Thursday, 23 March 2017

Will no-one rid me of this troublesome PMT?


The original version of this blog appeared on Development Pathways

Anyone who has worked in social protection knows that the thorniest issue of all is that of “targeting”. The recent polemics on these pages about the inadequacy of the Proxy Means Test (PMT) as a “targeting” mechanism have raised the more fundamental question of whether it is in fact the actual concept of “targeting” that is at the heart of the problem.

Amartya Sen began his famous and oft-cited paper on “The Political Economy of Targeting” (1995) like this:

The use of the term “targeting” in eradicating poverty is based on an analogy–a target is something fired at. It is not altogether clear whether it is an appropriate analogy. The problem is not so much that the word “target” has combative association. This it does of course have, and the relationship it implies certainly seems more adversarial than supportive.

Sen also worried that “targeting” treats the recipient as an object rather than as a human being (“The image is one of a passive receiver rather than of an active agent”). And Stephen Kidd, in his “Pathways Perspective” on “Rethinking ‘Targeting’ in International Development” (2013) argued that:

The concept of viewing recipients of public services as “targets” to be hit corresponds well with a neoliberal concept of social policy, in which people receive “assistance” as a form of government charity. It is much less appropriate within a paradigm in which public services are regarded as “entitlements” offered to “citizens”.

These are all fair points. But there is a further objection to the term. The concept of “targeting” carries with it an implicit assumption that accuracy is possible: the supposition is that we are able to hit a target. But such accuracy is, of course, impossible when we are trying to assess the comparative poverty of different households or individuals. There is no accurate way to capture the multiple facets of relative poverty: where a household lives, what it comprises, the age and capabilities of its members, what it owns, what characteristics and skills it possesses, what social bonds it has, and so on. To imagine otherwise is delusional. We need to be honest, and explain to policy-makers that PMT is not a “targeting” mechanism, it is one of several possible “rationing” mechanisms…and not necessarily the best one.

The concept of targeting should properly be reserved for the policy choice of which vulnerable group is to be the recipient of government support, for example the elderly, children, pregnant women, the working age poor. Here, at least some degree of accuracy of identification is possible. It is rather at the next stage that the problems arise: there is a subsequent choice about whether all of this group, or just some of them, should receive the benefit. Much the best solution of course is to provide the assistance universally, or at least only exclude the very wealthiest, as in South Africa’s self-declared affluence testing. But this decision is usually a function of the resources available: if they are inadequate, then some form of rationing will need to be applied to restrict them to just a subset of the group.

There exist a number of options. The major ones (in broadly decreasing order of acceptability, from best to worst) include:

  • Eligibility rationing – where eligibility criteria are highly restricted, for example by setting the eligibility age of a social pension very high (e.g. over 70, as in Lesotho), or of a child grant very low (e.g. under 2 years). The advantage of this approach is that it limits the numbers, but is nonetheless universal, which makes it popular, and which allows the eligibility criteria to be gradually expanded over time (as has happened for example with the Child Support Grant in South Africa).
  • Geographical rationing – where specific geographic areas are isolated for receipt of the benefit. Examples of this approach would include support to the riverine chars of Bangladesh, or to specific ethnic groups in tribal areas of Vietnam. The problem is that this is sometimes politically difficult, since it may create or exacerbate regional jealousies.
  • Random rationing – where beneficiaries are selected by lottery. This is the practice in a number of instances, for example on the SWAPNO programme in Bangladesh. It may seem a bizarre way to allocate social “entitlements”; but it has the advantage that people understand its arbitrariness, and it is at least honest and transparent.
  • Community rationing – where communities themselves are asked to ration the benefits of the programme. One advantage of this approach is that community members frequently opt for an inclusive approach, and simply distribute the total amount of the benefits equally among everyone in the community (as for example in Indonesia’s Raskin programme): universalism through the back door. On the other hand, there can also be significant challenges, where selection may reflect existing patterns of social exclusion within communities, or resources may be captured by the more powerful.
  • Temporal rationing – where groups of beneficiaries are selected to receive benefits for a short period (often 1 to 2 years), and are then removed from the programme and replaced by a new set of temporary beneficiaries, as occurs for example in Uzbekistan. This is the approach of so-called “graduation” programmes. It may be fair in that it rotates the benefits, but impacts are seriously constrained as a result.
  • PMT rationing – where a fallacious veneer of objectivity and transparency is applied to lottery rationing, permitting the selection process to be attributed to a computer rather than to pure chance. There is nothing inherently wrong with this: what is unethical – as we have seen from the evidence presented recently on these pages – is the charade that PMT results in accurate “targeting”.
  • Auction rationing – where potential recipients bid against each other for inclusion on a programme, as for example in Bangladesh, where some beneficiaries of the old age allowance have had to pay up-front bribes of up to one year’s worth of benefits to be included. The most common manifestation of such auction rationing is public works programmes, where “beneficiaries” bid their opportunity costs for the privilege of expending valuable calories on hard labour in exchange for a meagre transfer.
  • Patronage rationing – where community leaders or other worthies select beneficiaries based on their patronage relationships. Examples of this would include the constituency funds allocated in some Pacific Island states to Members of Parliament, for them to distribute as social assistance to selected constituents. The approach is not “fair”, but it is at least understandable, and transparent in its unfairness.

