Transit, Turnout, and Trust

Is the Red Line Circulator (a.k.a the “Mueller” route) the only politically feasible initial urban rail sequence? Some proponents of the preliminary urban rail route generated by the Project Connect process imply this is the case.

To evaluate this we have to revisit the failed 2000 rail referendum. Its specter haunts transit politics discussions and is the key piece of publicly-available evidence in rail political feasibility debates.

In that election, there were 124,479 votes cast for rail and 126,434 against.  The narrow 1,955 loss margin means that if just 979 ‘no’ votes changed to ‘yes’ then the proposition would have passed.

In 2000 George W. Bush won Travis County with 46.40% of the vote against Gore’s 41.42%. In 2012, Barack Obama won with 60.14%.  However, let’s assume (as many smart Austin politicos seem to do) that local rail politics are roughly the same as 13 years ago because of something inherent (enigmatic?) about transit politics in Austin.

The three basic strategies for changing the failed 2000 referendum outcome are: (1) persuade ‘against’ votes so that they vote ‘yes’ or skip the item, (2) boost turnout amongst supportive constituencies, and (3) a combination of the first two.  So how can we ‘build a winning map’ and what does it tell us about a given rail route’s political feasibility?

Here’s the Austin geographic vote distribution from the 2000 rail referendum courtesy of the City Demographer Ryan Robinson (cherry red precincts voted ‘for’ between 30% and 40%):


The Robinson visualization focuses on vote proportions; but to get a better sense of where the damage was done, let’s examine vote count margin. Proportions obscure that some districts have more registered voters, as well as differing rates of turnout for those voters.

The top ten ‘against’ precincts and their count are visualized below:


What both the original Robinson visualization and my vote count version highlight is that the 2000 referendum was lost – geographically, at least – in South Austin.  It’s certainly possible that the election was actually lost by some demographic segment that is over-represented in South Austin.  Independent men and Democratic-leaning women that don’t use transit have been mentioned to me. Perhaps these groups disproportionately resided in South Austin then.

But there’s no public polls that I could find that would allow for verification of a demographic segment explanation for 2000.  More importantly, in contemporary rail politics, the political feasibility argument is often geographic. “You can’t get the neighborhoods you need” behind proposal X or Y is a familiar admonishment.  So any route proposal seeking to persuade ‘against’ voters needs to deal with the South Austin issue.

As the Robinson map shows, support for the 2000 line was concentrated in Central Austin, and quite overwhelmingly so in almost all of the precincts the proposed rail would serve.   However, there were important differences in turnout and level of support that serve as the foundation of a ‘base boost’ strategy.

The top thirty supportive ‘for’ precincts are listed and plotted below.  There are countless opportunities to find the missing votes through modest improvements in turnout or margin across many combinations of precincts. For example, the highest positive vote margin count actually came from a precinct with slightly below average turnout. So there are many ways to build the ‘boost’ map.


Inevitably, any conversation about boosting the base from 2000 turns to the reliability of student turnout; ‘students don’t vote’ is argued by some.  But the focus on turnout rates confuses what we are really looking for, which is count of the contribution to the victory margin.

Which would you rather have of two equally-sized precincts: a precinct that has 70% turnout and a 51/49 split for your side or a district that has 45% turnout and 70/30 split for your side?  The answer is the latter because of the margin.  Student turnout (or any group for that matter) is less important if there is a route that engenders a lopsided margin.

The source data and my calculations for the above visualizations can be found here.

When testing the political feasibility of a rail alignment, the 2000 election seems to suggest two major obstacles: South Austin skepticism and expanding the Central base.  Certainly there are other obstacles like anti-rail skeptics (i.e. Skaggs) and ideological conservatives.  But those are individuals that will be opposed to rail regardless of the route. Hence it makes sense to focus on demographics and neighborhoods that can be persuaded or mobilized.

There are two major corridors presently getting a lot of attention amongst Austin transit advocates:  something similar to the 2000 line focused on substituting the “1” bus line on Guadalupe-Lamar (GL) and the evolving Red Line Circulator/Mueller route. In addition, South Congress and Riverside have been floated; and obviously, it makes sense to at least explain why existing MetroRapid bus candidates corridors from the most recent CapMetro bus service plan are not the optimal initial sequence for rail. A map of them follows with the MetroRapid candidates in blue:


Among the two routes that have gotten the most attention so far, it’s possible that the circulator could maintain and potentially expand Central enthusiasm, though one wonders about the precincts that would have been directly served by the original 2000 line.  But it doesn’t seem to have an answer to the South Austin issue.  On the other hand, it might be able to improve performance in the Northwest corner of the Austin map if advocates tell some kind of Mopac congestion story related to the Red Line.

