Targeted income programs are unsustainable, create social division, and disincentivize social mobility

Suppose first that targeting would be implemented by a perfectly efficient and functional state, without any internal issues of performance, incentives, or administrative cost. This is clearly unrealistic for developing countries but helps segment the various issues around whether to target or not. Targeting the poorest has the obvious advantage of transferring resources to people for whom the marginal value of money is highest, since they have the least money. Yet things are less clear in a world of imperfect markets. As we described above, in worlds where markets are imperfect recipients of transfers may use them to relax binding constraints on growth, for example, by purchasing assets they otherwise could not have financed or starting businesses they did not have the capital for. In this case, the welfare impacts of the transfers depend on variation in the opportunities and constraints each person faces as well as variation in their baseline standard of living (and hence their marginal utilities). It could well be optimal to make transfers to someone a bit better off in a community if this will enable them to make a transformative investment in that community. Existing analysis of optimal targeting has largely ignored this issue, and so we know little about whether targeting the poor actually optimizes impact on poverty. If interactions between households are important, then optimal targeting becomes even more complicated. Transfers to people embedded in existing networks of mutual support may be redistributed within those networks and generate a final distribution different from the one on paper (e.g. Angelucci and De Giorgi (2009)). As we discussed above, many basic income advocates argue that the effect of entire communities being “in it together” will be greater than the sum of the effects on individuals treated in isolation. For example, targeting may create bitterness and social division that might reduce willingness to contribute to the financing of local public goods (Kidd and Wylde, 2011); universality could have the opposite effect.19 More generally, if own and neighbor’s treatment status are substitutes then this strengthens the case for targeting, while if they are complements it weakens the case. As of yet we know very little about these interactions. Targeting may also alter the political viability of basic income schemes – there is a large theoretical literature showing this, reviewing it is beyond the scope of this paper (see Cremer and Roeder (2015) and Casamatta et al. (2000) for two examples). A program’s beneficiaries typically form the core constituency willing to fight for its effective implementation of a program, or its continued existence. If only a small number of the poorest, most disadvantaged and disempowered members of a society benefit from a program, it may be hard for them to exert much influence over its future. This is one reason some in India, for example, have fought for the expansion of eligibility in means-tested programs such as the Public Distribution System, arguing that it would benefit the poorest to include somewhat better-off people who are influential enough to push back against the corrupt middle-men who looted the system in the past (e.g. Dre`eze and Khera (2010)). Consistent with this argument, Klasen and Lange (2016) show using data from the Chinese urban Dibao cash transfer program that a percent increase in the share of people who get benefits increases the budget available for the program by a third of a percent. Using fixed budget arguments for targeting may therefore underestimate the benefits of universality. Finally, targeting may create disincentive effects. As we discussed above, targeting a basic income relocates where the disincentive effects may fall (though of course, they may be of small magnitude). Any redistributive scheme can create disincentive effects for some part of the population, depending on who pays and who receives. For example, in a universal basic income scheme, disincentive effects would be concentrated among the people who are marginal contributors to the scheme, likely the upper and upper-middle classes. If instead the basic income transfers were targeted to the bottom quintile, say, this would lower taxes at the top but create a new disincentive effect to move out of the bottom quintile. Fundamentally, then the question of targeting is one about where in the distribution of incomes to locate the disincentive effects and the magnitude of any such disincentive effects. Unfortunately, we know very little about the incentive effects of targeting regimes in developing countries, in contrast to work on public finance in developed ones for which this has been a major focus. They certainly exist; Imbert and Papp (2018) find that the regional targeting of India’s employment guarantee to rural areas has discouraged rural-to-urban migration, for example (which may be good or bad, depending on whom you ask). Similarly, Banerjee et al. (2018) look at the impact of asset based poverty targeting of a program in Indonesia on the purchase of specific assets (TVs and SIM cards) and find a temporary increase in misreporting of these assets, but no reductions in sales of these assets, implying disincentive effects for may be small. More work in this area would be of great value.