Posts about Peak Of The Peak

With a SmarTrip card, rail fares will range from $1.70 to $3.50 off-peak and from $2.10 to $5.75 during rush hours. Regular local bus fares will rise from $1.60 to $1.70 for SmarTrip customers, while customers paying in cash will have their fares rounded to the nearest dollar.

Since use of SmarTrip by visitors and non-regular riders is expected to increase, SmarTrip vending machines will be installed at more all stations.

With the elimination of the peak-of-the-peak fare, station fare tables will go back to having just two columns. But riders shouldn't expect to save a whole lot, since the "regular" fare has been increased enough to cover the difference. With the peak-of-the-peak surcharge, the current maximum fare is $5.20. It will rise to $5.75 under the proposal.

Parking at Metro lots and garages will increase by 25¢ per day, about a 5% increase. Bike locker fees will be cut from $200 per year to $120 per year, something we argued for based on low demand for lockers.

Most disappointing to me is that discussion of implementing some sort of flexible monthly pass has stopped for this budget cycle, meaning that Metro customers will likely have to wait at least two more years to have the flexibility of paying for their commute and getting their off-peak trips for free. The topic of monthly passes was briefly discussed during an October meeting of the finance committee, but by November had disappeared from the discussion.

The fare increases are expected to raise about half of Metro's $120 million shortfall for the coming fiscal year, with local jurisdictions expected to chip in the other half of the shortfall in order to balance the budget. Metro's finance committee will discuss the fare increase along with the rest of the budget on Thursday morning.

Below are brief descriptions of each idea, and our take on them.

Integrated monthly passes: Michael Perkins has long been advocating for this, which he calls Smart Passes. They resemble Sound Transit's ORCA system. Basically, it could be good to get most regular riders habitually buying monthly (or weekly) passes. But how much should a pass cost? There many different fares.

The idea is to let you buy a pass of any size. It will cost 40 times the regular one-way fare of your regular commute. You get to take that trip, or any other trip of equal or lesser cost, for free with the pass. For more expensive trips, you pay the difference.

Benefits of this option include encouraging more off-peak ridership ("free nights and weekends"). Mobile networks are least busy nights and weekends, which is why many now offer these times free. Metro and buses are similarly less crowded, which is why off-peak ridership should be encouraged.

Many riders will save some money, so it might cost Metro, but on the other hand it will stabilize the revenue. If there's a major snowstorm or government shutdown, Metro won't lose as much because people will still have bought their passes.

Zone fares: This is another Greater Greater Washington suggestion; the presentation notes that their evaluation is "based on blogger proposal." Though, actually, Matt and Michael proposed this specifically to explain that it might not be the best approach.

Zones make it a little easier to calculate your fare, since you just have to count zones. But you would still look up your fare in a table unless you opt to study the map and add. A zone system would still be complex, and some riders whose stations lie right near a zone boundary could end up paying a lot more.

Metro is studying different zone alternatives besides just the one Matt and Michael analyzed. One possibility would be to study a system where if you cross from, say, Zone 3 to the core (Zone 1) and back out to Zone 3 on the other side, you pay for 6 zones instead of 3. But that would still save some a lot of money and cost others greatly.

Flat fares: The simplest option is to just charge one fare. It gets suggested frequently, especially by people familiar with other subways like New York's.

However, that comparison actually isn't right. Metro is a hybrid of commuter rail and urban subway. In New York, Boston, Chicago and many other cities, the subway stays inside city limits, and a separate commuter rail system serves more outlying suburbs. Those commuter railroads all charge based on distance or zones.

Metro, and BART, combine the two. Someone commuting downtown from Shady Grove is more equivalent from coming in from White Plains on Metro-North. If DC's transit were like older cities, Metro would not go outside DC, Arlington, and Alexandria, and most people in Montgomery, Prince George's and Fairfax would take a larger MARC or VRE.

Anyway, a flat fare would mean everyone pays $2.70 per trip, much more than in other cities with urban subways that don't go to the suburbs. Now, a lot of short Metro trips cost only $1.60, much less than, say, New York's $2.25 regular fare.

