Public service anonymity is dead, long live public service anonymity

Wednesday, March 1, 2017

by Kent Aitken RSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Kent Aitkentwitter / kentdaitkengovloop / KentAitken

Very long read. Here's a print-friendly version. - Kent


Here’s my starting point:

  1. Public service anonymity is dying, and was probably always more dead than you’d think
  2. This will come with both upsides and downsides, and probably more of the latter than you’d think
  3. We should probably take steps to fill the current policy and culture vacuum

I find the concept of public service anonymity fascinating. That is, the convention that public servants work behind the scenes and should publicly receive neither blame nor praise for their work. It’s tied to the idea of Ministerial responsibility, for which we’ll turn to the Library of Parliament:

The essence of the theory of ministerial responsibility can be simply stated: ministers are responsible for the conduct of their departments. Reality, as C.E.S. Franks comments, "is much more complex and less satisfying than the theory."

Shielded from the publicity and politics of publicly accounting for outcomes, public servants can therefore provide honest and impartial advice to their Ministers, over a long career,  with the public’s trust to do so. They are "formally accountable internally - and only internally - to the ministers they serve." More on this later.

Public servants are clearly not anonymous in the general sense of the word. What we might call “public service Twitter” may not be obvious to those outside that bubble, but there’s an active community of public servants talking about government daily. But there are other trends:

  • Public servants’ emails, analyses, organizational charts, and and job descriptions are all subject to Access to Information laws
  • Anyone can look up who sits where in any government organization through the Government Electronic Directory Services
  • Public servants deliver speeches and presentations to public audiences

Examples like that are a part of, but not a full accounting for, the pressures against the anonymity convention. Within the concept, we can start making sub-divisions. Talking publicly about research, or trends within a field, is different from talking about one’s job and work. Talking about policy options pre-Cabinet is different from talking about implementation. And the general use of the term anonymity (the public not knowing who you are) is different from the meaning here: being held to account for policy advice and successes or failures: that is, answerability. Likewise, where a mistake is made in the bureaucracy, the Minister accounts for it, is responsible, but is understood not to be personally at fault. The end result is that public servants are theoretically kept out of the limelight.

There is, as you might imagine, a lot of gray area.

This post is going to be a fairly lengthy exploration of the concept, and I’m intending it as a conversation starter. I’d love feedback, challenges, or conversations. If it’s an interesting discussion but you’re not comfortable discussing it publicly, please feel free to re-post behind government firewalls and discuss there... while remembering, of course, that it’s still subject to Access to Information laws.

Origin stories


Where did this all come from, and why are we talking about public service anonymity anyway? You could be forgiven for thinking that it’s a meme, and only exists because public administration academics, journalists, and public servants keep talking about it. The concept doesn’t appear in, for instance, the Values and Ethics Code for the Public Sector.

We could skip to more recent history, but evidently such a piece wouldn’t be complete without mentioning Northcote-Trevelyan.

In 1853 the British Chancellor of the Exchequer William Gladstone (roughly equivalent to a combined Minister of Finance and President of the Treasury Board) commissioned a report on how the civil service should work. The resulting Northcote-Trevelyan report led to the creation of Her Majesty’s Civil Service, and set out many of the principles on which the UK (and therefore Canadian, though somewhat later) civil service operates: professional, non-partisan, and merit-based.

However, the flipside of merit is the reduction in patronage and politically-driven appointments, which meant a separation from political circles. Eventually this became a convention that made its way to Canada as well, cemented in 1918 through the Civil Service Act.

The convention got an in-depth treatment in 1995/1996 by the Task Force on Public Service Values and Ethics established by the Clerk of the Privy Council, in the form of  a report called A Strong Foundation.

That report refers to “discretion, anonymity, impartiality, and loyalty” as public service values, but also notes the fuzzy origins: “Rightly or wrongly, many public servants appear to believe that public service had always been based on an implicit bargain.” Donald Savoie characterizes this bargain as, essentially, less pay at senior-most levels and no public praise in exchange for a steady, anonymous career.

I’ll truncate and summarize the Strong Foundation points quickly, but they’re somewhat required reading for this topic. They pegged this as an emerging trend over two decades ago:

“Many public servants assume, rightly or wrongly, that the principles governing the relationships between themselves, ministers and Parliament are shifting, but they do not yet understand what the new principles are to be… [including] that the concepts of ministerial responsibility and public servant anonymity are under threat and lightly treated, and that this is undermining the foundations of public service…

[W]e do not see any reason, at this point, why it could not or should not evolve in ways that are largely consistent with the vital or essential principles of the past.”

