Sean Boots

Technology, public services, and people. But mostly people.

Coasian hecks, or, when the people in charge can’t change things either

Conversations around “state capacity” – can government organizations do big things? Can they meet the complex challenges of modern society? – have taken on a new urgency over the past few years. Pressing social and economic issues, confronting the consequences of climate change, and navigating disinformation campaigns and threats to democratic participation all require an effective, nimble, and robust public service.

One of the themes of my writing here is that I really want our federal public service to be excellent. There’s a lot of ways it could improve. Although I’ve recently begun working for a territorial government, I’m always thrilled to see more public conversations on how our federal institutions could do better.

Over the past year, one unexpected and welcome source of those conversations has been recently retired Clerks of the Privy Council sharing their perspectives on what the public service should do differently. The Clerk is the senior-most public servant in the Canadian federal government, the secretary of Cabinet meetings and the deputy minister in charge of the Privy Council Office (which is responsible for managing deputy minister and other senior-level appointments, responding to pressing national issues of the day, and public service renewal efforts, among other things).

Late last year, Michael Wernick (Clerk from 2016 to 2019) began writing a series of articles on how the federal public service should change for the better. The whole series is worth a read. More recently, Janice Charrette (Clerk from 2014 to 2016 and from 2021 to 2023) shared her perspectives in a CBC interview shortly after retiring:

“The public service is still working in what I would describe as kind of analog ways and the world has moved on. You can make a dinner reservation, you can book a cruise, you can move money in and out of your bank account, transfer between the two of us — it’s remarkable the things you can do in a digital world and the public service, and our service delivery infrastructure has not kept up with that.”

“In all humility, we know we have to do a better job there,” she said.

Other former Clerks, including Paul Tellier (Clerk from 1985 to 1992) and Kevin Lynch (Clerk from 2006 to 2009) have also spoken in interviews and published articles over the past couple years sharing their suggestions on how the public service should change.

One immediate question, of course, comes up: each of these former Clerks were, in turn, the senior-most public servants in the Canadian federal government. Given the problems they identify, and the solutions they propose, it isn’t surprising to ask: why didn’t they do something about these issues when they were in charge? Or, if they tried, what barriers limited their impact?

Abe Greenspon (one of my very favourite public servants) asked exactly that question in a LinkedIn reply to Michael Wernick:

As former Clerk of the organization about which you’re writing, perhaps you could consider writing about your efforts to address these issues. As much as I appreciate you advocating for change, it feels somewhat strange absent your personal reflections about how you worked to resolve these issues yourself when you had the chance. You were the most senior official of this very organization only a few years ago at a time when all of these issues were equally relevant. Perhaps something to consider for a future piece.

Essentially: the federal public service is a strongly hierarchical, top-down organizational environment. None of the problems – in procurement, in HR, in program design, in speed of decision-making – being discussed are new; practically all of them have existed through multiple governments and political environments. If the senior-most person in charge of that hierarchical structure couldn’t change things, who can?

A view of the Office of the Prime Minister and Privy Council (formerly Langevin Block) and nearby buildings, taken from the Peace Tower on Parliament Hill in early 2019.
Caption: The Office of the Prime Minister and Privy Council (formerly Langevin Block), where the Clerk of the Privy Council works.

Coasian hecks

The federal public service isn’t unique as an organization that operates in a way that is resistant to change. In some contexts, that’s seen as one of its virtues (particularly as a non-partisan institution that can continue to operate predictably through significant changes in political leadership).

In the context of operational (and non-partisan) challenges within the public service – like cumbersome procurement processes, or an inability to hire technical expertise – this resistance to change illustrates a different underlying issue.

Essentially: the public service is an example of an organization that has been structured in a way that prevents it from changing itself. (Or at least, from changing itself quickly enough or significantly enough to matter in the context of state capacity challenges.)

Alex Harrowell has a fascinating term for this phenomenon, “Coasian Hells” (named after famed economist Ronald Coase). I’ll call these “Coasian hecks” for the rest of this piece. It’s a family-friendly blog!

Alex Harrowell’s piece is about transaction costs and the shift (over decades) towards contracts and subcontracts instead of centralized organizations. He looks at privatized rail firms in the UK and healthcare in the US as useful examples of contract-based environments involving large numbers of actors. In these cases, if something goes wrong, legal claims and finger-pointing consume a lot of resources without providing much clarity or clear responsibility over outcomes.

The “cost” of determining what happened, or who was responsible, can become unfeasibly high because of the lack of information-sharing beyond the boundaries of each organizational group. The combination of diffuse actors and limited information sharing leads to negative outcomes for which no one actor is clearly responsible:

As we will see, in Coasian hell it is usually impossible to finger any particular guilty party, because its problems are system-level properties, driven by the interactions between firms in the system. Reductionism just leads to finger-pointing.

