Speakers:
Stephen Devereux - Research Fellow, Vulnerability and Poverty Reduction, IDS
Rachel Slater - Research Fellow, Protected Livelihoods and Agricultural Growth (PLAG) Programme, ODI

Chair:

Tim Waites -Livelihoods Adviser, Renewable Natural Resources and Agriculture Team, Policy Division, DFID

1. Steve Wiggins introduced the meeting as the second meeting in the 'From broad themes to practical policy' series being co-hosted by ODI and FAC this autumn. In particular this meeting looks at social protection: developing agriculture while protecting people.

Stephen Devereux
2. Stephen Devereux stated that he would be talking about the link between social protection and seasonality, in particular he noted that he would talk about seasonality as conceptualised around the entitlements framework, to look at the multidimensional aspects of seasonality, and to talk about seasonality as a crisis.

3. Stephen presented the entitlements framework to show seasonality as a sequence of food entitlements failures:

o Production based entitlements: failure of production self-sufficiency = food gap and hungry season
o Labour based entitlements: excess supply and low demand for labour = insufficient employment and low wages
o Trade based entitlements: Rising food prices and falling asset prices = hunger and irreversible impoverishment
o Transfer based entitlements: shrinking capacity of "moral economy" = inadequate informal protection and hunger.

4. Using information about coping strategies across different countries and time periods, Stephen showed that the most common coping strategies are to cut back on short term consumption and selling cash crops or livestock. Last resorts for coping are the most damaging and the least reversible such as migration. The poor adopt more strategies than rich households. Using these coping strategies is damaging because households are doing things they shouldn't be, such as selling assets and getting less for what they buy them for. In this way people are losing more than half the value of their assets and people are surviving at the cost of their future livelihoods.

5. Stephen then referred back to the framework to look at what types of programmes had been implemented in the past to address seasonal food entitlement failures:

o Productive-based entitlements: input subsidies to starter packs in Malawi (subsidies were then abolished under Structural Adjustment)
o Labour-based entitlements: public works programmes, e.g. Employment Guarantee Scheme (EGS) in Ethiopia (provided public works, ad hoc, not effective safety net); EGS Maharashtra (guaranteed employment under entitlement)
o Trade-based entitlements: Pan-seasonal pricing; grain reserves
o Transfer-based entitlements: formal transfers such as food price subsidies or food aid.

6. In comparison, types of programmes being implemented now are:

o Productive based entitlements: input subsidies are back! Rather than handouts
o Labour based entitlements: public works still with us
o Trade based entitlements: no revival of food price subsidies? No nostalgia for marketing parastatals? (which are currently ignored in current SP agenda)
o Transfers based entitlements: there has been a shift from food aid to cash transfer in last 5-8 years.

7. Stephen then looked at some of the claims as to how cash transfers will eradicate seasonal hunger, questioning some of them:

o Productive-based entitlements: the investment of cash transfers in agriculture will eliminate the hungry season food gap (however, such investment may be due to over funding or mis-targeting), we need to be careful about claims that cash transfers can promote investment in agriculture
o Labour-based entitlements: cash transfers will reduce excess supply of labour and force up rural wages - there is some evidence for India and Malawi
o Trade-based entitlements: cash will attract traders, markets will integrate and food prices will stabilise. However, why do some people in Ethiopia's Productive Safety Net Programme prefer food to cash transfers? More than 1 million prefer receiving food than cash. One reason for this is the increase in food prices.
o Transfer-based entitlements: informal redistribution of cash will extend social protection to everyone. However, redistribution within families or the community is more likely to occur with food than it is with cash.

8. However, Stephen added that he was not totally sceptical to the claims of cash transfers. In fact:

o Cash transfers do protect food consumption
o Cash transfers do reduce seasonal malnutrition
o Cash transfers do reduce the number of coping strategies adopted
o Cash transfers do finance investment in agriculture

9. One donor-funded programme increased the amount transferred to beneficiaries as food prices rose. But is this effective? Or does is in fact fuel food price inflation?

10. In conclusion, Stephen noted the advantages of cash transfers, but emphasised not to neglect investment in agriculture and other sectors; not to expect too much from a single instrument; and finally that social compensation should be a last resort, after addressing production, labour and trade based entitlements to food.

Rachel Slater
11. Rachel Slater started by saying she would present on the instruments for linking growth and social protection. She said that she would pick out what we're learning about instruments and their impacts on poverty reduction.

12. She started by questioning whether social protection and agriculture are alternative or complementary policies? Currently different institutions are responsible for these policies and there is a trade off between expenditure onsocial protection and agriculture: more for social protection or more for agriculture? This is a critical question for policy makers. She argued that poor conceptualisation of social protection and agriculture can lead to suboptimal growth.

