Recent African experience with input subsidies
Recent African experience with input subsidies
Abstract and Keywords
Resurgent interest in agricultural input subsidies has led a number of countries in Africa to implement agricultural input subsidy programmes focusing particularly on fertilizers. This chapter reviews available information on recent agricultural input subsidy programmes in Ghana, Zambia, Nigeria, Tanzania, Rwanda, Mali, and Senegal. These programmes are of different scales and pursue different, sometimes, multiple objectives, with the private sector playing different roles, and have varied implementation and programme outcomes. The chapter notes that while there are a number of objectives in different programmes, most focus on production objectives and producer welfare and ignore the interests of consumers and the processes by which subsidy programmes contribute to wider pro-poor economic growth. The review also shows that graduation from subsidies has not been an explicit design feature of these programmes.
Having considered key features of input subsidy programmes and a conceptual framework of their impacts in the previous chapter, we now turn to examine experience with the implementation of recent, large-scale initiatives in poorer agrarian economies. This leads to an emphasis on programmes in Africa. As noted in Chapter 2, there has been resurgent interest in input subsidies, in particular ‘smart subsidies’ for fertilizers in Africa. Unfortunately there are few detailed and rigorous evaluations of most of these programmes. In this chapter we review African programmes for which there is new information since an earlier review by Dorward (2009b)—but we explicitly exclude any discussion of the Malawi experience from 2005/6. In subsequent chapters we will, however, where appropriate compare Malawi’s post-2005 experience with observations from this review. This review does not cover all large-scale subsidy programmes in sub-Saharan Africa since 2005, only those for which some information is available. Thus, there is no review on achievements under Kenya’s National Accelerated Agricultural Input Programme (NAAIP), planned for implementation from 2008 (Dorward, 2009b). Similarly, there is very little information available on programmes in Mali and Senegal (Kelly et al., 2011) and these are therefore considered together. A number of authors comment on the lack of systematic and good quality information on subsidy programmes in Africa (Dorward, 2009b; Morris et al., 2009; Kelly et al., 2011; Druilhe and Barreiro-Hurlé, 2012).
We summarize here observations from an examination of nine fertilizer subsidy programmes in Africa selected on the basis of (a) availability of information and (b) relevance to our review of large-scale subsidy programmes aimed at boosting input use in staple crop production. We first briefly describe the main elements of each programme’s history before considering what they (p.47) can teach us regarding general patterns of subsidy programme objectives, design, implementation, and impacts.
In early 2008 Ghana faced high food prices and rising fertilizer prices and the government and large fertilizer importers (who had significant but high priced fertilizer stocks) discussed the potential and possible modalities for a national fertilizer subsidy programme. Press reports (Ghana News Agency, 9 June 2008, 3 July 2008) indicate that the programme was formally announced in June 2008 and operated from July to December. A total of 30,000 tons of four types of fertilizer was made available by three major importers, with pan-territorial farmer prices representing an approximate 50% subsidy, keeping farmer prices similar to the previous years’, in the context of rising international prices, at a total cost of around US$15 million.2 Large numbers of vouchers (over 1 million) were printed against planned subsidy sales of 600,000 bags. Deliveries were late for the cropping season in the south of the country (April to July), but were more timely for the north, and this may account for lower uptake and fertilizer sales in the south and use on a wider range of minor crops as compared with the north where there was more substantial uptake and use mainly on maize.
Vouchers were distributed by Ministry of Agriculture staff for redemption by distributors linked to the major fertilizer importers. There was wide variation in voucher distribution approaches, systems, and numbers across different areas, and limited information to field level staff on the total number of vouchers that they would receive for distribution. Redemption prices varied geographically to provide pan-territorial farmer prices in district capitals, but this tended to discourage suppliers from supplying fertilizers outside district capitals as neither redemption nor farmer prices covered costs of transport outside district capitals. No subsidy sales were made by (smaller) distributors independent of the major fertilizer importers (indeed in the north unsubsidized sales were reported to be banned completely) and the programme may have reduced competition and depressed sales and revenues for smaller retailers closest to farmers in rural areas. It may also have strengthened the position of importers participating in the programme (who gained from increased sales of previously imported stocks) at the expense of others. It was (p.48) also widely believed that sales without vouchers were illegal. Banful (2011) reports that both political and economic efficiency considerations appear to have influenced distribution of vouchers between districts. Receipt of vouchers by farmers is reported to be associated with access to distribution points, reliance on agricultural income, hiring of labour, wealth, and years farming (Vondolia et al., 2012). There is less specific and detailed information about subsequent programmes, which appear to have operated in a similar way. However, the scale of the programme had doubled by 2011—when 150,000 tons were subsidized at a cost of US$70 million and a 50% subsidy rate—and there are continuing concerns over late implementation (Ghana News Agency, 2012; Alassan, 2012).
3.3. Zambia Fertilizer Support Programme and Food Security Pack
We consider two input subsidy programmes in Zambia, the Food Security Pack and the Fertilizer Support Programme, both confusingly known as FSP.
The Food Security Pack as reviewed by Ellis (2007) has a number of differences in design from other programmes reviewed in this chapter, but displays many similarities in its actual operation.3 It was developed and has consistently been considered more as a social transfer programme than an input subsidy programme, aimed at ‘vulnerable but viable’ farming households, but attempting to address a cause of vulnerability: lack of access to productive inputs. Its main objectives have been to provide basic inputs to vulnerable households with some land and labour, but also to promote crop diversification and conservation farming. Its primarily social welfare objectives led to its implementation by a national NGO coordinating a network of district NGOs under the direction of the Ministry of Community Development and Social Services, with technical support from the Ministry of Agriculture and a domestically funded budget line from the Ministry of Finance and National Planning.
