Improving Public Records in Boston: Researching, Building, and Buying Solutions for Real People

What if we could make it easier for the public to request information from the City of Boston, and for the City to respond to those requests? What would that process look like, and how would it better serve the needs of outside requesters and City staff? These are the questions our team set out to answer over the spring 2018 semester as part of the Tech and Innovation in Government Field Course at the Harvard Kennedy School.

Back in January, we paired up with the amazing team from the City of Boston to rethink public records—essentially any document made or received by a government employee. Public records—like the City’s bid to attract Amazon’s second HQ—matter because they promote government accountability and trust in government.

While we didn’t start out as design professionals, our teaching team pointed us in the right direction to jump!

We’ve done a lot since then. We helped craft a Request for Proposal (RFP) with the City of Boston that lays out the goals and requirements for a public records solution that will meet both the needs of the public and City employees. Pending Boston City Council approval of the latest proposed City budget, we are proud to say that the RFP will soon go up on the supplier portal.

Historical Context: Learning from the 18F Methods

Our approach was inspired by the work of 18F, an innovation unit within the federal government that scaled rapidly after the healthcare.gov crisis. Among their many tech transformation projects within government (including work around federal public records as well!), 18F is helping other federal agencies buy technology in new ways. Traditionally, governments buy goods and services through complex, multi-year contracts with a long list of requirements. This is something that may work well for buying a tank or a building, but doesn't work as well for technology-driven services or software, which change over time.

Instead, 18F recommends using modular contracting: “breaking up large, complex purchases into multiple, tightly-scoped projects” to introduce technology incrementally. In this framework, people who use the service—i.e., “users”—have to be the north star, rather than the stakeholders. In the case of public records, for example, users include both the public making requests and the civil servants who work to make records public. 

As we started working on our RFP, we wanted to learn from 18F’s work and, contribute to the growing knowledge base around modular procurement.

Our Work with the City of Boston

Currently, almost all public records requests for the City of Boston are made, processed, and fulfilled via email. 

Shawn Williams is the Director of Public Records. We interviewed him and 40 other prime users.

The City set out to make this process more efficient, and created the position of Director of Public Records, hiring Shawn Williams. The City then started to consider technology solutions that could support the people and processes dedicated to making information public.

The City wanted to make sure the new system would be easy to use for both the public and City staff. That required learning more about who makes public records requests and better understanding their experiences. What was the current process like? What was working well, and what wasn’t?

That’s where our student team came in. We interviewed more than 40 people who interact with the records request system in different ways to learn about the service from the public, City staff, peer cities, and experts in the field.

We focused our research on two groups: the public requester (like a journalist, community activist, or just concerned resident) seeking public records, and the custodian within the City providing them. You can imagine them as two sides of the same coin.

Requesters and Custodians are two sides of the same coin, and share parallel frustrations

After interviewing both Requesters and Custodians, we asked three fundamental questions:

(1) Does the City need a records request tracker?

Analyzing our interview insights, we found that almost 70% of the frustrations expressed by both City staff and the public could likely be addressed by a public records tracking tool—i.e. Internet based software that would track the status of a public records request.  So our conclusion was yes, the City of Boston does need a tracker!

(2) What should a records request tracker do?

In our interviews, we heard about four key challenges—routing, scoping, timing, and trust—which we narrowed into three criteria for a record request tracker:

●      Routing: Does it clarify who could best fulfill this request? How do we make sure the request gets to that person?

●      Specificity: Can it help requests be more specific so that City staff can effectively fulfill them and requesters can get what they need?

●      Communication: Can it help provide a clear status for the progress of the response, setting clear expectations and making it easy to recognize great service delivery by the City?

For each of these criteria, we developed a prototype that we put in front of people to test how they interacted with it, and understand what could be improved. In addition to the “Tippy” prototype that helped narrow requests to be more specific, we tested a Department Suggestor and a Status Tracker

We tested paper prototypes designed to address each tracker criterion

We put our paper prototypes in front of City staff and the public to observe, listen, and learn

We listened carefully to the feedback we received in this testing phase. Here is some of what we heard:

●      Scrolling through 72 departments and matching one to your request was dizzying. The Department Suggester prompted people toward a suggested department, making it easier for them to route their request properly. We learned that the City is already doing similar work on their enhanced 311 system, which provides prompts for a suggested case type.

