Poverty-targeting: the social protection flaw?

The original version of this blog appeared as a Pathways Perspective on Development Pathways

As we start 2018, please may I have the temerity to suggest a common New Year’s resolution for all of us, as a social protection community: that we stop, definitively, the promotion of poverty-targeted approaches?

There are many, many disadvantages of poverty-targeting; and even its one supposed advantage – that it is cheap – is illusory, because low-cost social protection delivers few of the social, economic and political benefits that result from higher levels of investment. As with all things in life, you get what you pay for. Benjamin Franklin, a believer in egalitarian democracy, observed: “The bitterness of poor quality remains long after the sweetness of low price is forgotten”. And who really wants to be associated with something as tawdry as poverty-targeting?

Here are some of its many drawbacks:

·       It is impossible to do. First, of course, it is impossible to accurately identify the “poorest”, especially in contexts where the majority of the population live on low incomes. The charade that you can accurately measure and compare the poverty of different households is manifestly absurd, as has been frequently documented. We should all, at the very least, be open and honest that approaches to targeting on the basis of poverty (whether community-based or proxy means tested) are simply rationing mechanisms; and, in the case of the proxy means test, are as arbitrary as a lottery.

·       It adds cost and complexity. Even if a semblance of accuracy were possible at a given point in time, it is prohibitively expensive and complicated to maintain up-to-date information on comparative destitution, especially in countries where incomes are highly dynamic. In all low- and middle-income countries, a substantial proportion of the population moves in and out of poverty on a seasonal or annual basis. It is fanciful to think that a metric collected one year will have any validity in one year’s, three years’, or five years’ time.

·       It damages social cohesion. Because the outcomes are so random and unintuitive, poverty-targeting that chooses one household and excludes a nearly-identical neighbour will inevitably create jealousies and social tensions.

·       It is inequitable. In situations where many people are equally poor, to give an arbitrary selection of them a comparatively significant benefit will catapult the lucky few into a higher wealth category than the unlucky many. The “have-nots” will become the “haves” and will remain better off than the new “have-nots” until eventual re-targeting…when the iniquitous yo-yo will reverse.

·       It creates perverse incentives. As soon as people understand that they will only remain beneficiaries of a programme if they meet certain criteria of deprivation, they will be faced with the perverse choice between remaining poor and continuing to receive the benefit, or bettering themselves and losing it. Why pour a concrete floor, put a tin roof on your house, or save for your old age if, by doing so, you will be excluded from State benefits?

·       It rewards dishonesty. Again, with a growing understanding of the system, some people will be tempted to game it: they may borrow extra children from neighbours, feign disability, hide assets, or deny ownership of small livestock. The result is that dishonesty is rewarded and honesty penalised, which is damaging to the moral fabric of society.

·       It incites patronage. Giving anyone the discretion to influence the choice of beneficiaries for a programme puts that person in a position of power. This opens the door to abuse, patronage or exploitation. Even where such temptation is resisted, there may nonetheless be a perception of patronage, which again jeopardises the social compact.

·       It stigmatises. Poverty-targeting demeans programme beneficiaries, the polar opposite of the desired effect of social protection, which is to include and dignify. Posting the names of beneficiaries on walls, for example, or announcing them in public, has the effect of stigmatising vulnerable people, not empowering them.

But these defects, while egregious, are not the real issue. If you were an insensitive, extravagant, patronising bigot who didn’t care about dignity, ethics or equity, you could probably live with them. The real problem is the wider impact of such deficiencies on political support for social protection. National politicians shy away from poverty-targeted interventions. I challenge you to name one such programme that has led to an adequate – and sustainable – fiscal commitment to social protection, or has increased the real value of its transfers to beneficiaries over time.

Poverty-targeted programmes never become entitlements. As a result, they never generate popular demand (except among a voiceless minority) and, in consequence, they never gain political traction. Therefore, they never generate adequate domestic fiscal space for improved social protection.

Let’s look at a few examples: first, the much-vaunted – but poverty-targeted – Pantawid Pamilyang Pilipino Program (4Ps) in the Philippines. Figure 1 shows the evolution of the real value of the transfer since 2007: ever downwards. Worse still, as a result of this 32 per cent reduction over ten years, recent research by the World Bank has shown that some recipient children are now obliged to work to cover the costs of attending school in order to avoid being sanctioned or excluded from the programme.

Figure 1: Real value of the Philippines 4Ps transfer

Next, let’s look at a telling comparison between Malawi’s Social Cash Transfer (SCT) and Lesotho’s Old Age Pension (OAP). Both programmes started at around the same time and in response to the same problem: the ravages of HIV/AIDS. But Malawi’s was poverty-targeted, based on the (baseless) “10 per cent labour-constrained ultra-poor”, while Lesotho’s was near-universal, targeting all citizens once they reached the age of 70 years. Malawi’s has remained externally-driven, and is still – fifteen years on – more than 90 per cent funded by donors. Lesotho’s is (and has always been) fully domestically-funded and is very much part of the political landscape, with recent elections having been won and lost on competing pledges to increase the value of the transfer. As Figure 2 shows, the very different trends in the real value of the respective transfers is indicative of the difference between an inclusive “social protection floor” and a poverty-targeted “social protection flaw.”

Figure 2: Comparison between the real value of transfers in Lesotho’s OAP and Malawi’s SCT

Finally, look at the evolution in Zambia’s Social Cash Transfer scheme. Originally, like Malawi’s, this was tightly poverty-targeted at the “non-viable” poorest 10 per cent. For its first ten years it was almost exclusively donor-funded, and gained no attention from local politicians. It drifted unconvincingly, expanding from one district to five, then to eleven; from 1,000 households to 8,000, then to 24,000. But at this point (see Figure 3) it started experimenting with different categorical targeting approaches. Suddenly, politicians sat up and took notice…and domestic funding started to flow. In the five years since 2012, it has become essentially a near-universal old-age and disability pension…and it has expanded exponentially: to nearly 600,000 households in all 108 districts of the country. Donor funding has barely increased: the cost of the expansion has come from domestic resources.

Figure 3: Changes in coverage of Zambia’s Social Cash Transfer scheme

What we are striving for, as social protection practitioners, is programmes that are based on entitlement, that generate increased domestic funding, and that maintain (or even raise) the value of their benefits to the poor and vulnerable over time. We are not going to get this from poverty-targeting. So please can we make a common resolution to promote only inclusive approaches in 2018?