Housing First doesn’t and can’t ever work. Rough sleepers deserve far more than this hyperbolic cult

The only things wrong with Housing First are its housing offer and its support offer … and the refusal of its zealous adherents to face reality. Or in short its whole kit and caboodle.

This ‘model’ can never possibly work in the UK when looked at critically using fact and contexts and especially not in England. The HF model has more propagandists than the Stasi had informants and scrutiny and critical evaluation are nowhere to be seen except in those pesky objective things called facts. Those who are roofless deserve far, far better.

I briefly touch on why the Housing First model cannot possibly work in the UK in this 5000 word overview such is the extent of the deception and propaganda that surrounds this cult-like model. It cannot work, it is prohibitively expensive and its proponents do not even envisage it supporting any more than 1 in 9 rough sleepers yet say nothing about the 8 in 9 rough sleepers it will not support. HF will also create more rough sleeping than we have today WHEN we look critically at the facts. So let’s start to look at those facts and why it is dangerous madness to ‘scale-up’ this turkey …

The ongoing countrywide HF model in Scotland takes 149 nights on average to find a suitable property for a HF client – 5 months – with the best performing area, Stirling taking 101nights and the worst in Edinburgh taking 195 nights or 6 and half months to find a suitable property!  The term Housing FIRST is a chronic misnomer. A HF service in Sheffield closed after 18 months as a mere 10 suitable properties in either the SRS or PRS could not be found in a city of 220,000 properties. Such facts are glossed over in wave upon repeated wave of propaganda, hope and hyperbole.

Housing First is nothing more than a cult.

The Liverpool City Region is a large scale pilot area for the HF model and was said to be the ideal HF area by Crisis in the 168-page tome report it commissioned. Yet the six local authorities in LCR have 17% of all rented properties being the needed 1 bed type when the English national average is 24% which is just one example of basic yet critical fact not even being considered by the zealots ahead of their headlong blind bull in a china shop adventure.

The Greater Manchester Housing Partnership claimed huge success for its HF service providing 357HF properties over three years … between 24 of the largest providers and which is 5 rough sleepers per large housing association per year!  The current large-scale pilot programme across the Greater Manchester Combined Authority has supported 200 former rough sleepers in 2 years across its 10 LA areas or 10 per LA per year with £3m of government funding per year.  An average of £15,000 per rough sleeper per year just in support costs with rent costs being additional making the yearly cost per HF client being over £20k per year set against the average £9.5k pa of the hostel resettlement model that HF wants to replace and claims to be cheaper than in its propaganda when it is more than double the cost.

All of the example above typify the incredible hyperbole over fact of the Housing First model and its abject denial of the basic fact that the availability of the suitable 1 bed property for the Housing First client is as rare as hen’s teeth AND the selling of snake oil that surrounds the model.

Jon Sparkes the chief executive of Crisis is the propagandist in chief of the Housing First model with the Scotland example being typical.  He proclaimed that the 404 achieved cases in Scotland were a stunning example of the success of HF and we must roll it out across England. 

What Jon Sparkes and Crisis WANT you to see and believe
What Jon Sparkes and Crisis and other propagandists DON’T WANT you ro see and read!

The basic arithmetic fact reveals the Scottish example, the first wide scale HF service of 404 rough sleepers over two years, means the HF model managed to house just 4% of the 5,000 per year rough sleepers Scotland has (and its government accepted this Shelter Scotland estimate) and did not rehouse 96% of rough sleepers AND it took 5 months to find the suitable property on average with no comment on what rough sleepers do for 149 nights before they are housed! 

This Housing First with an average five-month wait model is not a catchy name yet it describes the reality.  It also means Housing First needs a pre-stage housing and support option for the 4% of rough sleeping clients it eventually deals with and the other 96% of those sleeping rough can stay where they are as roofless and not in the cherry-picked easiest cohort that is clearly going on in Scotland and the GMHP example, the latter having a perverse financial incentive of a Social Investment Bond (SIB) rewarding providers with bonuses the longer a rough sleeper stays in the accommodation.

