I. Findings


PROFILES OF
20 SUBWAY LINES

(click your line!)

1 A J/Z
2 B L
3 C M
4 D N
5 E Q
6 F R
7 G

What do subway riders want?

They want short waits, trains that arrive regularly, a little elbow room, a clean car, and understandable announcements that tell them what they need to know. That's what MTA New York City Transit's own polling of rider satisfaction measures.[1]

This "State of the Subways" Report Card tells riders how their lines service. We looked at six measures of subway performance for the city's 20 major subway lines do on these aspects of Most of the measures are for all or the last half of 2016.[2] We also made minor technical changes in one of the measures.

Our Report Card has three parts:

First, a comparison of service on 20 lines as detailed in the attached tables.

Second, we give an overall "MetroCard Rating"[3] to the 20 major lines[4].

Third, the report contains one-page profiles of the 20 lines. These provide riders, officials and communities with a simple summary of how their line performs compared to others. This is the 18th Subway Report Card released by the Straphangers Campaign since 1997.[5]

Our findings show the following picture of how New York City's subways are doing:

  1. There was an unprecedented three-way tie for best subway line, with the 1, 7, and L each garnering a MetroCard Rating of $2.05. In a year there was no one big standout, the three lines came out on top because:
    1. Tops in service: The 1, 7, and L (along with the 6) provided the most generous amounts of morning and evening rush hour service;
    2. Individually, these lines performed well on some features:
      1. The 1 is less crowded and cleaner than the average subway line
      2. The 7 had less frequent subway car breakdowns than the average subway line and the greatest percentage of clean subway car interiors.
      3. The L had a nearly perfect score for accurate and understandable subway car announcements.
  1. The 5 and A tied as the worst performers in the subway system, with a MetroCard Rating of $1.50 each.
    1. Bottoms in regularity: The 5 and A are two of the system’s most irregular lines.
    2. Individually, these lines performed below average on several features:
      1. The 5 was the second most crowded line in the system.
      2. The A provided less frequent midday service than the average subway line and its subway cars experienced breakdowns at an above-average rate.
  1. System-wide, for 20 lines, we found the following on four of the six measures we can compare over time: car breakdowns, service regularity, car cleanliness and in-car announcements.
    1. The car breakdown rate worsened from an average mechanical failure every 141,202 miles to every 131,325 miles, comparing the 12-month period ending December 2014 to December 2015 - a loss of 7%.
    2. Subway regularity decreased since our last report. Lines meeting the “wait assessment” standard—which measures the regularity versus gaps and/or bunching in service—decreased 1.4%, from 78.8% regular in our last report, to 77.4% regular in the period ending December 2015. We found that only four lines improved (2, 3, F, and G), while sixteen worsened (1, 4, 5, 6, 7, A, B, C, D, E, J/Z, L, M, N, Q, and R).
    3. Subway cars rated clean improved slightly, from 92% in our 2015 report to 93% in our 2016 report. We found that eleven lines improved (1, 2, 3, 5, 6, 7, F, G, L, N, and R) and nine declined (4, A, B, C, D, E, J/Z, M, and Q).
    4. Accurate and understandable subway car announcements remained the same—91% understandable and accurate—since our last report. We found five lines improved (7, C, D, F, and M), ten declined (1, 2, 3, 4, 5, 6, A, G, J/Z, and R) and five did not change (B, E, L, N, and Q).
  1. There are large disparities in how subway lines perform.
    1. Breakdowns: The Q had the best record on delays caused by car mechanical failures: once every 489,956 miles. Cars on the C were worst, with a breakdown rate almost eighttimes higher: once every 61,544 miles.
    2. Cleanliness:The 7 line had nearly a perfect score at 99%. The dirtiest line - the Q - had 15% of its cars rated moderately or heavily dirty.
    3. Degree of crowding: The 4 was the most crowded line, at 112% of the crowding guidelines at the AM rush peak load point. The R was the least crowded, at 72%.[6]
    4. Amount of scheduled service:The 6 and 7 lines had the most scheduled service, with two-and-a-half minute intervals between trains during the morning rush hour. The C line ranked much lower, with nine-minute-plus headways during the same time period.
    5. Regularity of service:The G had the greatest regularity of service, arriving within 25% of its scheduled interval 81% of the time. The most irregular line was the 5, which performed with regularity only 67% of the time.
    6. Announcements:Three lines – the E, M, and N lines - had perfect performance for accurate and understandable announcements made in subway cars, missing no announcements and reflecting the automation of announcements. The 1 and B lines were worst, missing or garbling announcements 23% of the time.