There is nothing very scientific about the list above, and the order of preference is strictly personal! But it demonstrates that there are other methods than the awful PMT to cut the cake. It may be that if the use of PMT for poverty-“targeting” is presented to policy-makers for what it really is – i.e. as one of many options for rationing limited social resources – then they will pay greater attention to alternative approaches…most of which are actually much better!

Wednesday, 22 March 2017

Poxy Means Testing: it’s Official!


The original version of this blog appeared on Development Pathways

“A prox on both your houses” 
[i]

The World Bank has recently – and some would say belatedly – undertaken a critical review of the Proxy Means Test (PMT)[ii], the approach to targeting that it has been advocating, uncritically, for the past decade.

The results are astonishing. Disguised beneath a splendidly econometric veneer, the raw findings that emerge demonstrate that the PMT is a wholly unsatisfactory targeting mechanism. Based on rigorous analysis of PMTs in nine sub-Saharan African countries (Burkina Faso, Ethiopia, Ghana, Malawi, Mali, Niger, Nigeria, Tanzania and Uganda), it finds the following: when using “Ordinary Least Squares results for Basic PMT” (the most common PMT approach), with a fixed poverty line of 20% of the population, “On average, the rate of inclusion errors implies that 48% of those identified as poor by the Basic PMT method are in fact non-poor”; and “The average exclusion error is sizeable, with 81% of those who are in the poorest 20% in terms of survey-based consumption being incorrectly identified as non-poor by the PMT method”.

Let’s just stop and think about this. What this means is that, if a country is encouraged to establish a poverty-targeted social assistance programme targeting the poorest 20% of its population, then its policy-makers will need to accept two facts: that almost half of the actual beneficiaries of the programme would be from outside the intended sub-group; and that fully four out of every five households intended to benefit would in reality be excluded from the programme. What kind of policy-maker would accept such lamentable targeting performance? In Mali, incidentally, not one single ultra-poor household was correctly identified by the PMT as being ultra-poor: an exclusion rate of 100%!

The paper goes on to suggest that certain refinements can improve the accuracy of econometric targeting. But the improvements are small, and the necessary refinements range from being unlikely to being wholly impractical in reality. At the unlikely end of the scale, one suggestion is to increase the coverage of such poverty-targeted programmes to 40% of the population. Yet there are practically no examples of this in Africa, and the reality is that the vast majority of PMT-based programmes target even fewer than the poorest 20%. At a more common level of 10% coverage, the targeting errors are likely to be significantly higher, especially since, as the paper states: “econometric targeting may have difficulty in identifying those who are very poor” and “PMT is missing many of the poorest households in all countries”. At the wholly impractical end of the scale, the proposal is to run an “Expanded PMT” with “far more data”. But here the paper itself accepts that: (a) “the improvement would have to be judged as modest”; and (b) “the field implementation of a PMT formula with many variables is expensive and difficult”.

