Every Tuesday, Jessica Ramgoolam heads down to the New Amsterdam branch of the New York City Public Library, sets up a small folding table, and takes a seat with her laptop. She lays out piles of paper flyers, and it’s clear she has information to share, like a fortune teller awaiting a passing seeker.
Just before 11 a.m., when the library opens, people may begin lining up for her assistance. With the aid of her team, she can communicate with people in nearly 20 languages, and her iPhone can help her manage many more.
Though she holds no unique powers of foresight, Ramgoolam represents for many the keys to the future. Sitting behind a bright yellow sign reading “GetCoveredNYC,” she’s there to help people—anyone—enroll in health care.
Determining what programs you might be eligible for, gathering the bewildering amount of information required for different applications, and navigating the submission process is a headache, even for the most administratively savvy.
That’s true even though most New Yorkers have already submitted information about their income and employment to the city many times over, and more and more residents get regular updates from and about the city government through websites, phone calls, chatbots, text messages, Twitter, email, Facebook and Instagram, livestreams, TV, and radio—all of which are used to communicate everything from emergency notifications to trash collection schedules. Not to mention the overwhelming volume of information online devoted specifically to the several public health-care plans available.
But even with those programs and a variety of tax credits, there are still hundreds of thousands of people in the city who do not have health insurance.
It’s a reality of politics that is often overlooked: once a law is passed, it needs to evolve from an idea into a plan with a budget and a staff, and from there it needs to actually reach the lives of millions of people. Moving from policy to implementation has always been a hard part of governing, but today it’s easy to assume technology can make it easier.
Yet even as technology presents unprecedented opportunities to bridge the gap between government programs and the people they serve, it also brings unprecedented challenges. How do we modernize without leaving people behind? How do we increase access without unduly burdening citizens? How do we increase efficiency and make services easier to use while still protecting sensitive data?
Today, technology is both an instrument and a medium of government, and in turn, it’s transforming the way citizens and states interact with each other. And it’s essential, even urgent, that governments understand this relationship—and how easily it can be broken, even by the tools meant to bolster it. After all, civic technology has the power to help, but not everything can be technologically simplified. Not everything can be automated. Bureaucrats can make forms all day long, but they are useless if people don’t know how to use them—or if they don’t even have the resources to access them or fill them out.
Which is why, every week, Ramgoolam supports uninsured New Yorkers as they navigate the ever growing, ever changing, always tangled web of online forms that promise access to affordable care.
“I’ve come across, in my lifetime, so many folks who have had many detrimental issues with the health insurance system,” she told me. “What motivates me is how great it makes me feel to know that I’ve succeeded in helping someone.”
New York City is something of a test lab for strategies to confront some big problems that plague the modern state. Akin to a country in the budget and bureaucratic complexity of its government, it is, and has been, dealing with the key question of how to make government work for people today. And through its experimentation, it is finding that sometimes the solution to doing big things also involves doing a lot of small things, sometimes with the lowest tech possible: a human sitting behind a table.
“Why can’t we just …?”
When President Barack Obama took office in 2009, his administration was heralded as more technologically savvy than any that had come before. At the dawn of Web 2.0 and with immense faith in the power of technology to do big things, it hired the country’s first chief information officer, started the US Digital Service to modernize the executive branch, and issued a directive to “build a 21st-century digital government.” Technology was envisioned as a key to the administration’s ambitious plan for expanding access to health insurance.
Yet when Healthcare.gov launched in 2013, after three years of work and a cost of more than $300 million, the website crashed. Fewer than 10 people were able to enroll on the first day.
In the years since, the Healthcare.gov fiasco has turned into a sort of parable for those working in policy implementation. The program’s tech-forward approach was meant to make it easier for people to compare the costs of health-care plans and enroll in one, but at least at first, the tech failed in spectacular fashion.
The crash was indicative of massive challenges that the US still faces when it comes to government use of technology. Jennifer Pahlka was serving as deputy chief technology officer of the White House Office of Science and Technology Policy at the time. As she explains in her book Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better, the failed site launch was a reflection of just how big the “glaring gap between policy intentions and actual outcomes” really is.
In the book, Pahlka—who also founded Code for America, a nonprofit that pairs engineers, designers, and product managers with government agencies to improve public services—lays out the problem. “Whether for good or for ill, the essence of the digital revolution is that it has become easier to implement ideas of all kinds—business, cultural, and social,” she writes. “Inside government, however, the digital revolution has played out very differently. Even as our expectations about the immediacy and accuracy of services have skyrocketed, the implementation of laws has become anything but easier.”
