![\"Map1\"](\"/assets/20250618_econ_tai.png\")
Danny Buerkli, 11 May 2026
What do we know about the current and future impact of AI on labor markets? Quite a bit, it turns out.
Every few weeks I put together a deck with some of the latest research that’s come out.
Think of it as Mary Meeker’s AI report, but for labor market impacts. In other words: a fun chartapalooza that pulls together some of the best research out there.
Download the PDF or read the online version.
", "url": "https://www.dannybuerkli.com/2026/05/11/what-we-know-labor-market-impact-ai", "image": "/assets/cards/essays/what-we-know-labor-market-impact-ai.png", "date_published": "2026-05-11T00:00:00-03:00", "date_modified": "2026-05-11T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2025/06/18/econ-impacts-of-tai", "title": "Understanding the Economic Impacts of Transformative AI: A Guide for the Perplexed", "summary": "A guide for the perplexed: how the major analyses of transformative AI's economic effects relate to one another, and where they genuinely disagree.", "content_text": "Danny Buerkli, 18 June 2025It is becoming increasingly evident that AI systems may have disruptive effects on labor markets and the wider economy. Despite this, the economic impacts of transformative AI remain oddly understudied. While there are disagreements over the pace of change and the magnitude of these effects, it is clearly worth taking these questions seriously.Over recent months, we have seen a surge in thoughtful publications on the topic. Given the breadth of this emerging field, however, it can be difficult to understand how different analyses relate to one another.This “guide for the perplexed” aims to highlight some of the most significant contributions, structured around a simple framework.Mechanical birds when?To make sense of the complex economic effects of transformative AI, it helps to break down the issue into three central questions: How long will it take us to see transformative AI? What will the economic effects be? What should we do about it?Additionally, two further questions will likely become increasingly important: What are the politics of this? How do governments need to change?Let’s explore each question in detail.How long will it take us to see transformative AI?A key concern is how much time we all have to prepare for a future with transformative AI. Debates over timelines are highly contentious. This section highlights arguments that clarify the core disagreements between competing perspectives.Two recent articles illustrate two starkly contrasting visions:“AI 2027” (April 2025), by Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland, and Romeo Dean, concludes that we may well see runaway superintelligence by the end of 2027. This enormous impact within a very short time frame, leaving almost no room for intervention, makes AI fundamentally different from other technologies we’ve had to contend with.“AI as Normal Technology” (April 2025), by Arvind Narayan and Sayash Kapoor, argues the opposite case: AI systems will take a long time to develop and diffuse through the economy. AI is, in other words, a powerful but fundamentally mundane technology. All previous lessons we have learned about handling technology still apply.The core disagreement between the two teams revolves around whether AI research itself will be automated. This would lead to rapidly self-improving AI systems.The AI 2027 team believes this to be not only possible but in fact likely. Arvind and Sayash don’t think so.Arvind and Sayash additionally believe that there exist important bottlenecks that will constrain AI diffusion. The AI 2027 is not materially concerned with diffusion, since a superintelligent system would be so capable that it would blow past any hurdles immediately.A helpful deep dive into the self-improvement aspect is presented in “Three Types of Intelligence Explosion” (March 2025) by Tom Davidson, Rose Hadshar, and Will MacAskill.Unfortunately, there isn’t yet a similarly detailed exploration of bottlenecks to AI diffusion. Tyler Cowen, in “Why I think AI take-off is relatively slow” (February 2025), comes closest and outlines reasons why AI take-off may take a long time.What will the economic effects of transformative AI be?Once transformative AI does diffuse through the economy, what will its economic impacts look like?This question is hard to reason about because it refracts immediately into a vast set of interlinked issues of wages, power, identity, social relations, and more. As a result, many of the articles mentioned below treat the economic effects of transformative AI as one question of concern, among others.We can group existing work into two categories: broad-stroke analyses outlining general economic and societal shifts, and more targeted investigations addressing specific economic questions.Let’s begin with the broad stroke analyses:Anton Korinek is one of the few macroeconomists to explicitly address transformative AI. In “Scenarios for the Transition to AGI” (March 2024) Anton and Donghyun Suh model potential trajectories for economic output and wages under different AGI scenarios. Later, in “Economic Policy Challenges for the Age of AI” (September 2024), Anton considers future roles for human labor and identifies production factors likely to grow in importance post-AGI.In a somewhat similar exercise, Epoch AI has published GATE, a competing macroeconomic model. The model, for example, predicts significantly accelerated economic growth and concludes that the “global economy can marshal enough effective compute to automate most tasks within two decades.”In their “Intelligence Curse” (April 2025) essay series Luke Drago and Rudolf Laine draw parallels to the resource curse. They argue that powerful entities may lose incentives to invest in ordinary people, similar to how many resource-rich states today neglect their populations. Their analysis highlights early warning signals, potential shifts in power structures, the erosion of existing social contracts, and possible interventions.“Gradual Disempowerment” (January 2025), by Jan Kulveit, Raymond Douglas, Nora Ammann, Deger Turan, David Krueger, and David Duvenaud takes a cautionary stance, suggesting that even in the absence of an intelligence explosion incremental increases in AI capabilities “pose a substantial risk of eventual human disempowerment”. Their economics chapter outlines how the human labor share of GDP might gradually asymptote towards zero, with attendant disempowering consequences.The Labor Market Risks chapter of the International AI Safety Report 2025 (February 2025) provides a concise, extensively referenced overview of the current evidence. As a scientific report, it remains cautious about speculative scenarios but offers a robust foundation for understanding labor market impacts.Now let’s look at the more focused explorations:In “What Will Remain for People to Do?” (March 2025) Daniel Susskind examines the effect of transformative AI on jobs, starting from the premise that AI could eventually perform all economically valuable tasks more productively than humans. His relatively optimistic conclusion is that comparative advantages, preferences for human-led processes, and normative considerations will preserve certain tasks for human workers.Similarly, Noah Smith, in “Plentiful, high-paying jobs in the age of AI” (March 2024), looks closely at comparative advantage. He argues that as long as compute is finite and not literally free, the use of AI systems comes with an opportunity cost, ensuring some tasks remain economically viable for humans.Finally, what do the leaders of leading AI labs say? While you may choose to not take what they are writing at face value, it is still instructive.In “Machines of Loving Grace” (October 2024) Dario Amodei, the CEO of Anthropic, sketches “what a world with powerful AI might look like if everything goes right.” He argues that 20% annual GDP growth is achievable in the developing world.Sam Altman, CEO of OpenAI, in “The Gentle Singularity” (June 2025) hedges his bets but writes that “2026 will likely see the arrival of systems that can figure out novel insights. 2027 may see the arrival of robots that can do tasks in the real world.”What should we do about the economic effects of transformative AI?Given the potentially profound economic changes caused by transformative AI, how should societies respond? Recent discussions focus on reforming our social and economic institutions to meet these challenges.Deric Cheng is currently editing an anthology titled the “AGI Social Contract” (June 2025). In the anthology’s introductory piece, he summarizes the state of play and outlines key questions for a new social contract.As part of this anthology, “The Missing Institution: A Global Dividend System for the Age of AI” (June 2025), by Anna Yelizarova proposes a global dividend system that would share the proceeds from concentrated AI wealth.This suggestion belongs to a broader set of proposals known as “benefits sharing,” which explores mechanisms for the fair distribution of AI-generated wealth.An earlier and influential proposal along these lines is the Windfall Clause, which describes an “ex ante commitment by AI firms to donate a significant amount of any eventual extremely large profits garnered from the development of transformative AI.” In their paper from January 2020 Cullen O’Keefe, Peter Cihon, Ben Garfinkel, Carrick Flynn, Jade Leung, and Allan Dafoe outline the mechanics of the idea. While the Windfall Clause as initially conceived has somewhat fallen out of favor, it remains influential as a foundational concept.Building on critiques of the Windfall Clause, “Predistribution Over Redistribution” (October 2024) by Saffron Huang and Sam Manning argues that we should focus on early structural interventions that prevent extreme wealth concentration before it occurs, rather than trying to redistribute wealth after the fact.Taking a macroeconomic perspective, Era Dabla-Norris at the International Monetary Fund (IMF) published a staff discussion note titled “Broadening the Gains from Generative AI: The Role of Fiscal Policies” (June 2024). This paper outlines fiscal tools and policies that governments could employ to harness generative AI’s benefits, mitigate negative labor market impacts, and promote fair distribution.What are the politics of the economic effects of transformative AI?Politics fundamentally is about “who gets what, when, and how.” Under any scenario the impacts of transformative AI are unavoidably political. So it is surprising that there has been relatively little analysis of the political dimensions.A notable exception is the recent work by Anton Leicht. In his essays “AI and Jobs: Politics without Policy” (June 2025) and “Three Fault Lines in Conservative Thinking” (May 2025), Anton offers helpful context on the policy landscape, especially in the US. These essays explore some of the structural and political challenges that will shape how the political debate about transformative AI might play out.How do governments need to change?If transformative AI arrives, it stands to reason that governments and public administrations will need to adapt significantly, not just in terms of specific policy decisions but also in their fundamental operations and structures. While the preceding sections considered which policies might be necessary, this section asks: how should government institutions themselves change?One of the few pieces to explicitly look at this question is Ed de Minckwitz’s “Government in the Age of Superintelligence” from June 2025. In his report for Policy Exchange, the London think tank, he looks at the impact of AGI on healthcare, education, infrastructure, defense, and politics.To close, four reflections:1) I have intentionally stacked the deck and sidestepped the question of whether or when we will have transformative AI. Instead, I have focused on asking what will happen once transformative AI is a reality.2) The rate at which thoughtful analyses and proposals are coming out is already high and will only increase. Specific articles and reports highlighted above may well go stale but the overall framework will hopefully remain useful.3) This guide is necessarily woefully incomplete. There is far more excellent work than can be featured in an overview like this.4) Much more thinking and preparation is urgently needed. Given the scale and importance of transformative AI’s potential impact on jobs and economies, the quantity and depth of thinking and policy preparation ought to increase fast.Thanks to my Windfall Trust colleagues Deric Cheng, Helen Stevenson, Anna Yelizarova, and Adrian Brown for their helpful comments.", "content_html": "Danny Buerkli, 18 June 2025
It is becoming increasingly evident that AI systems may have disruptive effects on labor markets and the wider economy. Despite this, the economic impacts of transformative AI remain oddly understudied. While there are disagreements over the pace of change and the magnitude of these effects, it is clearly worth taking these questions seriously.
Over recent months, we have seen a surge in thoughtful publications on the topic. Given the breadth of this emerging field, however, it can be difficult to understand how different analyses relate to one another.
This “guide for the perplexed” aims to highlight some of the most significant contributions, structured around a simple framework.
To make sense of the complex economic effects of transformative AI, it helps to break down the issue into three central questions:
Additionally, two further questions will likely become increasingly important:
Let’s explore each question in detail.
A key concern is how much time we all have to prepare for a future with transformative AI. Debates over timelines are highly contentious. This section highlights arguments that clarify the core disagreements between competing perspectives.
Two recent articles illustrate two starkly contrasting visions:
“AI 2027” (April 2025), by Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland, and Romeo Dean, concludes that we may well see runaway superintelligence by the end of 2027. This enormous impact within a very short time frame, leaving almost no room for intervention, makes AI fundamentally different from other technologies we’ve had to contend with.
“AI as Normal Technology” (April 2025), by Arvind Narayan and Sayash Kapoor, argues the opposite case: AI systems will take a long time to develop and diffuse through the economy. AI is, in other words, a powerful but fundamentally mundane technology. All previous lessons we have learned about handling technology still apply.
The core disagreement between the two teams revolves around whether AI research itself will be automated. This would lead to rapidly self-improving AI systems.
The AI 2027 team believes this to be not only possible but in fact likely. Arvind and Sayash don’t think so.
Arvind and Sayash additionally believe that there exist important bottlenecks that will constrain AI diffusion. The AI 2027 is not materially concerned with diffusion, since a superintelligent system would be so capable that it would blow past any hurdles immediately.
A helpful deep dive into the self-improvement aspect is presented in “Three Types of Intelligence Explosion” (March 2025) by Tom Davidson, Rose Hadshar, and Will MacAskill.
Unfortunately, there isn’t yet a similarly detailed exploration of bottlenecks to AI diffusion. Tyler Cowen, in “Why I think AI take-off is relatively slow” (February 2025), comes closest and outlines reasons why AI take-off may take a long time.
Once transformative AI does diffuse through the economy, what will its economic impacts look like?
This question is hard to reason about because it refracts immediately into a vast set of interlinked issues of wages, power, identity, social relations, and more. As a result, many of the articles mentioned below treat the economic effects of transformative AI as one question of concern, among others.
We can group existing work into two categories: broad-stroke analyses outlining general economic and societal shifts, and more targeted investigations addressing specific economic questions.
Let’s begin with the broad stroke analyses:
Anton Korinek is one of the few macroeconomists to explicitly address transformative AI. In “Scenarios for the Transition to AGI” (March 2024) Anton and Donghyun Suh model potential trajectories for economic output and wages under different AGI scenarios. Later, in “Economic Policy Challenges for the Age of AI” (September 2024), Anton considers future roles for human labor and identifies production factors likely to grow in importance post-AGI.
In a somewhat similar exercise, Epoch AI has published GATE, a competing macroeconomic model. The model, for example, predicts significantly accelerated economic growth and concludes that the “global economy can marshal enough effective compute to automate most tasks within two decades.”
In their “Intelligence Curse” (April 2025) essay series Luke Drago and Rudolf Laine draw parallels to the resource curse. They argue that powerful entities may lose incentives to invest in ordinary people, similar to how many resource-rich states today neglect their populations. Their analysis highlights early warning signals, potential shifts in power structures, the erosion of existing social contracts, and possible interventions.
“Gradual Disempowerment” (January 2025), by Jan Kulveit, Raymond Douglas, Nora Ammann, Deger Turan, David Krueger, and David Duvenaud takes a cautionary stance, suggesting that even in the absence of an intelligence explosion incremental increases in AI capabilities “pose a substantial risk of eventual human disempowerment”. Their economics chapter outlines how the human labor share of GDP might gradually asymptote towards zero, with attendant disempowering consequences.
The Labor Market Risks chapter of the International AI Safety Report 2025 (February 2025) provides a concise, extensively referenced overview of the current evidence. As a scientific report, it remains cautious about speculative scenarios but offers a robust foundation for understanding labor market impacts.
Now let’s look at the more focused explorations:
In “What Will Remain for People to Do?” (March 2025) Daniel Susskind examines the effect of transformative AI on jobs, starting from the premise that AI could eventually perform all economically valuable tasks more productively than humans. His relatively optimistic conclusion is that comparative advantages, preferences for human-led processes, and normative considerations will preserve certain tasks for human workers.
Similarly, Noah Smith, in “Plentiful, high-paying jobs in the age of AI” (March 2024), looks closely at comparative advantage. He argues that as long as compute is finite and not literally free, the use of AI systems comes with an opportunity cost, ensuring some tasks remain economically viable for humans.
Finally, what do the leaders of leading AI labs say? While you may choose to not take what they are writing at face value, it is still instructive.
In “Machines of Loving Grace” (October 2024) Dario Amodei, the CEO of Anthropic, sketches “what a world with powerful AI might look like if everything goes right.” He argues that 20% annual GDP growth is achievable in the developing world.
Sam Altman, CEO of OpenAI, in “The Gentle Singularity” (June 2025) hedges his bets but writes that “2026 will likely see the arrival of systems that can figure out novel insights. 2027 may see the arrival of robots that can do tasks in the real world.”
Given the potentially profound economic changes caused by transformative AI, how should societies respond? Recent discussions focus on reforming our social and economic institutions to meet these challenges.
Deric Cheng is currently editing an anthology titled the “AGI Social Contract” (June 2025). In the anthology’s introductory piece, he summarizes the state of play and outlines key questions for a new social contract.
As part of this anthology, “The Missing Institution: A Global Dividend System for the Age of AI” (June 2025), by Anna Yelizarova proposes a global dividend system that would share the proceeds from concentrated AI wealth.
This suggestion belongs to a broader set of proposals known as “benefits sharing,” which explores mechanisms for the fair distribution of AI-generated wealth.
An earlier and influential proposal along these lines is the Windfall Clause, which describes an “ex ante commitment by AI firms to donate a significant amount of any eventual extremely large profits garnered from the development of transformative AI.” In their paper from January 2020 Cullen O’Keefe, Peter Cihon, Ben Garfinkel, Carrick Flynn, Jade Leung, and Allan Dafoe outline the mechanics of the idea. While the Windfall Clause as initially conceived has somewhat fallen out of favor, it remains influential as a foundational concept.
Building on critiques of the Windfall Clause, “Predistribution Over Redistribution” (October 2024) by Saffron Huang and Sam Manning argues that we should focus on early structural interventions that prevent extreme wealth concentration before it occurs, rather than trying to redistribute wealth after the fact.
Taking a macroeconomic perspective, Era Dabla-Norris at the International Monetary Fund (IMF) published a staff discussion note titled “Broadening the Gains from Generative AI: The Role of Fiscal Policies” (June 2024). This paper outlines fiscal tools and policies that governments could employ to harness generative AI’s benefits, mitigate negative labor market impacts, and promote fair distribution.
Politics fundamentally is about “who gets what, when, and how.” Under any scenario the impacts of transformative AI are unavoidably political. So it is surprising that there has been relatively little analysis of the political dimensions.
A notable exception is the recent work by Anton Leicht. In his essays “AI and Jobs: Politics without Policy” (June 2025) and “Three Fault Lines in Conservative Thinking” (May 2025), Anton offers helpful context on the policy landscape, especially in the US. These essays explore some of the structural and political challenges that will shape how the political debate about transformative AI might play out.
If transformative AI arrives, it stands to reason that governments and public administrations will need to adapt significantly, not just in terms of specific policy decisions but also in their fundamental operations and structures. While the preceding sections considered which policies might be necessary, this section asks: how should government institutions themselves change?
One of the few pieces to explicitly look at this question is Ed de Minckwitz’s “Government in the Age of Superintelligence” from June 2025. In his report for Policy Exchange, the London think tank, he looks at the impact of AGI on healthcare, education, infrastructure, defense, and politics.
To close, four reflections:
1) I have intentionally stacked the deck and sidestepped the question of whether or when we will have transformative AI. Instead, I have focused on asking what will happen once transformative AI is a reality.
2) The rate at which thoughtful analyses and proposals are coming out is already high and will only increase. Specific articles and reports highlighted above may well go stale but the overall framework will hopefully remain useful.
3) This guide is necessarily woefully incomplete. There is far more excellent work than can be featured in an overview like this.
4) Much more thinking and preparation is urgently needed. Given the scale and importance of transformative AI’s potential impact on jobs and economies, the quantity and depth of thinking and policy preparation ought to increase fast.
Thanks to my Windfall Trust colleagues Deric Cheng, Helen Stevenson, Anna Yelizarova, and Adrian Brown for their helpful comments.
", "url": "https://www.dannybuerkli.com/2025/06/18/econ-impacts-of-tai", "image": "/assets/cards/essays/econ-impacts-of-tai.png", "date_published": "2025-06-18T00:00:00-03:00", "date_modified": "2025-06-18T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2025/06/03/end-of-bureaucracy", "title": "The End Of Bureaucracy As We Know It?", "summary": "Digital government only ever rearranged how information moves. AI reaches the actual work — reading, deciding, writing — and that changes the institution.", "content_text": "Danny Buerkli, 06 June 2025AI may be about to do to government what the power loom did to weaving.For decades, we’ve been digitizing government: moving paper forms online, connecting databases, speeding up communication. But even with digital transformation done right, we have only been rearranging how information moves around. The actual work of manipulation of symbols (reading materials, making decisions, writing responses) still happens exclusively the same way it always did: inside brains.That’s why, despite all our technology, a government office today would be instantly recognizable to a civil servant from a hundred years ago. If you brought in a capable civil servant from the early nineteen-hundreds they would be up and running before lunch. We have screens and keyboards now, but the logic of the wiring which shuffles information from one cranium to another remains essentially unchanged.This is not a dig at government bureaucracies. We simply have not had technology which would have changed the fundamental logic of public administration.It is however not how technology transformed other industries. When mechanical looms arrived in the 1800s, they didn’t just help weavers pass thread faster. They fundamentally changed what it meant to make cloth.We are arguably at that moment now with most of knowledge work, which includes public administration. And if we are not there yet we will soon be. AI can read, analyze, draft, and reason. In other words: it can think. If this strikes you as a bizarre claim, I strongly encourage you to play around with some of the most recent reasoning models.Of course AI systems will still need to be supervised in places. But that is no different from how less experienced civil servants have their work checked by more seasoned colleagues.The question before us is: what happens when every policy analyst has the analytical capacity of an entire department? When every caseworker can process applications with the thoroughness of a team? The binding constraint on public administration has always been the number of people who can read, think, and decide. That constraint is about to disappear.What would Lorenzetti's \"Allegory of Good and Bad Government\" look like for our age?This shift raises questions we’ve never had to ask before. Here are four implications worth considering:1) Against State Capacity?We might not want government to become too effective. Imagine tax authorities that can audit every transaction in real-time, or immigration systems that track every movement instantly. Public administrations aren’t just constrained by laws, but the number of people they have, how capable they are, and how well they work together. We may soon find ourselves in the odd position of defending bureaucratic inability to get things done as a feature, not a bug.2) But Do You Have A Choice?Yet we may not have much choice. Consider pharmaceutical regulation: once one country uses AI to review drug applications, others must follow or risk missing critical safety signals buried in complex data. The same logic applies everywhere. As citizens submit AI-generated documents and companies automate their compliance reporting, governments that don’t keep up will simply drown. The pension wave hitting public sectors worldwide only accelerates this: with boomers retiring en masse and fewer workers to replace them, automation becomes necessity, not choice.3) The Small Government ParadoxThis creates an intriguing opportunity for those who’ve long advocated for smaller government. One metric, though not the only one, is headcount. Why replace retiring civil servants when AI can do their work? The political narrative writes itself: no layoffs, just natural attrition plus “automation like we see in the private sector.” But: 100 regulators equipped with AI might wield more power than 10,000 working with paper. Is that really smaller government, or just government with a smaller headcount?4) The Stabilization DilemmaThe counter-argument runs deeper than efficiency. If AI displaces millions of private sector workers, government employment might become our economic ballast, a form of distributed universal basic income disguised as public service. We might need an “inefficiently” large public workforce precisely because the private sector becomes too efficient.The same technology that could make government monumentally more capable might also demand we keep it deliberately incapable. The same force that enables a radically smaller state might require us to maintain a bloated one.", "content_html": "Danny Buerkli, 06 June 2025
AI may be about to do to government what the power loom did to weaving.
