In the decade before that (so 2006-2015) we had four accidental collisions and two anti-satellite tests (China and US)
Those are not frequent events. And if you attribute any statistical significance to them, it seems we are getting much better at avoiding collisions
There are other reasons we don’t currently experience major problems with collisions in space, and why airplanes sometimes do, but it is not this.
Here's some math:
Average Earth diameter: 12742kms + 10km Average airplane surface area = 500m2 12752^2*pi = 510,865,389km2 Surface flight/plane = 1021730 planes
Starlink orbit height = ~500km Surface at orbit = ~551,712,377 so ~8% increase (which is non-negligible) Average Starlink satellite surface area = ~7m2 Surface LEO/satelite = 78816053 satellites (77x compared to airplanes)
Daily flights 50k-100k. Total number of satellites <20k.
And this is only for Starlink LEO. If you go for higher orbits the surface grows substantially. Also satellites have predictable paths, altitudes, airplanes maneuver and turn, gain altitude/lose altitude. They gather around points (airfields) etc...
Conversely:
> Daily flights 50k-100k. Total number of satellites <20k.
Those 20k satellites orbit roughly every 95 minutes, so they're doing ~15 orbits per day, and even the longest flights from conventional aircraft are about half that distance, so by distance each satellite in LEO is doing strictly more than the equivalent of 30 flights per day each.
Research I'm doing for a blog post has shown me that the exact position of a satellite is surprisingly variable compared to what you'd reasonably expect from a "Newtonian spherical Earth with a perfect vacuum" approximation of the orbits, enough so that it makes sense to treat 1 km as the "collision avoidance manoeuvres needed" threshold.