Content uploaded by Ronni Michelle Greenwood
Author content
All content in this area was uploaded by Ronni Michelle Greenwood on Sep 14, 2018
Content may be subject to copyright.
33Articles
Building A Way Home: A Study of Fidelity
to the Housing First Model in Dublin,
Ireland
Rachel M. Manning, Ronni Michelle Greenwood
and Courtney Kirby
University of Limerick, Republic of Ireland
\ Abstract_In Ireland, the numbers of individuals recorded as ‘out of home’
increases annually. In 2011, the government committed to ending long-term
homelessness and the need to sleep rough. As part of this, Dublin City Council
implemented a Housing First Demonstration Project with the goal to house and
support 30 chronically homeless individuals. In 2015, a consortium of two
homeless service organisations expanded and restructured the programme.
At the time of data collection for this project, Dublin Housing First had 16
employees and served 76 homeless individuals. Results from a fidelity self-
assessment of the team are described in the present paper. Team members
and team leaders (n = 12) completed a fidelity self-assessment. Five weeks
later, a conciliation focus group met to discuss and agree self-assessment
scores. The programme demonstrated higher fidelity on Housing Process &
Structure, Separation of Housing & Services, and Service Philosophy domains,
and lower fidelity on Service Array and Team Structure domains. Five key
stakeholders took part in a second focus group to discuss facilitators of and
barriers to fidelity in each domain. Thematic analysis identified facilitators and
barriers to fidelity across systemic, organisational, and individual ecological
levels and yielded nuanced insights into the establishment of social innova-
tions such as Housing First.
\ Keywords_Housing First, programme fidelity, enablers, barriers
ISSN 2030-2762 / ISSN 2030-3106 online
34 European Journal of Homelessness _ Volume 12, No. 3, February 2018
Introduction
In Ireland, homeless individuals have “no accommodation that they, and the people
they normally live with or who they might reasonably be expected to live with, can
occupy” (The Housing Act, 1988). They may sleep rough, stay in homeless hostels,
B&Bs or hotels, with friends or family, or in a squat. The number of homeless adults
residing in homeless accommodation increased by over 80% between June 2014
and October 2016, an increase of nearly 100% in Dublin and 60% outside of Dublin.
Between December 2014 and August 2017, the number of rough sleepers in Dublin
increased from 127 to 184 (O’Sullivan, 2016; Factcheck, 2017; Peter McVerry Trust,
2017). These counts may underestimate homelessness because they exclude
asylum seekers and people living in domestic violence refuges, institutions like
prisons and hospitals, or inadequate circumstances like overcrowded flats. While
some people still attribute homelessness to individual problems such as addiction
or mental illness, most now agree that rising rents and a social housing system that
does not meet demand fuel the homeless crisis (Peter McVerry Trust, 2017).
In 2009, Dublin City Council sought to reverse the homeless trend by reconfiguring
services to end long-term homelessness and the need to sleep rough (Dublin City
Council, 2009). The Pathway to Home document described resources and strate-
gies that would support a ‘housing-led’ approach. As part of this approach, SLÍ,
which is an acronym for to ‘Support to Live Independently’, and also means “path”
in Irish, was established. SLÍ was a visiting service to help people with moderate
support needs “move out of homelessness by sustaining independent living and
reintegration in the community” (Dublin Regional Homeless Executive, 30 April
2012). The plan, however, did not offer services for adults with high support or more
complex needs (Homeless Agency, 2008). To fill this gap, Dublin’s leadership moved
to implement Housing First.
Housing First is guided by a philosophy of self-determination; that is, homeless
individuals are believed to be competent to make their own decisions, with
support if required. Housing First provides immediate, affordable, permanent,
scattered-site housing. There are no sobriety, psychiatric stability, or transitional
housing requirements. There is, however, a focus on harm reduction, assertive
engagement, and person-centered planning. Conditions are minimal and flexible
(e.g., meet a support worker; pay 30% of income toward rent). Housing First
teams emphasise choice and the pursuit of various recovery goals, should service
users choose, including mental and physical health, integration, employment,
education, and meaningful activities. Importantly, Housing First consistently
shows better outcomes than traditional services, particularly in relation to housing
35Articles
stability and better outcomes in some studies in terms of quality of life (Greenwood
et al., 2005; Greenwood et al., 2013; Stergiopoulos et al., 2015; Aubry et al., 2016;
Padgett et al., 2016).
A Housing First Demonstration Project launched in Dublin in April 2011. It drew
resources from existing homeless services and had an initial target to house 30
adults with significant histories of homelessness and complex support needs. The
Outreach Team identified individuals who were high risk because they slept rough
in cold winter weather conditions. An Approved Housing Body (AHB) supplied the
team leader and the first four apartments. Two part-time key workers, a psychiatric
nurse specializing in alcohol and drug abuse counselling, and an education and job
specialist were provided from other community services. As a key ingredient of
Housing First, a programme evaluation by an external team commenced at the
same time to assess programme fidelity and client outcomes (Greenwood, 2015).
In the evaluation, the Demonstration project evidenced higher fidelity in Housing
Choice and Structure, Separation of Housing and Services, and Service Philosophy
domains, and lower fidelity in the domains of Service Array and Team Structure.
As a “microsystem of recovery” (Manning and Greenwood, 2018), it is important to
understand the ways in which particular aspects of the ecology affect the imple-
mentation of Housing First programmes. Multiple aspects of ecology affect model
fidelity, an observation reported by Housing First evaluators in other contexts, too
(e.g., Nelson et al., 2017). Landlords’ willingness to offer accommodation, as well
as stakeholders’ appraisals of the team as responsive, proactive, and attuned to
their concerns, facilitated fidelity, while a significant housing shortage, caused by
the interrelated economic and mortgage crises, was a barrier. Individual barriers
included gatekeepers’ reluctance to let units to clients in locations with low neigh-
bourhood-person fit. Stakeholders’ preferences for staircase or continuum of care
services, and scepticism that Housing First could deliver the necessary supports
were also barriers to fidelity in the Demonstration project phase. The evaluation
also yielded a number of recommendations that led to a reconfiguration of the team
(Greenwood, 2015). In April 2014, a consortium of two organisations that provide
continuum of care services in the region was awarded a three-year contract to
deliver the service. In September 2014, the Demonstration Project became Dublin
Housing First (DHF).
