Ethiopia’s Productive Safety Net Program: Chasing two rabbits, catching neither

Economic Commentary

By J. Bonsa, (PhD), special to Addis Standard

Prolonged humanitarian assistance and recurrent emergency relief have created a dependency syndrome in Ethiopia. For over four decades, Ethiopian history has been inextricably linked with famine and disaster. The country is often characterized as dependent on foreigners; and although successive Ethiopian governments have actively engaged in disaster risk management, disasters remain at the heart of Ethiopian politics.

In what sounded a very bold move to do away with dependency culture, the Ethiopian government launched the Productive Safety Net Program (PSNP) in 2005. This program has two key elements.  First, Ethiopia’s humanitarian aid has traditionally been channeled to beneficiaries through a large number of independently operating bilateral and multilateral agencies. However, the financing of the PSNP was designed to create a pool of funds coming from disparate sources and then channel them to target beneficiaries in a coordinated manner. Secondly, the expressions “productive” and “safety net” imply that the program will finance not only “consumption” but also “asset creation” and hence ultimately enhance Ethiopia’s food security.

In other words, the PSNP is designed to accomplish one additional objective, asset creation, with the same amount of resource that would have been committed to finance only final consumption of target beneficiaries.

This piece sets out to seek answers to two interrelated questions: to what extent has the PSNP contributed to reduction in Ethiopia’s dependency on humanitarian assistance? And has PSNP really achieved its two-pronged objectives? But first I will provide a brief program description and then proceed to its assessment.

 PSNP synopsis

The PSNP is an initiative by the Government of Ethiopia (GoE) to change from ad hoc emergency relief oriented toward welfare to a production inducing development activity.  The government designed PSNP in partnership with multilateral organizations, bilateral donors and non-governmental organizations actively engaged in humanitarian assistance.  The program did not replace emergency relief, but it supplemented the existing National Food Security Program (FSP).

PSNP has three specific objectives. The first one is smoothening household consumption by bridging the gap between production and consumption among chronically food insecure farming household.  The second is protecting household assets, by reducing asset depletion and hence vulnerability of poor households during droughts. The third objective is creating community assets by engaging participants in productivity-enhancing public works.  These are reinforced with two principles:  predictability of flows of funds and avoiding dependency by linking transfers to labor services in public works.  The program provides Direct Support to households who cannot provide “labor services”, mainly the elderly and disabled.

Chronically food insecure households have been identified in famine-prone areas of rural Ethiopia. The program started covering 192 districts and 4.5 million beneficiaries.  The size of the program expanded over the years both in geographic coverage and number of beneficiaries – 318 districts and 8.3 million beneficiaries in 2015.

Long hands

Be that as it may, a cursory look through PSNP’s organogram would prompt one to seriously question whether or not Ethiopia is really a federally constituted country. One expects regional authorities to spearhead programs like PSNP, because they are best placed to more quickly observe looming droughts and more effectively manage emergency relief operation.

However, Ethiopia’s federal government designed the governance of PSNP in such a way that its long hands would do everything, leaving no room for regional authorities in decision making. The Ministry of Agriculture is assigned the overall responsibility in planning and implementation; the Disaster Risk Management and Food Security Sector is tasked with overall program coordination; the Food Security Coordination Directorate facilitates the day-to-day management and coordination of the PSNP; the Natural Resource Management Directorate provides coordination and oversight of the public works; and the Ministry of Finance and Economic Development oversees financial management, including disbursement of funds according to plans supplied by the MOA.

These federal implementation arrangements are replicated within the eight regions and 318 districts covered by the PSNP. Regional and district bodies are also responsible for multi-sectoral coordination of the public works. One would normally expect that the overall planning and coordination in each region would be assigned to regional governments and then the sectoral ministries would be tasked to provide technical supports. This reality is put upside down – it is the regional states that receive orders from sectoral ministries.

Food insecurity worsened

How does the program logic that PSNP would contribute to food production and hence food security would apply in practice? The underlying conflicts in PSNP design can become clear if we look into ways in which the PSNP has actually caused food insecurity to move from bad to worse.

The public works are designed to contribute to environmental improvements, changing the ecology of areas that have suffered recurrent droughts and hence famine over several decades. However, the program designers have grossly underestimated the length of time required for investments in public works to yield fruit. The ultimate benefit from public works comes with long time lags. Environmental regeneration may gradually lead to some change in the microclimate to induce rain so that sustainable local food production would become a reality. It is likely that public works would not induce local food production in any immediate future.

The conflicts in food security objectives over different time horizons were the most fundamental flaws in the PSNP program design. PSNP probably enhances food security in the very long run, but it certainly reduces food availability in the short to medium run. This becomes clearer if we examine the adverse effects of PSNP on food security through re-allocation of agricultural labor. One does not need to be an agronomist to appreciate how any diversion of labor from farming activities can reduce agricultural output.