For decades, we’ve been digitizing government: moving paper forms online, connecting databases, speeding up communication. But even with digital transformation done right, we have only been rearranging how information moves around. The actual work of manipulation of symbols (reading materials, making decisions, writing responses) still happens exclusively the same way it always did: inside brains.
That’s why, despite all our technology, a government office today would be instantly recognizable to a civil servant from a hundred years ago. If you brought in a capable civil servant from the early nineteen-hundreds they would be up and running before lunch. We have screens and keyboards now, but the logic of the wiring which shuffles information from one cranium to another remains essentially unchanged.
This is not a dig at government bureaucracies. We simply have not had technology which would have changed the fundamental logic of public administration.
It is however not how technology transformed other industries. When mechanical looms arrived in the 1800s, they didn’t just help weavers pass thread faster. They fundamentally changed what it meant to make cloth.
We are arguably at that moment now with most of knowledge work, which includes public administration. And if we are not there yet we will soon be. AI can read, analyze, draft, and reason. In other words: it can think. If this strikes you as a bizarre claim, I strongly encourage you to play around with some of the most recent reasoning models.
Of course AI systems will still need to be supervised in places. But that is no different from how less experienced civil servants have their work checked by more seasoned colleagues.
The question before us is: what happens when every policy analyst has the analytical capacity of an entire department? When every caseworker can process applications with the thoroughness of a team? The binding constraint on public administration has always been the number of people who can read, think, and decide. That constraint is about to disappear.
![\"Map1\"](\"/assets/20250603_lorenzetti.png\")
This shift raises questions we’ve never had to ask before. Here are four implications worth considering:
1) Against State Capacity?
We might not want government to become too effective. Imagine tax authorities that can audit every transaction in real-time, or immigration systems that track every movement instantly. Public administrations aren’t just constrained by laws, but the number of people they have, how capable they are, and how well they work together. We may soon find ourselves in the odd position of defending bureaucratic inability to get things done as a feature, not a bug.
2) But Do You Have A Choice?
Yet we may not have much choice. Consider pharmaceutical regulation: once one country uses AI to review drug applications, others must follow or risk missing critical safety signals buried in complex data. The same logic applies everywhere. As citizens submit AI-generated documents and companies automate their compliance reporting, governments that don’t keep up will simply drown. The pension wave hitting public sectors worldwide only accelerates this: with boomers retiring en masse and fewer workers to replace them, automation becomes necessity, not choice.
3) The Small Government Paradox
This creates an intriguing opportunity for those who’ve long advocated for smaller government. One metric, though not the only one, is headcount. Why replace retiring civil servants when AI can do their work? The political narrative writes itself: no layoffs, just natural attrition plus “automation like we see in the private sector.” But: 100 regulators equipped with AI might wield more power than 10,000 working with paper. Is that really smaller government, or just government with a smaller headcount?
4) The Stabilization Dilemma
The counter-argument runs deeper than efficiency. If AI displaces millions of private sector workers, government employment might become our economic ballast, a form of distributed universal basic income disguised as public service. We might need an “inefficiently” large public workforce precisely because the private sector becomes too efficient.
The same technology that could make government monumentally more capable might also demand we keep it deliberately incapable. The same force that enables a radically smaller state might require us to maintain a bloated one.
", "url": "https://www.dannybuerkli.com/2025/06/03/end-of-bureaucracy", "image": "/assets/cards/essays/end-of-bureaucracy.png", "date_published": "2025-06-03T00:00:00-03:00", "date_modified": "2025-06-03T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2025/02/02/where-the-map-ends", "title": "Where the map ends: understanding Wardley maps’ limitations", "summary": "Wardley mapping is unusually versatile, but it rests on assumptions. Four limitations of the method, including what it cannot tell you about what comes next.", "content_text": "Danny Buerkli, 02 February 2025Every tool carries the context of its creation. Every tool therefore also has certain limitations. Wardley Mapping, though one of the most versatile analytical methods I have come across, inevitably rests on assumptions that shape what the maps can and cannot tell us.Below are four limitations, as I currently understand the method:First, Wardley maps describe how components evolve from genesis to commodity. However, they cannot explain which new components will emerge in the genesis stage. These tend to “pop into existence,” so to speak.At the same time the framework does indicate that innovation, which includes the genesis of new components, is facilitated by the commodification of other components.If we observe a move towards commodification, we can anticipate a wave of combinatorial innovation. Yet the map itself cannot tell us what new element will appear.Second, the development of components along the evolutionary axis is driven by competition.In the absence of competition, we would expect components to remain stationary because the evolutionary pressures aren’t there to move them along.Third, like a topographical map, a Wardley Map can only represent what its creators already know. You can’t map what you don’t see. As a result, we are inevitably constrained by how well — or how poorly — we understand the landscape.Fourth, the usefulness of a Wardley map also depends on which part of reality we choose to represent and which elements we decide to include. The right answer depends on the question we are trying to answer. While picking the appropriate segment of a system at the right level of resolution is crucial, the rules are difficult to codify.Getting the most out of Wardley Mapping requires experience, which is a way of saying that it involves implicit, somewhat difficult-to-codify knowledge. I am not sure whether this is a limitation, but some prefer more easily codified and thus “template-able” methods.These limitations are not defects. Any method must be understood in terms of the assumptions it carries and the constraints which these imply.The question is never whether a method is perfect; the question is whether, for what we are trying to achieve, it is more useful than the next best alternative.We can’t ever capture all of reality… (James Turrell’s “Meeting” at MoMa PS1)", "content_html": "Danny Buerkli, 02 February 2025
Every tool carries the context of its creation. Every tool therefore also has certain limitations. Wardley Mapping, though one of the most versatile analytical methods I have come across, inevitably rests on assumptions that shape what the maps can and cannot tell us.
Below are four limitations, as I currently understand the method:
First, Wardley maps describe how components evolve from genesis to commodity. However, they cannot explain which new components will emerge in the genesis stage. These tend to “pop into existence,” so to speak.
At the same time the framework does indicate that innovation, which includes the genesis of new components, is facilitated by the commodification of other components.
If we observe a move towards commodification, we can anticipate a wave of combinatorial innovation. Yet the map itself cannot tell us what new element will appear.
Second, the development of components along the evolutionary axis is driven by competition.
In the absence of competition, we would expect components to remain stationary because the evolutionary pressures aren’t there to move them along.
Third, like a topographical map, a Wardley Map can only represent what its creators already know. You can’t map what you don’t see. As a result, we are inevitably constrained by how well — or how poorly — we understand the landscape.
Fourth, the usefulness of a Wardley map also depends on which part of reality we choose to represent and which elements we decide to include. The right answer depends on the question we are trying to answer. While picking the appropriate segment of a system at the right level of resolution is crucial, the rules are difficult to codify.
Getting the most out of Wardley Mapping requires experience, which is a way of saying that it involves implicit, somewhat difficult-to-codify knowledge. I am not sure whether this is a limitation, but some prefer more easily codified and thus “template-able” methods.
These limitations are not defects. Any method must be understood in terms of the assumptions it carries and the constraints which these imply.
The question is never whether a method is perfect; the question is whether, for what we are trying to achieve, it is more useful than the next best alternative.
![\"Map1\"](\"/assets/20250202_Turrell.png\")
Danny Buerkli, 26 April 2024
This post was originally published on the staatslabor Linkedin page.
Many organizations find it challenging to respond to technological change. How do you distinguish between urgent necessity and hype? And how do you know what to do, once you believe something must be done?We recently worked with a national government agency on those problems. They felt disoriented by the speed by which they were being overtaken by technological developments.
This agency operates in a specific corner of the education space. The public service they provide is important for society, but could conceivably also be provided by the private sector – though there are good arguments for why this ought to be a government task.
Unusually for a government entity, they are competing with private providers. This meant that the organization would, over time, lose its mandate, if it did not manage to use technology appropriately and offer these services to the standard that their clients have come to expect.
We worked with a small team of government officials on a short project of just a few weeks to help them decide on their tech strategy.
To do this we used Wardley Maps, a method developed by Simon Wardley. His maps are a way of anticipating change and discovering useful strategies.The agency found the results compelling. And because what we did was useful and straightforward at the same time, we felt that by sharing what we did others might benefit too.
First, let’s get three quick disclaimers out of the way:
We will inevitably gloss over important details. It’s impossible to pack in all the nuance and keep it readable at the same time.
We are not naming the government entity. This means that the examples we can provide are unfortunately not as specific as you might hope.
Finally, if you are not familiar with Wardley Mapping, we suggest more detailed readings at the end. It is a remarkably useful way of looking at the world, so any time spent understanding the method more deeply is time well spent.
With all that out of the way, how did we help this government entity make sense of technological change?
First we drafted a (Wardley) map of what they do. This helped us ground all subsequent discussions in the specific details of their operation.
With that map in hand, we looked at where technological developments, such as Large Language Models (LLMs, e.g. GPT-4 or Claude 3), would impact their business.
Lastly, since we now had an agreed upon description of what they were doing and how technology would influence their operation, we discussed what they could and should do about it.
Let’s look at each of these steps in turn.
The first step was to draw a map of what the agency does, i.e. of the service it provides. We did this over the course of three two-hour sessions, together with a small working group of executives.
This map grounded all subsequent discussions over the impact of technology on the agency’s business, as described by the executives themselves.
In between the sessions we would clean up the maps, so that we would have a tidy basis for our next discussion.
This first step happened over the course of a few weeks. We could have gone faster and done it within a week or less, if required. Refining the maps over the course of several sessions helped everyone get comfortable with the method and, more importantly, what the map meant.
The aim was not to draw a perfect map. Like with any map, we wanted to hit the right level of abstraction. All we needed was an artifact we could use to guide our discussions.
![\"Map1\"](\"/assets/20240426_Map1.png\")
Quick interlude: You are looking at an (anonymized) Wardley Map. The general idea is this: The map shows the value chain of the business. The chains are made up of components. Each component depends on those below.
The vertical axis shows visibility to an end user. The higher up a component, the more visible it is to the user.
The horizontal axis plots the evolutionary state of each component. The basic claim is that under conditions of competition, components evolve from genesis to custom-built, from there to product/rental and finally to commodity.
See the recommended readings below for more detailed introductions to Wardley Mapping.
Now that we had a map the group felt was a good enough representation of their activities, we turned to the impact different technologies would have.
We took a list of key technological shifts they were interested in (think LLMs, robotics, blockchain, etc.), looked at each one in turn and asked “where and how will this impact your business?”.
This meant doing two things:
First, in many instances we already had an intuition where, for example, LLMs would impact the business. We then used the map to pinpoint the specific component that would be affected. This grounded everyone’s intuition and translated arguments into a shared, visual language.
Second, we went through every component on the map and asked how LLMs (or any of the other technologies on the list) might conceivably impact that element. This way we were sure not to overlook an important effect.
![\"Map1\"](\"/assets/20240426_Map2.png\")
As a result of these two exercises it quickly became apparent that LLMs were in all likelihood going to have a major impact on the agency. They would affect a large number of central components in the value chain. Virtual Reality, on the other hand, would only impact a minor, non-strategic element in the value chain.
The group quickly gained clarity on where it was worth spending time debating potential consequences of technological shifts and where we could safely move on. Because the maps helped anchor all conversations in the organization’s operational reality, we avoided spending lots of time debating questions that would have ultimately no business impact.
With the elements above in hand, we turned to strategy.
Depending on where on the map technology was impacting components, different responses were needed.
If the component in question was relatively mature (i.e. existing on the right-hand side of the map) this implied the organization needed to react quickly, follow “best practice” and make use of what the market had decided was the winning solution. In other words: there was nothing left to explore or figure out, the organization needed to get with the program and catch up as quickly as possible.
If the components where change was happening were placed in the middle of the map, there might be space to shape what would happen. This is where the organization could make a difference with shrewd, strategic moves.
If, lastly, the components affected by technological change were less evolved (i.e. towards the left-hand side of the map) this meant that there would be a lot of uncertainty around how the future would unfold and therefore a lot of room for exploration. Given the limited resources of the agency we felt that investments here would be too risky. Better to keep a watching brief and, given the considerable amounts of inertia in the environment, not worry about those components too much in the short term.
![\"Map1\"](\"/assets/20240426_Map3.png\")
Let’s take a closer look at the middle category where we saw potential for strategic action. In particular, one specific strategic gameplay stood out.
Without being able to name the specific agency and its service this is by necessity going to be somewhat cryptic, but the general idea was this:
There was a component which was in high demand, but could only be accessed by booking a series of appointments with a human interlocutor. The situation was structurally similar to how you can only get results from a blood test if you go see a doctor. We had, however, reason to believe that the commoditization of LLMs would soon allow this component to evolve to a commodity too.
![\"Map1\"](\"/assets/20240426_Map4.png\")
Because demand for this component got at present routed through a human and because humans are comparatively expensive, a commoditization would lead to a drop in cost and a corresponding increase in demand.
It seemed likely that the bundling with an expensive, but soon unnecessary complement would put pressure on the component to evolve.
The insight was this: what if we proactively commoditize the component and make it available online? This would release all the pent-up demand. That in turn would allow us to increase access to this sought-after component by several orders of magnitude.
The organization could serve this demand in a cost-efficient way, because it’s all commoditized, and route a share of that demand back to its traditional services.
If executed well, this play could be transformational. The agency would be able to serve a much larger number of residents in a cost-effective way (note that’s not the same as “cheap”). It could build organic reach and use that to drive demand for its other services.
In hindsight, after having pored over the maps for a few weeks, the move – and the rationale behind it – seemed obvious to everyone involved. I highly doubt, however, that we would have achieved the same level of clarity and conviction without the shared language of the Wardley maps.
Here are two excellent resources if you would like to learn more about Wardley Mapping:
A short video (the first in a series of four) by Simon Wardley, explaining the basics of Mapping
A blog post by Dave Hora that provides an excellent introduction to strategy and Wardley Mapping
I would like to warmly thank my colleagues at staatslabor, specifically Lynn Reinhart, Giulia Colella, and Dario Meloni who were part of the project team and to Lisa Brombach and Carla Mueller for helpful comments. Lynn and Giulia, in particular, did much of the essential work of drawing (and redrawing) the maps.
Our thanks, of course, also go to the unnamed government organization which kindly gave us permission to try a different way of doing strategy.
Finally, we owe all of what’s described above to Simon Wardley, the inventor of the eponymous method, and to his decision to publish all his work under a Creative Commons license (CC BY-SA). Accordingly, this article is also published under that same license.
", "url": "https://www.dannybuerkli.com/2024/04/26/wardley-maps-tech-strategy", "image": "/assets/cards/essays/wardley-maps-tech-strategy.png", "date_published": "2024-04-26T00:00:00-03:00", "date_modified": "2024-04-26T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2023/10/25/how-to-do-things-at-work", "title": "How to do things at work (if you're starting out)", "summary": "How to do things at work - a modest guide to gettings things done", "content_text": "Danny Buerkli, 25 October 2023This was originally an internal document meant to help out new joiners figure out how to be productive at work.I. What this article is and what it isn’tI occasionally get asked for advice on how to work, i.e. how to get things done at work. I claim no special insight into this. What I can offer, however, are a few things that have worked for me.If you are reading this, you are probably, like me, engaged in some kind of knowledge work.What makes our line of work at staatslabor interesting are the almost infinite degrees of freedom. The downside is, that – unlike artisans – our tools and habits are less well defined.This document offers a set of tools which, if you hone your use of them, will let you do more satisfying, more creative, more challenging, and – by virtue of all of this – more impactful work.My focus here is on individual contribution, not on management. Nevertheless it should all remain useful throughout your career.Everything you are about to see I have been taught by others, either in person or in writing. Much of it I practice regularly. Some of it I aspire to, but might still be working on.The usual disclaimers apply. Your mileage may vary. Keep what’s useful to you, discard what isn’t.II. First, some general advice on how to be productive at workWhat follows are a few principles that seem to work well in almost any context. Do the obvious thing. Much of the secret of being productive is this: doing the obvious next step that will advance what you are trying to do. Be two steps ahead. Think “what will be required once we complete this step?” Anticipate what it is and, if possible, already have that prepared or, even better, get it done. Output first. In academia we learn to work incrementally. We read for weeks (or months) on end and then produce a paper at once. Don’t do that. Start producing right away. Work in increments, not in big bangs. Signal intentions early and often. Overcommunicate. When you feel that you are sharing too much, you are probably approaching the level of communication you should be at. Drive to the conclusion (“so what?”). When doing analysis of any kind, always drive it to an action-oriented conclusion (rather than just preparing the facts). Always suggest a course of action. Whenever you can, offer your (reasoned) opinion on what should happen next. Aim for speed and precision. Speed matters. Most things can be done much more quickly than we think. Being quick will pay enormous (compounding!) dividends. But be also precise. Dot the i’s and cross the t’s. Aim for clarity, insight, and brevity. Make your points obvious, make them insightful and keep it short. Or, as Woodrow Wilson allegedly said “be brief, be brilliant, be gone”. Leverage existing work. Steal like a pirate. Reuse as much as possible. Underpromise and overdeliver. Do what you said you’d do by the time you committed to. If necessary, renegotiate. Ask for feedback. Ask for as much feedback as you can, from individuals whose opinions you respect. The secret to receiving feedback is this: listen and accept it graciously. That’s it. Later on you decide what to keep and what to ignore. Ask for help when help is needed. This is something many people, myself included, struggle with. Don’t. Get the help you need and get it early. Ask until you understand what is being asked of you. If someone is tasking you with something, it’s on them to make themselves understood but it’s on you to ask until you’re clear on what they want. Avoid surprises. Things will go wrong. You will make mistakes. Don’t hide them. Communicate early and actively. Maintain an internal locus of control. Having an internal locus of control means believing (correctly, I think) that you influence what happens to you. Recovery is a strategic must. You will fail at certain things and it will make you feel bad. Setbacks are inevitable. What you can influence is how you recover. III. How to know what to do whenThere are probably as many prioritization and time management methods as there are atoms in the universe. The one that has stuck for me comes from David Allen’s “Getting Things Done”.It can be summarized as follows:Have one “to do” list. Everything you need or want to do goes on there.That’s it. No categories, no colors, no order, no symbols, no intricate time allocation systems, nothing. Just a list.I use the Notes app on my Mac (which will sync effortlessly between my laptop and my phone), though anything that can store text will do.Knowing that every task I need to remember is on this one list gives me peace of mind. When I plan my week or my day, I simply look down the list and then do whatever is most important and/or urgent.IV. Taking note(s)Taking good notes and being able to retrieve them quickly is a superpower. It’s also indispensable.There is no way you will be able to remember everything if you don’t write it down. And unless you can find your notes quickly, they will be of no use. You should be able to locate your notes for any given meeting or topic within less than 30 seconds.All my notes go in one single Google Doc. The most recent entries go at the top. Every entry has a short title and a date. There are much more elaborate ways to take notes. This system works for me because it is incredibly simple.V. Being responsive on Email and SlackHere’s how I try to use Email and Slack: If it will take me less than two minutes to respond (or decide), I will do it right away. If it will take me longer, I’ll do it at a time that suits me.The most important habit to develop is taking immediate action on things. Don’t delay. Unless you genuinely need time to think or gather more information. But for many messages that won’t be the case. Just respond right away.Allen’s “Getting Things Done” has more on this, but that is essentially it.VI. Using your computer efficientlyYour computer is your most important tool. Most days you are probably touching it for eight hours or more. Learn how to use it like the serious professional that you are.The most important habit is to use keyboard shortcuts. It will make you a lot faster.VII. Further readingHere is a small selection of things I have found worthwhile:David Allen, “Getting Things Done”: A classic, and deservedly so. The only productivity book that has stuck for me. It’s old and very US-American, read past that if need be.Javier Grillo-Marxuach, “The Eleven Laws of Showrunning”: Funny and poignant. Your job has more in common with Hollywood showrunning than you might think.Robert Poynton, “Do Improvise: Less Push. More Pause. Better results”: A short and wise book on how to draw inspiration from improv theater at work.", "content_html": "Danny Buerkli, 25 October 2023
This was originally an internal document meant to help out new joiners figure out how to be productive at work.
I occasionally get asked for advice on how to work, i.e. how to get things done at work. I claim no special insight into this. What I can offer, however, are a few things that have worked for me.If you are reading this, you are probably, like me, engaged in some kind of knowledge work.
What makes our line of work at staatslabor interesting are the almost infinite degrees of freedom. The downside is, that – unlike artisans – our tools and habits are less well defined.
This document offers a set of tools which, if you hone your use of them, will let you do more satisfying, more creative, more challenging, and – by virtue of all of this – more impactful work.My focus here is on individual contribution, not on management. Nevertheless it should all remain useful throughout your career.
Everything you are about to see I have been taught by others, either in person or in writing. Much of it I practice regularly. Some of it I aspire to, but might still be working on.
The usual disclaimers apply. Your mileage may vary. Keep what’s useful to you, discard what isn’t.
What follows are a few principles that seem to work well in almost any context.
Do the obvious thing. Much of the secret of being productive is this: doing the obvious next step that will advance what you are trying to do.
Be two steps ahead. Think “what will be required once we complete this step?” Anticipate what it is and, if possible, already have that prepared or, even better, get it done.
Output first. In academia we learn to work incrementally. We read for weeks (or months) on end and then produce a paper at once. Don’t do that. Start producing right away. Work in increments, not in big bangs.
Signal intentions early and often. Overcommunicate. When you feel that you are sharing too much, you are probably approaching the level of communication you should be at.
Drive to the conclusion (“so what?”). When doing analysis of any kind, always drive it to an action-oriented conclusion (rather than just preparing the facts).
Always suggest a course of action. Whenever you can, offer your (reasoned) opinion on what should happen next.
Aim for speed and precision. Speed matters. Most things can be done much more quickly than we think. Being quick will pay enormous (compounding!) dividends. But be also precise. Dot the i’s and cross the t’s.
