When some disruption or unusual event causes people to shift from driving to transit, many never shift back. Ezra Klein Brad Plumer reviews the evidence in the Washington Post.
He's talking mostly about shifts caused by gas price shocks, but something similar happens in response to major disruptive events. For example, the 2010 Olympic Winter Games in Vancouver caused a major burst in ridership -- obviously a mix of Olympic visitors and residents who were trying to avoid Olympic traffic. But ridership never dropped to pre-games levels, and in fact, 2011 ridership was higher than 2010, despite the huge influx of Olympic visitors in 2010.
Klein goes on to lament that the very fuel price volatility that affects transit ridership also affects transit's funding, since federal funding is based on fuel taxes so they drop when fuel use drops. Unfortunately, US local operating funding (which is the real crux of the matter) is even more volatile, depending typically on payroll or sales taxes. Loss of a job equals a drop in payroll taxes, and causes drops in spending shortly afterward.
I wonder if we'll eventually create something like a property tax surcharge that captures some of the benefits of transit to a location -- possibly based on some future, vastly more objective Transit Score or index of transit access. Road funding could work the same way, but tending to fall more on the properties that benefit least from transit, since higher road use correlates to lower transit use. Property taxes (inevitably passed through to renters) are the least volatile funding source around, and if you want your transit agency to work on real service improvement, instead of endless cycles of cuts, adds, and cuts, we'll have to find our way to a more stable funding solution.