Always Under Construction

This article is part of Update or Die, a series from Future Tense about how businesses and other organizations keep up with technological change—and the cost of falling behind.

There was a time in New Orleans where you couldn’t get anywhere because of road closures. That time was yesterday. And will be tomorrow. And possibly forever. Or so it can feel to road-closure-weary residents. I say this as someone who calls New Orleans home, even when work keeps me on other cities’ roads.

Of course, those streets are bad too: According to a report from the American Society of Civil Engineers, Americans spent 6.9 billion hours delayed in traffic, wasting 3.1 billion gallons of fuel, to the tune of $160 billion lost, in 2014 alone. That’s before the costs of driving on roads that need repair, which hit an estimated $120.5 billion total in extra vehicle repairs and operating costs in 2015. And that is before you get into the time, aggravation, and personal insult people feel about living on streets chawed into states of destruction.

But New Orleans exhales this aggravation with every humid breath. Residents live in a bowl built on a swamp built below sea level. Our streets erupt faster than a teenager’s acne. We have potholes, craters, chasms, mudholes, manholes, sinkholes, and upchucked street guts. We lose semis in sinkholes. We find deer in potholes. We protest with infrastructure rallies and practice radical acceptance with pothole parties. In 2016, a report analyzing the quality of every street in New Orleans found two-thirds of them got a D or an F. Not one neighborhood got an A or a B.

Horrendous roads cause one set of problems. Horrendous road closures to fix said roads cause another. In the decade after Katrina, the federal government sent money to New Orleans to repair roads. That was good, except that together, planned road repairs plus unplanned closures made it impossible to track where you could get on any given day. It was like trying to cross town during Mardi Gras, minus the balls and beads. Oh, also minus the parade apps we use to track where floats are, so we know when, where, and how to navigate things like making groceries and getting to work.

So city technologists made a tracker of another kind. RoadWork NOLA launched in March 2015 as a map of planned road construction, updated every month, with basic layers you could turn on and off. In the city’s launch announcement, then-Mayor Mitch Landrieu noted, “This new website is another tool for our residents to be better informed of roadwork in their neighborhoods and plan ahead to avoid construction-related traffic disruptions when getting around town.”

On the surface, the website hit every civic-tech sweet spot. It partnered multiple departments. It improved back-end infrastructure. It offered a user-friendly front end. It used open data. A resident could see all roads, labeled with planned closures, and when. It seemed like a win. Minus one thing. “Almost nobody cared,” said Lamar Gardere, former chief information officer of the City of New Orleans Office of Information Technology and Innovation, or ITI. Residents didn’t actually use it.

I’ve worked in local and federal government, including in ITI. I also spent the past year interviewing government workers across the country and researching how government embraces innovation and service-delivery improvement. A ubiquitous point of failure: government upgrades in silos. Places often have overlapping problems (reporting issues, fixing roads, applying for permits). They often come up with overlapping solutions (apps, maps, one-stop shops). But individual cities, counties, and even agencies and departments often work in silos, making government upgrades all the more challenging.

We see the same thing happen again and again: The government announces an upgrade with a press release—because government loves itself a press release—and then there’s no follow-up. The upgrade is buried before it’s had a chance to live. Which is probably why you don’t know to use the USDA Ask Karen app on Thanksgiving.

Another problem we often see: In a hallowed room somewhere, a leader comments, Hey, why don’t we have one of those [insert tech solution]? By the time the conversation reaches wherever we’re storing government tech folk, the game of Telephone has garbled the comment into Governor mandate! Drop everything you’re doing, we need [insert tech solution]. Orders from the sky create huge pressure for civil servants not to say “No.” It also encourages government’s proclivities for chasing squirrels. Too often, this translates to solutions launched in search of problems, rather than identifying problems and crafting upgraded solutions to solve them. It also encourages a big, splashy debut instead of encouraging governments to launch in beta and iterate.

New Orleans actually avoided some of these pitfalls. It launched with a basic address search function, plus limited detail about work status and types of repairs. It knew it wanted to iterate and focused on how people could navigate routes around town. But the few residents who tested RoadWork realized it wasn’t useful for driving. RoadWork only showed planned closures. New Orleans, the city of sprouting sinkholes and unpredictable water breaks, faces brisk amounts of unplanned road closures. This led to two big questions: Could it help residents track ways to get around road closures in real time? And was that the actual value of the site for residents?

The answers turned out to be “No” and “No.”

The first question required an upgrade that appeared to be technical. Workers fixing temporary closures needed to send data about where to display them on the map. ITI prototyped an app workers could use to mark when they closed or reopened streets. “We called it the Easy Button Project,” said Gardere. The department wanted to make it as user-friendly as possible. “You hit a single button … and that means ‘I’m closing the street.’ “

That kind of collecting and sharing data sounds easy. In government, it is not. (Let the record show: Few things in government are easy.) In the case of RoadWork, the challenge was more than technical. It required upgraded technology capacity and enforcement. There wasn’t enough money, and there were no policies that mandated data collection.

