> "Countries with legalized prostitution are associated with higher human trafficking inflows than countries where prostitution is prohibited."
But:
(1) Can inflows really be measured comparably between legalized and non-legalized regimes? and
(2) Are inflows actually the figure of merit or a proxy?
(3) If inflows are a proxy for the figure of merit, is the relationship between that proxy and the actual figure of merit consistent between legalized and non-legalized regimes?
Its worth noting that key utilitarian arguments made for legalization (there are also more direct deontological arguments) include that legalization makes it easier for people subjected to abuses within sex work (including, but not limited to, trafficking) to identify themselves and/or be identified, and for them to escape from those abusive situations. If that is true, one would expect:
(1) Any mechanism for detecting trafficking levels (including inflows) to return a result higher relative to the actual level, ceteris paribus, in a legalized regime than a criminalized one, and
(2) Actual inflow levels in a legalized regime to be higher in proportion to impact levels (number of people subjected to trafficking at any point in time, number of person/years per year spent subjected to trafficking, lifetime average number of days spent subjected to trafficking, etc.) because shorter time to escape would require higher inflow levels [0] for any given impact level (or, equivalent, produce lower impact levels for any given inflow level.)
And this doesn't even address that the research uses a source that relies on local reports of human trafficking rather than consistent collection on common standards, and that legalizing regimes (and particular democratic legalizing regimes, which saw the greatest likelihood of increases) might simply tend also to use broader definitions of human trafficking in the first place, or expend more effort to either detect it or to distinguish it from non-trafficking prostitution.
[0] Well, where "inflow" is imports + people newly subjected to trafficking locally; the paper seems to consider only imports, which would ceteris paribus have the same relation, but there could be compounding effects from differences in local trafficking (OTOH, if legalization made it harder to traffic locals--because those who would be at risk would be more able to turn to legal sex work themselves--then that would further increase the ratio between import inflows and domestic impact levels.