Many of us who own cars are too lazy to move them every couple of days for street cleaning, so they are parked in attended garages or lots, which deters (though isn't a guarantee against) theft.
Some parts of NYC have higher theft rates that get lost when blended into the overall city rate. Brooklyn for example has a much higher theft rate than the rest of the city.
There is also a strong counter-argument to thieves preferring bad weather. Basically, it goes: when you are a thief, you want to do your trade with as few witnesses as possible, and to obscure any visibility of your actions to the witnesses there are. So, if you are thieving outside, you want it to be at a time when others want to be inside, and/or can't see you. Largely this means at night. Better if the weather is gross, so people are less likely to go out. Better still if they can't hear you over the rain. (If you are robbing suburban homes, you want to do it in the afternoon, after lunch, when most people are gone at work, and before the kids get out of school. For this example, the crime statistics I've seen, and anecdotes from cops back up the theory).
Interestingly, New York saw its car theft rate declare years after years since the 90s. San Francisco's theft rate also declined, but it does so rather unevenly.
Alone yes, you might however be able to say that it has meaning if you can show that the registered cars per capita stayed at the same level over that period.
I'm not positive that would be what you'd need to say that but It sounds right in my head, anyone have any better ideas?
http://newyork.areaconnect.com/crime/compare.htm?c1=san+fran...
Of course this was burglary or theft, not vehicle theft. But I wonder what makes for such an extreme difference.