A formal function

Before we dive in, let’s establish some definitions to help frame our discussion. In the infamous “vending machine” metaphor for government, forms serve as the primary means by which government collects citizen data in exchange for goods or services.

To ensure that government-provisioned goods and services align with public interest, government bears the responsibility of creating and maintaining information ecosystems. Information ecosystems describe an interrelated set of protocols, processes, platforms, policies, and principles designed to ensure that government-provisioned goods or services act in the public interest.

In this article we will limit our analysis to three of these elements: protocols, processes, and platforms.

• 

  Protocols are the interfaces by which government and citizens exchange information, goods, and services. This article will largely concern itself with the (paper) forms that citizens fill out to give governments information about themselves. Forms effectively facilitate a one-directional conversation from citizens to governments.

• Processes are any of a number of steps taken to either (1) aid in the function of a protocol or (2) transform data as it moves through an ecosystem. In practical terms: paper forms must be distributed, collected, validated, and stored. Government-sponsored goods and/or services must be distributed fairly and measured against success metrics.

  

• Platforms serve as the technological basis for the accrual and dissemination of data4. Platforms are difficult to architect because they are often a product of an information ecosystem’s protocols and processes—and vice versa. If a government employs paper forms to collect data, for example, they will likely distribute, process, and store those forms in a purely analog way. This limits government’s ability to retrieve and/or share data upon request.

  

Like all models, this conception of information ecosystems simplifies the reality it describes. Nevertheless, its utility lies in the perspectives it engenders. Many government agencies, for example, have struggled as they’ve worked to transition from paper-rich to digital-rich information ecosystems5. Thinking about this transition in terms of protocols, processes, and platforms effectively breaks it down into more manageable design problems. Additionally, this conception allows designers to draw upon the research and innovation already conducted in these discrete areas of study. Finally, the act of modelling information ecosystems encourages designers to consider the relationships between the various elements comprising the system. This, in turn, helps designers eschew form-centered approaches in favor of more people-centered ones.

Now that we’re speaking the same language, let's continue our analysis of “form design” by more closely considering how user-centered designers and civic hackers might work to change the individual elements constituting information ecosystems: protocols, processes, and platforms.

Protocols as interface to government

The ubiquity of protocols cannot be overstated. Whether we are signing up for food stamps, paying our taxes, or casting a ballot, every citizen is likely to, at one point or another, use a protocol—a form—as an interface to government. This ubiquity makes forms a highly desirable jumping off point for designers wishing to improve the citizen experience. (“By making the form better, we could potentially make the entire interaction better, right?”)

Not so fast. The user-experience design community has frequently noted the fallacies of approaching user-centered design by way of interfaces. Specifically, they warn that the user interface of software is just one of many levels of abstraction at which designers can solve problems for users; that it's often the case that the elements detracting from a user’s experience originate at “higher” levels within an organization, such as in its application architecture, its strategy, or its design principles. The history of design failures implies that designers hoping to improve the citizen experience by way of government forms will need to not only wield expertise in the principles undergirding form design itself, they will also need to understand and work to change the context in which forms are situated.

Designer Drew Davies shares a story that works to corroborate this sentiment. In 2011, the Federal Voting Assistance Program (FVAP) commissioned Davies and Dana Chisnell, a usability professional and Co-Director of the Center for Civic Design, to improve the design of the Federal Post Card Application and the Federal Write-in Absentee Ballot. Davies remarks: “It’s safe to say the[se] forms had not been critically reviewed in a long time (perhaps ever), and were some of the worst examples of confusing, difficult-to-use government forms.”

To improve the forms, Davies and Chisnell proceeded in the typical user-centered design way, conducting 28 usability tests throughout the year. Whenever the team noticed particular elements that confused voters, they created a new version of the form that ameliorated the cause for confusion (as McLeod did with her ballot’s instructions). They quickly learned by iterating, testing, and evaluating the designs of both forms in question.

