The UK has a rough sleeping crisis and a model called Housing First is being heavily promoted as the solution. The model cannot not just will not work and its widespread adoption will make the crisis worse for rough sleepers and all forms of homelessness. Below is just over 5,500 words on why the Housing First model cannot solve rough sleeping, note cannot not will not and which I say with almost 100% certainty.
Housing First as its name suggests gives the housing (bricks and mortar) first and then the support needs are attended to.
This reads so right and, as a principle, used to be common with emergency access and direct access hostels themselves admitting those who were street homeless and providing both the housing and support element. Yet that principle that has been replaced with gateway models demanded and operated by local councils that sees local authorities saying who could and could not even enter this form of fixed hostel provision let alone receive support.
The Housing First model instead of entering a fixed hostel type provision and receiving first-stage support there sees rough sleepers accommodated in dispersed housing in the community and receive a visiting support service with the level of support reducing over a much longer time period.
There are four major problems with Housing First of
- (1) housing availability,
- (2) its cost,
- (3) the cultural change needed, and
- (4) the visiting support model.
Within each of these four areas I make some necessary comments as to causation of the 169% increase in recorded rough sleeper numbers and debunk the nonsense of Heather Wheeler the Tory minister now responsible for rough sleeping and the government aim of eradicating it by 2027. In a piece in the Guardian today :
Wheeler was asked on Thursday about the reasons for the rise in rough sleeping, which has increased in England for seven consecutive years; official figures show 4,751 people slept outside overnight in 2017. The MP for South Derbyshire said: “In truth, I don’t know. That’s one of the interesting things for me to find out over the last eight weeks that I’ve been doing the job. We’ve looked at the different cohorts, and in London the number of veterans who are rough sleepers is down to about 2%.”
I emphasise the same myth which the Guardian reproduces as fact which says 4,751 people slept outside overnight IN 2017 – when the figures record that as the one night total and the same official figures reveal at least 8100 individuals slept rough in London alone! So apart from the fact the minister admits she does not know she does assert that it is not the fault of the Tories through not funding councils and through welfare reforms. How can she know that!!!!!
I have worked in homelessness for 25 years from a hostel worker to advising homeless providers including rough sleeper services and I developed and managed a Housing First service 20 years ago. Thus, I have no objections to the HF model and know how it and other models operate.
I do however have a major problem with the clamour for HF and how many who really should know better are promoting it as panacea for rough sleepers which is dangerous. When HF fails, and it is a question of when not if, I strongly suspect commissioners and funders and politicians will say we have tried as best we could and even adopted the radical HF model, so done all we can and the problem of rough sleeping is intractable and further adverse assumptions will be made for all single homeless provision.
A significant element of the clamour for HF is the argument that traditional hostel type provision has failed and funders will have to decommission hostels in order to fund the HF model, hence when the HF model fails there is a huge under supply of homeless provision making rough sleeping numbers increase markedly.
The four major problems with HF
1. Housing availability
The HF model is housing-led and is wholly dependent on suitable housing being available. IF suitable housing is not made available then the entire model collapses.
There are two possible landlords in the private rented sector (PRS) or the social rented sector (SRS) who can provide suitable properties and two immediate problems appear.
Firstly the PRS already houses single persons (which is the rough sleeper cohort) in bedsits and 1 bed properties and the question becomes why would a private landlord accommodate a rough sleeper with a much greater risk (actual and perceived) for no greater money? The private landlord is all about risk and reward and making the most profit possible from the housing asset. The HF model provides much greater risk of client yet has no greater financial reward so without the greater financial reward the PRS landlord will leave HF well alone. There are some possibilities and models that can work here and ones I have developed and used that result in giving the private landlord more in income and they are too complex to discuss here and even so are undermined by government intended housing benefit reforms.
In summary unless the HF provider provides the greater rent money that private landlords will demand for housing a more risky client group in rough sleepers then the PRS will not provide the housing element and to avoid any doubt PRS landlords will not touch this with a bargepole
Second, this then leaves the social rented sector of council and housing association landlords, to provide the suitable housing properties and at the scale needed. Yet SRS landlords are much more risk averse when it comes to housing single people generally and have always been loathe even to provide shared accommodation for low risk client groups (e.g. two nurses sharing etc) and now with uncertain housing benefit changes are even more risk averse.
