Saturday, 1 December 2012

Re: silly-ence, or the reinvention of vulnerability

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

“Resilience” is decidedly the development flavour of the month! Everybody’s doing it: it’s one of the three pillars in the World Bank’s “Social Protection and Labor Strategy 2012-2022”; the European Commission has put out a whole new Communication devoted to it; DFID has committed “to embed resilience-building in all DFID country programmes by 2015”; WFP (never one to miss a bandwagon) has a “Resilience Project” with the Swiss; and USAID celebrated a “Resilience Week” in April 2012. Academia is responding with delight: trenchantly erudite papers have appeared in the last few months with titles such as “The Relevance of ‘Resilience’?” (from the Humanitarian Policy Group), “Resilience: New Utopia or New Tyranny?” (from the Institute of Development Studies), “The Resilience Renaissance?” (also from IDS), “Resilience, a Risk Management Approach” (from the Overseas Development Institute), and “The Characteristics of Resilience Building” (from the Interagency Resilience Working Group, no less).

But what does it all signify? Does it mean anything? Does “resilience” advance the cause of development in any significant way? Probably not.

Resilience is no more and no less than the developmental antonym of “vulnerability”, a concept which has been around for a number of decades. Resilience, like vulnerability, has two dimensions: one internal and one external. You can increase resilience either by enhancing the inherent ability of an individual, a household, a community, a system or a country to withstand a shock, or you can act externally to reduce the potential for damage from that shock. Resilience can be increased, in just the same way as vulnerability can be reduced, in either of these two dimensions: you can strengthen an entity’s capacity to resist or you can reduce the likelihood of exogenous damage.

Think of driving a car on a busy motorway. There is always the risk of an accident, which could result in death or serious injury. There are two ways to protect against the damage that this might cause. In one dimension you can strengthen the capacity of the car’s occupants to withstand an accident: you can surround all passengers with airbags; you can legislate that all of them wear seatbelts; you can install revolutionary technology such as anti-lock braking systems (ABS) and traction control systems (TCS); you can capacitate the driver to drive better; and you can enforce adherence to the highway code. In the other dimension you can set up systems to implement variable speed limits according to weather conditions or traffic flow; you can position police and emergency services at strategic positions along the motorway; you can train ambulance crews to deal better with accident victims; you can educate the general population how best to react if involved in a collision. Through addressing both dimensions, you are significantly reducing vulnerability…and increasing resilience…to serious harm through a traffic accident.

One of the recent arguments in favour of resilience as a guiding development paradigm proposes that it assembles a broader church of practitioners than vulnerability: from climate change, ecology, disaster management and social protection. But much the same was said of vulnerability a couple of decades back. In practice (sadly) it may be true that different communities of practice still work in technical silos, but in theory both vulnerability and resilience are as broad or as narrow as you choose to make them: both are equally able to encompass climate-related factors, social dynamics, ecosystems, and so on…or not. Simply shifting the standpoint from vulnerability to resilience changes nothing.

Another paper sets up an assessment matrix to show how a commonly-used framework of social protection interventions to reduce vulnerability (with objectives of provision, prevention, promotion and transformation) maps to a suggested framework of resilience (with outcomes of coping/rehabilitating, adapting and transforming). The paper seems pleasantly surprised that the mapping works well. But if you accept that the two concepts are antonyms, then this is less surprising: it is like saying that a framework of increasing brightness maps well to a framework of decreasing darkness.

What is perhaps more useful, in linking social protection to the excitingly fashionable resilience model, is to consider how these four commonly accepted functions of social protection address different aspects of vulnerability to increase resilience. If we start from a definition of vulnerability as being represented by an equation: vulnerability = poverty + risk – agency – voice, then we can see how the four functions of social protection[i] act on each of the constants in the equation:

·       Poverty is reduced through the provision function of social protection, represented by social assistance (such as social pensions, child grants or disability benefits) – this may be likened to the provision of airbags and seatbelts in the motorway example above.

·       Risk is reduced through the preventive function of social protection, encompassing social insurance mechanisms (such as unemployment pay, burial societies and health insurance) – these are represented by the availability of ABS and TCS.

·       Agency is increased through the promotive function of social protection, comprising the kind of social empowerment embodied in asset transfer programmes, cash-for-training and (well-designed) employment schemes – equipping the driver to drive better.

·       Voice is increased through the transformative function of social protection, including interventions in the area of social justice, aimed at reducing discrimination, fulfilling rights, and so on – in our example, the standard application of the highway code, driving tests and driver education efforts.

At the same time, the motorway analogy also serves to underline that social protection is only one of many inter-dependent strategies that should be used to reduce vulnerability (and to increase resilience). Others might include interventions around:

·       Risk assessment, to anticipate the dangers and have plans in place to react to them – such as the ability to communicate weather and traffic information in our motorway scenario.

·       Early warning, such as tsunami sensors, earthquake detectors or drought monitoring systems – or the updating of weather/traffic warnings alongside the motorway in this scenario.

·       Pre-provisioning, in the form of emergency stockpiles or strategic grain reserves – or the pre-positioning of police and ambulance crews in our example.

