Building a Tech Policy Movement
We need to fill the pipeline of professionals who are bilingual in technology and public policy. Then, we need to give them a seat at the table.
On November 13, 2019, I gave a keynote address at the Aspen Tech Policy Hub’s Demo Day for its inaugural class of fellows in San Francisco, California. The following is based on those remarks.
In a rapidly changing world, it is sometimes hard to tell when you are at the start of a movement. Betsy Cooper and I have been talking for a long time about the urgent importance of developing professionals who are bilingual in both technology and public policy. Government needs people with the expertise to navigate a future in which technology is increasingly intertwined in our daily lives.
The Aspen Tech Policy Hub and similar programs around the country are part of a movement to bring such tech expertise into public service. As you end your fellowships here, I hope that you’ll continue your work to build a technological future that we all want to live in.
The Dance of Technology & Policy
I snuck into the tech industry over 20 years ago, as a liberal arts major, lawyer, policy wonk, technology fangirl, and a true believer in technology’s capacity to change the world for good. Today, I am hopefully the archetype of your future colleagues, enthusiastically sharing a vision with you about how technology can help make people’s lives better.
I am the one who may ask you to stay late to explain the output of that algorithm, where exactly that data is going to be stored, and whether we’ll be using differential privacy or some other technique to protect personal information.
I’m the one who wants to be your partner in this complicated dance between technology and policy to serve the public interest. I believe tech can change how we behave as humans, and humans influence how we build our tech. And, I believe policy can help shape how that relationship should work.
The Hub and tech policy fellowships get us a step closer to a future when that dance is real. A future when technology is not bolted on at the end of a policy process, but is an integral component throughout its development. A future when technology and technologists are continuing actors in how a policy unfolds over time. It is a future of policymaking that takes seriously not only the question of what technology do we build, but who do we build for?
The Problem & the Pipeline
But here’s where we really are: there aren’t enough of us. There simply aren’t enough people with tech expertise who want to put their talent into public service. We have a pipeline problem.
The need for modern technology in government is clear. Last year, our federal government spent almost $96 billion on IT. According to federal agency Chief Information Officers, about 60% of the projects today are at risk of being either over budget or behind schedule. The Social Security Administration, which pays $1 trillion in social security and disability benefits to 70 million Americans, relies on 60 million lines of COBOL, a computer language that was created in 1959. That is not merely inefficient — preventing us from taking advantage of things like open source databases and the scaling capabilities of cloud computing — it is clearly unsustainable. There are actually a diminishing number of programmers in the workforce who even know how to maintain this code. And that failure of our infrastructure and the modernization of our tech workforce puts the benefits of 70 million Americans at risk.
Government’s technology challenges are not typically, or even often, the splashy innovations showcased in the private sector. But here’s why they matter: our democracy is built on the fundamental promise that government will deliver for the people.
We are at the lowest point of public trust in government since the 1970s. And that’s not about what’s happening in DC right now. Those are statistics from a Pew Research study stretching from 1958 to 2015, before Donald Trump was elected. In 1964, 64 percent of Americans believed that “government is run for the benefit of all people.” In 2015, that number had dropped to 19 percent. Nineteen.
So, let’s pause the current conversation about the Russians and foreign interference and political polarization, and consider that trust in our government has been on a long decline. And one of the reasons is that government is not delivering for the people it serves. That’s why the non-glamourous work of modernizing a 70-year old system is critical as a matter of public policy, just as much or more as perfecting self-driving cars or putting a person on Mars.
We need DevOps and Site Reliability engineers and people who understand modern software development stacks. We need people with the full suite of skills for building technology from human-centered design to product management. We need people with deep expertise in data governance, information security, and privacy.
And those people also need to understand how their skills advance our policy goals. Yes, to upgrade old systems, but also to harness current technology to serve all people, and particularly those most in need.
Our technology priorities reflect our policy agenda. If we can order a gluten-free chocolate cake on our mobile phone while sitting in our living room and have it delivered in an hour, then we should be able to help a single mother get food stamps without having to take a day off of work to fill out paperwork and stand in line at some government office with limited hours.
We should be able to help immigrants renew their green cards in less than 6 months. This was the goal of a project that had been underway for 6 years at a cost of $1.2 billion, and was ultimately deemed to be less efficient than the paper system it was supposed to replace. In 2014, the US Digital Service sent in a 6-person team to help rebuild the system, and they launched a working process in 12 months.
We should be able to get benefits to our veterans — who have fought for our country and the world that makes our innovative, prosperous society possible — in less than 137 days, which was the average time it took for their benefits to be processed, until the VA launched a new streamlined system just earlier this year.
We can and must do better.
Our failure to serve the people who most need help is the root source of the growing divide in our country, and the challenge of our times. This is the most important technology problem before us.
The good news is that the trendline for digital service delivery in the public sector looks strong, although it’s hard to find precise numbers. In a 2018 Bloomberg Philanthropies survey of small, medium, and large US cities, 49% of mayors reported they had dedicated staff for managing data and 34% had staff for innovation. Indeed, some of these city officials are in “C-suite” roles to guide strategy across local government for technology, data, and analytics. In the last decade, we’ve gone from zero state-level Chief Data Officers to officers in 25 states, as well as in DC. It’s a large enough group that the Beeck Center at Georgetown just launched a state CDO network. There are 85 Code for America volunteer brigades around the country that collaborate with local governments and community organizations to build tech tools. Anecdotally, among those tens of thousands of Brigade volunteers, about 20–30% are actively seeking public sector tech jobs.
