 Solution

If gatherings causally raise infection risk and restrictions reduce birthday contacts, the birthday-event association should be smaller under strict restrictions and larger under lenient restrictions. A graded pattern across restriction intensity would be stronger evidence than a single binary contrast.