Wednesday, July 22, 2026

High Speed Rail Update: the alignment in northwest Harris County is (unofficially) DEAD, but could Terafab make a Robotaxi project possible?

Updated August 6 with new information about Terafab, and some speculation about another possible outcome for the corridor (see Terafab section). Additional minor update in the Terafab section on August 31.
Another excellent guest post from Oscar Slotboom, author of Houston Freeways. Honestly, I’m glad Texas HSR is dead because it was always going to end up a taxpayer boondoggle one way or another (eventual private bankruptcy and government rescue once partially or fully built). Here's a 2022 post on why the economics never made sense.
---
With the recent demolition of Northwest Mall, it's a good time for an update about the proposed high speed rail project between Houston and Dallas. Most of the Northwest Mall property is owned by an affiliate of Texas High-Speed Rail Holdings LLC (THSR), formerly called Texas Central, and is slated for the Houston station. In news reports (1, 2) a project representative was quoted saying the demolition "will allow the project to proceed as soon as we get the green light." At the minimum this statement is disingenuous, and most likely is a complete lie.
The findings of my investigation into the Harris County alignment lead to only one conclusion: the northwest Harris County alignment is dead. Multiple subdivisions have been built in the path of the approved alignment, and a new Blinn College campus with a large first building underway has been established directly in the alignment path. There is no way the project can be built on the officially approved alignment.
This doesn't necessarily mean the overall project is dead. It means there would need to be a supplemental environmental study to define a new alignment. With rapid development in northwest Harris County, it will be difficult to identify a new alignment. And looming over the alignment problem is the outrageous cost, which is now around $40 billion.
However, the situation is different in North Texas. Political leadership in the NCTCOG Regional Transportation Council believes the Houston-to-Dallas project will (eventually) be built. The council has spent $10.9 million since 2020 and ignited political strife with the City of Dallas attempting to obtain environmental clearance for extending the rail line from Dallas to Fort Worth.
The Planned Alignment in Houston
The Final Environmental Impact Statement (FEIS) for the project was released on May 29, 2020. It defines the officially approved alignment for the entire corridor. You can download it here. The detailed map showing the corridor in Harris County is appendix G-3, and the footprint map for the southern segment including Harris County is volume 5. You can also see alignment maps on the web site, although they don't match the FEIS alignment in far northwest Harris County (more below in the Binford Creek section).
This analysis is mainly focused on land development which has occurred in the path of the FEIS alignment since the release of the FEIS. Generally speaking, as soon as an FEIS is approved, corridor preservation and right-of-way acquisition should proceed at full speed, especially when development is occurring in the path of the alignment. As we'll see, the corridor has not been preserved except in two spots – Sanford Farms and part of The Grand Prairie.
In Houston, the FEIS alignment starts at the Northwest Mall site. Demolition of Northwest Mall took place in April and May (news reports 1, 2). As of July 21, there were large piles of leftover aggregate with processing and recycling apparently being done on site.
According to THSR's statement, the demolition is ostensibly for the rail project. This helps maintain a public perception that the project is alive and viable. As a practical matter, the mall needed to be demolished regardless of the future of the high speed rail project.
There is a complication with the Northwest Mall site. Most (45.25 acres) of the mall site is owned by "Cadiz Development Houston LLC", which is an affiliate of THSR created for station development. But the 10-acre north parking lot, 9500 Hempstead Road on HCAD, is owned by "NW JCP LTD" since January 2009, which predates THSR. As of July 21 there is a "Land For Sale" sign on the property, which strongly suggests the current owner does not have a partnership with THSR, or is trying to maximize the price of the property. Not having this tract is inconvenient for comprehensive site redevelopment, for example a sports arena. The FEIS footprint map (page 101) shows the entire site as part of the project, so THSR should be able to use eminent domain to acquire it.
Northwest Mall on April 25, 2026
Northwest Mall on May 15, 2026
10 acres on the north side of the site is not owned by TSHR and is for sale. Photo date 21-July-2026.
Leaving the mall site, the FEIS alignment proceeds northwest on an elevated structure above Hempstead Road. At West Little York, the alignment moves to the south side of the Union Pacific tracks and drops to ground level to go under Beltway 8, then returns to an elevated structure.
