Transit expansion always involves some tricky fortune-telling about revenue and ridership. The More MARTA crystal ball, however, could use another shot of Windex, with billion-dollar differences of expert opinion on the budget and no new ridership numbers at all.
MARTA says a new financial model shows “no shortfall” in the roughly $1.956 billion first phase of the More MARTA expansion and improvement program. But the agency’s recently fired deputy general manager says it’s more like an $852 million deficit, plus a contingency so low it’s a “cause for alarm” and debt that could weigh down the second phase.
Then there’s the question of how many people will ride new lines like the Clifton Corridor. Old projections are out the window in the wake of a pandemic that left nothing certain except that work from home is here to stay. MARTA says it’s still working on new numbers, and the Atlanta Regional Commission (ARC) won’t have its own data until 2024.
More MARTA projects – all entirely within the City of Atlanta – are funded by a 0.5 percent sales tax approved by voters in 2016. Amid renewed public controversy in recent months over the lack of delivery, MARTA has rearranged 17 projects into two “tiers,” or phases. The agency estimates it can deliver the first-tier projects by 2028.
However, skepticism has bubbled in such institutions as Atlanta City Council, the Mayor’s Office, and sometimes MARTA’s own Board of Directors. After a public spat, MARTA agreed to a council-demanded audit of More MARTA, which is underway.
Among the triggers of the audit was commentary earlier this year from former MARTA Deputy General Manager Josh Rowan — whose mysterious firing “without cause” after a brief tenure was another source of trust issues. Rowan claimed that an internal financial model showed More MARTA with a revenue shortfall of more than $1 billion, and a Clayton County transit expansion program with a shortfall over $160 million.
MARTA called those numbers inaccurate in the sense that the model was still being developed by contractor HDR, but never denied they were produced by it at some point.
Following meetings last month about the new More MARTA phasing, MARTA confirmed to me that HDR’s financial modeling tool is complete and its findings used to create the list. So, what were the final shortfall numbers?
“There is not and never has been a funding deficit for the Clayton County program,” claimed MARTA spokesperson Stephany Fisher, citing a sales-tax escrow account. As for More MARTA calculations, she said, “The inputs to the model result in no revenue shortfall projection in that program, either.”
Well, with an asterisk: “The model assumes $500 million in federal funding that is not yet committed, but is a reasonable assumption, based on the history of the Capital Investment Grant program,” Fisher said. “… With that federal funding, the revenues cover the expenditures over 10 years and there is no shortfall. Beyond Tier 1, we estimate the program will be cash-flow positive again in 2035. Tier 2 planning and phasing will commence in 2028.”
She provided me with a spreadsheet of HDR’s More MARTA model, which included that $500 million (actually marked as “potential federal and state funds”). She also pointed me to a recent presentation about the Clayton numbers – which also rely on an assumption of about $402 million in not-yet-secured, “potential” federal funds.
I bounced these numbers off Rowan. “MARTA just told you they need $1 billion from the federal government,” he said.
Rowan called the MARTA model a “good start,” but took issue with how the numbers are interpreted and, in one case, allocated. He whipped up his own version of the spreadsheet to “declutter” it and focus on uncertainties.
“The analysis provided by MARTA for the first ‘phase’ of the More MARTA Atlanta program forecasts a $852 million funding deficit,” he said.
He had another term for the $500 million in “potential” federal or state funds the model assumes as helping to plug that hole. “At this point, these are imaginary funds,” he said. “This could be a reasonable projection, but it must be achieved.”
Rowan’s concerns ran deeper. He noted the other plug in that deficit is the issuance of debt, “which results in the second ‘phase’ of More MARTA Atlanta inheriting a $350 million debt in 2033,” he said.
That’s if the hole doesn’t get bigger. “MARTA is showing a $6 million contingency for this first ‘phase,’” said Rowan. “This is extremely low for this early stage of project delivery and likely will not hold up to cost escalation over the next decade.”
He calculated that a 2 percent escalation in the overall program would increase its costs by $30 million. Make that 2 percent annually and it’s another $300 million.
“The next phase will inherit a lot of debt,” said Rowan. “… The low contingency is cause for alarm.”
I returned to MARTA with Rowan’s spreadsheet and commentary. “The HDR model was prepared by experts with decades of international experience in the transportation industry,” replied Fisher. “Although we are approaching the program in phases, the program is at least a 40-year program. Each project represented in the model includes cost escalation and contingency. The HDR model is a snapshot in time, but we stand behind it as it represents where we are today.”
She said the model will be updated annually, along with the agency budget. Rowan said it should be updated and published quarterly.
“I would encourage everyone to monitor two high risk items over the duration of this first ‘phase’ – MARTA’s ability to secure additional federal/state funding and MARTA’s ability to control project cost,” Rowan emphasized. “These items represent significant risk to the success of the first ‘phase.’”
Of course, the whole point of whatever gets built is to serve current riders and entice many new ones. But how many, where, and for what purposes? MARTA, like every transit and transportation agency, saw its existing projections go out the window in the wake of the pandemic, whose societal changes – especially around working from home – appear here to stay.
Fisher says MARTA is still working on updated projections. ARC, a regional planning organization that provides some of MARTA’s data, is still working as well. ARC spokesperson Paul Donsky said one thing they know is a “dramatic increase in people working from home” between 2019 and 2021, as shown in the latest U.S. Census data.
“We don’t have modeling data that shows any forecast for MARTA ridership that takes into account the pandemic, but we are working on it,” said Donsky. ARC hopes to have that in early 2024 for an update to the Metropolitan Transportation Plan. Next year, ARC also plans to update its household travel survey, “which should help provide much greater clarity about WFH [work from home] and transit trends,” he said.
Ridership projections can inform not just where transit goes, but what mode it uses – a hot debate in some More MARTA projects like the Clifton Corridor and Campbellton Road, as the transit agency increasingly leans towards “bus rapid transit” (BRT) rather than originally proposed rail. MARTA analyses often emphasize BRT as less expensive to build and operate than rail. But rail is often more efficient at handling increased ridership, as cars can be added to existing trains rather than buying entirely new vehicles and hiring more drivers.
“That’s one measure of flexibility,” acknowledged Fisher. She said BRT is more flexible in other ways, like driving around route blockages, and indicated MARTA’s existing, outdated projections already anticipated ridership increases.