In several conversations, Pahlka explained to me how well-intended policies morph between the time they pass a legislature and the time they finally trickle through the bureaucracy and down to the lives of everyday Americans. And today, of course, the way Americans interact with those policies is so often through technology—government websites, data management and record keeping, or benefit enrollment.
“Ultimately, we tell the American public we’re gonna do this thing,” she told me, “and then the actual outcome that was desired may or may not occur.” The reason, she argued, is that policy implementation has grown so complex—and technology often complicates it even further. What’s more, the American system isn’t designed to empower technology designers in this process. Instead, legislators are making the choices without necessarily understanding what technology would help carry them out most effectively.
“We need to rediscover what democracy offers to us and apply that in the context of building services, making decisions, and doing regulation that works for people in a way that’s less like ‘Everybody throws their stuff in the pot of soup and then that’s what the soup is,’” she told me.
Officials working on digital transformation and public services in New York, San Francisco, and Boston all told me that there is no silver bullet. Technology can be as much a part of the problem as it is part of the solution. As Cyd Harrell, the chief digital services officer of San Francisco, put it, the story of government technology is a story of the question “Why can’t we just …?” In other words, the contrast between the opportunities technology seems to offer and the challenges it often creates can make modern governing maddening.
Even if the technology is promising, deploying it takes money and talent. There are challenges with procurement and integrating new systems with legacy tech. There are the realities of budgets, bureaucratic red tape,election cycles, and ever-growing legal complexities. And getting the technology itself right is no simple task, especially when citizens are accustomed to easy-to-use interfaces and information management systems from the likes of Apple, Microsoft, and Google.
It’s all these things at once that make the problems with government technology so intractable. But at the same time, it’s never been more critical to improve government effectiveness.
“The stakes matter more at this moment than they ever have,” says Pahlka. “The [Inflation Reduction Act] is trying to save us from a climate collapse, the CHIPS Act is trying to save us from potential national security disasters, the infrastructure [law] is trying to save us from driving over bridges that might fall.
“These are all core issues where I think if the American public doesn’t see government deliver, I think it’s less that they get driven toward one party or another, and more that they get driven away from government altogether.”
In fact, according to recent survey data, trust in government is near record lows. Research has shown, too, that people who have had an unpleasant experience with government services are less likely to engage in civic activities like voting—and democracy depends on this kind of involvement from the people it serves.
People invest trust in their government when it works for them. And right now it isn’t working.
Bridging the gap between policy and implementation is just what Ramgoolam, the health-care specialist in New York, is doing at her table.
She is a staffer for the Public Engagement Unit of the New York City Mayor’s Office, which was first created by Mayor Bill DeBlasio in 2015 and was specifically designed to boost enrollment in underutilized programs. Before this, New Yorkers who needed help had to call 311 for assistance or physically show up at the offices of the Human Resources Administration.
“Unfortunately,” says Adrienne Lever, the executive director of the Public Engagement Unit, “there are resources that are underused, and that is just a waste. There is a resource, and there is a person in need. We just need to figure out how to make that connection happen.”
Lever told me that often those most in need of benefits are the least equipped to navigate a complex process required to access them, and the discrepancy becomes particularly acute when someone is in crisis.
“PEU’s target populations are often lower income. We work with a lot of seniors. Many of them don’t have access to computers, let alone the internet. Some are homebound and don’t have the ability to go out,” Lever explains. “So with those populations in mind, even if the technology is not flawed in and of itself, they may not have the resources or the information to be able to just fill out a simple Google form.”
And many applications are much more difficult to navigate than a Google form. Take New York City’s Senior Citizen Rent Increase Exemption (SCRIE) program, which enables people over 62 to have their rent frozen, depending on their income, even if a landlord raises the price. The city then reimburses the landlord through a tax credit. The PEU has reached out to 20,000 New Yorkers so far this year who might be eligible for a rent freeze.
Lever told me about one eligible New Yorker, whom she identified only by his first initial. D called the city asking for help renewing his enrollment in the program, but he was missing some required documentation, including a renewed lease. He also had severe cognitive and physical disabilities after suffering a stroke, which made it impossible for him to navigate the rest of the application online, or even with help over the phone.
Benefit programs like SCRIE and those related to health care are particularly troublesome. They’re often the product of complex regulation that has been chewed on by many policymakers and regulatory agencies with lots of legal requirements, stipulations, and definitions, necessitating lots of compromises.