Disconnects between technologists and policymakers in government may sound drier than a sack of desiccated cat kibble. In reality, they are a massive and underdiscussed roadblock thwarting governments from delivering critical services across the country. The technology challenge: Individual legacy systems rarely talk to one another. That means agencies can’t transfer data between the systems that are part of the same process. For example, the system that tracks someone who is found guilty in court is different than the system that tracks their transportation from court to jail, which is different than the system that tracks them in jail, which is different than the system that tracks them when they’re released to a halfway house. The solution to track one person over multiple agencies and systems? It involves a lot of faxing. Ditto, cyclical printing and retyping of information in static PDFs.

The contracts and requirements challenge: When government contracts for a service like road repair, it often fails to consider data collection. If it’s not in the contract, vendors have zero obligations to do it. The cover-your-rear challenge: Policies that do not give civil servants clear permissions to share data often mean they won’t.

New Orleans could have given up. But it knew successful upgrades required project iteration, not abandonment. In July 2016, then-Mayor Landrieu announced a hard-fought, finalized settlement with FEMA connected to Hurricane Katrina: an additional $1.2 billion, bringing total federal dollars he negotiated for road, water, and sewage repairs to $2 billion. With so many more road repairs coming, in October 2016, the city launched the next version of RoadWork, continuing to hopscotch over common government pain points before and after projects started.

The city’s Special Projects & Strategic Engagement Office tackled spreading the word and gathering information from residents with digital, social media, and in-person events, as well as the press releases. “We knew we needed to have a multifaceted comms strategy,” said Sarah McLaughlin Porteous, director of the Special Projects & Strategic Engagement Office. “We’re not relying on one platform.”

The city incorporated human-centered design, an upgrade often more critical to government projects than any single technology product. By getting feedback from residents, it learned people weren’t using the site to map routes. People wanted to check whether their streets would get shut down and understand how the city allocated dollars. What companies would get contracts? Were they local? Was my street going to be repaved? Why would it take so long? Why wasn’t my street chosen?

The city shifted to a communications focus. It created new RoadWork pages to address questions. This included a FAQ and information about vendors. The site’s usage grew. Between October 2016 and early June 2017, RoadWork received 113,935 page views, and 60,714 uses of the application where people can enter an address to receive construction updates. (Unfortunately, there’s no publicly available data about how often the first version of the site was used.)

Government upgrades often don’t invite trust. Government shows up slowly. Sometimes it doesn’t show up at all. It feels designed to amplify aggravation. Using technology can sometimes to reduce that impression and scale solutions. But it’s a double-edged sword. When tech upgrades fail, trust between government and residents breaks in equally spectacular, scaled ways.

Sometimes, these failures generate outstanding costs. Oregon spent $240 million on Cover Oregon, its online health care system, a spectacular failure that resulted in state government making residents fill out paper applications. Often, these “solutions” don’t solve problems. Frequent offenders include apps no one uses and civic-tech hack-athons where people with good intentions lock themselves in a room to build a product in a day that doesn’t exist a year (or a week) later. Too often, these failures are dangerous. France’s alert system, SAIP, failed during the Nice attacks in 2016. Earlier this year, South Australia’s Alert SA disaster app crashed while citizens at the height of bushfire watched. And no need to remind ourselves about Hawaii’s false missile alert this year.

Recently, I arrived in New Orleans, bumped into neighbors, and joined them for some porch sitting. In the week I had been gone, a new sinkhole emerged on our street. As darkness drifted down around our conversation, cars zipping down the street hit the crater and bit it over and over, ka-THUNK, ka-THUNK. I commented that I was writing a piece about RoadWork, the city’s planned construction map to fix streets. One neighbor responded with a rueful mix of chagrin and excitement, “Oh, yes, I saw that map. Our street is getting repaved in October 2019—and it’s going to last a year! Can you believe it?”

And so goes upgrade pain of another variety: from the chagrin of pothole creation to the consternation of repaving implementation. But unlike three years ago, she knows what’s coming—one sign of a successful upgrade.

New Orleans could have stopped when the site couldn’t support real-time navigation. It could have accepted permanent loss of trust that RoadWork could be useful. It did neither. It tested fundamental tenets of government upgrades: They are iterative, they are user-centered, they are ongoing, and their results are complex. Upgrade challenges and opportunities both remain. But like the residents RoadWork serves, the city continues to navigate the potholes as it drives toward the next destination. 

This article originally appeared in Future Tense, a collaboration among Arizona State University, New America, and Slate.