The potential of digital for city government — Part 2: Considerations and institutionalization
This is a continuation from my previous blog post which you can read here.
An interview with Jascha Franklin-Hodge
I followed up with an interview with Jascha, who is currently a fellow at the Harvard Kennedy School and is researching the need to embrace common standards in data between cities. Edited excerpts are below.
What should cities who want to use digital tools to transform themselves keep in mind?
It is tempting to build an ambitious digital infrastructure with all the latest gadgets but that poses a huge risk if you don’t deliver value. Above all, you need to find places to generate buy-in. Don’t hire a team of data scientists expecting them to reform the culture. Instead, get a contractor and a research fellow to work with your current IT team and duct-tape your way to a successful demonstration. With some patience you can start a positive feedback loop that encourages the entire organization to move to a more data-driven culture. In the case of Boston, we were lucky because when we started tackling these projects, the Mayor Office of New Urban Mechanics had already been planting the seeds of rapid experimentation for several years. When we proposed making an iterative website re-design nobody look at us funny, the authorities understood that not having a finished product right away was actually a good thing!
It is a political and a strategical process. You create value in one department, get more reliable data, other departments want to get in that train and you suddenly have the leverage to create a more elaborate data warehouse. It’s a slow and rocky process, and sometimes it backslides, but you have to keep at it.
You have recently been writing about the importance of the need for common data sharing standards and practices between cities in the topic of transportation. Why is this an important issue now?
When it comes to transportation, cities are moving to a future where they no longer have only individuals making choices about how they move, but rather a mix of private companies and public agencies offering services. It is not enough to just know where the roads are, we need real time data to set standards and provide outcomes that foster equity, speed and accessibility. We also want to design infrastructure that supports all types of services, be them car pooling, dockless electric scooter sharing or public transit. The problem is that companies such as Uber and Lyft can roll out a single framework that gathers data from cities all over the world.
When it comes to cities there is a lot of cross-regional overlap, so in Cambridge, Sommerville and Boston, things are overseen by the MBTA. But there may also be an opportunity to create economies of scale if cities set standards and cooperation procedures. This is a nascent and unproven field but there are a bunch of promising efforts being rolled out right now.
What excites you about the future?
I think that building a digital service team that can ultimately help create better interfaces for humans, but is still fundamentally rooted in a model offering services and people approaching city hall and asking us to solve X and Y and Z. In the long-term I think we should aim not only to get better at service delivery but also in creating two-way communication channels that allow for a better understanding at what’s going on deep down. What I mean is that we shouldn’t just get the message that we should fix the pothole. But rather, understanding that people feel unsafe in certain street and that is due to a variety of reasons that are not so easily quantifiable. The city can also become a platform for connecting people and institutions. An elderly person may not be able to get out to the street because of the snow in his front yard, but someone two blocks down the road may be willing to come and shovel. A lot of places speak about the importance of social cohesion and building a better social fabric, but very few put it in practice. I believe that digital can help.
Considerations
There is a number of issues that came up during this research that cities wanting to do a digital transformation should consider:
· To get started, focus on the people, not the process. Claudio Orrego was governor of the Metropolitan Region of Santiago in my home country of Chile, and as Mayor of one of the poorest urban municipalities (Peñalolen) was twice awarded with recognitions for its efforts to digitally transform the organization. When I asked him about how he managed to streamline the process of car registering in a Municipality that at the start of his term had one computer per every five workers, he replied that the most important part was that he never went over people. “We didn’t bring in an outside techie 20 year old expert to solve the problem, we put the Transit Director in charge of a project that made sense to him, gave him resources and a competent back-office staff, and we surpassed our goals quickly,” he added. Emily Monea, director of SomerStat at the City of Somerville had a similar point: “We have to become an asset and a partner to the rest of City Hall”.
· Complexity of Analytics. Although there is a lot of value to be gained by employing fancy machine learning algorithms, sometimes a regression analysis or even an API that more consistently gathers and distributes data is enough to create a large impact.
· Money will be an issue. Even for big cities such as Boston, data scientists may only be able to get a salary that is a fraction of what they earn in the private sector. And that would already be a controversial figure since even half of what a data scientist earns in a tech firm would put someone ahead of what most city hall head of departments earn. But as millennials seek not only money but meaning in their work, cities could emphasize the huge impact that they could have working for the institution. Smaller cities may consider partnering with local universities or NGOs.
Ethics in AI: When it comes to predictive analytics, a lot of warning alarms have been raised about the possibility of encoding systematic injustices through algorithmic vagueness. Properly used, data can help unmask and address inequalities. If they aren’t properly understood, they could become black boxes. Some efforts, such as the one undertaken by GovEx, seeks to give local government leaders a framework to understand some of these risks.
Institutionalization
“Yet when it comes to wider impact, the greatest shortcomings of innovation teams come from one of their biggest strengths: they are outsiders. Because the team is external to municipal budgeting structures, organization charts, and operations, they almost never have an impact on them. An innovation team can accomplish great things for a city, but when the project is over — and external funding is expended — cities still need to fully integrate the approach.”
A New City O/S, Kleiman and Goldsmith
To take a more comprehensive view at the tasks and tools being used by data officers throughout cities in the United States, the research by Jane Wiseman is a great starting point. As we can see there isn’t necessarily a single direction that cities can follow.
But another way to think about the different options that cities have is to see what level of depth these technologies and methodologies are penetrating the heart of government functions. A recent OECD report argues that innovation can happen outside of the organization (hiring a consultancy team for a specific goal), at its periphery (an innovation team inside the organization that works in specific hand-picked projects), or it can permeate the entire institution. In the following diagram, I’ve tried to graph how digital transformation can follow a similar pattern.
Cities may choose to adopt specific tools for a specific function, developed in-house or purchased from a civic tech company. Examples may include predictive data analysis for inspection or modelling through GIS technologies. At a second layer, they may move beyond specific tools and adopt the principle (say, data-driven decision making) across several different departments.
But we should also keep in mind that there are certain processes that sit at the core of government functions, and these are central and unskipabble. By this I refer to things such as procurement, as a big part of what a local government does is buying from other providers. We cannot forget the importance of Human Resource policies, often subject to old bureaucratic rules. But as much as we incorporate new technologies, the people overseeing processes and interacting with people will be at the center of a city government.
But this does not imply that for meaningful transformation we may only start at the bottom. There is not a definitive strategy applicable to every context. As the example from Boston shows, setting the ground by showing the capabilities of one data driven project may make other departments see the value of a collaborative data approach, thus slowly integrating principles across the organization. A performance management system that starts at one department may slowly change the way that the city rethinks its budgeting. Or IoT sensors in cleaning machines may be the kick needed to make the city rethink its procurement in a smarter way.
