Pew Study: Philadelphia’s 311 system shows progress after one year

(PHILADELPHIA)—A new study from The Pew Charitable Trusts’ Philadelphia Research Initiative finds that Philadelphia’s 311 contact system succeeded during its first year of operation in giving residents improved and easier access to information about city government. It did so at lower cost than other cities and won strong ratings from its customers.  But the system also mishandled thousands of service requests, and nearly one quarter of all service requests were not completed in the promised timeframe.

The study, A Work in Progress: Philadelphia’s 311 System After One Year, reviewed Philly311’s operations in 2009 in relation to 14 other communities that operate 311 systems.  Among those communities are most of the largest with such systems—New York, Los Angeles, Chicago, Miami-Dade County (FL), Houston, San Antonio, Dallas, Detroit, Charlotte-Mecklenburg (NC), San Francisco, Columbus (OH), Baltimore, Denver and Pittsburgh.

Almost all of the 311 centers in the study had reduced their budgets and staffs in the past year due to recession-driven city budget cuts.  Philly311’s own startup funding was reduced by at least 60 percent in 2008 and 2009, a cut that forced it to rely on a temporary computer system and delay hiring agents with outside call-center experience.

“The Philly311 contact center had a rough start but improved dramatically by the end of 2009, particularly in handling requests for information,” said Thomas Ginsberg, project manager of the Philadelphia Research Initiative and author of the report. “But the city struggled with service requests because it made only modest progress toward integrating the 311 center with the departments that actually do the tasks.”

Philadelphia Mayor Michael Nutter and Managing Director Camille Barnett have said that making the system better will require investments in technology at Philly311 and in other parts of city government. Nutter and Barnett have not said when or whether they would seek those investments. The Pew study found that achieving significant cost savings and efficiencies would require a major technological upgrade. Philly311 did not save the city money in its first year, nor was it expected to.

In a companion poll conducted for the Philadelphia Research Initiative, Philadelphians generally expressed satisfaction with Philly311, although only 28 percent of those polled could recall the number without prompting and only 15 percent had called it. Both results reflect the administration’s decision to do little or no promotion of Philly311 to save money and avoid overwhelming the fledgling system.

Among residents who called the number looking for information, 68 percent said they were satisfied and 29 percent dissatisfied. Among those calling with a specific complaint or request, 60 percent were satisfied and 33 percent dissatisfied.  And 53 percent of residents who were aware of the service said that Philly311 represents a “major step forward” for the city while 28 percent said it “won’t make much of a difference.” The poll was conducted in early January, after the first major snowstorm in the region but before the next two.

A Work in Progress also reviewed cities’ 311 performance across key measurement areas. For the year as a whole, out of 15 cities, Philly311 had the second highest percentage of calls abandoned before an agent answered (26 percent.) And out of 13 cities, Philly311 had the second longest average wait time to speak with a live agent (1 minute, 45 seconds.)  Much of that was due to start-up problems. As the year progressed, the experience for Philly311 callers improved dramatically. Callers spent less time waiting for agents and more time talking to them, with fewer calls being lost. While Philadelphia’s abandoned-call rate was one in four for the year as a whole, it dropped to just one in 17 by year’s end.

The contact center had several shining moments, providing fast and easy updates during spikes in calls in October during the World Series and a transit strike. The center also had recorded audio updates for the historic snowstorms in December 2009 and February 2010.

Between January and December 2009, about 1.1 million calls were made to Philly311, roughly the same number made a year earlier to the City Hall phone numbers it replaced. About seven in 10 calls were for general information. The others resulted in roughly 64,000 service requests, which Philly311 agents transmitted to the appropriate agencies through a patchwork of computer links and manual routines. In one major fumble, the system mishandled thousands of routine housing-inspection requests intended for the Department of Licenses and Inspections—requests that either were not submitted to the department or were submitted and not reported back as done.

One purpose of a 311 system is to reduce the load on the 911 emergency lines by giving residents with non-emergency requests one number to call. In its first year, however, Philly311 diverted fewer such 911 calls than city officials had hoped.

The Managing Director’s Office, which set up and runs Philly311, appeared to make the most of its limited funds in the startup year (2008) and first operating year (2009). It patched together low-cost software to emulate aspects of a more costly full-blown 311 system and assembled a large database with answers to hundreds of questions.  Philly311’s FY2010 operating budget of $2.8 million amounted to 0.08 percent of the city’s general fund expenditures, among the smallest proportions in the localities reviewed. The average 311 call cost Philadelphia taxpayers $2.20, compared with a median of $3.39 for the group.

About the Report

To prepare this report, Thomas Ginsberg, project manager of the Philadelphia Research Initiative, reviewed the history of 311 efforts in Philadelphia and analyzed contact-center data from Philly311 and other cities. The poll was conducted by telephone between January 8 and January 19, 2010, among a citywide random sample of 1,602 city residents, ages 18 and older. Interviews were conducted with 1,302 landline users and 300 cell phone users to reach a broad representative sample of Philadelphians. The final sample was weighted to reflect the demographic breakdown of the city. The margin of error for the entire sample is approximately +/- 2.5 percentage points. The margin of error is higher for subgroups. Surveys are subject to other error sources as well, including sampling coverage error, recording error and respondent error. Abt SRBI Public Affairs designed the survey and conducted the interviewing, working with Cliff Zukin, veteran pollster and professor of political science and public policy at Rutgers University.