Housing First  – A Zero Sum game

The rough sleeper has to compete with the general needs tenant for the rare as hen’s teeth 1 bed property as I pointed out here with the official statistics stating 46% of those on council waiting list are waiting for the 1 bed property.  The rough sleeper also competes with the single homeless hostel resident for the same 1 bed property to escape homelessness and every 1 bed property given to the HF client is one less 1 bed property available for the homeless hostel resident which means they stay in hostel longer and thus deny the hostel taking in 1 more rough sleeper from the street which highlights the perversity of the Housing First (after a 5 month wait) model. 

It also means the one-third of domestic abuse residents who are single and also require a 1 bed to finally escape domestic abuse are denied the 1 bed property they qualify for and need when one is given to the HF former rough sleeper client.  The room at the refuge they live in is similarly bed-blocked and denied to a domestic abuse fleeing individual or family who has to be turned away from entering the refuge. 

A finite number of 1 bed properties become available to rent each year and we have a zero-sum game.  Rent to a HF rough sleeper and that denies the hostel dweller, the refuge resident and the general public.

This zero sum game is most acute in England with the acute shortage of one bed properties. England has a decades old structural issue that will take decades to resolve as we need at least 100,000 new one bed properties per year at the social rent level that are reserved for single homeless cohorts. 

The structural 1 bed property problem

Many reports call for between 90 – 130,000 total new social housing properties per year of all sizes and none of these reports by CIH, NHF, JRF or the claimed Independent Commission of Shelter ever include a breakdown of the 90k to 130k figure by size or bedroom number. ALL of them are requests for government capital subsidy funding that do not even bother to tell government who they are for and directly because they do not break down the global figure.  This continues the post-war trend of the ‘great and good’ of social housing not building for objective housing need and simply calling on government for funding those properties that either make them the most money or are political acceptable. 

Landlord risk aversion

The zero sum game of a finite number of 1 bed properties being available each year sees the Housing First model simply create a new choice of prospective tenant for them – a distinctly unappealing choice with no incentive for a landlord to use the 1 bed property for a former rough sleeper.

Every landlord has a choice for each 1 bed property that becomes available of

  • rehousing a rough sleeper direct from the street (very high risk) or
  • a hostel resident (high risk) or
  • a single female from refuge (medium risk) or
  • a general needs (low risk) tenant

 – With no greater financial reward for accommodating the (perceived) very high risk rough sleeper than for the low or minimal risk general needs tenant. 

Why on Earth would any landlord take on much greater risk of tenancy failure and additional cost for no extra financial reward? 

This obvious practical question is yet another issue that the zealous HF model adherents conveniently sweep under the carpet.  That obvious question is further compounded given the acute shortage of 1 bed supply and further compounded by the affordable rent regime.

The affordable (sic) rent regime and Housing First

In Liverpool (which has no council housing) the average housing association rent for a 1 bed general needs property is £77 per week at the social rent level thus the landlord receives £77 per week in housing benefit or UC housing element.  The 1 bed maximum LHA rate in Liverpool is £92 per week so the private landlord can receive £92 per week in housing benefit.  However, Liverpool has the highest number of the misnamed “affordable rent” properties in the UK at over 6,000 and HAs charge £104 per week for a 1 bed AR property which means they receive £104 per week in housing benefit. 

This £104 per week AR level is 35% more than the HA receives in social rent and is 13% more than the maximum a private landlord can receive in housing benefit – and it is receivable for the general needs and lowest risk tenant.  The HA landlord can choose as a desktop exercise to convert a social rent level 1 bed to an affordable rent level 1 bed at is bidding.  The former tenant paying £77 per week leaves on a Friday to be replaced in the exact same property with a new tenant paying £104 per week on the Monday

I ask again … Why would the HA landlord accommodate the highest risk rough sleeper direct from the streets for no increase in rental income?

It is these basic and obvious questions set in the factual context that the Housing First zealots conveniently ignore when they zealously promote the absurdly theoretical HF model as the solution to rough sleeping.  Every industry works on the basis of higher risk must equal higher reward and is a given of any business so why do the HF zealots choose to overlook this obvious issue?

Housing First – The Comparative Bullshit

The same HF zealots are very fond of saying the HF model works so well in Finland (largely correct) so therefore it will work in England.  That is a chronic non sequitur which again ignores the critical facts that differentiate the English welfare system from the Finnish welfare system. 