The easement specified in the FEIS shows many displacements south of the Union Pacific tracks, mostly warehouses and lower-tier commercial properties which are politically easy to acquire, although some will be expensive. Looking at HCAD property listings between West Little York and Huffmeister (just west of State Highway 6), I don't see any properties with ownership listed as an entity which is obviously affiliated with the project, so it appears no property acquistion has been done.
West of Huffmeister, 63 homes in the White Oak Falls neighborhood are needed for the train corridor. THSR (then Texas Central) acquired 41 of these homes between December 2019 and March 2020. Channel 2 has reported twice about mismanagement of these properties, on August 6, 2024 and October 9, 2025 (video). Many homes owned by THSR were abandoned and not maintained, with squatters and criminal activity reported by neighborhood residents. I visited this area on July 20 and found the situation has been mostly cleaned up since the Channel 2 reports. Only four houses have visible damage or vandalism, mainly broken windows. About 13 homes are posted with no trespassing signs and appear to be unoccupied, and the rest are in good condition and/or occupied. It seems like a questionable decision to buy the homes first, instead of buying vacant land which was at risk for development. The situation was obviously mismanaged, causing grief for the neighborhood and avoidable expenditures for THSR.
Going west from the White Oak Falls neighborhood to Fry Road, I reviewed HCAD records and found no properties with ownership listed as THSR or an obvious affiliate. Plenty of displacements are anticipated in the FEIS, mostly easy-to-acquire lower-tier commercial properties.
After release of the FEIS, a warehouse was built in the path of the alignment at Barker Cypress. It will be easy to acquire but has a structure property tax value of $14.4 million.
At Fry Road the alignment shifts southward, away from US 290 and the Union Pacific railroad. Then the trouble starts. When the FEIS was prepared in the late 2010s, the land south of US 290 between Cypress and Hockley was nearly entirely vacant, which is shown in the alignment map. Substantial development has occurred in the path of the planned alignment in the 2020s
Dunham Pointe
In the FEIS alignment map (pages 69 and 70), the area which is now the Dunham Pointe development was still vacant land. Over 30 homes have now been built in the FEIS alignment, with homes costing up to $1 million currently being built directly in or very close to the path of the FEIS alignment by Toll Brothers and Tri-Pointe Homes. I spoke to representatives of both builders. They say the rail corridor is definitely not coming through the neighborhood, but they were unable provide any official documentation to support that claim. The development plan for Dunham Pointe on its web site does not include any provision for the high speed rail corridor.
At this location the FEIS alignment crosses Flowering Sage Way, which is full of high-dollar homes built by Toll Brothers.
This Toll Brothers home under construction at the corner of Elegant Azalea and Winding Angelica is directly in the path of the FEIS alignment.
Jubilee
Proceeding west from Dunham Pointe, the alignment remains clear of development for the next 4 miles (for now) and then reaches the rapidly expanding Jubilee development. The master plan (shown below with the alignment annotated) does not indicate any provision for the rail corridor. Housing construction has not yet reached the FEIS alignment corridor, but at the southern end of Jubilee Haven Boulevard I could see site work in progress in the distance, possibly in the FEIS corridor.
A home lot map available on the Jubilee site shows brown earth very close to the FEIS alignment path. With the new schools under construction north of the alignment, I think we can expect housing construction to start in the path of the alignment in the near future.
The Grand Prairie and Sanford Farms
Sanford Farms and The Grand Prairie are the only two spots where there is a provision for the FEIS alignment. Retention ponds are built along the alignment, which would allow the elevated structure to be built without any home displacements. But residents may not be thrilled to have the elevated structure coming through the middle of their neighborhoods. The map below shows the FEIS alignment through the area.
However, on the southwest corner of Warren Ranch Road and Baethe Road housing is being built in the alignment path. This photo looks along Prairie Rush Drive where the FEIS alignment comes through at a shallow angle (about 20 degrees from parallel to the street).
In The Grand Prairie, HCAD records show the retention pond land owner is mostly Lennar Homes with a smaller parcel owned by "SB-HS Lot Option Pool 01 LP". I spoke to a sales representative for Lennar Homes who told me that around the end of 2025 they stopped requiring homebuyers to sign a disclosure statement about the rail corridor. The plat for the area in the photo below says the retention pond is "Restricted to open space/landscape/incidental utilities/detention".
