Alon Levy challenged the assumption underlying my statement:
I'm willing to believe that the majority of New York City voters own cars, and even drive them to work. The city has a car-free majority, but I'd venture a guess that this majority is disproportionately immigrant and poor, and hence not likely to vote.
It's a good question, and it got me thinking about how you could test this. I checked the American Community Survey, and they have citizenship and poverty data cross-referenced with means of transportation to work.
|Borough||Percent of workers taking transit to work||Percent of citizen workers taking transit to work||Percent of workers above 150% of the poverty level taking transit to work|
|The Bronx||56 %||54 %||54 %|
|Brooklyn||60 %||59 %||60 %|
|Manhattan||57 %||57 %||57 %|
|Queens||51 %||47 %||50 %|
|Staten Island||32 %||30 %||31 %|
|New York City||55 %||53 %||54 %|
If you look at non-citizens and the poor, they're both significantly more likely to take transit to work, but since they're relatively small minorities, they don't affect the overall totals more than a few percentage points - and the spread tends to get bigger the sprawlier the area.
So the only boroughs where transit riders are a minority of non-poor citizens take transit are Queens and Staten Island. In the other three boroughs, and overall, transit riders are a majority. Even in Queens and Staten Island, though, I didn't look into how many of the non-transit riders were actually driving as opposed to walking, biking, taking taxis or telecommuting.
It's possible that the ACS is not very accurate for these communities, and if anyone knows about that, please post it in the comments.
Of course, when you get beyond citizens to voters, and beyond voters to campaign contributors and people who are likely to bring others out to vote for a candidate, then things start to look different...