As a side note, this is not quite a universal take, and some observers think the shift is more recent. In 2009 The Globe and Mail wrote “Few Canadians have heard of Kevin Lynch, which is as it should be with a member of Canada’s permanent public service.” Dennis Grube studied public speeches made by Clerks of the Privy Council - Lynch’s former position - as a new phenomenon in a 2013 paper.

Have the principles shifted? Let’s set aside Ministerial accountability and look at just anonymity. In 2013 Destination 2020, the most recent report on public sector renewal, called for “An open and networked environment that engages citizens and partners for the public good.” In 2015 a piece written by the former Clerk of the Privy Council included that “Canada needs policy-makers who are more networked, open and collaborative: our policy community needs to be set to ”open” by default.”

The devil, as always, is in the details; open doesn’t necessarily mean always, or two-way, or from every individual employee, or to the public. But I think it’s fair to suggest that the signposts are pointing in a particular direction.

I do not think, however, that the understanding and conventions on conduct, values, ethics, or culture have evolved in lockstep. This is territory for which there’s no map.

Government in 2017: from anonymous to open, networked, and anxious


We could ask whether the environment is changing - open government, digital media, collaboration, etc. - but I’d like to back up even a bit further and return to an opening line: public service anonymity was probably always more dead than you’d think.

It’s easy to point at Twitter and say that “anonymity is dying,” but it’s really hard to distinguish between genuinely new trends and trends that are merely new to the observer.

Public servants on Twitter is a novel behaviour, of course, because Twitter didn’t exist until 2006. However, it’s hard to say what the state of “collaboration with professional communities, stakeholders, and partners” would be in the absence of a given platform or tool. It’s likely that at least a portion of the weight we’re assigning this trend is because the digital age has revealed to us attitudes and behaviours that were, in part, always there. For example:

  • Public servants presented to external audiences before Twitter
  • Grube notes that public servants are increasingly called to Parliamentary committee hearings
  • The memoirs of Former Clerk of the Privy Council Gordon Robertson make it clear that people knew the senior public service community in the 1940s and 1950s
  • We could consider the “street-level bureaucrats” that Michael Lipsky studied in his 1980 book: front-line service delivery officials with a lot of leeway and little oversight in their interactions with citizens
  • Non-executives are named in media stories born from requests through the thirty-year-old Access to Information law

Anonymity was never a perfect arrangement, but the limits - and the behaviours that resulted - were more hidden from observers. Anonymity was always a more fragile and conflicted convention than we might admit, but I think it’s inarguable that the scope and and speed of the shift is increasing.

Anonymity, meet reality


Here’s a disjointed set of cases to consider.

First, the decline of anonymity is tied to larger shifts in governance practices. Starting in 2005, academics started painting a portrait of the shift from New Public Management (NPM) to Digital-Era Governance, which was as much built on the lessons learned from decades of NPM as it was from new digital possibilities. Some of the themes start to demand collaboration and interaction with actors outside government, beyond what’s possible through procurement, the political layer, or the creation of arm’s-length agencies:

  • Delivery-level joined-up governance
  • Client-based or needs-based reorganization
  • Co-production of services, especially in behavioural public policy (‘nudge’) fields
  • Citizens testimonials as substitutes for central regulation
  • Open book government and citizen surveillance as substitutes for central audit
  • Development of ‘social web’ processes within online government, and field services

Next: I’ve heard that social media is a thing. Almost two thirds of Canadians not only have accounts but log in at least once a month. While there’s a wide range of approaches to social media - attributed versus anonymous, professional versus personal, consuming versus contributing - it’s common to have a profile that is uniquely identifiable that references one’s professional position. LinkedIn is a good marker because it’s almost invariably both attributable and professional. I can find stats saying that 12% and 20% of Canadians use it - which includes public servants listing their roles, duties, and accomplishments.

Which is particularly important for knowledge workers, and as per the Government of Canada, “changes are occurring in the Public Service workforce with a shift towards more knowledge workers.” Having a digital presence is increasingly not only a benefit but an imperative for professional advancement. At its starkest, government simply couldn’t recruit new digital communications talent if those people couldn’t have a digital presence and portfolio, as that would mean nearly ruling out private sector and not-for-profit work later in careers.

