Experimenting with Policy Development

Friday, November 27, 2015
by Nick Charney RSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Nick Charneytwitter / nickcharneygovloop / nickcharneyGoogle+ / nickcharney

A few weeks ago I pitched a experimenting with open policy making at Policy Ignite here in Ottawa with my friend and colleague Todd Julie. The rationale for the presentation was pretty straight forward:
The policy monopoly of elites no longer exists. Opening up of policy process creates competition between options and questionable interpretation of the facts and contexts upon which they are based. In the future, data analytics, social media and citizen consultation should allow the triangulation of different perspectives in a way that respects factual analysis and arrives at common solutions. However in the interim, as policy tools, players and perspectives continue to change, governments must choose between competing policy priorities. Foreseeing the wider implications of these choices is critical to addressing today's "wicked" problems. However, very little is known about the different policy avenues available and the type and quality of policy advice that can be reasonably expected if it is pursued.  

In the presentation we argued that, if you accept the rationale than you are also likely to accept experimentation with policy development to facilitate learning; leading us to propose the following 6 step policy making experiment:

Step 1: Pick a wicked problem

Step 2: Frame the policy question

Step 3: Set a time limit

Step 4: A/B test different approaches:

  • Traditional, internal, institutionally led
  • Open, external, crowd driven
  • Contestable, outsourced, single private firm 
Step 5: Evaluate the results vis-à-vis common characteristics of any good policy:
  • It serves the public interest.
  • It follows appropriate laws and is enforceable.
  • It aligns with the organization’s mandate and direction and accountabilities are clearly defined.
  • It is evidence-based; assumptions, options, risks, and intended outcomes are clearly articulated.
  • Stakeholders were included in the development process and ideas have been tested prior to implementation.
  • It is historically informed and addresses both long-term interests and short-term concerns.
  • It is cost effective and there is capacity to evaluate outcomes.
Step 6: Socialize the findings to spread lessons learned and inculcate a wider culture of policy experimentation

Essentially, if you had to boil the presentation down to a TL;DR it would be:
We don't know what we don't know when it comes to the different policy-making approaches that are available to us; and if we want to know, then we ought to experiment.


There's a couple of related caveats worth mentioning re: the need for greater experimentation:

  • We also don't have a good sense of levels of effort related to different policy development approaches and evaluating/implementing their outcomes. This will require some practical study.
  • We might need to rethink our approach to online engagement as a policy input (See: Thinking, Fast and Slow about Online Public Engagement) because the current tool set may fall short.
  • Regardless of what the different outcomes are in a given experiment, the last mile always belongs to elected officials and their senior civil servants, the best the rest of us can hope for is a more robust evidence base to support existing and emerging tools (See: To Govern is to Choose).
Its worth mentioning that while there has been little activity on this front in Canada, the UK has some experience with it. If this is an area that interests you, I'd recommend looking at the details on the UK Contestable Policy Making Fundthe correspondong press release from its first use, and the Institute for Government's assessment of the fund.


Thank you to the Policy Ignite team for putting on a bang up event, screening us in, and moving us up to the first slot (at the last minute, presumably to warm up the crowd, which I think we accomplished).

Thank you to Minister Catherine McKenna for stopping by the event the day after her swearing in to offer words of encouragement.

If anyone is looking to pick up a good thinker and a talented researcher, you should get in contact with my co-presenter Todd Julie. While I've had a blast thinking through many of the issues facing governing institutions in the digital era with him, he's looking to relocate to Toronto for personal reasons and I'd love that to happen for him.

Impossible Conversations: Off and Running: The Prospects and Pitfalls of Government Transitions in Canada

Thursday, November 26, 2015
by Kent Aitken RSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Kent Aitkentwitter / kentdaitkengovloop / KentAitken

David Zussman’s Off and Running: The Prospects and Pitfalls of Government Transitions in Canada is written as a playbook for politicians and public servants going through government transitions, down to the length and nature of briefing material required, when to schedule particular meetings (e.g., between the Prime Minister-designate and the head of the federal public service), and how to prepare for them.

However, there’s a bit more to it: because he explains government transitions through real-world, recent examples, it also serves as a bit of a history book on the political-public service interface, an element of the literature we’ve felt missing from any other book we’ve read on Canadian politics. We’ve been through a few:

And with the exception of Savoie’s book, the role of the public service came up only occasionally. I suppose a book on government transitions affords a unique lens, as one of the key functions the public service plays in the larger governance ecosystem is shoring up the long game and providing continuity throughout successive elected governments.

