Reflections on six weeks: Digital Government: Technology, Policy, and Public Service Innovation

Before coming to Harvard Kennedy School, I worked at Propel — a mission-driven tech startup that is dedicated to making the social safety net more user-focused, starting with SNAP (food stamps). From lengthy forms to arbitrary recertification rules to work requirements, I was already familiar with many of the ways in which our government frustrates citizens by being slow to digitize and respond to user needs.

I’d dedicated significant headspace over the past few years to thinking about how we can improve how citizens, particularly the marginalized, are experiencing government services. I had not spent a lot of time thinking more conceptually about why improvement is so sorely needed — about why it is so challenging for us to improve and innovate within government.

I often leaned reflexively on two reasons: government is unable to attract talent comparable to the private sector as easily, and partisan interests create roadblocks to hamper programs or policies they don’t like but can’t eliminate through law (for more on this, see Administrative Burden by Donald Moynihan and Pamela Herd). While these reasons are unfortunately not untrue, they are also woefully simplistic and incomplete.

Taking this course has provided me an opportunity to step back and examine more intently what we are trying to do when we take it upon ourselves to digitize, improve, and “upgrade” government for the 21st century. The following are my main takeaways to this question that have provided a bit more light to me about why it is so hard to digitize government.

1. The government can’t be compared to Google
2. The government has run and will run on paper for a very long time.
3. The platform provides immense power, but the private sector is often better equipped to take advantage of it.
4. “Open data” is all the rage, but insufficient.

The government can’t be compared to Google

In one of our first classes, Professor Eaves derided comparisons between the government and top tech companies like Google, noting that there is massive survivorship bias and as such, it’s unfair to put the government on such a playing field. Our government is one entity that has remained intact; in the private sector, companies can go out of business if they don’t do well, and thousands of companies have done so. Nobody is comparing the government to AskJeeves, Geocities, ICQ, or CompuServe, or any number of old companies that were unable to innovate, keep up with the competition, expand their reach, or avoid obsolescence, and have now been mostly dropped from our daily lives — but that would actually be a more apt comparison!

Which leads into the next point:

The government has run and will run on paper for a very long time

In our first class, we talked about the medium of government — the fact that paper has been the organizing unit for about 2000+ years of systems, processes, and rules. The combination of paper + print creates power, and that was what made the early Roman and Chinese empires possible. Perhaps the first “open data portal” was the Doomsday book, Britain’s earliest public record, dating back to 1086.

This standardization has made governing and centralization so much easier, since they could now impose language and process. Over the centuries, other innovations that we take for granted now, such as ordering streets and giving people last names, have created all these other institutions that have made it possible and easier for the state to impose control.

The historical context made it easier to see that wherever we are in the journey to digital government, it’s early. Transitioning to paper was done because it was easier for the state to govern and impose control. However, transitioning to digital may ultimately be slowed down because it can help the state gain more control, but in many ways it can also cause them to lose control or be fundamentally destabilizing. (Yes, the state has more information about you, but conversely, you are also gaining more information about the state — whereas with paper, it was more uni-directional).

In class, we discussed numerous examples where both government AND citizens are operating using paper, despite other options — our ballots during elections, our tax returns, passports, and not to mention our COVID vaccination cards.

A key example that to me illustrates the control element is the paper trail for gun registration — the firearms transaction record, ATF4473, is always kept in paper form in the dealer shop, and it is explicitly agreed that these are in manual form so that there can be no computerized registry of these records. A digital record might provide the state with more control over crime and more power to enforce laws, but the paper record provides them with less control over ownership and private property. So, digitization is an enormous wedge in who has control and who doesn’t — and as such, is inherently political.

The platform provides immense power, but the private sector is often better equipped to take advantage of it

Another conceptualization from this class that has been valuable so far is the idea of a platform. A platform is something that enables others to build upon that foundation — it’s a fundamental piece of infrastructure, a launchpad for innovation. Examples of this include the federal highways, constructed because Eisenhower took a vacation in Germany and was inspired, but then became the foundation for all these other usages.

Our class examples showed how much power there is to an organization that is able to get this right — like Amazon, or Estonia. However, the class also revealed to me just how hard it is for the federal government to reach that kind of vision. Amazon is a relatively young private company, and Estonia is a relatively small and young nation-state. While there are some interesting examples of success stories in certain departments, city, or state governments, there’s nothing at the age and scale of the U.S. government that has successfully leveraged platform power, and I struggle to imagine that there will be. In comparison, we see many private companies that have successfully managed to build platforms, which has enabled all this other innovation — again, this is why we have to temper our expectations for government a bit because so much of what we see in the private sector nowadays as “innovation” is actually just finding and building out on top of a highly coveted platform.

“Open data” is all the rage, but insufficient

For my first job, I was working at a non-profit that was focused on tackling digital inclusion. At the time, there was a lot of interest in open data. Cleveland didn’t have an open data portal, so my organization thought that we would aggregate all the data in one place and create an open data portal.

I wish at the time that I had read about Government as a Platform and the principles, as outlined by Tim O’Reilly, because I would have known that open data was one of any number of ingredients that are necessary to help gain value from technology, and we may have been over-indexing on that particular aspect. Everyone has an open data portal — Saudi Arabia has an open data portal. But as we saw in the examples with Estonia and the UK, other factors, like simplicity, participation, and leading by example, are critical. In addition, leveraging opportunities to learn from the data, from hackers, and from experiments are really important as well. The question should not just be, do we have data and are we collecting it and is it available to people. The questions should be, have we created an environment and an ecosystem where we are gaining insights from data and returning that back in a way that we can implement changes and learn?

So what now?

We’re standing in the midst of a ton of opportunities to take this giant ship that we call government, and move it at least a centimeter or two in the right direction for the benefit of us all. We have to be patient, and earnest, and here for the long haul.

It seems like a key way to navigate this pivotal moment is just to keep your eyes open for opportunities — especially the ones that you can pitch yourself. Sometimes it doesn’t start by being formally asked to take on a project — it starts with one person just writing down the problems they’re seeing, and creating a bit of value, and seeing how far they can take that line of inquiry — as we saw in the CalFresh case.

By being open and looking for ways to insert platform thinking, being mindful of the power of data and who is gaining and losing control through its usage, and being patient and respectful of the gravity of what we are trying to do, I do believe that we can make a difference in digital government.