When It Comes to 311, the Customer Isn’t Always Right

As cities explore ways to use citizen complaints to enhance public services, research shows there are drawbacks to such data.
by | January 2016
An employee taking 311 calls in Washington, D.C. (Chuck Kennedy/MCT)

When Somerville, Mass., started waging war on an escalating rat population a few years ago, the city turned to data from its 311 nonemergency call center to devise a plan. Reports of rat sightings logged by callers to 311 showed where in the city the problem was most severe. Crews baited traps in those places, while every city property owner received new heavy-duty trash barrels.

Calls reporting rats in Somerville have since dropped by more than 60 percent. And the data has been useful in confronting a number of other issues. Every day, Mayor Joseph Curtatone checks a 311 dashboard highlighting trends and any anomalies across a range of city services. “It’s kind of like our pulse on the city,” says Daniel Hadley, the mayor’s chief of staff.

Like Somerville, many localities are finding novel ways to utilize 311 call data. At the same time, though, a growing body of research has highlighted drawbacks to the practice, most notably that not all residents use the system equally.

Hadley says problems can arise if the city relies too heavily on the data without taking into account the way various demographic groups use 311. Neighborhoods with more immigrants tend to call into the system at lower rates, for instance, so that’s taken into account when rat complaints are received from those areas.

Within any given city, some residents are comfortable knocking on their neighbors’ doors to resolve issues. In other neighborhoods, people are more likely to rely on public agencies to step in. In one recent study, New York University’s Joscha Legewie and Merlin Schaeffer of the University of Cologne sought to assess neighborhood tensions across New York by analyzing 311 complaints for loud noise, drinking or blocked driveways. The system was used most often, they found, where “fuzzy boundaries” separated ethnically homogeneous neighborhoods.

The volume of 311 calls doesn’t always reflect actual levels of need for a service, either. This point is illustrated in recent research by Columbia University’s Jonathan Auerbach and economist Christopher Eshleman. They reviewed calls to 311 for tree damage following six major storms in New York City and compared them to completed work orders and tree counts. Neighborhoods with more renter-occupied homes, buildings with 10 or more units and unmarried heads of household were less likely to report damaged trees.

What’s more, a city may receive multiple calls reporting the same issue. In smaller cities, frequent callers may skew the data in the same way that a few outspoken community members dominate local council meetings. Media coverage around an issue also prompts more calls.

All these factors must be considered when localities incorporate 311 calls into wide-scale policy decisions. “We want to make the distinction between data-based policy and evidence-based policy,” Auerbach says. Andrew Nicklin, open data director at the Johns Hopkins University Center for Government Excellence, expressed concerns that relying solely on 311 data could exacerbate inequalities if affluent households are overrepresented. “It’s important to understand what the gaps are,” he says, “and where you can use other data to shore up biases.”

 As an illustration, the New York City buildings department receives tens of thousands of complaints each year citing illegal conversion of rental units to unauthorized uses. By incorporating data from other city agencies, such as which landlords are delinquent in paying property taxes, the city better prioritizes the calls that need investigation. Chicago’s health department forecasts which restaurants are likely to have health violations by comparing 311 complaints with variables that include prior critical violations, how long the establishment has operated and nearby sanitation complaints.

One way for cities to reach a broad cross section of residents is to offer multiple lines of communication. Audrey Mathis, Chicago’s 311 director, says neighborhoods with more seniors may prefer calling an operator. Others can contact Chicago 311 via email, Web forms or text messaging. “We want to make sure that, across the board, people have a variety of ways to reach out to us that aren’t limited to a particular neighborhood or demographic,” Mathis says.

Many cities have just begun to explore the ways 311 data can enhance public services. Pittsburgh launched its 311 system nearly a decade ago, but it wasn’t until lately that the city began collecting and analyzing call data. The effort has already proven useful in identifying bottlenecks and ways to improve efficiencies, such as how to best deploy crews for its pre-winter pothole-filling blitz. Some departments cite 311 call data when they put together budget proposals for capital improvements, and it may soon be used to help measure performance.

Just like any data, 311 calls aren’t a perfect indicator of what’s happening on the ground. When the limitations are accounted for, though, the data can serve as an invaluable tool in policymaking.  “It’s almost like there’s not really an end game,” says Laura Meixell, a manager in Pittsburgh’s Department of Innovation & Performance. “The possibilities are endless as we expand and refine the process.”