In a recent article for Twin Cities Business, Adam Platt pours some cold water on all the hype over the new Borealis service, claiming that the service has not drummed up significant new ridership. He claims that old equipment and shoddy amenities have failed to draw in new riders, and that the vast majority of its ridership is crossover from existing Hiawatha and Empire Builder passengers.
Although I’ll admit the hype has gotten a little out of hand, I fear his article overcorrects and his bottom line is based on faulty logic. As such, I feel it necessary to clear the air.
At worst, the data is inconclusive. Considering ridership fluctuates over a yearly cycle, 40 days simply isn’t enough to go off of.
That said, the initial results are promising, not only surpassing initial estimates but also out-performing similar trains. Here’s why.
Some History
Mr. Platt first sells Amtrak short when talking about its history serving the MSP-Chicago corridor. He says the corridor only had dedicated service from 1978-1981, but Amtrak also operated the North Coast and Twin Cities Hiawathas over this corridor, with the former continuing on to Seattle by way of southern Montana.
Service between Chicago and MSP fluctuated from 1.5 to 3 trains per day, depending on the exact time. The North Coast Hiawatha ran daily during the summer months and three times a week the rest of the year, while the Twin Cities Hiawatha alternated between four times a week (covering the off days of the NCH) and daily, although sometimes disappearing altogether.
This arrangement started in late 1971 and lasted until 1979, when the NCH was axed. A year prior, the TCH was merged with the St Paul-Duluth Arrowhead to become the North Star, the latter two of which were state-supported rather than funded by Amtrak, as Mr. Platt claims. The North Star would later be truncated back to St. Paul. It ceased operations altogether in 1985.
What Is Borealis?
This is a matter of semantics, but I would argue that Borealis is a new Amtrak service. It requires a crew change at Milwaukee (or at least did last time I rode it) as Hiawatha crews aren’t qualified for the rest of the route. The existing Hiawatha trip had its numbers changed to prevent confusion, and it operates with equipment not seen on the Hiawatha, namely the cafe/business-class car.
Ridership
Yes, these numbers are still early, but at 648 riders per day (numbers have been revised upward since he published his article) the route is no slouch, putting it on track to surpass even the highest ridership estimate of 235,000 riders per year. Considering each consist Borealis train only has 314 seats, trains may actually be capacity constrained (more on that later). Also, compared with similar routes, it performs well. The two other similar trains are the Pennsylvanian and combined Missouri River Runner/Lincoln Service trains 318 and 319.
The Pennsylvanian is a daily train that runs 444 miles from New York to Pittsburgh via Philadelphia and Harrisburg, Pennsylvania completing the trip in about 9 hours and 15 minutes. It links the first, sixth and 26th most populous metro areas in the country and shares most of its route with both Amtrak and commuter rail services, particularly east of Harrisburg. It got 196,728 riders last year, and pre-COVID would hover in the 200,000 to 230,000 range.
Despite serving fewer people, the Borealis is set to match the Pennsylvanian’s pre-COVID ridership, and even a 17% drop would put it on par with the Pennsylvanian’s current ridership.
The Missouri River Runner/Lincoln Service trains 318/319 operate between Chicago and Kansas City via St Louis. The route is 567 miles long, serves roughly the same amount of people – about 13 million – as Borealis, and also has overlapping service in its entirety. Journey time is about 11 hours. Ridership data is again lacking as Amtrak only combined these trains in May of 2022, but a best guess based on 2023 ridership is about 210,000 riders per year, slightly below Borealis’s current trajectory.
The Milwaukee factor
There is a large amount of turnover in Milwaukee, but if we take Mr. Platt’s high-end estimate of 47% and subtract that portion entirely from the equation, the Borealis is still looking to get about 125,400 riders per year. This roughly matches Amtrak’s initial assessment as well as the Minnesota Department of Transportation and All Aboard Minnesota’s revised modeling (117,000 and 125,000, respectively), but their models still allowed for travel over the Milwaukee-Chicago segment. Theirs also added an additional frequency instead of extending an existing one. So even if ridership is artificially inflated, Borealis is still performing well.
Comparisons to the two trains mentioned earlier break down at this point. Minneapolis-St. Paul and Milwaukee combined only have about 5 million people, much smaller than the catchment areas on the other two trains.
Trains 311/316, the Missouri River Runner trips that don’t continue to Chicago, are the best comparison. The 283-mile route between Kansas City and St. Louis gets roughly 77,000 riders per year.
A less apt comparison would be to the Heartland Flyer between Fort Worth and Oklahoma City, but that train serves twice the population and only gets 73,000 riders on its 206-mile route, and so we have yet to find the elusive metric where Borealis is doing worse than expected.
These comparisons may even be moot. About 38.5% of Borealis rides were entirely within the Hiawatha corridor in May, according to the Wisconsin Department of Transportation. I asked WISDOT for an updated number for June, and they say it dropped to 34.5%. I’d expect this number to drop further as people adjust their travel habits to take cheaper and less crowded Hiawatha trips.
What portion of the remaining 65.5% is crossover from the Empire Builder probably won’t be known for another year or two. Anecdotally, podcast host Ian Buck and I both talked to about a dozen people on the first day who didn’t even know the service was new. They just saw a train that fit their schedule and price point and booked it. As any expert in any field should know, the average person knows much less about their area of expertise than one would think.
