The Westminster Tradition

We talk with Eleanor Williams from the Australian Centre for Evaluation about making evaluation fast, rigorous, and genuinely useful for public service decisions. We move from a 40,000-person employment services RCT to the practical steps that make testing possible, then unpack why transparency and mixed methods matter when the stakes are high. Along the way we discuss:
turning a mutual obligation debate into a 'fact question'  through a large three-arm randomised controlled trial
why a null result can be a powerful policy signal
the enablers that make responsive RCTs feasible, from systems to leadership to ethics and consent
how ACE partners with departments through a hub-and-spoke model, secondments, and capability building
planning evaluation from day one using program logic and theory of change
separating program cost from evaluation cost to make smarter investment decisions
publishing and normalising evidence through the ACE Evaluation Library
using developmental evaluation and test-and-learn without getting stuck refining measures forever
what decision makers value under time pressure, including “killer stats” plus narrative and qualitative insight
improving services with small, fast trials like pre-call SMS nudges at Services Australia
Mentioned in this episode:
Australian Centre for Evaluation
ACE Evaluation Library, hosted on the Australian Policy Observatory
The Magenta Book
Anthony Costello, The Social Edge
This podcast was recorded on Kaurna land, and we recognise Kaurna elders past and present. Always was, always will be.
Now for some appropriately bureaucratic disclaimers....

While we have tried to be as thorough in our research as busy full time jobs and lives allow, we definitely don’t guarantee that we’ve got all the details right.

Please feel free to email us corrections, episode suggestions, or anything else, at thewestminstertraditionpod@gmail.com.

Thanks to PanPot audio for our intro and outro music. 

'Til next time!

Three-quarters of government evaluations are commissioned to improve implementation. Two-thirds are commissioned for accountability.
They're not the same thing, and building a system that serves both turns out to be harder than it looks.
This episode is about the structural tension between backward-looking accountability — did we deliver what we promised? — and the forward-looking question of whether we're actually learning anything useful.
Caroline argues the current measurement architecture is structurally bad at three things: it locks in categories at the moment of maximum ignorance, it treats evidence as binary when almost everything useful lives in the middle, and it has no way of capturing what you found out that you didn't know you were looking for.
There's also a detour through lifts, chin hairs, tattoo-level decisions, and Marie-Antoinette's diningware.
Referenced in the pod:
The APS State of Evaluation report (2026) — the list of why government commissions evaluations
Andrew Greenway's article on Sir Olly Robbins and The Accountability Act
Adam Kucharski's article Do we need more 'aluminium-standard' evidence?
Inside the Box: How Constraints Make Us Better by David Epstein
Tom Loosemore Universal Credit TWT episode 
This podcast was recorded on Kaurna land, and we recognise Kaurna elders past and present. Always was, always will be.
Now for some appropriately bureaucratic disclaimers....

While we have tried to be as thorough in our research as busy full time jobs and lives allow, we definitely don’t guarantee that we’ve got all the details right.

Please feel free to email us corrections, episode suggestions, or anything else, at thewestminstertraditionpod@gmail.com.

Thanks to PanPot audio for our intro and outro music. 

'Til next time!

Budget season is here — and it’s tighter than many public servants have been used to. Across Australian jurisdictions, governments are landing budgets with real constraints: restrained spending, savings packages, efficiency dividends, and the ever-popular one-in-two-out hiring freeze.
In this episode, Caroline, Alison and Danielle get into what that actually means for the people who have to keep delivering. How do you lead a team through savings without poisoning the culture? What’s the difference between a genuine moral injury and just disagreeing with a prioritisation decision? And is the blunt force of FTE targets and advertising bans the problem — or just a feature of a system that has tried everything else?
They also get into the real work: Danielle on essential vs lovely and what service redesign actually looks like when the envelope gets tighter, plus the Stockdale paradox, why constraints might be good for your team (sorry), and some practical advice for junior staff trying not to doom-spiral.
Budget constraint isn’t persecution. It’s elected officials doing what they were elected to do with taxpayer money. Suck it up, and here’s how.
Mentioned in this episode:
Brené Brown and Adam Grant on the Stockdale Paradox — The Curiosity Shop
David Epstein, Inside the Box: How constraints make us better 
Alfred Lansing, Endurance: Shackleton’s Incredible Voyage

This podcast was recorded on Kaurna land, and we recognise Kaurna elders past and present. Always was, always will be.
Now for some appropriately bureaucratic disclaimers....