·       Health services, so that they strengthen people’s ability to withstand shocks – or, in our case, to correctly treat accident victims.

·       Education and training to raise awareness and increase people’s agency – or, in this instance, to persuade people to drive more responsibly and to respond effectively in case of an accident.

But the point is that to the same degree as these coordinated interventions reduce vulnerability, so they also increase resilience. Indeed they would suggest that a correspondingly accurate (and useful) equation for resilience would be: resilience = assets – risk + agency + voice.

The danger here is that resilience does not really advance the conceptual debate. It’s as if development partners are searching for a new buzzword around which to rally, which the poverty academics are all too ready to supply. The literature of resilience is thus self-inflating, and much of it self-defeating as it revolves not around the legitimate application of the word itself but around the baggage that other writers have loaded onto it.

As a parting concession to the resilience lobby, however, there is a valid argument that framing objectives as positives instead of negatives may be psychologically preferable: on that basis alone, it may be better to rally the development community around “increasing resilience” rather than around “reducing vulnerability”. To that extent, resilience may be a useful hanger; but to pretend it is anything more than that is, well, just plain silly!



[i] Not everyone accepts this P-P-P-T framework of the functions of social protection, nor that the different types of social protection intervention can be mapped directly to a single individual function (eg a pension can be used for provision, prevention and promotion). But it is widely used, and serves as a useful structure for grouping these different interventions.



Tuesday, 26 June 2012

Safety net ≠ social assistance

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

This is an appeal – primarily to the World Bank – to stop using (or abusing) the term “safety net”. For reasons best known to itself, the Bank now uses the term interchangeably with the term “social assistance”. But the two are not synonyms; and now, as it launches its new “Social Protection and Labor Strategy” for the next decade, is a propitious moment for the Bank to recognise this publicly, and to mend its semantic ways.

“Social assistance” defines a subset of social protection[i], comprising those social transfers that are non-contributory (ie which are funded from general government revenue, rather than from specific contributions by individuals). There are many definitions of social assistance[ii], but most people would agree that the key characteristics are regularity, predictability over the long term, government ownership, and entitlement.

“Safety nets”, on the other hand, is a dangerously elusive and baggage-laden term. It was originally coined by the Bretton Woods institutions in the 1980s and 1990s to refer to temporary measures to catch those who were transiently made vulnerable through structural adjustment and liberalisation. Most people continue to use it in this original narrower definition of a temporary social transfer project, usually operated for a finite period and often outside of Government structures.

The World Bank, however, has incrementally mutated the term “safety net” (often now prefaced with “social” – ie “social safety net”) to become synonymous with social assistance. Indeed the current draft of the “Strategy” makes this explicit, baldly stating on page 1: “Social assistance programs (also known as safety net programs…)”. In fact, social assistance programmes are not “also known as safety net programs” by the majority of people outside the Bank …nor should they be.

But, the Bank might say, terminology is allowed to evolve, and often does. Why should we object in this case? For three reasons, one definitional, one practical, and one Macchiavellian.

The definitional reason is that “safety net” is a clumsy and inadequate metaphor to capture what social assistance is meant to achieve. An actual “safety net” (eg one under a trapeze artist) is there to catch those who fall, no more, no less. Yet true social assistance, as the “Strategy” itself accepts in its belated adaptation of the “3Ps” framework, has three goals: (i) to catch people who fall (“protection”, or “equity” in Bank parlance), (ii) to prevent people from falling in the first place (“prevention”, or “resilience”), and (iii) to allow people, where possible, to raise themselves out of poverty (“promotion”, or “opportunity”). “Safety net” encapsulates only the first of these three goals. As a result of this deficiency, there has been a proliferation of slightly ridiculous terms such as “social springboards”, “social trampolines”, “safety ropes” and “safety ladders” to try to capture metaphorically the other functions of social assistance.[iii] There has also emerged the meaningless concept of “productive safety nets”: but how can a safety net be productive, either literally or figuratively? It is like talking about an “intelligent log” or an “emotional toothbrush”…or perhaps a “compassionate economist”!

The practical reason is that the use of “safety nets” as being synonymous with “social assistance” creates unnecessary confusion, because there is then no way of distinguishing temporary, short-term, discretionary measures implemented by external actors from guaranteed, long-term, entitlements offered by governments. And this is an increasingly important distinction in the social protection debate. There is certainly a continuing need to provide true “safety nets” in some instances, for example after a shock such as a drought, an earthquake or a dramatic rise in food prices, but this is not at all the kind of programme that the Bank is trying to promote in its new “Strategy” for 2012 to 2022, where the focus is on “the need to build a coherent portfolio of social protection and labor programs”, and whose “strategic direction is to help developing countries move from fragmented approaches to more harmonized systems”.

The Machiavellian reason is that it may suit the Bank to perpetuate the term to promote its own primary agenda: that of making loans. In the same way as subliminal advertising works, the use of the term “safety net” implies something temporary and finite, and may sub-consciously make governments more prepared to take on a loan than if the same package is dressed up as “social assistance” with its connotations of long-term commitment and rights-based entitlement. [This is the same logic as underpins the Bank’s promotion of conditional cash transfers, because it is much easier to sell loans for programmes that can be justified on their educational and health objectives than for child benefits or other forms of unconditional social assistance.] The only other explanation for the Bank’s reticence to change would be an inbuilt institutional inertia: in other words that so much has been invested in “safety net” websites, how-to tools, evaluations, reports, etc, that it now becomes too daunting to redo them all.