Advocates of GL might argue that running a better campaign that touts future expansion into the southern neighborhoods might be enough to get the line approved the second time around. They would point out that if the affordable housing bonds can be tried again after a narrow loss, it doesn’t make sense to abandon a route that came so close the first time even though it faced a uniquely hostile electorate.

Others might indicate that an initial sequence of GL in the South or some other route based on existing bus service that targets South Austin might be a ridership and political winner.

Presently, there is no body of publicly-available political polls that provide a sense of how voters would react to the ‘best message’ for each route.  And as already detailed, it’s not obvious that Mueller will achieve significant political lift relative to the GL route in the needed precincts.  A South Austin focused route might solve one problem but create lower support intensity in Central Austin.  The problem is compounded by the lack of detailed corridor/route comparison data.

At this time  no route seems like the clear, slam dunk political winner based on available polling or electoral analysis.  And if that remains the case, then other criteria such as ridership, cost effectiveness, economic impact and so on become the meaningful differentiators.

In general, Austinites should be cautious when someone argues for or against a rail route on political feasibility grounds but is unwilling to provide data to back up their assertions.  The experiences and theories of professional political and public affairs consultants should definitely be taken seriously, but need to be verified with independent, open data. Otherwise, we are handing over veto power of Austin’s transit future to rather opaquely-derived intuitions based on polls or electoral data analysis that are not submitted to public review.

This entry was posted in Transit. Bookmark the permalink.

14 Responses to Transit, Turnout, and Trust

  1. Jeb says:

    The MetroRapid routes are more than “candidates.” N. Lamar/S. Congress is under construction now and will be in operation in January. Months before we get to vote on any urban rail proposal. The Burnet/S. Lamar route will be in operation in August 2015.

    Rather than focus on just the political viability of the Hancock-Mueller proposed alignment, you should examine the complete Project Connect plan. Project Connect is the overall service plan. Hancock-Mueller would be just one piece of the plan, along with MetroRapid, managed lanes on Mopac, I-35 improvements, et al.

    • Julio Gonzalez Altamirano says:

      Thanks for stopping by, Jeb.

      The context for this blog post is the deliberation about Project Connect’s recommendation for the initial urban rail sequence. I think the proponents of GL, or Riverside, or South Congress rail as the initial sequence (as opposed to Mueller) are aware that there is bus service already and rapid is coming or likely to arrive soon. That’s already assumed, known, understood. What they then go on to argue is that even given that underlying bus service it still makes sense to do GL or Riverside or South Congress as the initial sequence for ridership or economic development or equity or what have you reasons. And some of the more cheeky ones would say that if MetroRapid is comparable, then why not just do rapid bus for the Mueller/Circulator before committing to expensive, fixed infrastructure.

      One of the arguments lobbed back at these folks is that it is pointless to even consider the alternative initial urban rail routes as they ‘can’t get the votes’. Here I was trying to figure out if that is evidently so, which it is not. Perhaps additional data will make it clearer that only Mueller can get the votes for some yet-to-be-articulated-reason. Moreover, from a political point-of-view (which is the perspective of this column), Project Connect will not face a comprehensive up or down vote. Or more to the point, the urban rail sequence and potential public-funding for it will be its own question and discussed that way, so we do indeed have to politically look for the votes for the urban rail choice independent of the overall regional vision.

      Now, the argument you might be making is about the GL route’s public value. In essence you are pointing out that GL already has MetroRapid, so there’s no public value to be found there and hence rail should be put somewhere else. That’s not a political argument, but rather a utilitarian one and a very important one. I think that proponents of alternate initial urban rail sequence routes would suggest that given the Measures of Effectiveness used by the studies that fed into Project Connect, that it is quite unclear if Mueller is the highest utility route based on things like total ridership, new ridership, fare recovery, and so on. Since there is no corridor comparison data, no consensus on this utility-driven calculation has emerged.

      One of the big issues is how different people assess the value of rapid bus; for some it is a near total substitute for rail. For others it is just a bit better than regular bus. So, sometimes when pro-Mueller people say to GL people: “But you already have MetroRapid there” it doesn’t make sense because they see that as a modest change from the status quo. So if they thought rail made sense there before MetroRapid, they still do. Having some comparison data for the corridors that assumes MetroRapid in there makes a lot of sense to me as a way of moving the discussion forward.

  2. Brad Absalom says:

    Good post Julio. It’s interesting that 2000 did so poorly in South Austin, since the 2000 plan did have rail going all the way to Slaughter on Congress, so at least a portion of South Austin was going to be served. I I don’t think that was going to be phase 1, but you could argue that the 2000 plan offered better future service to South Austin than Project Connect is, and if it had passed, I would hope we would be actively working on Phase 2 now.