●      Tippy, which provided timely tips about how to craft your request based on keywords, clarified broad requests. This type of functionality has the potential to significantly decrease the time it takes to fulfill a request.

●      The Status Tracker, modelled after the Domino's pizza tracker, provided the public with updates on the progress of their requests. This taught us that requesters like to know where their request has been and where it’s going.

(3) How does the City get a good records request tracker?

Enter procurement—the process through which the City buys goods and services.

We were lucky to have a tangible opportunity to bring our findings to life: the City had decided to issue a competitive RFP to buy a public records request tracker.

Our goal was to translate what we heard in our research and testing into specific evaluation criteria on the RFP. We hope our input will help the City find a vendor that will offer a system that truly serves user needs. We also provided the City of Boston with a number of recommendations about how they could include a human-centered approach within their procurement processes.

Based on what we learned in testing of “Tippy,” we added the pink text shown here to the RFP

Conclusions

Our work led us to three key conclusions:

●      A software system to track requests will be a huge win for both sides of the proverbial “coin”—that is, both requesters and custodians.

●      A best-in-class public records tracking system is entirely within reach for the City of Boston.

●      Integrating user testing into procurement will help the City make important procurement choices, for public records tracking and also for other software needs.

We look forward to seeing how the City of Boston, building on our insights and help, improves public records transparency!

Jackie Chea, Thad Kerosky, Jim Moffet, Erica Pincus, Jon Truong

Simplifying Public Records Requests

Twenty-nine interviews. 48 pages of notes. Hundreds of hours. Our Harvard student team working with the City of Boston spent the past two months talking to City staff, activists, journalists, and other Bostonians to understand their experiences finding information from the City.

Image of our problem statement “mad-lib”

How did all of this research lead us to playing mad-libs on a Saturday evening?

Our team assembled this past weekend to turn the stories we’ve been collecting into problems that we can solve. Playing problem statement mad-libs helped us fill in the blanks and put the core challenges of requesting public information into words. Here’s one example of a problem statement we came up with:

“A custodian [City staff] feels their time is disrupted by clarifying new requests, and needs a clear idea of what to look for, but faces a broadly-worded ask.”

After writing a handful of these problem statements for the various types of people who look for public information, we grouped the statements together into four key challenges:

1. Routing: Who could best fulfill this request and how do we make sure it gets to them?
2. Scoping: How can requests be more specific so that City staff can effectively fulfill them and requesters get what they need?
3. Timing: How can the City set and provide a clear timeline for the progress of responding?
4. Accountability and Trust: How can we help requesters trust the process, maintain open lines of communication, and recognize when requests are fulfilled well?

By synthesizing the research we’ve done into these four areas, we were ready to move into the next phase—generating solutions through prototyping.

Finding the Puzzle Pieces

Prototyping is a process of quickly developing ideas and trying them out. The goal is to go through many alternatives—both good and bad—find a few that might work, and then build them out further.

This was not a time to get too attached to any single idea.

Initially, we fell into this trap, thinking that the challenges we identified seemed to suggest only one obvious solution: a tracking system that would help everyone keep tabs on the status of requests. It would address the issues with accountability, offer updates to requesters, and ensure that anyone in the public could benefit from another person’s request. However, as we generated prototypes and researched commercial tracking systems, we discovered that this one solution appeared to be filled.

Rather than attempt to recreate what already existed, the team decided that we could offer value to the City of Boston by building upon and filling in the gaps of commercial products and translating our prototypes to support the City’s procurement process. We developed these tools to gain insights to better allow us to inform the City’s request for proposals and make recommendations about the short, middle, and long-term trajectory of the public records process.

As a team, we generated ten prototypes, and settled on four we want to test—including the idea below, that we called “Tippy.” 

Prototype: Tippy

Overly-broad public record requests can be disruptive to City officials, and often result in long wait times for requesters. Sometimes a broad request requires a lot of back-and-forth between requesters and City staff, to figure out what the requester needs. Other times, because the requester isn’t able see what the city actually has, they can’t be more specific, and the request takes a long time to fulfill.