13. She went on to say that engaging in agriculture is risky, for example with high levels of climate variation, HIV and AIDS and volatile inputs markets. As our understanding of risk and vulnerability improves, more innovative social protection approaches to dealing with risk and vulnerability emerge beyond smoothing household consumption to also promote people's livelihoods. However, social protection is still largely seen to fall in the domestic domain, and agriculture in the productive. Shocks and stresses differ between domestic and productive spheres, but one affects the other. How can social protection and agriculture complement one another?

14. Some initial findings include:

o The Productive Safety Net in Ethiopia: 80% of the participants are required to complete public works. There are three key things to note, i) small transfers can be used for productive investment only if transfers are timely and predictable; ii) the amount of work requirement is critical, it shouldn't take time away from people's own livelihood; iii) other investments are needed if cash is going to work, for example, investments in markets to ensure that food is available in markets to buy

o Employment Guarantee Scheme in India: rural households are entitled to 100 days of employment per household per year. It is self targeted by setting the wage rate below the prevailing wage rate, but what implications does this have? Evidence from the Maharashtra Employment Guarantee Scheme has shown that setting the wage rate below the prevailing wage rate has raised the agricultural wage. The guaranteed work offers insurance against problems of seasonality by stabilising work opportunities in the "off" agricultural season. However there is a question over whether the programme contributes to productivity through building assets which are actually of value to the poor households in the community, and furthermore, whether employment schemes can bring about rural transformation. It is focused on local employment insurance but not broader than that - does it trap rural workers?

o Some initial findings from CGE modelling in Cambodia looking at the impacts of transferring cash transfers, inputs subsidies and food aid on prices and shifts in demand and supply initially show that targeting the very poorest households in Cambodia is not the best mechanism for reducing poverty or promoting growth. The implication of this is that the impact will be bigger if resources are put into economically active households.

15. Some of the emerging lessons from these experiences point to the importance of:

o Timing and predictability of transfers
o Scale of the transfer
o The need for additional investment and access to complementary services
o The scope of social protection, where social protection overlaps with social programming and social policy, and also with economic policy and programming, social protection needs to be covered by a range of ministries. Furthermore, at the moment social protection is built on the back of disaster, but we need to flip this on its head, to expand horizontally and vertically if a disaster occurs
o There are different poverty groups
o We need to move away from instrument-focused social protection interventions to problem-focused approaches

16. Rachel concluded by laying out four outstanding questions:

1. How do we understand graduation beyond anecdotal evidence?
2. What are the real impacts of social protection on growth?
3. How can we answer a Ministry of Finance's question: how much growth will we get from a US$X million investment in social protection?
4. How can we understand more about politics and the political economy of social protection?

Tim Waites
17. Tim Waites, DFID Livelihoods Advisor, commented on the issues raised in the presentations. He said that picking up on what Rachel said about policy-makers wanting growth or protection is a key discussion being held in DFID at the moment. The new ministers have a renewed focus on growth. The agricultural policy paper focused on agriculture and growth, but it met with criticism from NGOs who questioned a strategy which didn't focus on smallholder farmers.

18. Tim went on to discuss target groups and instruments. He asked how much growth can you expect to get out of the bottom 10%? He reiterated that cash transfers do have a role but emphasised that addressing hunger and poverty is complicated and will not be fixed by one instrument.

20. He also emphasised that predictability and timeliness in implementing social protection programmes are key principles which need to be in place for farmers to plan around their risk.

21. Tim also brought up the issue of climate change: how do social protection mechanisms play out in an environment which is getting riskier? Do we have enough instruments in our tool box? Should we be thinking of other instruments and reacting faster? As we improve our understanding and prediction of climate change shocks can we do risk financing?

Discussion
22. One participant stated that in Cambodia (where the CGE modelling is taking place) one of the most important factors on the impact of cash transfers to poor households is regional differences, and the factor which has the biggest impact on growth and poverty is trade.

25. Another participant asked where gender analysis is when talking about livelihoods? She also wanted to know what had happened to the work on the transformative social protection agenda.

26. Stephen replied that gender issues are indeed crucial to the design of social protection. There are clear differences in preferences between women and men e.g. whether they receive food or cash. The transformative social protection framework recognises that cash transfers can't solve everything and there is a need for legislative changes to change structural inequalities. These should be brought on to the agenda. Rachel added that women's live in particular cross the dichotomy between production and domestic areas, but this dichotomy needs to be broken down to capture the resource flows between the two.

27. Another participant asked how the design of social protection programmes relates to cultural and social issues.

28. Rachel responded by saying that there is sometimes a trade-off between the value of local knowledge and acceptability versus what technical experts say you should do in the design of a programme. Stephen mentioned that targeting can create social unrest, e.g. when targeting by poverty, but if categories are used for targeting, they are much less controversial, but can create inclusion and exclusion errors.