The “Go Big” mobility bond will increase the single-occupant vehicle share of Austin work commutes.
Stuck in traffic
Austin’s sprawling, car-centric development pattern and unproductive public investment choices undermine the city’s ability to move away from a 73% single-occupant vehicle (SOV) commute share.
In growing regions, additional road capacity doesn’t generate sustainable reductions in travel times to work for motorists. New lanes support sprawl, which in turn generates more car trips. As a result, the previous congestion returns. This paradox is referred to as “induced demand”.
A more effective transportation strategy is to get motorists out of taking single-occupant vehicle trips by providing attractive mobility choices through transit, carpooling, telecommuting, biking, and walking. This concept is referred to as “mode shift”. Given Austin’s expected growth and constrained fiscal powers, mode shift is the only viable long-term approach for humane journey-to-work travel times.
In the past decade, other large, low-density cities have been able to make greater progress and close the gap with Austin’s relatively lower 73% SOV share. The other large Texas cities moved in the wrong direction, with increases in SOV commute share.
The story within Austin’s densest corridors is somewhat mixed.
For example, between 2000 and 2013, the census tracts adjacent to the Burnet corridor between Koenig Lane and MoPac experienced a 4% reduction in SOV as commutes shifted into transit and telecommuting. There was no net SOV reduction in the South Lamar corridor’s census tracts between East Riverside and US 290. And in East Riverside corridor’s census tracts between I-35 and SH-71, there was just a 1% SOV reduction as carpoolers shifted into other modes. So, even in Austin’s dense corridors, driving alone is the main commute choice, even if its share is threatened by both the perceived and real inconveniences of traffic congestion and car-ownership.
The proposed “Go Big” bond projects. Dotted lines indicate that only preliminary engineering and design will be conducted as part of the bond program.
With the current mode share context set, let’s proceed to scoring the expected mode shift from the $720 million “Go Big” bond. The calculations below can be found in this set of worksheets (Excel).
Bikes & Trails
The bond includes a $46 million allocation for the 2014 bike and urban trail plan; this is 31% of the $150 million originally sought by the Bike Master Plan (which includes urban trails). If fully implemented, the bike plan estimated a shift of 5,950 commuters into bike mode share. This bond’s portion would be 1,845. However, it’s fairly certain that in a similar fashion to the corridor portion of the “Go Big” plan, the original cost estimates are no longer an accurate reflection of the present-day implementation expenses. If we assume a similar 3x cost increase as in the corridor plans, then the expected bike mode shift would be closer to 615.
“Go Big” provides $37.5 million for implementation of the Sidewalk Master Plan. The plan found that Austin is missing 2,580 sidewalk miles in the public right-of-way. It would cost $1.6 billion to fill that gap. That’s a per mile cost of roughly $620,000.
The $37.5 million will generate somewhere between 40 and 60 new sidewalk miles depending on the the final costs and complexity of the targeted builds. The current sidewalk network is about 2,400 miles, which means an increase of 50 miles is a 2% augmentation to the existing built network. Austin presently has a 2% walking mode share. A boost of 3 and 5 basis points to the walking mode share leads to 157 and 261 new walking commuters, respectively.
In addition, the bond includes $2.75 million for each Council district to allocate towards improving the safety conditions of student and parents walking to school. This part of the bond, was, simply put, a political concession to West Austin Council Members. While school safety is worth investing in, the effect on mode shift is likely to be negligible due to the politicized and fragmented nature of the implementation.
The bond provides $482 million for road reconstruction targeting 26 miles of urban arterials (Airport Boulevard, Burnet, Guadalupe, North & South Lamar, East Riverside). These urban arterials are the so-called “smart” corridors. You can get the details of the previously-developed corridor plans here. However, the details contained in the drafted plans are not required by the bond language. The ultimate implementation will depend on the composition of the next City Council, the leadership style of the next City Manager, and future neighborhood politics.
This section scores the corridor portion of the bond based on Assistant City Manager Robert Goode’s memo to Council covering likely implementation projects for each corridor. That said, the Mayor’s public comments during the campaign – and the lack of response to said comments by the Council Members with more “progressive” transportation views – indicate a more car-centric implementation is most definitely within the realm of possibility. No elected officials or City Council candidates have urged a more aggressive mode shift implementation of the corridor portion of the investment.