DHF is the largest provider of Housing First in Ireland. It has a multidisciplinary team
of intake workers who engage with clients and follow them through to housing,
addiction workers, a counsellor, and a nurse. Clients are offered independent, long-
term, scatter-site homes procured from social and private rental markets. It takes
between two and four weeks to house a client, but the wait can be longer for people
with long histories of rough sleeping. It can also take longer to obtain social housing
36 European Journal of Homelessness _ Volume 12, No. 3, February 2018
than private rented housing. To be eligible, an individual must have a significant
history of rough sleeping or use of emergency services and complex support
needs. In the initial tender, DHF was to house 100 clients by the end of 2017. At the
time of writing this paper (August 2017), DHF had progressed significantly toward
this target by supporting 76 clients and employing 16 staff. Approximately 88% of
clients were housed in the programme for the past 12 months, while 89% were
housed at 1-year or 2-year follow-up. It is worth noting that in 2016, the Government
of Ireland launched an Action Plan entitled ‘Rebuilding Ireland’ wherein it promised
to increase Housing First tenancies from the original target of 100 to 300. This
significantly increased target was to still be achieved by 2017. During the research
period, informal conversations with Housing First staff suggested that meeting the
new target was challenging primarily due to its ‘unexpectedness’, which did not
afford consideration to the usual challenges of accessing housing, as well as the
time it takes to build relationships with homeless individuals.
The present study
DHF underwent significant expansion and reorganization since the Demonstration
project was replaced, particularly as most, but not all, of the team left the project.
In the present study, we assessed the extent to which the current DHF team
evidenced fidelity to the Housing First model and identified the facilitators and
barriers that affect model fidelity. Because we have the fidelity findings from the
original demonstration project, we also had the opportunity to compare the two
assessments and to look for similarities and differences in both periods. Thus, in
Autumn-Winter 2016, we conducted a fidelity self-assessment and focus groups
with current DHF team members and leadership.
Method
The fidelity assessment
Measure. The self-assessment is a programme-based, self-administered survey
used to evaluate fidelity to the Housing First model. Respondents rate programmes
on key domains (Housing Process and Structure, Housing and Services, Service
Philosophy, Service Array, and Team Structure). Items including: “What types of
psychiatric services, if any, are available to participants?”, “Which life areas does
the programme systematically address with specific interventions?”, and “What is
the programme’s approach to substance use among participants?” are coded on
a 4-point scale. Some items have only one answer, while others have several. A total
score is calculated as well as a score for each of the 5 domains.
37Articles
Procedures. A member of the research team (RMG) met with all interested DHF
team members who at that point had served on the team for six months or more.
RMG explained the self-assessment tool and the scoring procedure. She distrib-
uted information sheets, informed consent forms, and copies of the self-assess-
ment. Team members completed the self-assessment anonymously and individually
without discussion with their colleagues. The team leader collected and returned
the completed self-assessments to RMG, who then compiled the scores to identify
areas of convergence and divergence.
About five weeks later, RMG met with members of the team to conduct a concilia-
tion focus group. The meeting was attended by team members who did complete
the self-assessment and some who did not. The scores for each fidelity item were
presented to the team and they discussed each item for which there was not initially
consensus, until they agreed a final score. The meeting lasted about 2.5 hours and
was a lively discussion of the meaning and applicability of the fidelity items to the
Dublin context. RMG entered agreed scores into an Excel sheet that calculated
average scores for each domain.
At a third and final meeting, the domain scores were presented to a focus group of
five managers representing both organizations in the consortium. RMG used the
fidelity scores to guide a conversation about facilitators and barriers of fidelity. This
meeting lasted approximately two hours. Both meetings were digitally recorded
and a research assistant transcribed them verbatim.
Participants. Members of outreach, support services, and housing teams completed
the self-assessment tool (n = 12). Most, but not all, of those who completed the
self-assessment, plus other team members, participated in the conciliation meeting.
This included the Housing First Manager, Housing First Project Leader, members
of the outreach team, intake team, and housing and support services teams. The
focus group comprised of managers from each organization, including the Housing
First team leader (n = 5).
Data analysis
In this study, a programme developed by Roberto Bernad (and reported in Bernad et
al. 2018, this volume) was used to calculate the self-assessment score. We used
thematic analysis (Braun and Clarke, 2006) to code the two focus groups into relevant
and meaningful segments of information. Working from Nelson et al.’s categorization
scheme (2017), a postdoctoral researcher (RMM) and an undergraduate research
assistant (CK) identified factors that either facilitated or impeded fidelity (See Table
2). Within these two categories, subordinate systemic, organizational, or individual
level (with possibility of overlap between categories retained) were identified. The
coders also took an inductive approach to the transcripts to identify additional factors
38 European Journal of Homelessness _ Volume 12, No. 3, February 2018
that seemed relevant, which resulted in a third code, “methodological concerns”.
RMM collated the independent coding and discussed the codes with RMG and CK
until they reached 100% agreement. All discrepancies resulted from one coder iden-
tifying a text chunk missed by the other, rather than from disagreement.
Results
Fidelity assessment
Table 1 presents the standard scores for each item, the average domain scores,
and the total score. Previous research has set an overall total of 3.5 or higher as the
“benchmark” for high fidelity (Macnaughton et al., 2015). It was agreed by
researchers participating in the international Housing First project that a score of
3.0 or less reflected low fidelity. In Ireland, the total programme fidelity score was
3.4, indicating that overall the programme had close to high fidelity (i.e., 3.5 or
higher, Macnaughton et al., 2015).