    Obviously a lot has changed over the past 12-13 years, but I wonder if those neighborhoods simply don’t want improved transit. I have no data to base it on, but 2000 offered transit to at least a portion of the area and they overwhelming rejected it.

    • Julio Gonzalez Altamirano says:

      Our MetroRapid planning and bus service certainly seems to indicate that people in deep South Austin might be interested in transit investments. Hard to say without current, objective polling. It sure would be nice if some of the larger, civic-minded organizations (or a wealthy Austinite) did some thoughtful public opinion research on this and shared it with the public. Maybe one route is the clear winner, but if they all have essentially counter-balancing political issues, then we can focus on other criteria.

      I agree about the old data. I’d be the first to argue that relying on an election outcome from 13 years ago is not a good idea, but I felt I should address it since the debate just can’t seem to move on from that election. And I am also hoping that folks that use the ‘political feasibility’ argument at least start conveying to the community what data they are basing their warnings on. It will help speed up consensus.

      The other thing that I need to do a better job communicating is that we shouldn’t focus so much on what the politics were a decade ago, we should focus on making the case for what the best investment will be a decade from now when the rail actually has a ribbon-cutting, especially because the voter preferences seems so mysterious and and likely to be potentially malleable. Polling might change this sentiment in me, though.

      Now, maybe when we get the corridor comparison data, it turns out that one route is the clear winner on non-political criteria.

      • Brad Absalom says:

        I completely agree, with current data, or lack thereof, there’s no clear political winner, and in the absence of that data we should be looking at other metrics that you and others have mentioned. Hopefully more data will come as this process moves forward.

        I know neighborhood plans, don’t necessarily reflect majority opinion of a neighborhood, but could they give us some idea of which neighborhoods might be pro or con?

  3. This is interesting but it misses a key point. The City of Austin is not the same as the Capital Metro service district that voted for the 2000 plan. Take out the areas outside the City of Austin that won’t be voting for it and you get a different outcome.

    And counter to what Jeb says above, I do think its important to focus on the first leg. I’ve been hammering this home forever but no one seems to listen. The political viability of expansion for a whole system is DEPENDENT on a wildly successful first line. Look at what’s happened with the Red Line, it has given no political cover for expansion. You know what will? A successful first line just like Phoenix, Charlotte, Houston, Salt Lake City, Seattle, Denver, Minneapolis etc. Nashville did the same thing as Austin, cheap commuter line. Guess what? No future support or excitement for expansion. Tri Rail in Miami? No support.

    • Julio Gonzalez Altamirano says:

      I agree with both of your points: city limits make transit more viable and political path dependency as a key criteria. We need to do a better job at talking about the latter criteria.

  4. el_longhorn says:

    This is a great post. A few comments:

    1. As The Overhead Wire says, within the City of Austin light rail won the vote in 2000. Even ignoring the pro-rail demographic and political trends since 2000, a Guadalupe-Lamar light rail would very likely win today. So all this hand wringing about how will we win the vote is pretty silly, IMO.

    2. Demographic trends make it even more likely that a 2000 light rail plan would win. Population in West Campus and Downtown has grown very fast and both of those areas are likely to turn out heavily in support of a GL rail line. And there are literally dozens of new, large apartment complexes along GL that will have a strong incentive to vote for the rail line. I believe that South Austin has lost population relative to 2000, some places in absolute terms and other places just relative to the rest of Austin’s growth. At least that is what I remember from census numbers, most of which bode well for light rail.

    3. Traffic is becomin a more and more serious issue in Austin. People want something done about, probably a lot of somethings. There was not an urgency about traffic issues in 2000 the way there are today. And Austin has become a much more pedestrian/bike friendly city than it was in 2000, a much more pro-Democratic Party city than it was in 2000. Those trends will help a rail line election.

    4. The campaign money will be on the pro GL rail side. Property developers are already seeing the value of the corridor and capitalizing on it, light rail in that corridor is practically insurance on their investments. The anti-rail leadership is getting old and irrelevant. The constituency is not really there anymore. Hard to see an opposition campaign gaining much steam, especially with successful light rail plans in Houston and Dallas in the last decade.

    When I do some back of the envelope calculations, I see light rail along Guadalupe/Lamar/South Congress passing with 55-60% of the vote. The votes are there. All we are missing is the political leadership.

    • Julio Gonzalez Altamirano says:

      I think 1,2, and 3 make total sense. Less sure about the first statement about part 4 and I wonder about the second statement. Perhaps the ‘irrelevance’ is because we haven’t really put rail issues on the ballot since 2004. Have you seen some movement on the campaign money or recognize a past pattern?