And rather than simply figuring out which design tested well, Davies and Chisnell went one step further, seeking “to learn where voters had questions and made mistakes, and why.” This was a subtle goal with profound implications: By seeking to understand voters rather than forms, Davies and Chisnell came to understand those voter’s perceptions and rationalizations. This helped the team do more than critically assess the forms they were given, as the FVAP requested; it also gave them insight into how the instructions, the forms, and the FVAP website worked in tandem to affect the voter experience.

At the end of their research endeavor, Davies and Chisnell provided the FVAP with a report6 that included redesigned versions of the requested forms as well as recommendations for the instructions and the website that accompanied them. And that’s when the disconnect happened: While many of their visual design changes were adopted, their recommendations for simplifying the forms’ instructions were not. Why would the FVAP ignore recommendations that were based on user research?

Before we address that question it’s worth noting that this situation was possible in the first place, a fact directly related to the difference between agency and authority with regards to the design process. While Davies and Chisnell were given the agency to explore the design space, they were not given the authority to put their design into practice. That power rested with the FVAP. Davies’ story highlights a design truism: the difference between a designer’s agency and his authority helps determine the degree to which he can effect change.

Understanding the difference between agency and authority makes it easy to describe the ignorance willed on the FVAP’s behalf: Davies and Chisnell’s recommendations were likely ignored due to a mismatch in the design agency and design authority on the part of the person who initially commissioned the project. Davies even insinuates as much in his post: “a decision was made to ignore this recommendation.” Thankfully, two years later, the FVAP is under new leadership and is reconsidering the pair’s original recommendations.

Despite having acknowledged the problematic nature of approaching form design via forms themselves, let’s continue down this path. We will begin by addressing some of the inherent differences in presenting forms in print and online before reviewing some of the ways in which contemporary form design is currently shaped by the work of independent designers, developers and governments alike.

Processes and sensemaking

In late June 2014, I received a call from Michele Palermo, a state employee working with the Rhode Island Department of Education (RIDE). Michele asked if my teammates and I would be interested in helping the department in its management of the state’s pre-kindergarten (pre-K) lottery program. I instantly affirmed our interest and inquired as to the nature of the program itself.

Over the next thirty minutes, Michelle explained the situation in great detail. At that time, RIDE operated a small, pre-K program for which the demand greatly outstripped the supply. To account for this discrepancy, the department required families interested in the pre-K program to take part in a state-run lottery program. Families applied for the lottery by filling out and returning paper applications to any of the 18 classrooms in which pre-K was offered. RIDE then verified applicants and ran the lottery by randomly selecting applicants. As Michele described it to me, everything about the program seemed straightforward.

The only problem was the none-too-small matter of scale. Over the next four years, I learned, RIDE planned to quadruple the number of classrooms in which it offered pre-K. This would likely increase the number of people applying to the program by at least that amount, making their current processes untenable. Michele hoped to leverage RIDE’s partnership with Code for America to help design a pre-K lottery system that could accommodate this increase in scale.

After thanking Michele for her time, I hung up and immediately forwarded my notes to my teammate, Jeff, via email. Days later, he and I divided the processes of the program into two groups that just so happened to neatly correspond with opportunities to improve the the citizen experience and the government-employee experience: application/verification and selection.

As I mentioned, families applying to the pre-K lottery program did so using paper applications. This meant that if a family was interested in applying to more than one classroom, they had to fill out multiple applications and hand deliver them to each classroom. Jeff and I proposed that RIDE offer families a digital application whereby parents could apply to multiple classrooms with the click of a button. Using a digital application also allowed us RIDE to programmatically verify whether or not applicants were qualified to apply for a classroom in the first place.

I quickly sketched the application flow to give Michele and her colleagues a picture of what that might look like.

To improve the government employee experience, Jeff worked with Michele to hash out the details of the actual algorithm by which RIDE manually ran its lottery. Within a few weeks we had a working prototype that took a list of applicants and returned both a list of “accepted” students as well as a list of “waitlisted” students. Again, I sketched a series of pictures to explain what that process might look like when conducted through a web application.