Enter Tippy, our version of a smart suggestion tool that helps people narrow down their public record requests with more specificity. As a person types their request, helpful tips appear based on key “trigger” words. For example, when someone types “email,” tips pop up to remind the requester to add a date range or an email address they’re interested in. That way, requesters will be more likely to get what they need while city staff can more easily find the records in question

Mockup of Tippy interface when the word “email” is used.

Alternatively, when a word such as “aggregated” is written in the request box, tips that pertain to machine-readable datasets could appear to guide the individual towards a clearer request.

Mockup of Tippy interface with the word “aggregated”.

To help users find the right department for the request, the dropdown menu could make smart suggestions based on the types of words entered in the box. This helps requesters avoid having to look through a list of more than seventy departments to find the one best suited for their inquiry.

Mockup of Tippy suggesting the best department to respond to a request.

Testing Tippy

As we prepare to test our prototype with the public, we are asking ourselves the following:

• When a user interacts with this prototype, do they make requests clearer and easier to route to the right city official?
• Should we show tips before a user starts typing? If so, which ones?
• Which words should trigger which tips?

Although we have shifted in our process from collecting user stories to designing and testing prototypes, we maintain our focus on the primary challenge--keeping the experiences of the public at the heart and center of our design work.

While we generated many ideas, many of them ultimately ended up on the cutting room floor. Here are some examples of additional prototype ideas we had:

• Shawnbot 3000, the Artificially Intelligent public records custodian
• Simple CC tracker allows you to CC a specific email to track requests in a central location
• Email Review Time Estimator can magically tell requesters how long it will take to review
• Department Suggestor takes the keywords of your request and spits out recommendations of who might have the information
• City Data Organizational Chart that’s engaging and easy to navigate
• 10 Days Later automatically releases unflagged documents 10 days after their creation
• Collaborative Forum (think StackOverflow) for public records that could build a community around requesting and analyzing information from the city
• Domino's Pizza Tracker for the public records delivery process
• Data Map for City information instead of City streets that tells you which department has what information 

Jackie Chea, Thad Kerosky, Jim Moffet, Erica Pincus, Jon Truong

Why Government Transparency Matters

Screenshot of the website the City of Boston published with public information on their proposal to host Amazon’s second headquarters

In his first inaugural address in 2014, when Boston Mayor Marty Walsh laid out four pillars for his new administration, the final point was a pledge to “increase trust and transparency in city government.” Recent events surrounding the bidding process for Amazon’s second headquarters have brought this to life, suggesting that the City is capable of setting national standards for transparency and open government.

The City proactively created a website, making all of the information about its response to the Amazon headquarters Request for Proposal public. This response to public requests for information about the bid stands in stark contrast to the enthusiastically-redacted documents that investigative journalists at MuckRock received about the incentives offered to Amazon by Montgomery County, Maryland (pictured below).

Documents released to MuckRock by Montgomery County, Maryland

Our team of five Harvard Kennedy School researchers is taking amazon.boston.gov as a model for what the City of Boston can do when it puts its policy, economic development, press and technology teams to work making government decisions understandable and accessible. We’ve spent two months interviewing people on both sides of the public records process in order to figure out how Boston can successfully live up to its own example across more than 70 City departments and agencies.

Learning from City Officials and the Public

To learn more about the process, we talked to experts nationwide. “The process of responding to public records requests is basically a black box to most people,” said Leah Bannon, formerly of 18F, a user-centric digital services firm inside the federal government. This comment echoes what we’ve heard from journalists, activists, researchers, and constituents nationwide. Agencies in charge of records requests often fail to communicate the reasons that certain requests require additional time, leaving requesters in the dark.

“Maybe I didn’t structure my request very well and that’s why it’s taking so long?” said one journalist we interviewed. On top of that, requesters often aren’t given much information about the progress of their request. “Every so often, someone will write back to say they’re still looking,” the journalist added, referring to a particular request sent to the federal government agency.