Aim for clarity, insight, and brevity. Make your points obvious, make them insightful and keep it short. Or, as Woodrow Wilson allegedly said “be brief, be brilliant, be gone”.
Leverage existing work. Steal like a pirate. Reuse as much as possible.
Underpromise and overdeliver. Do what you said you’d do by the time you committed to. If necessary, renegotiate.
Ask for feedback. Ask for as much feedback as you can, from individuals whose opinions you respect. The secret to receiving feedback is this: listen and accept it graciously. That’s it. Later on you decide what to keep and what to ignore.
Ask for help when help is needed. This is something many people, myself included, struggle with. Don’t. Get the help you need and get it early.
Ask until you understand what is being asked of you. If someone is tasking you with something, it’s on them to make themselves understood but it’s on you to ask until you’re clear on what they want.
Avoid surprises. Things will go wrong. You will make mistakes. Don’t hide them. Communicate early and actively.
Maintain an internal locus of control. Having an internal locus of control means believing (correctly, I think) that you influence what happens to you.
Recovery is a strategic must. You will fail at certain things and it will make you feel bad. Setbacks are inevitable. What you can influence is how you recover.
There are probably as many prioritization and time management methods as there are atoms in the universe. The one that has stuck for me comes from David Allen’s “Getting Things Done”.
It can be summarized as follows:
Have one “to do” list. Everything you need or want to do goes on there.
That’s it. No categories, no colors, no order, no symbols, no intricate time allocation systems, nothing. Just a list.
I use the Notes app on my Mac (which will sync effortlessly between my laptop and my phone), though anything that can store text will do.
Knowing that every task I need to remember is on this one list gives me peace of mind. When I plan my week or my day, I simply look down the list and then do whatever is most important and/or urgent.
Taking good notes and being able to retrieve them quickly is a superpower. It’s also indispensable.
There is no way you will be able to remember everything if you don’t write it down. And unless you can find your notes quickly, they will be of no use. You should be able to locate your notes for any given meeting or topic within less than 30 seconds.
All my notes go in one single Google Doc. The most recent entries go at the top. Every entry has a short title and a date. There are much more elaborate ways to take notes. This system works for me because it is incredibly simple.
Here’s how I try to use Email and Slack:
The most important habit to develop is taking immediate action on things. Don’t delay. Unless you genuinely need time to think or gather more information. But for many messages that won’t be the case. Just respond right away.
Allen’s “Getting Things Done” has more on this, but that is essentially it.
Your computer is your most important tool. Most days you are probably touching it for eight hours or more. Learn how to use it like the serious professional that you are.
The most important habit is to use keyboard shortcuts. It will make you a lot faster.
Here is a small selection of things I have found worthwhile:
David Allen, “Getting Things Done”: A classic, and deservedly so. The only productivity book that has stuck for me. It’s old and very US-American, read past that if need be.
Javier Grillo-Marxuach, “The Eleven Laws of Showrunning”: Funny and poignant. Your job has more in common with Hollywood showrunning than you might think.
Robert Poynton, “Do Improvise: Less Push. More Pause. Better results”: A short and wise book on how to draw inspiration from improv theater at work.
", "url": "https://www.dannybuerkli.com/2023/10/25/how-to-do-things-at-work", "image": "/assets/cards/essays/how-to-do-things-at-work.png", "date_published": "2023-10-25T00:00:00-03:00", "date_modified": "2023-10-25T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2021/04/28/four-facets-of-failure", "title": "Facets of failure - lessons we might be able to learn from Switzerland’s pandemic crisis management", "summary": "What Switzerland's pandemic crisis management reveals about how states fail — four distinct kinds of failure that are usually mistaken for one.", "content_text": "Danny Buerkli, 28 April 2021This article was originally published on Medium.Not long ago I was invited to give a talk about some lessons we might be able to learn from Switzerland’s pandemic crisis management. The invitation provided a welcome opportunity to take a step back and reflect on what’s been happening since all of this started.Much of this piece is also based on a Twitter thread, which turned out to resonate quite strongly with many and helped me think through the argument.1 - We’ve got this pandemic business — or so we thoughtHeading into this pandemic, one could have thought that Switzerland should be able to nail this. And yet — judging by almost any metric — we did not nail this and we still are not nailing it.Why could one have had some confidence that as a country, as a society, and as a political system we should be able to respond effectively to this pandemic?We could make a long list here, but let’s just mention a few rather favourable initial conditions:1) Switzerland has one of the world’s highest GDPs per capita. That should be useful, as combating a pandemic is expensive. If there’s one thing Switzerland has in abundance, it’s financial resources.2) Government enjoys very high trust. Trust in government is the highest in Switzerland out of all OECD countries. Without trust, people won’t follow containment measures for any length of time. With high trust, they are more likely to do so.3) The country has the highest total expenditure on health per capita (compared to other OECD countries). Expenditures are a highly imperfect proxy for health care system capacity. Running the most expensive healthcare system in the OECD should still give us some confidence with regards to its abilities.4) The nation tops many science and innovation rankings. Switzerland for example ranks first in WIPO’s “Global Innovation Index”. So there should be plenty of capacity and ability to innovate and science ourselves out of this situation.We could go on, but you get the idea.So why did things not turn out so well, despite all these favourable initial conditions? That’s the question I’ll try to give some tentative answers to.2 - First, some humility is in orderBefore we proceed, I’d like to get a number of important caveats out of the way.First, there is a lot of nuance to this, which is difficult to express fully without rendering an essay clumsy and unreadable. I’ll try to hedge appropriately where necessary, but even without those explicit caveats: we should always remember that things are complicated. None of this is easy. Reality has a surprising amount of detail, as someone once said.Second, it’s easy to criticise when you’re not in the hot seat. Anyone in a position of responsibility who has stuck their neck out and has been taking decisions deserves our respect, even if we might disagree with them.Finally, “the owl of Minerva takes its flight only when the shades of night are gathering”. This is a pretentious way of saying that we might not truly know for a while who was right, who wasn’t, what worked, and what didn’t. So some humility is in order.My own pandemic story begins in Rome. I spent all of February 2020 there, as part of an extended holiday between jobs. As everyone knows, things started to go a bit crazy in the north of Italy around that time. Towards the end of the month I decided that it was time to bail and move back to Switzerland.These bricks have seen it allReturning felt like entering an alternate reality. In Rome everyone was already suitably alarmed about the situation and gently freaking out. In Zurich, the pandemic had barely registered. This is particularly remarkable, given that Bergamo is geographically closer to Zurich compared to Rome.So we waited and waited for what felt like an eternity, until on March 16th 2020 the Federal Council declared an “extraordinary situation” and introduced stringent measures.Since then, the country’s response has been patchy at best and horrifically ineffective at worst.Let’s explore some possible reasons why that’s the case.3 - Six strengths of the Swiss political systemBefore we turn to the hypotheses for why the Swiss political system failed to produce an effective response to the pandemic, I’d like to take a look at what it is good at. As we’ll see, those strengths may have also ended up hampering the response to the pandemic.Some of the strengths of the Swiss political system include:1) Producing legitimacyThe political system solves for legitimacy first. This is an invaluable asset. The ‘crisis of legitimacy’ which many other countries experience, isn’t a thing or at least it isn’t an issue in the same virulent way.2) Moderating the extremes and forging long-term compromiseThis is also generally A Good Thing. The Swiss political system has nifty mechanisms which will automatically sand off the edges of any piece of legislation or any policy. It also encourages repeated interactions, giving every player a strong incentive to take the long view and cooperate.3) Being responsive to every conceivable interest groupYou want to defend the interests of horned cows? Go right ahead, the Swiss political system has a way for you to submit your proposed legislation to a popular vote. This also, for better or worse, means high responsiveness to well-organised business lobbies and other interest groups.4) Allowing maximum regional varianceSwitzerland has 26 cantons, best understood as quasi-autonomous regions with vast decision-making powers. Their governments are, and this is important, elected directly by the voting population in those cantons.As a consequence regional variance is politically legitimated and isn’t seen as a “postcode lottery”. Compare and contrast that with the UK, for example, where regional variance isn’t grounded in the same sort of local legitimacy and therefore is understood as a problem to be resolved.5) Achieving perfection (while taking all the time in the world)The so-called “Swiss finish” refers to the obsessive nature with which The Perfect Solution will be crafted, refined, improved, and then polished again before it’s seen as maybe good enough. Usually no expense in either time, effort or money will be spared. There is an almost religious dedication to the achievement of perfection. It’s something that’s easy to make fun of, but when it works it’s really quite something.6) Non-action for the sake of avoidance of errorThis is, I believe, the most poorly understood and most generally underestimated feature of the system. Implicitly, the Swiss system follows a logic of “first, do no harm”.That’s meaningful because governments are perfectly capable of inflicting horrific damage on their populations, even when they hold the best of intentions. It’s why following a maxim of “hang on, first, let’s make sure we don’t screw this up” is possibly not that misguided.Many books have been written on the horrific errors that governments have committed. One of the classics of the genre is James C. Scott’s “Seeing Like a State — How Certain Schemes to Improve the Human Condition Have Failed”, in which he chronicles the immense human cost of high-minded, utopian government projects.Another more contemporary one is Anthony King’s and Ivor Crewe’s entertaining and yet depressing “The Blunders of our Governments” which chronicles the phenomenal blunders of various UK governments.Few if any such books have, so far and to the best of my knowledge, been written about Switzerland’s policy failures. And arguably that’s because the Swiss system is exceptionally good at avoiding errors of commission.A happy list of lovely attributesThe Swiss political system has, in other words, plenty of qualities.There’s just one problem.In a fast-moving pandemic, the same mechanisms that produce all those lovely attributes, don’t serve us well.Mike Ryan, head of the Health Emergencies Programme at WHO, put it this way back in March 2020: “If you need to be right before you move, you will never win. Perfection is the enemy of the good. Speed trumps perfection. Everyone is afraid of the consequence of error. But the greatest error is not to move”.And what are the two things that Switzerland absolutely doesn’t do?Speed and imperfection.Now, it would be a bit too easy to say “there you have it, Switzerland doesn’t do speed and imperfection, that’s why we’re doomed” and call it a day.That’s why I’d like to examine four specific facets of this failure which I think are interesting.4 - The four facets of failureUnderstanding the finer details of why things fail is important, because we can’t copy “best practices” from elsewhere in the world if they don’t fit the context. Because, to loosely paraphrase Tolstoy, all the successful countries are alike, but every unsuccessful country is unsuccessful in its own way.If we want to know what might be done in Switzerland, we need to understand the specific ways in which Switzerland is unsuccessful.Four facets of failure summarised in one neat tableHere are four facets of failure which have contributed to the mess that we are in:i) Usually the relationship between the federal level and the cantons is “loosely coupled”, now we are in desperate need of “tight coupling”.Information and decisions flow both ways between cantons and the federal level, but it’s not a rigid system. Usually that’s a good thing. “Tightly coupled” complex systems are much more prone to catastrophic failure, ‘loosely coupled’ ones are more forgiving.The problem is that now we’re arguably in desperate need of some properly tight coupling. Vaccine deliveries need to be moved at the right time to the right place, decisions about restrictions need to be coordinated, and all of this needs to be done at pace. The system seems to struggle heavily with achieving that tight coupling.ii) The dirty secret of the so-called “laboratory of ideas” is that while there is some experimentation there is barely any learning.It’s often said that the decentralised nature of decision-making between cantons allows for lots of experimentation and then cantons swiftly learn what works from each other.What’s not to like? In theory, it all sounds great. In actual fact, there seems to be some, rather limited experimentation, but next to no learning.If “policy diffusion” happens at all, it seems to happen at a glacial pace. That’s maybe fine under normal circumstances, but it’s very costly in a crisis.At the end of March the Canton of Bern decided to start a mass testing pilot with six schools. All fine and dandy because it makes sense to test the mechanics of this before you roll it out, doesn’t it?Yes, except that the Canton of Grisons had been running such mass tests for months. It’s entirely unclear why Bern shouldn’t be able to learn almost anything they’d need to learn from an identical program that’s been running for months and at scale in a comparable canton.I’m not familiar with the finer details of the school testing regime in Berne, so it’s possible that I am missing important detail here. The wider point however stands: no one currently gets the impression that cantonal governments are exceptionally quick to learn from others.iii) Compromise is the Holy Grail of Swiss politics and at the same time a broken heuristic.The Swiss system is, as mentioned above, truly exceptionally good at forging compromise. There’s a generally useful and accepted heuristic which says that — everything else being equal — aiming for compromise is a good thing.Compromise is seen as good not just because compromising is nice and friendly, but because it’s supposed to improve the quality of outcomes.This heuristic works in “concave” situations. Compromising between Option A (on the left) and Option B (on the right) will in fact lead to a superior outcome, relative to being on either end of the axis.The graph is taken from Vitalik Buterin’s blogpost on “Convex and Concave Dispositions”.It, however, does not work in convex situations.In a convex situation compromise leads to an inferior outcome. The properties of a fast-moving, exponentially growing pandemic are unfortunately such that we are in a convex situation.Here’s an example: An effective lockdown that pushes R to zero leads to an exponential decline of infections and therefore quick eradication. A half-hearted lockdown that only gets R to just below 1 will get you months of pain with not much to show for.iv) We suffer from a “self-delusion of competence” in public administration.We have dramatically underinvested in the digital transformation of our public admin. Before the pandemic too few believed this to be the case. Many efforts have begun, and they’re mostly led by excellent, committed, inspiring, and competent people.But we have only begun to grasp the sheer scale of the problem. And these people need vastly more resources (mostly: teams, budget authority, and power) than they are currently being given.Here’s a highly scientific depiction of the problem:The size of the problem vs. the current size of the effort to address it5 - A crisis of ambition and of imaginationThese four facets of failure explain, I’d argue, at least in part why the Swiss political system hasn’t covered itself in glory in the fight against this pandemic.The facets are, of course, somewhat arbitrary. There are other aspects which are likely just as important. But ultimately all of these explanations fail to capture some larger truth about what is happening.Ultimately this feels like a crisis of ambition and of imagination.It feels like we have just given up. It’s this defeatist position which I find deeply unsettling and sad.This should have been a moment for us to collectively shine, to deploy the vast resources and talent this place has in order to overcome a monumental challenge. Some of that did happen, just not nearly enough.The biggest hurdle seems to be a prevailing “business as usual” mindset. It assumes that we can handle this crisis within established channels and with established mechanisms.The countries that did comparatively well all shifted into an “all-out effort” mode. We have not.Just before Easter 2021 an official from the Canton of Basel-Stadt said that the vaccination center would remain closed over the long weekend despite the fact that a vaccine shipment had arrived just before. Why? Because asking people to show up at short notice “wouldn’t go down well”.Quote of an official in NZZ am SonntagI don’t wish to dunk on this one particular official, I’m trying to make a broader point: this attitude seems emblematic of the “business as usual” mindset and everything that is wrong about how we are handling this crisis.Why we weren’t able to shift into an ‘all-out effort’ mode is a question for another day. It may be too late for this crisis but if one good thing could come from all of this, it’s that we’ve been given a brutal reality check and some of our collective delusions will have been taken away from us.I’m hopeful that this can provide the ground for meaningful change and reform. There’s much work to be done and I’m encouraged by the many committed, capable, and creative teams and individuals having a go at changing the collective infrastructure we all rely on for our wellbeing, prosperity, and freedom.This essay builds on many sources, including many conversations I’ve had with friends and my colleagues at staatslabor. I am also grateful to everyone on Twitter who pointed out flaws and mistakes in the original thread.If you’d like to dive deeper into any of the ideas or concepts used in the text, in addition to the ones already mentioned in the text, I’d recommend having a look at the following books and essays:Charles Perrow, “Normal Accidents: Living with High-Risk Technologies” (1999), a detailed, book-length exploration of loose and tight coupling in the context of complex systems.Donella Meadows, “Thinking in Systems: A Primer” (2008), an accessible and yet substantial introduction to complex systems thinking.Danny Buerkli, “How we know when decentralization works in government” (2019), an essay which still informs my thinking on experimentation and learning in federal/decentralised structures.Georg Lakoff and Mark Johnson, “Metaphors we live by” (1980), an explanation of the primordial importance of metaphors for how we understand the world around us.", "content_html": "Danny Buerkli, 28 April 2021
This article was originally published on Medium.
Not long ago I was invited to give a talk about some lessons we might be able to learn from Switzerland’s pandemic crisis management. The invitation provided a welcome opportunity to take a step back and reflect on what’s been happening since all of this started.
Much of this piece is also based on a Twitter thread, which turned out to resonate quite strongly with many and helped me think through the argument.
Heading into this pandemic, one could have thought that Switzerland should be able to nail this. And yet — judging by almost any metric — we did not nail this and we still are not nailing it.Why could one have had some confidence that as a country, as a society, and as a political system we should be able to respond effectively to this pandemic?
We could make a long list here, but let’s just mention a few rather favourable initial conditions:
1) Switzerland has one of the world’s highest GDPs per capita. That should be useful, as combating a pandemic is expensive. If there’s one thing Switzerland has in abundance, it’s financial resources.
2) Government enjoys very high trust. Trust in government is the highest in Switzerland out of all OECD countries. Without trust, people won’t follow containment measures for any length of time. With high trust, they are more likely to do so.
3) The country has the highest total expenditure on health per capita (compared to other OECD countries). Expenditures are a highly imperfect proxy for health care system capacity. Running the most expensive healthcare system in the OECD should still give us some confidence with regards to its abilities.
4) The nation tops many science and innovation rankings. Switzerland for example ranks first in WIPO’s “Global Innovation Index”. So there should be plenty of capacity and ability to innovate and science ourselves out of this situation.
We could go on, but you get the idea.
So why did things not turn out so well, despite all these favourable initial conditions? That’s the question I’ll try to give some tentative answers to.
Before we proceed, I’d like to get a number of important caveats out of the way.
First, there is a lot of nuance to this, which is difficult to express fully without rendering an essay clumsy and unreadable. I’ll try to hedge appropriately where necessary, but even without those explicit caveats: we should always remember that things are complicated. None of this is easy. Reality has a surprising amount of detail, as someone once said.
Second, it’s easy to criticise when you’re not in the hot seat. Anyone in a position of responsibility who has stuck their neck out and has been taking decisions deserves our respect, even if we might disagree with them.
Finally, “the owl of Minerva takes its flight only when the shades of night are gathering”. This is a pretentious way of saying that we might not truly know for a while who was right, who wasn’t, what worked, and what didn’t. So some humility is in order.
My own pandemic story begins in Rome. I spent all of February 2020 there, as part of an extended holiday between jobs. As everyone knows, things started to go a bit crazy in the north of Italy around that time. Towards the end of the month I decided that it was time to bail and move back to Switzerland.
![\"\"](\"/assets/20210428_4things_a.png\")
Returning felt like entering an alternate reality. In Rome everyone was already suitably alarmed about the situation and gently freaking out. In Zurich, the pandemic had barely registered. This is particularly remarkable, given that Bergamo is geographically closer to Zurich compared to Rome.
So we waited and waited for what felt like an eternity, until on March 16th 2020 the Federal Council declared an “extraordinary situation” and introduced stringent measures.
Since then, the country’s response has been patchy at best and horrifically ineffective at worst.
Let’s explore some possible reasons why that’s the case.
Before we turn to the hypotheses for why the Swiss political system failed to produce an effective response to the pandemic, I’d like to take a look at what it is good at. As we’ll see, those strengths may have also ended up hampering the response to the pandemic.Some of the strengths of the Swiss political system include:
The political system solves for legitimacy first. This is an invaluable asset. The ‘crisis of legitimacy’ which many other countries experience, isn’t a thing or at least it isn’t an issue in the same virulent way.
This is also generally A Good Thing. The Swiss political system has nifty mechanisms which will automatically sand off the edges of any piece of legislation or any policy. It also encourages repeated interactions, giving every player a strong incentive to take the long view and cooperate.
You want to defend the interests of horned cows? Go right ahead, the Swiss political system has a way for you to submit your proposed legislation to a popular vote. This also, for better or worse, means high responsiveness to well-organised business lobbies and other interest groups.
Switzerland has 26 cantons, best understood as quasi-autonomous regions with vast decision-making powers. Their governments are, and this is important, elected directly by the voting population in those cantons.
As a consequence regional variance is politically legitimated and isn’t seen as a “postcode lottery”. Compare and contrast that with the UK, for example, where regional variance isn’t grounded in the same sort of local legitimacy and therefore is understood as a problem to be resolved.
The so-called “Swiss finish” refers to the obsessive nature with which The Perfect Solution will be crafted, refined, improved, and then polished again before it’s seen as maybe good enough. Usually no expense in either time, effort or money will be spared. There is an almost religious dedication to the achievement of perfection. It’s something that’s easy to make fun of, but when it works it’s really quite something.
This is, I believe, the most poorly understood and most generally underestimated feature of the system. Implicitly, the Swiss system follows a logic of “first, do no harm”.
That’s meaningful because governments are perfectly capable of inflicting horrific damage on their populations, even when they hold the best of intentions. It’s why following a maxim of “hang on, first, let’s make sure we don’t screw this up” is possibly not that misguided.
Many books have been written on the horrific errors that governments have committed. One of the classics of the genre is James C. Scott’s “Seeing Like a State — How Certain Schemes to Improve the Human Condition Have Failed”, in which he chronicles the immense human cost of high-minded, utopian government projects.
Another more contemporary one is Anthony King’s and Ivor Crewe’s entertaining and yet depressing “The Blunders of our Governments” which chronicles the phenomenal blunders of various UK governments.
Few if any such books have, so far and to the best of my knowledge, been written about Switzerland’s policy failures. And arguably that’s because the Swiss system is exceptionally good at avoiding errors of commission.
![\"\"](\"/assets/20210428_4things_b.png\")
The Swiss political system has, in other words, plenty of qualities.
There’s just one problem.
In a fast-moving pandemic, the same mechanisms that produce all those lovely attributes, don’t serve us well.
Mike Ryan, head of the Health Emergencies Programme at WHO, put it this way back in March 2020:
“If you need to be right before you move, you will never win. Perfection is the enemy of the good. Speed trumps perfection. Everyone is afraid of the consequence of error. But the greatest error is not to move”.
And what are the two things that Switzerland absolutely doesn’t do?
Speed and imperfection.