Table 1. Fidelity Assessment Item Scores and Domain Means
Domain / Item Domain Mean / Standard
Item Score (Out of 4)
Housing Process and Structure
1. Choice of housing
2. Choice of neighbourhood
3. Assistance with furniture
4. Affordable housing with subsidies
5. Proportion of income required for rent
3
3.0
3.0
4.0
3.0
4.0
6. Time from enrollment to housing 1.0
7. Types of housing 3.0
Separation of Housing and Services 4
8. Proportion of clients with shared bedrooms 4.0
9. Requirements to gain access to housing 4.0
10. Requirements to stay in housing 4.0
11a. Lease or occupancy agreement 4.0
11b. Provisions in the lease or agreement 4.0
12. Effect of losing housing on client housing support 4.0
13. Effect of losing housing on other client services 4.0
Service Philosophy 3.6
14. Choice of services 4.0
15. Requirements for serious mental illness treatment 4.0
16. Requirements for substance use treatment 4.0
17. Approach to client substance use 4.0
18. Promoting adherence to treatment plans 3.5
19. Elements of treatment plan and follow-up 2.0
20. Life areas addressed with programme interventions 4.0
39Articles
Service Array 3.5
21. Maintaining housing 4.0
22. Psychiatric services 4.0
23. Substance use treatment 2.4
24. Paid employment opportunities 4.0
25. Education services 4.0
26. Volunteer opportunities 4.0
27. Physical health treatment 4.8
28. Paid peer specialist on staff 1.0
29a. Social integration services 3.2
Programme Structure 3
31. Client background 2.7
33. Staff-to-client ratio 4.0
34b. Frequency of face-to-face contacts per month 4.0
35. Frequency of staff meetings to review services 4.0
36. Team meeting components 2.7
37. Opportunity for client input about the programme 0.7
Total 3.4
Figure 1 shows average ratings per fidelity domain according to the standardized
four-point scale. The highest score was in the Separation of Housing and Services
domain (4 out of 4), which includes access to housing, rights, and responses to
tenancy loss. Scores were also high in the Service Philosophy domain (3.6 out of
4), which includes choice and client-led practice, and in the Service Array domain
(3.5 out of 4), which includes housing and support. Scores were lower in the
Housing Process and Structure domain (3 out of 4), which includes type of housing,
rent subsidies, and neighbourhood. Scores were also low in the Programme
Structure domain (3 out of 4), which includes the programme’s target population,
client contact and meetings, and opportunities to give feedback or to hold paid
positions or seats on governing bodies.
Figure 1. Average scores for five fidelity domains
Extent of Fidelity to Housing First Model
Average Domain Rating on 4 Point Scale
Team Structure /
Human Resources
Service Array Service Philosophy
Housing Process
and Structure
Housing
and Services
3.0
4.0
3.0
3.5 3.6
40 European Journal of Homelessness _ Volume 12, No. 3, February 2018
Facilitators of Housing First fidelity.
The main facilitators of fidelity to the model are summarized in Table 2.
Table 2. Summary of Facilitators for Achieving Housing First Fidelity
Systemic Organizational Individual
• Mortgage crisis & economic
downturn access some
cheaper houses;
• Commitment to the
philosophy, incl. client-
centred, recovery-oriented
care;
• Work to build landlord
relationships;
• “accommodation finder”;
• Relationships with
community services;
• Pilot/ Demonstration project
successes
• Sense of reward/witnessing
success
Few systemic-level facilitators of model fidelity were identified in either focus group
transcript. Surprisingly, the mortgage crisis was the one systemic facilitator of fidelity
that participants mentioned, and that was in relation to the housing process and
structure domain. One team member described how they were able to “seize thirty
houses… [that offer] pretty secure tenancy…”. Key Stakeholder 3 noted that they “…
wouldn’t have gotten as many [houses] if the market had been much freer and easier”.
As a consequence, the team was able to quickly house many new clients.
Most of the facilitators of fidelity identified were at the organizational level.
Commitment to the Housing First philosophy of client-centred, recovery-oriented
care was discussed extensively. In relation to the housing process and structure
domain, Team Member 1 described commitment to the model in this way: “Housing
First, [it’s] core… everything circles around that, being and doing Dublin Housing
First as… an ethos [not just] a name of a service.” Team members also described
how they worked to “meet clients where they are at” and to encourage them to
develop adaptive, self-regulatory behaviours and build relationships with people in
their neighbourhoods.
When discussing factors that affected fidelity in the Separation of Housing and
Services domain, Key Stakeholder 3 noted how team members were “very clear
that… sobriety, mental health… were [not] a condition in respect of accommoda-
tion”. Emphasizing this point, Team Member 4 talked about their clients who are
unable, or unwilling, to engage in face-to-face meetings. Instead of punishing these
clients by taking them “out [of housing]”, the team used alternative or more creative
strategies, including “phone calls to link in some shape or form”.
41Articles
The importance of focusing on client-led goals and client-led responses was also
mentioned in discussions about the Service Philosophy domain. Noting the chal-
lenges involved in providing a client-led service, Key Stakeholder 4 described
how clients “find it hard to think beyond getting a home” and work on any other
goals before they are housed. However, once they are “in housing,” clients are
“much more able to communicate [about] their [other] goals”. This stakeholder
illustrated the kinds of client-led interventions the team employed with a story
about a client who was “banging the tables and… making noise”, which caused
his neighbours to complain. The team increased their visits to this client to “twice
a day…”, they “put carpet down… got slippers”, and linked him to “different coun-
sellors”. He added that in this case, the team did not “go in and… say stop
banging the tables and stop making noise”. Instead, they worked with the client
to help him understand that “if you do that [the neighbour will get annoyed] so
maybe it’s better to be… on carpet or… put some slippers on.” The team member
concluded the story by saying, “it’s not about stop doing what you’re doing, but
it’s about how to manage it.”
