Making the Invisible Visible
We have based most of the Databeats’ articles on collecting new data ourselves. Sometimes huge amounts of data. We then set about finding the story in the data by using our skills in data analysis and visualisation. The data collection is an important starting place, but it’s the supporting actor to the analysis that tells the story in the final article.
But with this Databeats’ article collecting the data is the story, because one of the richest countries in the world does not know how many citizens are dying on its streets due to homelessness.
This lack of data is odd considering the UK government, like any government, has departments who spend a lot of money collecting all sorts of data they do nothing with. Collecting data on the deaths of the homeless is difficult, but it is important. Apparently the Office for National Statistics (ONS) is currently producing experimental data on the deaths of homeless people, which it plans to publish later this year. Let’s see.
Why is it important? It is difficult to plan and target services if we do not understand the scale of the problem, and the small numbers involved can become hidden within the wider health and care system, effectively meaning most deaths are ignored.
It has fallen to a charitable organisation with limited resources, the Bureau of Investigative Journalists, to provide the first UK-wide count. It has done this through a network of local journalists, charities and grassroots outreach groups.
When using data, particularly within a corporate setting, you hear a lot people discussing the need for establishing “one version of the truth”, and only using data for analysis where the data quality can be relied on. Well, the final dataset fails on both counts. This data hasn’t been counted consistently, and it’s riddled with missing or partial data. It’s certainly not ‘Big Data’ at just 449 rows — no fancy database or machine learning algorithm is required.
Yet none of this matters because behind these figures are 449 unique humans who have died over the course of a year; that’s over one death a day. From teenagers to people in their 90s. A gardener, an astrophysicist, musicians, fathers and mothers, sisters and brothers.
To quote Crisis Chief Executive Jon Sparkes:
[the families of these homeless people face] “the injustice that their loved one was forced to live the last days of their life without the dignity of a decent roof over their head, and a basic safety net that might have prevented their death. No-one deserves this.”
How have we got here? The formation of this crisis lies in ten years of government welfare cuts since the financial crash of 2008, expensive and selective private renting, and a severe lack of social housing.
The Databeats’ team have no glib solutions. The triggers that have been reported in the literature to explain homelessness, both immediate and long-term, are huge.
We just want to emphasise the simple power to collect data to give people a voice, to uncover injustice and start a debate. In an age of amazing technology to collect and store ‘Big Data’, the simple act of collecting data in a spreadsheet can be the vital first step.
There’s an urgent need to make the invisible homelessness deaths visible. So that, as Crisis Chief Executive Jon Sparkes says, “in one of the wealthiest countries in the world, there is no excuse for this tragedy to carry on.”
Methodology and Contextual Data
Caveats on the figures
The The Bureau of Investigative Journalism (BIJ) spent months working with local journalists, charities and grassroots outreach groups to come up with the first ever UK-wide count.
The BIJ have recorded deaths where they heard of them but the list is far from definitive — they know this is likely to be a large underestimate.
These are not only rough sleeper deaths. The definition of homeless includes rough sleepers, those in temporary accommodation like hostels and B&Bs, or those long-term sofa-surfing. In Northern Ireland it includes all those on the Housing Executive waiting list, who are officially classified as homeless and are waiting to be permanently housed. Most are in some form of temporary accommodation while they wait.
Why not a bar chart?
The visualisation in Tableau could indeed have been a bar chart. But showing every data point matters. Why? As data Journalist Mona Chalabi posits, there is no such thing as an emotionless data visualisation. Chalabi argues, “It’s important that the visualisation itself reflects the subject matter and not just the numbers.” Having 449 individual marks represents the subject matter and not just the number.
Contextual Data on Homelessness in England
The vizualisation below shows the number of people accepted (or rejected) as homeless by each local authority since 2012. Remember these are not solely people sleeping rough. The figures also includes those in temporary accommodation like hostels, B&Bs or those sofa surfing while homeless. Note, these are not yearly totals of all statutory homeless, it is a database of the yearly statutory homeless application decisions.
There has been an 8% rise in those rejected for homelessness since 2012 by local authorities, although the breakdown by individual Local Authority provides a more complicated, mixed picture across England. There is thinking that applications involving cases unlikely to result in ‘acceptance’ are increasingly remaining uncounted in the statistics
Analysis by Crisis, the homelessness charity, has shown that homelessness acceptances due to mortgage repossessions or social sector rent remain historically low. Statutory homelessness is now far more associated with ejection from the private rented sector.
The next visualisation looks at the current data for 2017. A number of local authorities are outliers in the data, although it’s unclear from this data why that it.