Now, it would be a bit too easy to say “there you have it, Switzerland doesn’t do speed and imperfection, that’s why we’re doomed” and call it a day.
That’s why I’d like to examine four specific facets of this failure which I think are interesting.
Understanding the finer details of why things fail is important, because we can’t copy “best practices” from elsewhere in the world if they don’t fit the context. Because, to loosely paraphrase Tolstoy, all the successful countries are alike, but every unsuccessful country is unsuccessful in its own way.
If we want to know what might be done in Switzerland, we need to understand the specific ways in which Switzerland is unsuccessful.
![\"\"](\"/assets/20210428_4things_c.png\")
Here are four facets of failure which have contributed to the mess that we are in:
Information and decisions flow both ways between cantons and the federal level, but it’s not a rigid system. Usually that’s a good thing. “Tightly coupled” complex systems are much more prone to catastrophic failure, ‘loosely coupled’ ones are more forgiving.
The problem is that now we’re arguably in desperate need of some properly tight coupling. Vaccine deliveries need to be moved at the right time to the right place, decisions about restrictions need to be coordinated, and all of this needs to be done at pace. The system seems to struggle heavily with achieving that tight coupling.
It’s often said that the decentralised nature of decision-making between cantons allows for lots of experimentation and then cantons swiftly learn what works from each other.
What’s not to like? In theory, it all sounds great. In actual fact, there seems to be some, rather limited experimentation, but next to no learning.
If “policy diffusion” happens at all, it seems to happen at a glacial pace. That’s maybe fine under normal circumstances, but it’s very costly in a crisis.
At the end of March the Canton of Bern decided to start a mass testing pilot with six schools. All fine and dandy because it makes sense to test the mechanics of this before you roll it out, doesn’t it?
Yes, except that the Canton of Grisons had been running such mass tests for months. It’s entirely unclear why Bern shouldn’t be able to learn almost anything they’d need to learn from an identical program that’s been running for months and at scale in a comparable canton.
I’m not familiar with the finer details of the school testing regime in Berne, so it’s possible that I am missing important detail here. The wider point however stands: no one currently gets the impression that cantonal governments are exceptionally quick to learn from others.
The Swiss system is, as mentioned above, truly exceptionally good at forging compromise. There’s a generally useful and accepted heuristic which says that — everything else being equal — aiming for compromise is a good thing.
Compromise is seen as good not just because compromising is nice and friendly, but because it’s supposed to improve the quality of outcomes.
This heuristic works in “concave” situations. Compromising between Option A (on the left) and Option B (on the right) will in fact lead to a superior outcome, relative to being on either end of the axis.
![\"\"](\"/assets/20210428_4things_d.png\")
It, however, does not work in convex situations.
In a convex situation compromise leads to an inferior outcome. The properties of a fast-moving, exponentially growing pandemic are unfortunately such that we are in a convex situation.
Here’s an example: An effective lockdown that pushes R to zero leads to an exponential decline of infections and therefore quick eradication. A half-hearted lockdown that only gets R to just below 1 will get you months of pain with not much to show for.
We have dramatically underinvested in the digital transformation of our public admin. Before the pandemic too few believed this to be the case. Many efforts have begun, and they’re mostly led by excellent, committed, inspiring, and competent people.But we have only begun to grasp the sheer scale of the problem. And these people need vastly more resources (mostly: teams, budget authority, and power) than they are currently being given.
Here’s a highly scientific depiction of the problem:
![\"\"](\"/assets/20210428_4things_e.png\")
These four facets of failure explain, I’d argue, at least in part why the Swiss political system hasn’t covered itself in glory in the fight against this pandemic.
The facets are, of course, somewhat arbitrary. There are other aspects which are likely just as important. But ultimately all of these explanations fail to capture some larger truth about what is happening.
Ultimately this feels like a crisis of ambition and of imagination.It feels like we have just given up. It’s this defeatist position which I find deeply unsettling and sad.
This should have been a moment for us to collectively shine, to deploy the vast resources and talent this place has in order to overcome a monumental challenge. Some of that did happen, just not nearly enough.
The biggest hurdle seems to be a prevailing “business as usual” mindset. It assumes that we can handle this crisis within established channels and with established mechanisms.The countries that did comparatively well all shifted into an “all-out effort” mode. We have not.
Just before Easter 2021 an official from the Canton of Basel-Stadt said that the vaccination center would remain closed over the long weekend despite the fact that a vaccine shipment had arrived just before. Why? Because asking people to show up at short notice “wouldn’t go down well”.
![\"\"](\"/assets/20210428_4things_f.png\")
I don’t wish to dunk on this one particular official, I’m trying to make a broader point: this attitude seems emblematic of the “business as usual” mindset and everything that is wrong about how we are handling this crisis.
Why we weren’t able to shift into an ‘all-out effort’ mode is a question for another day. It may be too late for this crisis but if one good thing could come from all of this, it’s that we’ve been given a brutal reality check and some of our collective delusions will have been taken away from us.
I’m hopeful that this can provide the ground for meaningful change and reform. There’s much work to be done and I’m encouraged by the many committed, capable, and creative teams and individuals having a go at changing the collective infrastructure we all rely on for our wellbeing, prosperity, and freedom.
This essay builds on many sources, including many conversations I’ve had with friends and my colleagues at staatslabor. I am also grateful to everyone on Twitter who pointed out flaws and mistakes in the original thread.
If you’d like to dive deeper into any of the ideas or concepts used in the text, in addition to the ones already mentioned in the text, I’d recommend having a look at the following books and essays:
Charles Perrow, “Normal Accidents: Living with High-Risk Technologies” (1999), a detailed, book-length exploration of loose and tight coupling in the context of complex systems.
Donella Meadows, “Thinking in Systems: A Primer” (2008), an accessible and yet substantial introduction to complex systems thinking.
Danny Buerkli, “How we know when decentralization works in government” (2019), an essay which still informs my thinking on experimentation and learning in federal/decentralised structures.
Georg Lakoff and Mark Johnson, “Metaphors we live by” (1980), an explanation of the primordial importance of metaphors for how we understand the world around us.
", "url": "https://www.dannybuerkli.com/2021/04/28/four-facets-of-failure", "image": "/assets/cards/essays/four-facets-of-failure.png", "date_published": "2021-04-28T00:00:00-03:00", "date_modified": "2021-04-28T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2021/03/03/four-things-ive-been-surprised-by", "title": "Four things I’ve been surprised by in the context of the pandemic", "summary": "Four things a year of pandemic made surprising, including several the author had been confidently wrong about.", "content_text": "Danny Buerkli, 03 March 2021This short piece was originally published on Medium. It was originally a thread I wrote on Twitter in January 2021. It seemed to resonate quite a bit, which is why I’ve turned it into a lightly edited blog post.I’ve been cleaning out my notes and wanted to share a short note with four things I’ve learned, I’ve been wrong about and I’ve been surprised by in the context of the pandemic.The usual disclaimers apply: This is a short blog post (and formerly a Twitter thread), i.e. there isn’t a huge amount of nuance in what follows.1) Narratives are (much) more stable than I would have ever thought possible.Switzerland is, by any sensible metric, doing badly in this pandemic. And yet this hasn’t filtered through to public discourse. It boggles the mind.The logic seems to be “we usually do really well, therefore we must be doing well now”. Except we aren’t.Exhibit A:2) If our starting assumptions are wrong, we’re doomed (aka ‘mental models rule everything around me’).One of the reasons for Switzerland’s abysmal performance seems to be a ‘business-as-usual’ mindset. We don’t seem to have clocked that this a crisis of enormous proportions.Once the starting assumption is “we can deal with this the way we deal with any other problem” everything downstream is destined to fail.The starting assumption ought to be much closer to “how can we mobilize everything we’ve got?” rather than “let’s do this in BAU [business as usual] channels”.That is also why we witnessed silly debates last summer over the cost of the contact tracing operations. Investing massively in contact tracing would have easily been one of the most cost-effective investments available.And yet it wasn’t done because, as far as I can tell, politicians and (some) media were worrying about cost — arguably because they were trapped in a ‘business-as-usual’ assumption where we assess governmental expenditures based on their efficiency.Whatever looks efficient in the short run in a crisis situation, is bound to end up introducing fragility.It’s as if you argued with the fire department and told them to load fewer fire hoses because that way you’d save on fuel while driving to where the fire is.The problem is this: if you want efficiency, you need to know what you’re optimising for. In a complex or chaotic situation there are, however, no stable variables you can optimise for. This isn’t the time for efficiency, what you’re looking for is redundancy and effectiveness.The first chapter in Stanley McChrystal et al.’s book “Team of Teams” is a good and highly accessible introduction to the foolishness of looking for efficiency in crisis situations.The ‘business-as-usual’ starting point is also how we get these odd debates over the roll-out of the vaccines. Of course we should be vaccinating outside of business hours, if not 24/7.The human and economic costs of any kind of lockdown are astronomical. Shortening that by even just a few days by going all-out on the vaccination effort will easily pay for itself, not to speak of the ethical/moral imperative to do so.3) There seems to be a shocking lack of ambition and imagination at the top of (some/many of) our political institutions.This relates to point 2 from above. If it wasn’t for Israel, to pick just one example, we’d probably still have people telling us that it simply isn’t possible to vaccinate this many people in such a short period of time.This point has also been made over and over by people like Tyler Cowen.It’s hard not to read this lack of ambition as a sign of decadence and decay. We’ve become so comfortable and sure of our own success, that we seem to be unable to mobilise our institutions — even in the face of an enormous crisis.It’s fine to fail at an impossible task — and that’s how the narrative is currently being spun (“we maybe didn’t cover ourselves in glory, but this was fundamentally unwinnable”).Trouble is, this isn’t/wasn’t impossible. Hard, yes, but doable. Lots of excuses will be made, but — hopefully — everyone will ultimately recognise these for what they are.4) The best way to think about media products is to understand their (economic) incentive structure.This is a trivial point. I used to think of media products as somehow being different from other goods/services. I’ve been thoroughly disabused of that notion.Most media products are probably best thought of as belonging to the category of general entertainment. There’s nothing inherently wrong with that, it just seems important to recognise.", "content_html": "Danny Buerkli, 03 March 2021
This short piece was originally published on Medium. It was originally a thread I wrote on Twitter in January 2021. It seemed to resonate quite a bit, which is why I’ve turned it into a lightly edited blog post.
I’ve been cleaning out my notes and wanted to share a short note with four things I’ve learned, I’ve been wrong about and I’ve been surprised by in the context of the pandemic.
The usual disclaimers apply: This is a short blog post (and formerly a Twitter thread), i.e. there isn’t a huge amount of nuance in what follows.
Switzerland is, by any sensible metric, doing badly in this pandemic. And yet this hasn’t filtered through to public discourse. It boggles the mind.
The logic seems to be “we usually do really well, therefore we must be doing well now”. Except we aren’t.
![\"\"](\"/assets/20210303_fourthingsivelearned_a.png\")
One of the reasons for Switzerland’s abysmal performance seems to be a ‘business-as-usual’ mindset. We don’t seem to have clocked that this a crisis of enormous proportions.
Once the starting assumption is “we can deal with this the way we deal with any other problem” everything downstream is destined to fail.
The starting assumption ought to be much closer to “how can we mobilize everything we’ve got?” rather than “let’s do this in BAU [business as usual] channels”.
That is also why we witnessed silly debates last summer over the cost of the contact tracing operations. Investing massively in contact tracing would have easily been one of the most cost-effective investments available.
And yet it wasn’t done because, as far as I can tell, politicians and (some) media were worrying about cost — arguably because they were trapped in a ‘business-as-usual’ assumption where we assess governmental expenditures based on their efficiency.
Whatever looks efficient in the short run in a crisis situation, is bound to end up introducing fragility.
It’s as if you argued with the fire department and told them to load fewer fire hoses because that way you’d save on fuel while driving to where the fire is.
The problem is this: if you want efficiency, you need to know what you’re optimising for. In a complex or chaotic situation there are, however, no stable variables you can optimise for. This isn’t the time for efficiency, what you’re looking for is redundancy and effectiveness.
The first chapter in Stanley McChrystal et al.’s book “Team of Teams” is a good and highly accessible introduction to the foolishness of looking for efficiency in crisis situations.
The ‘business-as-usual’ starting point is also how we get these odd debates over the roll-out of the vaccines. Of course we should be vaccinating outside of business hours, if not 24/7.
The human and economic costs of any kind of lockdown are astronomical. Shortening that by even just a few days by going all-out on the vaccination effort will easily pay for itself, not to speak of the ethical/moral imperative to do so.
This relates to point 2 from above. If it wasn’t for Israel, to pick just one example, we’d probably still have people telling us that it simply isn’t possible to vaccinate this many people in such a short period of time.
This point has also been made over and over by people like Tyler Cowen.
It’s hard not to read this lack of ambition as a sign of decadence and decay. We’ve become so comfortable and sure of our own success, that we seem to be unable to mobilise our institutions — even in the face of an enormous crisis.
It’s fine to fail at an impossible task — and that’s how the narrative is currently being spun (“we maybe didn’t cover ourselves in glory, but this was fundamentally unwinnable”).
Trouble is, this isn’t/wasn’t impossible. Hard, yes, but doable. Lots of excuses will be made, but — hopefully — everyone will ultimately recognise these for what they are.
This is a trivial point. I used to think of media products as somehow being different from other goods/services. I’ve been thoroughly disabused of that notion.
Most media products are probably best thought of as belonging to the category of general entertainment. There’s nothing inherently wrong with that, it just seems important to recognise.
", "url": "https://www.dannybuerkli.com/2021/03/03/four-things-ive-been-surprised-by", "image": "/assets/cards/essays/four-things-ive-been-surprised-by.png", "date_published": "2021-03-03T00:00:00-03:00", "date_modified": "2021-03-03T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2020/03/08/gradually-then-suddenly", "title": "„Gradually, then suddenly“ — four concepts that may help us think clearly about the coronavirus epidemic", "summary": "Four concepts for thinking clearly about the coronavirus epidemic while it still looked small — written in the first week of March 2020.", "content_text": "This essay was originally published on Medium (08 March 2020). This essay is also available, in lightly updated form, on the website of the New England Complex Systems Systems Institute (NECSI) A German translation has been published by reatch (“research and technology in switzerland”) Massimo Conte of the Complexity Education Project has published an Italian translation The current situation around the coronavirus epidemic is evolving rapidly. Switzerland, where I’m based, as of March 7 2020, has 228 confirmed cases. The overwhelming odds are that this number will increase rapidly and with it will the number of seriously affected individuals.In some of my social circles the most pressing worry is, however, not centered on what we ought to be doing to get this under control. For many people the overriding concern is seemingly not to be seen as worried or, god forbid, as “panicking”.From what I can tell this doesn’t derive from a sober analysis of the facts but from some sort of magical thinking which says that bad things haven’t happened here in a long time and therefore surely this cannot be bad.Mental models, like the one just mentioned above, are the things which help us think about the world. That’s why, in situations like these, it may be helpful to examine and challenge our mental models.So here are four basic concepts (taken from complex systems thinking) which can help us make sense of what is going on. They also help explain why epidemiologists and public health experts are as alarmed as they are.Three important caveats before we continue:First, I’m neither an epidemiologist nor a public health expert nor a certified expert on complex systems. For all I know most of what follows could be wrong. I might well also be missing one or more crucial considerations.Second, all of what follows below might be painfully obvious to you. I also claim no originality here. All of this is basic, none of it is novel. I draw my inspiration from complex systems thinking but these concepts can be found in other disciplines as well.Third,what follows is — of course — not some sort of complete description of how we should think about any epidemic.There is also always an obvious risk of embarrassing yourself when writing about a situation which evolves rapidly but here we are.So proceed with caution. Also, wash your hands and follow your local public health advice.Exponential growth: “yesterday everything was still under control?!”Italy went from one diagnosed case to well over 3’000 in just over a month (34 days) with only three cases in the first 21 days. It’s safe to say that most intuitively did not expect the number of diagnosed cases to grow so quickly.We are all familiar with the story of the inventor who asked to be modestly compensated by his emperor by receiving one grain of rice on the first square of the chessboard, two on the second square, four on the third square, and so on until all squares are filled. About halfway into the chessboard the emperor realizes that the request wasn’t as modest.The point is that exponential growth keeps surprising us. And for some reason we don’t seem to have a good intuitive handle on it.Ray Kurzweil coined the expression of the “second half of the chessboard”. In the first half the effects are large but potentially manageable. In the second half things spiral out of control.\"Everything is fine\" -> \"It's growing but we got this under control\" -> \"Oh...\"What does that mean?A small number of cases COVID-19 cases combined with an exponential growth rate may not, as we might assume, be a small problem. Absent an effective intervention which prevents further exponential growth a small number of cases may already spell deep trouble. Better avoid getting into the ‘second half of the chessboard’ (see also this thread by epidemiologist Marcel Salathé on this topic).Of course the exponential growth of coronavirus infections cannot literally go on forever because there is only a finite number of humans. It is also true that diseases do not usually spread at an exponential rate for the entire duration of an epidemic.The point is, however, that we will tend to underestimate the danger that an initial small number of cases poses because we struggle to imagine how quickly that small number can turn into a very large one.The current doubling time (as of March 6, excluding China) is three days. Every three days the number of known coronavirus cases doubles.Such exponential growth is why intervening early and heavily may well be justified. The logic with all undesirable things with exponential growth rates is the same: you want to nip it in the bud early. Early action may well be orders of magnitude cheaper and easier compared to reacting later.Phase changes: Everything is fine until it suddenly isn’tComplex systems tend to have ‘tipping points’ and go through ‘phase changes’ once they reach one of those. This means that once a system reaches a certain threshold things can change rapidly.Take the provision of healthcare in hospitals as an example. An epidemic starts and hospital beds start filling up. Everything is basically fine. We can provide adequate medical care to everyone who shows up and needs it. No need to overreact or introduce costly containment measures, right?Maybe, except that at one point we will hit a ‘tipping point’ and run out of hospital beds (or ventilators, or masks, or any other finite resource).Once that happens, things shift suddenly.Survival rates might go down because we cannot provide care anymore to the same standard, infection rates among healthcare staff go up because there aren’t enough masks to go around, etc.Imagine a bathtub that keeps filling up with water. Water flows in at a constant rate. The tub slowly fills. Everything is completely fine. Until at one point the tub starts to overflow. The fact that the tub hasn’t overflowed yet doesn’t tell us that we shouldn’t worry about it overflowing.The same logic also holds for supply chains. Many firms will have some amount of material in stock.The fact that they are still able to function for a period of time in the absence of continued deliveries from their suppliers tells us very little about how resilient the system really is.Everything can seem to continue just fine for a while until it suddenly starts to fall apart because one or more critical components are simply not available anymore. For the most part we should be able to know in advance where things will fall apart — though that doesn’t necessarily mean that we can do anything about it.Delayed feedback cycles: “why are things still getting worse even though we introduced all these measures?”Imagine how hard it would be to drive a car where there was a lag of just a few seconds between you turning the steering wheel and the car doing what you want it to do. It would be phenomenally annoying.(This is also why cooks prefer gas stoves over electric ones: an electric stove has a built-in delay between me turning the knob up or down and the temperature changing. With gas stoves there is no delay, the temperature rises or falls immediately.)This is the situation we are finding ourselves in, both in terms of detection of the disease and in terms of understanding the effects of measures taken in response to the epidemic.The cases that are being diagnosed today are people who got infected two weeks ago or so. The current count of diagnosed cases captures a past reality, not the present situation. It’s like looking at a star at night — because light takes time to travel to us you’re looking at the past.The delayed feedback loops hold for our understanding of whether the measures that are being taken are effective. Italy introduced school closures on March 04. Whether these will curb the number of new infections will be known only two weeks after that (and quite possibly even later).This makes decision-making very difficult for governments. One implication is that any restrictions that are put in place will likely be lifted slowly and gradually rather than all at once in order to avoid wild swings.Another implication, particularly when you put this together with exponential growth, is that you need to act early because the results of the intervention will only show with a delay.Leverage points: (relatively) small change, (relatively) large effectAnother feature of complex systems is that there usually exist ‚leverage points’, interventions which can create an outsized effect because of their position and influence in the system.This cuts both ways, of course.If you wanted to accelerate the spread of SARS-CoV-2 quickly and cheaply you’d be well advised to encourage mass gatherings.If, on the other hand, you want to do the opposite you’d encourage ‘social distancing’, increased hand washing as well as other personal hygiene measures. Because they are — hopefully — ‘leverage points’ these may well make a much bigger difference than we might intuitively believe.“Social distancing” and handwashing might be two examples of such leverage points. They are both comparatively easy and inexpensive to implement yet they might make a big difference.\"Death of Achilles\" by Rubens --- Paris found a leverage point and... you know how the story endsIf these four concepts happen to apply in this situation then we’re well advised to act swiftly because that may well save us from a lot of pain.I offer these thoughts in the spirit of Donella Meadows — a pioneer of systems thinking — who said: “Remember, always, that everything you know, and everything everyone knows, is only a model. Get your model out there where it can be viewed. Invite others to challenge your assumptions and add their own.”Further reading: Three people you should follow on Twitter on this are Marcel Salathé, Marcel Althaus, and Yaneer Bar-Yam. For an accessible general introduction to ‚system thinking’: Donella H. Meadows, Thinking in Systems — A Primer and Mark Foden’s “The clock and the cat” podcast For a highly visual and interactive explanation of disease dynamics: Kevin Simler’s “Going Critical” For a short explanation of ‚base rate thinking’ vs. ‚exponential thinking’: Tyler Cowen’s „growthers vs. base-raters” article on Bloomberg Opinion For a more general introduction on how complexity relates to policy making: Andrew Haldane’s 2012 speech „The dog and the frisbee” I owe a lot of thanks to all the individuals who took the time to read and comment on initial drafts. I asked people with expertise in epidemiology and other related fields to look over it in an attempt to avoid obvious mistakes. Any remaining errors are, obviously, mine.", "content_html": "This essay was originally published on Medium (08 March 2020).
This essay is also available, in lightly updated form, on the website of the New England Complex Systems Systems Institute (NECSI)
A German translation has been published by reatch (“research and technology in switzerland”)
Massimo Conte of the Complexity Education Project has published an Italian translation
The current situation around the coronavirus epidemic is evolving rapidly. Switzerland, where I’m based, as of March 7 2020, has 228 confirmed cases. The overwhelming odds are that this number will increase rapidly and with it will the number of seriously affected individuals.
