In March, the Canada Revenue Agency’s website was shut down by its staff following an attack on Statistics Canada’s site. The precautionary measure was triggered by the identification of potential vulnerabilities related to the use of an open source web development tool. StatsCan and CRA were apparently the only two federal systems that had not updated their software to address that vulnerability.
The CRA/SC crash points to a lack of knowledge sharing. While the fallout seems to have been contained, the episode illustrates the benefits and barriers of open knowledge sharing in a large organization, especially one with lots of privacy and secrecy concerns.
Collaboration in organizations rests on the premise that they would be much smarter if they were able to take advantage of all of the knowledge embedded in them. They would be much more effective at solving difficult problems if people collaborated without first worrying about who the org chart says they should collaborate with. Any organization that is effective at sharing knowledge and collaborating across bureaucratic boundaries is much more effective than the most efficiently coordinated organization. As Lew Platt, a former CEO of Hewlett Packard, is reported to have remarked: “If HP only knew what HP knows, we could be 3 times as profitable.”
The message has not been heard. Research has consistently shown over the years that a misalignment of incentives makes people hesitant when it comes to sharing knowledge proactively. This hesitancy is grounded in a number of rational calculations: it’s not my responsibility to manage someone else’s website; knowledge is power, and sharing it reduces mine; no one is going to thank me for doing so, and I may just get criticised for it; I may be judged negatively by my colleagues and superiors for being a know-it-all; I may be advising something that is technically incorrect or an improper interpretation; it’s just much easier not to say anything.
One of the central challenges in management is how to encourage people to share the knowledge they have that could be of benefit to the organization. Knowledge management systems (KMS) have largely failed because they require that employees take extra steps to put their knowledge — tacit, contextual knowledge, difficult to express in written form — into the system. Coupled with the difficulty in finding useful knowledge once it was stored in the KMS, these elaborate systems have become impressive artefacts with little usefulness for emergent problems.
But there is good news: open collaboration platforms have been very successful. Across the Government of Canada, the GCTools suite (GCpedia and GCconnex) are shining examples of the power of open collaboration. A recent study found that GCpedia has over 65,000 Government of Canada employees as registered users who have collectively made over 1.5 million page edits to the site’s more than 28,000 pages – content that has been viewed over 50 million times. GCpedia contains items such as evergreen briefing notes, procedural instructions, and workplace knowledge resources.
Can this success be harnessed for deeper collaboration? In wrestling with this general problem of incentives against collaboration and knowledge sharing, I began to explore the potential for anonymous knowledge sharing platforms as a way to overcome the organizational cultural forces. The best systems — those that balance the freedom of anonymity with the desire to promote useful contributions and to ultimately reward contributors — provide what I call “accountable anonymity.”
Accountable anonymity involves contributions that are anonymous at the point of entry, with a longer term accountability that ensures responsible speech and that good ideas are rewarded. As not-so-good ideas are overwritten in the writing and re-writing of online collaboration systems, these disappear into the ether with no ill effect rebounding on the contributor. And that great idea that leads to the policy success of the year? A robust system will allow for the person who contributed that idea to be rewarded for doing so.
By providing a cloak of anonymity, participants experience the psychological safety that is required for knowledge to emerge from anywhere in the organization. But the equally important half of this solution involves accountability — both is the sense of the participant being responsible for their contribution (thus curbing the worst excesses of anonymous comment systems), and that they are ultimately rewarded for contributions that lead to a positive outcome.
GitHub, a digital project hosting web service which is primarily used for software development, may offer a model worth emulating in government. It is the dominant code-hosting platform on the Internet, but it also represents a new approach to collaboration that might have profound effects in the public service. GitHub features sophisticated accountability traces that could have profound implications for performance management in knowledge organization such as government departments. If individual public servants were recognized for their contributions in collaborative efforts, and given greater freedom to contribute to organization-wide objectives, their incentives to contribute might increase.
GitHub is certainly not the only platform currently available for organizational collaboration. Prominent examples include wikis, Google Docs, and SharePoint. GitHub is also not the easiest system to use: it has a steep technical learning curve and interface limitations. In addition, its original purpose as a software collaboration site makes it difficult to edit text documents.
But what distinguishes GitHub are its built-in social networking functions, back-end data capture and on-board reporting. It functions on the principles of distributed version control and openness that opens a window to the underlying architecture and that rigorously documents every change that is made. Throughout the long process of policy formation, the initial idea that ultimately becomes a policy success stays connected to the original contributor.
Tanya Kelley (a post-doc at the University of Michigan) and I examined 46 Canadian public sector institutions with an organization-level GitHub account. We found 180 public sector employees in Canada with a personal GitHub account, and invited them to complete a survey about their experience and expertise in using it (and similar platforms). We also conducted interviews with five Government of Canada employees to explore perceptions and experiences on the questions of collaboration and innovation within government.
Our data reveals that there is limited GitHub adoption and activity to date in the Canadian public sector, in part because of the tool itself but also because of a lingering government-wide ambiguity about collaboration in practice. GitHub is still resisted by all but the most technically savvy. With a peculiar terminology and work model that presupposes a familiarity with command-line computer operations and the language of software coding, using GitHub presents many barriers to the novice user. But while it is tempting to dismiss GitHub as ill-suited as a collaboration tool to support document writing, reflections on its use to date do provide useful lessons for considering the state of collaboration in the Canadian public sector and some barriers that remain to be overcome.
Most of the accounts used by public servants are held by Government of Canada employees. Within this participant community, the distribution of contributions per user follows a classic long-tail distribution: a small number of contributors are responsible for most of the work, a larger number of contributors do very little, and many users contribute nothing. While this may seem ineffective from an organizational management perspective, the central argument about the power of the long tail in collaborative systems bears repeating: the value of small contributions from a large number of minor contributors is unattainable in traditional, closed systems, but is a key virtue of collaboration. Rather than lament that a large number of contributors are not deeply engaged, collaboration allows those contributors who might have otherwise been excluded to contribute as much or as little as they wish.
And that person who makes only one contribution and is never heard from again? They may just point out that website vulnerability early enough to save you from having to shut it down for a weekend. If every federal government employee were potentially aware that the StatsCan and CRA web sites had not updated their web servers to guard against the known vulnerability, would this problem have been revealed more quickly through a site like GitHub than it did through standard channels? Maybe not. But the centralized approach promoted by initiatives such as Shared Services Canada calls to mind Joy’s law of management: “no matter who you are, most of the smartest people work for someone else.”
In our performance management cultures, there are the things we are supposed to do — and generally contributing to the shared knowledge of the organization is usually not on that list (or is not measured in any realistic way). And in our broader organizational cultures, there are unspoken strictures that discourage people from sticking their neck out. Add to this the phenomenon of organizational HiPPOs (the highest-paid person’s opinion) that tend to dominate conversations because of their position rather than the strength of their argument, and the causes of collaboration failure become clearer. The shutdown of key sites in mid-March 2017 shows the consequences and the urgency for public sector management to find the ways to draw on the knowledge of government employees from across the organization. It is a central challenge of our time, and GitHub’s model just might be the solution. It demonstrates what an accountable anonymity system might look like, one that recognizes contributors while still allowing them to stay in the shadow, content to simply share their strategic knowledge for the general good.
Justin Longo is an assistant professor in the Johnson Shoyama Graduate School of Public Policy at the University of Regina where he holds the Cisco Research Chair in Digital Governance.