Many City of Boston departments have responded quickly to public requests. William Wang, a journalist for The Harvard Crimson, requested an incident report from the Boston Police Department and received it in just over an hour. Is this example an anomaly or a frequent occurrence?  Unfortunately, until response times to public records requests are more thoroughly tracked and measured, we simply cannot know.

Mapping the Public Records Request Process

“You can’t have accountability without measurement,” says Michael Morisy, co-founder of Muckrock and a public records expert. It’s not just requesters: City staff that act as custodians of records also suffer from an opaque and unmeasured process. Our research indicates that both requesters and City staff  struggle with less common requests. Requesters are unaware of what records exist, how they’re stored or how to ask for them. Custodians are unaware of the precise intent of the request and may be limited by statute from asking certain questions in an attempt to gain clarity.

Both requesters and custodians are trying to peer through the walls of a black box from opposite sides. Mr. Morisy, who often finds himself in the middle of disputes between requesters and custodians, expressed sympathy for the custodians, stating, “it’s easy for agencies to get tired of journalists requesting 10,000 pages and only using one sentence.”

A public records request can take a wide variety of routes, and make a large number of stops along the way. In particular, the City of Boston has more than 70 departments and 4 autonomous agencies that all maintain records subject to public request.

The beginning of a public records search is familiar enough. “Almost all public records requests start with searching for the information in Google,” said Ms. Bannon, sharing a finding from 18F research for the recent foia.gov overhaul at the federal level. However, once your search engine of choice delivers you to an individual agency or government records portal, all bets are off. Current processes are highly decentralized and variable. Although helpful resources such as boston.gov/public-records are available, requesters often find a public email address without further guidance and begin the awkward process of writing a specific request for something that may or may not exist, or that might even be publicly available if they only knew where to look.

Requesters bring expectations for high-quality services based on their favorite commercial/private-sector products, and while these can be difficult standards to meet, the differences can help clarify points of frustration for both requesters and custodians. “It would be nice to have a portal where you track your request, see where you are in the process, like we have for packages,” stated one interviewee.

Many commercially available tools try to replace paper, email, and Excel to provide that type of transparent tracking. Boston has allocated resources for an IT investment in this area, although even low-cost tools may be out of reach for smaller jurisdictions. However, both Mr. Morisy and William Isaac, a data scientist for the Human Right Data Analytics Group, were even more excited about online communities that curate knowledge, such as Wikipedia, Genius and Stack Overflow. According to Mr. Morisy, “Most of the current crop of public records management software is far better than the legacy systems they’re replacing and are doing us all a great service, but to some extent they’re just porting processes developed in the 70’s to the web.”

In other words, our team has discovered the importance of updating not just the technology used in this process, but our understanding of how knowledge can be created and maintained. Our research has given us insight into the challenge of navigating a complex information system without a map or the ability to learn from others’ experiences. It is a challenge that digital tools are only just now coming online to tackle.

Here in Boston, we’re excited to continue to work with the City of Boston to figure out what technology and tools might help the City employee and public requester alike.

Erica, Jackie, Jim, Jon, and Thad

Teaming Up for Transparency

The Public Records homepage for the City of Boston (Source: boston.gov)

Have you ever had a question about how a decision was made by your city or town? In Massachusetts, by law, every person has a right to access the government’s public records.

In today’s climate of activism and citizen engagement, there is a strong need for government transparency and accountability. Many groups submit public records requests in order to engage directly with government and learn more about its work on issues that matter most to them.

"We value transparency as a way to keep constituents involved and engaged," said Jeanethe Falvey, Chief Digital Officer for the City of Boston. "We're committed to making the Public Records Request process as seamless as we can."

The federal equivalent of the Public Records Law—the Freedom of Information Act (FOIA)—has been lauded as “a vital tool for keeping government open and honest.” One effect of these laws has been to get government agencies in the mindset of posting their information publicly, from the Food and Drug Administration helping the public avoid contamination by distributing information about a listeria outbreak in ice cream, to the Department of Energy accelerating innovation by sharing details on recent scientific discoveries.

