FY 2025 Ridership Expansion and Fiscal Recovery

[NYP]

No- riders that travel any miles on a state-supported route are assigned to the state-supported route, not the NEC. A rider from New York to Lancaster would be counted as a Keystone rider.

 

[Septa9739]

This is correct. But if I understand properly, the same Keystone train carried a passenger NYP-PHL, it gets counted as an NER. If a passenger boards a regional in Charlottesville and gets off at WAS, it counts as the Roanoke service line. If he boards at Charlottesville bound for New York, it credits credited to the Roanoke service line. If he boards at Washington, however, he gets counted as an NER. That seems to be the only way the numbers make sense historically. Otherwise we would see dips in the NER every time a train gets extended into VA, which is not the case.

 

[Tom C]

The Vermonter has a rather complicated one. If you are on the Vermonter you may be assigned to 3 different routes depending on your northern terminus. If the northern terminus of your trip is New Haven or somewhere south of that on the NEC it’s a Northeast Regional trip, if the northern terminus is between Meriden and Springfield it’s a New Haven - Springfield line trip, if the northern terminus is north of Springfield it is a Vermonter trip.

 

[Septa9739]

After I posted last month’s update, I realized I didn’t like the way I formatted it. Winners and losers is too complicated, so I came up with a different program. Red, yellow, green, and pink. Red is below last year and pre-COVID. Yellow is above pre-COVID. Green is above both. I needed a color to reflect above last year, but still below pre-COVID. That is going to be pink, because I think that situation is more of a red than anything. This is NOVEMBER FY 2025 (CY 2024) Ridership. Percentages are of last year.

Green:

Northeast Regional 979,100 (108%)

1. Empire South 112,600 (105%)
2. Borealis 19,400 (first November ever!)
3. Illini/Saluki 28,300 (110%)
4. Cascades 81,600 (129%)
5. Missouri River Runner 15,400 (102%)
6. Pennsylvanian 20,900 (114%)
7. Piedmont 41,100 (121%)
8. Cardinal 8,700 (112%)
9. City of New Orleans 21,500 (108%)
10. Texas Eagle 31,800 (112%)
11. Floridian 33,400 (First November since Carter was president)
12. Palmetto 34,100 (102%)

Yellow

1. Heartland Flyer 6,600 (97%)
2. All FOUR Virginia services 112,100 (95%)
3. Carolinian 25,600 (80%)
4. Lakeshore Limited 30,000 (95%)
5. Crescent 26,900 (92%)
6. Auto Train 19,400 (97%)

Pink

1. Lincoln Service 52,700 (104%)
2. Illinois Service 12,800 (108%)
3. PacSurf 189,300 (103%)
4. Capitol Corridor 93,900 (102%)
5. Adirondack 6,700 (100%)
6. Silver Meteor 28,100 (108%)
7. CA Zephyr 27,400 (105%)
8. SWC 22,600 (101%)
9. Sunset Limited 7,200 (106%)

Red

1. Acela Express 261,000 (87%)
2. Ethan Allen Express 2,700 (36%)
3. Vermonter 5,900 (65%)
4. Downeaster 46,200 (96%)
5. Keystone 102,600 (88%)
6. Hiawatha 53,000 (91%)
7. Wolverine 35,700 (99%)
8. San Joaquin 81,000 (96%)
9. Blue Water 13,800 (97%)
10. Pere Marquette 7,200 (96%)
11. Silver Star 10,600 (31%)
12. Empire Builder 29,000 (96%)
13. Capitol Limited 4,000 (34%)
14. Coast Starlight 31,100 (96%)

 

[GariYaMoshi]

Ethan Allen Express 2,700 (36%)

Was Ethan Allen out of service for a while?

Full disclosure here - I once taught college level statistics. I'm thinking that Capitol Limited and Silver Star both deserve an asterisk given that they weren't in service the full month, losing their identities when combined for the Floridian.

 

[Septa9739]

As for commentary, the month looks pretty good on the whole. Big picture wise the railroad lost $45,100,000 over the last two months, which is dreadful, but half as dreadful as last year. Controlling operating losses may be key for Amtrak over the next length of time. Improvement was largely on the expenses side of the ledger. Revenue is soft compared to plan in line with ridership. YTD ridership is up over last year by 224,000. Ridership is up over all three segments, but 2 in three of them (on net) are NECSL passengers. Both the NECSL and SSSL are narrowly behind plan, LDSL is in line, but tilted ahead.

At the individual route level, the greens contain no surprises, expect for maybe the Cardinal, hitting first month of COVID recovery since February. Seven months of data in and the Borealis seems to be on track for 230,000 (ish) riders its first 12 months. We’ll see what the winter holds, but ridership has been very tight in this 18,000-22,000 range since June, most routes see bigger swings. The Cascades up 129% is about what we want to see going from 4 round trips to six.