In Finland the final HF 1 bed property is limited to one-third of the gross market rent which means using the Liverpool analogy a rent of £45 per week would be the maximum rent level.  I say final as the Finnish model is not a true HF model as rough sleepers are first placed in hostel shared provision until the 1 bed final accommodation becomes available and to those who have proved their readiness for semi-independent living. The notion that the Finnish HF system is rough sleeping directly to own property is a myth as this article reveals from the Guardian from which this comes

Read the article and it states the above self-titled “homeless hostel” is a stage before independent living

On a closer look the Finnish rough sleeper receives (2018/19 figures) €203 per week and crudely 3 times the English level of base benefit and the Finnish model guarantees by right the immediate availability of a social worker, probation worker, drug, alcohol, mental health and any other needed support worker on demand. 

The Finnish HF model also includes free breakfasts and free saunas too and main meals for a subsidised €1 cost.  The comparison is chalk and cheese … and yet another fact that the HF zealots in England very conveniently choose to ignore!

Many previous Housing First articlea I have written after extensive research summarises the main differences as

The same zealots are also fond of flippantly asserting that the HF model works in America so it must work here in the UK.  Yet the Housing First model does not work in the USA at all and see the Manhattan Institute’s scathing attack on it below. It does not include other facts for example HF was introduced in New York and New Jersey from the mid to late 1990s when New York City alone had 33,000 homeless persons (rough sleepers / night shelter residents) which by 2018 had increased to 78,000 despite the ‘panacea’ of the Housing First model. 

On a personal anecdotal basis I used to visit New York and New Jersey every year from the mid-1990s as the brother of my former partner worked in NYC and lived in New Jersey.  I walked into many homeless shelters that litter the centre of NYC and spoke with staff working there and I visited a number of HF services in New Jersey – a busman’s holiday if you will as a supported housing consultant – and have closely followed the HF and other homeless models in NY and NJ ever since and got my head around HUD and other funding of the US housing welfare system. Obama and even Trump tried the Housing First model and these two very divergent politicians could not make Housing First work and despite throwing more money at it than the UK government is willing to do by comparison too. 

I also set up what would now be called a Housing First model more than 20 years ago in advising a housing association with chronically difficult-to-let tiny 2 bed terraced properties to rehouse single people direct from a local detox service which raised a few eyebrows but was extremely successful and mutually beneficial as the landlord eliminated its void problem of its DTL’s and vulnerable detox residents got their own place of a far higher standard than they ever thought instead of going through the hostel system  and waiting for months for something inferior to eventually turn up. This single homeless cohort like all others saw this unexpected quality of provision as a huge incentive not to f*ck-up again and get their heads out of their backsides being their stated views. Note too this was at a time when social landlords were saying bedsits were no longer good enough for sheltered housing so let’s remodel them as being good enough for homeless persons!

Today such disgraceful arguments remain as common unfortunately

Housing First – The Bedroom Tax Problem

The detox example was one of many similar services that could no longer happen with the bedroom tax policy in place since 2013 and neither can a social landlord give a small 2 bed property to a single women from a domestic abuse refuge which was also quite common and again now unavailable due to the bedroom tax policy. The bedroom tax forces SRS landlords to only allocate based on immediate presenting need for bedrooms which means a newly married couple is given a 1 bed and has to move home if they start a family.  The single women moving out of refuge was pre Bedroom Tax offered a 2 bed property in order to free up a refuge room for a family fleeing domestic abuse is another example of good practice that is no longer available due to Bedroom Tax policy and alleged ‘reform’ which means to improve not make worse as has been the reality.

In short, many so-called and misnamed ‘welfare reform (sic)’ policies exacerbate the property availability problem and severely constrain the solutions for all single homeless cohorts not just the rough sleeper.

I could go on with many more nuanced issues that sees the Housing First model is delusional in believing that the ethereal 1 bed property – which the HF model is entirely based upon – is simply not available and the model has no chance whatsoever of ever succeeding in rehousing the rough sleeper.  The model also has a critical problem with its visiting support offer too and to which I now turn.

Visiting Support

In 2000 and as part of the Transitional Housing Benefit System (THBS) – the precursor to Supporting People (SP) – the then Labour government issued guidance on visiting support services to local authorities who adjudged claims for support funding.  The HB adjudication circular (HB A47/2000) stating that a visiting support service of up to 5 hours per week was a low level service, 6 – 21 hours per week a medium level visiting support services; and 21 hours or more per week of visiting support was a high level service.