Which is the theme, as in this quote from Jean Carle, Jean Chrétien’s first head of operations: “The day after the election you are the government, and the old government is gone.”

What happens then? Well, what’s already in the works? Can parts of party platforms be implemented in the short or long term term? With ease or with difficulty? Is the public service organized and geared to deliver what’s being asked? These are the kinds of questions that form the public service-political interface during transition.

On non-partisanship 

Are bureaucracies non-partisan, or promiscuously partisan? (See: Nick’s post on the topic.) Rock steady, or a bit too quick to play nice with incoming governments?

We make a lot of hay about the non-partisan idea, which is a central concept to Zussman’s book. Zussman collected a variety of views in the course of interviewing former members of transition teams. Some of these takes on non-partisanship are illustrative, such as that bureaucracies essentially learn to function under a certain government, or that bureaucracies can be too cozy or friendly to party interests. But Zussman mostly paints a portrait of a new normal emerging at each transition: each new leader and each new minister having a different style of accepting and considering public service advice, and different ideas on how to deliberate - and who to deliberate with.

“In my experience, most federal ministers don’t worry about partisanship. They recognize that public servants, like everyone else, have biases, they vote, and they believe.”
- David Zussman

“It doesn't mean you’re an empty vessel. There’s nobody I worked with who didn’t know where my biases were. And no one I worked with shared my bias entirely. I would give the best advice I had based on what I learned from the public service, what I learned from research, and from my own experiences, which included my biases. It was there, laid out honestly. That’s the advice portion. And then the democracy portion is when you implement the business of government. You build trust by demonstrating what the public service is all about.”
- former Clerk of the Privy Council Alex Himelfarb

As for reading

Be warned: this really is a playbook, primarily designed to be a step-by-step guide for the people wearing transition shoes. That said, it’s an interesting walk back through the last few elections’ transition periods, working through narratives and insightful stories from interviews with the people behind Canadian governance. These parts are impressive - Zussman has a deft hand balancing interesting detail without playing insider baseball. After all, he’s still in this space as an academic, working with the Ottawa scene.

Long  story short: if you’re in the federal public service and have been wondering what your senior executives have been up to lately? This is it.

by Nick CharneyRSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Nick Charneytwitter / nickcharneygovloop / nickcharneyGoogle+ / nickcharney

There’s really only one thing I’d add to Kent’s review is that reading the book when we did was timely, reading it now (post-transition) less so. That said, it’s likely to be required reading every 4 years (or less) by those working on or managing through government transitions.

Thoughts on how we think about online collaboration

Wednesday, November 18, 2015
by Kent Aitken RSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Kent Aitkentwitter / kentdaitkengovloop / KentAitken

I threw some of these thoughts out on Twitter the other day; if this is redundant,  you have my apologies.

In October I presented at the Conference Board of Canada's Public Sector Social Media Event on the topic of online collaboration, drawing heavily from a post called The Promise of Online Collaboration. The short version is the hypothesis that we are probably not good at online collaboration - we meaning everyone (not just government), and online collaboration meaning digital-based working groups, citizen engagement, or participation in online communities*.

In the talk I suggested that it's the responsibility of those convening online collaborations to become familiar with a much wider variety of formats so they can more appropriately design interactions (as Jared Spool has written, "design the design meeting").

In response, one of the questions in the Q&A was along the lines of "It sounds like you're suggesting more tools, which hasn't really worked for us in the past."

Fair point, though I'd say it's actually a handful of things:
  1. Breaking the standard mental model for online collaboration
  2. Being wary of the defaults that tools drive people towards

1. The mental model

There are some beautifully creative online systems, but as a general rule there's one core: threaded comments. This forms the core of most collaboration spaces (e.g. Github, Basecamp, GCconnex, blogs and websites, Jive, citizen engagement platforms). Which seems like a close analog to group discussion, but imagine this:

Let's say every time you wanted to work with a group, your only option was to open a large room, hand people a document to read, then invite them towards an infinite number of flipcharts. They can start their own, or add to others'. Depending on what time they get there. People can vote on which flipcharts are close to the front of the room. And sometimes, but not always, the person that invited everyone there would summarize the content and send it around.

Because at the heart of it, that's often how we work together online: "Here's a thing. What do you think? Feel free to add your own ideas, too."

That's crazy.

It might work. Heck, it's something we do on purpose in person sometimes. But we tend to assume that it'll work every time, for everybody.