The different ways that the programme works to build relationships with landlords
were also discussed. The “accommodation finder”, whose responsibility is to
obtain housing by building positive relationships with landlords, was mentioned in
regard to Separation of Housing and Services and Housing Process & Structure
domains. Key Stakeholder 4 described the accommodation finder as someone who
“knows the language… knows what to say [to landlords], things like ‘your house
won’t be destroyed’”. Key Stakeholder 2 added that landlords know that clients are
not required “to be… housing ready or any of that”, but that the team is “looking for
homes for people who are on the streets [and that they] have a support team”. He
felt that “honesty… gives credibility… and [helps]… within the local authorities and
the private rented [landlords]”. Key Stakeholder 1 noted that this work fostered a
reputation that Housing First will “manage the apartment for you” if you are reluctant
“to get into the property business.”
Building positive relationships with landlords is a responsibility taken up by all
members of the team, though, not just the accommodation finder. In discussion of
factors that facilitated fidelity in the Separation of Housing and Services Domain,
for example, Key Stakeholder 1 described how the programme strengthens rela-
tionships with housing sources through activities such as end-of-year “welcome
mornings”. Activities to build relationships with landlords were also viewed as
important to fidelity in the Housing Process and Structure domain. For example,
Key Stakeholder 3 described how they do “a ‘roadshow’, trying to get people to
understand the level of support that is with this programme, so that people
[landlords] can take a risk or a chance on housing someone”.
42 European Journal of Homelessness _ Volume 12, No. 3, February 2018
Positive relationships with community-based services were also identified as
important facilitators of fidelity to the model in the Service Array domain. Key
Stakeholder 1 described their existing connections in the community as “well-
matured”, which provided sources of education, volunteering, and social integra-
tion services for their clients. Some team members noted that these brokered
services are not accessible to all their clients, but that those who do engage with
them are afforded opportunities for increased community integration.
Key Stakeholders also noted the groundwork done by the Demonstration team was
an important facilitator of fidelity in the Housing Process and Structure domain. The
achievements won by the Demonstration team meant the current team “were
working off the back of an awful lot of work and engaging [with community
partners].” Key Stakeholder 4 described how the Demonstration team’s achieve-
ments were “successes that we can evidence” and that they were “pushing an open
door in as much as you possibly can.” The team’s success sustaining and building
relationships, as well as their clients’ positive outcomes, were identified as facili-
tating fidelity in the Team Processes and Structure domain. For example, processes
and procedures help new team members learn how to deliver services to clients in
accordance with Housing First philosophy and principles. Key Stakeholder 1
described their “buddy system” in which new members are paired with experienced
team members for the first three months on the job. Through this kind of shadowing
and learning-by-doing, the team builds cohesion and mobilizes commitment to the
Housing First model.
Although focus group participants did not mention many individual-level facilitators
of recovery, they did briefly note the sense of reward they get from working in a
Housing First team, compared to working in more traditional services. Despite the
challenges they face, the team does witness successes among their clients, which
sustains their engagement, motivation, and creativity, even in difficult times. As Key
Stakeholder 5 noted:
“the outcomes are so positive… there’s nowhere near the same reward for staff
having worked with them handing them their key and then seeing them in their
own place and whether they are doing well or not, they are there in a stable
surrounding and they have what they want.”
43Articles
Barriers to Housing First fidelity
Barriers to achieving Housing First fidelity are summarized in Table 3.
Table 3. Summary of Barriers to Achieving Housing First Fidelity
Systemic Organizational Individual
• Economic down turn,
mortgage crises, increased
rental prices;
• Conflicting client-led practice
& duty of care;
• Relatively young
organisation;
• Clients’ stages of change
In the previous section, we noted that some focus group participants identified the
economic downturn as a systemic-level facilitator of fidelity because it opened the
team’s access to some housing units. At the same time, as Ireland emerged from
the economic and mortgage crises, the housing market tightened and rental prices
spiked again, closing access to other sources of housing for new clients. Quite
simply, demand for housing outstripped supply. As Key Stakeholder 5 succinctly
said, “the number of rough sleepers is increasing and the availability of housing is
decreasing”. In discussion about the “housing process and structure” domain,
Team Members 2, 3, and 4 explained how the chronic housing shortage limits
clients’ choices in housing:
“they don’t really have a choice… we haven’t got the option to give people two
or three choices… if they say no, when is the next one to come up? They have
a choice to turn it down but the alternative [e.g., rough sleeping; emergency
accommodation] is usually enough to make them take it…”
The economic downturn was also associated with reduced tenancy security. For
example, Key Stakeholder 4 said:
“up until two years ago the security there was just not there for anybody -- not
just anybody who had problems, [but] for anybody -- because within a year…
rent would go up so there’s no security at all. So, anybody living in private rented
in Ireland never felt secure or unless they actually had a nice wad of money to
suppor t.”
Respondents identified the tight housing market as a barrier to fidelity in the
Service Philosophy domain, because it negatively affected the team’s ability to
re-house clients after housing loss. For example, Key Stakeholder 1 described
how the “really bad” housing market made it difficult to move clients when a
housing situation became unstable. Further, the tight market put the team in a
position where they felt they needed to encourage clients to take the first available
unit and prevented clients from being able to say “that’s not the right house for
me”. As a result, clients often had to choose flats or neighbourhoods that lacked
44 European Journal of Homelessness _ Volume 12, No. 3, February 2018
the characteristics clients preferred. The team worked to overcome the chal-
lenges caused by the housing market by trying to convince private “landlord[s] to
lease the apartment to [them so they could] convert [the lease to a] social housing
lease [that gives] the choice of location and the quality [with] social housing
security” (Key Stakeholder 1).
