LIRR reports of ridership and failures off track, comptroller's audit finds
Jan 7, 2020
The Long Island Rail Road uses misleading performance metrics to exaggerate the size of its ridership and reliability of its trains, according to a state report.
An audit from the Office of State Comptroller Thomas DiNapoli found “deficiencies and inconsistencies” in how the LIRR reports two key performance indicators — ridership and “mean distance between failures,” or MDBF, which measures how far trains travel before they break down.
“The MTA puts out mountains of data about ridership and service reliability, but too often its numbers are skewed or simply misleading,” DiNapoli said in a statement. “Riders deserve honest and accurate information on how MTA’s trains are running.”
The LIRR defended its system of measuring train reliability, but said it was open to considering new ways to count passengers.
One key flaw in how the LIRR calculates its train reliability, according to the report, is that it measures miles traveled by individual train cars, rather than an entire train, resulting in a “significantly higher … and misleading” estimates of mean distance between failures. The LIRR’s MDBF could be eight to 10 times higher, based on the number of train cars, the report said.
The audit, which analyzed Metropolitan Transportation Authority performance metrics between Jan. 1, 2015, and Aug. 22, 2018, also found that the LIRR excluded several mechanical failures when calculating its MDBF.
The report noted that, in February 2018 alone, the LIRR reported 202 train failures, but only counted 24 of them when calculating its MDBF. The others were left out for not meeting the LIRR’s criteria, which include that a failure was mechanical in nature and resulted in a delay of six minutes or more. But the report said 14 incidents were not counted, despite appearing to meet the criteria.
Auditors said, in some instances, the LIRR only counted the failure of single train cars breaking down, rather than entire train sets in apparent contradiction of its policy.
“LIRR officials explained it is their practice to only count a single failure for the train set and not a failure for each car in the train set,” the auditors wrote.
LIRR train reliability improved slightly in 2019 after falling steadily for several years. According to LIRR statistics, MDBF fell from 216,772 miles in 2016 to 185,217 miles in 2018. Through the first 10 months of 2019, LIRR trains traveled about 186,000 miles between breakdowns.
LIRR officials have posited that train reliability metrics will improve as the railroad replaces its 1980s-era “M3” model trains with its new “M9” trains. The LIRR does not expect to completely replace its antiquated fleet until 2024.
In a written response to DiNapoli, LIRR president Phillip Eng defended its MDBF calculations as “an accurate measure of car fleet reliability, which is consistent with the rail industry.”
The American Public Transportation Association, a nonprofit trade group, defines MDBF as “the arithmetic means of the distance traveled between successive failures of a repairable vehicle.”
The audit also took issue with how the LIRR measures how many riders it carries. The report noted that the LIRR calculates ridership based on ticket sales, and not trips or actual passenger counts, “and, as such, does not accurately reflect actual ridership.”
The LIRR’s formula for translating ticket sales into riders relies on a model developed from a 1983 Metro North passenger travel survey, the report noted.
“As it is based on demographics and commuting patterns from 36 years ago, the formula is of questionable value in producing reliable estimates today,” the audit said.
Through November, the LIRR was on pace to set another modern annual ridership record of more than 90 million customers, according to the railroad’s own statistics.
DiNapoli’s office made several recommendations to the LIRR, including that it consider other ways to measure and report train reliability, and that it update its formula for measuring customers to consider changes in travel habits and fare policies.
Eng, in his response to the audit, said that, absent a gating system that counts every passenger getting on a train, “the LIRR must make assumptions about how often, and precisely when, customers who purchase tickets travel.”
The LIRR will continue to reassess its methodology, Eng added, and research how other commuter railroads calculate ridership.
LIRR spokesman Aaron Donovan said the railroad’s methods of measuring performance have remained consistent. He said the railroad tracks MDBF by car over trains because each car is individually maintained. The railroad’s priority, he added, is continuing to improve service for its customers.