In some of my social circles the most pressing worry is, however, not centered on what we ought to be doing to get this under control. For many people the overriding concern is seemingly not to be seen as worried or, god forbid, as “panicking”.
From what I can tell this doesn’t derive from a sober analysis of the facts but from some sort of magical thinking which says that bad things haven’t happened here in a long time and therefore surely this cannot be bad.
Mental models, like the one just mentioned above, are the things which help us think about the world. That’s why, in situations like these, it may be helpful to examine and challenge our mental models.
So here are four basic concepts (taken from complex systems thinking) which can help us make sense of what is going on. They also help explain why epidemiologists and public health experts are as alarmed as they are.
Three important caveats before we continue:
First, I’m neither an epidemiologist nor a public health expert nor a certified expert on complex systems. For all I know most of what follows could be wrong. I might well also be missing one or more crucial considerations.
Second, all of what follows below might be painfully obvious to you. I also claim no originality here. All of this is basic, none of it is novel. I draw my inspiration from complex systems thinking but these concepts can be found in other disciplines as well.
Third,what follows is — of course — not some sort of complete description of how we should think about any epidemic.
There is also always an obvious risk of embarrassing yourself when writing about a situation which evolves rapidly but here we are.
So proceed with caution. Also, wash your hands and follow your local public health advice.
Italy went from one diagnosed case to well over 3’000 in just over a month (34 days) with only three cases in the first 21 days. It’s safe to say that most intuitively did not expect the number of diagnosed cases to grow so quickly.
We are all familiar with the story of the inventor who asked to be modestly compensated by his emperor by receiving one grain of rice on the first square of the chessboard, two on the second square, four on the third square, and so on until all squares are filled. About halfway into the chessboard the emperor realizes that the request wasn’t as modest.
The point is that exponential growth keeps surprising us. And for some reason we don’t seem to have a good intuitive handle on it.
Ray Kurzweil coined the expression of the “second half of the chessboard”. In the first half the effects are large but potentially manageable. In the second half things spiral out of control.
![\"\"](\"/assets/20200308_graduallythensuddenly_a.png\")
What does that mean?
A small number of cases COVID-19 cases combined with an exponential growth rate may not, as we might assume, be a small problem. Absent an effective intervention which prevents further exponential growth a small number of cases may already spell deep trouble. Better avoid getting into the ‘second half of the chessboard’ (see also this thread by epidemiologist Marcel Salathé on this topic).
Of course the exponential growth of coronavirus infections cannot literally go on forever because there is only a finite number of humans. It is also true that diseases do not usually spread at an exponential rate for the entire duration of an epidemic.
The point is, however, that we will tend to underestimate the danger that an initial small number of cases poses because we struggle to imagine how quickly that small number can turn into a very large one.
The current doubling time (as of March 6, excluding China) is three days. Every three days the number of known coronavirus cases doubles.
Such exponential growth is why intervening early and heavily may well be justified. The logic with all undesirable things with exponential growth rates is the same: you want to nip it in the bud early. Early action may well be orders of magnitude cheaper and easier compared to reacting later.
Complex systems tend to have ‘tipping points’ and go through ‘phase changes’ once they reach one of those. This means that once a system reaches a certain threshold things can change rapidly.
Take the provision of healthcare in hospitals as an example. An epidemic starts and hospital beds start filling up. Everything is basically fine. We can provide adequate medical care to everyone who shows up and needs it. No need to overreact or introduce costly containment measures, right?
Maybe, except that at one point we will hit a ‘tipping point’ and run out of hospital beds (or ventilators, or masks, or any other finite resource).
Once that happens, things shift suddenly.
Survival rates might go down because we cannot provide care anymore to the same standard, infection rates among healthcare staff go up because there aren’t enough masks to go around, etc.
Imagine a bathtub that keeps filling up with water. Water flows in at a constant rate. The tub slowly fills. Everything is completely fine. Until at one point the tub starts to overflow. The fact that the tub hasn’t overflowed yet doesn’t tell us that we shouldn’t worry about it overflowing.
![\"\"](\"/assets/20200308_graduallythensuddenly_b.png\")
The same logic also holds for supply chains. Many firms will have some amount of material in stock.
The fact that they are still able to function for a period of time in the absence of continued deliveries from their suppliers tells us very little about how resilient the system really is.
Everything can seem to continue just fine for a while until it suddenly starts to fall apart because one or more critical components are simply not available anymore. For the most part we should be able to know in advance where things will fall apart — though that doesn’t necessarily mean that we can do anything about it.
Imagine how hard it would be to drive a car where there was a lag of just a few seconds between you turning the steering wheel and the car doing what you want it to do. It would be phenomenally annoying.
(This is also why cooks prefer gas stoves over electric ones: an electric stove has a built-in delay between me turning the knob up or down and the temperature changing. With gas stoves there is no delay, the temperature rises or falls immediately.)
This is the situation we are finding ourselves in, both in terms of detection of the disease and in terms of understanding the effects of measures taken in response to the epidemic.
The cases that are being diagnosed today are people who got infected two weeks ago or so. The current count of diagnosed cases captures a past reality, not the present situation. It’s like looking at a star at night — because light takes time to travel to us you’re looking at the past.
![\"\"](\"/assets/20200308_graduallythensuddenly_c.png\")
The delayed feedback loops hold for our understanding of whether the measures that are being taken are effective. Italy introduced school closures on March 04. Whether these will curb the number of new infections will be known only two weeks after that (and quite possibly even later).
This makes decision-making very difficult for governments. One implication is that any restrictions that are put in place will likely be lifted slowly and gradually rather than all at once in order to avoid wild swings.
Another implication, particularly when you put this together with exponential growth, is that you need to act early because the results of the intervention will only show with a delay.
Another feature of complex systems is that there usually exist ‚leverage points’, interventions which can create an outsized effect because of their position and influence in the system.
This cuts both ways, of course.
If you wanted to accelerate the spread of SARS-CoV-2 quickly and cheaply you’d be well advised to encourage mass gatherings.
If, on the other hand, you want to do the opposite you’d encourage ‘social distancing’, increased hand washing as well as other personal hygiene measures. Because they are — hopefully — ‘leverage points’ these may well make a much bigger difference than we might intuitively believe.
“Social distancing” and handwashing might be two examples of such leverage points. They are both comparatively easy and inexpensive to implement yet they might make a big difference.
![\"\"](\"/assets/20200308_graduallythensuddenly_d.png\")
If these four concepts happen to apply in this situation then we’re well advised to act swiftly because that may well save us from a lot of pain.
I offer these thoughts in the spirit of Donella Meadows — a pioneer of systems thinking — who said: “Remember, always, that everything you know, and everything everyone knows, is only a model. Get your model out there where it can be viewed. Invite others to challenge your assumptions and add their own.”
Further reading:
Three people you should follow on Twitter on this are Marcel Salathé, Marcel Althaus, and Yaneer Bar-Yam.
For an accessible general introduction to ‚system thinking’: Donella H. Meadows, Thinking in Systems — A Primer and Mark Foden’s “The clock and the cat” podcast
For a highly visual and interactive explanation of disease dynamics: Kevin Simler’s “Going Critical”
For a short explanation of ‚base rate thinking’ vs. ‚exponential thinking’: Tyler Cowen’s „growthers vs. base-raters” article on Bloomberg Opinion
For a more general introduction on how complexity relates to policy making: Andrew Haldane’s 2012 speech „The dog and the frisbee”
I owe a lot of thanks to all the individuals who took the time to read and comment on initial drafts. I asked people with expertise in epidemiology and other related fields to look over it in an attempt to avoid obvious mistakes. Any remaining errors are, obviously, mine.
", "url": "https://www.dannybuerkli.com/2020/03/08/gradually-then-suddenly", "image": "/assets/cards/essays/gradually-then-suddenly.png", "tags": ["complexity","models"], "date_published": "2020-03-08T00:00:00-03:00", "date_modified": "2020-03-08T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2020/01/30/what-i-think-we-do-at-cpi", "title": "What I think we do at the Centre for Public Impact", "summary": "An exploration of how we can think about change in complex systems", "content_text": "This article was published on the Centre for Public Impact’s Medium page (30 January 2020).This is my last week at the Centre for Public Impact after five wonderful years. I feel privileged to have been part of an organisation with such an ambitious purpose.Part of the challenge of being in an organisation with a remit as broad as “shaping a new future for government” is that it can be difficult to wrap your head around what exactly you’re doing and why.As I’ve been reflecting over the past couple of weeks on my time here I came across a post on Medium by Cassie Robinson which introduced me to the Berkana Institute’s Two Loop model.The model resonated with me because it encapsulates what our work at CPI is about when we are at our best and why it all matters.The Berkana Institute’s Two Loop modelIn Cassie’s words, the model shows a dominant system that is dying, and an emergent system that has the potential to become the system of influence. As the dominant system reaches its peak, new pioneers emerge (1), recognising that the dominant system (however impossible and far away that might seem) is beginning to decline.In order to bring the Emergent System into the world, it is important that it is named and that the pioneers, the people and organisations building alternatives are connected together (2), and the work they are doing, illuminated.Thinking of the world we move in and of our work in the way of this model clarified a lot for me.We are indeed working — together with and alongside an entire community of changemakers, thinkers, and practitioners — towards bringing a new paradigm into being.Our “Manifesto for Better Government” is an explicit attempt to name this emergent system. Much of our work is equally about connecting people and organisations who are building alternatives to the current model of public administration. And, finally, many of the articles we publish on our website serve to illuminate the work of those pioneers.The Two Loop Model doesn’t tell us what to do or how to do it but it provides a useful way of thinking about the quest that we’re on. It clarifies the choices we face and the decisions we need to make in order to be effective.CPI might play a role as one of the “midwives” who do the naming, connecting, and illuminating — alongside and in partnership with many of the other excellent individuals and organisations who are working towards the same goal. At the same time, CPI might also have a role on the “hospice” side of things, helping the Dominant System die with grace and compassion.While I won’t be part of CPI anymore I’ll remain firmly in the community of people and organisations who are working towards a new paradigm in government. After a two-month sabbatical I’ll be joining staatslabor in Switzerland, a government reform lab I co-founded three years ago.I can’t wait to get started there and to continue working on this emerging paradigm for government which we at CPI — and many others — have begun to articulate and are working towards.", "content_html": "This article was published on the Centre for Public Impact’s Medium page (30 January 2020).
This is my last week at the Centre for Public Impact after five wonderful years. I feel privileged to have been part of an organisation with such an ambitious purpose.
Part of the challenge of being in an organisation with a remit as broad as “shaping a new future for government” is that it can be difficult to wrap your head around what exactly you’re doing and why.
As I’ve been reflecting over the past couple of weeks on my time here I came across a post on Medium by Cassie Robinson which introduced me to the Berkana Institute’s Two Loop model.
The model resonated with me because it encapsulates what our work at CPI is about when we are at our best and why it all matters.
![\"The](\"/assets/20200130_Berkana_TwoLoopModel.png\")
The Berkana Institute’s Two Loop model
In Cassie’s words, the model
shows a dominant system that is dying, and an emergent system that has the potential to become the system of influence. As the dominant system reaches its peak, new pioneers emerge (1), recognising that the dominant system (however impossible and far away that might seem) is beginning to decline.
In order to bring the Emergent System into the world, it is important that it is
named and that the pioneers, the people and organisations building alternatives are connected together (2), and the work they are doing, illuminated.
Thinking of the world we move in and of our work in the way of this model clarified a lot for me.
We are indeed working — together with and alongside an entire community of changemakers, thinkers, and practitioners — towards bringing a new paradigm into being.
Our “Manifesto for Better Government” is an explicit attempt to name this emergent system. Much of our work is equally about connecting people and organisations who are building alternatives to the current model of public administration. And, finally, many of the articles we publish on our website serve to illuminate the work of those pioneers.
The Two Loop Model doesn’t tell us what to do or how to do it but it provides a useful way of thinking about the quest that we’re on. It clarifies the choices we face and the decisions we need to make in order to be effective.
CPI might play a role as one of the “midwives” who do the naming, connecting, and illuminating — alongside and in partnership with many of the other excellent individuals and organisations who are working towards the same goal. At the same time, CPI might also have a role on the “hospice” side of things, helping the Dominant System die with grace and compassion.
While I won’t be part of CPI anymore I’ll remain firmly in the community of people and organisations who are working towards a new paradigm in government. After a two-month sabbatical I’ll be joining staatslabor in Switzerland, a government reform lab I co-founded three years ago.
I can’t wait to get started there and to continue working on this emerging paradigm for government which we at CPI — and many others — have begun to articulate and are working towards.
", "url": "https://www.dannybuerkli.com/2020/01/30/what-i-think-we-do-at-cpi", "image": "/assets/cards/essays/what-i-think-we-do-at-cpi.png", "tags": ["complexity","emergence"], "date_published": "2020-01-30T00:00:00-03:00", "date_modified": "2020-01-30T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2019/09/04/mistakes-to-avoid", "title": "Easy-to-avoid mistakes which will diminish your chances when applying for a job", "summary": "Practical advice for jobseekers", "content_text": "This article was originally published on the Centre for Public Impact’s Medium page (04 September 2019).I have been involved in a few recruitment rounds over the past years. There are recurring, easy-to-avoid mistakes people make which unnecessarily but very significantly diminish their chances.Sometimes those mistakes correlate with a lack of competence. But a lot of times it seems to be more due to a #hiddencurriculum issue. People may have never had the benefit of being told these things (or of being in an environment where these are taught as self-evident).A few remarks upfront:I’ve seen all of the things listed below happen a lot. These aren’t cherry-picked. All quotes are made up.Let me be super clear about my intention: I want people to avoid these traps so they can shine and show us what they’re capable of.We invest a lot of energy at @CPI_foundation into trying to hire in a fair and unbiased way. That’s why we use the @beapplied platform (which lets us review answers to work-related questions anonymously rather than review CVs and cover letters) and structured interviews.That’s why we care about how people perform once they decide to submit an application with us. We want to get the best candidates and that might well include people who don’t come across in the best possible way for reasons that don’t correlate with competence.For applications that we run through the @beapplied platform, we ask applicants practical questions (e.g. “how would you handle situation x?”). We then review the anonymized written answers. The way we do this is by reviewing all responses to question 1, then all those to question 2, etc.This is quite different from the usual CV plus cover letter system. While most items below apply universally in any job application or interview situation, some are more specific to systems that use anonymized written answers.With all of that out of the way, here we go: Always submit your CV/resume as a PDF, never as a Word file. Word files look unpolished and the formatting often breaks. This is a small thing but do not underestimate just how big of a difference it makes. Do not reveal confidential information about your current or past employers. Don’t. Resist the temptation. It’s done to build a connection and impress but it does the opposite. It immediately makes me worry about what you’ll say about us once you leave. Be kind when you talk about your current or former colleagues. There are ways of talking about difficult situations that don’t sound like you’re slagging off a colleague. Be specific about your contributions when asked. If you’re asked what your contribution to a team was, make sure we understand what you did. Don’t be shy, it’s not bragging, we want to know! Answer the question. If we ask how you would go about doing x, tell us how you would go about doing x. Don’t answer a different question. Keep it simple. Avoid jargon and abbreviations. This is particularly important in writing where we can’t ask clarifying questions. If you’re telling me about how “dealing with the IATR process required NFC buy-in” I will have no idea what you’re talking about. Practical tip: for “tell me about a time when you did x” questions use the ‘STAR’ framework: Situation, Task, Action, Result. Works like a charm because makes sure that you give a clear answer. It also helps because it makes you tell a story rather than just list facts. In written answers (I/III): check your grammar. Use the active voice (“results had to be communicated” -> “we/I communicated the results”). Use a clear structure (paragraphs, bullet points, section titles, …). In written answers (II/III): play it straight. Humour and/or quirky responses rarely work. You can be as witty as you like in person but it often comes across as ill-judged in text. In written answers (III/III): keep it short. Don’t use more words than necessary. We are often impressed by concise answers (both spoken and in writing) that do the job. It demonstrates your ability to think and communicate clearly. Specifically for answers on the @beapplied platform: avoid cross-referencing between your answers (“as I’ve mentioned above…”). Because we batch review all responses to the same question together and because it’s all anonymized we won’t know what you’re referring to. On your CV/resume: use a straight-forward email address (usually some combination of your first and last name), avoid anything that isn’t that (e.g. “joebrumby96@gmail.com”). Outside of a formal job application process, when asking for an informational interview or a ‘coffee chat’: avoid emailing multiple people on the team separately. Even if you customise your emails, don’t forget that we tend to sit next to each other and we talk 😃 That’s it for now. This isn’t a comprehensive list and it’s always hard to get all the nuance across in short posts but hopefully, this is of use to some out there. If you recognize having done some of these things yourself don’t worry too much, I’ve certainly done the same! :)This post started out as a thread on Twitter. It seemed to resonate and I’ve turned it into a Medium post to make it easier to refer to it.I was inspired to share this by @jhaushofer (who recently shared a #hiddencurriculum doc he gives to Research Assistants and students: https://twitter.com/jhaushofer/status/1156853146985086976…) and by @patio11 who made a different-but-related point that there’s lots of value in capturing implicit knowledge and making it explicit (https://twitter.com/patio11/status/1069422308526239744…)Many thanks to my colleagues @MargotGagliani, @_AdrianBrown, and John Burgoyne who helped me pull this together.", "content_html": "This article was originally published on the Centre for Public Impact’s Medium page (04 September 2019).
I have been involved in a few recruitment rounds over the past years. There are recurring, easy-to-avoid mistakes people make which unnecessarily but very significantly diminish their chances.
Sometimes those mistakes correlate with a lack of competence. But a lot of times it seems to be more due to a #hiddencurriculum issue. People may have never had the benefit of being told these things (or of being in an environment where these are taught as self-evident).
A few remarks upfront:
I’ve seen all of the things listed below happen a lot. These aren’t cherry-picked. All quotes are made up.
Let me be super clear about my intention: I want people to avoid these traps so they can shine and show us what they’re capable of.
We invest a lot of energy at @CPI_foundation into trying to hire in a fair and unbiased way. That’s why we use the @beapplied platform (which lets us review answers to work-related questions anonymously rather than review CVs and cover letters) and structured interviews.
That’s why we care about how people perform once they decide to submit an application with us. We want to get the best candidates and that might well include people who don’t come across in the best possible way for reasons that don’t correlate with competence.
For applications that we run through the @beapplied platform, we ask applicants practical questions (e.g. “how would you handle situation x?”). We then review the anonymized written answers. The way we do this is by reviewing all responses to question 1, then all those to question 2, etc.
This is quite different from the usual CV plus cover letter system. While most items below apply universally in any job application or interview situation, some are more specific to systems that use anonymized written answers.
With all of that out of the way, here we go:
Always submit your CV/resume as a PDF, never as a Word file. Word files look unpolished and the formatting often breaks. This is a small thing but do not underestimate just how big of a difference it makes.
Do not reveal confidential information about your current or past employers. Don’t. Resist the temptation. It’s done to build a connection and impress but it does the opposite. It immediately makes me worry about what you’ll say about us once you leave.
Be kind when you talk about your current or former colleagues. There are ways of talking about difficult situations that don’t sound like you’re slagging off a colleague.
Be specific about your contributions when asked. If you’re asked what your contribution to a team was, make sure we understand what you did. Don’t be shy, it’s not bragging, we want to know!
Answer the question. If we ask how you would go about doing x, tell us how you would go about doing x. Don’t answer a different question.
Keep it simple. Avoid jargon and abbreviations. This is particularly important in writing where we can’t ask clarifying questions. If you’re telling me about how “dealing with the IATR process required NFC buy-in” I will have no idea what you’re talking about.
Practical tip: for “tell me about a time when you did x” questions use the ‘STAR’ framework: Situation, Task, Action, Result. Works like a charm because makes sure that you give a clear answer. It also helps because it makes you tell a story rather than just list facts.
In written answers (I/III): check your grammar. Use the active voice (“results had to be communicated” -> “we/I communicated the results”). Use a clear structure (paragraphs, bullet points, section titles, …).
In written answers (II/III): play it straight. Humour and/or quirky responses rarely work. You can be as witty as you like in person but it often comes across as ill-judged in text.
In written answers (III/III): keep it short. Don’t use more words than necessary. We are often impressed by concise answers (both spoken and in writing) that do the job. It demonstrates your ability to think and communicate clearly.
Specifically for answers on the @beapplied platform: avoid cross-referencing between your answers (“as I’ve mentioned above…”). Because we batch review all responses to the same question together and because it’s all anonymized we won’t know what you’re referring to.
On your CV/resume: use a straight-forward email address (usually some combination of your first and last name), avoid anything that isn’t that (e.g. “joebrumby96@gmail.com”).
Outside of a formal job application process, when asking for an informational interview or a ‘coffee chat’: avoid emailing multiple people on the team separately. Even if you customise your emails, don’t forget that we tend to sit next to each other and we talk 😃
That’s it for now. This isn’t a comprehensive list and it’s always hard to get all the nuance across in short posts but hopefully, this is of use to some out there. If you recognize having done some of these things yourself don’t worry too much, I’ve certainly done the same! :)
This post started out as a thread on Twitter. It seemed to resonate and I’ve turned it into a Medium post to make it easier to refer to it.
I was inspired to share this by @jhaushofer (who recently shared a #hiddencurriculum doc he gives to Research Assistants and students: https://twitter.com/jhaushofer/status/1156853146985086976…) and by @patio11 who made a different-but-related point that there’s lots of value in capturing implicit knowledge and making it explicit (https://twitter.com/patio11/status/1069422308526239744…)
Many thanks to my colleagues @MargotGagliani, @_AdrianBrown, and John Burgoyne who helped me pull this together.