The yellows aren’t terribly exciting. They are all within the striking distance. I think the only news story is that Virginia may just have a ceiling after all. Nothing can be straight growth forever.

The pinks are a little more interesting. The Lincoln service is right there to break through it’s COVID dip. It’s still a long way from its 2013 record. FY 2026 might see the record. All the California services are going to take years to rebuild. None of them are close. None are growing with any real speed. The Zephyr and Chief are getting close. The Adirondack just deserves a medal for surviving. Brutal schedule, long term track work, state non-commitment, and historically soft ridership usually spell the end (looking at the Hoosier State).

Reds are mostly easy to explain. Acela=pulled equipment+soft business demand+failing reliablity. At this point, the only reason I’d ride an Acela is because I never have. Otherwise, there’s no guarantee you’ll actually save the time advertised versus the NER. The Vermont trains collapsed out of nowhere and I don’t know why, but would like to. The Downeaster is a matter of a couple riders per train. The rest of the SSSL trains are repeat offenders. The Empire Builder is right there, and the Starlight is still like half the train it was in 2019. The Star and Capitol Limited only ran 10 days, which brings me to what everyone wants to know about.

 

[Bernie]

Interesting that the Keystone is down, but the Pennsylvanian is (way) up. Keystone is probably down because of track work requiring busing from Lancaster

 

[Septa9739]

That and also I think there is probably growth in the Western side with the change at Pittsburgh to Chicago more possible. Because the Pennsylvanian is a one-a-day with fairly low ridership in a absolute sense, it doesn’t take a whole lot to tilt the needle.

 

[John Bredin]

I'm curious about the Midwest routes.

*Illini/Saluki is green while Lincoln Service and Illinois Service is pink. Is there better recovery in demand for going to UofI (Champaign) and SIU (Carbondale) but not ISU (Normal) or WIU (Macomb)? Or fewer people going to Springfield on government business or lobbying since COVID?

*Missouri River Runner is green while Lincoln Service is pink. Similarly, Borealis is green while Hiawatha is red. Those just seem counterintuitive to me.

If Lincoln Service is approaching pre-lockdown ridership as you commented, that may be part of the explanation.

*All Michigan services are red. What's the deal there? My instinct when I see the Michigan schedules is that there aren't enough trains; there should be 2 Pere Marquettes and 2 Blue Waters to allow bidirectional travel, and definitely more than 3 Wolverines with Ann Arbor and the sheer size of metropolitan Detroit. But the ridership doesn't seem to be there for the service as-is.


And a non-Midwest question: what were the numbers for Empire West and Maple Leaf?

 

[Acela150]

Interesting that the Keystone is down, but the Pennsylvanian is (way) up. Keystone is probably down because of track work requiring busing from Lancaster

This is the best answer for the conundrum.

 

[thefirewarde]

I'm curious about the Midwest routes.

*Michigan services are red. What's the deal there? My instinct when I see the Michigan schedules is that there aren't enough trains; there should be 2 Pere Marquettes and 2 Blue Waters to allow bidirectional travel, and definitely more than 3 Wolverines with Ann Arbor and the sheer size of metropolitan Detroit. But the ridership doesn't seem to be there for the service as-is.

The Wolverine was hurt by the extended reduction to 2 daily trains for track improvements, and hasn't recovered quickly since. I believe Michigan wants to add Windsor and additional frequencies - the Windsor/Detroit to Chicago corridor can definitely handle the same frequency as Raleigh to Charlotte - but getting the momentum back until those additional frequencies and the equipment to run them materialize may be quite tricky.

We've seen that even relatively substantial speed increases in the Midwest don't make massive ridership gains.

 

[ryanwc]

Lincoln service has been hurt by state work-from-home and the movement of some of the workforce to Chicago. Hiawatha was basically a long commuter hurt by WFH and by a big increase in monthly pass prices. SIU enrollment has rebounded, helping the Illini/Saluki.

I think a year of full service will pull the Wolverines back to par.

 

[Septa9739]

Meant to post this with the rest of the stuff, but I got distracted by reality. This is what I think is extremely interesting.

The Floridian: It’s really to early to tell, but my read is it’s working. It carried 33,400 riders in 20 days. That means we can guess at about 50,000/month. That leads us to fact one. It will be the busiest LDSL train in the system by a mile, replacing the Lakeshore Limited. This makes sense because the Star was a close second. This is a little huge. A “new” long distance train is going to report 500,000+ passengers for the first time since 2013. If it reports 600,000, it might be a 21st Century continental record. That’s remarkable PR, especially at a time when knives are out for funding.