And there's the question of equity. A flat fare would mean people taking shorter trips would pay more to subsidize people taking longer trips. Is it fair to charge someone traveling from Union Station to Rhode Island Avenue the same as someone traveling from Shady Grove to Metro Center?

Flat fares for paper farecards: One way to simplify things for tourists would be to charge a flat fare for using a paper farecard, either everywhere or just inside a core zone. That fare could be the highest fare you'd pay for any trip in that zone. That costs tourists (and those without SmarTrips) more, but is very simple.

Free bus-rail transfers: During every fare debate, some suggest raising bus fares substantially. They point out that bus fares are much lower than rail fares and the farebox recovery rate on bus is much lower than on rail. Many other cities, including New York, charge the same flat fare for bus as for subway trips.

The typical counterarguments note that bus riders are generally lower income, meaning the difference serves a social equity purpose, and that buses are slower than the trains. There's one more important point to remember: in New York, for instance, while one pays the flat fare for a bus ride, they can then transfer for free to rail, and vice versa.

Metro, instead, charges the full bus fare minus a 50¢ transfer "discount" when riding both modes. Metro could instead make the transfers fully free, letting anyone get on a bus who's just gotten off rail without paying, and deducting a bus trip from the rail fare if someone rides bus and then rail afterward.

Higher fare for faster buses: If part of the reason for lower bus fares is that buses are slower, what about the faster limited-stop buses? Maybe those should cost more?

This makes some sense. One potential drawback is that riders who don't understand the more confusing system could end up in altercations with bus drivers. The limited-stop buses are supposed to be blue while the local buses are red, but the bus garages don't always follow the rules properly. It would be much more important to get them right if one type of bus costs more.

Get rid of "peak of the peak": We pushed for instituting peak of the peak last year, but it hasn't worked as hoped. The idea was to charge more at the busiest times, to either encourage people to shift when they commute or shift to bus.

However, very little behavior shifting happened. Metro made the peak-of-the-peak period longer than advocates had suggested, and decided to apply it to all trips instead of ones in the congested core, peak direction. Also, it was just more confusing. Few people will change their behavior if it's not easily evident to them when they're paying more versus less. And many people can't change their travel times.

Therefore, it's probably wise to eliminate this. That will, though, mean that regular peak fares or all fares have to rise more than they would otherwise.

The list of ideas includes a brief reference to "Rail reverse-commute discounts," which seems to mean the possibility of charging less for the trips against the rush which just use mostly-empty capacity instead of adding to the crush.

Minimum on-board SmarTrip reloads: One of the biggest sources of bus delay is people loading SmarTrips on the bus, which takes a long time. The best solution is to find more stores that can offer SmarTrip loading, or install kiosks. Meanwhile, Metro could require riders load a certain amount on their SmarTrips.

Some riders might not have that much money at one time, so it would be very important to couple any minimum with other alternatives for those riders who need to load small amounts of money.

Others: A slide lists other options without much explanation, such as rounding all rail fares to the nearest quarter, setting rail fares based on number of stations traversed (unfair; would the Orange Line between Metro Center and Farragut Square cost twice as much as the Red Line?), a bonus for storing more value on a SmarTrip, more passes, and others.

Staff expect to make a recommendation in January. Hopefully they will also be able to release details of the analysis of all of these options, to help the public better understand why they made the choice they did and advocate for or against the choice.

One way to simplify fares for tourists and people who don't have SmarTrip cards would be to simply make a fixed fare, equal to the peak-of-the-peak fare, that applies all the time to paper farecards. With this, Metro could also eliminate of the paper farecard surcharge.

Such a move would not be without drawbacks. SmarTrip card usage is high for bus (78%) and rail (82%), but there are still barriers for people with disabilities and limited incomes that prevent them from using SmarTrip cards. Any move to reduce the discounts available on paper farecards should be implemented alongside improvements that allow everyone to use SmarTrip.