The FEIS alignment comes through at this spot on Mesquite Field Drive. If built, it will be an elevated structure over the detention pond going straight ahead.
The FEIS plan shows an area for a "Rail systems site" (electrical substation), and a large "Temporary construction area" southeast of the intersection of Betka and Kickapoo roads. This area is now filled with houses.
It is surprising that Sanford Farms is designed to accommodate the FEIS corridor, because the subdivision infrastructure was platted and built very recently in 2025. From publicly available information it's not possible to know if there was coordination with TSHR; it's possible the land developer acted independently and may have designed the subdivision years ago when it was more likely the project would be built. A sales manager for Meritage Homes told me homebuyers are not provided with a disclosure about the project. He told me internal communications say the rail project is not going to be built, so they don't need to be concerned about it. Retention ponds in Sanford Farms are jointly owned by Meritage Homes and "Century Land Holdings of Texas LLC".
The FEIS alignment is straight ahead through the retention pond in Sanford Farms. If the rail project is built, an elevated structure will be built here.
North of Sanford Farms the path remains clear, but a large structure has been built in a planned temporary construction area.
The New Blinn College Waller Campus
There is a construction site just getting started along FM 2920 with a large sign announcing "Future Home of Blinn College District". There are deep piles being sunk, so work is clearly for a large building. It turns out this is the site of the new 100-acre Blinn College Waller Campus, and this is the first of many planned buildings.
This image shows the site map from the Blinn College press release, with annotations added to indicate the construction zone and the FEIS alignment through the middle of the property. The land was purchased in October 2024. (video)
This is a depiction of the first building on the campus, with a cost of $68 million (press release).
Here are views of construction, with work just getting started
I was not able to find a master plan for the campus, and at this early point in construction I could not determine if the building is in the path of the FEIS alignment. If it's not in the path, its eastern edge will be very close. It is unlikely the campus is intended to accommodate the rail corridor, which in the FEIS (pages 58 and 59) is elevated at FM 2920 but on embankment on the north half of the Blinn property. At the groundbreaking ceremony the Chancellor of Blinn made the following statement (see video at link)
The academic building we celebrate today will be the first structure on this new 100-acre campus. It will serve as the foundation for future campus development ...
This view looks north from FM 2920 along the FEIS alignment. The pine tree ahead is in the middle of the FEIS alignment, and the building construction site is to the left (west). Foundation work is starting on its west end and progressing in the direction of this view.
Binford Creek
Just north of the Blinn campus site is the Binford Creek subdivision. The area of Binford Creek was vacant land in the FEIS maps, but is now a subdivision filled with houses. The FEIS footprint map (page 56, shown below) shows most of the area of the Binford Creek subdivision was targeted to be a permanent maintenance-of-way facility (MOWF) for the rail project, but the image below shows that the entire area is now a subdivision.
The photo below looks north along Willow River Canyon Lane. All the recently built houses on the right side of the street are directly in the path of the FEIS alignment.
There is a discrepancy in the alignments shown on the official web site. The alignment map shows a path west of the FEIS map path. I sent an inquiry to the THSR general inquiries email, but did not receive a response. I'm assuming the FEIS alignment is correct, but it makes no difference — both paths are now built up with houses. In the map you can see an easement running approximately parallel to Binford Road. At first glance I thought this might be a corridor preserved for the rail, but it turns out it is a 100-foot-wide easement owned by the Southeast Regional Management District utility agency and 75 feet of the width is earmarked for Lonestar Gas.
Solar Farms
Farther north in Waller County, a large solar farm has been built in the path of the FEIS alignment. Needed property can be easily acquired, but will be another substantial expense.
I reviewed the path going north all the way to Dallas. I did not find any new development in the path of the planned alignment except two solar farms. At this location there is a new electrical substation and a small area of solar panels in the path, and here there is a new solar farm.
North Texas
While there has been minimal political attention relating to high speed rail in Houston, the situation is very different in North Texas. Since around 2020, the North Central Texas Council of Governments, their equivalent to H-GAC, has coordinated studies to define an alignment to extend the high speed rail from Dallas to Fort Worth, including a station in Arlington. There is an abundance of documents on the study web site. As May 2026, NCTCOG reported spending $10.9 million on the studies (page 48). Separate from NCTCOG, three economic impact studies have been done, including a study by Fort Worth and Arlington completed a few months ago.