", "url": "https://www.dannybuerkli.com/2019/09/04/mistakes-to-avoid", "image": "/assets/cards/essays/mistakes-to-avoid.png", "date_published": "2019-09-04T00:00:00-03:00", "date_modified": "2019-09-04T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2019/08/23/a-tale-of-two-paradigms", "title": "A tale of two paradigms — meet the ideas which changing public administrations around the world", "summary": "We are currently witnessing in a kind of ‘quiet revolution’ the emergence of a new paradigm in public administration. It’s one that relies on shared power rather than control.", "content_text": "This essay was originally published on the Centre for Public Impact’s Medium page (23 August 2019). “Most people imagine that the present style of management has always existed, and is a fixture. Actually, it is a modern invention — a prison created by the way in which people interact.” W. Edwards DemingWe need paradigms to understand the world around us. There is no such thing as “simply seeing what’s in front of us” or “just using common sense”. We always require some frame through which to interpret what is going on.One such paradigm —which exerts much influence in business just as in public administration — is managerialism, the belief that all kinds of problems, irrespective of their context, can be solved by applying control, hierarchy, evidence, and “best practice”.While the problems with managerialism have been apparent for a long time, it still exerts a strong influence on our imagination, including in the public sector. Unfortunately, managerialism is, however, not well adapted to many of the challenges public administrations face.To understand why that’s the case it’s important to understand the historical context from which this paradigm emerged.The origins of managerialismLet’s go back to the England of Queen Victoria. Before Frederick Winslow Taylor introduced the idea of “scientific management” in the late 19th century, every artisan would have their own way of, say, weaving a piece of cloth.Taylor broke up any activity into its constituent parts, figured out The One Best Wayfor each and did away with all that variation. Once the ideal, e.g. most time-saving way of picking up a hammer had been found, Taylor made sure no worker deviated from it.Taylor’s method of so-called “scientific management” allowed firms to literally do more with less. Around the year 1900, the time required to produce a cannon projectile dropped from ten hours to a mere ninety minutes.Today, this way of optimizing production seems so obvious that it’s hard to imagine a world in which we would organize things differently. But to the artisans of the late 19th century, the introduction of these “best practices” felt anything but obvious. Workers who thought of themselves as skilled individuals felt offended by the fact that they were reduced to being human cogs in an optimized machine.Optimize this! (photo by Ant Rozetsky on Unsplash)Taylor’s peers, on the other hand, were deeply impressed. Henri Le Châtelier, a chemist, praised Taylor’s method as “a model which every young engineer will have to study”. Many took this advice and studied Taylor’s ideas and those of his contemporaries — like Henry L. Gantt, of Gantt chart fame — and collectively built the modern discipline of management.Times have changed but Taylor’s influence still looms large — “put the pen in your left shirt pocket”Management thinking has, of course, evolved a great deal since the early 20th century. But in many ways, Taylor’s insights are still as influential and useful as ever. Amazon’s close monitoring of its fulfillment center workers using wristbands or the way in which UPS instructs its drivers in which shirt pocket to put their pens (“right-handed people should use the left pocket, and vice versa”) is indistinguishable from how Taylor optimized steel plants.Taylor’s fundamental insight was that people we would come to call “managers” could deconstruct activities into their constituent parts (e.g. how a delivery driver pulls out his pen to get a receipt signed) and would identify The One Best Way to do it.\"I've calculated that putting your pen in your left pocket will save us three seconds\" (photo by Jude Beck on Unsplash)This method of optimising work excels, but only under a rather specific set of circumstances. It works when we have clear objectives, when those objectives are measurable, when we understand what drives performance, when things work in linear and deterministic ways and, finally, when none of these things change.A different way of saying it is that Taylor’s practices work in situations that are complicated but not complex. But what does that mean?Interlude: what is the difference between “complicated” and “complex”?Something that’s complex isn’t just something that’s very complicated. The two concepts are entirely distinct. Here’s a simple yet accurate analogy which I owe to Mark Foden (once you’re done reading you should click through and listen to his podcast):⏰Complicated things are like clocks. You can take one apart, study the many pieces, understand how they interlock and from that learn how a clock works. Clocks operate in predetermined ways.Clock = complicated (photo by Sonja Langford on Unsplash)🐈 Complex things are like cats. They don’t work in predetermined ways and opening up a cat will get you only a limited (and gory) insight into why it behaves the way it does. Cats are organisms and their behaviour emerges from interactions between what’s happening inside a cat as well as between that and a cat’s environment.Or, in the words of Nassim Nicholas Taleb, “the main idea behind complex systems is that the ensemble behaves in way not predicted by the components. The interactions matter more than the nature of the units.”Cat = complex (photo by Borna Bevanda on Unsplash)Any social benefits system, for example, has aspects of both “clock” and “cat”. The transactional parts that deal with moving money from government to beneficiaries are complicated, i.e. clock-like. The aspects of the system that aim to change beneficiaries’ behaviour are complex, i.e. cat-like.Much of what public administrations deal with is complex, not complicatedThere are many bits of government that are predominantly transactional, relatively unchanging and complicated. These are activities that may have easily measurable outcomes with clearly understood mechanisms behind them.Printing passports might be one example. This doesn’t mean that running a well-functioning passport office is easy — it is not. It simply means that managerial practices will work. By using targets, KPIs, best practices, and a strict distinction between managers and workers we stand a reasonable chance of achieving a well-functioning passport office.Printing passports --- complicated but, for the most part, not complex (Photo by Kelly Sikkema on Unsplash)There is, however, a wide and important range of things governments do that are predominantly complex.Think, for example, of education or healthcare. How a child learns or how patients maintain or regain their health isn’t a process that can be understood by closely looking at its different parts in isolation.These outcomes can’t be “managed” or “produced” in any meaningful sense of the word. Instead, it is better to think of them as emergent properties of complex systems.Bringing “clock” practices to bear on “cat” problems is a bit like trying to get a flower to bloom by pulling at leaves instead of tending to the surrounding conditions the plant needs. At best it’s ineffective, at worst it’s destructive.Emergency room physicians at a medical center in Oregon, USA recently complained about their hospital administrators refusing to admit very ill patients. What had happened? The management was under pressure to improve the hospital’s performance metrics. Those metrics are negatively impacted by people dying in the hospital. Entirely predictably the hospital started sending critically ill patients away.There exist an enormous number of examples where public administrations attempt to solve “cat” problems with “clock” practices. The problems than then promptly ensue are sadly entirely predictable and they are the norm, not the exception.How should we go about changing this?“If you’re in a hole stop digging”, learn from what’s already here, and lead by exampleIt’s time to replace or at least complement the dominant managerial paradigm with something that’s more adapted to complexity.1) The First Rule of Holes: “If you’re in a hole, stop digging”No amount of tweaking and refining will make practices borne out of complicatedness fit for complexity. If you’re in a hole, you need to stop digging. A different shovel made of sophisticated materials isn’t going to help. Doing less of the wrong thing is not the same as doing the right thing. This might mean, for example, not having fewer targets but having no targets.2) We should learn from what’s already hereIt’s important to remember that none of this is fundamentally new. Elinor Ostrom won a Nobel Prize for documenting some of these mechanisms in her 1991 book “Governing the Commons”.James C. Scott described the importance of tacit knowledge in complex situations his 1999 classic “Seeing Like a State” and Herbert Simon wrote in 1969: “when we come to the design of systems as complex as cities, or buildings, or economies, we must give up the aim of creating systems that will optimize some hypothesized utility function”.We should be careful not to reinvent the wheel and to learn as much as possible from the pioneers who have worked on these questions for a long time.3) “You can’t reason someone out of something they haven’t been reasoned into” — that’s why we need to lead by exampleParadigms don’t change easily. The old ones usually put up a fight before we all collectively shift to a new paradigm. Too much — companies, careers, reputations — is at risk for this to happen quickly. Also, when they’re strong enough we forget that they exist in the first place. Established practices become “common sense” and, as a result, are impervious to criticism.That’s why it’s almost impossible to win a theoretical argument against the dominant paradigm. The way forward then is to proceed by practical action, to demonstrate the effectiveness of complexity-conscious ways of working, and to champion the many examples that already exist out there.The emerging contours of a complexity-conscious paradigm for public administrationWe are currently witnessing in a kind of ‘quiet revolution’ the emergence of a new, competing paradigm. It’s one that relies on shared power rather than control.Buurtzorg, a Dutch home care organisation, for example, delivers outstanding care with no managers and no targets (see our detailed case study here). Wigan Council in the north of England implemented The Deal, a radically holistic, ethnographically based approach to public service delivery.Movements like the Losing Control Network or OneTeamGov are working in vastly different contexts but are gaining traction with their ideas that are highly attuned to complexity.The same themes are also resonating in the development world. Dan Honig, Professor at Johns Hopkins SAIS, used a database of over 14’000 development projects to show that “tight controls” and a “narrow focus on reaching pre-set targets” undermined the performance of foreign aid projects in complex environments. The Doing Development Differently movement has been arguing for a more complexity-informed approach for years now.All progress happens at the periphery first and then moves, over time, into the mainstream. This will not be any different. While we don’t yet exactly know what this new paradigm will look like we can discern some emerging contours.In our recent discussion paper on the Shared Power Principle, we describe four patterns that characterize this paradigm: Subsidiarity as a guide: putting decision-making power into the hands of people or organisations with the greatest knowledge of an issue and helping them exercise this power as effectively as possible. Relationships first: embracing human complexity and acknowledging that we should not seek solutions with the “average” citizen in mind. We need more bespoke approaches, which require deep and trusted relationships that go beyond mere transactions. New forms of governance and leadership: fostering the emergence of more diverse accountability structures that give people time and space to make and discuss decisions. Building a culture of continuous learning: creating a culture of continuous learning within teams and across different organisations. The paper contains much more detail on all of these. Much remains to be discovered and figured out but it is encouraging to see so much movement and experimentation.One day this new, complexity-informed paradigm will no doubt seem blindingly obvious, just like the managerial paradigm of efficiency and “best practice” feels obvious right now. Until then we need to keep chipping away at the old paradigm and leading by example to bring about a new one.Special thanks go to my colleagues Lena Kuenkel, Adrian Brown, John Burgoyne, and Bridget Nurre Jennions for their help with this post.", "content_html": "This essay was originally published on the Centre for Public Impact’s Medium page (23 August 2019).
“Most people imagine that the present style of management has always existed, and is a fixture. Actually, it is a modern invention — a prison created by the way in which people interact.” W. Edwards Deming
We need paradigms to understand the world around us. There is no such thing as “simply seeing what’s in front of us” or “just using common sense”. We always require some frame through which to interpret what is going on.
One such paradigm —which exerts much influence in business just as in public administration — is managerialism, the belief that all kinds of problems, irrespective of their context, can be solved by applying control, hierarchy, evidence, and “best practice”.
While the problems with managerialism have been apparent for a long time, it still exerts a strong influence on our imagination, including in the public sector. Unfortunately, managerialism is, however, not well adapted to many of the challenges public administrations face.
To understand why that’s the case it’s important to understand the historical context from which this paradigm emerged.
Let’s go back to the England of Queen Victoria. Before Frederick Winslow Taylor introduced the idea of “scientific management” in the late 19th century, every artisan would have their own way of, say, weaving a piece of cloth.
Taylor broke up any activity into its constituent parts, figured out The One Best Wayfor each and did away with all that variation. Once the ideal, e.g. most time-saving way of picking up a hammer had been found, Taylor made sure no worker deviated from it.
Taylor’s method of so-called “scientific management” allowed firms to literally do more with less. Around the year 1900, the time required to produce a cannon projectile dropped from ten hours to a mere ninety minutes.
Today, this way of optimizing production seems so obvious that it’s hard to imagine a world in which we would organize things differently. But to the artisans of the late 19th century, the introduction of these “best practices” felt anything but obvious. Workers who thought of themselves as skilled individuals felt offended by the fact that they were reduced to being human cogs in an optimized machine.
![\"steel](\"/assets/20190823_two_paradigms_a.jpeg\")
Taylor’s peers, on the other hand, were deeply impressed. Henri Le Châtelier, a chemist, praised Taylor’s method as “a model which every young engineer will have to study”. Many took this advice and studied Taylor’s ideas and those of his contemporaries — like Henry L. Gantt, of Gantt chart fame — and collectively built the modern discipline of management.
Management thinking has, of course, evolved a great deal since the early 20th century. But in many ways, Taylor’s insights are still as influential and useful as ever. Amazon’s close monitoring of its fulfillment center workers using wristbands or the way in which UPS instructs its drivers in which shirt pocket to put their pens (“right-handed people should use the left pocket, and vice versa”) is indistinguishable from how Taylor optimized steel plants.
Taylor’s fundamental insight was that people we would come to call “managers” could deconstruct activities into their constituent parts (e.g. how a delivery driver pulls out his pen to get a receipt signed) and would identify The One Best Way to do it.
![\"stopwatch\"](\"/assets/20190823_two_paradigms_b.jpeg\")
This method of optimising work excels, but only under a rather specific set of circumstances. It works when we have clear objectives, when those objectives are measurable, when we understand what drives performance, when things work in linear and deterministic ways and, finally, when none of these things change.
A different way of saying it is that Taylor’s practices work in situations that are complicated but not complex. But what does that mean?
Something that’s complex isn’t just something that’s very complicated. The two concepts are entirely distinct. Here’s a simple yet accurate analogy which I owe to Mark Foden (once you’re done reading you should click through and listen to his podcast):
⏰Complicated things are like clocks. You can take one apart, study the many pieces, understand how they interlock and from that learn how a clock works. Clocks operate in predetermined ways.
![\"clock\"](\"/assets/20190823_two_paradigms_c.jpeg\")
🐈 Complex things are like cats. They don’t work in predetermined ways and opening up a cat will get you only a limited (and gory) insight into why it behaves the way it does. Cats are organisms and their behaviour emerges from interactions between what’s happening inside a cat as well as between that and a cat’s environment.
Or, in the words of Nassim Nicholas Taleb, “the main idea behind complex systems is that the ensemble behaves in way not predicted by the components. The interactions matter more than the nature of the units.”
![\"cat\"](\"/assets/20190823_two_paradigms_d.jpeg\")
Any social benefits system, for example, has aspects of both “clock” and “cat”. The transactional parts that deal with moving money from government to beneficiaries are complicated, i.e. clock-like. The aspects of the system that aim to change beneficiaries’ behaviour are complex, i.e. cat-like.
There are many bits of government that are predominantly transactional, relatively unchanging and complicated. These are activities that may have easily measurable outcomes with clearly understood mechanisms behind them.
Printing passports might be one example. This doesn’t mean that running a well-functioning passport office is easy — it is not. It simply means that managerial practices will work. By using targets, KPIs, best practices, and a strict distinction between managers and workers we stand a reasonable chance of achieving a well-functioning passport office.
![\"passport\"](\"/assets/20190823_two_paradigms_e.jpeg\")
There is, however, a wide and important range of things governments do that are predominantly complex.
Think, for example, of education or healthcare. How a child learns or how patients maintain or regain their health isn’t a process that can be understood by closely looking at its different parts in isolation.
These outcomes can’t be “managed” or “produced” in any meaningful sense of the word. Instead, it is better to think of them as emergent properties of complex systems.
Bringing “clock” practices to bear on “cat” problems is a bit like trying to get a flower to bloom by pulling at leaves instead of tending to the surrounding conditions the plant needs. At best it’s ineffective, at worst it’s destructive.
Emergency room physicians at a medical center in Oregon, USA recently complained about their hospital administrators refusing to admit very ill patients. What had happened? The management was under pressure to improve the hospital’s performance metrics. Those metrics are negatively impacted by people dying in the hospital. Entirely predictably the hospital started sending critically ill patients away.
There exist an enormous number of examples where public administrations attempt to solve “cat” problems with “clock” practices. The problems than then promptly ensue are sadly entirely predictable and they are the norm, not the exception.
How should we go about changing this?
It’s time to replace or at least complement the dominant managerial paradigm with something that’s more adapted to complexity.
No amount of tweaking and refining will make practices borne out of complicatedness fit for complexity. If you’re in a hole, you need to stop digging. A different shovel made of sophisticated materials isn’t going to help. Doing less of the wrong thing is not the same as doing the right thing. This might mean, for example, not having fewer targets but having no targets.
It’s important to remember that none of this is fundamentally new. Elinor Ostrom won a Nobel Prize for documenting some of these mechanisms in her 1991 book “Governing the Commons”.
James C. Scott described the importance of tacit knowledge in complex situations his 1999 classic “Seeing Like a State” and Herbert Simon wrote in 1969:
“when we come to the design of systems as complex as cities, or buildings, or economies, we must give up the aim of creating systems that will optimize some hypothesized utility function”.
We should be careful not to reinvent the wheel and to learn as much as possible from the pioneers who have worked on these questions for a long time.
Paradigms don’t change easily. The old ones usually put up a fight before we all collectively shift to a new paradigm. Too much — companies, careers, reputations — is at risk for this to happen quickly. Also, when they’re strong enough we forget that they exist in the first place. Established practices become “common sense” and, as a result, are impervious to criticism.
That’s why it’s almost impossible to win a theoretical argument against the dominant paradigm. The way forward then is to proceed by practical action, to demonstrate the effectiveness of complexity-conscious ways of working, and to champion the many examples that already exist out there.
We are currently witnessing in a kind of ‘quiet revolution’ the emergence of a new, competing paradigm. It’s one that relies on shared power rather than control.
Buurtzorg, a Dutch home care organisation, for example, delivers outstanding care with no managers and no targets (see our detailed case study here). Wigan Council in the north of England implemented The Deal, a radically holistic, ethnographically based approach to public service delivery.
Movements like the Losing Control Network or OneTeamGov are working in vastly different contexts but are gaining traction with their ideas that are highly attuned to complexity.
The same themes are also resonating in the development world. Dan Honig, Professor at Johns Hopkins SAIS, used a database of over 14’000 development projects to show that “tight controls” and a “narrow focus on reaching pre-set targets” undermined the performance of foreign aid projects in complex environments. The Doing Development Differently movement has been arguing for a more complexity-informed approach for years now.
All progress happens at the periphery first and then moves, over time, into the mainstream. This will not be any different. While we don’t yet exactly know what this new paradigm will look like we can discern some emerging contours.
In our recent discussion paper on the Shared Power Principle, we describe four patterns that characterize this paradigm:
Subsidiarity as a guide: putting decision-making power into the hands of people or organisations with the greatest knowledge of an issue and helping them exercise this power as effectively as possible.
Relationships first: embracing human complexity and acknowledging that we should not seek solutions with the “average” citizen in mind. We need more bespoke approaches, which require deep and trusted relationships that go beyond mere transactions.
New forms of governance and leadership: fostering the emergence of more diverse accountability structures that give people time and space to make and discuss decisions.
Building a culture of continuous learning: creating a culture of continuous learning within teams and across different organisations.
The paper contains much more detail on all of these. Much remains to be discovered and figured out but it is encouraging to see so much movement and experimentation.
One day this new, complexity-informed paradigm will no doubt seem blindingly obvious, just like the managerial paradigm of efficiency and “best practice” feels obvious right now. Until then we need to keep chipping away at the old paradigm and leading by example to bring about a new one.
Special thanks go to my colleagues Lena Kuenkel, Adrian Brown, John Burgoyne, and Bridget Nurre Jennions for their help with this post.
", "url": "https://www.dannybuerkli.com/2019/08/23/a-tale-of-two-paradigms", "image": "/assets/cards/essays/a-tale-of-two-paradigms.png", "tags": ["complexity","public","administration"], "date_published": "2019-08-23T00:00:00-03:00", "date_modified": "2019-08-23T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2019/07/31/how-self-organisation-has-changed-the-way-we-work", "title": "How self-organisation has changed the way we work and what I have learned along the way", "summary": "Four things we learned from moving towards self-organisation as a team", "content_text": "This essay was originally published on the Centre for Public Impact’s Medium page (31 July 2019).In January this year we, the London-based team of the Centre for Public Impact, began moving from being a hierarchical, conventional organisation to one that’s built on self-organisation. We collectively determine the projects we take on as an organisation, we individually decide where our skills and judgment can contribute most and each project team is empowered to make all decisions it needs to. There’s more detail to it, of course, but this is the gist of it.Before moving to self-management we operated the way most organisations do: we had a hierarchy, line managers, and all the other features you’d find in most places. Why? Because that’s how work works, doesn’t it?Self-organisation offered a radical and promising alternative. We got interested in self-management through seeing a handful of public service organisations doing it and sensing that there was a lot of promise in this.When I was first confronted with the idea of self-organisation (or self-management, as it’s also called) I was skeptical — somewhere on the spectrum between “ha, no way” and “beautiful idea but this can’t possibly work for us”. I completely changed my mind. Having lived it now I know that I wouldn’t want to go back.The way to self-management has, however, involved considerable pain and we’ve made more than a few decisions which turned out to look a lot like mistakes. It’s, of course, possible that in some oblique way these helped us get to where we are today. But if we were to do this again I would want to think long and hard about these and almost certainly do things differently.We initially weren’t aligned on whether we had in fact made the decision to move to self-management and, if so, what it would mean: in Fall last year we had a team away day where we discussed our collective desire to move to a less hierarchical, more self-directed management model. Out of that grew the decision to move to self-management. We have, however, different recollections of how that decision came about, who was involved in it and whether the group had wanted this in the first place.We had a lack of a clear mandate coupled with a disagreement over who gets to make the decision to move to self-management (is it ‘the last hierarchical decision ever taken’ or is it the team? If it’s the team, does it need to happen by consensus or is a plurality enough?). This turned out to be a formidable obstacle which cost us a lot of time and energy.Moving to self-organisation is not necessarily a decision that needs to be made by consensus, though that’s an option. What it needs is clarity around the decision to do it and what self-management means. We had neither.We tried to reinvent the mechanics of self-organised working without learning from others that had gone before us: we took a radical approach and removed (almost) at once all structure and procedures that we had in place previously. The idea was that this would show what we actually needed. It works in theory but it didn’t in practice. We ended up discovering lots of things which no doubt others already knew. This approach caused a significant amount of unnecessary pain and anxiety for many on the team because of the high degree of uncertainty that came with it.Every organisation is unique and I would never advocate copying an existing model wholesale. Not building more purposefully on what was already out there was, however, not the best choice.We’ve also come to understand that self-management does not imply getting rid of structures and processes. Quite the opposite, it requires an intentional effort to craft structures which create clarity and allow everyone to get on with the work.We struggled to distinguish management from leadership and abdicated responsibility: one of the most common failure modes when moving from traditional management to self-organisation is for managers to disengage and abdicate responsibility in an attempt to create the space for others to step into. I walked right into that trap. The only solace is that I wasn’t the only one to do that…If you’re reading the above and are thinking ‘this sounds hard and while Danny is being candid I suspect there’s lots more emotional conflict and anxiety behind what he’s writing” then you’d be exactly right.This is easily one of the hardest things I’ve ever done professionally. But it is also easily one of the most rewarding undertakings. And I would not want to go back to what we had before.The relationships I have with the people who I used to ‘line manage’ are deeper and more genuine: Previously people I work with would show me their work or get my opinion not necessarily because they thought it was useful but because they knew I expected them to do so. Now I’m pulled into meetings or asked for advice on something because people believe it’ll be useful, not because they have to. Apart from the considerable amount of time that we’ve stopped wasting this makes for much more genuine and productive personal relationships.I spend no time ‘managing’ and much more time adding value: A fair amount of my time would be spent keeping up with how a project was going, getting to grips with some detail I’d need to know so I could explain it myself, reviewing materials, etc. etc. etc. This will sound familiar to almost anyone in a managerial position. I still do some of that — but never with the intention of “checking” or “approving” but with the intention of providing support. It makes a huge difference.I feel liberated from needing to conform to a particular role that isn’t mine: we used to have fairly rigidly defined ideas of what our jobs were. Self-organisation has allowed me to feel comfortable with two things:Firstly, my job is made up of several roles and these can include things which traditionally would be seen as not appropriate for a manager. Let me give you a silly example: I tremendously enjoy coding and automating recurring business processes. Doing that is most definitely value-adding but it’s something that as a manager I’d traditionally be discouraged from doing because it’s a ‘hands-on’ type of activity that could, at least in theory, be done by someone else. I’ve now embraced this as a (minor but very enjoyable) part of what I do.Secondly, the roles I play and their relative balance will change shift regularly — and that’s great. In a traditional model, any change in roles can feel threatening and we tend to evaluate them against questions like “how does this change my status?”. In self-organised environments, this is much less of a concern.We are only seven months in and far from having ‘figured it out’. It does, however, feel like the initial hurdle is maybe behind us. There’s tremendous energy that comes from us all collectively having generated clarity around what we’re trying to do and why we’re doing it.I wanted to take a step back and share some of what we have learned. This is a personal reflection — and one that in all likelihood I would have never written if it weren’t for this environment which allows for more openness and vulnerability. I’m confident that my colleagues would agree with much of it but I’m equally certain that they’d bring up other aspects as well and tell some of the stories very differently.I also wanted to express our gratitude to those who have and are still helping us figure this out.Vishal Jodhani — a master facilitator and possibly the most attentive listener on earth — gave us a spin in the right direction with the facilitation training he ran with us more than a year ago. Therese Wallin — trainer at Tuff Leadership Training and a role model for clear and frank communication — and Lisa Gill — also with Tuff and host of the wonderful Leadermorphosis podcast — provide encouragement and much of the mental underpinning for what we’re doing. Rich Bartlett and Nati Lombardo — founders of The Hum and part of Enspiral — shared their expertise and inspiring examples. Belinda Harrison-Wong — no-nonsense coach — worked with what used to be the “management team” to help us with this transition. Finally, Helen Sanderson — inspiring role model, founder of Wellbeing Teams and fearless facilitator — pointed out what was in front of us at just the right moment in time.", "content_html": "This essay was originally published on the Centre for Public Impact’s Medium page (31 July 2019).
