Ridership in Riverside

A reader, Helen, sent me a link to a Planetizen article showing that the Riverside-San Bernardino-Ontario Metropolitan Statistical Area (which covers the entire counties of San Bernardino and Riverside) ranked third in the nation for growth in public transit usage between 2006 and 2008, according to the American Community Survey of the U.S. Census Bureau. The article reports a 26% increase in transit commuting in the IE, and favourably quotes RTA's Brad Weaver. (I should note that RTA isn't the only transit agency in the area, which is also served by Omnitrans, MARTA, VVTA, Pass Transit, Sunline, MBTA, Palo Verde Valley Transit, Barstow Area Transit, Needles Area Transit, Corona Cruiser, Metrolink, and on the edges by Foothill Transit and OCTA.) These data are undoubtedly indicative of a positive trend, but the measures used by the study are questionable, and the sampling size must be outrageously small. We know that only a few percent of the residents of our area actually take transit, and this study is based on the self-reported commuting habits of survey-takers. The likelihood that, say, a thousand randomly-sampled IE residents contains more than 10 or so transit riders is rather small, and so the 25% increase they're reporting could very well be from 10 to 12. Also, the article mentions the margin of error for the study, which they say is under 15%. So that means the actual increase in transit ridership could be anywhere from 10% to 40%.

Now, I happen to have found all of RTA's ridership reports (with the curious exception of September 2007) from December 2004 until now, and took some time today while watching Maddow to plug real ridership numbers into a nice, chunky spreadsheet, and then into my statistical analysis software (which produces really nice graphs). Let's see what we see, shall we?



Here's the ridership data from the study period, January 2006 through December 2008. The blue line is the reported totals, and the red line represents what we call a "regression line", which is a way of representing the average relationship between two variables, in this case time (in months) and ridership. What we see here in this graph is actually consistent with the reported Census Bureau data- ridership increases from ~520,000 to ~650,000 (looking at the regression line, which smooths out the seasonal variation), which is a 25% increase. Hurray!



If we look at the data from the entire period I have available to me, however, we see some disappointing mitigating facts. Here's the graph for 2004-2009, and you'll notice that the regression line is shallower, and that it has approximately the same two endpoints. Before 2008, the data essentially vary around a mean of 520,000. You can see that there was a large increase in ridership through the summer of 2008, ending in a huge spike in October 2008 (probably exacerbated by UCR U-Pass riders returning), and then settling in to vary around a new mean in the ~650,000 range. What this says is that, while RTA seems to have managed to hold on to many of the riders they gained in 2008, ridership has not continued to grow at the same pace. All of the variation in the data is explained by that one huge increase in the summer of 2008 (when gas prices were exorbitantly high), and RTA hasn't maintained that momentum.

Of course, this isn't to say that this is entirely unexpected. Since the summer of 2008, we've had a fare increase and significant service cuts from RTA. That they've maintained the ridership gains they made two years ago is impressive, and it shows how powerful just getting people to try taking public transit is. However, it's important to note that big percentage increases don't mean all that much when you're starting from such a low point. A 25% increase from 1% only gets you to 1.25%, and I'm not even sure transit had 1% mode share in 2006. I'm glad we're going somewhere, but we've got a long way to go.