The availability of participation income is broad enough to avoid most conventional issues with means-testing BI.

Whereas some see universal basic income as a catch-all alternative to both means-tested welfare and social insurance, Atkinson carefully positioned PI as complementary to unemployment insurance and state pensions. Social insurance would remain the default social security for many, enhancing the affordability and political feasibility of a PI that would ‘cut dependence on means-tested benefits’ (Atkinson, Reference Atkinson1996: 67). Atkinson’s key objection to means-testing was its application at household level in male-breadwinner systems. This denies women individual entitlement, a prerequisite for autonomy and equality. Income tests also create stigma and barriers to income support and lend themselves to paternalistic narratives of ‘deserving’ and ‘undeserving’ poor.

Ensuring an absence of stigmatising barriers is an important principle. However, both fiscal sustainability and political tactics may require that access to PI is limited by some financial threshold, while also ensuring stigma is at least mitigated by how any such threshold is applied. A PI income threshold, if not universal, can be designed so that it functions to selectively ‘screen out’ the most affluent citizens rather than as means of reserving support for only those suffering poverty and extreme deprivation. One could, for example, base circumstance thresholds on estimated minimum income or living wage standards, rather than close to the relative poverty line, as under conventional means-tested payments.Footnote 2 PI entitlements could also be gradually tapered on an individualised basis when a person’s (rather than household’s) income exceeds the relevant threshold. So, while a PI need not be universal, it must be widely available, individual, transparent to access, and set at a level that, in conjunction with access to universal basic services, enables people to meet essential needs.