The environmental process is especially difficult from the bureaucratic perspective, needing to get environmental clearance from the Federal Transit Administration while also being accepted into and being in compliance with the Federal Railroad Administration "Corridor ID" program (page 47). In 2024, after considerating a huge universe of alternatives, the study announced its recommendation, which almost entirely follows the Interstate 30 corridor. Looking at the map, you can see that the recommended alternative has long tunnels in Fort Worth and Arlington, but in Dallas it was on a high elevated structure through the southwest corner of downtown. (newsletter with image shown below)
 
The original recommendation for the Fort Worth extension included an elevated structure through southwest downtown Dallas. This incited strong opposition from the City of Dallas and Hunt Realty Investments, resulting in the a new alignment farther west.
Then the trouble started. The City of Dallas was furious that the recommended design has tunnels for Fort Worth and Arlington, but a high elevated structure in downtown Dallas. The area of the downtown Dallas alignment is targeted for future redevelopment. Hunt Realty Investments, owner of Reunion Tower and adjacent Hyatt Regency, unleashed the wrath of their lawyers on NCTCOG, sending a continuous barrage of opposition and threatening letters to the NCTCOG transportation council. (examples page 74, page 57)
By August 2024 the alignment study was working to define a new alignment which avoids the Hunt property and placates the City of Dallas. This map shows the two alignments, with the east (dark blue) the original recommended alignment and the new west (light blue) alignment which mostly avoids southwest downtown Dallas but still has a short section inside the downtown Dallas freeway loop.
In January 2026 Dallas City Council reaffirmed a previous 2024 resolution which opposed any elevated structure inside the downtown Dallas freeway loop. As of May 2026 (page 49), it was believed that Dallas would oppose any elevated structure anywhere inside the downtown Dallas freeway loop, which would preclude the planned western alignment since it briefly goes inside the freeway loop and could potentially entirely kill the extension to Fort Worth. To get a feeling for the opinion of Dallas, listen to the commentary from a Dallas City Council member starting at 2:18:57 in this video. She summarizes by saying "It's really outrageous the way [the alignment study process] has gone forward." Negotiations were on hold during the summer due to the World Cup, but should resume soon.
I can't find an official cost estimate for the Dallas-to-Fort Worth section, but anything with tunnels is certain to be outrageously expensive. During a NCTCOG meeting a cost of $12 billion was mentioned. (1:00:00 in this video)
All of this effort in North Texas is contingent on the Houston-to-Dallas project actually being built. Does NCTGOG and its politically powerful director Michael Morris know something that I don't know? Maybe. But I'm also inclined to wonder if Morris is aware of the alignment becoming nonviable in Houston.
A New Wildcard: Terafab (Updated Aug. 6, 2026)
The FEIS alignment includes a station east of College Station called the Brazos Valley station. This station seemed politically motivated to help gain support from that area. It is located in the middle of nowhere (at Roans Prairie near the intersection of state highways 30 and 90), 25 miles from the Texas A&M campus.
But the Brazos Valley station may have suddenly become relevant. In May it was revealed that SpaceX selected the Gibbons Creek reservoir site for its proposed massive Terafab semiconductor factory, with economic incentive documents listing a $55 billion initial investment and a potential total investment of $119 billion. The proposed size of this facility is truly staggering, potentially 10 times larger than the massive Tesla factory in Austin, with a target production area around 100 million square feet (3.6 square miles).
SpaceX quickly received needed agreements in June and July, including a large tax investment zone called SpaceX Reinvestment Zone No. 1, which is reported to be 22,000 acres (34 square miles) in addition to the 4000-acre reservoir. I added the proposed station location to the map of the zone included in a Houston Business Journal article.
On August 6 SpaceX posted a video and depiction of the site on Terafab.ai. (HBJ report) The depiction below shows the massive factory on the southeast side of the reservoir, on the side of the property closest to the proposed train station. Of course this is preliminary and tentative, but the concepts are truly impressive. A post on X stated "Terafab looks more like a sci-fi city than a chip factory", to which Elon Musk responded, "Sci-fi city is what we're aiming for".
This rough size estimate from blogger Joe Tegtmeyer shows the approximate footprint, with the east side of the factory about 2 miles from the proposed station at Roans Prairie.
