Monday, 4 February 2013

Transformative Graduation?

Photo of Stephen Devereux, IDS Vulnerability and Poverty Reduction Research Fellow By Stephen Devereux

In social protection programmes, “graduation” refers to participants crossing an income or asset threshold, after which they are no longer eligible for support. In January 2013 the Centre for Social Protection (CSP) hosted an online discussion event on the topic of graduation, under a partnership agreement between Irish Aid and IDS on ‘Hunger Reduction and Climate Change Adaptation’. The debate was lively and helped clarify my ideas on graduation (which are personal and do not necessarily reflect the views of CSP, IDS or Irish Aid).

I discovered that I am a graduation sceptic. Of course graduation might be possible for some households at some times in some contexts, but probably only a minority under very specific conditions. It bothers me greatly that graduation is increasingly driving social protection programmes, because this is not what social protection is all about.

 A basic distinction should always be drawn between three functions of social protection. In the “3 Ps” terminology, social protection should “protect” the chronically poor, “prevent” the vulnerable from being decimated by uninsured shocks, and (thirdly and lastly) “promote” the livelihoods of the entrepreneurial poor.

The first function is social assistance for the "ultra-poor", such as social pensions and disability grants. Pensioners and people with severe disabilities are not expected or required to "graduate out" or exit from this social assistance after a few years – nor should they be.

The second function is the original (now regrettably much maligned) safety net role – ensuring that people who are vulnerable to shocks (because they lack savings, insurance or social security) don't collapse into destitution or worse when their livelihoods are disrupted by drought or unemployment or AIDS. Again, nobody – poor or rich – graduates out of the need for safety nets. No-one who can afford insurance cancels it after 5 years.

The third function does not have a "social something" label, because it actually isn't social protection; it's livelihood support for poverty reduction (the "promote" P). The people who are likely to graduate from grants and subsidies to micro-credit and eventual self-reliance are not the labour-constrained "ultra-poor", they are low-income smallholder farmers and self-employed workers in the informal economy – people with labour power whose livelihoods need a boost. They are the “entrepreneurial poor” – but, as we know from the microfinance experience, not all poor people are entrepreneurs, so even in this category, 100% graduation is 100% unrealistic.

Nonetheless, this is the poverty reduction agenda into which social protection has become co-opted. It is politically sexier than social grants and it is massively popular with donors and governments who fear the cost and commitment of permanent programmes, who want an "exit strategy", “value for money” and a "return on their investment", and whose vision of success is driven by the MDGs. So graduation looks like a "win-win": programme participants get their poverty reduced, programme implementers and policy-makers hit their poverty reduction targets. No wonder some social protection programmes are holding graduation ceremonies.

But what if you don't hit your graduation targets? I was disheartened to hear an Ethiopian policy-maker in a conference in 2011 describe the Productive Safety Net Programme as a “failure”, because “we are expecting most of the beneficiaries to graduate but instead the numbers just go up every year”. If political support for programmes that are fixated on unrealistic graduation targets starts to dwindle, it isn’t just the “Productive” components that will be closed down; the “Safety Net” will go too.

Graduation is not social protection. Although "promotion" was included In the "transformative social protection" framework it was never intended to stand alone. Our argument was that some forms of social protection provide transfers that simultaneously alleviate poverty ("provision") while at the same time invest in useful physical infrastructure (e.g. public works = wages for subsistence + economic assets for income generation) or human capital formation (e.g. conditional cash transfers, or school feeding = cash/food today + education for enhanced livelihoods tomorrow), which have the potential to generate future streams of income and thereby reduce poverty sustainably (i.e. poverty alleviation + poverty reduction).

The danger is that seeing social protection as a poverty reduction instrument loses sight of its primary functions – to provide social assistance to alleviate the poverty of the labour-constrained "ultra-poor" who have no livelihoods (and negligible prospects of graduating); and to install effective and reliable safety nets or social insurance mechanisms against downside risk for people with low and/or unpredictable incomes. In both cases social protection is a permanent commitment; it is not a 3-5 year project. You don't put a safety net in a circus until you believe the trapeze artist won't fall, and then remove it. We must not replace the critical core functions of social protection with objectives drawn from other agendas.

But I am not a total graduation sceptic. I do believe that social protection can build resilience and self-reliance, but it might take a decade or longer – maybe even a generation, which is why many social protection programmes invest in the human capital of children, in an attempt to break the intergenerational transmission of poverty. We should surely aim for ‘sustainable graduation’ over time, rather than ‘threshold graduation’ based on income or asset ownership at a point in time. This requires a broader assessment of the problem that social protection is trying to address.

The impacts of a social protection intervention will always be heavily influenced by the context into which it is introduced. There needs to be a favourable economic “enabling environment” – good infrastructure, strong public services, well developed markets, financial services, vibrant economic activity, complementary linkages to other economic sectors – for graduation outcomes to be achievable and sustainable. The problem with naive forms of graduation thinking is that is tries to propel people out of poverty simply by throwing cash and assets at them, while leaving them trapped with the adverse personal characteristics (illiteracy, etc.) and structural constraints (weak infrastructure, imperfect markets, etc.) that constructed their poverty in the first place.

Beyond these economic “enablers or constrainers”, if the socio-political context is unfavourable to the extent that it is driving poverty, then no injection of cash or assets is going to lift people out of their poverty trap. Social exclusion of minorities, discrimination against stigmatised groups, xenophobia – these are all social rather than economic drivers of poverty and vulnerability, and they require interventions at the social and political levels – social transformation, not income or asset transfers.

“Transformative social protection” recognises that social change and political interventions can remove the structural barriers and trigger virtuous cycles of income generation and poverty reduction. Progressive legislation or anti-discrimination campaigns could be enough on their own to “graduate” entire sections of society. One example is abolishing discrimination in the labour market against persons with disability or HIV-positive job applicants, thereby “graduating” those who can work, from social grant recipients to wage-earning employees. We could go further and argue that, where socio-political constraints are significant drivers of poverty and vulnerability, transformative social protection is the only assured pathway to “transformative graduation”.