The frequent upshot is that these programs are implemented only partially or with so many barriers that they are inaccessible to people most in need. As a result, many policies lose their impact. The SCRIE program, for example, had nearly 76,000 people enrolled as of 2019, though it’s estimated that around 135,000 New Yorkers were eligible, according to an October 2022 status report. Many benefit programs in the city—including Fair Fares, which offers lower public transportation prices for eligible travelers, and NYC Care, which increases access to low-cost and no-cost health care—are also underenrolled.
Making matters worse, the system is always growing as more laws are written and more programs are started—but different public benefit programs are administered by different agencies, each with its own databases and registration processes. When people are eligible for a number of separate programs, which is common, they have to work through each of these agencies individually to enroll. New York doesn’t currently have a centralized database that manages city benefits, in part because of regulatory constraints that limit data sharing and in part because siloed processes and legacy technology make it difficult to stitch all these processes together.
Virtually every government office across the US faces or has recently faced a similar problem. In 2015, for instance, there were over 450 different websites just for veteran services before the US Digital Service swooped in to overhaul the online registration processes through a redesign of Vets.gov.
As the world moves online, policy implementation that doesn’t center citizen accessibility will increasingly lead to undersubscribed benefits programs or laws that, in practice, look very different from what their drafters intended.
Vivek Kundra, who served as the first chief information officer of the United States in the Obama administration, told me that government is working, even if slowly, to adapt to this new reality. “I think we have to reimagine and even rethink what we mean when we talk about policy,” Kundra said. “There’s going to be a massive impact on the regulatory front that we haven’t even conceived yet.”
Door knocking for benefits
New York City’s Public Engagement Unit has found that it needs to deploy low-tech interventions to bring people into the high-tech ecosystem. Consistent outreach through multiple channels is the most effective way it’s found to support people eligible for city programs as they cope with the bureaucratic complexity. Above everything else, the unit’s staffers aim to take some of the burden, technological or otherwise, off average city residents.
Lever told me she believes it’s the government’s responsibility to “help people break through that struggle and find the resources they need to get access to the services that they deserve.”
So the unit applies what it calls “campaign tactics” to policy implementation, proactively engaging with New Yorkers through door knocking, phone banking, text messages, emails, and public events to share information about city services like rent assistance, public transportation subsidies, and—of course—health care and help people sign up for them.
“My goal and my team’s goal is to limit the technical complexity and, as much as possible, also minimize the amount of times that you have to provide the same piece of information.”Matt Fraser, NYC chief technology officer
The specific outreach approach depends on the population involved. For young people in the city, texts alone might do the trick. If the unit wants to target seniors, it might also start with a mass text campaign, since most people are comfortable with cell phones, and quickly move on to door knocking and in-person support for those who don’t respond to texts. To reach those who are not accessible by phone or at home, staffers work with community-based organizations and in public spaces like libraries to meet people in person.
I recently tested the PEU’s system, texting the unit to ask for help with my health insurance options. I received an immediate text back and two follow-up calls the same day. When I didn’t reply, I continued to get texts and calls consistently throughout the week until I informed them that I did not need help any longer. It was almost annoying, but it was effective.
The PEU has seen that people are significantly more likely to sign up for government programs when the city comes to them, whether it’s through texts, calls, or some other approach. In one study of a campaign to enroll New Yorkers in the Fair Fares program, the PEU targeted people already registered in the Supplemental Nutrition Assistance Program (SNAP), since the eligibility requirements are similar. It found that people it texted were 46% more likely to sign up for Fair Fares than those it didn’t reach out to. And eligible New Yorkers who texted back were 168% more likely to enroll.
Avoiding techno-solutionist traps
The PEU is proving that more, or more complicated, tech is not always the answer. Shiny tech-savvy government projects touted by politicians can prove to be radical letdowns. Take blockchain voting, which West Virginia briefly piloted during the 2020 election; after much media attention, the experiment was abandoned once it was clear the technology couldn’t provide any increased security for electronic voting.
Or consider the rise, and rapid fall, of education technology programs during the pandemic; at first, Zoom and personalized online lessons seemed like a great way to replace in-person teaching, but core learning metrics dipped dramatically across the country.
In many cases, advances in technology meant to help implement public policy have actually harmed people they were supposed to help. Think of electronic health records, which have led to infringements on patient privacy, and even deaths, caused by data errors. Or the use of facial recognition in policing, which is less accurate for Black and brown people, leading to false arrests and actually decreasing public safety for large swaths of the population.