Likewise, government digital teams are a noticeable outlier for both their level of public presence and their propensity to push against what used to be limits for how public servants talk about their work. The culture may be a smidge outside government norms:

[Department of Defence digital lead Chris] Lynch was originally recruited to the USDS by [US CTO Todd] Park. “I swear to God, I had no fucking clue who Todd Park was,” Lynch recalls. “Then I got this email from the White House. So then I looked and I was like, Oh shit, Todd is the CTO of the country!”

Basically, partially because of the “tour of duty” approach to bringing people into government as a sojourn from longer private sector careers, and partially because of the need for top talent, the rules were bent for digital. For example: The White House Gives Up on Making Coders Dress Like Adults.

The UK digital team started their blog in 2010 and started attributing posts to team members in 2011. (It took us until 2016 to get there.)  Notably, the common maxims of those digital teams, expressed through those blogs, are often essentially a (small-p) policy shift. Particularly in the case of calls to use open source software where possible.

The digital talent example also hints at the reality governments will face creating a “porous” public service with regular movement between public and private sectors, which is an increasingly common goal. The top talent in a field will have a public record of statements, presentations,  blogs, and social media expressing their views from their time outside the public sector. An interesting characteristic of the internet is that content has two audiences: the current, intended audience and an infinitely large hypothetical, unpredictable, future audience.

For career public servants, the Government of Canada’s 2013 Policy on Acceptable Network and Device Use was explicit about what employees could do publicly, including:

  • Participate in a video or audio conference with colleagues or clients from other organizations or jurisdictions through tools such as Skype or Google Hangouts;
  • Develop and share code repositories in collaboration with departments, other jurisdictions and private sector organizations via code sharing tools such as GitHub;
  • Maintain an up-to-date profile on professional networking sites such as LinkedIn;
  • Read, contribute to, or edit articles in work-related wikis, online forums or discussion groups;
  • Discuss professional issues or participate in professional associations via online forums or social networking sites

Leadership and policy is changing too. In the Government of Canada, a handful of high-priority government-wide initiatives (Blueprint 2020 most notably) have directly solicited comments about public servants’ work - and done so in public spheres. The community-led #LeadersGC tweet chats, while unofficial, get as close to an official blessing as possible through the participation of senior executives. In these cases and others, public servants get rewarded (or their wrists go un-slapped) for what could be seen as breaches of the anonymity convention far more often than they get training or reminders about what it means to be a professional in the public service.

Meanwhile, public service executives have been regularly named in media stories about the Phoenix pay system, and people have been putting public pressure on whether or not they receive performance bonuses. This goes beyond anonymity and well into answerability and accountability.

Why should we care?


Taking stock: the principle of anonymity exists precisely because it has value for public sector governance in Canada. But it never worked perfectly, and the pressures weighing on that convention are increasing. Which means what, exactly? As per Grube, again:

“I argue that the reason these changes matter is because the traditional anonymity of civil servants is linked in important ways to the impartiality of the civil service. To dispense with the former is to endanger the latter in ways that re-shape the core role of civil service leaders in a Westminster system.”

Then, quoting Peter Aucoin:

“The anonymity of public servants, as invisible to parliament or the public, disappeared some time ago…[M]inisters, sometimes explicitly, usually implicitly, expect those public servants who are seen and heard in countless public forums to support government policy, that is, to go beyond mere description and explanation…[W]hen the public value of what the government is doing is disputed, they expect public servants to rise to the challenge. To the degree that ministers can expect public servants to do so without instruction, the culture is infested with the norm of promiscuous partisanship.”

There’s a range of ways public servants can talk about government. Lindquist and Rasmussen put the ends of the spectrum as “neutral competence” and “political instrument.” Alternatively, consider these differences, taking public sector science as an example:

  • “There were X number of fish in the Fraser River in 2012 and Y number in 2013.”
  • “There were X number of fish in the Fraser River in 2012 and Y number in 2013, and the impacts will be Q.”
  • “There were X number of fish in the Fraser River in 2012 and Y number in 2013, and to correct this trend government could do Z.”
  • “There were X number of fish in the Fraser River in 2012 and Y number in 2013, and government should do Z.”

Subtle but important.

