Specification is especially required for a UBI – it determines implementation.

Interest in UBI is on the rise in the United States and other advanced countries. Decades of wage stagnation and concerns about automation, robots, and job destruction, as well as discontent with the current social safety net, provide the foundation for interest in this area. Support for UBIs has led to several pilot programs and policy proposals in the United States, Canada, Finland, and Switzerland. Despite all of this, there is a lack of clarity on what makes a UBI, what problem it is meant to solve, whether the social safety net can provide or is providing these benefits, and what (if anything) can be learned from the pilot programs that we do not already know from the decades of existing research on individual and household responses to the social safety net, and on wages and income opportunities more broadly. Our article seeks to fill this gap. A pure UBI (providing a set benefit to all regardless of income, age, etc.) funded to meet basic needs for a household without earnings would be extremely expensive, about twice the cost of all existing transfers in the United States. Funding it would require substantial new revenue. The source of the new funds is a first-order issue and will have substantial impacts on the distributional effects of the policy and its ability to target those most in need of assistance. In particular, replacing existing antipoverty programs with a UBI would be highly regressive, unless substantial additional funds were put in. Much about the effects of a UBI on labor supply, income, and family well-being can be gleaned from the existing research, which we have briefly reviewed here. We have identified a few outstanding questions, such as the impact of a truly universal program (presumably without stigma) as well as the effects on human capital and, hence, labor supply in the longer run. Unfortunately, the planned and ongoing pilots are not well suited to answer these questions. Experimentation aimed at identifying parameters and mechanisms (Ludwig et al. 2011, Rothstein & von Wachter 2017, Deaton & Cartwright 2018) would be more useful than evaluations of small UBI pilots.