Why don’t we just send aid money directly to poor people’s cellphones?

Just before Christmas I had a thought-provoking discussion on the BBC World Service with Paul Niehaus, who has set up GiveDirectly, a US-based startup NGO pioneering a new financing model based on cash transfers. The idea couldn’t be simpler:

1. People donate through GD’s webpage
2. GD locates poor households in Kenya (see below)
3. GD transfers your donation electronically (through the M-Pesa mobile payments system) to a recipient’s cell phone (they send each household $500 per year for two years)
4. The recipient collects the transfer

GD reckons that in this way, it can get 90 cents in every donated donor into the hands of poor people. Step 2 is interesting: ‘We do this in three steps.  We first select regions of Kenya with high poverty rates using census data.  We then identify villages with low-quality housing and access to an agent providing mobile-phone-based payment services. Finally, we identify the poorest households in these villages using simple, transparent criteria: we target all households living in homes made out of mud, wood, and grass. These criteria effectively identify relatively poor households and are generally perceived by the community as fair. We record eligible households’ phone numbers or, for those that do not have a phone, provide them with a SIM card. We follow up initial identification with a rigorous process of audits to prevent mistakes or fraud.’

What’s innovative about this is the coming together of cash transfers (CTs) and mobile payments systems to make the CT option available to individual donors, rather than (as previously) being exclusively a government, big aid donor or large NGO activity (Oxfam does lots of them – in fact it was our cash for coffins project that partly gave Paul the idea).

There was a high level of agreement in the BBC discussion (doubtless to the horror of the producer – arguments make much better radio). This kind of approach is exciting, but only relevant to part of the aid and development story – for example in the Horn of Africa, we are doing cash transfers, but also have to work to get market traders to re-establish supply chains in the worst-hit areas or there is nothing for people to spend the transferred money on.

While they help with short term consumption and investment, cash transfers don’t directly tackle the kinds of systemic problems that underpin poverty and inequality – dealing with those requires a more complex approach based on partnering with local civil society organizations, and all that brokering and convening stuff I write about on this blog. And what about gender – who owns the phones and gets access to the $500? It would be interesting to see if there’s a difference between how men and women phone-holders spend the money – I wonder if GD have included that in their monitoring and evaluation?

Finally the approach seems inherently individualistic – there is no obvious way to fund community organizations in this model. At least not yet. I talked to our fundraisers prior to the interview and they linked this to a generational shift. Younger people are less trusting of institutions than older ones, so the pressure for this kind of person-to-person ‘disintermediation’ (sorry) is only likely to grow. People only believe their money is doing good if they can see it drop into the hand (or cellphone) of a recipient. Oxfam has already responded to this with schemes like Projects Direct, and new initiatives like Kiva and now GiveDirectly are addressing the same disquiet.

An alternative approach is to do a better job in explaining why we need to use people’s donations to tackle the underlying structural causes of poverty, through a more complex (and expensive) engagement with the state, companies, civil society organizations etc.

But another might be to put the two together. If the new generation is both more activist (Occupy, Arab Spring etc) and more sceptical of institutions, how about adapting the GiveDirectly model to ‘sponsor an activist’? Your $10 a month would go straight to the cellphone of a named HIV activist, or a land rights organizer. In return you would get regular tweets, blogs or whatever so you can follow what they’ve been getting up to. Paul says he’s going to think about the idea, but is anyone already doing it? If so, how’s it going?

You can listen to the piece here, with Paul Niehaus of GiveDirectly, Mike Jennings of SOAS and me, (although it might have got a bit truncated at either end of the 7 minute piece).

Poverty reduction v well-being: a cash transfer experiment from Malawi

What difference does it make to development interventions if you worry about well-being rather than income poverty? A rather neat example has just come through from some new research by Sarah Bair, Jacobus de Hoop and Berk Özler for the World Bank Poverty and Inequality team. They looked at the impact on girls’ mental health of cash transfers in Malawi (why do so many researchers work on Malawi? They must be reaching plague proportions – a researcher poll tax would probably solve the country’s problems overnight).

The researchers used a randomized control trial (no surprise there, then) involving nearly 4000 girls to compare the psychological impact of unconditional cash transfers with making them conditional on school attendance and uncovered some striking differences.

“The provision of monthly cash transfers had a strong beneficial impact on the mental health of school-age girls during the two-year intervention. Among baseline schoolgirls who were offered unconditional cash transfers, the likelihood of suffering from psychological distress was 38 percent lower than the control group, while the same figure was 17 percent if the cash transfers offers were made conditional on regular school attendance.”

The researchers concluded that the main psychological benefit stemmed from improved family income, rather than school attendance, and that making them conditional on school attendance put sufficient stress on the girls to undo a lot of the benefits. ‘When an important source of income for the family depends on the actions of the adolescent girl, it might place a heavy burden on her and to cause adverse effects on her mental health’ eroding over half the psychological benefits of the cash transfer.

