Is it time to re-imagine Houston's re-imagined transit network?

I've spent a lot of time being critical of Houston's new transit system. I suppose it's because I'm the rare person who likes both transit and Houston; I just don't much care for the transit system that exists in Houston.

The month the new system was unveiled, I wasted no time in criticizing it, not so much for what the system was designed to be but that those who planned the system failed to understand the history and context of what led to Metro's ridership decline in the first place (hint: it involves land use and demographic change, coupled with a massive expansion in the highway network). A few months later, I started digging into the nuts and bolts of the system by testing it out, and I declared it a flop. By mid-2016, I was able to dig into some of Metro's performance statistics. While transit geeks around the country were still admiring the new and improved system, the preliminary data and my own misadventures with the system suggested otherwise.

And because I really would like to see transit in Houston succeed - and by transit, I mean that I am mostly focused on the bus system - I decided to take a look at the data now that the new network has been in place for over a year.

One of the reasons Metro wanted to re-imagine its system was to stop losing ridership. Did Metro accomplish that goal? The weekday and monthly boarding counts suggest not. The table below shows Metro's ridership a year before the service change (2014), immediately after it took effect, and a year later (2016). In fact, the only thing that saved Metro's poor performance was growth in the LRT system (an extension plus two new lines, coupled with continued ridership growth on the original line). 

Table 1: Ridership Change

More telling than the ridership data is Metro's farebox revenue. In theory, the old radial system offered more one-seat rides to the primary downtown transit hub than the new, more grid-like system of frequent routes which required people to transfer between more routes to make the same trip. Even if the number of riders remained the same, the number of boardings made on the system should have increased, while farebox revenue while would remain flat (mathematically, it reflects the same number of passengers that pay for the first trip but transfer between an increased number of routes at no charge from origin to destination). Even if you accept Metro's 3.1% growth over two years, it is still losing ground in a city that's growing by 1.8% annually (that's the city's modest growth rate, not the region's much higher rate of sprawl).

However, monthly farebox revenue dropped from $6.7m in September 2014 to $6.2m in 2015 and $5.4m in 2016 (source: Metro Monthly Financial Reports). This is a staggering 19.4% loss in fare revenue from the old to new system that suggests there has been a huge decline in the actual number of people using the system. Meanwhile, system operating costs (excluding allocations to non-operating expenses) grew during that same period from $41.9m to $50.7m, or 21.2% increase over the same two-year period.  Metro does not separate costs by mode in its monthly reports, but this change is largely a result of an expanded rail system. With a loss in fare revenue, lower ridership, and higher operating costs, the average fare collected per passenger fell by nearly 22%, and the operating subsidy per passenger increased by 25%.

Even as many praise Metro's new system, the reality is nothing short of horrifying. Other performance measures also show that Metro's system continues its state of decline. One thing to note: Metro's "new" and re-imagined system only changed the system of local and express buses; the commuter bus system saw relatively few changes in the 2015 "re-imagined" system, and the LRT network had been planned long before Metro decided to shake things up.

Table 2: Bus System Performance

Source: Metro Monthly Ridership Reports

Why has Metro failed to keep up?

• Is population a factor? Not likely. The Census shows 1.8% growth in city population from April 2014-2015, while the City's own estimates shows 4.0% growth between January 2015-2016. The two sources differ from each other in total population by nearly 5% in 2015, but both show robust growth within Houston's city limits during as Metro's ridership was falling.
• Are oil prices a factor? It is possible, and probably likely, that ridership on the commuter bus system has fallen with lower oil prices. Obviously, lower oil prices make commuting by car cheaper, but the single largest transit-using employment sector in downtown Houston are employed in "mining" - that is, they are oil company office workers, and low oil prices likely fueled a cut in the workforce. The same does not apply to the local system, where buses serve primarily a transit-dependent population.
• A change in service levels? Possibly. Metro does not publish how much service it puts on the street, but it does provide data that allowed me to figure that out. Metro's monthly reports show boardings and route productivity measures, and I built a table for each route in each of the three years to compare how much service was going out on the street. My results in this area were somewhat inconclusive.

Table 3: Service Levels

Source: Calculated from route-level ridership and productivity, as provided in the Monthly Ridership Reports.

• Increased service on commuter bus routes: As shown in the table above, Metro increased both miles and hours of service on the commuter bus service, but ridership declined (again, oil prices and related downtown employment may be to blame).
• Changes in the local bus network: The picture is less clear for the local bus system, where Metro has increased its revenue miles of service but reduced revenue hours of service. The change in proportion of distance to time implies that Metro's average operating speed should have increased, but it actually fell. Alternatively, Metro could have increased the ratio of express to local services, but that ratio actually fell, too. So how did Metro increase miles of service while reducing hours of service? As I noted in one of my earlier posts, the new Metro network eliminated semi-express portions of local routes (e.g. 53 Westheimer/Briar Forest, 56 Airline, and 88 Hobby operated partly in freeway/HOV service), while today's local routes operate only on urban streets.  Moreover, some popular express routes that were part of the local system (e.g. 163 Fondren with 10,300 weekday boardings) were eliminated. These routes had operating speeds higher than the local system average, and eliminating them may have helped reduce Metro's average speed from 12.9 mph (2014) per hour to 12.1 mph (2016) for the overall local bus network. But miles increased while hours fell, suggesting that average speeds should also have increased. The only possible explanation I have here is that Metro reallocated local bus services from urban areas that are more congested to suburban areas that are less congested. If true, this would represent a shift of service from higher demand areas to lower density, lower demand, and pedestrian-unfriendly suburbs. In an earlier post, I noted one example (Alabama) as a kilometer-wide higher density central city corridor without a transit route, but the data suggest that perhaps more transit service has left the central city than I previously realized. A shift from urban to suburban transit would represent another failure to understand the land use-transit connection and may explain why local bus system ridership and farebox revenue continue to fall at an alarming rate.  Metro really needs to undertake a comprehensive review of its local bus network to understand its declining performance.

Metro's own data seem prove that its re-imagined system is slower, costs more to run, and attracts even fewer riders than its old system. Houston deserves a better transit system, and it's probably time for Metro to re-imagine its re-imagined local bus network. Hopefully, Metro will address the challenges with shifting demographics, a changing urban context, and the fact that maybe it should start using its "general mobility" funds on pedestrian infrastructure rather than building more roads. Houston deserves better.