As problems spread and involve more moving parts they generally do become harder to solve, though the reason is usually due to complexity rather than size alone.

Indeed many massive, complex problems don’t become unsolvable because of the problem itself. They become unsolvable because of the scaffolding we build around them.

When a social crisis emerges, our instinct is to build support systems: initiatives, funding streams, non-profit networks, and oversight agencies. In theory, this scaffolding is temporary infrastructure designed to hold a collapsing system together while we fix the problem. But over time, something else happens. Scaffolding becomes an industry. It develops its own bureaucracy, its own executive payrolls, and its own political constituency.

Eventually, the scaffolding stops being a temporary fix and becomes a permanent, self-reinforcing ecosystem. Many of the systems and institutions of today were originally envisaged as temporary safety nets.

How Los Angeles Built a $2.4 Billion Non-Solution

Skid Row in Los Angeles is the largest homelessness encampment in the USA accommodating 5,000+ people. Surely it should be relatively simple for the world’s biggest economy to end street homelessness?

What began as an urban management plan in 1975 mutated over five decades into a massive, entrenched “homelessness industry.” Surrounding neighbourhoods, state prisons, and regional hospitals routinely dumped vulnerable people directly into Skid Row, knowing the scaffolding was there to catch them.

When the crisis exploded, the response wasn’t to rethink the model—it was to throw eye-watering sums of money into expanding the scaffolding. In 2016, voters approved Proposition HHH, a $1.2 billion housing bond. But instead of housing people efficiently, the massive administrative web swallowed the outcomes:

• Astronomical Cost Escalation: Building a single unit of supportive housing soared from an estimated $375,000 to between $600,000 and $837,000, dragged down by 3-to-6-year development timelines and complex funding patchworks.
• Administrative Paralysis: An independent audit revealed that the region’s main agency, the Los Angeles Homeless Services Authority (LAHSA), could not accurately track $2.4 billion in spending. Millions were paid to vendors without verifying service delivery, while contracts lacked basic performance requirements.

When a problem gets big enough, the ecosystem of providers, auditors, and bureaucracies becomes part of the problem. The crisis isn’t solved—it just gets permanently institutionalised.

Breaking Down the Scaffolding

If massive, centralised scaffolding leads to stasis, how do we dismantle it? You don’t build a bigger machine; you break the problem back down into digestible, human-sized chunks.

I tried for a long time to solve big problems through small experiments, but soon realised the scaffolding itself was the barrier.

In 2016 Bromford, one of BFL’s three founding organisations, replaced its traditional housing officers with neighbourhood coaches. Patch sizes fell from 600–800 homes to 175–200, so that each coach could know every customer on their patch and the community around them: the school, the health centre, the PCSO, the youth group. The role was deliberately generalist. Coaches carried out most housing management functions themselves. They took a strengths-based approach, working with residents as people with agency and aspirations rather than as a set of needs to be managed. It became the cornerstone of Bromford’s first approach to place.

What we did not change was the system around the coach. Repairs, income, lettings and investment stayed organised by function, each with its own targets, budgets, reporting lines and data.

We had put a new part into an old system and expected the system to behave differently.

It did not, and the limitation lay in the design, not in the coaches. An organisation’s behaviour is shaped less by any single role than by the connections between roles: the targets people answer to, who holds budgets, who signs decisions off and what information reaches whom. When a role changes but those connections do not, the system gradually pulls the role back to its previous shape.

Attempts to bring the Dutch Buurtzorg model of self-managing neighbourhood nursing into the NHS show the same pattern. Independent evaluations of the pilots found that many patients preferred the new teams to their previous experience of district nursing. The organisations around the teams did not change. In Tower Hamlets, nurses had clinical autonomy but little control over operational decisions such as the budget. Caseloads at times rose well above the level the model relies on, and some staff reported that the pilot’s high profile brought more oversight, not less. The pilot ended in 2018. The teams worked; the systems around them did not change.

Designing for 2031

Our own move to around 121 semi-autonomous teams is therefore not a reorganisation of frontline services. It depends on changing the functional system itself: how targets are set, how budgets are held, how line management works, how data flows and where decisions are made. Some of that system has to be dismantled, not adjusted. Wave 1 went live in summer 2026. That is the start of structural deployment, not evidence of a mature model. Our strategy to 2031 is the beginning of this transformation, not its completion.

The move to place also carries a risk of its own. Devolving decisions raises legitimate concerns about consistency, compliance and visibility. The instinctive response is to add management: more oversight roles, more reporting. Without discipline, that would burden the new system with controllers rather than enablers. Place Teams would have the language of autonomy and the daily experience of supervision.

This is more than a management tweak—it is a systematic technique for slowly removing the heavy administrative scaffolding.

We cannot solve our most stubborn social challenges using the same centralised, bureaucratic machinery that profits from managing them.

Systems don’t leave of their own accord; they have to be integrated, mitigated against or removed.

Place-based working and decentralisation isn’t a solution, it’s a bet that you can make things small enough to be solvable.

Dismantle the scaffolding, solve the problem.

Photo by Brandi Alexandra on Unsplash