But this hasn’t stopped political leaders from pinning their administrations’ fates on new technology, even in New York City.
In December 2022, toward the end of his first year in office, Mayor Eric Adams told Politico: “It blows my mind how much we have not embraced technology, and part of that is because many of our electeds are afraid. Anything technology, they think, ‘Oh, it’s a boogeyman. It’s Big Brother watching you.’
“No, Big Brother is protecting you,” he added.
The comments have somewhat defined Adams’s style in office since. He has supported the deployment of police tech, including facial recognition, and he has prioritized incorporating technological solutions into city programs. This includes finally building a centralized database residents can use for city services—a potential one-stop shop for benefits access.
“The newly launched MyCity online portal will allow New Yorkers to go online [and] easily search, apply for, and track city services and benefits right from their smartphones or computers,” Adams said in March 2023. “We are using the power of technology to reduce the bureaucracy and red tape in our government, to help New Yorkers get the services their taxes pay for, and to get stuff done for the working people of this city.” (The mayor’s press office did not respond to requests for comment.)
NYC chief technology officer Matt Fraser has high hopes for the project, which will focus first on child-care benefits. (It’s a particularly daunting initial target; infant care in New York City costs over $21,000 per year on average, and according to the federal affordability standard, a household would need a combined income of over $300,000 in order to afford that.)
The city offers several subsidized child-care programs, which are administered by at least three separate agencies; the sign-up process previously started with a separate paper form for each of them. In March, MyCity launched a child-care benefit portal that can screen applicants for two of the programs online and at once.
“My goal and my team’s goal is to limit the technical complexity and, as much as possible, also minimize the amount of times that you have to provide the same piece of information,” Fraser told me.
The ability to go to one website, be screened, and submit one application for all city programs they may be eligible for would be a major upgrade for New Yorkers who struggle to navigate so many disparate, confusing applications today.
The Adams administration isn’t the first to try to achieve this, though. In fact, he’s the third mayor to attempt to centralize and streamline city benefits enrollment online. And while some more limited projects have had considerable success, like the DeBlasio administration’s redesign of the central screening tool Access NYC, no one managed to create and sustain the technology for a comprehensive centralized registration portal.
Ariel Kennan, a product designer and government tech researcher who led the redesign of Access NYC in 2016, told me that MyCity’s success depends on both political will and an internal investment in designing human-centered technology. The work of building the portal has been contracted out, as is common with government technology projects, even though Kennan notes that many similar projects have failed after outsourcing. Hiring contractors can lead to slow and expensive procurement cycles, high turnover, and minimal investment in technology and design teams within government, which ultimately makes it hard to turn digital services into sustainable, evolving solutions.
Noel Hidalgo, cofounder and executive director of BetaNYC, a civic technology organization, echoes these sentiments. “Technology is a manifestation of bureaucracy and its complexities,” he told me. “These systems are built over decades, and we need technologists and designers to go work inside of city government.” (Fraser said that government employees “remain very involved” in MyCity.)
For his part, Fraser recognizes the bleak history of government’s digital services, but he told me he’s committed to making MyCity a success; he sees the project as part of a greater mission.By expanding access to benefits through an easy online interface, MyCity will help “bring equity to government,” he said, adding that other initiatives to increase connectivity, digital access, and online literacy in largely offline communities are helping the city close the digital divide.
Still, there are New York residents like D, the senior citizen who was trying to renew his SCRIE benefits. For him, technology simply couldn’t replace in-person assistance. After he had an unsuccessful phone call with the PEU, one of the unit’s specialists, Hakim Hamsi, showed up at his door and walked him through the forms. Hamsi expedited D’s application, and D’s rent dropped from $1,000 a month back down to his original rate of $850. D also introduced Hamsi to a neighbor, who now helps him stay on top of his renewal forms.
“All of this takes time,” says Hidalgo.
“Government doesn’t work at the speed of the internet, and that’s fine—so long as it’s working to actually address these problems for New Yorkers.”
What’s next for AI regulation in 2024?
The coming year is going to see the first sweeping AI laws enter into force, with global efforts to hold tech companies accountable.
Three technology trends shaping 2024’s elections
The biggest story of this year will be elections in the US and all around the globe
Four lessons from 2023 that tell us where AI regulation is going
What we should expect in the coming 12 months in AI policy
The FTC’s unprecedented move against data brokers, explained
It could signal more aggressive action from policy makers to curb the corrosive effects that data brokers have on personal privacy.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.