II. Summary of Methodology

The NYPIRG Straphangers Campaign reviewed extensive MTA New York City Transit data on the quality and quantity of service on 20 subway lines. We used the latest comparable data available, all from 2015[7]. Several of the data items have not been publicly released before on a line-by-line basis. We then calculated a MetroCard Rating — intended as a shorthand tool to allow comparisons among lines — for all 20 subway lines, as follows:

First, we formulated a scale of the relative importance of measures of subway service. This was based on a survey we conducted of a panel of transit experts and riders, and an official survey of riders by MTA New York City Transit. The six measures were weighted as follows:

Amount of service

  • scheduled amount of service 30%

Dependability of service

  • percent of trains arriving at regular intervals 5%
  • breakdown rate 5%

Comfort/usability

  • crowding 15%
  • interior cleanliness 10%
  • adequacy of in-car announcements 10%

Second, for each measure, we compared each line’s performance to the best- and worst-performing lines in this rating period.

A line equaling the system best in 2015 would receive a score of 100 for that indicator, while a line matching the system low in 2015 would receive a score of 0. Under this rating scale, a small difference in performance between two lines translates to a small difference between scores.

These scores were then multiplied by the percentage weight of each indicator, and added up to reach an overall raw score. Below is an illustration of calculations for a line, in this case the F.

Figure 1

Indicator

F line value including best and worst in system for 5 indicators

F line score out of 100

Percentage weight

F line adjusted raw score

 Scheduled service

AM rush—4 min, 15 sec;

noon—7 min, 30 sec;

PM rush—4 min, 45 sec

69

30%

21

 Service regularity

72% (best—81%; worst—66%)

38

22.5%

8

 Breakdown rate

329,862 miles (best—489,956 miles; worst—61,544 miles)

63

12.5%

8

 Crowding

95% of max guideline (best—72%; worst—112%)

43

15%

7

 Cleanliness

93% clean (best—99%; worst—85%)

45

10%

6

 Announcements

97% adequate (best—100%; worst—77%)

87

10%

9

 Adjusted score total

     

F line—58 pts incl. rounding

Third, the summed totals were then placed on a scale that emphasizes the relative differences between scores nearest the top and bottom of the scale. (See Appendix I.)

Finally, we converted each line’s summed raw score to a MetroCard Rating. We created a formula with assistance from independent transit experts. A line scoring, on average, at the 50th percentile of the lines for all six measures would receive a MetroCard Rating of $1.75. A line that matched the 90th percentile of this range would be rated $2.75, the current base fare. The F line, as shown above, falls at a weighted 58th percentile over six measures, corresponding to a MetroCard Rating of $1.90.

New York City Transit officials reviewed the profiles and ratings in 1997. They concluded: “Although it could obviously be debated as to which indicators are most important to the transit customer, we feel that the measures that you selected for the profiles are a good barometer in generally representing a route’s performance characteristics… Further, the format of your

profiles… is clear and should cause no difficulty in the way the public interprets the information.”

Their full comments can be found in Appendix I, which presents a more detailed description of our methodology. Transit officials were also sent an advance summary of the findings for this year’s State of the Subways Report Card.

For our first five surveys, we used 1996 — our first year for calculating MetroCard Ratings — as a baseline. As we said in our 1997 report, our ratings “will allow us to use the same formula for ranking service on subway lines in the future. As such, it will be a fair and objective barometer for gauging whether service has improved, stayed the same, or deteriorated over time.”

However, in 2001, 2003, 2004, 2005, 2008, 2009, 2010, 2011, 2013, 2014 and 2015, Transit officials made changes in how performance indicators are measured and/or reported. Transit officials rejected our request to re-calculate measures back to 1996 in line with their adopted changes. As a result, in this report we were forced to redefine our baseline with current data, and considerable historical comparability was lost.