But a table with just one, instead of three, columns would simplify the system particularly for those visiting DC, who are very likely to buy paper farecards. Since everyone knows (or should know!) what station they're going to, it makes the decision easy. You no longer need to know what time you're going or returning. If you're traveling during the peak of the peak, there's even a small discount compared to today because the fare for paper and Smartrip are the same.

There's still an incentive to get a Smartrip card, since they're convenient and offer discounts for off-peak rides. They protect your balance and offer transfer discounts to buses throughout the region.

There are currently six different fares between any two stations. This change would reduce that to three (Peak of the Peak/Paper farecard, Regular, Reduced). Only one of these fares would actually be listed on the fare machines. That's much simpler.

The presentation also points out that as WMATA switched from SmarTrips to a more modern "commodity card" which will be much cheaper, it will become easier to drop the price of the cards. Based on this, just keeping SmarTrip prices the way they are until new cards come in seems increasingly persuasive. More commenters supported this option than any other in our previous discussions.

At Thursday's Board meeting, the finance staff will also present number of draft "fare policy principles," high-level guiding theories to guide any decisions around fares. The Board often makes its fare decisions based on immediate circumstances, but hasn't stepped back to think about the broader goals that could shape those decisions.

The proposed principles are:

1. Develop fares so they are easily understandable
   
2. Charge fares relative to level of service (people who get more service, like longer distance riders, or faster service, like rail over bus, should pay more)
   
3. Optimize use of existing system capacity (the fare structure should encourage more riding at times or in places where the system already has room)
   
4. Maintain adequate cost recovery while maximizing ridership
   
5. Facilitate movement between modes and operators throughout the region
   
6. Encourage use of cost-effective and efficient fare media (like SmarTrip or contactless credit cards)
   
7. Ensure fares comply with federal regulations (particularly civil rights laws around low-income riders and riders with disabilities)

According to the presentation, riders in focus groups thought the existing fare charts are hard to understand and that distance fares are too complex. They do not like peak-of-the-peak fares, but like daily or weekly passes.

The customer focus groups wanted a simpler time and distance based system with open payments and additional SmarTrip features. Some of the most popular features are likely automatic reloading, and being able to load additional fare from the Internet, both of which WMATA is working on.

These principles do give some guidance, though they are also general enough that many different policies could fit into them. For example, monthly passes like those Michael has recommended would satisfy all of the principles. Peak-of-the-peak, on the other hand, does a lot to meet #3 and #4 but isn't so good on #1. The presentation says that the current passes and peak-of-the-peak will be evaluated against these fare principles.

Metro is also considering approving a $351,000 grant to provide remote detonated bomb jamming equipment. The capability will allow Metro to deploy a vehicle-mounted jammer and portable jammer to reduce the likelihood that a detected bomb could be detonated remotely. After the initial three-year grant expires, Metro will have to apply for continued maintenance grants.

Board members entered the April 27th meeting with a wish list of items they'd like to change. DC wanted to keep late night service going until 3 am and not charge a flat $4 fare after midnight. Fairfax wanted to get rid of parking increases.

DC's Jim Graham started out with a concrete proposal. He'd keep the late night service and charge a rush hour fare instead of a $4 fare. In exchange, he recommended increasing the peak-of-the-peak charge from 10¢ to 20¢. The late night service mostly benefits DC, Arlington, and Alexandria, though it also benefits suburban riders who ride to locations in DC, Arlington, and Alexandria.

The peak of the peak also hits DC, Arlington, and Alexandria riders a bit heavier than others, since being a flat fare, it's a greater percentage for those who ride short trips. The peak of the peak, as formulated, also will miss some riders from Shady Grove, Vienna, and other stations with long rides to the center, because many of those riders get on the train before 7:30 and are still on it when it's crowded downtown. But it's a reasonable tradeoff to make to pay for something that inner jurisdictions want.

Graham also suggested increasing the maximum fare, which does hit suburban riders. CFO Carol Kissall said that wasn't necessary to pay for his suggestions, so he didn't push the idea.