      Some proponents of non-GL routes (whether it be Mueller, Eastside, SoCo) probably all believe that if a pro-transit majority is baked, it doesn’t matter what route we pick so we should use some other criteria. For example, pro-Mueller people will say that future economic development impact and TIF financing opportunities and (maybe) new riders are criteria the route shines on relative to GL. Hard to know for sure without shared corridor comparison data.

      I think when discussing the political feasibility criteria we have to be clear about whether we think the voters are there in general or for a specific route. If I read between the lines, I think you are saying that you feel secular trends definitely bolster GL given the 2000 Austin base but are not certain about other routes, right?

      • el_longhorn says:

        I think GL is by far the best line and will do the best job of maximizing the turnout in a light rail election, but I think even a Mueller/Red River alignment or other similar alignment will pass. The vote is going to be strong for this. The trends are all in one direction. People want transportation options. The link below is another pro multimodal transport trend.

        Taking into account all the quantifiable factors, the non-GL routes don’t even come close to GL in any metric. The GL route has the most VMU and TOD possibilities, most projected ridership, is more accesible to more peoples residences and work. In what quantifiable criteria could another route match GL? I can’t think of one. GL is the backbone route on which a strong transit system can be built.

        The campaign money comment is just a hunch. Great blog post. This is the most important issue in Austin, IMO. A GL line would transform the city.

  5. An important missing piece to your analysis is Federal support for the rail project. Without the FTA dollars, it doesn’t get built regardless of the political- (voter support) or data-driven (ridership projections) rationale. The “GL” corridor has already won a significant FTA investment covering 80-85% of the capital costs of the forthcoming rapid bus system BASED on our grant application that the proposed service would adequately serve the corridor. It’s unlikely that the FTA would invest redundantly in the same corridor before the rapid bus system has at least exceeded its minimum service life. It is unfortunate that the 2000 election forced us to develop a Plan B (rapid buses). The other option would have been to sit on our hands. As it stands today, the rapid buses will finally come online in 2013, after nearly 9 years of planning. The proposed urban rail on the Mueller corridor is projected to come online in 2021. If we had sat on our hands, we would have neither service until 2021.

    • Julio Gonzalez Altamirano says:

      We should definitely consider the benefits of MetroRapid in GL and the FTA funding risks when optimizing for the initial urban rail sequence. If we don’t, we are not making a data-driven decision.

      Here’s the thing though…people in the past have mentioned the FTA funding risk but don’t quantify it – there’s a big different between 80% likelihood of denial risk and 8%.

      If GL does prove to be the objectively superior initial urban rail sequence even considering MetroRapid on different measures, how much do we think going back to FTA for New Starts money (in the hundreds of millions to begin with) hurts us after winning the Very Small Starts (~$30 million from the FTA, $9 million locally) money? Because if the risk is a few percentage points, it might be optimal to maximize the public value. Related to this, is how confident we feel about winning a grant quickly, because as you point out, the later the New Starts win is, the closer MetroRapid would be to a minimum service life point. And those funds are already extremely competitive and the federal government in not exactly presently headed in a fiscal or political trajectory to make New Starts funding abundant. So, if GL is substantially better (which it might not be at all), we should talk concretely about the funding probabilities. If not, then a huge misallocation and mispricing of risk is hiding behind terms like ‘unlikely’.

      By the way, can you point me to a document that speaks to the minimum service life for MetroRapid vehicles on GL? I’ve been looking and can’t find one. I would really appreciate it.

      • Jeb says:

        I’ve heard that the MetroRapid vehicles are expected to have a useful life of 12 years and that the FTA expects an investment to remain in place for 20 years. It’s possible FTA may be open to replacing MetroRapid on a shorter timeline, but I doubt that you’ll find any document that gives you a definitive answer. For instance, there will be a new administration in four years, and we can only speculate on what the FTA’s policies may be then.

  6. The problem is that the FTA likely (I did it again!) does not have a table of exact probabilities for making a decision any more than the rest of us in our jobs. It will come down to whoever is assigned to work on our application. If you Google “rapid bus minimum service life”, you should find 2 relevant DOT documents. One is a white paper recommending that the government make provision for re-evaluating minimum service life requirements. The other is a comparison of rapid bus to light rail. Both estimate minimum service life at 12-15 years.

    I agree, we should probably have ballot language that states upgrading GL as soon as procedurally and financially possible, If that “clause” passes in 2014, it would likely take 10 years to complete the planning and funding process, which could start sometime around 2016-2018.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s