Also due to changes in the measuring and/or reporting of data by Transit officials, it was necessary to make modest adjustments to the MetroCard Rating calculation and scale—as was the case in several earlier State of the Subways reports. In selecting this scale, we attempted to create a single measure which we felt accurately and fairly represents the relative performance priorities listed in our original 1996 poll of riders, community leaders and independent transit experts.

III. Why A Report Card on the State of the Subways?

Why does the Straphangers Campaign publish a yearly report card on the subways?

First, riders are looking for information on the quality of their trips, especially for their line. Our profiles seek to provide this information in a simple and accessible form. Our profiles have pressed NYC Transit to do likewise.

In recent years, the MTA has both gone forward and stayed stagnant.

For example, in 2011, when current MTA Chairman Tom Prendergast was the President of New York City Transit, he created a "key performance indicator." The KPI is a line-by-line composite that evaluates three basic measures of service. Unfortunately, the KPI has not caught on, as the measure does not allow for meaningful comparisons and gives too much weight to the weight assessment statistics.

In 2009, the MTA began posting monthly data on subway car breakdowns by each of the 20 subway lines. However, sometime in 2013, the MTA stopped reporting this information, saying they found the blended measure inaccurate. It was not replaced.

Our Report Card has coincided with an uptick in the public's desire for "transparency" in government. The MTA's "big data" sets—on-time service, for example—allows for complex analyses only recently possible. Unfortunately, these transparency initiatives have moved slowly. For example, in four years we have not succeeded in getting data on the nature of transit riders' complaints. The Straphangers Campaign believes that should change.

Second, our Report Card provides a picture of how the subways are doing. Riders can consult our profiles and ratings and see how their subway line compares to others. For example, this report warns riders of the steady deterioration of subway car breakdown rates. Future performance will be a challenge given the MTA's tight budget. We will be watching.

Lastly, we aim to give communities the information they need to win better service. We often hear from riders and neighborhood groups. They will say, "our line has got to be worst;" or "we must have the most crowded trains;" or "our line is much better than others." For riders and officials on lines receiving a poor level of service, our report will help them advocate.

Our reports can be found online at www.straphangers.org, as can our profiles. We hope that these efforts - combined with the concern and activism of many thousands of city transit riders - will win better subway and bus service for New York City.



[1] New York City Residents' Perceptions of New York City Transit Service, 1999 Citywide Survey, prepared for MTA New York City Transit.

[2] The measures are: frequency of scheduled service; how regularly trains arrive, delays due to car mechanical problems; crowding at morning peak period; car cleanliness; and in-car announcements. Regularity of service is reported in an indicator called wait assessment, a measure of gaps in service or bunching together of trains.

[3] We derived the MetroCard Ratings with the help of independent transportation experts. Descriptions of the methodology can be found in Section II and Appendix I. The rating was developed in two steps. First, we decided how much weight to give each of the six measures of transit service. Then we placed each line on a scale that permits fair comparisons. Under a formula we derived, a line whose performance fell exactly at the 50th percentile in this baseline would receive a MetroCard Rating of $1.75 in this report. Any line at the 90th percentile of this range would receive a rating of $2.75, the current base fare.

[4] We were unable to give an overall MetroCard Rating to the system's three permanent shuttle lines - the Franklin Avenue Shuttle, the Rockaway Park Shuttle, and the Times Square Shuttle - because data is not available.

[5] No Report Card was issued in 2013 given concerns about the impact of Superstore Sandy on the subway system. That was also the case in 2002 following the attack on the World Trade Center. As a result, the Straphangers Campaign has issued subway Report Cards twenty times in eighteen years.

[6] In previous reports, the Straphangers Campaign calculated crowding conditions using New York City Transit’s annual Cordon Count. In the 2016 report, we cite a new data set—Sixty-minute Weekday AM Rush Peak Load Point Summary—recently made available by New York City Transit.

[7] See Appendix I for a complete list of MTA New York City Transit data cited in this report.