A Seat at the Table
Ok, so let’s say we solve the pipeline problem with programs like the Hub’s fellowship and other training grounds. We must not squander the talent of the technologists brought to public service. We must stop consigning them to the back office. They are not here to solve our email problems.
What do we need to do? Here are a few modest proposals:
- We need to take recruitment seriously, and not just for tech folks but for a diversity of tech folks. That means proactively finding the best talent — particularly among women and communities of color — and convince them of their place in public service. The big names in the industry are pursuing this talented generation of techies with free food and massages at work, and we’ll never be able to match that. But we can offer them the ability to make a difference at scale on the critical issues they care about. We can create the type of flexibility that allows these folks to do limited public service commitments for 6 month fellowships or one-year tours of duty, thereby attracting amazing private sector talent and regularly refreshing tech knowledge in government. Moreover, we must invest in this skilled and highly sought-after work force, providing them with effective mentorship, challenging projects, and clear career ladders.
- We need to send up a signal that government is safe for nerds, as my friend and the founder of Code for America Jen Pahlka says. We need to create an environment that ensures techies have modern tools to work with, and the ability to do the type of experimentation and iterative design — and sometimes failure — that is commonplace in the tech industry, but still viewed warily in government. We need to actively create a space in which technologists’ talent is not only welcomed, but properly resourced.
- We desperately need to change how we do procurement. Today, most public sector procurement processes are designed for waterfall instead of agile development. These processes assume that, from the start, the government completely understands what it needs in a project and can specify its requirements in excruciating detail. They assume that no new ideas, knowledge or conditions will arise during a project, which virtually never happens in software development. This disconnect between the process and the reality of technology development makes it enormously difficult to actually achieve desired outcomes.
- Finally, and most importantly, tech talent needs to be at the decision-making table. We cannot make good public policy on the growing list of technology issues — from encrypted communications to artificial intelligence to autonomous drones — without technical experts involved at the highest levels. We deserve better than policymakers who have only the barest understanding of how the Internet works, much less the complex choices of data science or technical infrastructure or cybersecurity. We can expand our toolset and make better policy decisions if we involve those with a sophisticated understanding of how technology works and what it can and can’t solve.
We are in the midst of a moment and a movement of tech savvy people coming into public service. But to sustain this movement, we need to institutionalize our place at the table. We are seeing this happen in small corners of government, like the Boston Mayor’s Office of New Urban Mechanics or the US Digital Service and 18F, which provide tech and other experts for new approaches to building government services. We are also seeing it in nascent local governance bodies, like Oakland’s Privacy Advisory Board which reviews surveillance technologies before they are adopted by the city, or New York City’s Automated Decision System Task Force which was created to oversee the use of algorithmic decision making in city services. These are good starts, but we need to do more.
Because here’s the thing: a movement can slip back into being just a moment, unless we institutionalize its power.
We, as a society, and certainly as policymakers, must ensure technologists’ seat at the table, value their input, engage them as the critical contributors that they are.
What do we build and who do we build for?
Ok, so, let’s assume we assemble an army of tech policy wonks, and we create a place for them at all levels of government and in municipal, state, and federal positions. Let’s even say we have the resources to put behind them.
Which problems are we going to solve? Now we get to the existential question: what do we build and who do we build for?
I am conscious of the fact that this is the month and year in which Ridley Scott’s techno-dystopic film Blade Runner takes place, and I may sound like I’m channeling Roy Batty.
From the tech perspective, I have never seen the industry more in crisis in 20+ years of doing this work. It is embattled by governments here and abroad, trolled by both mischievous and malign actors, and deeply uncertain about its own role in the world.
From time to time, I revisit John Perry Barlow’s adamant Declaration of the Independence of Cyberspace from 1996, in which he wrote to the governments of the world:
“I declare the global social space we are building to be naturally independent of the tyrannies you seek to impose on us. You have no moral right to rule us nor do you possess any methods of enforcement we have true reason to fear.”
I remember first reading Barlow’s words in Wired Magazine, and thinking he was a Grateful Dead lyricist who had gotten caught up in a weird techno-fever dream. But it does capture the (what we now recognize as naive) optimism of the time: that the cyberspace community would stand for purer values than those of the physical world; that it could self-govern; and that earthly governments would only corrupt its democratic essence.
This, we can see now, was all wrong.
As technologists, whether we build tech for the public or private sector, it is our responsibility to give proper attention to the people who will live — willingly or not — with the technology. We all live in a context of laws and governments, social norms and biases, economic, physical and personal challenges that defy being captured in code.
So, a few final words of advice:
- Build ambitiously and inclusively, to cover the full spectrum of people and their needs.
- Build systems that are responsive and accountable to the people who live with the impact of those systems.
- Be awesome at tech — and be honest about the things that tech can’t solve, and can sometimes make worse.
With that said, our most important tech policy milestones in the next few years will not be any particular project or system. The work of the next few years is to change the culture in the public sector around tech’s role, and the processes and decision frameworks that support it.
Our challenges in this century are big. The security of our infrastructure, from telecommunications to electricity to transportation. The integrity of the Census that will start next year. The introduction of more and new surveillance technologies in our cities, and the use of that data in our name.
At the end of the day, this government is still just us. No one guarantees us a functional democracy. We have to fight for it every day, across every level of government, and with every tool at our disposal. As technologists, it is our responsibility to care for our future, to insist on being part of its development, and to take seriously technology’s role in our lives, our communities, and our governments.
So, congratulations, fellows. Good luck and thank you for your service.