Author Archive

Off-Peak Fare Discounts and their Behavioral Implications

June 8th, 2016 3 comments

Metrorail riders facing a high percentage off-peak discount are motivated to delay their travel and save.

This post is guest-written by Sam Winward, a behavioral economist and Metrorail commuter who lives and works in the District. His analysis of Metro ridership patterns sheds light on the influence of off-peak pricing.

If you’ve ever waited a few extra minutes before swiping through a Metrorail faregate to qualify for the off-peak fare, you’re not alone. Within the afternoon peak / off-peak cusp period, when riders may be somewhat time flexible, Metrorail data from AprilMay 2015 confirms that some riders are delaying entry for savings.

What riders probably don’t know, is that their off-peak pricing discount could be substantially different from the riders around them. In fact, off-peak percent discounts can range from a 19% to 40% reduction of peak fares. Given the varying degree of off-peak incentive, riders are responding as we’d expect when weighing their opportunity cost of time.

As percent savings increase, riders are more likely to delay travel for off-peak fares. The graph below, derived from WMATA data, displays this ridership tendency.

Note the ridership uptick just after 7 pm. This runs counter to the natural pattern of declining ridership over this period, and the uptick steepens as the percent discount facing riders increases.

Exhibit 1_Ridership Uptick

So why is there so much variation in off-peak discounts, and who are the lucky riders allowed large savings?

Particular trips, based on mileage between origin / destination stations, are subject to off-peak fare “caps.” WMATA introduced the caps in 2012 when off-peak fares switched to incremental fee-per-mile pricing. Off-peak fares for trips of just under 7 and 10 miles were capped, as these trip distances would have otherwise seen very large increases between the old and new fare structure. A maximum off-peak fare cap of $3.60 was also implemented, affecting riders with trips lengths greater than 11 miles.

The effects of these fare caps remain in the system today, and are not mirrored in peak fares.

As displayed below, the gap between peak and off-peak fares increases over regions subject to off-peak fare caps. This makes off-peak travel more enticing for riders’ whose trip length falls within a green zone.

Exhibit 2_Fares

Notably, small tweaks in the fare structure have large behavioral effects.

Even if you decide that added off-peak savings aren’t worth the wait, at least your decision can be informed. To see the off-peak discount applicable to your route, and how it compares to the rest, check out the interactive off-peak discount calculator on my website.

The full story, including the off-peak discount calculator and a formal analysis of the delayed ridership patterns, can be found at Sam’s site.

Rail Ridership Data to Help the Region Plan for SafeTrack

May 27th, 2016 3 comments

To help the region and our partners plan alternatives and mitigate the impacts of SafeTrack, Metro releases rail ridership data applicable to this important maintenance effort.

SafeTrack example 2Rail Link Volumes: This data describes the number of customers on board trains between two contiguous stations, for a given hour of the day, then assigned to rail links.  For example, the link volume from Bethesda to Friendship Heights is the sum of everyone who boarded upstream, minus those who exited at or before Bethesda. This can be useful for planning SafeTrack mitigation efforts because it gives a first-order estimate of the potential demand for a bus bridge, for example.

Online Interactive Map of Metrorail Link Volumes, average weekday and Saturday in May 2015

Metrorail Link volumes, average weekday in May 2015, by hour, by line color (580kb, .xlsx)

Read more…

Metrobus Ridership During the Rail Safety Shutdown

May 23rd, 2016 2 comments

When Metrorail closed on March 16th, tens of thousands of rail riders switched to bus, including almost 20,000 riders who took their first bus ride in over a month! Bus-to-bus transfers spiked 45%, and ridership surged in downtown and central areas but fell in the suburbs.

Ballston Bus

Special thanks to the Systems & Performance Analysis team in Bus Planning for their help developing this analysis.

When Metrorail closed on Wednesday, March 16, Metrobus braced for impact as over 700,000 displaced rail trips sought alternatives. But there was little time or capacity to significantly alter bus service. What happened to bus ridership?

Overall ridership as tallied by the farebox came in at just 5% over the monthly average, or about 20,000 additional trips. So the changes look fairly small given the volume of displaced rail trips.

But don’t be fooled by the bottom line.  Underneath that total, a seismic shift was taking place. Read more…

Data Download: Metrorail Ridership by Station by Month, 2010-2015

March 24th, 2016 6 comments

Data_clipSee how seasons, land use, and service drive trends in rail ridership at the station level, in this new data download.