Next, Jeff McKay of Fairfax proposed cutting the parking fees. To pay for it, he suggested reducing the bus-rail transfer discount. Making transfers more expensive would have been a terrible idea, both for Fairfax and DC. It would discourage bus riding and push more Fairfax riders to drive to rail instead of taking bus to rail, even though riding the bus creates less congestion. And it would have harmed many inner jurisdiction residents who ride bus to rail and live nowhere near parking or don't even own cars.

Graham said he was happy to work with Fairfax to find a solution to the parking, but that the bus-rail transfer idea wasn't going to be the answer.

To summarize, now we have something Graham wants, the late-night changes, and something he's willing to do to pay for it that got general asset, the peak of the peak. And we have something McKay wants, the parking, and nothing specific to pay for it that's got broad support. When the meeting adjourned, it sounded like Graham was going to get late night changed and the higher peak of the peak, and McKay was going to need to find a funding source for his parking that wasn't unfair to DC, Arlington, and Alexandria.

Staff promised to research some of the ideas. They did so, and did a nice job of analysis. If they'd just presented a slide showing the costs of each change, the members could have resumed horse trading. But instead, they summed up only three items: the late night, the peak of the peak, and the parking, and ended up not surprisingly with a deficit. They then allocated that deficit to all of the jurisdictions, making DC and Arlington pay just as Fairfax and Maryland were.

This makes no sense. Inner jurisdictions get something and pay something, and outer jurisdictions get something, and everyone pays.

No wonder Jim Graham said that DC would veto the budget as is. To solve this, the Board needs to go back to the Sarles budget, and start horse trading again from there. The peak of the peak seems a fair way to cover the late night. Then, what would cover the parking? There's surely a deal to work out.

Next: Why is Fairfax so obsessed with parking?

The option to charge "to" congested stations would apply to trips to stations on the Red Line from Dupont Circle to Union Station, Orange from Rosslyn to Capitol South, Blue from Crystal City to Capitol South, Green from Navy Yard to Mount Vernon Square, and Yellow from Crystal City to Mount Vernon Square.

The option to charge "to and through" would only go as far as Farragut North on the Red Line and L'Enfant Plaza on the Orange, Blue, and Green Lines.

"To" has the advantage that it doesn't hit reverse commuters, like commenter inlogan who goes from McPherson Square to Tysons Corner. He's not using up any precious capacity as Orange Line trains westbound empty out at Metro Center. On the other hand, people who ride from Grosvenor to Suitland and use up a lot of room on crowded trains through the center would also be exempt.

"To and through" gets the through commuters but also hits the reverse ones. WMATA staff felt that my option 3, using congested segments only in the busier direction, was too complex.

To avoid conflict, the plan would actually based the charge on the time people entered the system in the morning, not the time they exit, since riders might get frustrated if they entered early enough to finish their trip before 7:30 only to be kept in the system by a delayed train. The delay is bad enough that they shouldn't also get hit with an extra charge.

Unfortunately, this also hurts people who take the rail system for short trips in the congested core at the very start of the peak-of-the-peak period, when it's not crowded in the core yet. For example, someone who rides from Mount Vernon Square to L'Enfant Plaza from 7:45-8:00 is not taking up precious capacity since the crowding is still mostly out at the fringes. The core itself gets crowded later.

Ideally, there could be 2-3 "zones" with staggered start times, such as an entry time of 7:15 at Shady Grove, 7:30 at Bethesda, and 7:45 at Cleveland Park, but this would complicate the implementation and potentially confuse passengers.

Here are the revenue and ridership implications of each:

"To" seems fairer, but raises less money and causes less behavior shifting while also less ridership loss. Of course, these elasticity estimates are pretty rough, since staff don't have any experience with "peak of the peak."

DC Board members want to maintain late-night service to 3 am, which will cost $2.1 million more than the 2 am closing in Sarles' latest budget, and also charge a peak fare instead of a $4 fixed fare late at night, which costs $1.5 million more. That totals $3.6 million.