Yet, Grube writes that the core of it isn’t even “promiscuous partisanship,” but just the possibility of “perceptions of partisanship,” which he says only anonymity can solve. Any public explanation of programs or policies, however factual, may sound like support. Which is also true for social media: a public servant amplifying official government channels or spreading the word about events (I hesitate to use the word “promoting”) is likely going to be conflated with supporting that material. And on the flip side, public servants are very limited by the Values and Ethics Code in publicly addressing challenges in implementation, or expressing empathy towards affected stakeholders.

Grube concludes that “If governments are moving towards expecting civil servants to play a more public role, then some reconsideration of what we expect of administrative leaders is required. Westminster tradition needs to catch up with the new realities of practice and evolve some conventions that allow civil servants to fulfill their public roles without being targeted with allegations of partisanship.”

I’d suggest that the “more public role” exacerbates the issue, but barely matters. Government has to reconsider anonymity and the public service’s ability to provide fearless advice regardless. Cabinet documents are still held in confidence, but it’s not hard to piece together the advice to government from the ecosystem of inputs, and people outside government now know far more about what’s happening inside government: Access to Information requests have increased by 3,000% since the law was passed: from 2,229 in 1984-85 to 68,193 per year in 2014-15. Requests are currently increasing at roughly 13% per year.

I’d also suggest that governments don’t have to wonder whether they “[expect] civil servants to play a more public role.” I’ll invoke the argument from inevitability: whether it’s public servants being officially sent out to events as representatives or simply the combined portrait of their public activities, public service anonymity is good and dead. To resuscitate it, the entire contract for public service professionalism would have to be rewritten for the modern age, and very principled lines around public statements drawn and enforced. Which would have to be for both personal or professional contexts; the line between the two is disappearing.

And, while this may sound unlikely - particularly coming from me - I think that should be on the table.

But as it stands we’re in a bit of a policy vacuum, and it’s easy to imagine downsides:

  • Actors in the ecosystem could put pressure on one another to “not put that in writing” and hamper fearless advice
  • Public servants could paint themselves into corners through politically benign public statements that get viewed as partisan by observers
  • The impartiality of the public service could come into question (more than it already is) if advice is seen to change between political transitions
  • Public servants could asymmetrically build both relationships with, or understandings of, different stakeholder communities (e.g., what if the Director of a policy shop has a relationship with Google but not Facebook? What if they gain insight into stakeholders’ needs but they’re only connected to an unrepresentative sample of stakeholders?)

And let’s be honest, the word “could” is doing a lot of work in those bullets. For instance, research from South America has shown that transparency initiatives do indeed make officials more reluctant to share information.

So anonymity is dead, it’s going to cause some issues for the public service’s ability to provide advice to Ministers, and we should probably re-professionalize the public service in a new context and put some serious legwork into policies, principles, cultures, and training: to protect both the public service as an institution from perceptions of partisanship, and to protect public servants from pressure against fearless advice.

No individual will ruin the standing of the public service as a politically neutral, professional institution. But it could happen as death by a thousand cuts, a slow culture and perception shift driven by thousands of actors exploring an increasingly public world - and with many more citizens and journalists noticing that exploration. Which also means that it’s a tragedy in the commons situation, and no one actor has a reason to change behaviour. Particularly when a digital presence can help people’s careers. And the public service, as an organization, rarely drives people’s training and education; it’s mostly driven by individual employees, which makes it hard to reinforce messaging on values and ethics repeatedly throughout careers.

The good news


So you weren’t planning to renovate your kitchen, but some pipes froze and you have to rip out the counters anyway. Might as build it the way you want it this time. As for the public service contract, we should at least consider if there are changes we want to make, and I’ll suggest some possible benefits.

One, if the nature of accountability is changing, we could reconsider the relationship between accountability, responsibility, authority, and expertise. Right now we hold Ministers accountable for the minutia and complexity of vast portfolios, including things for which they cannot be reasonably expected to have control or expertise. Likewise, we make public servants with subject matter expertise responsible for activities over which they have no authority. (And somehow measure their performance on that.)

I cannot stress enough how important this concept is: that we hold Ministers accountable for the minutia and complexity of vast portfolios. Consider:

The anonymity convention was born (roughly, subject to debate) in 1854.

Skipping ahead almost a century, in 1937 the entire Ottawa contingent of the External Affairs department (now Global Affairs Canada) was eleven officers. In 1939 there were only six missions abroad. Now it’s 5,890 employees.