Conclusion?

‘Overall, the results presented in this paper indicate that mental health among adolescent girls can substantially improve when they experience positive income shocks. However, if these income shocks are administered as part of a cash transfer intervention such as the one examined here, these mental health benefits can also be quickly eroded if sufficiently large payments are made to the parents conditional on the actions of their adolescent daughters.’

So if you focus on income poverty, conditional cash transfers offer you a double bonus – direct poverty alleviation, and greater school attendance leading to lower poverty in future generations. But if you focus on wellbeing, the equation becomes more complex. Interesting.

How to halve world poverty – ignore J. Paul Getty

An intriguing new paper from Paul Segal at the University of Sussex kicks off with a great quote from Mr Getty “The meek shall inherit the Earth, but not its mineral rights”. Segal wonders what might happen if governments defied that prediction, and just handed over the income from oil, mining etc directly to poor people as unconditional cash transfers to all citizens (i.e. a universal benefit).

He calls the scheme the Resource Dividend (RD) and actually it may not be that crazy or utopian – Alaska has been doing it for years, handing over $600-$1500 a year to every Alaskan citizen, and Iran is currently considering introducing such a system to pay out $60 per person per month. When Britain first discovered North Sea oil, two FT journalists, Samuel Brittan and Barry Riley wrote, “The simplest and also the wisest answer to the question ‘What should we do with the state’s oil revenues?’ is ‘Give them to the people’”.

Why now? Segal argues that ‘two developments make its more general application particularly relevant today. First, resource nationalism and resource ownership rose in importance amid the dramatic rise in resource prices up to mid-2008. Second, the first Millennium Development Goal, adopted by the United Nations in 2000, is to halve global poverty at the $1 a day line from its 1990 level by 2015. I estimate the global impact of the policy on poverty and find that if enough poor countries were to adopt the RD then it would be sufficient to achieve the first Millennium Development Goal: extreme global poverty would be cut by half.’

Segal calculates the impact using World Bank data on resource rents and $1.25 a day poverty, for two different cases. In the first, governments simply replace the lost revenue with taxes on non-poor people. In the second, everyone, including those below the poverty line, pay taxes proportional to their post-RD incomes (so the impact on poverty is slightly reduced). The overall results are shown in the table.

He sets out the advantages of an RD scheme:

“First, it would substantially reduce poverty. Second, by being levied only on rents, the scheme implies none of the economic distortions or efficiency loss that other redistributive schemes may risk. Third, it provides an incentive to informal workers and individuals with little or no formal interaction with the state to register with the fiscal system. Finally, there is a moral and legal argument that by the nature of rents, no individual has a special claim to them, so the only morally defensible distribution is an equal distribution.” Universal benefits are also relatively simple to administer (no means testing) and the RD has an automatic stabilizer quality, as food prices tend to vary with oil and minerals, so the increased RD would cushion people during food price spikes.

In volume terms, the RD would come in at under 6% of GDP in most countries – less than the cash benefits in the EU15 (6.6%). It’s like a jump start to a welfare state.

He contrasts the simplicity and ready-to-go nature of the RD with the complexity of boosting economic growth – the standard recipe for reducing poverty. The RD scheme does not require growth, or even new taxes – merely the redirection of existing resource rents to poverty reduction.

Interestingly, even countries not known as major commodity exporters would benefit. India, where poverty now affects nearly half a billion people despite high growth rates, would see poverty fall from 42% to around 20%, even though its RD would amount to just $2.90 per person per month.

The obvious flaw is political – why would governments do it, especially if they have to replace the lost revenue with more unpopular taxes? Segal briefly considers this, arguing that “Incumbent governments are likely to be reluctant to give up an easy-to-collect source of revenue. One can imagine it being proposed in a democracy by an opposition party in order to get elected, or by a government (democratic or not) that decided that such a popular measure may increase its chance of political survival.”

‘Just Give Money to the Poor: the Development Revolution from the Global South’, an excellent overview of cash transfers

Cash transfers (CTs – regular payments by the state directly to poor people) are all the rage at the moment, prompting heated debates across the development sector. As its title suggests, a new book, ‘Just Give Money to the Poor’ has no doubts about their merits. But Joseph Hanlon, Armando Barrientos (see his blog on the book here) and Hulme are academics with a long track record on development, and their conclusion is not a crude polemic, but an excellent and readable compendium of country experiences, arguments and research findings.

Summarizing the research, they say ‘Four conclusions emerge repeatedly: These programs are affordable, recipients use the money well and do not waste it, cash grants are an efficient way to directly reduce current poverty, and they have the potential to prevent future poverty by facilitating economic growth and promoting human development.

