Last month, I ran into a friend at a government conference. “I am so tired of innovation,” he said to me. “You know? The word. Innovation. Ugh. What does it even mean any more?”
“Something shiny,” I replied, sighing.
“Exactly what successful ‘government innovation’ is not,” he replied, heaving a sigh of his own.
Government loves innovation, the shinier the better. It gets to feel modern (“We have an app for that!”). It generates headlines (“[City name] will use blockchain to solve food deserts!”). And it feels like a real solution. (“Why yes, Amazon Prime, I would like that 40% off-deluxe-espresso-drip-coffee urn crafted by empowered Italian grandmothers by 10 a.m., thank you.”)
Anchorage, Alaska, experienced this firsthand last year. It tried to solve a problem by asking a simple question: “Can we help more residents avoid going to third-party collections by changing the final notice letter we send out about delinquent debt?” Nothing it did will sound revolutionary. It brought together people from different agencies to brainstorm. It came up with a series of design changes in standard government letters. It tested them and measured the impact.
How did Anchorage do it? Small unshiny, changes.
Oh, and reams of pink paper.
For years, Anchorage followed the same process to collect delinquent debt. Each month, the Department of Delinquent Criminal Civil Fines and Fees (DCE) received a list of around a thousand people whose unpaid debts were heading to collections. DCE would send each person a final notice letter. If Treasury doesn’t receive a payment within 10 days, the letter warned, your debt will go to a third-party collections agency. If it goes to collections, you’ll owe extra money.
The letter wasn’t working. Treasury had amassed $49 million in unpaid delinquent debt, dating as far back as the 1980s. Sometimes a better address would turn up and it would send a reminder letter. But it didn’t mail reminders systematically. And it didn’t track their impact. It relied on the final notice letter as the one consistent touchpoint for residents.
One small problem. The letter hadn’t changed in 20 years.
In early 2015, now Mayor Ethan Berkowitz was so committed to tackling challenges and empowering civil servants that he reached out to Bloomberg’s What Works Cities to learn more about the program—before he became mayor. Brendan Babb, then a member of his local Code for America brigade, and now Anchorage’s chief innovation officer, helped write the successful application.
In April 2017, the city held a one-day workshop with agencies from across the ticketing process, from Department of Health and Human Services, to Parks and Recreation, GIS, Solid Waste Services, and the Mayor’s Office. The challenge: How could Anchorage take core ideas from behavioral science to create a letter that was easy, attractive, social, and timely (EAST)? This was not about creating a new process. It was about testing improvements to an existing one. People got these letters. What would encourage them to open the letter? Read the letter? Understand the letter? “We assume folks will read the entirety of letters that we send them,” Kalin, the behavioral scientist, noted. “People don’t look at a letter as the great American novel. You have five seconds to make an impact when you send out the letter. How can we communicate its essence in five seconds? Make it memorable five minutes later?”
In the morning, city staff learned behavioral science basics from Kalin. In the afternoon, they did research on the collections process. They talked to Treasury staffers. They hopped on the phone with the third-party collections agency. And they studied the old letter, identifying all the problems someone might have when reading it. Then teams competed to redesign it, presenting their solutions at the end of the day to a panel of “celebrity judges,” city managers, and leadership.
“What happened was cool changes from all three groups. We ended up combining all the changes together into a new letter,” said Babb. “Some changes were on the outside of the envelope, adding text that said, ‘You really need to open this’ in red ink. They changed the letter to pink paper, and added the due date in a box in the upper right-hand corner that said, ‘This is due June 30.’” It also explained what the citation was for and what was owed (information the old letter didn’t include . . . because, government) and made info about payment plans more conspicuous.
Next came testing. Treasury sent 10,000 old letters to one group as a control, and the new design to 10,000 others. When the results came in, it made some surprising discoveries. Overall amounts of revenue increased, but immediate revenue decreased slightly, from 3.4% to 2.5%. However, payment plan enrollment almost doubled, from 8.2% in the new letter compared to 4.5% in the old letter. So what did this mean? For starters, a new set of questions Treasury could choose to explore, like why they had a payment plan. Was it important to residents?
That started all kinds of changes. They put all their urgent notice letters on pink paper. They took up the challenge issued by the flurry of returned letters from unknown or wrong addresses. Treasury had access to a database that residents kept up-to-date because government used it to mail checks. Treasury began checking more addresses through that database. They used the insights from the workshop to test ideas with other people.
As small changes rolled out, bigger benefits started to roll in. The initial test returned $63,000, more than paying for the $55,000 cost of mailing out the test letters. By the end of 2017, Treasury would collect just under $1 million in new revenue from all the small changes they made. Equally compelling, the staff was energized. They continue to try and test new changes. Hundreds of residents cleared up their debt.
For government workers, making small, unshiny changes can be even harder than making big, bold ones. Leadership loves big changes. It can be easier to find money from taxpayers or from outside partnerships. Any change at all often counters what civil servants have been taught for decades: “Success is when we are in compliance and check all the boxes,” rather than encouraging people to ask, “Are the boxes on our list the right ones?”
First, it created space for civil servants to ask questions and to brainstorm across departments. If you’ve never worked in government, the accomplishment of “we got people who don’t talk to each other in a room to talk to each other” may sound like a recipe for a yawn or a jeer. But getting people who don’t, won’t, or (think they) can’t get in a room to talk to each other is possibly the single most under-recognized and powerful tool when improving how government works. It is also one of its greatest challenges.
Second, it offered a specific approach–behavioral science–for considering a problem in a new way. “There’s this perception that experimentation and innovation are really shiny and new, as opposed to simple but vigorous methods to evaluate what might work,” Kalin noted. “[Behavioral science] is all about incremental improvement . . . and no other options come close to the rigor of an experiment.”
Third, the solution wasn’t specialized, custom-built, expensive, or technical. It was replicable. Behavioral science and testing of small changes to improve services is something we can use across problems, agencies, and governments. As Kalin noted, “[Cities] have to collect money. To do that, they have to interact with citizens or send letters out. There’s often a gap in compliance. That gap is measurable. Cities are already keeping tabs on who owes what. This means we have all the numbers to see if we can increase compliance by even just a couple of percentage points.”
Fourth, it empowered local civil servants, the people who knew most about the process, and the community. After the results from the initial test, Treasury could have done another test with a control group and a small process change. Instead, a representative from Treasury noted: “We got so involved after that workshop that we decided to really evaluate what our whole portfolio looked like. How could we apply what we learned and make more changes?” Energized by evidence that change could be made without specialized tech skills or big shiny tools plus their leadership’s clear support to test doing things differently, employees began examining their work and processes differently. What else can we do better? How? “This became our baby,” the team reported. “We got excited to see what we could do, you know?”
Examples like Anchorage illustrate the importance of stepping away from the shiny, redefining what success looks like in government, and updating what celebrate when celebrate “innovation.” There are over 2.5 million civil servants working in government. When was the last time we heard about government work igniting joy? For the people doing it or for the communities served? Where do we find celebrations of small victories hard won? We must update the “how do we define success” algorithms we use as we push for upgrades to government services that are critical to everyday ability to survive, and to thrive.
That is how we move forward. Step by unshiny step.
With the occasional ream of pink paper.