But whatever the explanation, the time has come to change. So, please, before the Bank goes final with its “Strategy”, could it (i) replace that definitional entry on page 1 with the wording: “social assistance (formerly, and sometimes still incorrectly, referred to as safety nets)…”, and (ii) replace all 100 or so other occurrences of the term “safety net” with the term “social assistance”, except in the rare instances where it is actually referring to a short-term response to transitory poverty, operated substantially outside of government social protection frameworks?



[i]  Many would argue that “social security” (the term more commonly used in OECD countries) should also be used globally, in preference to “social protection”.

[ii]  There is some debate about whether “social assistance” itself is the best term for such non-contributory transfers, since it may have connotations of charity rather than entitlement.

[iii]  The previous iteration of the Bank’s strategy paper (for 2000-2008) was even called “From Safety Nets to Springboards”.



Monday, 19 September 2011

RHVP: RIP!

So it's official: RHVP's controversial Wahenga Comments "may occasionally have over-stepped the mark and caused offence"!

This is one of the findings of a recent assessment of RHVP, facilitated by the Overseas Development Institute (ODI), which attempted to learn lessons about the degree to which RHVP had influenced social protection policy over the six years of its lifetime.

As RHVP draws to a close, this seems an appropriate opportunity to apologise to those we may have offended; but also, as the ODI assessment does, to consider the more positive influences to which the Programme may have contributed in some small way.

First, we tender our apologies. We were perhaps too strident in some of our campaigns, but they were ones that we genuinely believed in, and where – as a small programme trying to induce major shifts in policy – we were often pitted against entrenched opinions, vested interests and a stolid status quo. We apologise to World Food Programme (WFP), an early target in the food/cash debate ... though we are still not convinced that an organisation with such global expertise in emergency response should be diversifying into the development of national social protection policies. We apologise to the World Bank and its academic associates, especially for airing Sissy Teese's choleric rants ... though we still harbour serious reservations about a number of the Bank's prescriptions in an African context, for example its advocacy of proxy means testing and its fondness for conditionality. We apologise to UNICEF for tarring them with the "ten-percent" brush ... though we still reject a poverty-targeted approach in situations of widespread poverty such as those prevailing in sub-Saharan Africa. We apologise to the International Labour Organisation (ILO) for questioning the foundations of their social protection floor ... though we continue to believe that more work is needed to burnish its credentials as a support to truly national policies and priorities. And we apologise to any others we may have offended inadvertently along the way ... though – as the saying goes – you can't make mealie-meal without milling mealies!

Second, there is evidence of more positive RHVP impacts, even arising out of such transgressions! The ODI assessment suggests that RHVP's "strong and consistent messaging was a key strength in terms of taking a strategic approach to policy influencing ". It goes on to state that "the Programme stuck to strong, clear messages on social protection and took an ‘uncompromising’ approach which sometimes involved controversial or critical stances. In many cases this drew defensive reactions, but over the course of the Programme, not only were many of those messages vindicated, but the messages delivered can be linked to a number of concrete policy changes". The report cites a number of specific examples where "RHVP identified very early on a number of key tenets around social protection, and was consistent in adhering to those tenets throughout its lifespan". These include:

·       The supremacy of cash over food aid in most cases - e.g. Wahenga Comments as early as 2006; strong advocacy for cash-based responses in Lesotho and Swaziland in 2007 and Malawi in 2009; the organisation of an influential regional workshop on cash transfers in 2006.

·       The potential for innovation in delivery systems - e.g. our "Upwardly Mobile" Brief and Comment in early-2006 (before MPESA had even started!), predicting a key role for mobile phones; and the "Lesotho Ladies" initiative and Briefs beginning in 2006 through to the end of the programme.

·       The need to move away from piloting to supporting national programmes - e.g. our "Stop Experimenting" Comment in mid-2006, when pilots were still all the rage among donors and INGOs; and our criticism of two World Bank experiments in Tanzania in 2008 and 2010.

·       The deficiencies of public works programmes as a response to chronic hunger - e.g. our "Public Works Don't" Comment in early-2007.

·       The potential for direct transfers to individual households - e.g. our "Direct Aid" Comment in early-2007 (long before the 2010 book by Joseph Hanlon et al "Just Give Money to the Poor").

·       The weaknesses of poverty targeting, especially in situations of significant poverty - e.g. our "One out of Ten" and other poverty-targeting Comments in 2008; and our Frontiers of Social Protection (FoSP) Brief in 2009.

·       The uncertain evidence on the benefits of imposing conditions (NB not "conditionalities"!), especially in Africa - e.g. our "What's Wrong with Conditionality" Comment as early as 2006; an article in the Institute of Development Studies (IDS) Journal in 2007; and the infamous "Sissy Teese" Comments in 2010 and 2011.