Despite such benefits, many people do not know how to file a public records request with their own city, or find the existing process to be daunting. How might the City of Boston reimagine the public records request experience for advocates, journalists, and the public to advance good governance? And how can the City proactively surface the most relevant information, making it easy to access without a formal request process?

Our Team

Our first meeting with the City of Boston team

Enter our team. We are students from the Harvard Kennedy School Tech and Innovation in Government Field Class, taught by Adjunct Lecturer and former U.S. Deputy Chief Technology Officer, Nick Sinai. Our diverse backgrounds provide us with an interdisciplinary approach to this challenge:

• Jackie Chea is currently a junior at Harvard College studying applied math and biology, and will be a Program Manager Intern for Microsoft’s Windows team this summer. She is excited to devote her efforts to the realm of civic technology.

• Thad Kerosky is a Master’s student at the Fletcher School at Tufts University, coming into public policy from a background as a full-stack developer. He has worked on new sustainable energy, agriculture, open data and education social enterprise projects from East and West Africa, D.C. and Cambridge over the past decade. Thad has been a Code for Boston enthusiast since 2013 and is now part of its core team.

• Jim Moffet is a Master’s student at the Harvard Graduate School of Design who is passionate about discovering the strengths and weaknesses of the use of technology as a lever for social change. After five years as a product manager for SEIU, he spent a Fulbright fellowship in Budapest teaching design thinking at public fablabs and co-founded the Cambridge civic tech start-ups Commit.vote and Outvote.io.

• Erica Pincus is passionate about increasing access to opportunity and solving persistent challenges in new and innovative ways. She previously served as a Policy Advisor and Special Assistant for Social Innovation for the Obama White House, where she supported the development of data-driven and technology-enabled policy solutions. Erica is currently pursuing a dual MPA/MBA degree between Harvard’s Kennedy School of Government and Stanford’s Graduate School of Business.

• Jon Truong became interested in using digital tools to advance positive social change in urban settings through his work as a housing and tenant rights researcher and advocate in Los Angeles. He wants to explore opportunities for collaboration between the private, public, and non-profit sector to increase civic engagement, elevate public discourse, and address inequality in historically marginalized communities. He is currently pursuing a Master of Public Policy at the Harvard Kennedy School of Government.

We are excited to be working with Laura Melle (Senior Procurement Lead, DoIT), Shawn Williams (Director of Public Records), and Reilly Zlab (Product Management, DoIT) from the City of Boston to reimagine the public records request experience.

Requesting a Record: How Things are Done Now

Massachusetts law defines public records to include all documentary materials or data made or received by any officer or employee of any Massachusetts governmental entity, with a few narrow exemptions. Police records, rally permits, construction updates, and city events calendars are all considered public records.

Some city information is already readily available to the public, such as the information housed on Analyze Boston, the city’s open data portal. But for other information, how would a person know which of the city’s 72 Departments to contact?

The homepage for the City of Boston’s open data portal (Source: data.boston.gov)

Currently, to formally request a record, one would contact the Records Access Officer (RAO) with a description of the desired information. The RAO, in compliance with the Massachusetts Public Records Law, then has ten business days to respond to the request--either providing the requested record, providing a fee estimate (where applicable), or denying access if the record meets one of the narrow exemptions.

Sample public records request (Source: www.sec.state.ma.us)

The Challenge

The City’s Director of Public Records is working to ensure compliance with applicable laws and regulations, while improving the experience for both requesters and records custodians. In support of that mission, our team’s goals are to:

1. Make public records easier for the public to find and use

2. Make it easier for the City of Boston to provide individuals seeking information with the relevant public records

Next Steps

Our approach to this project will be guided by the technology sector's best practices, including designing with rather than for records custodians and requesters via user-centered design.

As we enter our user research phase, we will be seeking out different types of users, including, but not limited to, researchers, activists, civil society organizations, individuals making requests out of personal interest, and staff of various City departments. We wish to find out what information they seek, and whether that information is readily available. We are looking forward to listening to their stories and hearing more about how Boston’s public records request process can better serve their needs.

If you want to learn more or have thoughts on our project, please contact us at innovategov.bostonrecords@gmail.com. Stay tuned!

Erica, Jackie, Jim, Jon, and Thad