While we have tried to be as thorough in our research as busy full time jobs and lives allow, we definitely don’t guarantee that we’ve got all the details right.

Please feel free to email us corrections, episode suggestions, or anything else, at thewestminstertraditionpod@gmail.com.

Thanks to PanPot audio for our intro and outro music. 

'Til next time!

In 1997, Tony Blair’s government inherited a problem: tuberculosis in cattle was rising, farmers were furious, and nobody agreed on whether badgers were responsible. The solution was to commission a gold-standard randomised control trial — 30 sites across the southwest of England, three conditions, run by an independent scientific group. Proper science. No cutting corners.

Eleven years and £49 million later, the trial produced findings that made things more complicated, not less. Reactive culling of badgers made TB rates worse. Proactive culling helped inside cull zones but increased rates in surrounding areas. Two expert panels reviewed the same data and reached opposite conclusions. And by the time the final report landed, the minister who commissioned the review had left office, the department had been restructured, and the politics had moved somewhere else entirely.
This episode is the first in our three-part series on evaluation in government. It’s not an argument against evidence — it’s an argument for being honest about what evidence can and can’t deliver, and what happens when government treats a long-run trial as a substitute for judgment rather than an input into it.
Also: there are quite a lot of badgers.
Referenced in the show 
Angela Cassidy’s book — Vermin, Victims and Disease
The New Yorker article on placebos in RCTs, The Power of Nothing by Michael Specter
Australian Centre for Evaluation’s paper on RCTs

This podcast was recorded on Kaurna land, and we recognise Kaurna elders past and present. Always was, always will be.
Now for some appropriately bureaucratic disclaimers....

While we have tried to be as thorough in our research as busy full time jobs and lives allow, we definitely don’t guarantee that we’ve got all the details right.

Please feel free to email us corrections, episode suggestions, or anything else, at thewestminstertraditionpod@gmail.com.

Thanks to PanPot audio for our intro and outro music. 

'Til next time!

Inspired by the new podcast The Curiosity Shop, Alison, Danielle and Caroline take on the things they might never agree on — welcome to TWT Fight Club. 
In the ring: 
Do academic and conceptual frameworks actually help public servants do their jobs, or are they a privilege that most people simply don't have time for?
Central agencies: great idea, but are they delivering? The trio debates whether they're connectors and coordinators — or arrogant secret-keepers who love a template.
Delivery units get their moment in the ring too, with strong views on the difference between a compliance-heavy traffic light report and genuine brokerage between agencies.
And the big one: would they go back to the public service? Caroline misses it in her bones, Danielle has a very petty list of reasons why probably not, and Alison is delighted to never write another bona fide.
Referenced in this episode: 
James Plunkett, The Centre is from Mars, the Edges are from Venus:
https://medium.com/@jamestplunkett/the-centre-is-from-mars-the-edges-are-from-venus-abca86f66bb8 
This podcast was recorded on Kaurna land, and we recognise Kaurna elders past and present. Always was, always will be.
Now for some appropriately bureaucratic disclaimers....

While we have tried to be as thorough in our research as busy full time jobs and lives allow, we definitely don’t guarantee that we’ve got all the details right.

Please feel free to email us corrections, episode suggestions, or anything else, at thewestminstertraditionpod@gmail.com.

Thanks to PanPot audio for our intro and outro music. 

'Til next time!