——— Work

Three classes of problem. One move.

The three classes below are not topics; they are the three ways a problem can be hard — each the case where one dominates. Public policy is where complexity rules: many actors and levels, none holding the whole. Problems without a map are where uncertainty rules: no model exists, and the terms of the problem are unsettled. Problems that resist solutions are where divergence rules: there is often no agreement that a problem exists at all. Every real problem carries all three; naming the one that dominates is where the work begins.

How I read them

In each case, the move is the same: the problem is usually a missing architecture, and the criterion for building it usually comes from what the whole exists for. From that criterion follows the architecture, and only then the instruments that make it work.

a. Public policy

Designing national policy is a problem of high complexity: it demands integration rather than addressing a single topic, bringing together levels, actors, and evidence that rarely meet. Chile’s upcoming National Mental Health Plan is designed so that monitoring flows from its architecture rather than being added after the plan is written. That architecture distills the long tradition of mental health plans into three levels: how a country understands mental health and its care; how care happens across all settings, both health and intersectoral; and what defines good clinical care. These are not three Chilean choices — they are the questions any mental health policy must answer, in any country, and each country decides its answers. What you can’t do is skip the questions. Ordering Chile’s tradition is where I found them, not where they end.

Building it meant bringing together bodies of work that rarely intersect: the literature on policy design, a theory of change borrowed from impact evaluation, and a participatory process placed on equal footing with scientific evidence and international standards.

b. Problems without a map

When no model exists yet, one has to be built. Ordering indicators is a recurring case: a system can hold a great deal of information and still be unable to tell whether it is doing what it sets out to do.

I designed the monitoring system for the National Mental Health Action Plan 2019–2025. The problem was not the indicators but the lack of a standard to hold them together: I brought a large, highly varied set into a single, legible whole that was workable within the everyday reality of the ministry’s teams. That work convinced me that the next plan had to be built, from day one, with its monitoring and evaluation logic already embedded in its architecture.

Brought in to “measure the impact” of the Universidad San Sebastián’s third-mission work, I reframed the request: the problem was not measurement but a missing architecture to organize it. I built it from three bodies of theory developed for different purposes, and the university carried it forward on its own, publishing its own model that credits this initial work. Since 2025, Chilean accreditation has required impact evaluation of every university. The work anticipated it [Read the full case].

Earlier still, the innovation institute iCubo at Universidad del Desarrollo asked whether students passing through its program became more innovative — a question with no model to answer it and no certainty it could be answered at all. We built one and found that they did — and that the more diverse the team, the more innovative its work. That finding became a method for composing interdisciplinary teams and the starting point for everything I later did in diversity management.

c. Problems that resist solution

These are not topics but a class of problems: the disagreement even extends to what the problem is, and the institution has no language to hold it — so it holds nothing. My work is to give it that language. It is almost always the language of quality of care.

Stigma toward sexual and gender diversity. Here, the divergence runs deepest: there is often no consensus that stigma toward LGBTI+ people is a problem the health system should address. I designed Chile’s national strategy for mental health care related to sexual and gender diversity in the public system, reframing stigma not as a contested rights debate but as a technical failure in care quality and patient safety — the terms on which a health system can be made to act.

The end of life. Accompanying someone as they die runs counter to centuries of health culture. I contributed the psychological dimension to Chile’s national palliative care guidance — the reframing that enables a system to care for a life it can no longer save.

Where the method was tested

Fifteen years of teaching one subject at three levels of depth.

Cultural competence. How culture shapes what people think, feel, do, and how they interact, and how that occurs under oppression and power asymmetries.

Managing diversity. The same applies within a working group, where differences improve solutions but also create conflict. In interdisciplinary settings, disciplines behave like cultures, each with its own language, goals, and tools.

Complexity. Diversity is not a virtue — it is what a problem demands when it exceeds the capacity to solve it.

Each level came from what I was doing at the time, and the classroom was where the models were tested before they were written.