Today is election day, who will win the cyclist vote? Well, I don’t know for sure, but based on Bikeside’s online survey, I can guess: Eric Garcetti.
Bikeside administered an online survey at the beginning of May, and we closed the survey 1 week ago. We received 641 responses to the survey, giving us a fairly rich picture of cyclist opinion in Los Angeles. I’m grateful to all the people who helped us get the word about this survey, and to everyone who took a few minutes out of their day.
Here’s a cold breakdown of how cyclists plan to vote:
Garcetti is the the big winner. Undecided is in the hunt. This chart really surprised me: I guessed that cyclists leaned Garcetti, but I didn’t expect it to be so dramatic. I was also shocked by the number of undecideds.
Let’s pull out all the non-voting folks and “neither” voters (see my quotes in the LA Weekly about that voting option) and clear up the picture:
It’s hard to know how close this is to actual cyclist opinion (see the discussion of limitations below.) Online surveys are pretty clumsy. Bikeside’s survey is no exception. Our sample was very lopsided
, with our 9 most common City of LA zip codes coming from Silver Lake (and adjacent), West LA, Northeast LA, Hollywood (and adjacent). And Greuel is supposed to be relying on the Valley vote, so this sample probably punishes Greuel. Unfortunately, with the existing data, there’s not much we can do about that.Worse, our survey is extremely lopsided ethnically. The picture tells the story:
Do cyclist demographics match city demographics? We don’t know, but this chart shows that survey demographics don’t match our survey demographics. For our picture of who cyclists’ are supporting to be any good, we need to try to match survey demographics to voter demographics. To make matters worse, voter demographics don’t match city demographics, for example: Latinos make up 48.5% of City of LA residents (2010 Census) but the USC Price/LA Times Poll indicates that Latinos will be about 24%
of the vote turnout. It’s a headache.There are ways to adjust for this. For example, if you conduct a presidential poll, and you get two male responses for every female response, but you know that likely voters are 50% men, and 50% women, you can compensate. All you do is value the responses of women to your poll twice as much as men. Suppose we compensate for our lopsided racial demographics in the same way, adjusting our results to reflect what would happen if the racial composition of voting cyclists matched the City of LA’s racial composition? It’s not likely to turn out that way, but let’s try it for fun. Here’s the predicted cyclist vote:
Essentially nothing changes! Making this chart was a fun math exercise for me, but turns out to be not useful.
Another question I had: is a cyclist’s perception of bike friendliness connected to their vote.
It appears that whether a cyclist felt LA is bike friendly or not is not strongly connected to their vote. I dug into a bunch of other ways cyclist opinions might be connected to their choice for Mayor. I haven’t found much right now, although it appears that Garcetti has an advantage, above and beyond his already considerable advantage with cyclists, with people who value CicLAvia especially highly. But it’s not huge. I’ll keep digging.
One last chart! Greuel supporters are partly counting on women to come out big for Greuel. Will women support Greuel decisively over Garcetti? I don’t know, but I can tell you that’s not true among women cyclists… at least those that took the survey. However, women cyclists do swing more toward Greuel than men:
It’s still not enough to win the cycling vote though. By the way, we had two respondents who gave “transgendered or other” as their gender, 100% for Garcetti.
There are huge limitations to online surveys. The first is that you capture mostly frequent and sophisticated internet users. Another is that our survey was in English only, but simple observation suggests that there are a lot of Spanish-only speakers who are cyclists. We’re not capturing their opinions here at all. As well, given the way we promoted this survey, primarily through social media, we’re mostly getting responses from people who are tied into my social network, or one near to it. That misses a lot of cyclists.
Huge limitations means huge opportunity for improvement. I’ll be writing up the basic results of the survey – is LA bike friendly in the eyes of LA cyclists, and how has that changed since August 2011 – in my next installment.