This study of shops in downtown Vancouver did find a net decrease in sales after the implementation of a separated bike lane. But the analysis relied on business surveys, rather than actual sales data, which might have led to a response bias among the merchants who took the biggest hit. The little sales data that was received “indicated that the estimated loss in sales was not as high as reported in the surveys.”
Despite efforts to increase response with follow-up telephone calls, there is some degree of uncertainty about the randomness of the results obtained.
Surveys were conducted with 61 merchants and 538 patrons on Bloor Street in Toronto. It was found that only 10 percent of patrons drove to the shopping area, and that those arriving by foot and bicycle spent the most money per month. Report authors concluded that converting street parking into a bike lane in the area was “unlikely” to have a negative impact on business and that, on the contrary, “this change will likely increase commercial activity.”
All case studies here.