The food price crisis and its impact on the Ethiopian Productive Safety Net Programme in 2008

The Ethiopian government’s Productive Safety Net Programme (PSNP) is the biggest social protection instrument in Africa and one of the biggest programmes of its type in the world, providing cash or food transfers to nearly 7.2 million people in 2007. Ethiopia has seen unprecedented inflation in 2008 coinciding with a period of global food price growth. This, together with rain failure, is having a significant impact on people’s ability to meet their basic food needs. This article looks at how the PSNP has responded to these challenges.

The PSNP

The objectives of the PSNP are to ‘provide transfers to the food insecure population in chronically food insecure woredas in a way that prevents asset depletion at the household level and creates assets at the community level’. The full programme, which began in 2005, operates in four of Ethiopia’s Regions, with pilots currently implemented in two others. The vast majority of PSNP resources (approximately 85%) are provided as payment for participation in public works, with the remaining support provided to households without access to adult labour. Households in receipt of PSNP benefits (cash, food or a combination) receive transfers from January to June each year. These are typically the ‘hunger’ months in Highland areas. As a result, most of the tables and figures in this article refer to the months January–June.

Food price rises in Ethiopia

Food prices across Ethiopia have increased significantly since mid-2007. Figure 1 shows the absolute rise in food prices in four markets between January 2007 and June 2008.

In Highland areas, normal seasonal variations mean that food prices on the market are typically significantly higher in July/August than they are in January/February. Between 2005 and 2007, the PSNP wage rate (6 Ethiopian birr or $0.42) enabled a household in Amhara Region to buy about 3.6kg of maize. However, seasonal variations each July between 2005 and 2007 meant that households were only able to buy 2.8kg of maize with 6 birr. This constituted a 21% loss in purchasing power in a ‘normal’ year. Between early 2006 and mid-2007, the price of maize in Gabalafto market in Amhara remained fairly stable at just below 2 birr per kilo.

However, between July 2007 and February 2008 the price of maize rose to 3 birr/kg. In a ‘normal’ year before 2007, households would expect to see the price of maize decreasing at this time, not increasing. The Ethiopian government responded by increasing the PSNP wage rate from 6 to 8 birr ($0.56) per day in early 2008. However, maize prices on local markets continued to rise, to the point where, in February 2008, 8 birr purchased just 2.77kg of cereal. Six months later, at the end of the 2008 PSNP transfers in July, the 8 birr PSNP wage rate only enabled a household to purchase 1.2kg of cereal. As Figure 2 shows, this represents a loss in purchasing power of 56% for the poorest and most food-insecure households in Amhara Region, in the space of seven months. Between April and July 2008, the reduction in purchasing power of PSNP wages (and therefore the effective value of the wage rate) continued at nearly 20% a month.

Purchasing power among PSNP households varies significantly between Regions. Figure 3 shows the amount of maize that could be purchased with the PSNP wage rate of 6 birr per day in four Regions between 2005 and 2007. Taken as an average, purchasing power in Tigray was significantly lower than among comparable households in Southern Nations, Nationalities and Peoples Region (SNNPR).

Purchasing power within regions varies too, as Figure 4 demonstrates. This shows that poor households living close to the cheapest 5% of markets in Tigray have better purchasing power than those living near to the most expensive 5% of markets in SNNPR. Therefore, chronically food-insecure households that live close to cheaper markets in Tigray are better off than their counterparts living near expensive markets in SNNPR (despite the average Regional differences). From this it follows that one national size of PSNP transfer does not meet all the requirements of poor households in different Regions. The Ethiopian government has, however, repeatedly stated that varying wage rates between Regions is politically unacceptable.

How has the PSNP responded to food price rises?

In different parts of the PSNP coverage area, some food-insecure households receive food, while others receive cash in exchange for their labour. Maintaining parity between the cash and food components of the PSNP has proved an ongoing challenge, particularly since the end of 2007. At present, three wage rates are operational in the PSNP, depending on geographical location and the implementing agency involved:

• A cash wage rate, currently 8 Ethiopian birr ($0.75) per day or 40 Ethiopian birr ($4.20) per person per month.
• A full food basket, based on the Sphere standards, which state that those receiving emergency assistance should get the equivalent of 2,100 kcals per day. In Ethiopia this is calculated as 15kg of cereal, 1.5kg of pulses and 0.5 litres of oil.
• 15kg of cereal. The government’s Programme Implementation Manual (PIM), which guides the implementation of the PSNP, states that payments in kind will be ‘3kg of grain per day plus pulses and oil’. However, the government views pulses and oil as a bonus which will only be provided when available.

In 2005, there was parity between the first two of these ‘wage rates’. Indeed, one of the aims of the PSNP design was to investigate which type of transfer had the most impact on smoothing consumption, protecting assets and stimulating markets. However, food price and availability fluctuations since 2006 have resulted in considerable changes in the comparative value of the three different wage rates, distorting the impact of the PSNP.

The relative values in Ethiopian birr of each of the three wage rates since 2006 are presented in Figure 5. Also indicated in Figure 5 – in the shaded area – is Save the Children UK’s calculation of the Minimum Cost of a Healthy Diet for parts of Amhara Region. This new methodology uses linear programming to generate the financial value of a diet that meets 100% of nutritional requirements, based on local availability. In other words, this is how much it costs a household in this area to guarantee against malnutrition. The cost of a healthy diet in 2006 in Amhara Region was approximately 66 birr per month. In Figure 5, this has been increased on an annual basis in line with the rate of food price inflation.

Figure 5 shows how the financial value of the different wage rates in the PSNP has contributed to meeting the nutritional requirements of household members since 2006 in Gubalafto woreda. Although all three wage rates have responded differently, cash has failed to maintain the same contribution to households’ nutrition as a full food basket. This goes some way to explaining the likely outcomes of escalating food prices on the 57% or so of poor food-insecure households that receive their PSNP wage rates in cash.