This model is a useful description of the federal public service, which ultimately is a collection of more than two hundred organizations, each with highly-structured internal divisions and sectors. Each of these organizations reports to an individual deputy head (in most cases, a deputy minister) who is responsible for pursuing their specific minister’s priorities and for complying with government-wide policy requirements.

In practice, doing anything substantial almost always involves a wide range of participating actors. Different sectors of a department (corporate IT, HR, and accounting; strategic policy; communications; specific operational and program areas) as well as separate institutions (Shared Services Canada, for IT infrastructure; related departments for horizontal initiatives) all need to collaborate. This environment ultimately takes the same form as the “Coasian heck” that Alex Harrowell describes: diffuse actors, limited information-sharing, complex interactions, and no clear ultimate owner if something goes wrong.

More importantly, no one in this kind of environment is clearly empowered to structurally change the environment itself. Each participating actor (each sector, silo, or individual department) has a clear picture of their existing (previously-established) responsibilities, even if responsibility for the overall outcome is diffused and unclear. Each actor’s funding and prestige might also depend on maintaining or defending their current responsibilities against changes, making them predisposed to be hostile to system-wide changes.

Strategies for operating in a Coasian heck

It’s probably accurate to say that most public sector work involves operating in a Coasian heck of some kind. As a shorthand, if you’re operating in a dysfunctional (or partly dysfunctional) structural environment, and no one involved is able to change that structure, you’re in a Coasian heck.

(Structural dysfunction could mean: burdensome or counterproductive policies and compliance requirements; broken feedback loops; ineffective service provider or contractor dependencies, and so on. There are many other kinds of organizational dysfunctions – interpersonal dynamics, for example – but we’ll leave those for another day!)

If you find yourself there, here are some strategies you could try:

  • Create protected spaces for pilots and experimentation. In many organizations, “pilots” are a mechanism to (at least temporarily) opt-out of established policies and procedures. In best-case scenarios, this approach can create enough organizational flexibility to deliver a prototype (or even ship it), allowing teams to gather feedback that can help make the case for maintaining that flexibility.
  • Gather and publicly disseminate data (the more embarrassing the better). Many structural dysfunctions are widely known and acknowledged within organizations, but not beyond. In Coasian hecks – where organizations are unable to change themselves – publicly reporting on service response times, transaction success rates, project costs, and other metrics can create external pressure to implement systemic changes.
  • Make decisions that increase future autonomy and flexibility to change. As Dave Thomas writes, when faced with an uncertain set of options, “take the path that makes future change easier”. Delegating authority down to lower levels of hierarchy and creating or improving feedback loops within organizations are both important ways of doing this.
  • Find allies and reduce external dependencies. A lot of inefficiency in government originates from organizational dependencies, and the resulting communications and management overhead of parallel and competing reporting structures. (Separate program teams and IT teams, or separate service provider departments, are both examples of this.) At the same time, finding teams with similar objectives in other sectors and organizations can help create a broader internal push for systemic changes. Researchers and advocacy organizations outside of the public service can also be important allies.
  • Shift the Overton window of expectations within and outside institutions. Changes within a Coasian heck result from one of two things: external pressure, or sufficiently-broad consensus on the need for change. Communicating the impact of structural issues, and the benefits of different ways of working, can help shift the “Overton window” of decision-makers’ expectations over time.

Challenging the orthodoxy

Challenging the status quo within public service organizations is hard and often lonely work. This can frequently lead to staff burnout within change-oriented digital government organizations, and within the public service more broadly.

One broader reason that change is difficult – and this is a structural issue that takes place on a personal level – is that relationships are an important form of social capital (for decision-makers and working-level public servants alike). The more that you push for change, the more you diminish your relationship-based capital. Fighting certain fights and burning certain bridges may not be worth it; at the very least, there are strong social incentives against doing so.

Public service organizations – with their job security, pensions, and comparatively long careers – tend to produce change-averse, tradition-oriented organizational cultures. They tend not to reward creative or unconventional thinking. Their size and scale, combined with a political and media environment that is unforgiving of risk-taking and failure, tends towards more blockers and fewer enablers.

Overall it’s …pretty heckish. But I’m glad that we’re talking about it!

I hope that, both inside and outside public service organizations, there’s enough advocacy taking place for things to change for the better. It’s hard not to feel pessimistic about the public service’s ability to respond to future social issues – social instability, environmental collapse, geopolitical events, at-scale disinformation campaigns, future pandemics – without substantially changing how we work.

No one person can do that, not even current or former Clerks. It will take all of us.

For more on decision-making in large organizations, Lorin Hochstein’s Moral Mazes book review and Avery Pennarun’s What do executives do, anyway? are both excellent and thought-provoking reads.