However the main problem with SRS properties forming the HF housing element part is a chronic lack of supply within the social rented sector of bedsit and 1 bed properties that is the housing need of single persons who make up the rough sleeper cohort. Leaving aside the much greater demands in the SRS for smaller properties that the bedroom tax gives, the SRS has just 24.03% of all its properties being 1 bed (22.66%) and bedsit (1.37%) as detailed in the English Housing Survey and Statistical Data Return and is very reluctant to give these up for such a perceived and actual high risk client group in rough sleepers.
That 24.03% proportion is significant when it comes to the 3 pilot areas for the HF model in Birmingham, Liverpool and Manchester as assuming that no reluctance to give up properties existed and further assuming that cost and culture change and the HF support model worked, all of which are very naïve, then having 24% of SRS housing stock being 1 bed or bedsit equates to a 100% chance of success.
Yet Liverpool has just 76% of this English average and therefore has a 24% less chance of HF succeeding due to lack of suitable properties (and Manchester just 83%) thus they are curious or is that ignorant choices for pilot areas?
Much of the hype surrounding the Housing First model as panacea stems from a 2017 report commissioned by Crisis for the Liverpool City Region and which said that LCR is perfect for a large scale HF pilot … without any consideration of the stock profile in the six councils in LCR at all in its tome length report of 170 pages … which I remind again never even asked the question of the availability of housing for this housing-led Housing First model!!
Liverpool has 76% of the English average of 1 bed and bedsit properties
Knowsley has 70% of the English average
Sefton has 78% of the English average
St Helens has just 56% of the English average
Wirral has 75% of the English average of 1 bed and bedsit properties
This leaves Halton which has 95% of the English average of 1 bed and bedsit properties which stands out in the Liverpool City Region as having a much higher proportion than the other 5 councils yet still has less than the English average.
This is more than a comment that proves the ignorance and hyperbole of the Crisis claim that LCR was perfect for a large-scale Housing First pilot, in fact the LCR area is one of the last places that is ‘perfect’ for such a model given the facts of its housing stock profile which I repeat was not even considered in this Crisis report! It has practical and disturbing implications.
LCR has a Metro Mayor in Steve Rotheram who enthused over this Crisis report given its (potential) political brownie points of solving rough sleeping and especially for a new political construct of a Mayor for a devolved region. Yet, if the Housing First model is implemented across LCR it would mean rough sleepers from the 5 council areas of Knowsley, Liverpool, Sefton, St Helens and Wirral would be exported to the Halton area as that is the only area with anywhere near suitable housing stock for the rough sleeper housing need!
Good luck with that Mr Rotheram and just for added irony I note that you have engaged Crisis as your adviser on homeless issues (!) and also that you have engaged Barbara Spicer chief executive of Plus Dane as you housing advisor … and Plus Dane have the lowest percentage of 1 bed or smaller properties at just 8% of all social landlords with over 500 properties in Halton! And good luck with the fact that LCR at just over 19.5% of all social tenants affected by the bedroom tax, which is 50% more than the English average, and thus has 50% more demand on these smaller properties than the average English area.
I don’t mean the above comments to be flippant which they might read as they are very pertinent issues of fact that act as huge practical barriers to Housing First solving rough sleeping in the Liverpool City Region area.
For example what if Halton Housing Trust who own 65% of the total 1 bed or less properties in Halton decide they are not going to get involved in any Housing First service? What is KHT in Knowsley say they are not going to get involved in HF and they have 85% of all 1 bed properties there, and/or OVH in Sefton say the same with the 73% of all 1 bed properties in Sefton?
There is no council housing across LCR and only housing associations who can and will choose to get involved or not based on whether this is the best use of their stock of 1 bed properties yet even that obvious practical issue has not been considered in the will Housing First even find the housing necessary question and without which the model dies
Facts that have not been considered and facts which mean the hyperbole over solving the rough sleeping problem in LCR only reveal that the problem has NOT been identified which you are seeking to solve.
Housing availability – quantity needed
I despair at anyone who says that the rough sleeper need is the 4,751 number of rough sleepers in England that the official rough sleeper count recorded on one night in the Autumn of 2017. This is a count for one-night only and is a chronic underestimate of the number of rough sleepers and thus anyone who advocates the HF need for properties across England as 4,751 is a deluded idiot.
