Targeting refers to any mechanism to identify eligible individuals, households and groups, for the purposes of transferring resources or preferential access to social services (Devereux et al., 2015: 7). ‘Popular targeting mechanisms include means testing, proxy means tests, categorical, geographic, community-based, and self-selection’ (ibid.: 3).
The foremost rationale for targeting ‘is to direct programmes to those who will most benefit’ (White, 2017: 145). Targeting may have other aims: to maximise poverty reduction; to ensure no one is ‘left behind’; to contain the costs of provision; and to make the most efficient use of resources when faced with budget limits; or for political gains (Devereux et al., 2015: 7–8; Kidd & Althias, 2019: ii). While ‘targeting of benefits to those most in need is widely practiced’ (Ulrichs & White-Kaba, 2019: 17), there are ongoing debates about targeting approaches – the most cost-effective methods for reaching those most in need – and the appropriate degree of targeting.
Universal social protection includes schemes that aim to reach every citizen passing a basic criterion, often categorical schemes for all people of a certain age (e.g. a social pension where eligibility is only restricted by age and therefore reaches all older citizens) or status (e.g. all children under five years old) (Devereux et al., 2015: 9). A universal basic income would provide benefits to each individual: ‘such schemes are rare’ (Kidd & Althias, 2019: 6).
Poverty incidence may reach a level at which it is ‘not worth the cost of targeting’ and investing that money in universal programmes may be preferred (White, 2017: 158; Ulrichs & White-Kaba, 2019: 17). A universal benefit may be intrinsically self-targeting; for example, if for some beneficiaries the cost of the benefit – such as queuing or participating in a public works programme – is too high and they choose not to take part (White, 2017: 158). Universal programmes ensure all in need are reached and can increase buy-in from all sections of the population (Ulrichs & White-Kaba, 2019: 17). However, few countries can afford to provide social protection to all. The Universal Social Protection 2030 Initiative aims to facilitate countries’ progressive expansion of social protection to achieve universal coverage, as ‘resources and politics permit’ (Devereux, 2016: 14). See What is social protection? Analytical concepts and Coverage, spend and systems.
Given financial limitations, social protection programmes and systems are often targeted to some extent. Options to target social protection include:
- narrow the geographical coverage;
- limit the categories selected (e.g. old-age pension or child grant);
- narrow the category selected;
- direct resources at those living in poverty (by means testing or proxies); or
- use a combination of approaches (e.g. a poverty-targeted child grant).
Narrowing the category involves limiting the age of eligibility or, in the case of disability-specific benefits, selecting those with more severe disabilities. In addition, governments can then choose to restrict the programme further by targeting those living in poverty.
The use of targeting is contested, criticised for both pragmatic (as all targeting mechanisms generate errors and costs) and ethical reasons (as it can lead to ‘social divisiveness and perceptions that excluding some people from benefits is socially unjust’) (Devereux, 2016: 1; Devereux et al., 2015). Targeting mechanisms face design and implementation difficulties in reaching those that need the assistance most; consequently, some of the most vulnerable can be excluded (UNDP, 2016: 41; Kidd & Althias, 2019). Typically, interventions using proxy means testing[1] feature ‘inherent 30–40% inclusion and exclusion errors’ (World Bank, n.d.), while poverty data collection and analysis to inform targeting, and keeping this information up to date, can be expensive (Ulrichs & White-Kaba, 2019: 17). In addition, targeting can potentially increase social tension (Devereux et al., 2015: 34). Evidence on the impact of broader targeting on social cohesion is limited, with mixed findings. Ellis (2012: 212) finds that universal (or categorical) transfers are socially popular. However, Babajanian (2012: 31) highlights that ‘Social categorical targeting in fragile states can exacerbate social divisions and inequalities by including specific groups and leaving out others (e.g. in Sierra Leone and Liberia).’
Key texts
White, H. (2017). Effective targeting of social programmes: An overview of issues. Journal of Development Effectiveness, 9(2), 145–161.
This paper reviews the issues involved with targeting. It notes that the choice between a universal benefit and a targeted scheme is ultimately a political decision, but sets out some technical criteria to take into account when making this decision.
Devereux, S. (2016). Is targeting ethical?. Global Social Policy, 16(2), 166–181.
This article examines ‘targeting’ versus ‘universalism’ debates, drawing on three principles of redistributive justice – equality, equity, and need. It concludes that ‘social assistance should be targeted at those who need it, especially when budgets are constrained, moving progressively towards “categorical universalism” when resources and politics permit’.
Devereux, S., Masset, E., Sabates-Wheeler, R., Samson, M., Rivas, A.-M., & te Lintelo, D. (2015). Evaluating the targeting effectiveness of social transfers: A literature review (IDS Working Paper 460). Brighton: IDS.
This paper reviews empirical evidence on targeting mechanisms from a range of social protection programmes. It considers evidence on errors (inclusion and exclusion, by eligibility and by poverty) and associated costs (administrative, private, social, psychosocial, incentive-based and political).
See also:
Kidd, S., & Athias, D. (2019). Hit and miss: An assessment of targeting effectiveness in social protection (Working Paper). Orpington: Development Pathways.
Ellis, F. (2012). ‘We are all poor here’: Economic difference, social divisiveness and targeting cash transfers in sub-Saharan Africa. Journal of Development Studies 48(2), 201–214.
Coady, D., Grosh, M. E., & Hoddinott, J. (2004). Targeting of transfers in developing countries: Review of lessons and experience (Vol. 1). Washington, DC: World Bank.
Other resources
Article/blog: Kidd, S. (2016, 9 February). Social protection: Universal provision is more effective than poverty targeting. Ideas for Development (ID4D).
Article/blog: Yemtsov, R. (2016, 16 August). Social protection: Universal and poverty targeting approaches are not in contradiction. Ideas for Development (ID4D).
[1] Proxy means testing is a targeting methodology that uses ‘observable characteristics of the household or its members to estimate their incomes or consumption, when other income data (salary slips, tax returns) are unavailable or unreliable’ (World Bank, n.d.).