Will the end of the pandemic lead to a new fairer society or will status quo reign supreme?
Week 5 of working from home amidst the COVID-19 pandemic is coming to an end. The more time that passes the less that time is a concept that has meaning. Days are blending together and the only difference in the days is whether it is a day to work or a day to not work. Otherwise, I’m starting to form new routines. It will be interesting to see how the routines need to change once society begins to gradually re-open.
Before I go into the weeknotes for week 58, I wanted to call attention to the absolutely amazing work done by thousands of public servants in Canada and around the world who are putting in ridiculous hours to roll out new programs and benefits and are managing the government response to COVID-19. My work and much of my thinking over my entire career has been in the public sector innovation and digital government space. My experience has taught me that change in the public service is incremental, a “long game” where things will change but probably not at the pace you want things to change. The idealistic early 20s view of “blow the whole damn thing up” theory of change gave way to a more pragmatic and strategic approach to change focused on taking victories small or large as long as the goal posts kept moving. COVID-19 has proven that the public service can move mountains when it needs to. It has shown the professionalism and skill of the public service to roll out such ambitious benefits and programs in the midst of a global pandemic. I think the following quote from a recent article illustrates the point perfectly: “So easy I thought it was fake”. This is the new gold standard for government digital services. This is what every service delivery team across the entire public service needs to strive for. A service so easy, so seamless that Canadians think it is fake.
Post COVID, every public sector innovator has a moral obligation to make sure the system cannot return to the “status quo”. The pandemic has disrupted many things in society and the public service is not exempt. The pandemic has proven that many public servants can do their jobs at home. It has proven that face to face meetings are nice but not necessary. It has proven the ingenuity of a talented workforce to get shit done even in the midst of a pandemic. It has revealed weaknesses in policies, programs and services that the government offers and provided insights into how we can do better. If you are a public sector innovator, it is time to start thinking about what you can do so we don’t return to status quo once our offices open back up. How can we port the best of the pandemic response to other areas of our work especially when the incentive and sense of urgency caused by a global pandemic is no longer there.
Glorification of Over Working
The pandemic has led to other insights for me. Extra time with the family (and especially my daughter who is now 19 months old), has opened my eyes to how lucky I am to get so much time to watch as she grows up right in front of me. It has also highlighted how broken our system of family and work really is. We have built a system where we ask parents to maintain full time work schedules while trying to raise children based off daycares and schools which do not operate around office hours. We are outsourcing the raising of our kids to others so we can operate off an archaic 5 day/40 hour a week work schedule. On top of that, there is either a direct or indirect cultural influence (at both the organizational and societal) level which glorifies overwork. Many see working through their lunch, having side gigs, working long hours etc. as badges of honour, something to humblebrag about. I would know as I’ve done it in the past. For example, how often do you hear during an “awards ceremony” or “awards of excellence” the phrase “through long hours and tight timelines”? On the surface, this sounds like a point of pride but if you take a moment to think critically about the statement you realize that working long hours under tight timelines shouldn’t be the quality we focus on. What about the quality of the work they did? What about how it was received? If they worked long hours and with tight timelines, is that something to be proud of or does that mean the organization under resourced the team? It says a lot about the culture of an organization how they decide to present the best of the best when it comes to award season.
So my resolution post pandemic is stop the glorification of overwork. It is not a good thing that you had to work through lunch or that you had to stay at the office until 7PM. Life is short and family is important. I hope that you can take pride in your work and I will continue to admire those who do really important and high quality work. However, this doesn’t mean that good work = long hours. If you can’t get your work done without burning yourself out, then that means your management has failed, you have failed and we should take pity on you rather than praise you. Strong point of view? Yes absolutely! But the only way to turn the tide on a culture of overwork is to admit that overwork is not something to be proud of but rather speaks to organizational failure.
Regulations as Data Part 3
Over my past two weeknote entries, I’ve talked about the concept of Regulations as Data. This week will be my final entry on Regulations as Data for now. In this entry, I want to briefly touch on the importance of developing a proper process for the creation of metadata and labeling of regulatory data.
The challenge of creating Regulations as Data is not the concept itself but rather it is the how. Regulations are owned by multiple federal departments and agencies. To generate metadata on regulations will take a collective and unified effort which will need to include horizontal collaboration and careful consideration of how to consistency tag and label the regulations. While individual regulations are owned by a single Department, the regulations may be enforced or controlled by multiple Departments. Additionally, any labeling or tagging of the regulation will need to be done in a standardized way that is applied consistently. For example, if I am holding a red apple in front of a group of 15 people, do all 15 people agree that it is indeed a red apple and just as importantly do they know about and agree upon what the different kinds of red apples actually are? Would all 15 tell me that I have a red delicious apple or would a few say it’s actually another kind of apple? When we are trying to label and tag a dataset as diverse as regulations with many Departments and Agencies this becomes an important concept. The way we apply labels (e.g. which industry does this regulation apply to) must be understood across all Departments and Agencies. Otherwise, are we applying these labels differently and what does that mean for the quality of the analysis we can do?
Our strategy to overcome these challenges involving coming up with labeling guidance, having subject matter experts label a handful of their own regulations and then centralizing labeling with one entity for the entire system with quality control checks at the back end of the process (e.g. spot check of 100 to verify accuracy). Is this perfect? Likely not. But we need to mitigate the numerous risks with such a diverse dataset that is controlled by a dispersed set of entities.
AI Demonstrator Projects (Incorporation by Reference, Regulatory Evaluation Platform, Rules as Code)
Regulatory Evaluation Platform: The proposal for the next phase of this project is taking shape. With the focus on Regulations as Data, we are looking at building the foundations needed to support advanced regulatory analysis. How can we analyze the impact of regulations on a specific industry if we do not know which regulations impact which industries? Creating the solid foundation to enable metadata creation, tagging and labeling is crucial if advanced regulatory analysis is going to take place.
Incorporation by Reference: As mentioned in my last weeknotes, this project is completed. We have been busy talking to a number of key stakeholders so we can understand what went well, what did not go so well and what could be done differently the next time we do a project like this. One of the key lessons learned is the importance of breaking up AI work into small chunks of work (like you would do with agile development and sprints) so you can test your assumptions, see if things are viable and pivot as needed. While it seems an obvious finding, it is still a critical one especially for an AI project. Many AI projects tend to leap to the conclusion that AI = solves all the problems without deconstructing the work into more manageable chunks. So in the Incorporation by Reference project that would mean the first part of the project would be testing whether AI can detect references made in federal regulations within a reasonable rate of accuracy. If that is successful, then we can see if it can identify Incorporation by Reference and do another check-in. We had a good grasp of the problem we were trying to solve, a good sense of what the solution could be but we did not know whether the solution would work or not. By breaking up the work into smaller chunks, it would give us the space to build, test, iterate and re-focus without being married to a larger vision which may not make sense as we learn new information.
Rules as Code: We are working on a final presentation and a video summerizing our Rules as Code project. I am hoping to share through my weeknotes soon. We are also working out the details of future Rules as Code projects that cover a wider variety of use cases. I can’t share details yet but do expect to see more Rules as Code in this space over the next year.
In other news, Pia Andrews and I are looking into starting a monthly Rules as Code community chat. Given the COVID-19 situation and the international nature of the community, we will be aiming for 4PM EST (so more can join) and holding it virtually. Stay tuned for more details in the weeks to come!
Week 58 is in the books. Hope you enjoyed reading. Wash your hands, stay safe and then wash your hands again (just to be safe).