Regrettably only yesterday I read an article in the Guardian from Nicholas Pleace the professor of social policy at the University of York on Housing First co-authored by Joanne Bretherton a research fellow there and co-director of the Women’s Homelessness in Europe Network.
In England, the government plans to use Housing First to reduce rough sleeping, including a £28m programme to pilot Housing First projects in three areas of England, announced last November. There are about 4,750 rough sleepers in England: a relatively small proportion of the 61,000 homeless families in temporary accommodation.
When even academia accepts the nonsense of the rough sleeper count in discussing and considering the Housing First model you start to get what I mean about the clamour for HF and how it is being remarkably ill-thought through even by academics. Below is that piece from the Guardian Housing Network that while arguing HF is a part of a solution, which it is and only can be, still makes ignorant naive assumption over rough sleeper numbers and once again fails to mention cost of HF services.
To even attempt to answer the question Will the Housing First model solve rough sleeping you must ask first ask the question of How many rough sleepers are there in England yet it has not been asked and the scale of the assumption, ignorance and naivety in not asking how many and accepting the 4,751 figure as the answer is breathtaking in its error!
I posited a simple yet nonetheless valid estimate of there being close to 34,000 rough sleepers in England here and which means that the Housing First model in order to solve rough sleeping needs to find 34,000 properties. Further, if you accept the arguments that the hostel type single homeless model has had its day, which is what many HF adherents are saying, you then need to add the 36,000 or so who reside there to this figure which then becomes a HF need for 70,000 properties – and that also assumes you halt completely the flow of new rough sleepers too!
When the official rough sleeper count figures for England were released in January 2018 I posted about how absurd they were and stated:
Note too that the figures are for one night only and the 1,137 recorded in London belies the total number of 8,108 rough sleepers recorded in the capital in 2016/17. If this figure is replicated across England then there are 33,880 who sleep rough each year in England and not the headline one night number of 4,751.
That is why, despite being simplistic, the rough sleeper numbers and thus the immediate rough sleeper housing need is not 4,751 but much closer to and may even be above some 34,000 in number. As I said in opening if you can’t get the housing properties for Housing First then Housing First as a model dies!
Note well that my crude number estimate was said in the context of Theresa May saying she was going to eliminate rough sleeping by 2027 at Prime Minister Questions in December 2017 and she quoted the same £28 million pilots in Birmingham, Liverpool and Manchester as evidence of this. Thus the Prime Minister has been similarly duped into taking on board the woeful and risible Crisis research into Housing First hook, line and sinker!
I also posted last week some musings over rough sleeping in Liverpool as I walked around the city centre last Tuesday in the middle of the day and personally witnessed 46 rough sleepers yet the official count for Liverpool is just 33 in total. This was in the middle of the day and 3 of those 46 were female yet Liverpool’s official rough sleeper count is 1 female only!
If that is too anecdotal for you ask yourself why Liverpool City Council opened a new rough sleeper shelter in late 2017 and last month announced they were opening up a new floor of this building (Labre House) to accommodate an additional 40 rough sleepers doubling its capacity to 80. An additional 40 and 80 in total yet the totality of rough sleepers in the city is officially just 33!
One final point on the official rough sleeper count figure for England is that 4,748 of the 4,751 claimed figure are over 18! Can anyone possibly believe that just 3 of the total rough sleepers in England are 18 or under!! Yet that is what the official rough sleeper count figure says and below is a screenshot from the official tables.
The above are just some of the real practical issues of whether housing properties will become available for HF or not. While I have concentrated on the Liverpool City Region to illustrate the paucity of suitable properties I have figures for every local authority area and using the critical English average of stock of suitable properties reveals that HF has a much better chance of working in Brighton which has 149% of the English average 1 bed properties but no chance at all in Barking which has just 46% of that average – and they are not even the polarities of local council areas beginning with the letter B!
2. Housing First costs
Yes all of the above is just some of the issues of point 1 of housing availability, just some and has taken 2500 words to describe some of the main housing availability issues of the housing-led Housing First model. There are a great many more issues around the lack of housing availability.