A 50¢ "to" peak-of-the-peak (PotP) would only get $2.4 million over the previously-budgeted 10¢ across-the-board PotP. Approximately a 30¢ "to and through" would do it.

Another option is to also charge a bus PotP, which staff estimate would bring in $1.2M at 10¢ and $2.2M at 20¢, or increase the bus SmarTrip-cash differential to 25¢, for a gain of $1.5M.

So a 10¢ bus PotP and a 50¢ "to" PotP on rail would do it, or a 30¢ "to" PotP ($1.2M over the baseline) plus a 10¢ bus and a higher cash-SmarTrip differential, or other combinations.

The system is particularly crowded during a single hour in the morning and evening peak, and in a small core section. Therefore, it makes sense to charge extra for trips in that area, during that time, but not for those people who are riding from Franconia to King Street and aren't taking up the most precious space.

Where should the charge apply? Based on Matt Johnson's terrific analysis of system crowding, I'd suggest applying it to trips that start, end, or pass through stations from Dupont Circle to Union Station, Rosslyn to Capitol South, and Pentagon to Gallery Place. Below are several alternatives and a rough analysis of their value.

1a. Charge at the busiest stations

Ideally this charge would only hits riders on the most crowded segments of the system, since the goal is to target only those riders who are using up the most precious space and who, if they switched to another time of day, would better spread out ridership across the system.

One approach would be to only apply it to the busiest stations. Looking just at the stations with the most passenger traffic (scroll down to the table), means that Shady Grove and New Carrollton would be included. But the trains aren't at their most crowded there, it's just that those stations have a lot of boardings.

1b. Charge at the busiest stations by morning exits

How about looking at exits in the AM peak? Click on "Exits" in that table to sort in that way. The top 19 stations are basically contiguous, which would apply the congestion charge to anyone getting off the system at any stations from Dupont to Union Station on the Red Line, Rosslyn to Capitol South on Orange/Blue, Crystal City to Gallery Place on Yellow and Blue/Yellow, or Navy Yard to Gallery Place on Green.

Anyone boarding at one of those stations in the AM peak, like a bus rider who transfers at Pentagon to go to King Street, wouldn't pay. Likewise, the evening rush would only apply to those who board at one of the stations in question. You could also argue for removing Crystal City, since the trains aren't particularly crowded along the Yellow-Blue segment there.

However, this doesn't hit riders who ride all the way through the congested core. They are taking up room on the busiest trains as well, and it makes just as much sense to create an incentive for a Grosvenor-Suitland rider to avoid the busiest times as a Grosvenor-Dupont rider—more, actually.

2. Charge for trips to, from, or through core stations

Another approach, therefore, would be to apply the charge to riders whose trips pass through the busiest stations as well as start or end at them. Then it wouldn't be necessary to count some more outlying stations like Navy Yard, since anyone going there from or through downtown would pay anyway, and people riding there from Congress Heights aren't contributing to congestion.

This is the approach I recommended above, and would apply to 17 stations, from Dupont to Union Station on Red, Rosslyn to Capitol South on Orange/Blue, L'Enfant Plaza to Gallery Place on Yellow/Green, and Pentagon on Yellow or Blue.

This would catch the through riders. It would also catch some reverse commuters who aren't contributing to congestion, like anyone going outbound from Dupont Circle, Pentagon, Rosslyn, or Capitol South, where the trains in that direction aren't crowded at all.

3. Charge for trips through the busiest segments

A more complex alternative would be to look at the segments themselves; we have Matt Johnson's analysis of which segments have the highest crowding. We could charge only people who are riding on one of those busiest segments, in the direction(s) where they are busy but not in the reverse direction.

Based on that, we could apply it to Glenmont-bound Red Line trips from Cleveland Park to Gallery Place, Shady-Grove bound trips from Union Station to Farragut North, Orange or Blue Line eastbound trips from Virginia Square to Capitol South, westbound Orange/Blues from Eastern Market to Metro Center, or Yellow trips from Pentagon City to Gallery Place.