Remember too that these underwhelming reported outcomes reflect only the inherent statistical inaccuracy of the PMT approach. As other papers have emphasised[iii], the overall performance of a PMT will inevitably be further compromised by a range of other factors. Many of these are touched on, but not explored, in the World Bank paper. Actually implementing the complex and unintuitive PMT approach is bound to introduce further errors (as the paper coyly admits “Field implementation may introduce idiosyncratic mistakes”, and “Most likely the methods will perform less well than our calculations suggest.”). And there are still further problems: (a) with a PMT’s perverse incentives (e.g. households not wanting to acquire assets or improve their dwelling for fear of being excluded); (b) with moral hazard (e.g. households being encouraged to lie about their situation in order to qualify); (c) with the actual costs involved in the targeting process; and (d) with the damage to social cohesion of an improperly understood and seemingly arbitrary selection procedure. As the paper acknowledges, “We present and compare the best-case results for the various methods reviewed, unaffected by potentially differential costs, ease of implementation, and susceptibility to manipulation and corruption”.

One final reason for the PMT’s inaccuracy – which the paper does explore – is that of its inability to respond to the dynamics of poverty. There is always a degree of churning in and out of poverty, and a PMT is very static: most PMTs are only re-run every five to ten years. The paper looks at the implications of this on targeting accuracy by using panel data and running the analysis with lags of one to two years. This shows that – even with such a small lag – inclusion error increases from 48% to 55%, and exclusion error from 81% to 90%. On this basis (which would become still worse over a longer time-lag), we would now need to be telling our putative policy-maker that his or her poverty-targeted social assistance will consequently include more unintended than intended beneficiaries; and that nine out of ten of the intended beneficiaries will be excluded from the programme. This is crazy: imagine trying to persuade a policy-maker to adopt a criminal justice system that resulted in more than 50% of all jail inmates being innocent, and nine out of ten criminals being found not guilty!

So what are the alternatives to PMTs? Well, the paper helpfully goes on to explore some options. It looks at various permutations of simpler, more transparent and more intuitive targeting approaches, premised on a basic income transfer either to all, or to selected categories of, the population (children, the elderly, widowed, disabled or orphaned). It assumes the same overall budget for all the options (though it doesn’t allow for the additional costs involved in running a PMT), and it looks at the comparative poverty impacts of each. The verdict: “even under seemingly ideal conditions, the ‘high-tech’ solutions to the targeting problem with imperfect information do not do much better than age-old methods using state-contingent transfers or even simpler basic income schemes. We find that an especially simple demographic ‘scorecard’ method can do almost as well as econometric targeting in terms of the impacts on poverty. Indeed, on allowing for likely lags in implementing PMT, the simpler categorical targeting methods perform better on average in bringing down the current poverty rate. This conclusion would undoubtedly be strengthened once the full costs of fine targeting are taken into account”.

The paper thus demonstrates conclusively that, in terms of poverty reduction in the real world, PMT performs worse than simpler categorical approaches or even basic income schemes…as well as being administratively costly, morally reprehensible and socially divisive.

Hurrah! But why has this taken so long? And what are the implications for those countries that the World Bank has already persuaded to sign up to such an execrable model?


[i] To misquote Mercutio in “Romeo and Juliet” by William Shakespeare.

[ii] Brown, C, Ravallion, M and van de Walle, D (December 2016), “A Poor Means Test? Econometric Targeting in Africa”, World Bank Policy Research Working Paper 7915, Washington DC.

[iii] See for example, Kidd, S and Wylde, E (September 2011), “Targeting the Poorest: An assessment of the proxy means test methodology”, AusAID, Canberra; and Kidd, S, Gelders, B; Bailey-Athias, D (2017) Exclusion by design: An assessment of the effectiveness of the proxy means test poverty targeting mechanism, International Labour Office, Social Protection Department (SOCPRO), Geneva.

Come on and open up your heart!

  This blog originally appeared on Development Pathways I very much enjoyed Stephen Kidd’s humble and courageous admission that he is a refo...