It is one thing to create employment opportunities for people who remain idle during disaster periods, when farmers could not undertake farming activities because of lack of rain.  However, it is an entirely different matter altogether to engage millions of farmers, whose job is food production, with public works even during seasons with favorable weather. Confusing these two conditions seem to have proven a fatal mistake.

It is claimed that attempts have been made to minimize public work during sowing seasons, July to September, by restricting activities to the post-harvest seasons, January to May. However, this only displays astonishing extents of ignorance among program designers about Ethiopian agriculture.  Farmers do not just walk to the fields to sow seeds.  Land preparation and threshing the soil takes place months before sowing season arrives.  Besides, the months immediately after sowing are busy seasons for weeding crop fields.

Farmers would always experience acute shortage of labor during peak harvest or sowing seasons. It is untenable to argue that participants in public works are vulnerable farmers who may have smaller plots and hence may not contribute much to food production.  The diversion of farm labor can still cause damage even in this case because it is common in rural areas for the relatively better off farmers to employ local laborers. Community level food production critically depends on a pool of labor locally available.

Here we are not talking about a marginal diversion of labor away from food production.  The average number of working days that PSNP participating households have annually spent on public works ranged between 104 in Oromia to 155 in Tigray. Critically, even during sowing season, on average participation in public works were 21 days in 2007 and 31 in 2009.

It is not straightforward to establish the exact number of working days, particularly given the regional variations due to religious and cultural holidays.  Elsewhere in the world, the total number of working days in a year is around 220.  If we go by this conservative estimate, then it means PSNP beneficiaries have spent 47% to 70% of their labor time on public works.  Since the number of program participants is given in several millions, exorbitant agricultural person-days have been diverted from food production every year since the launch of the PSNP. No wonder then the country’s food security has worsened.

A further complication in food security issues lie in possible inflationary effects of PSNP.  This is related to disturbances to food prices due to induced imbalances between food supply and demand. PSNP inevitably reduced food production. Cash transfers to beneficiary households inevitably turning food sellers to food purchasers.  The end result is food inflation. It defies belief that such obvious facts escaped from the attention of the donors who keep pouring billions of dollars on this program.

Welfare deteriorated

The analysis so far has been confined to country level outcomes of the PSNP program.  Now we turn our attention to examining how “beneficial” the program has been to the “beneficiaries”.  It is useful to look into extents of unfairness in the payments for their labor contribution.  For participants in public works, daily wage rate started with birr 6 in 2005 and remained stagnant at that level until 2008, when it increased to birr 8, and finally rising to birr 10 in 2009.

Stagnant nominal wage rate means rapidly declining real wage, given the alarmingly increasing inflation over the years. Given program fund comes in hard currency terms, it is legitimate to anchor wage payments to variations in the value of the local currency with the USD.  Between 2005 and 2016, the exchange rate of the birr against the USD declined by 163%, but the wage payment barely moved from its original position.

The above analysis is based on data obtained from program documents and then used by independent program evaluators. However, if we use survey data and calculate average person-days with total payments for specific years, then we get a different figure.  For instance, a survey conducted by IFPRI indicated that daily wage rates varied significantly across the regions, ranging between birr four and seven in 2007 and four and nine in 2009.

The miserably low wage rate paid under PSNP is perhaps a tenth of the prevailing labor market rate in rural Ethiopia. There is an opportunity cost to households supplying labor to the government for public works.  This comes not just in terms of missed opportunities to engage in their normal duty as farmers, but also the alternative of supplying their labor to the private sector at the prevailing wage rate.  Program participants are highly likely to have experienced incalculable welfare losses.

Program evaluation dodged

There is a very strong rationale to subject the implementation of PSNP to some scrutiny through independent evaluation.  It is not enough to report the number of people who have had “access” to the program. What would need to be evaluated are: “concrete and measureable” changes in the livelihood of beneficiary households, particularly change in their behavior in terms of seeking humanitarian assistance; and changes in environmental outcomes, visible effects of the program on the natural environments of the districts covered by this program.  PSNP is meant to detach Ethiopia from dependency syndrome. The program has been in place for over a decade and it is time to show concrete outcomes indicating that Ethiopia is becoming less and less dependent on humanitarian aid.

However, there has been fundamental flaws in the way PSNP monitoring and evaluation has been approached.  Not only is the responsibility of administering and governing PSNP lies in the hands of the Ethiopian government, but also, bizarrely enough, the donor community has effectively left the responsibility of monitoring and evaluation to the Ethiopian government. A Program appraisal document by the World Bank to justify PSNP 4 (2015-2020) stated that:  “… impact evaluation surveys will be done using single source selection.  The Central Statistics Agency (CSA) is the only agency qualified and capable to undertake such big national surveys in Ethiopia. … [T]he development of the survey methodology is likely to require raw data from several prior CSA surveys, both for selecting sample Kebeles and for calculating household sample size. The types of data needed to design the survey are made available only for use in surveys carried out by CSA…. CSA shall be engaged to carry out the baseline and impact evaluation surveys under PSNP 4.”