In January this year we, the London-based team of the Centre for Public Impact, began moving from being a hierarchical, conventional organisation to one that’s built on self-organisation. We collectively determine the projects we take on as an organisation, we individually decide where our skills and judgment can contribute most and each project team is empowered to make all decisions it needs to. There’s more detail to it, of course, but this is the gist of it.
Before moving to self-management we operated the way most organisations do: we had a hierarchy, line managers, and all the other features you’d find in most places. Why? Because that’s how work works, doesn’t it?
Self-organisation offered a radical and promising alternative. We got interested in self-management through seeing a handful of public service organisations doing it and sensing that there was a lot of promise in this.
When I was first confronted with the idea of self-organisation (or self-management, as it’s also called) I was skeptical — somewhere on the spectrum between “ha, no way” and “beautiful idea but this can’t possibly work for us”. I completely changed my mind. Having lived it now I know that I wouldn’t want to go back.
The way to self-management has, however, involved considerable pain and we’ve made more than a few decisions which turned out to look a lot like mistakes. It’s, of course, possible that in some oblique way these helped us get to where we are today. But if we were to do this again I would want to think long and hard about these and almost certainly do things differently.
We initially weren’t aligned on whether we had in fact made the decision to move to self-management and, if so, what it would mean: in Fall last year we had a team away day where we discussed our collective desire to move to a less hierarchical, more self-directed management model. Out of that grew the decision to move to self-management. We have, however, different recollections of how that decision came about, who was involved in it and whether the group had wanted this in the first place.
We had a lack of a clear mandate coupled with a disagreement over who gets to make the decision to move to self-management (is it ‘the last hierarchical decision ever taken’ or is it the team? If it’s the team, does it need to happen by consensus or is a plurality enough?). This turned out to be a formidable obstacle which cost us a lot of time and energy.
Moving to self-organisation is not necessarily a decision that needs to be made by consensus, though that’s an option. What it needs is clarity around the decision to do it and what self-management means. We had neither.
We tried to reinvent the mechanics of self-organised working without learning from others that had gone before us: we took a radical approach and removed (almost) at once all structure and procedures that we had in place previously. The idea was that this would show what we actually needed. It works in theory but it didn’t in practice. We ended up discovering lots of things which no doubt others already knew. This approach caused a significant amount of unnecessary pain and anxiety for many on the team because of the high degree of uncertainty that came with it.
Every organisation is unique and I would never advocate copying an existing model wholesale. Not building more purposefully on what was already out there was, however, not the best choice.
We’ve also come to understand that self-management does not imply getting rid of structures and processes. Quite the opposite, it requires an intentional effort to craft structures which create clarity and allow everyone to get on with the work.
We struggled to distinguish management from leadership and abdicated responsibility: one of the most common failure modes when moving from traditional management to self-organisation is for managers to disengage and abdicate responsibility in an attempt to create the space for others to step into. I walked right into that trap. The only solace is that I wasn’t the only one to do that…
If you’re reading the above and are thinking ‘this sounds hard and while Danny is being candid I suspect there’s lots more emotional conflict and anxiety behind what he’s writing” then you’d be exactly right.
This is easily one of the hardest things I’ve ever done professionally. But it is also easily one of the most rewarding undertakings. And I would not want to go back to what we had before.
The relationships I have with the people who I used to ‘line manage’ are deeper and more genuine: Previously people I work with would show me their work or get my opinion not necessarily because they thought it was useful but because they knew I expected them to do so. Now I’m pulled into meetings or asked for advice on something because people believe it’ll be useful, not because they have to. Apart from the considerable amount of time that we’ve stopped wasting this makes for much more genuine and productive personal relationships.
I spend no time ‘managing’ and much more time adding value: A fair amount of my time would be spent keeping up with how a project was going, getting to grips with some detail I’d need to know so I could explain it myself, reviewing materials, etc. etc. etc. This will sound familiar to almost anyone in a managerial position. I still do some of that — but never with the intention of “checking” or “approving” but with the intention of providing support. It makes a huge difference.
I feel liberated from needing to conform to a particular role that isn’t mine: we used to have fairly rigidly defined ideas of what our jobs were. Self-organisation has allowed me to feel comfortable with two things:
Firstly, my job is made up of several roles and these can include things which traditionally would be seen as not appropriate for a manager. Let me give you a silly example: I tremendously enjoy coding and automating recurring business processes. Doing that is most definitely value-adding but it’s something that as a manager I’d traditionally be discouraged from doing because it’s a ‘hands-on’ type of activity that could, at least in theory, be done by someone else. I’ve now embraced this as a (minor but very enjoyable) part of what I do.
Secondly, the roles I play and their relative balance will change shift regularly — and that’s great. In a traditional model, any change in roles can feel threatening and we tend to evaluate them against questions like “how does this change my status?”. In self-organised environments, this is much less of a concern.
We are only seven months in and far from having ‘figured it out’. It does, however, feel like the initial hurdle is maybe behind us. There’s tremendous energy that comes from us all collectively having generated clarity around what we’re trying to do and why we’re doing it.
I wanted to take a step back and share some of what we have learned. This is a personal reflection — and one that in all likelihood I would have never written if it weren’t for this environment which allows for more openness and vulnerability. I’m confident that my colleagues would agree with much of it but I’m equally certain that they’d bring up other aspects as well and tell some of the stories very differently.
I also wanted to express our gratitude to those who have and are still helping us figure this out.
Vishal Jodhani — a master facilitator and possibly the most attentive listener on earth — gave us a spin in the right direction with the facilitation training he ran with us more than a year ago. Therese Wallin — trainer at Tuff Leadership Training and a role model for clear and frank communication — and Lisa Gill — also with Tuff and host of the wonderful Leadermorphosis podcast — provide encouragement and much of the mental underpinning for what we’re doing. Rich Bartlett and Nati Lombardo — founders of The Hum and part of Enspiral — shared their expertise and inspiring examples. Belinda Harrison-Wong — no-nonsense coach — worked with what used to be the “management team” to help us with this transition. Finally, Helen Sanderson — inspiring role model, founder of Wellbeing Teams and fearless facilitator — pointed out what was in front of us at just the right moment in time.
", "url": "https://www.dannybuerkli.com/2019/07/31/how-self-organisation-has-changed-the-way-we-work", "image": "/assets/cards/essays/how-self-organisation-has-changed-the-way-we-work.png", "tags": ["self-organisation","self-management","teal"], "date_published": "2019-07-31T00:00:00-03:00", "date_modified": "2019-07-31T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2019/05/13/how-we-know-when-decentralization-works", "title": "How we know when decentralization works in government", "summary": "There is a lively debate raging in the public policy world over the use of evidence, the role of innovation and experimentation, and the importance of \"place\". But how do we know when to centralize and when not to?", "content_text": "This essay was originally published on the Centre for Public Impact’s Medium page (13 May 2019).There is a lively debate raging in the public policy world over the use of evidence, the role of innovation and experimentation, and the importance of “place”. Some public policies clearly benefit from being steered from the national level. Others are much more effective when key decisions are made locally, e.g. in cities.But how do we know when to centralize and when not to? When is evidence of “what works” useful and when is it not? When is local experimentation crucial for success and when is it a waste of time and resources? When is the “enablement mindset” most needed?We develop a simple model here that helps us think about these questions in a structured way. Like all models, it is a simplification but it hopefully helps illuminate the questions we should be asking ourselves and others in this debate.One important caveat: we are looking here at public policies and their effectiveness from a purely empirical perspective. In other words, we are not exploring normative considerations for or against, e.g. decentralization or the use of evidence.Policymaking is about changing the world in a purposeful way“Policies” or “strategies” or “interventions” in government are fundamentally about changing the world from the current state into a more desired state, to move it from where it is today to where we want it to be.There are of course many more reasons why governments sometimes embark on certain policies. Often it is about needing to be seen doing something. But let’s assume for now that our motives are pure and the objective of a policy is to change the world from A to B in some purposeful way.Assuming that we have a vision or desired state in mind the question becomes: given that we’re here, how do we get there?The problem of the infinite policy design spaceAny intervention or policy will always have multiple dimensions about which we need to make decisions. This is what some people call the “design space” of a policy. This design space is very large, even for simple policies.Take conditional cash transfers as an example. Conditional cash transfer programmes seek to reduce poverty by making welfare payments conditional on some criteria, e.g. whether the recipient’s children are enrolled in school.What is the design space for a conditional cash transfer programme?We need to decide on the amount of money we transfer to beneficiaries. Should it be 10$ per year? 100$ per year? 1000$ per year?We also need to decide on the frequency of transfers. Should money be given weekly? Monthly? Quarterly?Next, we need to decide on the conditions attached to the transfer and on how we’ll determine whether the conditions are met. And it goes on and on.The number of possible combinations is enormous. You could have a programme that transfers $10 weekly and demands that a recipient’s child be vaccinated. Or you could have a programme that transfers $50 on a monthly basis and requires that the children be sent to school. Or you could go for any other conceivable combination of these factors.Design space for a policy with only two dimensionsThe graph above shows a policy with just two elements. Each dot represents one possible combination. With ten choices to make on each dimension, we would get 100 possible versions of that policy.Given that most policies involve not just two but hundreds if not thousands of dimensions the “design space” of any policy is — for all intents and purposes — infinite.What policymakers are interested in is which configuration will yield the best result. In other words, which policy design will be the most effective?We could, in theory, try out every single conceivable combination in the design space. That would mean trying out a huge number of variants of any policy. Given the number of dimensions of most policies and the combinatorial explosion involved we would quickly be faced with a near-infinite number of policies to try out. That seems impractical, particularly once you factor in that most policies take time to show their results.Systematically trying out every single combination of parameters is clearly not an option.What should we do instead then? The answer depends on two things: the shape of the response surfaceand on how much that response surface differs between different places.Smooth and rugged policies — for some policies small changes don’t make a difference, for others it changes everythingThere exist policies for which the response surface is relatively smooth. Small changes in the design space won’t make a big difference.The sketch below shows such a policy. The red dot shows the optimal configuration. You can imagine this as a hill on a map with the red dot being the highest point.Policy with a \"smooth\" response surface --- small changes in its configuration won't make a big differenceBecause the surface is smooth we can systematically search our way to the best solution. For policies with smooth response surfaces, we are much more likely to have either found the optimal place in the design space (the red dot in the sketch) or be somewhere near it.For policies with a \"smooth\" response surface, we can iterate in small steps and find our way to better configurationsFurthermore, because the surface is smooth there is one right answer to the question of “which configuration of parameters will maximize the effectiveness of this policy?”. Once you found it you should stay there and not change anything. If you deviate from it, you’re doing it wrong.But what if the response surface for a policy isn’t smooth but rugged?First, we now have multiple “right” answers, i.e. multiple configurations which will produce good results.Policy with a \"rugged\" response surface - small changes in its configuration will make a big difference hereSecond, finding our way to a good configuration (a red dot in the sketch) is harder because the landscape is more confusing. For those policies, we are much less likely to have either found a global maximum or be near it. This is typical for policies which are complex (rather than just complicated).To summarize, a policy can either have a smooth or rugged surface. This is, of course, a spectrum with perfectly flat surfaces on one side and crazily spiky surfaces on the other.There is, however, a dimension beyond rugged-vs-smooth that we need to consider.The importance of place — some policies work the same way everywhere, others don’tSo far we implicitly assumed that we are implementing this policy only in one place. But almost any policy will, of course, be implemented in more than one place, city, town, region, etc.If we implement a policy in more than one place each place will have its own response surface. If we look at those local response surfaces and compare them to each other we’ll see one of two situations: it’s possible that they all look the same. Or it’s possible that they look very different.If they all look the same we have low variability between surfaces. If they all look radically different we have high variability between surfaces.Response surfaces can either be similar (low variability) between different places or very different (high variability)If we have a smooth surface and low variability it means that no matter where this policy is implemented, the effect will be the same. Changes in the specification of the policy will also have the same effect everywhere.Smooth response surfaces and low variability between placesIf the surface is rugged and the variability is low we will get the same effect no matter where we implement the policy. Again, we can “scale” policies to our heart’s content.Rugged response surface and low variability between placesIf the surface is smooth and the variability is high we will see some differences between places. This means that a policy’s effectiveness will be different in different places — it’ll work everywhere but somewhat better in some places than others.Smooth response surfaces and high variability between placesThe most challenging case is, however, when the response surface is rugged and the variability is high.Rugged response surfaces and high variability between placesIf we implement the policy with the same specification in different places we’ll get radically different results. To achieve similar results we will need different specifications.What this means for evidence, experimentation, innovation, management, and local decision-makingWhat does this all mean in practice? This model gives us a useful framework to think about the role of evidence, experimentation, innovation, management, and local vs. central decisionmaking.Let’s look at the four stylized types of policies in turn.If we’re looking at a policy that has a smooth response surface and low variability between places life is easy. Evidence about “what works” will travel well. “Scaling” works well because what worked in one place will work in another. That’s why we can pilot a policy in one place then scale it to many more. Top-down control and deciding things at the highest possible (e.g. national) level will also serve us well. In this scenario, there is nothing to be gained from allowing local variation or experimentation.Consider now a policy with a rugged response surface but still low variability between places. Evidence about “what works” will also travel well. Because the response surface rugged we are, however, less likely to have found the best way of configuring the policy. This means that we would benefit from local experimentation since that increases the chances of hitting on new and better ways of doing things. But if we were to decree this policy and roll it out in an identical fashion across a country we could expect to see the same results no matter where we implement it.What about a policy that has a smooth response surface but high variability between places? High variability means that unless we have worked out the absolute best way of configuring this policy then the same policy implemented the same way will generate different results in different places. Again this implies that we would benefit from local experimentation.Finally, let’s look at policies that have a rugged response surface and show high variability between places. Those are policies for which small changes will make a huge difference and for which the effects will vary drastically across places. Evidence about “what works” is not going to tell us much if anything at all in this scenario. Local decision-making and experimentation are essential because anything else will guarantee inferior results.“Local” or “place” in this context can mean a city, a region or an individual. For some of the most complex social services which governments provide the unit that we should be looking at is the individual. Saying that individuals with complex needs require individualized solutions to their problems is simply saying that the response surface is rugged and the variability between individuals high.The bottom line is this: unless we’re looking at a policy with a smooth response surface and low variability across places local decision-making and experimentation will always beat control from higher up.Some things certainly fall into the smooth surface/low variability category but most policy problems arguably don’t. This provides the justification for subsidiarity, the idea that everything should be dealt with at the lowest level possible.In other words, we should start from the assumption that a problem can be dealt with at the lowest level and only hand it up if there is compelling evidence that it’s necessary. Unless we are very sure we’re in a situation with a smooth response surface and low variability localism will generate better results.Sources and inspirationThis post draws on many different ideas. Two were particularly influential:The idea of a response surface is inspired by Lant Pritchett’s talk titled “The debate over RCTs is over: We won, they lost.” (Slides)The discussion of the role of evidence is influenced by Nancy Cartwright’s and Jeremy Hardie’s book “Evidence-Based Policy: A Practical Guide to Doing It Better”.", "content_html": "This essay was originally published on the Centre for Public Impact’s Medium page (13 May 2019).
There is a lively debate raging in the public policy world over the use of evidence, the role of innovation and experimentation, and the importance of “place”. Some public policies clearly benefit from being steered from the national level. Others are much more effective when key decisions are made locally, e.g. in cities.
But how do we know when to centralize and when not to? When is evidence of “what works” useful and when is it not? When is local experimentation crucial for success and when is it a waste of time and resources? When is the “enablement mindset” most needed?
We develop a simple model here that helps us think about these questions in a structured way. Like all models, it is a simplification but it hopefully helps illuminate the questions we should be asking ourselves and others in this debate.
One important caveat: we are looking here at public policies and their effectiveness from a purely empirical perspective. In other words, we are not exploring normative considerations for or against, e.g. decentralization or the use of evidence.
Policymaking is about changing the world in a purposeful way
“Policies” or “strategies” or “interventions” in government are fundamentally about changing the world from the current state into a more desired state, to move it from where it is today to where we want it to be.
![\"Policies](\"/assets/20190513_decentralization_a.png\")
There are of course many more reasons why governments sometimes embark on certain policies. Often it is about needing to be seen doing something. But let’s assume for now that our motives are pure and the objective of a policy is to change the world from A to B in some purposeful way.
Assuming that we have a vision or desired state in mind the question becomes: given that we’re here, how do we get there?
Any intervention or policy will always have multiple dimensions about which we need to make decisions. This is what some people call the “design space” of a policy. This design space is very large, even for simple policies.
Take conditional cash transfers as an example. Conditional cash transfer programmes seek to reduce poverty by making welfare payments conditional on some criteria, e.g. whether the recipient’s children are enrolled in school.
What is the design space for a conditional cash transfer programme?
We need to decide on the amount of money we transfer to beneficiaries. Should it be 10$ per year? 100$ per year? 1000$ per year?
We also need to decide on the frequency of transfers. Should money be given weekly? Monthly? Quarterly?
Next, we need to decide on the conditions attached to the transfer and on how we’ll determine whether the conditions are met. And it goes on and on.
The number of possible combinations is enormous. You could have a programme that transfers $10 weekly and demands that a recipient’s child be vaccinated. Or you could have a programme that transfers $50 on a monthly basis and requires that the children be sent to school. Or you could go for any other conceivable combination of these factors.
![\"Design](\"/assets/20190513_decentralization_b.png\")
The graph above shows a policy with just two elements. Each dot represents one possible combination. With ten choices to make on each dimension, we would get 100 possible versions of that policy.
Given that most policies involve not just two but hundreds if not thousands of dimensions the “design space” of any policy is — for all intents and purposes — infinite.
What policymakers are interested in is which configuration will yield the best result. In other words, which policy design will be the most effective?
We could, in theory, try out every single conceivable combination in the design space. That would mean trying out a huge number of variants of any policy. Given the number of dimensions of most policies and the combinatorial explosion involved we would quickly be faced with a near-infinite number of policies to try out. That seems impractical, particularly once you factor in that most policies take time to show their results.
Systematically trying out every single combination of parameters is clearly not an option.
What should we do instead then? The answer depends on two things: the shape of the response surfaceand on how much that response surface differs between different places.
There exist policies for which the response surface is relatively smooth. Small changes in the design space won’t make a big difference.
The sketch below shows such a policy. The red dot shows the optimal configuration. You can imagine this as a hill on a map with the red dot being the highest point.
![\"Policy](\"/assets/20190513_decentralization_c.png\")
Because the surface is smooth we can systematically search our way to the best solution. For policies with smooth response surfaces, we are much more likely to have either found the optimal place in the design space (the red dot in the sketch) or be somewhere near it.
![\"For](\"/assets/20190513_decentralization_d.png\")
Furthermore, because the surface is smooth there is one right answer to the question of “which configuration of parameters will maximize the effectiveness of this policy?”. Once you found it you should stay there and not change anything. If you deviate from it, you’re doing it wrong.
But what if the response surface for a policy isn’t smooth but rugged?
First, we now have multiple “right” answers, i.e. multiple configurations which will produce good results.