Fact Two: It put up a respectable .52 operating margin. That’s huge because the Capitol was a bit of a bruiser on the money side, even when it was properly equipped. 50% has been a respectable ratio as long as I’ve looked at these numbers. If the third sleeper comes in March, we’ll see further improvement. 60% would be huge. Not many trains have crossed that lately. In 41 days of running, the Capitol and Star cost $16,200,000 to run. For 20 days, the Floridian was $6,100,000. This represents a roughly 25% reduction.

Fact Three: It added riders. You can dispute my methodology here. This month the Capitol+Star+Floridian carried 48,000 riders. Last year it carried 46,100. In CY 2019, it carried 46,400 riders. There seem to be 1,900 new riders in 20 days (now yes some of this is unmet demand over the Capitol’s routing). That’s 85/day. That isn’t an insignificant move. Looking at last month also provides some insight. Last month Capitol+Star was 42,000, down from 42,500 the year prior, down from 43,500 in 2019. In 2019, the pair gained 2,600 riders from October to November. Last year it gained 3,600. This year the triplet gained 6,000, or 200/day, which suggests maybe 2,400 more than might be expected. Like I said this data is very new, but something’s happening here.

Fact Four: It takes a mostly empty train out of the congested trans-Hudson tunnels. In 2022 (with the whole Star/Meteor thing still ongoing), the Star carried about 121 passengers per frequency to/from New York with WAS-NYP being the second busiest station pair, so most of those NYP passengers were on 91, with 125 on its markers the whole way. The Cardinal is the same way (and there’s plenty good reason that shouldn’t operate over the NEC either) and so is the Crescent (but the reports indicate strong demand from other NEC cities to the South). The Meteor handled a larger portion of its traffic in NewYork relative to its total ridership and is the preferred train from the NEC to Orlando, with most on the Star motivated by fare difference, so it makes sense to keep the Meteor into NYP over the Star.

The way I interpret this data: The Floridian has, by early indicators, increased ridership and reduced costs. It has improved utilization massively, facilitated the return of traditional dining service to about 15,000 riders a month (few of whom use it, but still), and improved capacity over strained sections of the WAS-CHI route. It seems to have been an excellent business move, and, at this juncture, I would hope Amtrak leave this arrangement for the foreseeable future.

A note about headwinds: There seem to be two major concerns re the long term operation of the Floridian, timekeeping and sleeper capacity. Concerning timekeeping, in the first 20 days Norfolk Southern kept its delays to 667 minutes per 10,000 miles over its 518 miles. That’s enviable for some routes. CSX reported a less stellar 1194 minutes per 10,000 miles over its 1413 miles, but that figure is certainly in the realm of workable. And this includes that many of the delays are Amtrak responsible. There have been hours long mechanical failures, that have snowballed into massive delays that cause the next day’s run to be delayed on origination. If these issues get resolved and the train becomes more able to hold a slot, these numbers will likely improve. As for the sleeper situation the Capitol had two 5-14-1-1 sleepers, for 10-28-2-2. These were replaced by two 2-11-1 sleepers, for 4-22-2. This is a serious capacity reduction. 42 rooms have been replaced by 28. The third sleeper coming in March (we hope) will raise totals to 6-33-3, or 42 rooms, restoring capacity, but offering fewer high yield Bedrooms.

 

[NES28]

We've seen that even relatively substantial speed increases in the Midwest don't make massive ridership gains.

I think that frequency increases are more important than increasing speed from 79 to 110 mph.

 

[BlackDiamond]

act Three: It added riders. You can dispute my methodology here. This month the Capitol+Star+Floridian carried 48,000 riders. Last year it carried 46,100. In CY 2019, it carried 46,400 riders. There seem to be 1,900 new riders in 20 days (now yes some of this is unmet demand over the Capitol’s routing). That’s 85/day. That isn’t an insignificant move. Looking at last month also provides some insight. Last month Capitol+Star was 42,000, down from 42,500 the year prior, down from 43,500 in 2019. In 2019, the pair gained 2,600 riders from October to November. Last year it gained 3,600. This year the triplet gained 6,000, or 200/day, which suggests maybe 2,400 more than might be expected. Like I said this data is very new, but something’s happening here.

It would be interesting to know how many riders are taking advantage of the through routing, say Pittsburgh - Orlando or Chicago - Richmond VA for example.

 

[railiner]

It would be interesting to know how many riders are taking advantage of the through routing, say Pittsburgh - Orlando or Chicago - Richmond VA for example.

It would help save a segment for those on a Railpass, but perhaps add a segment for others (between NEC and Tampa)