CSA is a government agency and its staff would have a vested interest in collecting and organizing information to show “success”.  CSA would collect data from civil servants working with branches of sectoral ministries at different levels in PSNP governance structure.  The assessment of staff promotions would strictly depend on their ability to successfully implement programs like PSNP.  Therefore, regardless of the extent of sophistication in survey design, the content of information collected by CSA is bound to be biased toward success.  What matters is who has the power to influence the way data is collected and organized, not who would analyze the data and write reports. PSNP needs genuinely independent and impartial program evaluators who would handle the process of information collection, organization, analysis, and interpretation.

Dependency syndrome aggravated

In the absence of credible monitoring and evaluation mechanism, we resort to using deductive methods to infer the extent to which PSNP has led to reduction in Ethiopia’s dependency on humanitarian aid. This can be assessed in a very straightforward manner – by examining key indicators such as changes in humanitarian aid metrics –  year-on-year changes in the number of people dependent on humanitarian aid, geographical coverage, and sizes of funds.

PSNP was launched in 2005 with USD $203 million budget, supporting 4.5 million target beneficiaries in 192 districts. The program has steadily expanded over the years with the number of beneficiaries reaching 7.6 million in 2012 and 8.3 million in 2015. Similarly, annual program budget averaged around $310 (2005-2009) and $460 (2010-2014).  The corresponding figures for PSNP 4 (2015-2020) is planned to be more than double the amounts at the launch of PSNP – number of beneficiaries further rising to 10 million and budget moving up to a whopping $720 per annum.



Source:2005-2011:Venton, Shitarek, Coulter, and Dooley, 2013, 2015/reliefweb.

Has emergency aid declined since the launch of PSNP?  Fig. 1 indicates that there is no sign of ad hoc emergency aid falling. If anything the situation has gotten worse over the years.  In 2005, the number of people relying on humanitarian assistance was about 8 million (PSNP and relief were 4.5 and 3.5 respectively).  These figures rose to 8.3 million and 10.2 million respectively in 2015.  In other words, the number of PSNP dependency has doubled while that of relief dependency tripled.

In a PSNP assessment study published in 2013, Desalegn Rahmato of Forum for Social Studies compared year-on-year changes in emergency fund requests and actual donations between 2005 and 2011. The receipts were on average only 1% of the total funds it requested.  This means the authorities have displayed an insatiable appetite for aid money. This totally contradicts the spirit with which the PSNP was launched.

Back to square one

At PSNP design stage there were intense debates and significant disagreements among the donor community on the basic principles of the Safety Net Program. It was reported that “the most prevalent debate centered on whether the safety net should be primarily protective or productive.  Some donors opposed the idea of a productive safety net that forced poor people to labor on public works projects, while the Government opposed unconditional social transfers out of concern that this would create disincentives for households to improve their livelihoods and ultimately foster dependency on external assistance,” according to a WB document.

The Ethiopian government seems to have forced the donor communities to give in.   There was a hidden political ploy behind a façade of Ethiopian government’s “steadfast commitment” to shift from “protective” to “productive” humanitarian aid.  Desalegn Rahmato concurs with Human Rights Watch (2010) report that the Ethiopian government used PSNP to reward its supporters or punish those who rally behind the opposition. It is highly likely that the Ethiopian government’s motivation to initiate the PSNP was to use it as a political weapon.  It would prove useful to remember the timing of the PSNP launch – 2005, a watershed year in Ethiopia’s political history during the EPRDF era.

At the time, donors resolved to stop providing development aid to the Ethiopian government due to the debacles in the aftermath of the general election.  However, donors declared they would continue to provide emergency relief aid to save lives when disasters occur. It should be recalled that the ominous prospect of development aid drying up infuriated the EPRDF regime, who suddenly shifted to an aggressive mode. This was expressed in terms of banning local NGOs from using donor funds and restricting the operations of foreign NGOs, that is to say a pre-emptive strike that made “protective” aid unworkable.

PSNP was simply a variant but scaled up version of such actions, with which the Ethiopian authorities intimidated multilateral and bilateral donors to allow “productive” or development aid by other means. Clearly, donors were caught between a rock and a hard place. It was clear that the Ethiopian authorities would not budge on their position, even if that would mean lives would perish. Consequently, donors had to yield to the demands of the Ethiopian government, effectively handing over aid money and letting the authorities use it more or less unconditionally. That is how PSNP funds have begun to be used as a source of financing wages and salaries of Ethiopia’s bloating civil service. The double-pronged strategy to simultaneously achieve “protective” and “productive” goals using PSNP funds was nothing more than a window dressing.


ED’s Note: J. Bonsa is an economist by training. He can be reached at

Show More

Related Articles

Back to top button