Panarchy: The Level We Weren't Looking At

Thirty-seven seconds left in the breakout, Tyson Heaton said what several of us were thinking: “The challenge everyone’s having with this exercise, including me, is like, what’s the point?” He said it while producing the most useful thing anyone did all morning.

What we were trying to do

On August 13, thirteen of us spent ninety minutes on Zoom as pre-work for the California Improvement Summit. The question is awkward on purpose: how do we improve improvement in California? Not fix healthcare or housing — improve the capability.

Improvement people like to focus, so the pre-work kept snagging: on what? Ten arenas came out of session one, no two people naming the same. Fernando Murray and Carolina Almeida, Liberating Structures practitioners on Starlink near Sequoia, suggested we stop choosing.

The idea, in one paragraph

Levels move at different speeds. Down low things change fast; up high they move slowly, conserving what worked. Fernando’s example, from Keith McCandless: MRSA. The bacterium, a clinician’s handwashing, hospital practice, Medicare policy — and above it all, the belief that superbugs are inevitable. So the question isn’t “where is the problem?” but which levels have had our attention, and which are neglected?

Panarchy Level Framework

What happened

We dot-voted. Business competitiveness and housing tied at seven, higher education six. Three groups worked two steps: list factors, name levels.

The myth level did the most work. Mapping housing, we landed on beliefs nobody would put on an agenda: high-density living is un-American, a house is the middle class’s retirement plan. Janice Hodge named the heaviest — “the kind of individualism that if you can’t afford it, it’s somehow your fault.” Fifteen minutes later, permitting queues, where Tyson found the neglected lever: “The county throws theirs in, the state throws theirs in, the city throws theirs in, and it doesn’t move.” Beliefs and queues in one conversation — the span the structure exists for.

What the structure handed back

Two of us said mid-exercise we couldn’t see the point. Kristin Kielich named it afterward: “It was the What, the So What. And now I’m thinking the Now What?” Panarchy delivers “so what”; “now what” lives in the distinction between how we do improvement and where we apply it — the session’s key takeaway.

One group saw it coming. Mark Rosenthal and Michael Letson spent their breakout on the headwinds against improving anything — each level owned by a different faction, “and the crossover from one level to the next is a really high wall.” The sharper challenge came in the margins of this draft. Mark: “They are all places where improvement could be applied, but none of them are about HOW WE DO IMPROVEMENT, which is the core question on the table here.”

He is right, and it may be the most useful thing the session produced. We ran an experiment in exploring three arenas; the answer is that the arenas are not the subject. The improvement process is. Then he separated the two:“‘How we do it?’ and ‘Where we do it?’ are both legitimate, but separate, questions, so it is important to be clear which one we are trying to answer in any given discussion.”

Mark also pressed a question from the chat that nobody answered: who is “we”? My initial answer — everyone I talk to — was naive, as I said at the time. And he clarified that an “engagement,” as a noun, is the entire interaction of bringing improvement into an organization, distinct from the common plea that leaders need to engage: change agents trying to get line leaders to own the work rather than hand it to the CI team and expect results without changing how they lead. Their standing question: “How do I get the leaders to engage?” His diagnosis: “practitioners focus on the technical aspects of the physical process, and not so much on the ways humans have to interact with that process.”

And the constraint I keep turning over: “Whoever your target is, part of the solution is giving them the ability to more effectively engage the environment they’re faced with.” A Panarchy insight arriving without the chart.

What we are carrying to August 30

Steven Mummy, on the myth level aimed at the practitioner rather than the system: “If you’re not looking at the dogmas and myths you carry around, and challenging them, you’re part of the problem.”

Two questions to keep apart — how we do improvement, and where we apply it — and a commitment to name which is live. A force field analysis in the current-state phase, aimed at the headwinds Mark and Michael mapped. And the “they,” still undefined, now in the A3’s first box.

Come argue with us

The California Improvement Summit runs August 30 through September 1 at the Octagon Barn in San Luis Obispo, and three other pre-work options fed into it. If you have run Panarchy, or watched a good improvement effort die a level above where it was working, we would rather argue with you than agree with each other. The full record is in the Option 4 detailed report.

“Option 4 detailed report” → )

Knowledge Map

Process keywords

Panarchy; Liberating Structures; nested Ecocycles; levels of scale; fast–slow dynamics; myth level; identify factors; translate factors to levels; neglected levels; cross-level leverage; Social Network Webbing; dot voting.

Context keywords and reader pain points

Improvement efforts that stall above the team level; picking a focus when nobody agrees; systems too big to scope; sponsorship that disappears with a leadership change; workshops that produce maps but no action; practitioners stuck inside their own domain.

Application triggers

If your improvement work keeps hitting a barrier owned by someone at another level, then map the levels before proposing a countermeasure. If a group cannot agree what to focus on, then let people self-select an arena and compare maps rather than negotiating one. If a belief keeps reappearing as a constraint, then name it as a level, not an attitude. If a policy layer is crowded with activity and nothing moves, then look one level down for the queue nobody owns. If participants ask “what’s the point” mid-exercise, then frame the payoff earlier next time — and keep going.

Related continuous-improvement themes

Technical versus adaptive change; going to the gemba across organizational boundaries; process thinking in public administration; the limits of tool-first improvement; scaling improvement beyond a single organization.

Developed from the Zoom recording and chat, the Miro board, and five AI note-taker records — including the only one that captured the after-party — and synthesized with Claude AI assistance, reviewed by a human. Quotations lightly cleaned from automated transcripts. Editorial contributions by Mark Rosenthal. Published by Central Coast Lean. Corrections and pushback: eric.o@centralcoastlean.org

eric.o@centralcoastlean.org→