I have touched on the main areas of housing-cost by positing that the PRS landlord will not touch the HF housing element with a bargepole as there is no more money in it. The PRS landlord can make the same income return whether in bedsit, 1 bed or shared HF housing from other single persons as they do now so there is no reason why they will take a higher risk client group in rough sleepers for no extra reward.
The same is the case for social landlords as they too have the same profit motive of the PRS landlord it is just called the making better use of stock or MBUS issue, which was the name given to a CIH inspired programme many years ago. So even if social landlords soften their reluctance to high risk client groups or to shared housing they now have the option to consider rough sleepers in the HF model with say two nurses sharing a two bed property each paying the affordable rent level (or the Magenta model which transfers properties from social housing stock to a private sector subsidiary and charges 59% more in rent for a former council house.)
In summary on the housing costs element of HF all renting landlords can get the same financial returns from much lower risk client groups than rough sleepers and that is precisely what they will do. Even where we have council landlords operating we will see the private landlord organisations they are creating in Local Housing Companies will prioritise housing stock for client groups that save the local council budgets and not rough sleepers who now cost them nothing. Priorities such as foster carers, homeless families but not single homeless (though that may change to some degree with the Homeless Reduction Act coming in from April which essentially gives single homeless persons more rights to housing and thus creates more cost for local councils because of the additional duties.)
The REAL issue with Housing First is the support cost which is excessive and much higher than the current support costs paid into hostel type provision.
The Crisis report for LCR in 2017 made reference to 2014 examples of HF in low to medium rent areas in England and had a unit cost – the cost of one hour of support – at £34 in 2014. A 2018 report by St Mungos had a unit cost of £39 and £40 per hour as I discussed here for Crisis and here for St Mungos that was authored by the same Nicholas Pleace of York University I mentioned above which accepted the 4,751 rough sleeper count figure as correct!
Yet the unit cost in hostel provision is invariably less than £20 per hour and its English average will be less than £20 per hour of support. In broad terms the cost of one hour of HF support is double the cost of one hour of hostel support.
Surprise, surprise this doubling of cost of HF support is claimed to be cheaper than hostel unit cost of support in the Crisis report with sophistry and sleight of hand aplenty and no mention of this doubling of cost of HF support and its comparison with hostel unit cost is considered in the St Mungo’s and University of York study either; rather both reports infer strongly that the hostel model has failed and use that inference and assertion to say we must try the Housing First model instead. This is a less than subtle narrative of all who promote HF and all such studies and articles blame and cast aspersions on the hostel model.
There are failings in the hostel model, many failings in fact yet they all come down to funding and its lack of funding to be precise. I will address that issue shortly but the key issue here with this pro HF / anti hostel narrative is that there is a finite amount of funding that will be spent on single homeless services to include hostel and HF model. So in order to begin more HF services councils need to defund and close hostel services as consequence which is strongly in my view a policy of throwing out the baby with the bath water.
There is also a huge difference between the hostel and HF models in that the HF model is a single stage model while the hostel model is a multi-stage resettlement model and the clamour and hyperbole of the HF research only sees a false comparison between the entirety of HF compared with just the first-stage hostel service.
The hostel model is immediate and high level need catered for in a set fixed building, the hostel. Once support needs have been assessed a second-stage service is in play with either a core and cluster type linked accommodation to the hostel or straight into the residents own property with visiting support. The HF model only has this one last stage and purports to be able to provide effective support in this one stage only.
Prior to welfare reforms (aka cuts to housing benefit entitlement) and even prior to the Supporting People funding regime it was notoriously difficult to find such ‘move-on’ properties from hostels within the social rented sector. As such hostel providers turned to the PRS for these move-on properties as their only real option and then only for hostel residents who were aged 25 or over due to the SAR policy. In 2011/12 the government changed the age limits on the shared accommodation rate or SAR to those aged 35 or over and those below 35 were restricted to this shared accommodation rate of housing benefit which meant that the move-on route from hostel to the PRS collapsed and thus no exit route for hostel residents became the norm.
Hostels thus ‘failed’ through circumstances beyond their control and instead of being the first-stage to resettlement in the community, hostels became holding pens that in turn led to single hostel residents becoming more volatile as there was no light at the end of the tunnel. This also led to LA commissioners saying hostels are not working as their throughput rates are increasing (residents staying there longer) and since 2010 a loss of around 6000 hostel beds has taken place due to the alleged ‘failure’ of the hostel model caused outwith hostel providers control.