If Terafab moves forward as envisioned, it will create an entirely new city as suppliers and workforce move into the area. The rail project's prospects could receive a boost from the potential ridership resulting from Terafab.
It also causes me to speculate about another possible outcome for the FEIS corridor. Could SpaceX take over the project and convert the alignment to a paved roadway for high-speed self-driving Robotaxis and possibly freight? This would drastically lower the cost of the project, perhaps to around five billion dollars (the cost at $20 million per mile), and make alignment adjustments much easier, since robotaxis can accommodate more curving alignments compared to a train. A high-speed roadway would allow entrances and exits at just about any point and could provide service to rural areas, which could potentially extinguish rural oppostion, since they would receive a benefit from the project. While Robotaxis would be slower than 200mph trains, Robotaxis would provide on-demand service with no waiting, no security screenings and potential point-to-point service if the Robotaxis also drive on streets, likely closing most of the time gap compared to 200mph trains. If there's one thing we can say about Elon Musk and his companies, it's that bold visions are possible.
UPDATE August 31, 2026: A drone video of the Tesla test track in Austin shows Cybercab vehicles going around the track in a platoon-like configuration, something we would expect to see on a dedicated Cybercab roadway. Go to timestamp 28:30 in this Youtube video by Joe Tegtmeyer. By the way, if you have an interest in Tesla developments, follow Joe Tegtmeyer because he has exclusive drone access and posts top-quality videos.
Screenshot from the video of Cybercabs going around the test track in a platoon-like configuration on August 31, 2026.
Terafab is also close to corridors under study for Interstate 14 and a new loop around Bryan-College Station, as shown in this map from the Interstate 14 study web site with annotations added.
The Grim Financials
There is a reason the project has been on hold since approval of the FEIS in 2020. The cost of building high-speed rail is absurdly high. The FEIS (page 45) estimated the capital cost to be between $16 billion and $19 billion in 2019.
When federal Department of Transportation canceled a $63.9 million grant to Amtrak on April 14, 2025, the press release stated that the cost is expected to be over $40 billion.
In a July 2025 editorial, John Kleinheinz, chief executive of the firm that owns Texas High-Speed Rail Holdings, stated that the cost of the project is $35 billion. The cost of highway construction has actually declined 16.1% between May 2025 and May 2026. However, high speed rail requires extensive electrical infrastructure, and the data center boom has caused equipment shortages and cost increases, especially affecting transformers. So the 2026 project cost is likely around the same as 2025. I'll use $40 billion for illustrative calculations.
Even with the high cost of THSR, it's chump change compared to California High Speed Rail. That 500-mile project was estimated to cost $33 billion in 2008, but in 2026 the official cost estimate is $231 billion (page 18), and likely to go higher if history is a guide.
In England, the HS2 project going north from London was drastically scaled back due to outrageous costs. The entire project with a Y-shaped alignment serving London, Birmingham, Manchester and Leeds was estimated at 32 billion pounds ($43 billion) in 2011. As the cost escalated, the two lines north of Birmingham (to Manchester and Leeds) were canceled, and as of May 2026 the 140-mile line between London and Birmingham is expected to cost up to 103 billion pounds ($138 billion) with completion between 2036 and 2039. This is nearly $1 billion per mile — truly mind-boggling!
Even at the bargain-basement price of $40 billion for Texas high speed rail, let's consider the cost per rider. THSR claims they can get 6 million riders per year. A recent study by the Boston Consulting Group said 6.5 million is possible.
Federal 30-year treasury bonds have a yield of 5.14% as of July 21. Since bonds for a high speed rail project would be far more risky than treasury bonds, a 7% interest rate seems reasonable. Interest alone on a $40 billion bond issue at 7% is $2.8 billion per year. Divide that by 6 million riders, and the annual interest cost per rider is $467. (!!!)
The annual payment including principal for a 30-year bond would be $3.223 billion, which is $537 per rider. Of course, this doesn't include other costs like operations, maintenance and security.
For comparison, United shows a cost of $249 for a one-way ticket for immediate travel and between $220 and $249 for travel within 3 weeks. Advanced purchase prices (three weeks or more) are as low as $82. Southwest shows similar fares, between $230 and $249 for immediate or near-term travel, with the price dropping to $82 for a 24-day advance booking.
