Quasar - galaxy associations

There is controversy about the measurement of statistical associations between bright quasars and faint, presumably foreground galaxies. We look at the distribution of galaxies around an unbiased sample of 63 bright, moderate redshift quasars using a new statistic based on the separation of the quasar and its nearest neighbour galaxy. We find a significant excess of close neighbours at separations less than about 10 arcsec which we attribute to the magnification by gravitational lensing of quasars which would otherwise be too faint to be included in our sample. About one quarter to one third of the quasars are so affected although the allowed error in this fraction is large