2. Be wary of defaults

I originally had three sections, one of which would have been about choosing between platforms. For today, let's assume we have the tools we already have.

The problem is that the design of the tools pushes you towards a small set of default approaches. An obvious example would be meeting lengths. Outlook's default is 30 minutes, thus we end up with a ton of 30-minute meetings. I love getting 15- and 45-minute meeting invites, if only because it shows that the person organizing it has thought through the discussion to be had.

A collaboration environment built around threaded comments leads us to the default of posting a document, blog, or some framing thoughts, then asking people to respond.

But there are options. The question - as it always is - is what serves the goal. What'll work. For example:

Blaise Hebert's approach to a GC-wide collaboration on Red Tape Reduction was to throw the "What do you think" approach out the window and ask people simple, short questions wrapped in pop culture, once a week: "If you could personify your challenges at work, what would such a villain be called?" Once people launched in with creative answers, then he'd dig into why, and what their reactions might mean.

Or, author Sam Sykes recently spent an evening turning Twitter's poll system into a crowd-based roleplaying game, starting at this point:
Even with a very vanilla platform, there are a raft of design questions:

  • Ask for ideas first, or post a document for discussion?
  • Ask broad questions that require long answers, or short specific ones?
  • Organize an activity in stages and build on the discussion that takes place, or launch into everything all at once?
  • Allow people to go in any direction, or provide parameters?
  • Stick to one space, or supplement the main platform with other channels?
  • Get involved in the conversation and try to spark interactions, or stay back and listen?
The obvious approach may not be best, and there's probably more room for creativity, design, and intentionality than it first appears.

*In fairness to our digital prowess, we don't get off scott-free in in-person forums, either. Try Thinking, Fast and Slowor Wiser: Getting Beyond Groupthink of Make Groups Smarteror Too Dumb for Democracyor the wealth of research on our subtle gender- and ethnicity-based biases that persist in group decision-making.

What Innovation Feels Like (Part 3: Try Honesty)

Monday, November 16, 2015
by Melissa Tullio RSS / cpsrenewalFacebook / cpsrenewaltwitter / creativegov

Actual conversation.
Me: We need to project manage this thing; it's pretty complicated. The problem is we don't have tools to collaborate beyond our own team. Why don't we use Google Drive for the project tracking?

Them: It's against IT policy.

Me: There's nothing confidential or sensitive in a tracking doc, and we won't have a public link.

Them: It's better to follow policy. Let's just use what we have.
Let's just leave that here for now.

When saying "no" to absurdity is not an option

I've written in this space many times (and on other spaces internally) about our risk averse culture, which is rooted in fear. Nick has a blog post from several years back on how risk has become an excuse to do nothing (see: Risky Business: Deputy Minister or Bust). This passage describes a culture that still exists today:
I think we have created a cultural safe haven for poor decision-making at all levels. A metaphorical space where public servants of every type can throw up their hands in resignation, claiming they didn't do it, like children standing over a broken cookie jar in the kitchen attempting to absolve themselves of their responsibilities.
The thing is, a cover-your-ass culture not only supports bad decision making by managers/leaders, but has the added perk of completely disengaging staff in the worst way: over time, it takes away our natural instinct to ask "why" when being asked to do things that are absurd. When staff become afraid – or worse, apathetic – to critically questioning things that don't make sense, the cycle of stupid decision making is never broken.

There's a great blog post by Matthew Taylor, who provides advice to the UK government, that talks about the outcomes of a cover-your-ass culture: "you could see the officials wrestling with the need to provide a reality check but all too often deciding it was better to nod sagely than look career-threateningly unhelpful."

The pernicious impact of self-censorship at management levels is how it seeps down to staff. I felt compelled to write this post because I've noticed that I'm catching myself self-censoring more often these days. I'll be tasked with writing a communications strategy for some announcement, and while drafting the key points about the announcement, I'll find myself writing things that I know our minister's office wants to hear, because I know that if I try to get the plan through approvals without that language, my manager or someone above him will change it anyway. This is pernicious because at that point, there's a real possibility that my role as an apolitical public servant is being compromised; I'm unconsciously weaving in political language into a non-political product.

Seek Forgiveness, Not Permission

When asking "why" makes you the thorn in people's side, or the odd-man-out, or the "complainer" – and/or when you ask "why" and people agree about the absurdity of things, but nothing changes – something is deeply wrong. A culture that makes "why" a dirty word is a place that has stopped thinking critically.