As part of the PSNP design, there is a contingency fund which can be tapped by Regional governments in the event of a crisis. The value of the fund represents 20% of the value of PSNP transfers in the Region. Unfortunately, due to the severity of the crisis and the lower than normal harvest after poor rains, the contingency fund in Amhara had been exhausted by June 2008. The funds were used to support those households suffering from food shortages but who were not enrolled in the PSNP.

This made it more difficult for the PSNP to respond to its target group as food deficits emerged. From June 2008 onwards, everyone in need was reliant on humanitarian assistance alone.

The anticipated impact of food price rises on the PSNP in 2009

From the end of the scheduled PSNP transfer period in July 2008, food prices in rural markets appear to have stabilised, albeit at (or near) their peak. In Gubalafto and Sekota markets, Amhara Region, the price of maize peaked at 6.60 and 7.15 birr respectively in July (usually the seasonal peak price for staple foods). In November 2008, prices were 5.45 and 5.50 birr respectively. Despite the appearance of cheaper maize on the markets, this still represents just under a 300% increase on prices in the period 2005–2007 (when they were stable at around 2 birr per kilo). Thus, in Gubalafto market purchasing power had begun to increase for PSNP households – but only marginally. Figure 6 plots market prices against purchasing power for PSNP households in Gubalafto woreda.

Implications for the PSNP

The PSNP is designed primarily as a cash-based safety net to protect chronically food-insecure households and ensure that they meet their food requirements without having to deplete household assets. However, the cash provided through the PSNP is inadequate to enable recipients to purchase foods similar to those provided in a food basket. Perhaps unsurprisingly, the vast majority of PSNP participants consulted in a Save the Children UK study in 2008 indicated that they would prefer to receive food, rather than cash, under the PSNP. PSNP participants indicated that they distrusted cash because they had seen its value deteriorate so rapidly; food was more tangible and reliable, and did not lose its value.

Using data from Save the Children UK final evaluations of prior programmes, the government’s Medium Term Expenditure and Financing Framework for the PSNP and market prices, Figure 7 shows the differences in programme cost between imported food, locally procured food and cash transfers between 2002 and 2008. In July 2008, it was (abnormally) more cost-effective for the PSNP donors to transfer food rather than cash.

Conclusions

Events in 2008 provided the PSNP with its first real test. For the first time since the programme’s inception, PSNP target households faced a large-scale, significant shock. Many of these households’ usual means of response had already been exhausted by the time the extent of the crisis was fully appreciated. Between 2005 and 2008, the cash portion of the PSNP largely succeeded in meeting the objectives of the programme: asset accumulation and smoothing of food consumption. However, this was largely not the case during 2008.

An underlying factor in the problematic response has been a delayed understanding of the real impact of inflation on the rural poor, due to inappropriate analysis and delayed responses to manage risk. Because chronically food-insecure and poor households typically purchase the majority of their food on the market – the poorer households are, the more they rely on the market – those who received cash under the PSNP lost as much as 56% of their purchasing power as a result of the food price rises during the period February–July 2008. Although absolute food prices are now slowly declining, if the PSNP had restarted transfers in November 2008 the purchasing power of chronically food-insecure and poor households would still be around 45% below its level a year earlier.

Several implications flow from this, both for the government and for donors funding the PSNP. First, the system of three different wage rates needs to be rationalised and parity between them achieved. This will entail significantly increasing the value of the cash wage rate.

Second, as demand is outstripping supply the PSNP must grow in scale and scope, to allow more chronically food-insecure households to enter the programme. While growing in scale and scope, an expanded PSNP must also ensure that it is tailored to the livelihoods of those it is providing assistance to. Replicating PSNP ‘blue-prints’ for other livelihoods systems are unlikely to be appropriate: a PSNP pilot is under way in some lowland, pastoral areas and a tailored safety net instrument that relies on the strengths of pastoral livelihoods should be explored with relevant stakeholders in more detail.

Third, steps must be taken to ensure that the type and quantity of resources put into the PSNP are flexible enough to enable the programme to meet its objectives. If a cash-based instrument is still favoured by the government, then flexibility in the delivery and quantity of cash could address concerns about poor households being unable to purchase sufficient food with PSNP transfers. Finally, risk management strategies have to be reviewed so that future crises can be identified and addressed in a timely fashion. The early-warning system currently operational in Ethiopia’s lowland, pastoral areas must be extended to cover the whole country. The fundamental function of such a system would be to trigger preventativeresponses to losses of the productive assets the PSNP is meant to help households accumulate. This requires pre-positioning resources and the completion of the administrative and legal necessities during times of ‘normality’. Preparedness ‘audits’ should be completed for each agency and administration, and contingency plans at district level should be prepared, with locally agreed ‘triggers’ for response.

For donors, the priority must be making good on commitments in 2005 to fund a PSNP that delivers timely, predictable and appropriatetransfers (a 1 birr increase in the PSNP cash wage rate means an additional $25 million). Donors must also recommit to supporting the PSNP, rather than reverting to the old, untimely system of annual needs identification, humanitarian appeal, drip-fed funding and the eventual (usually late) delivery of humanitarian assistance. Lastly, donors should make extra resources available to fund an effective early-warning system that enables communities to manage the risks they face.

Matt Hobson is Head of Hunger Reduction at Save the Children UK. His email address is Matt.H@scuk.org.et. This article is based on a Save the Children UK report published in September 2008, entitled Cash, Food, Payments and Risk in the PSNP, by Judith Sandford. Interested readers should contact Save the Children UK for a copy. Matt Hobson would like to express his thanks and appreciation to Judith Sandford, without whose work this article would not have been possible.