If these numbers tell us anything, it's that the project can only be done with substantial government funding — some combination of direct payments for construction and low-interest loans. I'm sure the project proponents have known this all along. They want to position the project to be "shovel-ready" in the event there is a large federal spending binge or stimulus, something like the inflation-causing $1.9 trillion American Rescue Plan Act of 2021. But with a negligible amount of right-of-way actually acquired, the project is nowhere near shovel-ready.
It will take a large government subsidy to overcome the sky-high construction cost. For example, suppose federal funding covers 50% of the $40 billion construction cost and provides a 2% interest rate loan for the remaining $20 billion. The annual bond payment including principal is $893 million and the capital cost per rider $149. Overall cost per rider (including operations) could plausibly be around $200, which would be competitive with airlines.
Can we get some facts and truth about this project?
Texas High-Speed Rail Holdings LLC needs to come clean about the project status. If the northwest Harris County alignment is dead, they need to make an announcement. If they plan to do a supplemental environmental impact study to identify a new alignment, they need to start efforts as soon as possible because the clock is ticking as new development proceeds quickly. An SEIS would be an indication that the project is not dead.
If they still plan to build on the FEIS alignment and bust through the neighborhoods and Blinn campus, they need to make their intentions clear and start protecting the corridor.
If the FEIS is obsolete, the Federal Railroad Administration page with the official documents needs to be updated with a disclaimer or clarification of the FEIS status.
If the project is dead, it would be proper etiquette to inform the folks in North Texas so they can stop going at each other's throats.
Without having inside information, it's difficult to know the true status of the project. But the prohibitively high $40 billion cost and development in the path of the planned alignment in northwest Harris County suggest the rail project is going nowhere.
But Terafab may create new possibilities for the approved FEIS corridor. I'll be tuned in to find out if Terafab has any influence on the future of the project.

Labels: , , ,

Tuesday, June 02, 2026

Houston Metro vs. Dallas DART: Houston wins as costly light rail nearly destroys DART

Another fantastic and devastating (for DART) guest post from Oscar Slotboom.

---

Updated 6/3/2026 to include the official announcement of the Dallas Stars moving to Plano
2026 has been very bad for downtown Dallas. On January 5, AT&T announced it will leave downtown to build a new $1.35 billion headquarters complex in Plano. In May, Fifth Third Bancorp announced it will leave downtown, vacating 222,900 square feet of space in a prominent towner. On June 1, the Dallas Mavericks announced they will leave American Airlines Center, with its DART light rail station, to move to the former Valley View Mall site along Interstate 635 (LBJ Freeway) in North Dallas with no light rail service. Like a one-two punch, the next bomb dropped on June 2 when the Dallas Stars officially announced they will leave downtown to move to the Shops at Willowbend site in Plano, which has no light rail service. These events are the culmination of years of decline for downtown Dallas (more below).
How can this be happening? Dallas Area Rapid Transit (DART) has 93 miles of high-quality light rail focused on downtown Dallas. Shouldn't light rail be an asset which attracts and retains downtown tenants? The typical response would be to say these departures are happening in spite of the light rail system serving downtown. But the more likely reality is that downtown lynchpins are fleeing partly because of the light rail system, and its affiliation (justified or unjustified) with unsavory elements and perceived high crime rate.
Earlier this year, DART had bigger problems than the ongoing loss of potential customers at its downtown hub. Six member cities, fed up with DART's high costs and unwillingness to share sales tax revenue, threatened to leave DART. In February a compromise was reached, preserving DART at least through 2031.
With all the news about DART, it's a good time to do an analysis of Houston's Metro vs. DART, to find out how the two agencies compare on key performance statistics. Spoiler alert: Houston performs far better in nearly every key statistic. The reason is because DART spent massively on downtown-focused, high-cost light rail. Houston has far less light rail, even with the cringe-worthy Green and Purple lines, which has enabled Metro to have much better financial performance.
This map shows the 119-mile DART light rail system and its connections to other rail service. The system is heavily focused on downtown Dallas, which is in serious decline while almost everywhere else in North Texas is booming.
Key Statistics
All data is for 2025 except as noted. (Metro, DART) Unusually bad statistics are highlighted with red, good statistics are highlighted with green.