Barriers to asking higher order questions means the policies you're writing might be solving the wrong problems because nobody spoke up to ask whether we're asking the right questions. It means the key messages you're writing might be totally off base, because nobody spoke up to ask what audiences we're really trying to influence, and what outcomes we want to achieve from the messaging. You can't move hearts and minds when your heart's just not into it because you've been beaten down into the role of some ineffectual yes-man.

The Google Drive example I started this post with is just one small manifestation of the inability (or, perhaps, some learned behaviour of willful blindness) of many people in this system to weigh the cost-benefit of taking certain risks. In truth, there were valid reasons for not using Google Drive (e.g., cultural problems – you'd have to train people on what a "google" is, most likely). Still, the message was "we don't want IT to find out and..." what, exactly? Tell us to stop using it? *Feigned shudder and best Hazel McCallion impression*: Do I look scared to you?

Seeking forgiveness rather than permission should be our new guiding principle inside government. In a place where IT policy doesn't give us the ability to set permissions for our own shared folders so we can work collaboratively beyond our own teams, there's room for some honest to goodness personal judgement about when workarounds are the only way to function effectively. I know that this mildly anti-establishment attitude might be a lot to ask of managers/leaders. But in general, a little bit of it would be hugely beneficial, because by bending a few rules – and demonstrating that the world inside can be a more productive place if our hands were just marginally less tied – maybe we could prove to The Rule Making Overlords that their rules are kind of stupid and need to be re-written. Or, at the very least, we could use modern tools to do our work within our small silo and do things more effectively without anybody being the wiser.

"Stop Whispering; Start Shouting"

When I started this series of blog posts, the language I used to describe how to break down these barriers to innovation was a bit startling to one of my former colleagues. He wondered why I described it as a "battle" or some kind of virtuous "fight." He cautioned me about using such language, I think because it implies the kind of relationship where I'm pitting myself against something, and in a way, that sort of plays right into the hands of the system. Instead of pitting against, or being a yes-man, maybe there's an alternative. Maybe the third way is creating the marketplace for new ideas/mindsets within the system, in some parallel structure that acts as one big workaround to the unnecessary layers of bureaucracy (people who know me know I'm doing this, too).

But the truth is that I tend to use combative language because that's exactly what the experience feels like, day in and day out. The process of bringing new mindsets into a very entrenched system is an ongoing challenge. Sure, it's a bit of persuasion, a bit of generating goodwill/allies, and a bit of art. But mostly, it feels like a bloody fight, and some days I'm so exhausted I want to give up.

And the real truth is that the work of changing cultural norms within the system isn't one person's job. There's no way I can do this alone. I wonder sometimes if anyone else is up for this fight.

Golden non-rules for government communications

Wednesday, November 11, 2015
by Kent Aitken RSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Kent Aitkentwitter / kentdaitkengovloop / KentAitken

We’re past the days of people calling for government writ large to be “more like the private sector.” It might still come up for certain areas or practices, but the view that private sector techniques are a panacea is fortunately out of date.

But private sector practices have a more subtle, distributed influence too. In fields like marketing, communications, or community management it’s far easier to find books and resource directed at the private sector. The software we use is almost invariably designed for the private sector (e.g., customer relationship management, web publishing, and analytics platforms). We have to mentally convert terms like “sales leads,”disregard some irrelevant ideas, and bootstrap some missing ones.

With the private sector comparison in mind, I’d like to look at some reasonable rules for government communications - and their limits.

Measure engagement

Last year I was getting advice on analytics from a friend in the private sector. She suggested mapping website pages out on a 2x2 grid to start analyzing content against views and time-on-page. It looked like this:

You can see how the logic applies to a company’s blog or product page. If people rarely find a page, but when they do find it they spend a lot of time on it, perhaps it should be promoted better. But what if your “best content” - i.e., where visitors spend the most time-on-page - is an FAQ explaining how a government program works? Or guidance on how to complete a transaction with government? If they’re reading text because they can’t figure out how to do something, that’s not good content, that’s bad service.

Consider the idea of a retail sales funnel, below. The idea is that a number of people will keep taking steps towards buying something, and you want to minimize the number of people you lose at each stage of the funnel (via losing interest, bad UX, etc.):

This isn't what we do, but the logic behind it is pervasive. I’ve heard people talk about “driving traffic” to government websites. Why? We’re not selling anything. We're not necessarily trying to get more people into the funnel. 

