Straphangers are waiting longer for the next train to arrive than the Metropolitan Transportation Authority's (MTA) reported numbers would have them believe, according to an audit released today by New York State Comptroller Thomas P. DiNapoli. The audit found that the MTA's New York City Transit (Transit) published "wait assessment" shows a decline in performance. Wait assessment measures the train's ability to meet goals for the length of time between trains.

"The MTA is very clear that it considers its wait time assessment to be its most important measurement of the reliability of subway service and riders' experience," DiNapoli said. "It turns out the way Transit calculates this measurement obscures the reality of straphangers' wait times. New York's subway riders deserve better."

Wait Times Getting Worse

Transit sets targets for the number of minutes between trains in an effort to provide evenly spaced service. Transit also sets a goal for meeting the weekday target of time between trains — measured by averaging all lines together — which it calls the wait assessment. The goal in 2013 was to have 79.4 percent of all trains meet their target and 80.7 percent in 2014 and 2015. Transit met this overall goal in 2013, with 80.3 percent, but did not meet the goal established for 2014 with 78.8 percent. Subway overall wait time performance declined between March 2013 and August 2014 as well. As of June 2015, Transit's year-to-date assessment performance was 78.4 percent, or 2.3 percent below its goal.

Excluding shuttle lines, among the numbered lines, the 1 line performed best and the 5 line the worst in both 2013 and 2014. Among the letter lines, the best trains were the C in 2013 and the D in 2014, and the A line was the worst in both years (73.7 percent in 2013 and 69.6 percent in 2014). See a chart of each subway line's performance.

See a chart showing which lines' are getting better and which are getting worse.

Questionable Math

By the Transit's own measure, wait times are getting worse. But even Transit's own measure may make subway performance appear better than it is in reality. DiNapoli's audit determined that Transit's wait assessments are potentially misleading for several reasons. First, Transit uses a score that is an average of averages. It calculates a simple average of the combined annual average performances of all 23 subway lines and shuttles.

Transit's wait time score gives a shuttle train the same value as a bigger line. For example, Brooklyn's Franklin Avenue shuttle, which makes 119 round trips each weekday, received the best score in the 12-months ending August 2014 by meeting its target wait time 96.2 percent of the time. That score was given the same weight as — and averaged in with — the 5 train, which had the worst wait assessment in the system during that period, meeting its goal just 68.3 percent of the time. Averaging lines like these together results in a rosier image of subway reliability.

Also, the MTA's wait assessment is a moving average, based at any given time on an average of the previous 12 months. A moving average smooths out short-term highs and lows. If there is any major improvement or spate of problems in a given month, it gets muted.

Fixes Far Down the Line

Riders expecting wait times to improve any time in the near future will have to keep waiting. DiNapoli's audit concluded that while Transit has 11 committees and two task forces that address subway delays, it does not have a comprehensive plan to deal with the long-term causes of service disruptions.

The MTA's 2015-2019 Capital Program cites signals as being the leading cause behind subway service delays and describes plans to modernize signal and communication systems. Modernizing signal technology could let the MTA run trains closer together. Currently, Transit plans to convert to Communications Based Train Control (CBTC), a proven technology that permits trains to operate closer together. However, progress toward CBTC will likely be slow. The MTA's own assessment is that antiquated signals will be replaced and half of the subway system will be modernized to CBTC by 2034 – nearly two decades away. Additionally, the current capital plan reduced other improvements needed to allow for CBTC and it is unclear how this impacts the current timeline.

As a result, the MTA has no formal long-term plans detailing when CBTC installations would begin, how much they would cost, when they would be finished, and the estimated performance improvements (including wait time reductions) that would be achieved.

DiNapoli's audit called for the MTA to:

• Revise and disclose how it calculates and reports its performance when it comes to wait times.
• Address the decline in wait time performance with a detailed and comprehensive plan that includes structural and information technology improvements and cost estimates.
• Develop and implement a process to advise riders of delays and alternative routes based on real-time data obtained by traffic checkers to help alleviate congestion.

Transit disagreed with the audit findings that it has not developed a full and comprehensive plan to deal with long-term causes of service disruptions. It also disagreed that its method of using a system wide average overstated its wait assessment results, but will revisit the issue as full data becomes available. Transit officials did agree that wait time performance has declined. They also said they expect to revisit how it collects and reports wait assessment data.

Read the full audit "Subway Wait Assessment" or go to http://www.osc.state.ny.us/audits/allaudits/093016/14s23.pdf.

See charts of subway line's individual performance:

http://web.osc.state.ny.us/reports/mta/MTA_Wait_Times_2f.pdf 
http://web.osc.state.ny.us/reports/mta/MTA_Wait_Times_4c.pdf

Read DiNapoli's audit "Train On-Time Performance," released Aug. 2015, or go to www.osc.state.ny.us/press/releases/aug15/081315.htm