But two areas remain the subject of intense debate: targeting and conditions. Should smaller grants be given to many people, or larger grants to a few? Should recipients be asked to satisfy conditions such as sending their children to school or doing voluntary labor?’ Intriguingly, there is very little evidence that conditions have any actual impact, by the way.

The authors try to locate these discussions in the big historical picture, arguing that CTs represent a paradigm shift on a par with the creation of the welfare state in the North, and moreover constitute a ‘genuinely Southern revolution’, in response to the efforts of a failed northern aid industry. They have been pioneered by the new generation of G20 leaders – South Africa, Brazil, China, India, Mexico and Indonesia. The big shift is away from the insurance schemes preferred in northern welfare states, to direct government-funded schemes that recognize that self-insurance isn’t a viable option for poor people. Posing this as a crude ‘North v South’ argument feels rather retro in these multipolar times, but it’s certainly an interesting frame.

The book summarizes the current state of CTs as follows:

‘At least 45 countries in the Global South now give CTs to more than 110 million families. Every program is different, from universal child benefits in Mongolia to pensions in Africa to family grants in Latin America. Some grants are tiny – only $3 a month – whereas others give families more than $100 a month; some cover more than one-third of the population, and others aim only for the very poorest. The size of public spending varies from 0.1% of GDP to 4%, although most programs fall in the range of 0.4% to 1.5%.’

The four main goals of CTs are social protection and security (for the young, old, disabled); development and economic growth (CTs give poor people the security they need to invest in higher risk/return options like new crops, or migrating in search of work); breaking intergenerational poverty (by ensuring children are better nourished and educated than their parents) and rights and equity (reducing income inequality and promoting the status of women).

And one important caveat: CTs increase demand for things like schools and hospitals, but that can be pointless if their quantity or quality does not respond in turn – the book argues that ‘co-responsibilities’ between citizens and state is a better frame than ‘conditions’ on grants. The same holds for ensuring that decent jobs await this new generation of health, educated people.

The discussion I found most interesting was on the politics of CTs – when/why are they introduced? How do they spread? They usually start small, both in terms of target group and geography (eg in Brazil, India and China, they started off as regional initiatives and were then picked up by national governments); often politicians introduce them either to head off unrest during a crisis, or as vote winners before an election, but they then take on lives of their own (it would have been good to know more about when this does/doesn’t happen – what distinguishes CTs from old school patronage? Why are similar programmes introduced by governments of very different political stripe?); pensions and child benefits are the best entry points, as they rapidly win political consensus and public support.

The authors tie CTs’ new popularity with southern governments to the resurgence of interest in the developmental state, and the decline of the Washington Consensus. CTs are the basis for a new social contract between citizen and state, in situations where informalization of the economy make taxation (the historical basis of the contract) an unrealistic foundation.

There is often a fascinating conflict between the technocrats’ urge for tight targeting to ‘maximise efficiency’ and public rejection in favour of universality – in Mongolia this led to a targeted programme being replaced by a more popular (but more expensive) universal child benefit. Targeting appears particularly divisive when only a few cents divides the poor and the non-poor, as in many low income countries. In those circumstances, the nominally poor suddenly become much richer than their non-poor neighbours, which, in a lovely throwaway line ‘raises questions of witchcraft and nepotism.’

The path has been smoothest when CTs are nationally owned and driven – donor pressure appears to be counterproductive in many cases, being interpreted by local elites as having their arms twisted to give cash to lazy, feckless wasters.

How to fund it all? The authors are aid-sceptics and argue that domestic revenues should be the preferred source, but also raise an interesting alternative for the poorest countries. CTs require a big pot of cash, without donor conditions or loads of reporting red tape. What better source than a financial transaction tax?

For a comprehensive survey of practice or the policy dilemmas and choices surrounding CTs, I have yet to find a better place to start than ‘Just Give Money to the Poor.’

Why conditional cash transfers can prevent HIV

Conditional Cash Transfers, in which poor families receive regular payments from governments or aid donors on condition they keep their kids in school, or get them vaccinated, are all the rage at the moment. They are seen as effective ways to reduce poverty, cushion poor families against shocks and get kids into school, but it seems their benefits extend even further. New research from the World Bank even suggests they reduce the risk of HIV infection. ‘The Short-Term Impacts of a Schooling Conditional Cash Transfer Program on the Sexual Behavior of Young Women’ summarizes the results of a randomized control trial in Malawi, involving nearly 4,000 young women, comparing those who received a $10 a month CCT for a year, with a control group in the same community who did not.

Main finding: ‘After one year, the program led to large increases in self reported school enrollment, as well as declines in early marriage, teenage pregnancy, sexual activity, and risky sexual behavior. The evidence presented here suggests that as girls and young women returned to (or stayed in) school, they significantly delayed the onset (and, for those already sexually active, reduced the frequency) of their sexual activity. The program also delayed marriage – which is the main alternative for schooling for young women in Malawi – and reduced the likelihood of becoming pregnant.’