Similarly, the federal public service was 46,000 people in 1939. Today it’s more like 260,000 people.

And compared to the mail-based eleven-person External Affairs of 1937, modern governments have several orders of magnitude more contact points with stakeholders and citizens. We recognize the complexity of issues and the need for collaboration: both of which require significant time investments and on-the-ground, user-centric knowledge. We designed public institutions around a long-standing belief that public policy issues were far simpler than they are, and for policy shops, meaningful connections to stakeholder communities are a vital source of insight, understanding, and reality about how policy decisions affect people. And they’re likely to be two-way connections. It seems reasonable to at least consider where individual accountability might meet the original reasons for anonymity halfway. As the head of the UK Government Digital Service, Mike Bracken, put it:

[The Permanent Secretary at the Department of Business] says that the best way to get impartial advice is to ask an impartial adviser, if there is such a thing. These days, sadly, those impartial advisers don’t know enough. Advice must come from the doers – the people at the coal face that understand the shortcomings of their services and who are brimming with ideas for how to make them better.

This, in my view, is the response to what I’d consider the best argument for maintaining anonymity and bolstering the primacy of the Minister as the person responsible for a department. To suggest a bigger role for the public service could easily be conceit, where Ministers are the democratically legitimate representatives and, of course, should be speaking for the portfolio and setting policy direction. I agree, but I also humbly submit that to consider that principle in an absolute form in 2017 is somewhere between nuts and impossible.

Two, there are benefits that will flow from the move towards open governance and citizen engagement that are simply in conflict with the ideas of anonymity. Governments have been consulting extensively online and in-person. So for starters, we can throw anonymity straight out the window when a public servant is representing government by sitting in a room with stakeholders; they’re going to associate that person with the eventual policy decision by default. As well, the value proposition for participants is in knowing that they’re giving their input to someone who can do something about it. There’s no engagement without a degree of authority, and no reason for trust without a degree of accountability.

Third, governments have long relied on their authority in communications, perhaps at the expense of credibility; telling, not showing. The exact phrase “The government is committed to” gets 1,200,000 results in Google. The idea is that if it came from government, people should trust it. Yet what we’re seeing is a decades-long decline in trust in government, and Canada was an outlier in the 2017 Edelman trust barometer: a -10 percentage point drop from 2016. This is also a tragedy in the commons: governments should be more forthcoming, credible, and honest in their communications and comments to media, but no individual should. One misstep that hurts the government outweighs a hundred moments where an honest interaction with a journalist would add quality and context to a story. As Deloitte’s Bill Eggers puts it:

“Bureaucracies tend to punish failure, which is memorable, and reward mediocrity, which doesn't produce a metric you can pin to a file.”

The current construct also means that governments can’t actually ever use social media for anything than a broadcast channel. People interact with and have relationships with people, not governments. (See: Why Government Social Media Isn’t Social.)

Yet. As per Edelman: the most trusted source of information was the employees of governments and firms - on multiple topics.

It’s not like there won’t be downsides to pushing more authority to officials to interact with citizens and stakeholders. Too far, to some extent, that way madness lies. But the question should be on the table - largely because it's already happening, but without guidance.

So where does that leave us?


For those who currently have a public presence, and get benefit from it: don’t lose sight of the fact that this is not an unambiguous win for government writ large. There are risks both short and long term. The public service is built out of the collective expertise and experience of public servants - but also their relationships and biases. For better and worse.

Public service anonymity is dead, and perhaps always was. If so, it’s extra dead now.

We may be able to maintain the value - in particular, the ability of the public service to provide fearless advice - but there’s work involved. It will require a hard look at the public service contract and culture, and a legitimate effort to create and embed a new set of principles for public service that honour the professional, non-partisan role while being realistic about an unavoidably public public service. The good news is that there’s a lot of potential for better governance, if this is done well.

I dislike analyses that end with  “Governments have to think about X” rather than providing concrete recommendations. In this case, however, the issue is too fundamental to our democratic institutions. This cannot be meaningfully explored without elected officials, and so if pursued, it - anonymity, but within the larger question of accountability - would have to be a question for a commission, task force, or a Parliamentary Committee. My recommendation for this piece is that it should be. In the meantime, organizations' heads are responsible for making sure that their employees have an appropriate understanding of public service values.