To be fair, I don't consider my research all that great. This is not a good methodology because it can be affected by hyperlocal effects. I just used the same approach as in papers that looked at the _benefits_ of bike lanes.

There is very little good-quality research on wider impact of bike lanes on business in the US. One pre-COVID study found little effect in either direction when a large LA neighborhood added bike lanes.

The only definitive result for the US is that bike lanes have very little impact on the absolute percentage of bike commutes (it stays around 2-3%).