Any trip passing through one of those segments would get charged. If WMATA implements this charge by simply looking up the trip in a table of start and end points, it's technically easy; for any station pair, we can tell whether that trip does or does not pass through that area in the proper direction. However, it'd be confusing to explain to riders.

Therefore, the simplest yet fairest approach would probably be to designate the 17 stations I listed and charge for any trips to, from, or through those stations in the busiest time periods.

• Charge 50¢ for peak-of-the-peak trips (during the busiest 1½ hours in the morning and evening rush) instead of 10¢ as proposed by the General Manager. However, only apply this to rail trips that traverse the congested core. This would encourage peak riders in the busiest areas to ride when trains are a bit less crowded.

• Charge 25¢ for peak-of-the-peak trips on bus during the same time as the peak-of-the-peak rail surcharge. This would avoid riders to shifting from rail to bus and also spread peak bus ridership. To avoid double-charging customers who transfer, increase the transfer discount by 25¢ during the peak-of-the-peak time.

• Increase the differential between cash and SmarTrip on buses beyond the current and small 10¢. This would discourage cash payment and significantly speed boardings.

• Concentrate parking price increases at stations that get filled up and which have waiting lists for Monthly Reserved Parking, rather than simply raising charges across the board by 50¢ per day and $5 for reserved. For example, charge $1.15 more for lots with 100% or higher utilization, 50¢ for lots with utilization rates between 85 and 100%, and zero for less filled lots. Similarly, increase Monthly Reserved Parking by $10 where there are waiting lists, but zero elsewhere.

• Scale back the proposed $4 flat fare for Metrorail after midnight, which creates a strong disincentive for short-range late-night riders to take rail when cabs would be cheaper, especially for large groups. Instead, return to the Catoe proposal to charge the peak rush-hour fare at night.

The letter does commend some fare proposals, including creating a differential between paper farecards and SmarTrip on rail, reducing transfer periods to two hours, raising fares on airport buses, and MetroAccess fare changes.

With the revenue raised from these proposals, WMATA could reduce other fare hikes, including off-peak fares on rail and, especially, bus:

We believe that off-peak fare increases, for rail and bus, should be kept to a minimum, below the levels proposed by the General Manager, as the magnitude of the proposed increases would result in substantial ridership losses per dollar raised during times when there is excess system capacity. Substantially increasing off-peak bus fares, in particular, would hit some of Metro's poorest riders the hardest.

Bus fares are increasing by a greater percentage (20%) than rail fares (15%), and long-distance rail fares even less (11%). Like the parking, the bike locker rentals are similarly untargeted. At the very least, Metro should release bike locker utilization rates before and after any increase so we can determine whether that hike is beneficial or harmful.

Some have raised concerns that peak-of-the-peak bus fares would generate arguments between riders and bus drivers. The solution would be to set the SmarTrip-cash differential on bus higher than the peak-of-the-peak charge, so that cash users always pay the same flat rate, SmarTrip users get a discount, and if riding off-peak, SmarTrip users get an even larger discount.

All fare buckets have increased from the GM's recommendation, and the items I highlighted in yellow are the items added or modified in addition based on advocate suggestions: the peak-of-the-peak, a differential between SmarTrip and paper farecards on rail, and parking rate increases.

They did not evaluate a peak of the peak charge for bus or increasing the SmarTrip differential for bus, as some suggested.

The rightmost column gives the estimated ridership losses for each item. While WMATA has data on the ridership losses that come with small fare increases, once we get to items of this magnitude the ridership loss numbers become even more uncertain. I doubt that charging a differential for paper farecards would cost 4.1 million rides, though we really don't know.

At the same time, they don't list any ridership loss for increasing parking prices. At some lots which fill up, that might be true, but other lots don't fill up and we would expect some loss of ridership if those lots go up in price.

All three options also increase the MetroAccess fares for service beyond the required service area of ¾ miles from a transit station.

Here is the first chart, fare increases in the General Manager's proposed budget:

Here is the second, scaling up the fare increase to remove service cuts: