What has tackling homelessness and rough sleeping got to do with testing jet engines? Not much, you might say. But for one highly specialist homelessness scholar these radically different tasks have key things in common. “They both have complex systems, are hard to understand, and can benefit from simulations or laboratories to test interventions,” says Patrick Fowler, a professor at Washington University, in St. Louis, Missouri, in the United States. 

Jet engines have long been put through their paces with simulations. But using a virtual laboratory to tackle homelessness and rough sleeping is a relatively new idea. Prof Fowler is one of very few scholars with expertise in the field and has devised such a simulation to test homelessness interventions as part of an ambitious project being co-ordinated by the Centre for Homelessness Impact.

Prof Fowler has used a “stocks and flows” model to simulate the homelessness system in England with all its complexities in a virtual environment. The model’s building blocks are boxes, pipes, and valves. Put together, these bits of computerised plumbing can track the passage of large numbers of households through the twists and turns of a virtual homelessness system. It measures the speed at which people “flow” through the system and pinpoints any bottlenecks. Crucially, it can also assess the changes a new intervention might make to the virtual passage of homelessness populations through this system and whether it makes any difference to where they get stuck.

The stocks and flows model has been fed with reams of data on homelessness and rough sleeping from the Government official bank of statistics and other reliable sources dating back to 2018 as part of a systems-wide evaluation of homelessness and rough sleeping led by CHI and commissioned by the Ministry of Housing, Communities and Local Government. This will help deepen understanding of the impact of homelessness interventions across the system.

“The data collected by the  Ministry of Housing, Communities and Local Government has been really useful for this simulation,” Prof Fowler says. “With any simulation you are trying to reliably say what is going on in a system. All this historical data allows us to test whether any experiments we do in 2025 line up with how the system actually worked in 2018.”

Prof Fowler’s simulation has already been put to the test by assessing the impact of several theoretical interventions on his virtual homelessness system.

Ultimately, Prof Fowler hopes the model will help the UK Government to work out its best plan for tackling homelessness and rough sleeping. But he is keen to stress that these kinds of simulations have limits. “They are not oracles; they can’t predict the future”, he says. “But they can be very useful when planning an intervention. They can help you to understand the trade-offs, unintended consequences, and missed opportunities.”

Prof Fowler hopes the CHI project will help the UK shift towards more preventative services and so slow the relentless flow of people needing help with rough sleeping and homelessness. It’s an overwhelming pressure he witnessed himself as a therapist in a homelessness shelter in the early 2000s.

“I was doing therapy in homelessness shelters with 130 people in the same room all dealing with their own challenges. They would show up for one session and not the next. It wasn’t their biggest problem. I wanted to help them but became acutely aware of my limitations,” Prof. Fowler says.

“It was a pivotal moment when I realised I was able to do more by working in public policy and on these stocks and flows models,” he adds. “I want to prevent people from drowning in rivers by helping them further upstream or stopping them from jumping into rivers at all.”

Keith Cooper is a freelance journalist