According to Goode’s memo, a total of 20.4 corridor miles of bike lanes, as well as improved sidewalks will be delivered with “Go Big” funding. Since there are roughly 72 total miles (both sides of the arterials) necessary for 100% coverage of the corridors, the Goode memo’s implementation approach provides 39% bike and sidewalk coverage for the urban corridors.
According to City staff, approximately 17% of Austin population lives within a quarter-mile from the arterials. We’ll use this quarter-mile radius as the catchment area for new transit, bike, and walk commutes. The 2015 Austin Community Survey indicates that Austin presently has 522,623 workers over the age of 16, and that would make the corridor share 88,846. Of those, approximately 34,855 would have new bike and sidewalk lanes implemented within a quarter-mile of their home.
Corridor bike mode share – based on the Bike Master Plan’s overall estimate and the sample corridor mode shift performance between 2000 and 2013 in the earlier tables – is likely to grow by 2%, which would put most areas near corridor bike lanes at 3% bike mode share. This would generate 697 new bike commuters. Portland, Oregon has a 6% citywide bike mode share, the highest share amongst large U.S. cities.
Besides Guadalupe and the Northern part of South Lamar, none of the corridors directly connect their residential density to major employment centers. The main downtown census tract has a 15% walk-to-work mode share. The adjacent East Austin census tract, has a steep drop-off, with a 5% walk-to-work mode share. The three relatively dense corridors profiled above struggled to make gains, in part because they are not directly adjacent to a dense jobs center. A 1% boost in the walk-to-work mode in the corridor catchment areas provides a net shift of 349 commuters.
The Goode memo does not indicate any investment in separated transit lanes on Burnet (where they are explicitly recommended in the corridor plan) or any other corridor. CapMetro’s response to Goode’s recommended implementation is included as an appendix in his original memo to Council. The CapMetro executives repeatedly point out the lack of funding for dedicated transit lanes. In a later memo, Goode makes clear that Burnet transit lanes are not included in his recommendation due to cost:
Goode’s implementation does mention bus shelter construction and remains vague about the precise mix of bus pullouts versus queue jumpers on both Lamar corridors and Burnet.
The Goode recommendation is non-committal about the nature and size of bus infrastructure spending from the corridor pot. It is quite possible that the ballooning costs of corridor reconstruction have reduced bus-focused investments to a dozen bus shelters and some bus pullouts. There’s just not enough planned in the Goode memo to award the corridors any mode shift points from transit.
Finally, we have to estimate what impact the access management initiatives and new signals will have on delaying mode shift by making SOV trips more convenient. Some of the plans refer to a 20% increase in vehicle volumes; others refer to maintaining the current “level of service” for car flow. Across the plans, intersection wait-time reductions from 11% to 49% are touted. There are also anti-mode shift effects outside of the corridor area. The “congestion relief” efforts will be implemented from the most congested corridor areas to the least; this will impact the attractiveness of SOV trips across the entire corridor.
If these changes undermine a Burnet-style transformation, that’s a 4% mode shift loss, or 3,554 commuters across the quarter-mile band of the corridors. If we use East Riverside’s 1% mode shift figure, that’s 888 commuters retained in SOV.
Combining the different modes, the net corridor mode shift ranges from a decrease in SOV mode share by 158 commuters to an increase of SOV mode share by 2,508.
The “Go Big” bond includes $101 for traffic-unclogging “improvements” to West Austin roads. In addition, FM 969 lane expansion is part of the corridor program bundle.
These are the most recent daily traffic counts available for these roads from the Texas Department of Transportation and the City of Austin.
The funds include a third lane for Parmer, a grade-separated interchange for Loop 360, and a new bypass road at 620/2222 . The most comprehensive memo from City staff on FM 969 implementation indicates a $40 million dollar investment from the corridor pool, which would be allocated towards interim expansion to six lanes for a 1.8 mile stretch of the corridor.
The table above models a 10% boost in trips and also 20% boost (based on the intersection impacts from the corridor studies). The latter percentage reflects the potential for a accelerated expansion of FM 969 through matching state funds. According to the National Household Travel Survey (NHTS), 17% of private vehicle trips are for work commutes. Let’s assume two trips per SOV commuter. Factoring these estimates together leads to a projection of an additional 2,139 to 4,277 commuters from the road portion of the bond.