Houston Metro DART
Year created 1978 1983
Service area 1307 sq miles 700 sq miles
Service area population 4,038,849 (2024, source) 2,646,340
Light rail miles 22.7 119
93 as of 2025
Light rail standards Low High
Employees 4,779 3,512
Sales tax rate 1% 1%
Sales tax collection $1,086 million $900.4 million
Per capita taxation $268.88 $340.23
Sales tax sharing with member cities 25% target
27.1% since 2002
23.3% since 2016
none until 2026
5% in 2026
rising to 10% in 2031
Ridership (boardings) 78,642,439 54,175,258
Ridership per capita 19.5 20.5
Year of peak ridership 2006 2013
Ridership loss since peak -23.5% -24.0%
Ridership loss since Covid (U.S. average -15.2%) -12.6% -21.6%
Operating expense $1,076 million $1,021 million
Operating loss $1,027 million $959.1 million
Average taxpayer subsidy per boarding $13.07 $17.70
Long term debt $758.8 million (page 99) $4.027 billion (page 81)
Long term debt per capita $188 $1522
Interest on debt $28.1 million (page 100) $144.4 million (page 11)
Interest + principal $98.2 million (page 100) $210.8 million (page 83)
Ridership
DART steadily expanded its light rail system after the first opening in 1996, reaching 93 miles in 2014, all focused on downtown Dallas. (A 26-mile line from Plano to DFW airport opened in fiscal 2026 and is not included in ridership data.) Houston opened the first section of the Red Line in 2004 and the system reached 22.7 miles after expansion in 2014.
Houston Metro ridership peaked in 2006 and in 2025 was 23.5% below the peak and 12.5% below the pre-Covid 2019 ridership. DART ridership peaked in 2013 and in 2025 was 24.0% below the peak and 21.6% below 2019 ridership. In March 2026 the national average for ridership vs. 2019 is down 15.2%, so Metro is performing better than the national average and DART is worse.
The plot shows that DART's ridership between 2002 and 2019 was basically flat, even with huge spending on light rail. Of course ridership collapsed in 2020 due to Covid.
Metro and DART are now in a very similar position in terms of ridership. Per capita ridership is close, 19.5 for Metro and 20.5 for DART (see table). Metro is 23.5% below its peak and DART is 24.0% below its peak.
DART's extensive and costly light rail system, including DFW airport service, does not provide better ridership performance.
Sales Taxes, Revenue Sharing and the DART revolt
A huge flow of sales tax revenue goes into the bank accounts of Metro and DART. Inflation-adjusted tax collections are up 73.1% for Metro and 66.1% for DART since 2003.
While the population of the DART service area (2,646,340) is only 65.5% as large as Metro's (4,038,849 in 2024), DART sales tax collection is 82.9% as large as Metro's. DART's per capita taxation is $340.23 per year, and Metro's is $268.88 per year.
Houston Metro has had a formal policy of sharing sales tax revenue with member cities since 1988, traditionally called "general mobility" funding and, in financial statements, "local infrastructure assistance". The target for infrastructure assistance is 25% of sales tax revenue. 2025 was 30.1%, the 10-year average since 2016 is 23.3%, and the average since 2001 is 27.1%.
DART, however, never had a sales tax revenue sharing program. (There was a one-time transfer in 2022.) Certain member cities, particularly Plano, Farmers Branch and Irving, became increasingly frustrated about their high sales tax payments to subsidize high-cost DART, leaving them with no resources for other transportation needs. In recent years DART consistently refused to return sales tax revenue to member cities. Legislation was introduced in the Texas Legislature to mandate DART return 25% of its sales tax revenue to member cities, but the legislature was unwilling to get involved in the local issue. The situation came to a boiling point in 2026, when six member cities - Plano, Irving, Farmers Branch, Addison, University Park and Highland Park - scheduled elections to withdraw from DART. After months of negotiation, an agreement was reached for DART to share some sales tax revenue with member cities, starting at 5% in 2026 and reaching 10% in 2031. Plano, Irving and Farmers Branch canceled their withdrawal referendums. Addison, University Park and Highland Park proceeded with elections, and only Highland Park voted to withdraw.
DART's high costs and unwillingness to share revenue nearly resulted in the collapse of the agency. In contrast, Metro's longstanding revenue sharing targeting 25% has empowered member cities to meet their local mobility needs.