While our collaboration with RIDE is ongoing, the story thus far provides good insight into the procedural aspect of the design of information ecosystems. Notice, for example, how little the design of actual forms has come into play: Most of the design work on our part has involved helping RIDE identify and articulate its own processes. By scrutinizing and digitizing these processes, we have provided the Department with a way to scale its program with minimal effort.

So how can form designers apply this approach more generally? Through the identification and reorchestration of an information ecosystem’s processes.

A platform for innovation

In April of this year, the Federal Department of Education (DoE) published on its website a Request for Information (RFI) asking developers how they might use a DoE API in the context of higher education data and student aid programs. Word of the opportunity—to essentially weigh in on the DoE's data policies—spread quickly around the Code for America office and soon enough a small cadre of designers and developers formed a working group to consider future use cases of such an API.

In our estimation, the DoE’s RFI posed the following thought exercise: What applications might developers dream up if given an API that allowed them to access the DoE's data about the institutions of higher education as well as student aid programs? While the pragmatist in me longed to understand the technical, monetary, and political details of how such an API might come into existence, the DoE’s RFI allowed our group to conveniently ignore those issues for the time being. Regardless of its feasibility, the DoE wanted to know what the private sector might dream up.

And dream we did. As an organization built around the conception of “government by the people, for the people in the 21st century,” the Code for America working group focused its efforts on the needs of college-bound students. After an hour-long discussion, we arrived at two problems confronting students for which access to data might help: finding the “perfect” college and completing the Free Application for Federal Student Aid (FAFSA).

If the Department of Education provided developers with a window into the data they collected about colleges and universities—for example, the algorithm powering a college’s net price calculator, the average scholarship awarded to students, the average cost of attendance, the average class size, etc.—then developers could create interfaces to colleges and universities that employed robust filtering mechanisms. Interfaces similar to, say, Kayak.com would allow students to more effectively “window shop” based on the metrics they deemed valuable.

The Code for America working group also saw ways to improve the experience of completing the FAFSA. Our first thought was to provide a simpler interface for the application, similar to the wizard-like interface employed by Intuit's TurboTax software. Upon further reflection, though, we wondered if it was possible to avoid making students file the FAFSA altogether. Why not? The IRS already holds much of the information that the FAFSA requires, and families can already download their tax information as a PDF from the IRS website. Why couldn’t the FAFSA come to students pre-completed?

Eventually our working group sketched and documented the interactions that publicly available data might inspire and sent our report off to the DoE. And ever since then I've wondered, “What might civic engagement look like if government were able to provide increased access to everything it knew about its citizenry? What interfaces might we create?” In order to find out, government will inevitably need to deal with the technical, political and monetary debt that the DoE was able to conveniently ignore in their Request for Information.

Many companies are rising to the challenge, including Seamless Docs and Captricity. Seamless Docs helps governments migrate their forms online with an interface that is heavily inspired by Google Docs. Users simply upload a PDF form and Seamless Docs automatically detects its form fields. What's more, Seamless Docs provides a tool that translates PDF forms into responsive webpages, providing citizens the ability to access and submit government forms from any device. While the responsive translation is not 100% perfect, Seamless Docs does help governments seize the opportunities that digital media presents—no change agents required.

Captricity, on the other hand, is conceptually the reverse of Seamless Docs. While it also offers government employees the ability to upload a PDF form and tag its fields, Captricity's real strength lies in its ability to quickly and easily pull data from stacks of paper forms. As the company's CEO, Kuang Chen, explained to me over the phone: “Paper forms have affordances that are important for government: they're cheap, they're reproducible and they don’t run on batteries. We're working to bridge the gap between the accessibility of paper forms vs. the accessibility of the data those forms collect.” I personally have not had a chance to put Captricity through its paces, but Kuang claims that it is as easy as tagging a form and providing Captricity with a way to access completed copies of paper forms. Governments can provide that access, he explains, by either uploading scanned images, routing their mail to a PO box (where Captricity scans them) or having end users send in pictures of their completed forms.

