A summer in Sioux Falls I:
Fighting to increase transit ridership in the American Midwest
During the summer of 2019 I will be in Sioux Falls, South Dakota, helping City Hall reform its public transportation system.
The largest city in America’s third least populous state has 12 bus routes and a paratransit system ran by the Sioux Area Metro (or SAM for short). Transit runs roughly from 6 AM to 9 PM, with limited service on Saturdays and no service on Sundays. Bus operations are contracted out to a company based in Ohio called First Transit.
The problem is that over the past few years the city has been spending more and more cash while less and less people ride the buses.
Sioux Falls bus ridership is steadily decreasing
Bringing Human Centered Design to Transit
Mayor Paul TenHaken is a business leader and entrepreneur who recently began serving his first term in office. As part of a collaboration with the Harvard Bloomberg City Leadership Initiative, he directed a team of municipal employees, including a transportation planner, several financial analysts, the city’s fire captain, among many others, to conduct a human-centered design innovation project to find new ideas for the transit system. Since November, the group has been hard at work doing user research, holding ideation sessions, riding along the different bus routes, interviewing local stakeholders, and reflecting on the insights gained.
The underlying question that drove the group was “how can we improve the economic benefit of the public transit system for both the residents and the city?”. And after several months of looking at the system’s strengths and weaknesses, they landed on six different innovations that require further testing:
· Using artificial intelligence and analytics for transit planning
· Provide on-demand information about the bus system
· Partner with businesses to extend service
· Get employers to provide transit incentive programs
· Increase the budget allocated to transit
· Expand fare payment options (i.e., not needing exact change for a bus ticket)
Agile methods in government
In preparation for the summer, I have been reading The Lean Startup by Eric Ries, which was required reading for David Eaves class on digital government and public sector innovation at the Harvard Kennedy School as an introduction to Agile management. Although Ries is a digital entrepreneur type, he is also quick to point out that the ideas in the book are supposed to help anyone that wants to steer any kind of organization into embracing innovation. Could Agile methodologies allow public sector organizations to be more willing to try new ideas that create tangible public value? And what obstacles may they find along the way?
Here are three ideas to focus on during my 10 weeks at Sioux Falls:
The Idea — Fast Iteration. We can think of a local government as a collection of established enterprises (that must deliver the same services year after year as efficiently as possible) and startups (that must come up with new services or improvements). Ries writes that “because startups often accidentally build something nobody wants, it doesn’t matter much if they do it on time and on budget. The goal of the startup is to figure out the right thing to build” (emphasis mine). The public sector is not safe from creating services that nobody wants and in the face of extreme uncertainty, a more experimental approach to new ideas (fast iteration) would help us figure out if we are indeed building the right thing.
How do we do this? Through quick “Build, Measure, Learn” cycles. Instead of spending months planning a policy that is then validated by a pilot, the lean startup method demands lower fidelity prototypes that allows us to “focus on what costumers want (without asking them)”. Because it is hard for costumers (or citizens for our purposes) to accurately gauge before-hand if a new service will provide value for them, focus groups and surveys are limited when ideating novel solutions. Prototyping involves building rudimentary and imperfect versions of our proposed solution, measuring with concrete data its results, and learning from that experience whether to “pivot or persevere”. The faster you can make this cycle, and the more cycles you go through, the more you can ensure that the innovation can deliver value.
Auditable innovation accounting. The author is weary, though, of vanity metrics. If the measure phase of your prototyping is not looking at the right indicators or, put it another way, you are not able to make meaningful decisions based on the new information, then you risk only giving the appearance of learning. Innovation requires a rigorous approach to measuring, where you set a baseline and then see how different alternatives improve an outcome or not (thus helping you make the decision whether to pivot or persevere). A/B testing (taking two subsamples of our target population and measuring one tweak against the status quo) is king.
Concerns
Could the lessons from The Lean Startup help give weight to the prototyping phase of Sioux Falls transit innovations? I will keep you updated, but for the mean time I do have some lingering concerns:
· “Economic benefit” is too vague as an initial problem statement and could lead to contradicting answers. Is the city spending less money an economic benefit? If people spend more money but spend less time in traffic, would they perceive it as a benefit or as a loss? There is a large academic literature on people’s willingness to pay to reduce travel time. Still, precision in what the project wants to accomplish would give a direction and a goal. While allowing the team to anticipate potential trade-offs such as decreasing waiting time to citizens while costing more money to the city.
· Not all the innovations proposed by the team so far seem to be at the same level of maturity. Applying artificial intelligence (i.e., training algorithms using vast amount of data) is a tool that only becomes an innovation depending on what you use it for. Increasing the budget for transit, on the other hand, is not so much an innovation as it is a policy choice, and thus not really fit for prototyping.
· Finally, I worry that throughout this laborious work by the Sioux Falls team, not enough attention has been given to some of the disruptions that threaten transit systems, specially in rural areas. Lyft and Uber have only arrived at the city during the last couple of years, coinciding with declines in bus ridership. Once autonomous vehicles roll out and bring down the costs of private rides even further, the pressure on the system would be even higher.