We can’t build high-impact systems on underpriced expectations 

The Centre for Excellence in Child and Family Welfare has been doing some thinking about the future for services supporting children, young people, families and carers. This piece is part of a series aimed at sparking conversations about opportunities to innovate together and create the change we want to see. 

Overheads have long been treated as a necessary evil in social services: something to reduce, conceal, or apologise for. The myth runs deep that real value lies in the ratio that low administration equals high impact, and that every dollar not spent directly on “the work” is waste. 

But the story doesn’t hold, and it’s doing real damage.

The same patterns keep surfacing in conversations across the Centre’s member network and with stakeholders: organisations are expected to scale, innovate, and professionalise while keeping core costs flat. Services face demand for robust evaluations, without funding for data infrastructure. Hope for improved outcomes, without investment in the workforce. Drive to ramp up early intervention, with budgets that sustain little more than frontline delivery.   

This isn’t about bad intent – rather, about tired assumptions.

Somewhere along the line, an accounting term was mistaken for a performance metric. What counts as “overhead” often includes the very things that enable a service to function: quality systems, secure data storage, staff development, rent, insurance, professional supervision, reporting tools, policy expertise, administrative coordination, fundraising, translation, design, and increasingly, digital infrastructure. These are the mechanisms through which services become safe, responsive, and effective.  

Yet, the obsession with driving down overheads persists.  

Funders speak of “realistic unit costs” but rarely cite the data behind them. The result is a scarcity loop, where services are locked into underinvestment, high turnover, and compliance regimes that favour process over purpose. The incentives reward short-term output over long-term capability, and they make it harder to direct funding to what works. Worse, they normalise fragility, embedding instability into the systems services rely on to deliver impact. This is a dynamic long recognised in what the Stanford Social Innovation Review called the “nonprofit starvation cycle”. 

This isn’t a new insight. A 2013 TED Talk with nearly six million views forced the issue into the open, arguing that a double standard had taken hold: one rulebook for business, another for charities. Globally, charity watchdogs have since walked back their emphasis on overhead ratios. In Australia, the conversation has been slower. Efforts like the Pay What It Takes initiative have started to shift the frame, but much of the system still defaults to old ways of thinking. 

Being more transparent might help. There’s growing evidence that when organisations go beyond basic financials and share how operational investments support impact, supporters respond positively. But transparency alone won’t resolve a flawed funding logic, and it shouldn’t fall to providers to continually prove their legitimacy in terms that don’t reflect reality. 

It may be time to ask better questions, instead. 

What does it cost to run a high-performing intervention service? How do those costs shift by region, cohort, model, or maturity? What infrastructure is essential, and what would it take to build it collectively? 

The sector needs a clearer baseline: a shared view of the real costs behind the work – not as a compliance tool, but as a foundation for smarter investment and fairer design. A baseline that captures not only delivery costs, but the overheads tied to safety, continuity, and quality. A standard that reflects the actual drivers of impact, not just the line items that are easiest to explain. Because until we specify what’s required, we’ll keep designing systems that pretend those needs don’t exist.  

So, before the next round of reforms or procurements, we need to be bold enough to ask ourselves: If we continue to underwrite impact, what are we really measuring? And who gets to decide what’s worth the cost? 

Related articles

Last updated: 13 Jan 2026