Yet another reason for the above is another instance of false comparison and one that will be very relevant, when and not if, more HF services come online and more hostel places close by consequence and that is age. The vast majority of single homeless hostel residents are under 25 while the vast majority of rough sleepers are the direct opposite and over 25.
As the same screenshot of the alleged 3 only rough sleepers under 18 reveals it also say 81% of rough sleepers are over 25 so when HF services replace single hostel beds it also replaces current rough sleepers with much younger rough sleepers as those who are victims of hostel closures will be younger single people … who will have even less of a chance of securing any form of housing provision!
Put simply, if landlords will be reluctant to let to over 25 year old rough sleepers then they will be twice as reluctant to rough sleepers under the age of 25!
Within the broad area of costs of the HF model we have to look at what cost has done to the hostel model with which HF is falsely being compared to and cost through the change to housing benefit entitlement with the SAR age change and the widespread uncertainty over how housing benefit goes forward now and the additional issue of the effective banning of housing benefit to all those under 22 years of age are not just cost factors, they are causal factors in the 169% increase in rough sleeping numbers since 2010.
3. Culture change
As discussed earlier the responsibility for deciding who gets Housing First accommodation will pass back from local councils who now control it through gateway services and go to HF providers and this is a significant culture change. Local authority gateway services all tend to restrict homeless accommodation and support provision to those who have a local connection with that locale. This acts as a huge constraint to rough sleepers who invariably and by definition do not meet local connection criteria such as having lived in the borough for 3 or even 4 years out of the last five years.
The nexus of control passing from local authority gateway teams to HF providers will meet huge resistance from local councils and local councillors – as it always has with local councillors saying and particularly so in core cities why should we fund supported housing residents from outside our area.
The pilots of Liverpool, Manchester and Birmingham is a good example of what I mean by core cities in cities that are surrounded by many neighbouring authorities in Merseyside, Lancashire to Liverpool, with the other 9 local councils in the Greater Manchester area to the City of Manchester and the West Midlands councils to Birmingham.
Manchester council and its councillors have always objected to residents of Bolton, Bury,Oldham, Rochdale, Salford, Stockport, Tameside, Trafford and Wigan making use of the much greater supported housing (accommodation-based) services available within the City of Manchester local authority borders. Not any more so than Liverpool or Birmingham councils and councillors I hasten to add but I use here for illustration. Core cities in all regions have always had more provision than satellite neighbouring councils and especially when it comes to single homelessness of which rough sleeping is a subset.
Funders and commissioners of any HF service will be local councils not regional devolved mayors and for example (a) Manchester city council will want to only fund Manchester rough sleepers and not rough sleepers in Stockport or Salford or any other bordering local council. Then add on (b) potential joint funders of HF services such as health and social care and whose boundaries are not co-terminous with local councils either and will also want to ensure that NHS area X only co-funds rough sleepers from NHS area X and not from NHS area Y. So even if suitable properties materialise a part of suitable housing in this housing-led HF model is location.
Even if the funding for all supported housing services (of which HF is a part) devolves to a regional level and say the LCR holds the budget not the six individual constituent councils within LCR then that in itself is a massive cultural change in practical terms (and will mean bunfights between those 6 councils too!!)
Councillors and councils will be very reluctant to move to a devolved regional budget holder as it (a) sees loss of local council control; (b) sees loss of local councils ability to meet their local mandated housing strategies which they have even if they have no council housing; and (c) has significant political fall out over power, influence and importance generally. As I said previously local councils will be resentful and not give up their gateway control they now have even to providers in their local council area let alone give that control up to a devolved regional construct such as a Metro Mayor.
You see why the term bunfight has been used!
A massive cultural factor is also at play with landlords as I have discussed earlier who because they have the properties needed to make HF succeed or not, hold the power. They will find just as much profit in the PRS from client groups other than rough sleepers in the PRS and in the SRS the landlords will find in their view better use of stock than the high risk actual and perceived that the rough sleeper client group holds.