This latest data download shows Metrorail ridership by station, by month, for the last five years or so. It hints at the complex factors driving rail ridership – from short-term effects like weather or service changes, to long-term trends like real estate development and office relocations.

We see a few tidbits in this data:

  • Seasonal trends: rail ridership follows a predictable pattern each year – peaking in the summer and around Cherry Blossoms, and reaching lows around the holidays.  Compare the high seasonality of Arlington Cemetery to a more commuter-oriented station, for example.  Ridership in the summer at that station can quintuple over its winter base.
  • Weather impacts: see how the blizzard this past January lowered the average for the month as service was shut down.
  • Service changes: See how the opening of the Silver Line shifted riders from Orange to Silver in July-August 2014.
  • Land Use is key! Look at the recent growth rates at stations like Navy Yard and NoMa (formerly New York Ave.), reflecting the new jobs and residences near those stations.

Metrorail Ridership, by Station by Month, 2010-2015, Average Weekday (.xlsx, 120kb)

Metrorail Ridership, by Station by Month by Period, 2010-2015, Average Weekday (.xlsx, 630kb) (Added 3/30/2016)

Notes: these numbers are raw entries for an average weekday in the month, including snow days, excluding holidays when we did not run a weekday schedule.  The numbers are for trend analysis and will differ slightly from those we report in financial statements, which undergo additional data scrubbing and normalization.

What do trends you see?

Ridership Before and After the Blizzard of 2016

February 22nd, 2016 3 comments

What happens to ridership when a major snowstorm closes Metro?

The blizzard at the end of January dumped around two feet of snow on the Washington region over the course of a few days, making it the fourth-largest snowstorm in D.C. in over a century. To protect equipment, for the safety of employees and riders, and to focus on cleanup efforts, Metro ceased all operations for over three days. The Federal Government and schools were closed.

The chart below tells the story of the blizzard from a ridership perspective.  Actual bus and rail ridership by hour is shown as blue and red, with the orange dotted line showing a typical pattern from two weeks prior in January.

image012

Even before the snowflakes began falling on the afternoon of Friday, January 22, Metro’s ridership was impacted. That day, riders adjusted their normal routines by staying home, teleworking, or leaving early. The morning’s peak was half of normal. By Friday night, bus and rail operations ceased. No service operated Saturday or Sunday. Very limited bus and rail service was offered for a few hours on Monday, January 25 as cleanup and recovery began. Fares were not charged on rail, so we recorded very few transactions.

We recovered gradually on Tuesday and Wednesday, and nearly fully by Thursday, even as many school systems remained closed.

In the end, total ridership losses were estimated to be about 2.9 million boardings over five days, 1.6 million on rail and 1.3 million on bus. Bus took a proportionately bigger hit than rail given service restrictions from icy and unsafe road conditions.

What do you see in these numbers? Also, feel free to download the underlying data (.xlsx, 40kb) for yourself.

Four Key Questions about Metro’s Future with the Federal Worker (5 of 5)

December 9th, 2015 No comments

The answers to these four questions will shape the future of Metro’s federal customers, and the region’s transportation future. (Fifth and last in a series of posts on Metro’s Federal customers – see posts 1, 2, 3, and 4)

The ATF headquarters adjacent to NoMa Metrorail station has helped grow ridership there significantly.

1. Will the transit benefit be restored to parity with parking? When Congress cut the transit benefit in half, it hurt Metro riders hard. 42% of Metro’s ridership – around 500,000 rail and bus trips per day – comes from riders who use the Federal Transit Benefit, including private-sector workers. At Metro, 22% of all ridership comes from commuters who spend over $130 per month on transit. Following the changes to the SmartBenefits program, Metro saw ridership losses concentrated on these riders hit the hardest, and federal employees overwhelmingly pay with SmartBenefits.