This matters because young women in sub-Saharan Africa are nearly three times more likely to be HIV positive than young men. It’s received wisdom that girls’ education is a ‘social vaccine’, but according to the paper, the evidence for this comes from snapshots in time and comparisons between countries. By tracking a single community over a year, the Malawi study builds a much firmer case not just for the health benefits of girls’ education but also of CCTs. All this in addition to the intrinsic benefits of education, of course.

Giving cash to poor people and reducing inequality: lessons from Latin America

Two interesting ‘one pagers’ from the consistently excellent International Policy Centre for Inclusive Growth, run by the UNDP and based in Brazil. In ‘Do Conditional Cash Tranfer (CCT) Programmes Work in Low-Income Countries?’ Simone Cecchini of ECLAC takes the well-known successes of cash transfers in large middle income countries such as Brazil (Bolsa Familia) and Mexico (Oportunidades) and evaluates efforts to replicate them in their much poorer neighbours in Central America. CCT programmes pay poor families a monthly amount provided they meet certain conditions, such as keeping their kids in school or getting them vaccinated.

The scale of CCT is much greater in the middle income countries. In Brazil, Bolsa Familia reaches 22% of the total population and accounts for 2% of total government spending, in Mexico the numbers are even higher (see table); by contrast, in Nicaragua CCT reaches just one in forty people and accounts for only 0.4% of public spending. Cecchini argues that this is explained by some of the institutional weaknesses that characterize small, low income countries. For CCT programmes to work:

- because of their multidimensional approach to poverty, CCT programmes require coordination between different ministries and levels of government (eg local and national). In low income countries weak state machineries struggle to do this.

- state policy mustn’t change every time a new government is elected, or aid donors change their minds, both of which are more likely in poorer countries without strong bureaucracies

- CCT programmes require solid data and payment systems, but low income countries often have weak statistical capacity and fragile banking systems – in Guatemala, payments are often made in cash at public events attended by the first lady, which feels more like old school patronage than CCT.

Cecchini concludes that cash transfer programmes may have to be adapted to low income countries, for example putting more emphasis on targeting poor regions rather than means testing, which is expensive and not much use when nearly everyone is poor. Insisting on conditioning cash payments on school attendance or health check-ups may not make sense when schools and clinics are either absent or of dismal quality.  In that case, the government has to sort out supply of essential services rather than just focus on increasing demand. For her full paper (in Spanish) see here.

Meanwhile IPC’s ‘What Explains the Decline in Brazil’s Inequality?‘ returns to the explanations of Brazil’s success in reducing inequality. From 2001-2007 Brazil’s gini coefficient, a standard measure of inequality, fell sharply from 0.59 to 0.53. Over that period, the poorest 60% of Brazilians, who receive only 18% of national income, accounted for 40% of total income growth.

The authors, Degol Hailu of the IPC and Sergei Suarez Dillon Soares of the Brazilian parastatal thinktank IPEA, crunch the numbers and find three factors driving the improvements:

- a third of the fall in the gini index comes from improved access to education (presumably through its knock-on effect on incomes).

- a third comes from the kind of social protection programmes discussed earlier.

- a third comes from broader improvements in social and economic policy, falling unemployment etc, which have increased domestic demand and consumption.

See my previous post for more on Latin America’s significant progress in reducing inequality in recent years.

Cash for Coffins? What happened when Oxfam gave poor Vietnamese a lump sum

I’ve just been reading the latest evaluation of an Oxfam project I’ve started to call ‘cash for coffins’ in Viet Nam. From mid-2006 Oxfam GB directly disbursed non-emergency cash grants to 550 poor and near poor households in An Loc commune, a poor rice-growing community on the Central coast of Viet Nam. Not only is this one-off cash transfer (aka ‘just give them the money’) pilot project unique to Vietnam but it is also distinctive given its establishment outside of an emergency situation and with only minimum conditions (no alcohol, gambling or drugs) placed upon how people spent the money. (I’ve previously blogged on cash transfers in emergencies here.) Read More …

What happens when you give people money (rather than food or blankets) after a natural disaster? Some evidence from Zambia

When disaster strikes in the shape of floods or droughts, aid agencies traditionally ship in food and blankets, often over great distances. But increasingly, people are trying out a novel alternative – give people envelopes full of cash and let them buy what they need. I’ve just been reading an evaluation of two such exercises in response to floods in Western Zambia in 2007, by Oxfam GB and Concern Worldwide. Concern distributed money to 1,700 households in the immediate aftermath of the floods, while Oxfam waited a couple of months and then helped 2,100 households to rebuild. What happened? Read More …