in sheltered housing has led to more solutions being offered within the social relief sector than outside it. From the underlying municipal policy assumptions it is clear that there is a lack of experience with new care and support concepts and that there remains a growing emphasis on security.
In Amsterdam, at the time of research (2013), decentralizing trends were playing a role. The policy network for social relief consists of a large number of, mostly similar, partners. Predominantly missing, however, are housing players. The closeness of the formal policy network seems to get in the way of a more integrated policy: the current supply is based on a historically grown question, one that is apparently unable to respond to current broader social issues. Amsterdam's administrative tradition is characterized by balancing interests, and by historically grown structures and relationships, and sees the flexible management style as relevant. This administrative tradition also explains the pace at which new paradigms and policies can be implemented. The civil service has a relatively large discretionary space. In terms of accountability, there are relatively few conditions placed on the non-profit part of the social relief sector. The instrumentation to do so is too much under the influence of politics and society, from which a resistance to change towards a more corporate relationship comes.
The governance arrangement appears to be effective where it concerns the quality of the services provided. The outputs of the policy appear to be effective in respect of the high-risk groups, which the current tools are still engaged with. PMHC criteria are decisive for the inclusion of high-risk groups in the Amsterdam chain. This explains why the homeless population within the provisions often has mental health issues as well as why mental health services are often offered to the homeless population. The central gateway to homeless shelters in Amsterdam appears able to distinguish prioritized care demands such as evidence of mental health and addiction issues among the homeless who report there. This is something that the municipality has also set as a target so, in this respect, this governance arrangement is effective. Finally, the continuing exclusivity of social-relief budgets in the broader decentralising trend raises questions regarding municipal efficiency.
The outcomes vary. The number of homeless people is decreasing. The proportion of homeless people with less severe care needs is, however, increasing. This trend may explain the observed increase in the number of rough sleepers in Amsterdam. A possible side effect of the method of monitoring and reporting in Amsterdam, where the focus is on safety and prioritizing the clients with the most severe health needs, is that individuals and groups are unnecessarily criminalized and medicalized.

The results of the three case studies can – in respect of the organisation of the governance arrangements – be summarised in a graphical manner as below. In these, a figure, an ordinal grouping ranging between zero and two, has been allocated per element to the most relevant aspect of the element. To the most positive organisation of the relevant aspect a maximum of two points has been awarded.

The results of this study found variations in governance arrangements in the cities selected in how they approached the problem of homelessness. Using these findings I was able to make connections between the similarities and differences in the three cases in respect to the quality of the services offered and the related societal effects. In this way I have empirically tested the three hypotheses from the theoretical framework. This empirical testing of the hypotheses results in the following picture of the impact of the organisation of a governance arrangement on the efficacy of the arrangement. The general conclusion on the basis of this is that governance really does matter. The first hypothesis from the theoretical framework is thus confirmed. What the authority does or does not initiate in conjunction with other parties to address homelessness does indeed have an impact. So, it appears that levels of integrated service delivery are dependent on the condition of whether or not the setting of internal policy goals is part of the governance arrangement. With the help of internal policy goals, multi-level fragmentation of responsibilities and budgets can be prevented. Also, the degree of multi-level fragmentation or, in other words, where (at what level) responsibilities and budgets are allocated, appears to account for the quality of this output. In addition, we may conclude that the successful outputs for the various cases in this respect are mostly attributable to successful coordination (cf. Bouckaert et al., 2010). In reaching this conclusion, I have also found that decentralizing trends do not impact positively on the realisation of higher levels of integration, such as more integrated approaches and customized services (cf. Fleurke en Hulst, 2006).
A second conclusion, which results from the testing of the second hypothesis, is that the efficacy of a governance arrangement – in terms of the improved housing situation of homeless persons – appears to depend on the network structure that is part of the governance arrangement. A more heterogenic constellation of the network, which means a more mixed composition of specialists and generalists in the network, does appear to have better outputs in terms of the housing situations of homeless persons.
Third, it appears from the testing of the third hypothesis that it is predominantly elements of structure and management, rather than policy, that are decisive for the efficiency of the arrangement. Again, multi-levelness is a decisive factor, but only because this has an impact on the room for variations in elements of management to account for efficiency of output.

The above conclusions offer explanations for the quality of the services provided: they concern the quality of the outputs. Meanwhile, the societal effects (the outcomes) of a governance arrangement can be explained less precisely. There is a wider range of explanations for the societal effects that are possible and these explanations are more centred on policy and management elements of governance arrangements than on structure. Beliefs and social traditions would have more influence on social results than means or structures.

The research has made clear that it does matter whether governments do or do not do something, and that the way in which efforts are constellated in each arrangement also matters. More insight has been obtained into exactly which aspects of administrative and political arrangements contribute to the effectiveness of the arrangement. In this way this study makes an important contribution to the academic debate on the optimum organisation of governance arrangements.
It does not offer support for an approach that is about networks as diffuse and liquid entities within an increasingly globalising context. Instead, what it does do is provide important information that will enable scientists to better understand relationships between elements of governance, output and outcome. This research also clarifies exactly which aspects are relevant in the study of these relationships. In addition, it provides a methodological tool to study a complex social phenomenon such as homelessness from a governance angle. Finally, the results of this research offer ideas for the development of policies in areas of socially complex issues, such as a strategy on homelessness.

This study is a plea for a corporate, instrumental approach towards governance arrangements on homelessness. Interventions to improve the functioning of the social relief sector should focus on elements of structure and on the governance levels at which the adjacent responsibilities for homelessness are organised – such as the responsibility for mental health policy, youth policy and learning disability policy. The main focus of homelessness policy and sheltered institutions should then be to locate trends and refer persons – or in other words ‘push them back’ – to the adjacent, primarily responsible areas. The sector of social relief should be functioning as a trampoline, not as a last resort.