If Congress restores the maximum transit benefit to parity with parking, it would be a huge boon to Metro’s federal customers and Metro’s bottom line. Read more…

Why We Care About GSA’s Location Decisions: Lessons from the History of Metro’s Federal Customers (4 of 5)

December 2nd, 2015 4 comments

Data show that where GSA chooses to locate federal office buildings has a huge impact on Metrorail ridership from federal commuters.  But in the meantime, non-Federal riders in the inner jurisdictions are driving up ridership outside of the usual commute market. (Fourth in a series of posts on Metro’s Federal Customers – see posts 1, 2, and 3)

Growth in Metrorail Ridership from Feds by Time of DayBetween 2002 and 2012, rail ridership from federal employees has grown 15%, the same as from non-federal riders.  (N.B. this post focuses on rail only; no comparable survey data for bus is available.)  Federal employees have remained about a third of total ridership, as overall ridership ebbed and flowed. Most of these new federal riders live in the inner jurisdictions of D.C., Arlington, and Alexandria – ridership from federal employees has been much slower in the outer jurisdictions, particularly Fairfax County (growing at 5-15%, vs. 25-40% over ten years).  The growth from federal riders has mimicked existing riders – they are focused on the peak commute too, with a moderate amount of off-peak travel as well.

But over the same timeframe, non-federal customers drove up ridership much faster in the PM Peak and Off-Peak times.  These riders similarly come from the inner jurisdictions. Read more…

Where Are Metro’s Federal Customers Going? (2 of 5)

December 1st, 2015 No comments

Employees of the federal government comprise 27% of Metro’s weekday ridership, but what rail stations and bus routes are they using?

(Second in a series of posts on Metro’s Federal Customers – see post 1)

On Metrobus, federal workers are about 10-20% of most bus routes’ ridership, with a few logical exceptions. Federal ridership is higher on bus routes that are more “peaked” and commute-oriented, and/or on routes that directly serve federal facilities.

Top Metrobus Lines for Feds, by Pct of Riders

Read more…

How Metro’s Federal Customers Pay Fares, and Why It Matters (3 of 5)

November 30th, 2015 3 comments

Metro’s federal customers pay fares a little differently than other riders. Why is that such a big deal for Metro’s financial future? 

(Third in a series of posts on Metro’s Federal customers – see posts 1 and 2)

SmartBenefits Feds vs NonCompared to other riders, Metro’s federal customers are much more likely to pay their fare using SmartBenefits.  SmartBenefits are the type of funds you load onto your SmarTrip card – usually through your employer, either as direct subsidy, or a certain amount of pre-tax dollars you set aside for transit fares. Metro riders in the Washington region may know the program manager WageWorks.

 (Importantly, SmartBenefits is not exclusively a federal government benefit!  On the contrary, non-federal workers are nearly half the overall enrollment in the SmartBenefits program, and 42% of Metro’s overall ridership comes from riders paying with SmartBenefits.  We’ll make clear the difference between SmartBenefits and SmartTrip in an upcoming Metro 101 post – stay tuned.)

84% of federal employees on Metro pay their fare with SmartBenefits, compared to 27% of non-federal customers. In addition, federal customers tend to pay higher fares on rail – because federal workers typically take longer trips and ride more at peak times (average peak fare $3.00, vs. $2.87 non-federal customers).

Finally, 87% of Metro’s federal customers pay using “stored value” (pay-as-you-go funds), rather than a weekly or a monthly pass. Very few still use paper tickets or passes – but this is similar to other riders.

Read more…

Metro’s Federal Customers: A Snapshot (1 of 5)

November 19th, 2015 No comments

Think Metro is all about getting the federal commuters to work? Think again!

(First in a series of posts on Metro’s customers who are Federal Government employees)

Just as the workforce in the Washington region has a sizeable share of federal workers, so has Metro’s ridership.  Metro serves major federal employment centers downtown, and even boasts stations named for the federal sites they serve, like Federal Triangle, Medical Center, and Pentagon. But while Metro has a long supported the federal government, it’s a myth that Metro is all about federal government commuters and nothing more. Federal workers are a minority of riders and have been for years, and federal funding is playing an increasingly smaller role in Metro’s finances.

So just who are Metro’s federal customers?  When and what do they ride? Where are they coming from and going to, and how has this changed in the last decade? The next series of posts seeks to answer just that, using passenger survey data (bus and rail) where customers identified as employees of the federal government or not (contractors excluded).

How Many, Where, and When? About 27% of all Metro weekday trips are made by federal workers – a total of 317,000 boardings across bus and rail.  These federal employees can be anyone from a nurse at Walter Reed Medical Center, to a military officer at the Pentagon, to a Congressional staffer on Capitol Hill.  The majority of these trips (255,000) are made on Metrorail, where federal workers make up 35% of all boardings (all-day).  The remainder – just over 60,000 boardings from federal workers – happen on Metrobus, where riders are generally less likely to be federal workers (14% of all bus boardings are federal).

Pct Fed Workers by Mode and Period Read more…