·       The considerable potential for South-South learning - e.g. our early contacts with International Poverty Centre in Brazil; drawing lessons from Bangladesh in 2009; and our promotion of social protection study tours within Africa.

·       The importance of promoting political buy-in whilst also building technical capacity - e.g. the "Institutional and Policy Context" Briefs in 2008; our initiation and support of the civil society Africa Platform for Social Protection from 2006; a targeted focus on the media from 2007; and our SADC Parliamentary Forum initiative and handbook aimed at parliamentarians from 2008 onwards.

As the ODI report makes clear, "many of these were controversial – even radical – ideas at the time they were first aired, and part of RHVP's visibility and impact has probably derived from its provocative (but unswerving) stance". And, even where RHVP itself may not have been directly responsible for changing policy, the report makes a compelling argument that "by taking this ‘radical’ and visible stance in the debate, RHVP opened the door for other more ‘moderate’ voices to find traction within institutions where they would otherwise have not". This in itself would represent a significant achievement of the Programme.

Finally, on this valedictory note, we would like to thank our paymasters, DFID and AusAID, who have made the Programme possible. Their staff (almost without exception!) have provided us with unstinting support and excellent guidance. They have also – more importantly – allowed us the freedom and independence to chart our own course and to fight our own battles. We are very grateful to have had these opportunities over a tumultuous six years, during which there has unquestionably been a paradigm shift in thinking around social protection. This shift cannot be arrogated to RHVP, of course; but it has been an appropriate (and exciting) time for the Programme to have been in existence, and we hope we have contributed in some small ways. Very many thanks to my fantastic colleagues on RHVP who have shared the adventure; and thanks - above all - to all of you that we have worked with along the way!


Sunday, 31 July 2011

Social Protection and the Four Horsemen of the Donor Apocalypse

A version of this blog was later published in “Social Protection in Developing Countries: Reforming Systems”, eds Bender, Kaltenborn and Pfleiderer, Routledge, Oxford, UK (2013)

Social protection has come a long way in the decade since it first entered the development discourse. There is now generally a recognition that the old development paradigm ("the poor are the problem") is bankrupt. This paradigm focussed development on promoting economic growth, waited for economic growth to reduce poverty, provided residual interim safety nets to those who could not benefit immediately, and left donors to step in with expensive "emergency" assistance when necessary. This has not worked.

The emerging development paradigm ("the poor are the solution") is predicated on the provision of comprehensive social protection which will allow the poor themselves to participate in the generation of economic growth; this in turn will reduce poverty and the cost of providing social protection, and will also reduce the need for emergency assistance, thus freeing donor resources to help fund further development.

Over the last decade, donors and their international development partners have indeed done a lot to promote and support these significant advances. But they could have done better. There is a serious, and continuing, problem that donors and IDPs fall into four competing camps, which this paper characterises as the "Four Horsemen of the Donor Apocalypse": the Productivists, the Ten-Percenters, the Instrumentalists and the Universalists.

The White Horse

I watched as the Lamb opened the first of the seven seals. Then I heard one of the four living creatures say in a voice like thunder, "Come and see!" I looked, and there before me was a white horse! Its rider held a bow, and he was given a crown, and he rode out as a conqueror bent on conquest.

The Productivists are led by the World Bank and the IMF. The characteristics of their vision of social protection are that it must be targeted at the poorest; that the poorest should preferably be identified through "proxy means testing"; and that no-one should be given "something for nothing". This means that they favour public works programmes, social action funds and conditional cash transfers (for all of which it is of course much easier to persuade Governments to take substantial loans, since such programmes are also demonstrably producing physical, social and human assets). Examples of Productivist programmes abound, for example in Latin America, in the Philippines, and in Ethiopia, Rwanda and Tanzania.

There are some concerns with this "poor relief" approach, however. The first is that the incremental benefits of conditionality, pushed hard by the Bank, are unproven (or at best controversial) . The second is that proxy means testing is, by the Bank's own admission, "most appropriately used where a country has reasonably high administrative capacity", and it has been shown elsewhere to be highly inaccurate (watch this Wahenga space). And the third is that, while the Bank has (belatedly) adopted the "3 Ps" framework for social protection, it has bizarrely dropped the concept of "provision". Its "3 Ps" are promotion, prevention and protection, with the latter introducing a worrying semantic circularity. But the more fundamental problem is that, the way it is defined, "protection" (ie protection " from destitution and catastrophic losses of human capital ") is little different from "prevention" (ie protection "against drops in well-being from income and expenditure shocks") - in that both presuppose that people have something to start with, which needs to be "protected" or its disposal "prevented". And what we have lost from the original "3Ps" is any concept of "provision" - the idea that there are groups of people (who many would argue should be the core target groups for social protection) who have virtually nothing to "protect" or "prevent", and who rely on social protection just to survive. This might include the elderly, the disabled, street-children, marginalised groups. The worry is that by excluding this "provision" function, the Bank is inching SP away from the poorest and most vulnerable, and towards those with economic growth potential ... and therefore better prospects for their "lending operations".

The Red Horse

When the Lamb opened the second seal, I heard the second living creature say, "Come and see!" Then another horse came out, a fiery red one.