Maybe a stakeholder seeing a tweet and simply knowing that there’s a policy that says government data and information is open by default is exactly the ideal outcome. For many people, I don’t think clicking on that tweet and reading that policy adds to their experience. They actually shouldn't go any further into the "sales funnel". Which means that Twitter’s engagement metrics might be precisely backwards sometimes; maybe it means we didn’t write the tweet clearly enough and we forced people to waste their time finding an explanation.

Community is king

“Kent, we haven’t seen you in a while.” “Kent, X, Y, and Z people are all talking about #exln42.” “People are looking at your profile.” Etc. We get emails like that every day, plus tons of straight-up content marketing. Community is king, they say. Create a relationship with your customers. Every business wants a relationship with you. Buying show tickets in a city you don’t live in? Heck, may as well make you choose a password and verify your account for the next time you happen to be swinging through Sao Paulo. Why not.

“Kent, check out these shows that are happening this weekend [8,000 km from you]!”

But really:

“I’m not here to enter into a relationship. I just want to buy something.” [source]

In government, we want to maintain regular communications, write engaging social media content. Of course.

But maybe, we actually want to design for transience, not community. Maybe, we don’t actually want people coming back regularly, because maybe they don’t want to come back regularly. Maybe they don’t want to “check back” or “stay tuned” or “keep watching this space.” Maybe we should contact them only when we have something really meaningful to share, while erring on the side of being forthright for the sake of transparency.

Don’t make me think

Don’t Make Me Think is a wonderful book that everyone should read, regardless of whether you work in digital communications or not. It’s about designing websites, and the golden rule is that everything should be easy for visitors. They shouldn’t have to search hard for the information or task that they’re most likely to be looking for, and things should work intuitively.

But in government we have to nuance even this bible. Nick wrote last week about how government’s role may appropriately be a platform for slow, deliberate dialogue, and he’s right (see: Thinking Fast and Slow About Online Public Engagement). That will often be the approach we want. In a crowdsourcing world, a high volume of people that click Like is probably not our goal. In the private sector it's about lead generation. For government, it’s about better decisions, people feeling involved, and decisions being legitimate.

Are single-click upvotes the right way to gauge sentiments on, say, where we build power plants? It’s a sensitive, complex topic.

An example: the Brooklyn Museum asked their community to help curate an exhibit called Click!. They designed the engagement platform to slow people down. No Likes, no Favourites, no Upvotes, no thumbnails of art to scroll through. Participants had to view each piece of art at full size, one-by-one, consuming their screen, and enter a number out of 100, slowing them down just enough to make the process deliberate and thoughtful.

To be clear: the functionality was still smooth. The task was still easy to complete. They just defined the task as “thoughtfully review and rate art” instead of “rate art.”

The future is visual

Tweets with images get more engagement. Posts with videos get linked more. Visual content drives engagement. On Facebook, photos get more Likes and Comments than other content types. Youtube is apparently a whole big thing. (Here’s a bunch of stats.)

40% of people will respond better to visual information than plain text.”

What does that even mean?

Yes, of course, if we as government officials have important information and visual content is how we get people to view it or understand it, great. How to file taxes or awareness campaigns about online bullying are good examples.

Here’s a framework for how we interact with information (h/t):

We can present information such that people interact with it in a purely conceptual, abstract, rational, and language-based way. Or, we can present it in a very perceptual, visual, and even visceral way.

The rise of visual content in political campaigns and product marketing is often based on the desire to elicit an emotional, rather than rational, reaction. The images and music in some political ads (particularly in the US) are designed precisely to avoid having our language processors kick in.

More people will click on content that fits in the bottom right of that graph.

But maybe, the future of government content will often still be to elicit deliberate and rational reactions. Which means that we might find ourselves in positions where we could build an infographic, but we rightly decide not to. The future might occasionally still be text-heavy, and we’ll be okay with that, knowing that it’ll serve the needs of the country.

Lastly, plain language

Just kidding. This rule stands. Plain language is the way. Stop writing like that. (That goes for me, too.)

Non-rules are made to be broken

I'm not suggesting that we don't try to engage people, or that we skip visual content, or that we should be protecting citizens from our web content. Not at all. But we live in world where the state-of-the-market and much of our professional guidance is an imperfect analog for government's goals - and citizens' goals interacting with their government - and we must be conscious of that.

This post owes a lot to last weekend’s CanUX conference, particularly the presentations from Marsha Haverty, Leisa Reichelt, and Shelly Bernstein. Feel free to check out my rough notes.