![\"Policy](\"/assets/20190513_decentralization_e.png\")
Second, finding our way to a good configuration (a red dot in the sketch) is harder because the landscape is more confusing. For those policies, we are much less likely to have either found a global maximum or be near it. This is typical for policies which are complex (rather than just complicated).
![\"\"](\"/assets/20190513_decentralization_f.png\")
To summarize, a policy can either have a smooth or rugged surface. This is, of course, a spectrum with perfectly flat surfaces on one side and crazily spiky surfaces on the other.
There is, however, a dimension beyond rugged-vs-smooth that we need to consider.
So far we implicitly assumed that we are implementing this policy only in one place. But almost any policy will, of course, be implemented in more than one place, city, town, region, etc.
If we implement a policy in more than one place each place will have its own response surface. If we look at those local response surfaces and compare them to each other we’ll see one of two situations: it’s possible that they all look the same. Or it’s possible that they look very different.
If they all look the same we have low variability between surfaces. If they all look radically different we have high variability between surfaces.
![\"\"](\"/assets/20190513_decentralization_g.png\")
If we have a smooth surface and low variability it means that no matter where this policy is implemented, the effect will be the same. Changes in the specification of the policy will also have the same effect everywhere.
![\"\"](\"/assets/20190513_decentralization_h.png\")
If the surface is rugged and the variability is low we will get the same effect no matter where we implement the policy. Again, we can “scale” policies to our heart’s content.
![\"\"](\"/assets/20190513_decentralization_i.png\")
If the surface is smooth and the variability is high we will see some differences between places. This means that a policy’s effectiveness will be different in different places — it’ll work everywhere but somewhat better in some places than others.
![\"\"](\"/assets/20190513_decentralization_j.png\")
The most challenging case is, however, when the response surface is rugged and the variability is high.
![\"\"](\"/assets/20190513_decentralization_k.png\")
If we implement the policy with the same specification in different places we’ll get radically different results. To achieve similar results we will need different specifications.
What does this all mean in practice? This model gives us a useful framework to think about the role of evidence, experimentation, innovation, management, and local vs. central decisionmaking.
Let’s look at the four stylized types of policies in turn.
If we’re looking at a policy that has a smooth response surface and low variability between places life is easy. Evidence about “what works” will travel well. “Scaling” works well because what worked in one place will work in another. That’s why we can pilot a policy in one place then scale it to many more. Top-down control and deciding things at the highest possible (e.g. national) level will also serve us well. In this scenario, there is nothing to be gained from allowing local variation or experimentation.
Consider now a policy with a rugged response surface but still low variability between places. Evidence about “what works” will also travel well. Because the response surface rugged we are, however, less likely to have found the best way of configuring the policy. This means that we would benefit from local experimentation since that increases the chances of hitting on new and better ways of doing things. But if we were to decree this policy and roll it out in an identical fashion across a country we could expect to see the same results no matter where we implement it.
What about a policy that has a smooth response surface but high variability between places? High variability means that unless we have worked out the absolute best way of configuring this policy then the same policy implemented the same way will generate different results in different places. Again this implies that we would benefit from local experimentation.
Finally, let’s look at policies that have a rugged response surface and show high variability between places. Those are policies for which small changes will make a huge difference and for which the effects will vary drastically across places. Evidence about “what works” is not going to tell us much if anything at all in this scenario. Local decision-making and experimentation are essential because anything else will guarantee inferior results.
“Local” or “place” in this context can mean a city, a region or an individual. For some of the most complex social services which governments provide the unit that we should be looking at is the individual. Saying that individuals with complex needs require individualized solutions to their problems is simply saying that the response surface is rugged and the variability between individuals high.
The bottom line is this: unless we’re looking at a policy with a smooth response surface and low variability across places local decision-making and experimentation will always beat control from higher up.
Some things certainly fall into the smooth surface/low variability category but most policy problems arguably don’t. This provides the justification for subsidiarity, the idea that everything should be dealt with at the lowest level possible.
In other words, we should start from the assumption that a problem can be dealt with at the lowest level and only hand it up if there is compelling evidence that it’s necessary. Unless we are very sure we’re in a situation with a smooth response surface and low variability localism will generate better results.
Sources and inspiration
This post draws on many different ideas. Two were particularly influential:
The idea of a response surface is inspired by Lant Pritchett’s talk titled “The debate over RCTs is over: We won, they lost.” (Slides)
The discussion of the role of evidence is influenced by Nancy Cartwright’s and Jeremy Hardie’s book “Evidence-Based Policy: A Practical Guide to Doing It Better”.
", "url": "https://www.dannybuerkli.com/2019/05/13/how-we-know-when-decentralization-works", "image": "/assets/cards/essays/how-we-know-when-decentralization-works.png", "tags": ["decentralization","policymaking","evidence","public","administration"], "date_published": "2019-05-13T00:00:00-03:00", "date_modified": "2019-05-13T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2019/04/08/what-gets-measured-gets-managed-drucker-never-said-it", "title": "“What gets measured gets managed” — It’s wrong and Drucker never said it", "summary": "“What gets measured gets managed” is one of the most widely received wisdoms but it's deeply flawed", "content_text": "This article was originally published on the Centre for Public Impact’s Medium page (08 April 2019).Out of all received wisdoms “what gets measured gets managed” must be one of the most widely accepted as obvious. After all, how could we ever manage something that isn’t being measured?The quote is usually attributed to Peter Drucker, the brilliant management theorist. A bit of digging reveals, however, two surprising things.First, Drucker — according to the Drucker Institute no less — never said it.Let that sink in. All those Harvard Business Review articles attributing the quote to him? Apparently not so.Second, the fact that something is wrong with “what gets measured gets managed” has been apparent for a long time.V. F. Ridgway published a paper in 1956 criticizing the measurement mantra. Simon Caulkin, a columnist, neatly summarized Ridgway’s argument as: “What gets measured gets managed — even when it’s pointless to measure and manage it, and even if it harms the purpose of the organisation to do so”.Indeed the paper’s title is “Dysfunctional Consequences of Performance Measurements”.Ridgway clearly was onto something in 1956. Not everything that matters can be measured. Not everything that we can measure matters.It’s also obvious that there are plenty of organizations that achieve outstanding results without measuring some outcome of interest and “managing” for it. The Dutch home care organization Buurtzorg is just one example. They achieve higher patient satisfaction at lower costs without measuring and managing for either in any meaningful way.To summarize: Drucker never said the infamous sentence and we know already in 1956 that the “what gets measured gets managed” was deeply flawed.Photo by William Warby on UnsplashThere is a mountain of evidence out there that Ridgway was right. That’s why it’s hard to understand why we are still having this debate more than 60 years after his paper.Why would Ridgway say that it may be pointless to measure and manage something? Because measuring what we actually care about — rather than what’s easily measured — is fiendishly difficult and often impossible.The difficulty of precisely specifying the outcome we are truly after is well known to AI researchers. There’s an entire list with hilarious examples of algorithms being given a metric to optimize for and finding entirely unexpected ways of ostensibly achieving their objective while violating common sense.This is entertaining in the case of a toy algorithm. It’s less so when organisations and the people they serve are affected like when “staff at an outsourced Police control room were making 999 calls at quiet times in order to meet their target of answering 92% of calls within 10 seconds”. Or the time when “doctors [in the UK] were diverted from treating seriously ill patients to ones with minor problems in order to meet 4-hour waiting time targets” (details here).What’s depressing about this is how predictable these problems with measurement and management are, how long we’ve understood them, how everyone involved in pursuing some meaningless metric knows it’s a sham and how “what gets measured gets managed” somehow is still accepted as gospel.Note: The quote from Simon Caulkin above was previously inaccurately attributed to V. F. Ridgway. Many thanks to Scott B. Weingart for spotting the error and kindly pointing it out.Hat tip to Bill Hennessy whose blog post on this topic sparked my investigation into this — note though that the V. F. Ridgway quote is misattributed there.", "content_html": "This article was originally published on the Centre for Public Impact’s Medium page (08 April 2019).
Out of all received wisdoms “what gets measured gets managed” must be one of the most widely accepted as obvious. After all, how could we ever manage something that isn’t being measured?
The quote is usually attributed to Peter Drucker, the brilliant management theorist. A bit of digging reveals, however, two surprising things.
First, Drucker — according to the Drucker Institute no less — never said it.
Let that sink in. All those Harvard Business Review articles attributing the quote to him? Apparently not so.
Second, the fact that something is wrong with “what gets measured gets managed” has been apparent for a long time.
V. F. Ridgway published a paper in 1956 criticizing the measurement mantra. Simon Caulkin, a columnist, neatly summarized Ridgway’s argument as:
“What gets measured gets managed — even when it’s pointless to measure and manage it, and even if it harms the purpose of the organisation to do so”.
Indeed the paper’s title is “Dysfunctional Consequences of Performance Measurements”.
Ridgway clearly was onto something in 1956. Not everything that matters can be measured. Not everything that we can measure matters.
It’s also obvious that there are plenty of organizations that achieve outstanding results without measuring some outcome of interest and “managing” for it. The Dutch home care organization Buurtzorg is just one example. They achieve higher patient satisfaction at lower costs without measuring and managing for either in any meaningful way.
To summarize: Drucker never said the infamous sentence and we know already in 1956 that the “what gets measured gets managed” was deeply flawed.
![\"Three](\"/assets/20190408_ruler.jpeg\")
There is a mountain of evidence out there that Ridgway was right. That’s why it’s hard to understand why we are still having this debate more than 60 years after his paper.
Why would Ridgway say that it may be pointless to measure and manage something? Because measuring what we actually care about — rather than what’s easily measured — is fiendishly difficult and often impossible.
The difficulty of precisely specifying the outcome we are truly after is well known to AI researchers. There’s an entire list with hilarious examples of algorithms being given a metric to optimize for and finding entirely unexpected ways of ostensibly achieving their objective while violating common sense.
This is entertaining in the case of a toy algorithm. It’s less so when organisations and the people they serve are affected like when “staff at an outsourced Police control room were making 999 calls at quiet times in order to meet their target of answering 92% of calls within 10 seconds”. Or the time when “doctors [in the UK] were diverted from treating seriously ill patients to ones with minor problems in order to meet 4-hour waiting time targets” (details here).
What’s depressing about this is how predictable these problems with measurement and management are, how long we’ve understood them, how everyone involved in pursuing some meaningless metric knows it’s a sham and how “what gets measured gets managed” somehow is still accepted as gospel.
Note: The quote from Simon Caulkin above was previously inaccurately attributed to V. F. Ridgway. Many thanks to Scott B. Weingart for spotting the error and kindly pointing it out.
Hat tip to Bill Hennessy whose blog post on this topic sparked my investigation into this — note though that the V. F. Ridgway quote is misattributed there.
", "url": "https://www.dannybuerkli.com/2019/04/08/what-gets-measured-gets-managed-drucker-never-said-it", "image": "/assets/cards/essays/what-gets-measured-gets-managed-drucker-never-said-it.png", "tags": ["measurement","managerialism"], "date_published": "2019-04-08T00:00:00-03:00", "date_modified": "2019-04-08T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } }, { "id": "https://www.dannybuerkli.com/2016/05/18/ten-government-innovations-and-their-place-in-the-hype-cycle", "title": "10 Government Innovations and Their Place in the Hype Cycle", "summary": "10 selected hot topics in government innovation and their place in the Hype Cycle", "content_text": "This article was originally published in the Huffington Post (18 May 2016).The Hype Cycle, created by Gartner, provides a helpful and entertaining way to map how a technology develops from conception to maturity to adoption. As new technologies and ways of doing things come into the world, it takes experimentation to fully grasp the potential (or lack thereof) of something new. And much like in the private sector, where the Hype Cycle originated, it can be applied just as well to government.New technologies start in the technology trigger phase, climb the peak of inflated expectations, fall into the trough of disillusionment, and traverse the slope of enlightenment before - finally - arriving at the plateau of productivity. Things go from new and shiny to mature, from hyperbolic to realistic expectations, from courageous first users to mainstream adoption. Some of course never make it to the end and fall into the graveyard of discarded ideas.New Public Management (NPM) was at one point at the peak of inflated expectations but is now firmly and, many would argue, productively embedded in the practices of many governments. Along the way governments have learned, sometimes painfully, how to make best use of it.With that in mind, and without being too serious or scientific about it, here are 10 selected hot topics in government innovation and their place in the Hype Cycle.1) Artificial Intelligence: out with the humans, in with the machinesTransitioning out of technology trigger phase, scaling peak of inflated expectationsIf you’re a policymaker, you should be very afraid. DeepMind has beaten the world’s best Go player, something all experts thought was at least another decade out, and governments are only beginning to grasp what might be possible using machine-learning and AI. When will you be replaced? Expectations are high but will rise even more as governments start experimenting in earnest with this technology.2) Blockchain: one replicated, shared ledger to rule them allPeak of inflated expectationsBlockchain is the technology behind Bitcoin. While worthwhile government applications for blockchain will surely emerge, hype notwithstanding, few people seem to genuinely understand what blockchain can and cannot do. Everyone is still trying to figure out where applications of this technology really make sense for governments and their citizens and what it’s really good for. No doubt worthwhile and potentially radical applications will emerge, but some disappointments seem inevitable along the way.3) Design Thinking: Post it-sized, citizen-centred insightsPeak of inflated expectationsOriginating in the world of product and service design, this method borrows heavily from ethnography and is particularly valuable to improving governmental services. Design is not likely to be the solution for all of our governmental problems, but you wouldn’t be able to tell if you only listened to some of the star speakers at design conferences. It’s a solid method that delivers great results if used thoughtfully and in the right context. We should stick with it through the disillusionment that may be coming as governments learn about the limits of the method the hard way.4) Policy Labs: creative laboratories for better governmentJust past the peak of inflated expectations, about to descend into the trough of disillusionmentPolicy Labs are the hottest property in government innovation land. They bring together design thinkers and other creative problem-solvers to tackle government’s hardest problems. Organisations such as MindLab, OPM Lab and the UK Policy Lab are leading the way. The trough of disillusionment will likely come as more governments adopt the model, but it’s worth sticking with it since they, if set up and run well and given the necessary political backing, can make a difference.5) Randomised Control Trials: randomise this!Trough of disillusionmentHaving originated in development economics many countries, regional governments and cities - developing and developed alike - are warming up to the idea that randomised studies of interventions can help them figure out what works and what doesn’t. After an initial hype practitioners and policymakers are now getting better at understanding where and how this method can deliver most value.6) Social Impact Bonds: pay for success onlyTrough of disillusionmentSocial Impact Bonds, SIBs for those in the know, were first trialled in Peterborough, UK, in an attempt to reduce prisoner recidivism by using an innovative financing mechanism where private investors would get a return if a new intervention performed better than a conventional one. After much initial, and justified, excitement it today seems clear that they should work in theory, but it’s less clear how well they work in practice. SIBs are difficult and expensive to run, they likely have a place in the social services landscape, but it may be more limited than was initially thought.7) Open Data: spray and prayTrough of disillusionmentTo the casual observer, it may not be immediately obvious if “open data“ has produced much of practical value to citizens, other than better public transportation apps and free pizza at government-sponsored open data “hackathons”. Making government data easily accessible should improve accountability and allow smart outside actors to spot opportunities for improvement, but so far actual results have been somewhat below expectations.8) Behavioural Insights: gently nudging citizens to better decisionsSlope of enlightenmentBehavioural science has given governments a new tool not just to better understand the sometimes odd decisions their citizens make but to find ways to gently nudge people towards better choices. The method is well on its way to mainstream adoption in some (Anglo-Saxon) countries, but remains under-utilised in others.9) E-Government: bringing government websites into the 21st centurySlope of enlightenmentAllowing citizens to conduct their interactions with government online is a somewhat unexciting but absolutely essential service to citizens. No one should be subjected to long, pointless queuing to pick up a piece of paper anymore. Government Digital Services in the UK, 18F in the US and the entire Estonian government are showing how it can be done. Building user-friendly government online services can also act as a gateway to a more digitally enabled government in other areas.10) Lean in government: cut the slack the Toyota wayPlateau of productivityAs a structured way to cut waste in all kinds of processes, “lean” has long established itself as a useful tool for government operations too, not just for car factories. While “lean” doesn’t generate any particular excitement anymore, it’s become a productive instrument to improve government services.Not all countries, regions or cities are in the same place at the same time on this journey. Inflated expectations in one place co-exist with bitter disillusionment in others.The movement of new ideas, methods and technologies through the Hype Cycle is inevitable, as we collectively take promising new approaches and figure out what exactly they can and cannot do. Asking about impact, rather than appearance, can help us get there faster and discern the useful from the hopeless a bit quicker.Some amount of experimentation where the end result is unknown is, however, inevitable and indeed desirable. Odds are that all of these methods have some valuable contribution to make to better government. It is worth persevering all the way through to the plateau of productivity if we want citizens to reap the benefits.", "content_html": "This article was originally published in the Huffington Post (18 May 2016).
The Hype Cycle, created by Gartner, provides a helpful and entertaining way to map how a technology develops from conception to maturity to adoption. As new technologies and ways of doing things come into the world, it takes experimentation to fully grasp the potential (or lack thereof) of something new. And much like in the private sector, where the Hype Cycle originated, it can be applied just as well to government.
New technologies start in the technology trigger phase, climb the peak of inflated expectations, fall into the trough of disillusionment, and traverse the slope of enlightenment before - finally - arriving at the plateau of productivity. Things go from new and shiny to mature, from hyperbolic to realistic expectations, from courageous first users to mainstream adoption. Some of course never make it to the end and fall into the graveyard of discarded ideas.
![\"steel](\"/assets/20160518_hypecycle.png\")
New Public Management (NPM) was at one point at the peak of inflated expectations but is now firmly and, many would argue, productively embedded in the practices of many governments. Along the way governments have learned, sometimes painfully, how to make best use of it.
With that in mind, and without being too serious or scientific about it, here are 10 selected hot topics in government innovation and their place in the Hype Cycle.
Transitioning out of technology trigger phase, scaling peak of inflated expectations
If you’re a policymaker, you should be very afraid. DeepMind has beaten the world’s best Go player, something all experts thought was at least another decade out, and governments are only beginning to grasp what might be possible using machine-learning and AI. When will you be replaced? Expectations are high but will rise even more as governments start experimenting in earnest with this technology.
Peak of inflated expectations
Blockchain is the technology behind Bitcoin. While worthwhile government applications for blockchain will surely emerge, hype notwithstanding, few people seem to genuinely understand what blockchain can and cannot do. Everyone is still trying to figure out where applications of this technology really make sense for governments and their citizens and what it’s really good for. No doubt worthwhile and potentially radical applications will emerge, but some disappointments seem inevitable along the way.
Peak of inflated expectations
Originating in the world of product and service design, this method borrows heavily from ethnography and is particularly valuable to improving governmental services. Design is not likely to be the solution for all of our governmental problems, but you wouldn’t be able to tell if you only listened to some of the star speakers at design conferences. It’s a solid method that delivers great results if used thoughtfully and in the right context. We should stick with it through the disillusionment that may be coming as governments learn about the limits of the method the hard way.
Just past the peak of inflated expectations, about to descend into the trough of disillusionment
Policy Labs are the hottest property in government innovation land. They bring together design thinkers and other creative problem-solvers to tackle government’s hardest problems. Organisations such as MindLab, OPM Lab and the UK Policy Lab are leading the way. The trough of disillusionment will likely come as more governments adopt the model, but it’s worth sticking with it since they, if set up and run well and given the necessary political backing, can make a difference.
Trough of disillusionment
Having originated in development economics many countries, regional governments and cities - developing and developed alike - are warming up to the idea that randomised studies of interventions can help them figure out what works and what doesn’t. After an initial hype practitioners and policymakers are now getting better at understanding where and how this method can deliver most value.
Trough of disillusionment
Social Impact Bonds, SIBs for those in the know, were first trialled in Peterborough, UK, in an attempt to reduce prisoner recidivism by using an innovative financing mechanism where private investors would get a return if a new intervention performed better than a conventional one. After much initial, and justified, excitement it today seems clear that they should work in theory, but it’s less clear how well they work in practice. SIBs are difficult and expensive to run, they likely have a place in the social services landscape, but it may be more limited than was initially thought.
Trough of disillusionment
To the casual observer, it may not be immediately obvious if “open data“ has produced much of practical value to citizens, other than better public transportation apps and free pizza at government-sponsored open data “hackathons”. Making government data easily accessible should improve accountability and allow smart outside actors to spot opportunities for improvement, but so far actual results have been somewhat below expectations.
Slope of enlightenment
Behavioural science has given governments a new tool not just to better understand the sometimes odd decisions their citizens make but to find ways to gently nudge people towards better choices. The method is well on its way to mainstream adoption in some (Anglo-Saxon) countries, but remains under-utilised in others.
Slope of enlightenment
Allowing citizens to conduct their interactions with government online is a somewhat unexciting but absolutely essential service to citizens. No one should be subjected to long, pointless queuing to pick up a piece of paper anymore. Government Digital Services in the UK, 18F in the US and the entire Estonian government are showing how it can be done. Building user-friendly government online services can also act as a gateway to a more digitally enabled government in other areas.
Plateau of productivity
As a structured way to cut waste in all kinds of processes, “lean” has long established itself as a useful tool for government operations too, not just for car factories. While “lean” doesn’t generate any particular excitement anymore, it’s become a productive instrument to improve government services.
Not all countries, regions or cities are in the same place at the same time on this journey. Inflated expectations in one place co-exist with bitter disillusionment in others.
The movement of new ideas, methods and technologies through the Hype Cycle is inevitable, as we collectively take promising new approaches and figure out what exactly they can and cannot do. Asking about impact, rather than appearance, can help us get there faster and discern the useful from the hopeless a bit quicker.
Some amount of experimentation where the end result is unknown is, however, inevitable and indeed desirable. Odds are that all of these methods have some valuable contribution to make to better government. It is worth persevering all the way through to the plateau of productivity if we want citizens to reap the benefits.
", "url": "https://www.dannybuerkli.com/2016/05/18/ten-government-innovations-and-their-place-in-the-hype-cycle", "image": "/assets/cards/essays/ten-government-innovations-and-their-place-in-the-hype-cycle.png", "date_published": "2016-05-18T00:00:00-03:00", "date_modified": "2016-05-18T00:00:00-03:00", "author": { "name": "Danny Buerkli", "url": "https://www.dannybuerkli.com", "avatar": null } } ] }