However, in a no-build scenario, the majority of these commuters would still opt-out of carpooling, transit, or active transportation. Let’s discount the expected commuter growth by the existing SOV share, which is 73%. This results in a loss by alternative modes back to SOV that ranges between 577 and 1,155 commuters.
The “Go Big” bond features considerable capital opportunity costs, as well as street right-of-way opportunity costs.
Council chose to finance $250 million of the bond with available debt service capacity that does not require an increase in the property tax rate. This “free” capacity is the result of population growth coming in at a preset tax rate, previous debt service obligations expiring, as well as the defeat of the 2014 road & rail bond.
In part, this free capacity is why such a relatively large bond can be credibly sold to the public at such a comparatively low cost to other recent bond propositions. In my estimation, this category of debt capacity is a sort of political propellant that improves the odds of decent bond proposals passing from, say, 50-50 to 85-15. The $87.5 million difference in expected approved value between conventional bond capacity and this tax-increase-avoiding capacity is the capital opportunity cost of the election.
For $87.5 million, a plain mix of targeted bike, pedestrian, and targeted housing subsidy investments can generate a net mode shift between 725 and 967 commuters depending on the productivity of the investment mix. In recent elections, Austinites have passed bond propositions with these types of investments: 2006 and 2013 for housing, and 2010 and 2012 for bike and pedestrian infrastructure.
The major street right-of-way opportunity cost is productive light rail. The planned median on South Lamar effectively eliminates that stretch in the short- and medium-term as a light rail corridor. CapMetro and the City have repeatedly proved unwilling to invest in rebuilding recently-updated infrastructure. The design of dedicated bus lanes in the Guadalupe corridor also pose a threat to productive light rail. If they are highly-permanent, expensive, and incompatible with light rail, it unleashes the same bureaucratic resistance to tearing up recent infrastructure that rail advocates witnessed during the 2013-14 rail route selection process. The design of transit lanes on East Riverside is another risk to productive light rail, though much smaller since the Goode memo doesn’t allocate funds for implementation.
In my estimate, the current likelihood of a productive light rail proposal emerging from Project Connect 2.0 is around 40%. I expect the preferred route will once again be a version of East Riverside-Highland or the Mueller circulator concept.
Previous calculations found that an unproductive route could easily destroy about 10,000 bus trips per day. According to the NHTS, 30% of transit trips are for work commutes, so, the unproductive routes will reduce 1,500 transit commuters, with about 1,050 opting for SOV commutes. Factoring in the 60% route probability of adoption and a 50% chance of bond passage, the expected impact of the unproductive rail route is a loss of 315 commuters to SOV.
No estimate of a productive light rail’s ridership has been done since 2000. The 2000 FTA proposal featured an estimate of 37,400 daily trips. Let’s assume a contemporary estimate would conservatively yield 20,000 new trips. That’s 3,000 new transit commuters. Factoring in my 40% route adoption probability and a 50% chance of bond passage yields a net shift of 600 commuters.
Without the passage of this bond, the present-day expected light rail shift is 285 commuters into transit. This figure combines the productive and unproductive routes. If the right-of-way risks posed by the bond shift the expected outcome towards an 80% chance of an unproductive route selection, the present-day expected light rail shift is now a loss of 120 commutes to SOV (-420 plus 300) . This negative mode shift of 405 commutes (from positive 285 to negative 120) is the right-of-way opportunity cost I’m assigning to the “Go Big” bond.
If we combine the capital and right-of-way opportunity costs, the bond’s passage carries a mode shift opportunity cost between 1,130 and 1,372 commuters.
Something for everybody
The chart above visualizes the 243 cross-combinations of the best-case, worst-case, and average estimates for each of the major categories of the bond discussed above. It’s a simple way of representing “what if everything goes right”, “what if a everything goes wrong”, and the permutations in between.
The median outcome scenario is the addition of the average for each category. That expected outcome is a net increase of 1,853 SOV commuters. This is a reduction in mode shift compared to the status quo.
Even if one were to recalibrate estimates, it is very difficult to get the projects funded by the bond to generate a positive mode shift away from SOV of 3,600 commuters – the point at which it would match the per commuter mode shift cost-efficiency ($200,000 per commuter) of a high-quality mix of bike, sidewalk, housing, and transit projects. The unproductive nature of the overall bond allows small plans (i.e. in the tens-of-millions) that are focused on productive mode shift investments to severely outpace the “Go Big” package’s mode shift potential.