Taxpayer Subsidy per Boarding
Average subsidy per boarding is the cost covered by taxpayers every time someone steps onto a bus or train. A transit roundtrip is two boardings. This subsidy is calculated by dividing the operating loss by the number of boardings.
While both Metro and DART have high average boarding subsidies, DART's $17.70 subsidy is significantly worse than Metro's $13.07.
The DART web site scorecard page (click Finance button) reports the bus boarding subsidy is $13.93, the light rail subsidy is $11.73, and the paratransit subsidy is $72.92 in April 2026. In my view the average subsidy per boarding for all services is the best indicator of agency cost and the best number for comparison.
Debt and Interest Payments
DART incurred substantial debt to build its light rail system. The 2025 financial statement shows $4.027 billion in long-term debt.
Houston Metro's financial statement shows $758.8 million in long-term debt. Metro's per-capita debt of $188 is only 12% of DART's $1522.
Metro's interest payment is only $28.1 million per year, compared to $144.4 million for DART.
Houston is in a vastly better position than DART, with far lower debt and lower interest payments. Page 18 of the Metro annual report states, "The debt payable balance has been declining during the last several years as principal payments were made, and the amortization of premium/discount occurred." Metro acheived this superior financial position because it built much less light rail, even with the ill-advised Green and Purple Lines and their disastrous ridership.
Light Rail Doesn't Save Downtown Dallas
An extensive light rail system focused on downtown should help ensure a vibrant and growing downtown, right? No, not in Dallas. Downtown Dallas has generally been stagnant or in decline since light rail service started 1996.
After AT&T, the Dallas Mavericks and the Dallas Stars dropped their departure bombs, the outlook for downtown Dallas became increasingly grim. The downtown office vacancy rate is projected to reach 35% after AT&T and other departures.
On December 28, 2025, the Wall Street Journal reported on downtown Dallas with an article "Dallas Is Booming—Except for Its Downtown"
DALLAS — This city is a hotbed for commercial property. The metro area’s population is booming and financial-services firms are flocking here, earning the area the sobriquet “Y’all Street.”
Yet at its heart is one of the country’s worst-hit central business districts: Downtown Dallas.
Companies are abandoning this neighborhood and its aging office towers. They are heading to the Uptown district or the thriving suburbs, often over concerns about crime and homelessness. Left behind are defaulted loans, foreclosures and deeply discounted property sales.
Dallas’s downtown has the second-highest office vacancy rate of any in the nation, behind Seattle. Downtown Dallas, the central business district, was at 27.2% at the end of the third quarter, according to real-estate data firm CoStar. In Dallas’s Preston Center district, 7 miles away, the office vacancy rate was only 5.9%.
The Wall Street Journal recently reported on downtown Denver, another city which built extensive train service to downtown, only to see its downtown become a wasteland as employers fled to the suburbs. "Can This Guy Get People to Live in America’s Emptiest Downtown?"
America’s downtowns are suffering a crisis. Office work has migrated into the suburbs, leaving abandoned buildings and blighted conditions behind. Central business districts from St. Louis to Dallas and Portland, Ore., are fighting to escape a death spiral.
Downtown Houston has had and continues to have struggles, with office vacancy at 28.9% in March. (The Wall Street Journal report used older or different data.) I think downtown Houston is in a better position than downtown Dallas, but that subject is beyond the scope of this post.
Conclusion
Here are the consequences of of DART's high-mileage, high-cost, downtown-focused light rail system
  • Inability (and unwillingness) to share sales tax revenue with member cities, leading to a revolt which put the agency's future at risk
  • High debt, 5.3 times higher than Metro, with per-capita debt 8 times higher than Metro
  • Downtown Dallas is among the worst performing downtowns in the United States, with lynchpins AT&T, Dallas Mavericks and Dallas Stars announcing departures this year.
  • Long-term ridership comparable to Metro, but worse than Metro and the national average since Covid
  • A sky-high $17.70 average boarding subsidy ($35.40 for a roundtrip)
  • High per-capita taxation, $340.23 vs. $268.88 for Metro
In the Houston vs. Dallas transit performance contest, Houston wins by a landslide, mainly because Houston has built much less light rail.
As I've posted in the past, the future of public transit is new, low-cost technology: automated Robotaxis and low-cost tunnel transit built by The Boring Company.

Labels: , , , , , ,