One of us has been banging on about this so often and for so long that it has been termed Worstall's Fallacy. The point being that we cannot decide how much more we must do about something until we account for what we already do about that thing.

For example, the American poverty numbers are calculated before taking account of much of what is done to reduce poverty. Only cash income is considered. But much American poverty alleviation is delivered through the tax system (the EITC) or goods and services in kind (Medicaid, Section 8, Snap and so on). The published official poverty rate is thus around 14% of the population, the number after alleviation more like 4% or so.

We cannot then go around shouting that 14% in poverty is terrible thus we must spend more on the EITC, Snap, Medicaid and Section 8. It may well be true that more should be spent - possibly even less - but the relevant number to be looking at is the 4%, after the effects of what is already done.

And so it is with Shelter's estimations of how many homeless people there are in the UK. They tells us there are 250,000 unfortunates in that position:More than a quarter of a million people are homeless in England, an analysis of the latest official figures suggests.
...

For the very first time, Shelter has totted up the official statistics from four different forms of recorded homelessness.

These were:

national government statistics on rough sleepers
statistics on those in temporary accommodation
the number of people housed in hostels
the number of people waiting to be housed by social services departments (obtained through Freedom of Information requests)

Temporary accommodation, hostels, those are things we do about homelessness as most of us understand the word - that rough sleeping. This is to commit Worstall's Fallacy, to count as not yet done the very things we are doing to solve the problem.

It's entirely true that hostel living isn't as desirable as a safe and secure home to call one's own. But it's an entirely viable solution to rough sleeping, or homelessness.

The importance of the fallacy is that only by proper measurement can we work out what the next step in our solution should be. Absent serious mental health or addiction problems (and the occasional expression of choice) the problem of rough sleeping is largely solved. For we do have those hostels and people don't have to sleep on the streets, something that was not true say two centuries ago.

What we don't have is those safe and secure homes that all can call their own. And thus that is the problem that should be concentrated upon. And fortunately we know the solution there, blow up the Town and Country Planning Act and successors.