To a lesser extent, participants noted some organisational-level barriers to fidelity
in Service Philosophy domain, particularly conflicts between client-led practice and
their duty of care. Participants described how clients’ behaviours that threaten
tenancies can set “alarm bells… ringing [because it might be] due to… poor mental
[health, which is] under the contract of a duty care [because clients could be]
harming themselves or others” (Key Stakeholder 3, Key Stakeholder 4).
Participants also identified organizational barriers to fidelity in the Service Array
domain. This domain includes clients’ opportunities for meaningful participation in
the programme, perhaps by means of employment as paid peer specialists. Focus
group participants felt that that the programme scored lower on items in this such
as, Does the programme have a paid peer specialist on staff who provides services
directly to participants? because the organisation was still young. Key Stakeholder
1 said that these programme elements are only supposed to “kick in… around
now”, suggesting that, if the assessment been completed at a later time, then the
programme might have scored higher in this domain. According to participants, the
programme does not offer many avenues to service users’ input into the programme
operations and policy. When asked to comment on the barriers to fidelity on this
item, Key Stakeholder 3 responded that “we are too young as a partnership or as
a project to have that fully implemented, and I think the longer we go on, the more
you get aspects that [service user input] creeping in to the programme”.
Key Stakeholder 1 also added that there are also individual-level barriers to service
user input and suggested that “the nature of the customer group” means that they
need intensive case management and are not ready to participate in programme
operations. Key Stakeholder 2 suggested that the programme leadership believed
that, at this stage, service users’ integration into their communities was more
important than their input into the programme.
Participants noted some individual-level barriers to fidelity in the Service Array
domain, particularly the availability of education, volunteering, and social integra-
tion supports. They noted that a number of their clients are not yet ready to engage
in these areas, suggesting that clients’ early stages of change (Prochaska et al.,
1994) explains the programme’s lower fidelity in this domain. Acknowledging the
longitudinal and often non-linear nature of client recovery, Team Member 4
described how “it takes somebody that has been living on the street for twenty
years… more than six months before they decide, ‘actually I want to be a doctor’”.
45Articles
Although their clients might not necessarily be prepared to work toward those kinds
of goals just yet, Key Stakeholder 3 was optimistic, suggesting that “over the next
year, two years, three years [they] would have stuff like that coming in”.
Methodological Concerns
Some methodological challenges arose during data collection. For example, for
Item 7, a programme evidences ‘high fidelity’ if 60% or more clients are in “inde-
pendent apartments rented from community landlords with outside support”. Our
participants found it difficult to answer this question because, in the Dublin context,
private rented apartments are not always the best option:
… [with] a private rented apartment [the] big worry is this ‘is a home for life?’ If
we get through the… first 6-months… you get some security of tenure for the
next three and a half years… 99% of the time… if we had an independent
apartment rented from community landlords and a social housing apartment
rented from local authority housing body… the citizen would choose the social
housing,” (Key Stakeholder 1).
Individually, team members’ responses to Item 18, which measures the extent to
which the programme uses coercion to encourage treatment compliance,
suggested that clients were required to attend daily meetings with the team. In the
conciliation meeting, however, Key Stakeholder 3 explained that meetings provide
“more intensive support” when a client’s “mental health is maybe deteriorating or…
an addiction [is] really escalating. Meetings were to encourage clients to “re-engage”
or for the team to gain insight into what might be going wrong. Key stakeholder 1
further explained, “there isn’t a real consequence” when clients choose not to meet
with the team. Therefore, it seems that these meetings are a form of assertive
engagement and not punishment or coercion. The team also struggled with the
wording ‘systemic interventions’ in Item 20, suggesting that this does not reflect
their work because client-led care cannot be done according to a fixed plan or
system. Finally, for Item 29, the team were unsure if the social skills and training
they provided informally on a day-to-day basis counted as ‘social integration’
services.
46 European Journal of Homelessness _ Volume 12, No. 3, February 2018
Discussion
Overall, the team evidenced a moderate to high degree of fidelity in each of the five
domains, with scores ranging from 70% to 100% (i.e., 3 to 4 on the 4-point scale).
Results indicate that despite being a relatively young programme working in a
challenging housing market, the programme embodies the key principles and
practices that define the Pathways Housing First model. In order to achieve this
result, the programme benefitted from a number of facilitators and overcame a
number of barriers across different ecological levels.
Commitment to core aspects of the Housing First philosophy, as well as the team’s
relationships with landlords, were identified as important organisational facilitators
of fidelity, while the relative newness of the team and conflicting duty of care were
identified as organizational barriers. The main systemic barrier identified by our
participants was access to housing, which limited client choice and security of
tenancy. Individual-level facilitators or barriers were rarely mentioned, although the
limiting influence of clients’ stages of change on Service Array was noted. Findings
indicate that most of the items in the self-assessment tool (Gilmer et al., 2013) were
applicable in the Dublin context, but that the North America-centric terms used in
some items were ambiguous and were the source of some disagreement among
the team about the extent of programme fidelity on some facets of some domains.
By identifying facilitators and barriers to fidelity across ecological levels, our
findings replicate and extend knowledge about fidelity to the Housing First model.
Housing First can also be conceptualized as an innovative health and social care
model, and so, more generally, our findings contribute to understanding of fidelity
in these contexts, too (Greenwood, 2015; Nelson et al., 2017). At the organisational
level, we found that commitment to the model was particularly important, which
reflects Nelson and colleagues’ (2017) findings, who noted the importance of
‘organisational champions’ who enhance organisational learning, performance,
and transformation (Ferlie and Shortell, 2001). Together, team members are a
stronger lever for change than one individual, and so programmes should strive to
maximise team commitment (Maton, 2008).
