A new report from the UCLA Institute of Transportation Studies finds that the main cause of declining ridership in southern California is poor people buying cars. Between 1990 and 2000, when ridership was growing, the Los Angeles region grew by 1.8 million people but only 456,000 cars, or about one car per four people. Between 2000 and 2010, when ridership was shrinking, the region grew by 2.3 million people and gained 2.1 million cars, or nearly one car per new person.
There is certainly something to this, but other factors are probably more important than the report estimates. The report says that neither ride sharing nor changes in transit service and fares have played an important role, and I suspect these conclusions are wrong.
The report shows that transit trips per capita peaked in 2007 and have declined in most years since then. Certainly the decline before around 2012 or 2013 was not due to ride sharing. But the decline steepened after 2014, and I suspect much of that decline is due to ride sharing.
With regards to service levels and fares, the report concludes that “Changes in transit service and fares have mostly followed and not led falling ridership.” The authors base this on combined data from the region’s transit systems as a whole. But most of the decline has been in bus ridership on LA Metro, the region’s largest transit system, and a close look at the Federal Transit Administration’s historic time series reveals that LA Metro bus ridership in fact followed changes in service levels and fares.
Vehicle-miles of service peaked in 2006 and dropped just 0.7 percent in 2007. Despite the drop, 2007 ridership grew by 3 percent. But then, after several years of fairly constant fares, Metro hiked fares by nearly 14 percent in 2008, leading to an immediate 7 percent decline in ridership. This was followed by a 2 percent drop in service in 2009. Riders responded by reducing trips by 5 percent in 2010. It appears that fare increases led to an immediate decrease in ridership while ridership responses to service cuts lagged by about a year.
At first glance, data seem to back the report’s conclusion that increasing auto ownership played a role in ridership declines, but the timing is curious. In 2000, 11.4 percent of Los Angeles urban area households lived without a car. By 2007, this had dropped to 8.5 percent. Renters (who tend to have lower incomes than homeowners) without cars declined from 18.4 to 14.7 percent. Between 2006 and 2016, those without cars fell a little more to 8.1 percent while renters alone fell to 12.7 percent. So while the increase in car ownership was considerable, most of it was before the ridership peak in 2007.
Table B08119 of the American Community Survey shows that, between 2006 and 2016, the number and percentage of low-income commuters taking transit declined while the number of high-income commuters increased. In the Los Angeles-Anaheim urbanized area, the share of workers earning less than $35,000 a year commuting by transit declined from 9.5 to 9.0 percent (representing a loss of 57,200 commuters), while the share earning more grew slightly from 2.20 to 2.23 (representing a gain of 23,600 commuters).
Table B08119 doesn’t break down bus vs. rail transit. But FTA data show Los Angeles Metro increasing rail service as it decreased bus service, with increasing rail ridership offset several times over by decreasing bus ridership. It is not hard to imagine that the growth in higher-income riders is mainly on rail while the decline in low-income riders was mainly on buses. The agency apparently made a conscious decision to sacrifice low-income riders so it could gain more high-income riders.
My conclusion is that keeping fares low helped maintain ridership in the face of increasing auto ownership before 2007. But a stiff fare hike in 2008 led to a vicious cycle of declining ridership and service cuts that mainly affected low-income bus riders. The rapid growth of ride sharing after 2013 is probably eroding high-income riders, but this hasn’t shown up in commuter numbers yet so if it is true it must mainly affect non-work-related travel.