Considering both trains are still nearly sold out on most days, I’d wager many more riders are people discovering Borealis for the first time rather than regular users switching from one to the other. Prices remain elevated into October, suggesting some trains are already about 40% sold even three months out.
Plane slayer
The biggest issue with Mr. Platt’s back-of-the-napkin math about how much plane travel Borealis might replace is that someone’s already gone through the effort of compiling the patronage of the various flights on this corridor to compare it against – thanks Edwin! This is where the 15.5% number comes from.
I doubt Edwin has picked up on every little puddle jumper between small towns, and some of the Amtrak numbers are fudged, but I would be surprised if this number varies by more than a few percent. While not the best performer, the Chicago-MSP corridor is still above average for its journey time, not just for North American trains but worldwide.
Profitability
Yes, Amtrak includes subsidy in its revenue, artificially inflating its operating ratio, but that’s standard practice. Borealis isn’t unique in this regard. Looking at 2018, state-supported services as a whole recovered nearly 90% of their operating expenses from “revenue”, with some routes “earning” over 50% more than they spent! Amtrak’s accounting in general is a bit suspect, but again, it’s nothing unique to the Borealis.
Prospects for a third train
Contrary to Mr. Platt’s claims, work on a third train between MSP and Chicago (i.e., a second Borealis) should begin soon, if it hasn’t already.
WISDOT began seeking consultants for this and three other projects back in May, using funds from the Corridor ID program. Furthermore, a deal was recently reached with the Canadian Pacific Kansas City railway that expedites the process for said train. This basically mirrors the deal for the first train, including extending a Hiawatha round trip and allowing startup before construction starts.
If everything goes well, a second round trip could easily start in the next two to four years. This is lightning speed by Amtrak standards! It is strange that Amtrak’s own spokesperson wasn’t aware of said agreement, so I don’t blame him for this one. On the other hand, MnDOT’s statement that “it would be premature for us to discuss adding a third train at this time,” seems more noncommittal than actively cynical.
Bottom line
I understand trying to temper people’s expectations about a literal hype train, but I fear Mr. Platt has overcorrected. Even after butchering the existing data into a worse-than-worst-case scenario, where there’s only about 100 or so new riders per direction per day on the new segment of the Borealis, the train is still performing as or better than expected.
Even if things were so dire, the correct solution isn’t just to wait for new riders to show up. It’s to find ways to improve service and attract new riders. This could be in the form of new or faster trips, or smaller improvements like better Wi-Fi and more bus connections to Madison and other nearby cities, like Mr. Platt suggests.
What About Borealis Could Amtrak Improve?
Speaking of possible improvements, let’s talk about some I’ve seen in this article and elsewhere online.
Newer equipment
For those who don’t know, the Borealis operates with some of Amtrak’s older equipment, mainly Horizon and Amfleet coaches, as opposed to the new Venture coaches seen on other Midwest routes.
My hot take is that I’d rather have the older equipment on this route. The new seats aren’t the best, and the five hours from Chicago to St. Louis was about my limit. My friend Joel, who was with me on that trip, seemed to hold the same opinion. The old coaches have different seats with much thicker padding.
That said, I haven’t tried the new business class yet.
Better WiFi
West of Milwaukee, at least, the WiFi on Borealis is painfully slow. I only ever managed to pull about 0.5 megabytes per second and since it’s cellular, it tends to drop out when switching towers or where the signal is obscured. There’s really no reason to use it over my own data plan.
Checked baggage
Amtrak stated this will come after the long-distance trains get theirs back, but I would like to take a moment to re-up this request. Right now, the only train on this corridor that allows bikes is the Empire Builder and those slots sell out quickly. I’d love to be able to bring my bike to some of these smaller towns to help me explore.
Longer trains
As I mentioned before, the Pennsylvanian is running with longer trains (and checked baggage) despite garnering fewer riders. Lower capacity on the Borealis means higher prices; higher prices mean fewer bookings. While Borealis trips aren’t quite selling out anymore, we’re seeing increased prices out to October, suggesting those trips are already about 40% sold. Sure, changing the pricing regime would do the same thing, but all those new riders will need seats. I know, I know, equipment shortage, but surely Amtrak has two extra coaches somewhere.
Additional stops
I’ve heard some people clamoring for express and sleeper trips already, but I think we’re a long way from either of those. We’d exclude more potential riders than we’d gain with these options.
If anything, we should add infill stops to maximize the number of new riders. Places like Cottage Grove, Hastings, Lake City and Wabasha in Minnesota and Watertown, Pewaukee and Truesdell in Wisconsin would all fill significant gaps on the current route. Since much of the line tops out at 65 miles per hour, the time it takes to make these stops can easily be made up with track upgrades.
One Step Beyond
I could go on all day about ways to improve service, like adding a dining car, extending the train to Minneapolis or, gosh, wouldn’t it be nice if we bought those Talgos that were supposed to go to Wisconsin? But then I’d probably end up fueling the rumor mill that both Mr. Platt and I are trying to get under control. Only time will tell what and when improvements will come, but the initial results for Borealis are quite promising.