Expect Beltline rail to exceed $2.8 billion
ABI cleared trees earlier this year on the Southside Beltline. It’s among many costs that aren’t usually counted as Beltline rail expenses. Photo by Walter Brown.
The price tag is likely to be Beltline rail’s undoing. But you’re not alone if you have no idea what that price tag might be.
“We do not have an estimate for building rail on the entire Beltline as there are too many variables to consider at this time,” a MARTA spokeswoman told us.
Fair enough. But Atlanta residents and policy makers still ought to have some idea of what we’re getting into.
So we decided to calculate a reasonable estimate. Our aim wasn’t to detail expense categories or to come up with a rough budget. It was simply to use a transparent process to land on a conservative, credible number .
The segmented nature of the ambitious project masks its full cost. Officials and journalists talk about the Streetcar Extension East (which includes the first 1.3 miles of Beltline rail) as a $230 million project. But Beltline rail supporters view the Extension as a downpayment on 22.8-mile project. They base their arguments for the rail line on benefits they’re convinced the city will reap once the entire loop is completed.
It’s as if we should expect all the (purported) benefits of the entire 22.8-mile line, while only paying for 1.3 miles of it.
At the same time, transit agencies and their contractors are notorious for underestimating the cost of big capital projects. One of many studies on the topic found that actual costs typically exceed estimates by nearly a third. And these weren’t early guesses — they were cost projections relying on contractors’ estimates well into the design phase. In general, the earlier in the process the projection, the more extreme the lowballing.
MARTA’s track record appears to be on the high side of those benchmarks. Summerhill bus rapid transit, a project that Better Atlanta Transit supports, is a good example. Initially budgeted at $58 million (in 2024 dollars), it was revised upward in 2022 to $91 million. And the BRT line only recently began construction. Granted, Summerhill BRT’s budget revisions were influenced by post-Covid inflation. But that doesn’t account for anywhere near a 57 percent increase.
We arrived at our Beltline rail estimate via an objective, real-world measure: We compared the proposed loop to the actual cost of similar projects in recent years.
The Eno Center for Transportation’s Transit Cost Database uses authoritative spending data on transit projects. It adjusts for inflation to 2020 dollars. It’s the most complete and carefully curated database of transit projects that we could access.
Relying on it presents a few challenges. The database doesn’t have a “streetcar” category per se, but some modern streetcar projects do appear under the “light rail” heading. The dividing line between modern streetcars and light rail has become muddled anyway, and nowhere more than in Atlanta. Beltline rail is being planned as a hybrid that leans closer to light rail. Both Atlanta Beltline Inc. and Atlanta Regional Commission planning documents refer to the project as ‘light rail.” But it will also be an extension of the Atlanta Streetcar, which does run along streets and uses slightly shorter vehicles than light rail typically has.
Light rail generally ends up being more expensive than streetcars. Greater speeds and longer distances, require communication systems to be more complex. Stops may be more elaborate and boarding platforms longer. There also may be a need for more electric substations to provide additional power. To reflect the potential lower expenses for these items on the Atlanta project, we discount Beltline rail’s per-mile estimate by 10 percent from the actual costs of our comparables.
But the main cost differential between streetcars and light rail is real estate. Streetcars usually run on streets, while light rail typically requires right-of-way acquisition. In that key respect, Beltline rail is a light-rail project. It requires a 35-to-40-foot-wide right-of-way purchased with public money rather than repurposed from car lanes.
The Streetcar Extension East only includes about one-seventeenth of the planned Beltline rail loop. Map credit: Atlanta Beltline Inc.
Because land acquisition differs so much from project to project and is budgeted separately, the Eno Center exempts real estate expenses from its database. So our conservative projection doesn’t include real estate expenses for Beltline rail. The most recent Atlanta Beltline Inc. Annual Report notes that 38 percent of Beltline rail’s right of way remains to be purchased.
Eno also excludes early site preparation from its data where possible, because some sitework for many projects is done prior to actual design and construction, and often by a different agency. Beltline rail is a good example. ABI has been clearing trees, performing cleanups, building retaining walls and otherwise prepping the transit right of way for well over a decade.
We want more transit. We love the Beltline. Check out how Beltline rail would damage both.
The same goes for interest expenses, which would be substantial in the case of Beltline rail. But again, the Eno database exempts financing because it can differ so widely from project to project.
Finally, rolling stock is excluded from the Eno Database. It typically falls outside of the construction budget.
The existing Atlanta Streetcar, which will be integrated with the Beltline streetcar, runs a streetcar version of the Siemens S70 light-rail vehicle. It’s slightly shorter than the light-rail version and isn’t required to go as fast. Before cancelling its own modern streetcar line, Broward County, Fla., was set to spend $31.4 million on five similarly outfitted S70s six years ago — and that was before Covid inflation. If MARTA needed 25 vehicles to operate Beltline rail (and it probably will need more), rolling stock alone would add more than $150 million to the cost in 2017 dollars.
So it’s important to remember that major cost categories – land, financing, vehicles and much of the site prep – aren’t even included in our already conservative estimate.
With those caveats, we set out to find comparable projects. We limited our search to light-rail projects in the United States completed 2005-2022. We struck projects from the list that required any tunneling at all or elevated tracks totalling more than 3 percent of the project’s length. The Beltline streetcar isn’t expected to rely on tunnels, but it will require new bridges and stretches of elevated tracks. (If you wish, follow along on our Beltline Rail Cost Estimate Worksheet.)
That left us with a cohort of 18 projects. Then, to prevent outliers from pulling our numbers up or down, we excluded the two projects with the highest per-mile costs and the two with the lowest per-mile costs.
From there, we came up with an average cost per mile. The mean cost per mile of the 14 remaining comparable projects came to $80.5 million in 2020 dollars. Our 10 percent discount for Beltline rail brought that down to $72.5.
By that measure, a rough estimate for the entire 22.8-mile loop would come to $1.65 billion (aside from land, financing, vehicles and site prep). But that’s before inflation.
To account for inflation, we first adjust our dollars from 2020 to 2026, the year construction is slated to begin on the first leg of Beltline rail. We relied on Ed Zarenski’s authoritative Construction Analytics inflation index. Unfortunately, construction costs for “non-building” construction spiked as we came out of the pandemic. In 2022 alone, infrastructure pricing rose by more than 17 percent. So our rough estimate of $1.65 billion in 2020 rises by 2026 to $2.47 billion.
Next, we had to adjust for inflation during the construction period. For the purposes of this exercise, we assumed that the entire loop would be completed in 15 years. That’s optimistic. Generally, costs rise the longer a project takes to complete. But we settled on 15 years in the interest, again, of making our estimate as conservative as possible.
So we spread the spending evenly across 15 years and assumed an historically low annual inflation rate of 2 percent annually rolling forward. Even such a low inflation rate has a big impact over a decade and a half. It drove our final estimate for actual spending to $2.85 billion.
The full cost of Beltline rail should be expected to come in significantly higher than. Land acquisition, financing, rolling and early site prep would require hundreds of millions of dollars. Much of that has already been spent on land acquisition, funded by Invest Atlanta (the city’s development agency) through the Beltline districts Tax Allocation District.
We hope to get a better handle on those numbers at a later date. If we do, we’ll add them into our cost worksheet. For now, however, it’s safe to characterize the full cost of Beltline rail as well over $2.8 billion.
For a more complete explanation of our methodology and math, please see our Beltline Rail Cost Estimate Worksheet.
