CapMetro’s Board should utilize transit ridership and transit mode share to evaluate the performance of CapMetro’s CEO. Evaluation should focus on the public value created by the executive’s actions.
How does one determine “value”?
There’s a general consensus amongst transportation academics, practitioners, and activists that there are factors external to a transit agency that determine ridership and transit mode share performance (R&MS). There are also internal factors (i.e. management, local policy choices) that impact R&MS. It is impossible to do a metro-level randomized control trial on this topic, so determining the weight of each possible factor is left to sophisticated statical processing of data across metro areas.
Oft-cited research (PDF) by Brian Taylor et al. tackles this problem and finds that 74% of the variation across US metros can be explained by external factors such as population density, median income, Democratic-lean, and related levels of car-independence (e.g. bike, walk commutes).
About 26% of the observed variance in per capita transit ridership across metros is explained by internal management choices, specifically service frequency (frequency is good) and fare levels (cheap is good). Taylor et al. calculate service frequency by examining the National Transit Database and dividing an agency’s total vehicle miles traveled by its total route miles.
To determine the value added by agency management, one needs a model that spits out expected ridership or mode share given validated predictive factors. If an agency’s outcomes exceed the expected R&MS, then it’s likely the management is creating value. If the agency outcomes are substantially below the expected R&MS, then it’s likely the management is destroying value. Taylor et al.’s research indicates that developing such a model is feasible.
Can additional public value be created in Austin?
Many Austin leaders are skeptical of the notion that CapMetro can perform better. While they might not publicly state it, privately they stress that the agency is already entrenched in excellence.
Recently, CapMetro officials argued that in addition to the drag created by conventional external factors such as Austin’s sprawl, part of the reason for their ridership and mode share stagnation was a decline in the amount of bus service hours they could provide. This line of reasoning implies that the decline in service hours was a funding reduction of some kind beyond the control of management. However, a closer examination reveals that hours for bus service have declined as a result of management actions.
Real spending has actually been increasing but it has purchased fewer hours per dollar. The ‘Growth Index’ below expands on this previous analysis. All of the data referenced can be found here (Excel).
It’s possible that CapMetro’s deterioration in purchasing power is some kind of widespread trend across transit agencies. Examining the real dollar bus service costs for the largest Texas transit agencies undermines the validity of that explanation. The data below is only for bus and rapid bus/BRT expenses and vehicle revenue (‘in service, taking fare revenue’) hours.
The diversity of cost trajectories for bus operations shows that each transit agency is unique. At CapMetro, the period between 1999 and 2006 featured the agency losing its low-cost delivery structure as its per-hour cost ballooned past the average cost, where it’s remained. This loss of purchasing power due to contractual and management choices – from $80 per hour in 2015 real dollars in 1999 to $110 real dollars for the present hour cost – is a major explanation for the difficulty in providing additional service hours.
Further reducing CapMetro’s productivity is the shift of resources into the high-subsidy Red Line commuter rail. As the chart below shows, bus operations receive a smaller share of CapMetro’s budget once the Red Line launches. All of those commuter rail dollars trade-off with the frequency that Taylor et al. cites as boosting ridership; and they also contribute to the need for higher fares that Taylor el al. found to decrease ridership.
Austin’s City Council and the CapMetro Board each have their role to play in ensuring the CapMetro CEO delivers the most public value possible.
Council’s role is providing transit-supportive land use and prioritizing ridership-supporting public investments. It’s not glamorous, but building neighborhood corridors and employment nodes with enough density to support frequent bus service is the most feasible, fiscally-sound move to boost ridership and transit mode share. Taylor’s admonishment to local officials is relevant:
The CapMetro Board can shape a performance management culture and a CEO evaluation process that emphasizes public value created on R&MS. As part of this, they must settle on a rigorous model of what is expected ridership and mode share given the land use provided by Council, the economy provided by the private sector, and the service region’s other relevant demographic factors. The Taylor et al. model is one place to start. They must also empower the CEO to take on new tools – for example enable CapMetro to be a more active developer – as one way of solving Council’s recalcitrance in providing transit-supportive density.