Psychology proves – conditional tests increase the popularity of social assistance programs, especially PI.

The second policy design dimension, conditionality, refers to so-called ‘conditions of conduct’ that are attached to benefit receipt (Clasen & Clegg, 2007). The ideal-typical definition of basic income is very clear on this: once entitlement and eligibility are established, the cash payment should be fully unconditional, with no requirements involved whatsoever. However, as discussed in Chapter 1, there have been alternative proposals for basic income varieties that do impose some form of conditionality on the recipients. The best-known of these is arguably the participation income originally formulated by Anthony Atkinson (1996) that is conditional on a number of agreed-upon, socially appreciated activities. In its most comprehensive form, a participation income can cover a very broad range of activities, including caring, training, working, volunteering and so on. In its strictest form, only the performance of paid work suffices to receive the benefit. Although many basic income proponents reject the idea of having any conditions attached to it, and especially any work-related conditions, some may accept such conditional types of basic income because they see them as a promising stepping stone policy towards the fully unconditional version.

This was not the case for Atkinson (1996), who proposed his participation income because he believed it would stand a better chance of being publicly accepted and thus being politically feasible. In fact, there are several signs in existing opinion research that suggest that Atkinson might have been correct. First, the literature on welfare deservingness shows that both control and reciprocity are hugely important normative criteria that people take into account when deciding who should get what from the welfare state (Laenen, 2020; van Oorschot, 2000). More concretely, this means that people are generally less supportive of welfare provisions for those who they see as not having contributed sufficiently (reciprocity) or as being responsible for their own neediness (control). Some even go as far as suggesting that it is a fundamental psychological trait of human beings to categorize help-seekers (and thus welfare claimants) into two distinct categories: first, those who are willing to contribute to the community but have been affected by unlucky circumstances beyond their control (the ‘reciprocators’), and second, those who are unwilling to do so out of laziness (the ‘cheaters’) (Petersen, 2012). In this co-called ‘deservingness heuristic’—which supposedly applies universally across countries, time periods and social groups (Aaroe & Petersen, 2014)—free-riding cheaters should be denied access to any welfare provision, presumably therefore including a basic income. Second, prior studies have shown that work ethics, understood as a set of values centred on the importance of employment in a productivist society, continue to be key in mainstream public opinion (Stam, 2015). Although post-productivist streams (which aim to de-emphasize the importance of wage labour in particular) are certainly emerging, these appear to be undercurrents at present, making it unlikely that the general public will readily accept a fully unconditional basic income (Kozàk, 2021).

Prior studies have consistently found overwhelming support for imposing work-related obligations on the (mostly able-bodied) recipients of current welfare benefits, and this is known to be strongly rooted in deservingness considerations and work ethics (Buss et al., 2017; Laenen & Meuleman, 2019). Accordingly, it makes sense to assume that conditional varieties of basic income will be more popular than their fully unconditional counterparts. By and large, this is also confirmed by the evidence from surveys, although there are important exceptions. Let us start with the BNES experiment conducted in Belgium, because it gauges (on a scale from 0 to 100) support for three types of schemes with varying degrees of conditionality: first, a fully unconditional basic income, second, a participation income that requires people to perform care tasks or to volunteer (if they are not in paid employment), and third, a basic income that is conditional on job-seeking efforts. Overall, the findings show that there is somewhat greater support for the conditional schemes (the mean is 53 for both schemes) than for the fully unconditional one (the mean is 47)—both before and during the COVID-19 pandemic (the reported means are pre-pandemic, see Fig. 3.2). There are nevertheless some interesting regional differences (see Fig. 3.3). While support for the three basic income varieties was relatively stable in the Dutch-speaking region of Flanders before and during the pandemic (except for a small decrease in support for the scheme that is conditional on job-seeking), there was a clear increase in support for all three types in the French-speaking part of Wallonia. This increase was larger for the unconditional variant (from 44 to 57) compared with the schemes that are conditional on job-seeking (from 54 to 63) or participation (from 55 to 60). However, as explained above, it is unclear whether these COVID-19 induced changes in support were also long term.