Our work also builds on previous findings from the area of homeless service delivery
that highlight the importance of sharing evidence of programme effectiveness with
community partners such as landlords and other services on an ongoing basis
(Steadman et al., 2002; Nelson et al., 2017). We know innovation involves stages of
sharing knowledge and evidence, persuasion, decision, implementation, and
confirmation (Rogers, 2003). That is, innovators learn about a new programme, are
persuaded by evidence of effectiveness, and then decide to implement it and see
if it works in their local contexts. In Dublin, the stages of innovation were not linear;
instead, the team delivered their programme, shared their knowledge of success
47Articles
as evidence (e.g., “the level of support that is with this programme”) and persuaded
landlords to buy-in, all in a cyclical, iterative process (Swan and Newell, 2000). This
process of evidence sharing also contributed to a positive reputation in the
community. As such, evidence sharing is an important component of broader
innovation processes.
Interestingly, individual-level factors did not feature strongly in our findings. This
finding differs from those reported by Nelson and colleagues (2017), who noted that
staff turnover and change are important influences on fidelity. A number of indi-
vidual-level barriers to fidelity were, however, identified at the phase of the
Demonstration project when staff turnover was high and disrupted team func-
tioning. In fact, in the early days, staffing was such a significant barrier to fidelity
that substantial reconfiguration of the team and its management was recommended
(Greenwood, 2015). It would be wrong, then, to say that Housing First in the Dublin
context has been immune to individual-level barriers to fidelity. Rather, the current
programme is delivered with a clear structure, strong leadership, and effective
management and oversight, so staffing barriers did not feature in this particular
assessment. The finding that barriers and facilitators change over time makes
sense, given that Housing First is a dynamic and adaptive programme model.
Overall, assessments at regular intervals might be useful to monitor and understand
current facilitators and barriers to fidelity, and also to assess effectiveness of activi-
ties intended to increase or maintain fidelity, and to guard against threats to fidelity
such as model drift or dilution.
Comparing the present findings to the Demonstration evaluation (Greenwood, 2015)
offers further insight into the relationships of systemic, organizational, and indi-
vidual facilitators and barriers to fidelity. As in the Demonstration, higher fidelity
was shown in Housing Choice and Structure, Separation of Housing and Services,
and Service Philosophy domains in the present study. Areas of lower fidelity were
found in the Service Array and Programme Structure domains. Also, as in the
Demonstration, participants in the present study repeatedly mentioned the impor-
tance of landlords. However, the Demonstration team identified their pro-active
responsiveness to landlords as a facilitator of fidelity, while the current team
emphasized the importance of maintaining relationships with landlords. This
suggests that, over time, Housing First has established a positive reputation with
landlords which must now be maintained and capitalised on.
Some barriers to fidelity that were observed in the earlier Demonstration programme
evaluation were also observed in the current evaluation, especially the systemic
problem of housing shortages and low neighbourhood-participant fit. Other barriers
that were observed in the Demonstration evaluation included key stakeholders’
preferences for traditional homeless services over Housing First and scepticism
48 European Journal of Homelessness _ Volume 12, No. 3, February 2018
about the Housing First model (Greenwood, 2015), were not found in the present
study. The fading influence of these factors on fidelity indicates that key stake-
holders in Dublin have become more convinced about the efficacy of Housing First.
However, since this study did not include interviews with private landlords, repre-
sentatives of approved housing bodies, or local authorities, we cannot draw any
firm conclusion about changes in attitudes over time.
Practical contributions
Most decisions to implement Housing First are “top-down” policy decisions, and
are sometimes met with resistance from those who feel the model is being imposed
on them. However, bottom-up, employee-led participation is important to imple-
mentation success and sustainment, because commitment to the model and its
philosophy is embedded in and enacted through providers’ daily practices (Ferlie
and Shortall, 2001). The Dublin Housing First team’s practices, such as their “buddy
system,” serve to empower staff to participate fully in a programme that has
certainly influenced the wider community. These findings reflect the broader litera-
ture on mentoring, which is described as a means to share power and develop
leaders (Maton and Salem, 1995; Maton, 2008). Mentoring creates empowering
settings that motivate team members to participate in actions and decisions
(Peterson and Zimmerman, 2004). Empowering settings, in turn, can exert influence
over the wider community and society. As such, shared leadership and mentoring,
as a route to empowering and empowered settings, should be an important consid-
eration for the development and running of Housing First programmes.
Like Nelson and colleagues (2017), who emphasized integrated knowledge transla-
tion strategies, we note the importance of evidence-sharing in our study. It is widely
acknowledged that evidence for promising innovations, such as Housing First, is
often difficult for practitioners and providers to access because it is published in
specialized outlets, written in technical language, or without the level of detail
necessary for implementation in practice. These issues make effective translation
of findings critical to a programme’s success. In Dublin, evidence of Housing First’s
efficacy was transmitted via word-of-mouth, the media, and the team’s accom-
modation finder. Information about how Dublin Housing First manages apartments
was crucial for landlord buy-in, even more so than evidence of clients’ recovery
outcomes or public savings. These are just some real-world examples of how
evidence can be synthesized and attuned to the priorities and concerns of potential
community partners for effective programme dissemination and implementation.
49Articles
Further considerations and future research
We believe our findings offer important insights into the facilitators and barriers to
Housing First fidelity. However, readers should keep certain aspects of the methods
and procedures in mind when drawing inferences from our findings. For example,
the order in which domains were presented for discussion may have resulted in
different emphases on the various aspects of the context of implementation
(Shaughnessy et al., 2002). Participants did not, for example, mention access to
housing when talking about separation of housing and services, although inde-
pendent, scattered-site housing can be presumed essential for fidelity in this
domain. Because participants had already discussed housing at length in relation
to the housing process and structure domain, which was discussed first, they may
have felt they had already exhausted this topic. Researchers should be aware that
the order of topics may influence the extent to which participants emphasize or
discuss information that is relevant to a range of topics over a long interview or
focus group.