Companies like Seamless Docs and Captricity are imminently applicable because they address a known problem: government needs to digitize its current workflow. But these companies only scratch the surface of the potential provided by today's communication technologies. In order to see conceive of new futures, we have to be brave enough to ask open-ended questions.

All together, now

The old ways we were taught to interact with government—by voting, writing letters, calling our representatives—aren’t enough anymore. It’s time to break the mold altogether, to embrace a new, technologically aided peer-to-peer way of solving our problems.

Gavin Newsom, Citizenville

I have a confession to make—one that I hope hasn’t been too apparent up until now. When I began drafting what would eventually become this article, I didn’t quite know where I wanted to end up. I knew that I wanted to write about forms because, as far as I could tell, nearly every Code for America project utilized them. More selfishly, I wanted to write about forms because “form design” gave me a pragmatic lens through which to continue my study of digital literacy. In the end, however, I have written this article in an attempt to look past forms to the ways in which user-centered design aids in the practices of government and citizen sensemaking. By breaking form design down into “protocol, process, and platform” design14, I hope to have shown something about how these elements relate to one another that might inform future design endeavors.

In his doctoral thesis “Making Visible,” designer Timo Arnall argues that designers cannot competently work with a medium until they thoroughly understand that medium's physical properties. Industrial designers, for example, will test a material's bendiness, its strength, its resistance to heat, scratching, etc. As far as I can tell, there are no equivalent tests for civic technology. In going about our work, civic technologists must borrow from user-centered design in order to determine the degree to which our work has utility and cultural resonance (i.e. “popularity”). If design is a rendering of intent, then user-centered design is a way of determining whether or not those intentions are coherently rendered.

Are forms really, then, the best way to help people manifest their understanding or share their perspective, one directional as they are?

Numerous people, from both the academie and the public sector, have acknowledged the curious ways in which millennials approach civic participation. A paper authored by Neta Kligler-Vilenchik and Sangita Shresthova describes this more formally as “participatory culture civics.” From their paper: [These are] organizations rooted within 'participatory cultures:' cultures which share a strong sense of community, have relatively low barriers to participation, informal mentorship structures and support for creating and sharing one’s creations (Jenkins et. al., 2006). Participatory civic cultures are exemplified by organizations like Code for America as well as designers like Molly, Jake and Dave. Taken together, these actors are indicative of a larger movement in which citizens celebrate the cultural change capacities of digital, media and design literacy.

This is something that the future of government forms hangs upon. In a white paper released earlier this year, Dana Chisnell tells a story in which Washington State Ballot Designers invited representatives from companies that manufactured voting machines to observe usability tests of existing ballot designs:

After some negotiating, representatives from vendors accompanied a staff designer on a tour of counties to conduct usability test sessions on existing ballot designs. From this evidence, the vendors implemented support for better ballots. The importance of design literacy in the elections division in Washington is not to be missed: the project took persistence, planning, and commitment from both sides.

Dana Chisnell

In this case, Washington ballot designers leveraged not only their design literacy but also their position as potential customers to become de facto change agents in relation to their vendors—the opposite of the normal relationship between government and its vendors.

The way to ultimately improve the design of government forms, then, is to both invite participation and take part in design-mediated conversations. Examples of this abound. Dana Chisnell’s recently launched Center for Civic Design applies design thinking not only to ballots themselves but to all aspects of the voting experience—essentially improving the entire ecosystem in which ballot design is embedded. Moreover, Chisnell is not content with keeping her knowledge to herself. Chisnell’s “field guides” encourage participation by distilling the salient, actionable elements from her research and fieldwork.

Sharing the best practices of form design and promulgating design literacy should not be the remit of any one entity—in fact, it can’t be. The only way for citizens to participate in the design conversations necessary to responsibly govern a 21st century democracy is to acknowledge that these conversations are happening; join in; and share their findings. Through the creation of participatory civic cultures centered around design literacy, civic hackers and government employees alike can work to improve the citizen experience of forms once and, quite literally, for all.