Bikes ... slow traffic, which creates more pollution.
removing auto lanes or narrowing them to accommodate dedicated cycling infrastructure tends to have a traffic calming effect, which often significantly increases traffic congestion.
There is, of course, a large amount of evidence to the contrary.
In one study, "Research On Bicycle Conversion Factors,” bikes were calculated to cause as much congestion as 0.28 passenger cars (less on trails, more when turning left).
In an AASHTO study, it was calculated that the passenger car equivalent for a bicycle depended on the width of the lane, but can be as low as 0.
A third study, by Heru Sutomo, places the PCE for bikes in developing urban areas at 0.19.
The Victoria Transport Policy Institute translated this into a per vehicle mile cost. They calculated the cost of bicycle-induced congestion as 0.2 cents per vehicle mile. Passenger cars are more than ten times as high at 3.5 cents per mile. Even taking into account the 1.6 passengers per car and assuming one passenger per bike (probably a little low) they per passenger cost becomes 0.2 cents per passenger mile for bikes and 2.2 cents per mile for cars - 11 times larger.
As for bike lanes, even when they result in closing whole roads, they might - by themselves - reduce congestion. And at an 11 time improvement, you don't have to induce many people to bike to cut down on congestion. Bikes don't cause congestion, they reduce it.
Cars are another issue.