The Housing First model typically sees just 3 hours of visiting support PER WEEK when the only ever guidance document on reasonable, realistic and justifiable visiting support for a high and complex need client group such as the rough sleeper was for a minimum of 3 hours PER DAY of visiting support.  It is immaterial as to whether you believe that 3 hours per day (21 hours+ per week) is right, the issue is how can ANY model of visiting support make any sense or frankly work, when it is one seventh of the guidance of 20 years ago?

Another factor is in 2005 the Audit Commission produced a baseline report of actual in-payment costs of support per hour – or unit cost – which found the cost of one hour of visiting support was 34% higher than the cost of one hour of support in an accommodation-based service such as a homeless hostel. 

With a finite budget always being the case in terms of support (which is 100% discretionary and not a right as some forms of care funding is) the cost of support is always a critical issue.  Under Supporting People 2004 actual figures of £17 to over £23 per hour, the accommodation-based support cost at a hostel funded 6 hours of support every £100 invested yet the visiting support model to homeless persons bought just 4 hours of support for each £100 invested.

Recently the oft-stated cost of visiting support in the Housing First model is said to be £40 per hour by its adherents who also claim this is good value (in fact they sell HF as being cheaper than the hostel resettlement model when it is far more expensive) yet the average cost of visiting care service funding is less than £18 per hour as an English national average.  Why is the visiting support cost under Housing First more than twice the cost of visiting care services is one question the Homeless APPG who are currently determining HOW to scale up Housing First (and not whether they should or whether it can work, which it can’t!) need to consider: Yet they too are swayed by the hyperbole and propaganda of the HF model and, presumably, as Crisis the propagandists-in-chief of the HF model also provide the secretariat for the Homeless APPG!

The APPG being all party means politicians from all parties are convinced about the Housing First model, which the above part detail discussion reveals to be delusional, and frankly smacks of government wanting to be seen to be doing something rather than doing anything of substance.  It’s not the first time an apolitical cock-up has happened and won’t be the last yet moving forward with Housing First is a very dangerous pursuit.

We have thrown millions at what the ‘experts’ (Crisis et al) said would eradicate rough sleeping so rough sleeping must be unsolvable and therefore we are not going to throw any funding its way in the future … is the position that all future governments will adopt WHEN the Housing First model fails as it inevitably will.  That is the real danger in going ahead with a model that has been proven NOT to work in the UK and which all of the facts, factors and context all say it cannot possibly work.

It is entirely understandable that everyone wants the HF model to work yet that is extremely improbable as the facts demonstrate and is no reason to move headlong into adopting the HF model which appears to be the case from the Homeless APPG given its call for papers solely on how we scale up and not whether we should or whether HF will work. 

The Political Economy Problem with Housing First

The new Boris Johnson administration declared in December 2019 shortly after winning the last general election that they would ‘eradicate’ rough sleeping by the end of the current parliament that is by 2024.  The previous Theresa May administration had a policy to halve it by 2022 and eradicate by 2027, a policy aim hastily stated in response to the severe winter known as the Best from the East, that the current government with the extremely ambitious Jenrick at MHCLG brought forward and gave £29m in support only costs to three large-scale pilot areas in Birmingham, Liverpool and Manchester. 

£29m for 1000 rough sleepers over three years and just for support costs with rents being met by housing benefit gave a target figure of £9,666 per rough sleeper for support per year and a figure that equated to the average cost of hostel provision for both accommodation and support of circa £180 per week.

Simple arithmetic tells us that the 100,000 homeless hostel residents England has each year (circa 40k hostels with an average length of stay at 21 weeks)  would need 100,000 lots of £9,666 per year of support only costs or £966 million per year in Housing First supply only cost for the HF model to replace the failing hostel resettlement model – failing as they have been bed-blocked for decades due to the same lack of 1 bed move-on provision to provide the escape from homelessness, the same structural non-availability that Housing First is now seeing and which the HF zealots and propagandists chose to ignore and believe was not at play.