The Ten-Percenters are led by UNICEF (though - in fairness - not all of UNICEF). The principal characteristic of their vision of social protection is that it should use community-based targeting to identify the most labour-constrained households among the "ultra-poor", which - somewhat mysteriously - appears to equate ubiquitously to ten percent of the total population. The identified beneficiaries are then given unconditional cash transfers, sometimes with additional incentives if they have school-going children. This approach began with the tiny, but much-feted, Kalomo pilot in Zambia, and it has since been rolled out, with little modification to adapt to differing national circumstances, in Malawi and Liberia. More recently, Zimbabwe and Uganda have escaped with variants on the original.

Here the concern, rather than being with the "3 Ps", is that it has "no Ps": no practical basis, no political support, and no potential! On the contrary, there are multiple problems, many of which have been discussed in earlier RHVP Comments and Briefs. First, it is almost impossible to identify the poorest and most-labour constrained ten percent, especially in countries where over seventy percent of the population are churning in and out of poverty. It is difficult enough even on a pilot scale with extensive external resources and additional "project" funding: it is totally impractical on a national scale. This can lead to problems of moral hazard and even social conflict. Second, even if it were possible, the income from the transfer to the poorest households will result in them rapidly "leap-frogging" up the wealth distribution, so that they will very soon no longer be the poorest. This can lead to problems of equity and the need to retarget frequently. Third, the arbitrary choice of ten percent - essentially a budget rationing tool - cannot be applied universally with any semblance of fairness: there are huge variations in the extent of poverty between, and even within, different districts and communities. As a result, none of the cited examples has been successfully scaled up: and in Zambia, showcase of the Ten-Percenters, the Government has recently expressed its clear preference for a categorical child benefit over the ten percent model that has been pushed so hard and for so long.

The Black Horse

When the Lamb opened the third seal, I heard the third living creature say, "Come and see!" I looked, and there before me was a black horse! Its rider was holding a pair of scales in his hand.

The Instrumentalists are led by ILO, which has rallied an impressive array of UN and other agencies behind its Social Protection Floor(SPF) initiative. Emerging from an employment and "decent work" perspective, the SPF comprises four main components: access to health services; child or family support; income support for the unemployed; and income security for elderly and disabled persons. If you believe the propaganda, examples of the successful implementation of the SPF are everywhere, even in countries where social protection has been evolving at its own pace and in its own way for many years before the term SPF was even coined: these include Mexico, Brazil, Argentina, South Africa, Bolivia, Sri Lanka, Ghana, Haiti, East Timor, etc, etc, etc.

The primary concern with the SPF is to know "which P?": is it a process or is it a prescription? The correct answer is that it is a process. But - for one reason or another - it has been mis-sold, and now suffers from an image problem. Many stakeholders, including some countries and some key donors, perceive it as being too prescriptive, with a mandate to roll out a social protection carpet of standard instruments: health insurance/fee waivers, child grants, unemployment benefits and old-age/disability pensions. To some extent, this misperception in understandable: the ILO has itself recently made a number of presentations describing the SPF as "four essential social transfers" and "four essential social security guarantees", in both cases specifically citing the instruments of "family/child benefits", "social assistance", and "pensions". The ILO public relations machine will have to work hard to retrieve the situation and demonstrate convincingly that the SPF is what it says it is: a recognition that "each country has different social needs, development objectives and fiscal capacity to achieve them, and will choose a different set of policies"; and that the role of the SPF "through a coordinated country response [is to] facilitate and accelerate the introduction or strengthening of sustainable context-specific social protection systems".

The Pale Horse

When the Lamb opened the fourth seal, I heard the voice of the fourth living creature say, "Come and see!" I looked and there before me was a pale horse!

The Universalists are a diverse (and fluid) crowd: they include a number of the big international NGOs, a handful of bilateral donors including the Nordics and DFID ... and - let us be honest - RHVP. The characteristics of their vision of social protection are that it should be rights-based, that it should use predominantly universal approaches and categorical targeting, that it should favour demand-led employment guarantee schemes over supply-led discretionary public works projects, and that it should deliver unconditional cash transfers with no strings attached. Examples can be found in many OECD countries, but also in South Africa, Botswana and Nambia, and (at least partially or incipiently) in lower income countries such as Lesotho, Nepal, Uganda and Zambia.

The primary concern here is again with "3 Ps": how do we pay, pay, pay? There is no doubt that such universal schemes will cost more to implement, and this is often raised as an impossible hurdle. Yet experience has shown that when such schemes are introduced, they are more sustainable: more people benefit from them, understand them, and accept them as fair; so more people vote for them; so they become more popular politically; so they attract a greater share of the national budget. And so, ultimately, it can be argued that the poorest get more benefit from universal schemes than from programmes targeted more narrowly at them. One option to defray the costs is to introduce such schemes progressively: in other words, to set eligibility for a social pension at a relatively high age such as 70 and reduce it gradually, or to start a child benefit at a young age such as 2 and then increase it over time (like South Africa has done), as the evidence of beneficial impacts generates increased public and political support. Experience shows that governments will devote much more substantial shares of their budgets to universal programmes than to poverty-targeted ones. For example, the universal social pensions in Mauritius, Lesotho and Nepal command more than double the resources as a share of GDP than the means-tested pensions in Cape Verde, Argentina and India; and it is often overlooked that the budget of the vaunted poverty-targeted Bolsa Familia in Brazil is dwarfed by that of the country's near-universal rural and urban pension programmes.