A further issue that straddles cultural change is the suitability of the HF support model to which I now turn
4. Suitability of HF support element
As stated above the support element of the HF model is visiting support and the non-accommodation based type and the opposite of the hostel model. Visiting support, often called floating support, is (a) a 100% planned model in theory and is (b) less reactive to support need than an accommodation based service at a hostel with support staff on site 24/7, and (c) assumes in its planning and design a permanent reducing need for support basis with little if any design for the proverbial hitting the fan for the (former) rough sleeper.
Straight away we see the idea of selling the HF model to landlord being much more difficult as they would need to accept that support need and the risks that go along with that is one of off-site support that is less qualitative and slower to respond when (and not if) the proverbial hits the fan. All landlords will be minded of anti-social behaviour aspects at this that already exacerbates their perceptions of rough sleepers as already having a huge risk to ASB concerns. This is another reason why landlords in the PRS or SRS will NOT be falling over themselves to offer up housing that the housing-led HF model demands and dies without. Given that there is no extra financial reward for this greater risk makes that even more obvious.
One of the other reasons why the traditional hostel model has been argued to be ‘failing’ is that commissioners and budget holders tend only to see planned support as the only form of support for single homeless client groups yet I would argue that at least 40% of the necessary support for single homeless residents in hostels is reactive support and for the rough sleeper subset of single homelessness is 60% or even more.
It is fairly well documented that rough sleepers have a high level of mental health problems and similarly that many have and use / misuse alcohol, illegal and legal drugs to cope with mental health issues and the many other support needs they have. Physical health problems such as TB and many other medical needs are prevalent in the rough sleeper cohort as are a complete lack of trust in others as are scores of other social and health problems relating to why they came to live on the streets in the first place such as abandonment, the care system, abuse they have suffered and are fleeing (an issue we readily see in domestic violence and abuse yet tend to never see in rough sleepers!) and so on and so on.
All of these issues we are interrelated and complex flare up with regular frequency and occur perhaps more frequently in this client group than in any other (allegedly homogeneous) client group.
To explain that in lay terms the HF support model is visiting support mostly weekdays and not weekend and a resident will have planned visits and possibly an out of hours hotline. Yet place all of that into the time frame between 5pm Friday and 9am Monday and the proverbial hits the fan for the former rough sleeper. He may be able to call his support worker or hotline yet he has no credit on his phone as the model assumes that a former rough sleeper will be able to manage his financial affairs the second a roof is put over his head! An hugely errant assumption that Universal Credit will see become a huge issue when all benefit recipients move to monthly payment of social security benefits!
Whatever the proverbial is at 5.05pm of a Friday evening you can rest assured will have more sever impact than for any other client group due to the complexity, mulltiplicity and interrelated very high support needs that the typical rough sleeper will have. For a former rough sleeper running out of electricity in the meter and having no mobile phone credit is a much greater issue than a resident in a hostel or refuge who has on-site support staff with access to a telephone. The social isolation the proverbial hitting the fan at 5.05pm of a Friday is also far worse in context.
Then we see one of the first if not THE first role of the HF support worker is to get the former rough sleeper in the social security benefit system and with recourse to welfare benefits. Yet that also moves the former rough sleeper into the ambit of the sanction and an immediate move of the former rough sleeper into a routine of job search and all other UC conditionality when he has been used to coming and going as he pleases and having no UC or other master. I wish all HF support workers well in seeking to persuade DWP and Job Centre staff that the former rough sleeper should have extra allowances made for his receipt of benefit because he is a former rough sleeper and not used to any form of conditionality or routine!
I could give dozens (literally) of other proverbial examples and how the HF support model assumes that just putting a roof over a former rough sleepers head will solve anything other than the client being pissed on by the rain … which also happens in Summer time too not just winter!
In summary, the hyperbole over HF is dangerous and dangerously misguided. It can work as I know and have witnessed but when we have homeless lobbies and academics believing as many of the public believe that merely putting a roof over a rough sleepers head is panacea and will solve the rough sleeping problem then abject delusion has set in.
The chronic lack of available and suitable housing and the cost of both the housing and support element of HF along with the massive cultural changes necessary all mean that HF cannot and not just will not work.
HF as a proposed solution to rough sleeping needs all current nonsense and research to be ripped up and be started again from a practical and pragmatic or even cold as a fish perspective.
Rough sleepers deserve far better than being set up to fail and that is what the Housing First model does.