England has some 50,000 rough sleepers per year as a cautiously low estimate and something that apparently shocked the government and its ‘experts’ in Crisis et al when the pandemic brought about the need to get all rough sleepers off the streets as a public health issue. 29,000 in the first 6 months were put up in temporary hotel style accommodations leading the National Audit Office to say rough sleeping numbers were 8 times what the government previously estimated and believed.  The costs of ending rough sleeping spiralled in the minds of the bean counters of Whitehall which led to the right wing CSJ think tank set up by Iain Duncan Smith seconding a Crisis employee to come up with the Close To Home report published in February 2021.

Close to Home asserts that just 16,500 rough sleepers require the Housing First support model and require HF visiting support for a target duration of three years.  However, the English system produces 50,000 rough sleepers per year so 16,500 out of 150,000 or 11% of the rough sleepers over three years is the government seeking to reduce costs by only supporting 11% and not funding the support of 89% of the rough sleepers over the next three years. Note well this is not a case of calling for 16,500 rough sleeper per year being supported but 16,500 being supported for a three year period. 

16,500 per year would be 16,500 in Year 1, then 33,000 in Year 2 and then 49,500 in Year 3 and each subsequent year yet the funding proposal is just for 16,500 every year of this 3 year period – a very different thing altogether as the numbers reveal, and even arguing against myself it would mean Crisis and their CSJ bedfellows believe 2 in every 3 of the 50,000 yearly rough sleepers England has each year can transition from the street to a tenancy of their own without the need for any support!  Those inclined to believe the right wing CSJ think tank’s hyperbolic nonsense over Housing First may wish to read the right wing Manhattan Institute view of Housing First effectiveness in the USA here which is scathing and argues:

 “…that proponents overstate the ability of Housing First to end homelessness, the policy’s cost-effectiveness, and its ability to improve the lives of the homeless.”

I am now being called a noted critic of the Housing First model by right wing commentators and academics – a more diplomatic term than the usual scaremonger-in –chief and other blame the messenger to deflect strategy. In 2002 I was a noted and publicly vociferous critic of the Supporting People programme and the direction it was taking in practical terms.  The theory of SP like the Housing First theory is fine in theory yet a dangerous nightmare in practice as has been proven with SP.  This resulted in central government offering me a wide-ranging position as an advisor to their SP programme with a very lucrative yearly consultancy contract that I flatly refused. It was an attempt to shut me up and instead have to promote the propagandist Labour government line on Supporting People and I continued advising scores of supported housing providers in this very niche area. I had to turn away 9 out of every 10 providers who approached my firm from our marketing budget of zero by challenging LA commissioning and decommissioning decisions. 

This relatively short 4000+ word article on the absurdly theoretical and unworkable Housing First model is not a pitch for business either it just needs to be said in broad overview to counter the hyperbolic propagandist nonsense of the Housing First model which is dangerous twaddle and closely resembles a cult.

The fears I had about SP all came true when the Labour government took away the SP ringfence – the only effective time support was anywhere near non-discretionary – and which led directly to the rough sleeper, homeless, and social care crises we have today.    The Labour governments SP ringfence removal was piloted in Liverpool, the only administering local authority who failed its Audit Commission inspection AND its reinspection to add perverse insult to absurd injury. 

It was also the first real version of doctrinaire Localism and this abjectly failed too and led to a national crisis seen for example in the Dementia Tax issue at the last before one general election.  Short –term cuts to preventative support services led to crises of care as they always will do and large scale cuts to subjectively perceived ‘non-deserving’ support services, such as the closure of 7,000 hostel beds in England since which has led to the crisis in rough sleeping and other homeless numbers we have today.

My fear that WHEN the Housing First model fails governments will say we funded to what the ‘experts’ of industry lobbies and academia said would work thus ending rooflessness / reducing homelessness is a lost cause governments will say and a case of throwing good money after bad will also happen IF the blind zeal for the unworkable Housing First model sees the ‘scale-up’ that its propagandists want.  It is the government ‘out’ for their realisation that the financial cost of eradicating rough sleeping is not the petty cash funding level they want to throw at it.  Ending rough sleeping will need multi-billions per year for the next ten years and government is not prepared to commit such funding levels for a mere pledge.

If something isn’t working expand it and no-one will notice or get blamed is an old saw or adage and the Housing First model today is the epitome of that adage.  Those who are roofless and the only visible manifestation of homelessness deserve far, far better than the unworkable Housing First model.