Angels and Demons

They were given power over a fourth of the earth to kill by sword, famine and plague, and by the wild beasts of the earth.

The funny thing about the Four Horsemen of the Apocalypse is that no-one really knows what they mean. The popular, more dramatic view is that they symbolise tribulation, and are the harbingers, respectively, of conquest, war, famine and death. But there is another school of thought that equates them with the angels of the four winds (Michael, Gabriel, Raphael and Uriel), who are often associated with the four cardinal directions. And this seems more appropriate in terms of our Four donor Horsemen: the danger is that, whatever the strengths and weaknesses of their competing visions, they are pulling in four opposite directions. And this can have disastrous consequences at national level. Let us look at how this has played out in one poor country: Malawi.

Malawi has been trying, since 2007, to develop a comprehensive national social suport policy. During that time:

1) The Productivists (aka the World Bank) have been (a) pursuing their investments in public works through the Malawi Social Action Fund (MASAF), and (b) experimenting with a pilot Conditional Cash Transfer in Zomba;

2) UNICEF has been pushing its Ten-Percent model, with a pilot first in Mchinji, and then expanding  to a handful of other districts;

3) ILO and its Instrumentalist partners have been assiduously fixing the joists for their Social Protection Floor;

4) DFID has commissioned some like-minded Universalists - HelpAge, EPRI and (it must be admitted) RHVP - to argue the case for a universal social pension.

The result: stalemate. No wonder Malawi's Cabinet has still not approved the policy, five years on!

In the meantime, the Government has shown very clearly where its own "social support" policy priority lies: in its agricultural input subsidy programme. So what should the donors have done? Instead of  declaring that they didn't approve of subsidies, instead of insisting that the national social support policy be approved before they did anything, and instead of pursuing their opposing policy agendas, they should rather have:

·       jointly supported the Government;

·       explored integrated ways to extend social transfers to the poor with no land or labour (as in the direct welfare support components of Ethiopia's PSNP and Rwanda's VUP);

·       explored options for cash transfers to those who routinely sell their agricultural input vouchers, on the basis that there must be cheaper and more efficient ways to provide them with the cash they need;

·       facilitated the inclusion of civil society and of beneficiaries in the national social protection debate;

·       invested in necessary upfront expenditure, such as national identity systems;

·       helped to build capacity at national and sub-national levels;

·       promoted the institution of independent grievance and appeals procedures, and of social audits;

·       explored innovative delivery systems;

·       leveraged maximum private sector participation in the programme;

·       funded a futures option on maize imports (as DFID and the World Bank had done successfully in 2005) to provide insurance against the next year when (not if) the harvest fails;

·       invested in monitoring and evaluation to learn lessons and inform the evolution of the programme.

What a missed opportunity!

In conclusion, social protection still has a huge role to play in development; but it will only achieve its full potential if the Four Horsemen of the Donor Apocalypse reconcile their differences and unite to ride in the same direction as individual national Governments. 

Tuesday, 20 July 2010

And now for something completely different: Contraceptional Cash Transfers

This blog first appeared on the Wahenga website, under the pseudonym of Sissy Teese

The World Bank can only be doing this to provoke me! Just when I had vowed never to take up my pen in anger again (and when Wahenga had vowed never to air my pseudonymical rants) … they do this!!

We already know, because the Bank has told us, that Conditional Cash Transfers (CCT) can do everything! First we had Conspirational Cash Transfers. Then we had Connubial Cash Transfers. And now we have Contraceptional Cash Transfers. A paper just launched describes an experiment - “Rewarding STI Prevention and Control in Tanzania” (with the horrendously contrived acronym of RESPECT, no less!), which provides quarterly cash transfers, each equivalent to nearly one-tenth of average annual income, to those who avoid unsafe sex!

We hear that the study “has three separate arms [all the better to hold you with, my dear!!] – a control arm and two intervention arms (low-value cash rewards and high-value cash rewards).  Study participants were randomly allocated across the three study arms. All participants have been monitored on a regular basis (every 4 months over a 12 month period) for the presence of common sexually-transmitted infections (STIs) that are transmitted through unprotected sexual contact and therefore serve as a proxy for risky sexual behavior and vulnerability to HIV infection.  A small payment has been provided to all participants (regardless of arm assignment) to minimize attrition from the study.  Anyone testing positive for an STI (again, regardless of arm) received free STI treatment and counseling. Individual pre-test and post-test counseling was provided to study enrollees [the word “enrollees” has a wonderful ring to it in this context!!] at each testing interval, and monthly group counseling sessions were also made available to all study participants in all villages.”

And it works! And, what is more, it seems that the more you pay, the better it works!

By the end of one year, “study participants who were randomly selected to be eligible for a $20 payment every 4 months if they tested negative for a set of curable STIs, experienced a 25% reduction in the incidence of those STIs. After one year, 9% of individuals in the group that received the $20 quarterly payment were positive for one of the STIs, compared to 12% in the control group”.

