Precision—‘basic income’ is a term of art—universal and unconditional. Things like the (Aff/PIC) which affect who is eligible or attach preconditions are distinct from the variation among experiments along vectors like frequency or size of payments. The latter are fair game, the former are not.

Malcolm Torry, London School of Economics, A Research Agenda for Basic Income, Elgar Publishing, 2023

The problem with the word ‘guarantee’ is that to guarantee an income can either mean to guarantee that a particular income will be paid as a Basic Income, or that a stated income level will be reached by some means that might not imply a Basic Income. The two possible meanings of ‘guarantee’ in this context can result in ambiguities in the literature that can leave the reader not knowing whether what is being discussed is a guaranteed Basic Income or a Minimum Income Guarantee (Grevc, 2017: 95, 127; Torry, 2021a: 7-9): and so, for example, a sentence such as The idea is to guarantee every citizen in the country an unconditional income sufficient to meet some minimal threshold’ (Walker, 2016: 3) can be read either as meaning that an unconditional income is given to every individual, or that every individual or household is given an income-tested benefit to enable them to reach a prescribed threshold. If the reader is aware of the potential ambiguity then they might be able to work out what the author intends in a particular context, but if they are not, and they bring to the text a presupposition that to guarantee an income implies a Basic Income, then they might completely misunderstand what the author has in mind: for instance, if a Minimum Income Guarantee experiment is what is being discussed.

Particularly problematic is the use of the term 'Basic Income’ for some- thing that is not a Basic Income. For instance, Pitts et al (2017) regard the Speenhamland reforms of 1795 as ‘a kind of Basic Income’, and argue that a Basic Income would encounter the same problems as those reforms, whereas the Speenhamland experiment was with household-based and income-tested with an annual dividend (Canadian Centre for Economic Analysis, 2020), and the recent short-lived Ontario experiment called a household-based and income-tested benefit a ‘Basic Income’ (Mendelson, 2019: 3, 22; Ontario, n.d.). Duverger has called it a ‘revenu de base sous condition de ressources’, ‘a resource-conditional basic income’ (Duverger, 2018: 129): a valiant attempt to encompass what it called itself and what it was. A similar attempt to qualify the noun ‘Basic Income’ can be found in the Australian proposal for an ‘affluence-tested Basic Income’: a means-tested benefit that constitutes a Minimum Income Guarantee (Spies-Butcher et al., 2020). It is of course perfectly legitimate to qualify ‘Basic Income’ in relation to those characteristics that constitute variants: for instance, in relation to territorial area, amount, and periodicity. An example would be ‘a weekly partial European Basic Income’. What is not legitimate is to qualify it in relation to an unpermitted condition. A ‘means-tested Basic Income’ is not a Basic Income; a ‘household Basic Income’ is not a Basic Income; and a ‘work-tested Basic Income’ is not a Basic Income.