Participants’ familiarity with some topics may have led them to emphasise on
certain facilitators or barriers over others. For example, participants spent much
more time discussing organizational-level factors than systemic-level or individual-
level factors. It may be that organizational factors were the most important facilita-
tors in Dublin, or it may be that these factors were simply most salient to our
participants. Although they received little attention from our participants, systemic
and individual facilitators are often critical to programme fidelity. In the case of
Housing First, political will to solve homelessness (Nugent and Rhinard, 2015) and
public willingness to help (Toro and McDonell, 1992; Agans et al., 2011) are critical.
We may have obtained different findings had we used a differently structured
interview that probed deeper into systemic and individual factors, with different
sets of stakeholders (e.g., service users, landlords), in different stages of programme
development, or in a different context. Comparison of Dublin’s findings with other
international programmes and with programmes at different developmental stages
will shed additional light on this topic.
Finally, ecological forces may exert bi-directional influences on fidelity. Kidd and
colleagues (2007), for example, showed that vicarious exposure to homeless
youths’ trauma led to burnout among service providers. Our findings also indicate
that there are links between facilitators and barriers across ecological levels. For
example, commitment to client-led care, as an organisational facilitator, is likely to
reflect positively on the team’s reputation and, in turn, enhance relationships with
landlords, both of which are systemic facilitators. Future research will need to
confirm this conclusion, but we believe that our findings indicate that, rather than
being mutually exclusive, facilitators and barriers of fidelity influence each other
across multiple ecological levels in iterative, cyclical, and non-linear ways. In future,
50 European Journal of Homelessness _ Volume 12, No. 3, February 2018
researchers might examine these relationships more closely. We recommend that
programme leaders and team members also consider the ways in which the actions
and choices they take to affect fidelity at one level may have either positive or
negative consequences for fidelity at another level.
Cross-national implications & generalizability
We believe our findings are applicable and relevant to Housing First stakeholders
across different contexts. First, our findings, alongside the evidence for the spread
of Housing First internationally, show that successful innovation in homeless
service delivery is possible when policymakers and programme leaders consider,
develop, and impl ement plans for long-te rm positive change. Key to the programme’s
success were their responsiveness to early staffing challenges and their sustained
efforts to build community relationships. An unstable team and doubt about change
to the status quo are often features of any change processes. Thus, our findings
about the importance of building a cohesive team that is committed to the model
philosophy, as well as establishing positive relationships with community partners,
particularly landlords, are most likely to be critical to success in any context.
The prevailing challenge for DHF nowadays is structural in nature, namely a lack of
affordable housing. Again, homeless services across contexts are all working to
manage similar challenges. Thus, our findings highlight the importance of antici-
pating and planning for challenges related to housing shortages. Moreover, our
findings should be taken as direction to activism in the relevant arenas, particularly
toward lobbying for the provision of adequate and affordable housing. Overall, the
fidelity research presented here, particularly the nuanced insights into barriers and
challenges, is a crucial precursor to effectively disseminating the Housing First
model and establishing a strong evidence base in the European context (Greenwood
et al., 2013).
51Articles
Summary and Conclusions
In sum, our findings suggest that higher fidelity to key aspects of the Housing First
model in Dublin was facilitated by commitment to core aspects of the philosophy,
as well as the team’s positive relationships with landlords. Our findings also provide
practical examples of how these facilitators can be embedded in organisations
(e.g., coffee mornings with landlords, buddy systems for new staff). At the same
time, fidelity was challenged by housing shortages, an issue that is not unique to
Ireland. Shortages in affordable housing make it difficult to find homes for new or
existing clients. Programme implementers should not assume “if we build it
[Housing First], they will come [housing units]”. Securing pathways to housing
should be an important preparatory step in the implementation of any new Housing
First programme. Overall, we hope that by identifying facilitators and barriers to
Housing First fidelity, the current study findings, combined with those from others
in this special issue, will provide direction and inspiration for innovators in homeless
and other human service contexts.
52 European Journal of Homelessness _ Volume 12, No. 3, February 2018
\ References
Agans, R. P., Liu, G., Jones, M., Verjan, C., Silverbush, M., and Kalsbeek, W. D.
(2 011) Public Attitudes Toward the Homeless. Paper presented at the 66th Annual
Conference of the American Association for Public Opinion Research, Phoenix,
Arizona.
Aubry, T., Goering, P., Veldhuizen, S., Adair, C. E., Bourque, J., Distasio, J.,
Latimer, E., Stergiopoulos, V., Somers, J., Streiner, D. and Tsemberis, S. (2016) A
Multiple-city RCT of Housing First with Assertive Community Treatment for
Homeless Canadians with Serious Mental Illness, Psychiatric Services 67(3)
pp.275-281.
Braun, V. and Clarke, V. (2006) Using Thematic Analysis in Psychology, Qualitative
research in Psychology 3(2) pp.77-101.
Dublin City Council (2009) Pathway to Home (Dublin: Homeless Agency).
Dublin Regional Homeless Executive (2012) Support to Live Independently (SLI)
Review, Retreived 29 July 2017 from http://thehomelessagency.newsweaver.co.
uk/newsletter/1l6v1yv3ch8? opc=false&s=
Factcheck (November 15th, 2017) Factcheck: Does Ireland Really Have a Low
Rate of Homelessness by International Standards? Accessed from: http://www.
thejournal.ie/fact-check-homeless-2-3693945-Nov2017/
Ferlie, E. B. and Shortell, S. M. (2001) Improving the Quality of Health Care in the
United Kingdom and the United States: a Framework for Change, The Milbank
Quarterly 79(2) pp.281-315.
Gilmer, T. P., Stefancic, A., Sklar, M. and Tsemberis, S. (2013) Development and
Validation of a Housing First Fidelity Survey, Psychiatric Services 64(9)
pp.911-914.