We also learn that “the two year study cost $1.8m”, which must surely make it the most expensive condom in African history. And maybe this is where the reality check comes in: the finances. Let us look at this from two different perspectives.

First, the experiment. We have a sample of 2399 (why not 2400?). Of these, let’s assume one-third (remember the three arms?) received the higher transfer value - i.e. 800 people. Of these, 3% (i.e. the difference between 9% and 12%) avoided STIs as a result of the transfer - that is 24 individuals. So it cost probably 800 x $60 in transfers, plus all the costs of testing and administration - let’s say conservatively $100,000 in total annually - to prevent 24 individuals from getting an STI. That is over $4,000 each. Condoms would have been a cheaper option.

Second, consider this programme scaled up, nationally, to cover the whole of Tanzania. To be effective, it would need to be targeted universally, at all who were sexually active, so between the ages of 12 and 60. That is more than 22 million people. The pilot cost $750 per person, but we can assume some economies of scale, so let’s say $500 per person. That is a total programme cost of $11 billion, equivalent to 50% of GDP at official exchange rates. By way of comparison, the ILO calculated that Tanzania could afford a full “basic social protection package”, comprising a universal pension for all over-65s and all persons with disabilities, a universal child benefit to all under-15s and universal access to basic health care … all for 17% of GDP. Alternatively, $11 billion could buy you 110 billion condoms, more than five times the number currently produced worldwide annually!

But - since money is no object - what next for the World Bank? How else might they influence people’s behaviour for the better? My vote is for some form of Conflagrational Cash Transfer, providing nicotine vouchers to all who refrain from smoking cigarettes … or maybe Congregational Cash Transfers, distributing multiples of the collection plate to all who faithfully attend church every Sunday … or a Consumptional Transfer of unlimited maize meal to all who eschew junk food for the month … or a Convivial Cash Transfer to those who moderate their alcohol intake … or perhaps Constipational Cash Transfers, providing monthly supplies of suppositories to all those performing their ablutions regularly on a daily basis.

Maybe my writing days aren’t over yet …



Monday, 31 May 2010

An appeal against "Conditionalities"

As RHVP - and perhaps Wahenga - draw to a close, I would like to use our pages to make a personal and heartfelt appeal to the social transfers community: please can we stop using the horrible word “conditionalities”?

I would like to stress that this plea has nothing to do with the debate about the relative merits of conditional and unconditional cash transfers that has been raging on Wahenga’s pages recently (Sissy Teese v World Bank): it is aimed squarely at writers at both ends of that seemingly unbridgeable spectrum. But let me at least draw on those Wahenga exchanges to quote a couple of examples:

“why bother with the moral hazard, additional cost and complexity of imposing, monitoring and enforcing conditionalities, when unconditional programmes appear to have the same effects?” (Ms Teese) 

“[CCTs] combine three key mechanisms: grants that increase the income of poor households, awareness promotion that emphasises the importance of human capital, and conditionalities that link the two” (EPRI “Designing and Implementing Social Cash Transfers” [2006], quoted by Ms Teese)

The danger is that this ugly terminology is becoming firmly entrenched in the literature: it is used over 200 times in the EPRI manual cited above[i], it has sneaked into the World Bank’s seminal “tome” on “Conditional Cash Transfers” (2009), and it is being replicated in countless papers, articles and journals.

What is wrong with the word “conditionalities”? First, it is not a word; second, if it were, it would mean something quite different; and third, it is wholly unnecessary, since we already have a perfectly good word that we can use in its place.

1)    It is not a word because it does not appear in any dictionaries, and because the irreproachable Microsoft Word gives it a red squiggly underline, in whatever language you are using.

2)    Even if were to be coined as a neologism, it would be meaningless: we can apply a “condition”, or multiple “conditions”. We can thereby make things “conditional” (ie “dependent on the fulfilment of one or more conditions”); and we can therefore have a system that is based on the concept of “conditionality” (ie the “state of being conditional”). But we cannot have multiple “conditionalities” - and, if we could, it could only mean “multiple states of being conditional”.

3)    So, if we can’t use the word “conditionalities”, what can we use instead? We need a word that has the definition of “actions stipulated as requirements before the performance or fulfilment of other actions”. Does such a word exist? Miraculously, it does: the word is … “conditions”! Try it for yourself in each of the examples cited above: you must agree that it improves them!

So this is my plea: let us have no more “conditionalities” in social transfers! But - lest this sounds as if I am aligning myself with the sassy Sissy - let us equally have no hesitation in applying “conditions” to cash transfers wherever they are agreed to be appropriate!

It is important to nip this awful usage in the bud now, because the use of conditions in social transfers is clearly a ubiquitous and fast-growing phenomenon, both in the World Bank’s portfolio and elsewhere. Or to put it another way: without wishing to cause confrontationalities, in order to have confirmationalities of the proliferationalities of the utilisationalities of conditionalities in social transfers, in both connurbationalities and rural locationalities, you need only to make random observationalities in the World Bank’s “Manual of Operationalities”!



