AGN outflows and feedback twenty years on

It is twenty years since the seminal works by Magorrian and co-authors and by Silk and Rees, which, along with other related work, ignited an explosion of publications connecting active galactic nucleus (AGN)-driven outflows to galaxy evolution. With a surge in observations of AGN outflows, studies are attempting to test AGN feedback models directly using the outflow properties. With a focus on outflows traced by optical and CO emission lines, we discuss significant challenges that greatly complicate this task, from both an observational and theoretical perspective. We highlight the observational uncertainties involved and the assumptions required when deriving kinetic coupling efficiencies (that is, outflow kinetic power as a fraction of AGN luminosity) from typical observations. Based on recent models we demonstrate that extreme caution should be taken when comparing observationally derived kinetic coupling efficiencies to coupling efficiencies from fiducial feedback models.