I find it hard to believe we waste more money administering welfare than we will transfer to those who really don't need it under the need-blind UBI.
Under UBI we'll be giving money to many, many people who need it less than the current welfare recipients, and even UBI will require administration.
A solution (UBI) that claims to help a group (the actual needy) that is not tailored to administer to that group is unlikely to benefit them, especially since it's now going to split the fixed amount of benefits to others not in the group.
>see, e.g. research on GiveDirectly's impact
What works in Kenya does not necessarily work at every location and economy. It's not like we have not studied this question in the US for nearly a century, and recent research on foreign third world cases does not overturn a century of research on first world economic aid.
And this is not an argument for UBI. At best it would be an argument to change current welfare programs to give cash aid to those already eligible for welfare.