[i] The otherwise excellent EPRI manual is to be republished soon. Please can we ensure that it drops the use of the term “conditionalities” in its second edition?



Monday, 15 February 2010

Connubial Cash Transfers

This blog first appeared on the Wahenga website, under the pseudonym of Sissy Teese

With some reluctance, I take up the cudgel again!

I have recently had the good fortune to read the Bank’s latest Policy Research Working Paper (No 5259)[i]. This is another fascinating, meticulously researched and well-argued paper about the same cash transfer programme in southern Malawi that was cited in our earlier wahenga exchanges about the impact of conditionality on schooling - the Zomba Cash Transfer Program (ZCTP). Entitled “Cash or Condition”, the Paper’s specific objective is to disentangle the impact of the condition from the impact of the cash transfer by reporting on the first “ideal experiment to answer this question – i.e. a randomized controlled trial with one treatment arm receiving conditional cash transfers, another receiving unconditional transfers, and a control group receiving no transfers” [emphasis in original].

Not even I could fail to be impressed by the refreshing candour of the Abstract, which states unequivocally that:

 “the authors find that the program reduced the dropout rate by more than 40 percent and substantially increased regular school attendance among the target population of adolescent girls. However, they do not detect a higher impact in the conditional treatment group” [emphasis added].

After such an admirable display of openness, it might seem churlish to complain! But there are a couple of more worrying findings which emerge from the Working Paper that shed further significant light on the debate between CCTs and UCTs. One of these is clearly recognised in the paper (although, noticeably, it doesn’t make it into the Abstract!), while the other is tucked away unobtrusively in a footnote.

The first is that UCTs dramatically reduce the probability of early marriage, while CCTs do not:

 “the unconditional treatment reduced the probability of marriage by 2.7 percentage points (or by 56%), whereas the marriage rate in the conditional group was identical to that in the control group”.

The paper suggests a dispassionate economic argument to evoke “the possibility that a CCT offer could actually trigger marriage by presenting the girl with an untenable schooling alternative and forcing the household into a decision [of early marriage] at the time of the offer” [emphasis added]. But might there not be another, more human, explanation: that beneficiaries of unconditional transfers feel more empowered, and more independent, and more confident of making sensible life-choices than recipients of conditional transfers from an intrusively paternalistic nanny-state? This would after all, conform to standard economic theory, helpfully reiterated in this Working Paper, that “in the absence of externalities, conditional cash transfers are worse than distributing an equivalent amount of unconditional cash”. 

The second, perhaps even more serious, concern is tucked away in a footnote:

 “the program led to substantially elevated stress and psychological morbidity among adolescent girls in the conditional group relative to the unconditional arm”.

Again, the implication is clear. The imposition of conditionality places additional stress on the recipient, and creates tensions within the household between those that must adhere to the conditions (in this case, the school-age girls) and those that benefit from the transfer (siblings, parents, carers, grandparents). What is worse, the authors “also find that the mental health of the CCT recipients worsen [sic] when the transfer amount offered to the parents is larger, while the mental health of UCT recipients is uncorrelated with the transfers offered to the parents”. In other words, the higher the value of the transfer that rests on fulfilling the conditions, the greater the stress for CCT recipients.

Surely these are serious concerns, and a reason to argue more forcefully in favour of unconditional transfers? What this impressive study tells us is not only the already important message contained in the conclusion that: 

“given that the marginal impact of imposing a schooling conditionality is at best low [it is in fact absent!], and that monitoring school attendance to enforce the conditionality is costly, it seems that policymakers can consider unconditional cash transfers as a viable alternative”.

It also raises two potentially damaging criticisms of conditionality, and therefore two significant advantages of UCTs, which - at the very least - merit critical attention and further research. These are:

1)    That, “while unconditional cash transfers nearly eliminate marriage in our study population, the conditional cash transfers have no effect on it”. Early marriage is seriously detrimental to girls’ education (in Latin America and elsewhere, as well as in Malawi), so reducing it, through whatever mechanism, would be a highly desirable outcome in building human capital and reducing the inter-generational transmission of poverty.

2)    That - as might indeed rationally be expected - imposing conditions can inflict psychological damage and create intra-household tensions. Instead of feeling empowered and independent, recipients may feel bullied, pressured and ultimately overwhelmed by the conflicting demands imposed on them, forced into making behavioural choices (like early marriage) that they would otherwise eschew.

Should these two issues not have been given greater prominence in a truly balanced Working Paper, intended as it is to “inform policymakers as to which combination of contract parameters might allow cash transfer programs to deliver the largest impacts per dollar spent”? Should policymakers not be told, upfront, that, on the evidence emerging from Malawi, unconditional transfers have identical benefits in terms of school enrolment and attendance to conditional transfers, while being far less complex and less expensive to implement; and that they appear to be much more effective in reducing disruptive early marriage and minimising psychological stress?



[i] Baird, S., McIntosh, C. and Özler, B., ‘Cash or Condition? Evidence from a Randomized Cash Transfer Program’, World Bank, March 2010



All social protection interventions are equal, but some are more equal than others

  This blog originally appeared on  Development Pathways  ( with apologies to George Orwell, Animal Farm [1945]) I recently came across a ...