Greenwood, R.M. (2015) Evaluation of Dublin Housing First Demonstration
Project: Summary of Findings. Retrieved 28 July 2017 from http://www.
homelessdublin.ie/sites/default/files/publications/HOUSING FIRSTirst_
Evaluation2015.pdf
Greenwood, R. M., Schaefer-McDaniel, N. J., Winkel, G. and Tsemberis, S. J.
(2005) Decreasing Psychiatric Symptoms by Increasing Choice in Services for
Adults with Histories of Homelessness, American Journal of Community
Psychology 36(3-4) pp.223-238.
53Articles
Greenwood, R. M., Stefancic, A., Tsemberis, S. and Busch-Geertsema, V. (2013)
Implementations of Housing First in Europe: Successes and Challenges in
Maintaining Model Fidelity, American Journal of Psychiatric Rehabilitation 16(4)
pp.290-312.
Homeless Agency. (2008) Counted In, 2008 (Dublin: Dublin City Council).
Kidd, S. A., Miner, S., Walker, D. and Davidson, L. (2007) Stories of Working with
Homeless Youth: On Being “Mind-boggling”, Children and Youth Services Review
29(1) pp.16-34.
Maton, K. I. (2008) Empowering Community Settings: Agents of Individual
Development, Community Betterment, and Positive Social Change, American
Journal of Community Psychology 41(1-2) pp.4-21.
Macnaughton, E., Stefanic, A., Nelson, G., Caplan, R., Townley, G., Aubry, T.,
McCullough, S., Patterson, M., Stergiopoulos, V., Vallée, C., Tsemberis, S., Fleury,
M. J., Piat, M. and Goering, P. (2015) Implementing Housing First Across Sites
and Over Time: Later Fidelity and Implementation Evaluation of a pan-Canadian
Multi-Site Housing First Programme for Homeless People with Mental Illness,
American Journal of Community Psychology 55(3-4) pp.279-291.
Manning, R. M., & Greenwood, R. M. (2018) Microsystems of recovery in
homeless services: The influence of service provider values on service users’
recovery experiences. American Journal of Community Psychology, 61(1-2) pp.
88 -10 3.
Maton, K. I. and Salem, D. A. (1995) Organizational Characteristics of
Empowering Community Settings: A Multiple Case Study Approach, American
Journal of Community Psychology 23(5) pp.631-656.
Nelson, G., Caplan, R., MacLeod, T., Macnaughton, E., Cherner, R.,Aubry,
T.,Méthot, C., Latimer, E., Piat, M., Plenert, E., McCullough, S., Zell, S.,
Patterson, M., Stergiopoulos, V. and Goering, P. (2017) What Happens
After the Demonstration Phase? The Sustainability of Canada’s At Home/Chez
Soi Housing First Programs for Homeless Persons with Mental Illness,American
Journal of Community Psychology59(1-2) pp.144-157.
Nugent, N. and Rhinard, M. (2015) The European Commission (Basingstoke:
Palgrave Macmillan).
O`Sullivan, E. (2016) Ending Homelessness in Ireland: Ambition, Adversity,
Adaptation? European Journal of Homelessness, 10(2) pp.11-39.
54 European Journal of Homelessness _ Volume 12, No. 3, February 2018
Padgett, D., Henwood, B. F. and Tsemberis, S. J. (2016) Housing First: Ending
Homelessness, Transforming Systems, and Changing Lives (New York: Oxford
University Press).
Peter McVerry Trust (2017) Facts & Figures (Accessed on December 18th, 2017
from: https://www.pmvtrust.ie/news-media/facts-and-figures/)
Peterson, N. A. and Zimmerman, M. A. (2004) Beyond the Individual: Toward a
Nomological Network of Organizational Empowerment, American Journal of
Community Psychology 34(1-2) pp.129-145.
Prochaska, J. O., Velicer, W. F., Rossi, J. S., Goldstein, M. G., Marcus, B. H.,
Rakowski, W. and Rossi, S. R. (1994). Stages of Change and Decisional Balance
for 12 Problem Behaviors,Health Psychology13(1) p.39.
Rogers, E. (2003) Diffusion of Innovations, 5th Ed. (New Yrok: Free Press).
Shaughnessy, J. J., Zechmeister, E. B. and Zechmeister, J. S. (2002) Research
Methods in Psychology (New York: McGraw-Hill).
Steadman, H. J., Cocozza, J. J., Dennis, D. L., Lassiter, M. G., Randolph, F. L.,
Goldman, H. and Blasinsky, M. (2002) Successful Programme Maintenance When
Federal Demonstration Dollars Stop: The ACCESS Programme for Homeless
Mentally Ill Persons, Administration and Policy in Mental Health and Mental
Health Services Research 29(6) pp.481-493.
Stergiopoulos, V., Hwang, S.W., Gozdzik, A., Nisenbaum, R., Latimer, E., Rabouin,
D., Adair, C.E., Bourque, J., Connelly, J., Frankish, J., Katz, L.Y., Mason, K., Misir,
V., O’Brien, K., Sareen, J., Schutz, C.G., Singer, A., Streiner, D.L, Vasiliadis, H.M.
and Goering, P.N. (2015) Effect of Scattered-Site Housing Using Rent
Supplements and Intensive Case Management on Housing Stability Among
Homeless Adults with Mental Illness: a Randomized Trial, JAMA 313 pp.905 –915.
Swan, J. and Newell, S. (2000) Linking Knowledge Management and Innovation,
Proceedings of the 8th European Conference on Information Systems
Proceedings
Toro, P. A. and McDonell, D. M. (1992) Beliefs, Attitudes, and Knowledge about
Homelessness: A Survey of the General Public, American Journal of Community
Psychology 20(1) pp.53-80.