The first year of operation of the Food Security Pack was 2000/1, and it was intended to reach 200,000 households per year for three years. It has in fact operated for a much longer period on a much smaller scale as regards the number of beneficiaries served, but has been thinly spread across all 72 districts in Zambia. Budgetary provision, beneficiaries reached, and input packs have varied widely across the years (Ellis, 2007; Kodamaya, 2011), as has the composition of input packs. These were supposed to contain inputs for 0.25 (p.49) ha of cereal (maize) production, for 0.25 ha of cassava or sweet potato production, and for 0.25 ha of legume production. Packs are provided free, but after the first couple of years the programme farmers are expected to repay 50% of the value in kind after harvest with the proceeds partly stored food security and partly sold to finance public works (Jorgensen and Loudjeva, 2005).
There have been particular difficulties with the sourcing of inputs for root crops and legumes, and in 2005/6 for example, only maize inputs were provided. Funding has often been inadequate, unpredictable, and late, and thus reached a smaller than budgeted number of beneficiaries with common late delivery of inputs. The small numbers of beneficiaries in each district have often been concentrated in small geographical areas for logistical reasons. Within these areas, targeting by local committees has been affected by local elite capture and also by widespread splitting and sharing of packs. Beneficiary selection was supposed to use primary targeting criteria focused on land-operating but land-poor households with unemployed labour, together with secondary criteria focused on favouring particular vulnerable groups (for example, households of the elderly, disabled, orphaned or unemployed youth members). However, Jayne et al. (2006) report that in 2002/3 beneficiaries under the Food Security Pack had average per capita incomes of less than half of those of beneficiaries of the Fertilizer Support Programme discussed below. There is, however, very little information available on the programme’s impacts.
Turning now to consider the much larger Fertilizer Support Programme, this was initiated in 2002 as the successor to a long history of fertilizer subsidies and alongside the smaller and socially targeted Food Security Pack discussed above. Since that time the programme has grown and evolved: from 2002/3 to 2009/10 beneficiaries rose from 120,000 to 500,000 farmers, subsidized inputs rose from 48,000 to 180,000 tons of fertilizer and from 2000 tons to nearly 9000 tons of maize seed, and the subsidy rose from 50% to 75% of input costs (Ministry of Agriculture and Cooperatives, 2011). However, the same basic approach has been maintained, with fertilizer imports by private companies under government tender with distribution to farmers through cooperative societies. Our review draws on a number of studies which have examined different aspects of the programme over time.
Minde et al. (2008) report that the objectives of the programme have changed over time with it being intended at first to assist smallholders accessing inputs in remote areas which it was thought were not served by private traders. However, private sector market and service expansion, in these areas particularly, has also been stressed, with reduced government involvement in input supply. Thus, the Ministry of Agriculture and Cooperatives (2011) lists the 2011/12 programme overall objectives as ‘to increase private sector participation in the supply of agricultural inputs to small-scale farmers and (p.50) contribute to increased household food security and income’ with specific objectives to expand private sector input markets and involvement, reduce government involvement, ensure timely and effective supply to targeted farmers, improve these farmers’ access to inputs, promote competitiveness and transparency in input supply and distribution, reduce risks in farmers’ inputs use, and promote rural institutional development. However, the Zambia Agricultural Consultative Forum (2009) proposed that the overall objective of the programme should be ‘to increase small scale farmers’ productivity in order to contribute to improved household and national household food security’ (p. 14) with (as above) specific objectives to improve small-scale farmers’ access to inputs with increased participation of the private sector, an expanded agro-dealer network, and timely, effective, and adequate input supply.
These multiple objectives lead to a number of different criteria by which the programme is judged. As regards private expansion, there are persistent criticisms that the programme has discouraged (or at best had mixed effects on) private trader participation due to the restricted number of companies supplying fertilizers, and unpredictable location, timing, and composition of subsidy supplies leading to uncertainty over commercial demand (for example, Jorgensen and Loudjeva, 2005; World Bank, 2010b). This is associated with concerns that there has been significant leakage of supplies from the subsidy supply chain into the commercial supply chain (Muleba, 2008; Minde et al., 2008; World Bank, 2010b; Mason, 2011) and displacement of unsubsidized sales by subsidized sales to larger farms in less remote areas where the bulk of supplies have gone (Minde et al., 2008; Xu et al., 2009a; Mason, 2011; Jayne et al., 2011).
Displacement and targeting, and widespread reports of late input delivery, also affect the programme’s effectiveness in raising maize production and in improving the food security and livelihoods of poorer farmers, and again this is the subject of concern in a number of studies (Civil Society for Poverty Reduction, 2005; Jorgensen and Loudjeva, 2005; Minde et al., 2008; Zambia Agricultural Consultative Forum, 2009; World Bank, 2010b; Mason and Ricker-Gilbert, 2012). While these issues reduce the effectiveness of the programme in raising maize production, the programme’s impact on increased maize production is recognized, though these increases may be smaller than the effects of good rainfall and smaller than official estimates of the programme’s benefits (Minde et al., 2008; World Bank, 2010b; Mason et al., 2011). There are, however, also concerns that yield responses are limited by soil acidity and other complementary investments are needed for fertilizer subsidies to be effective (Burke et al., 2012).
There are also concerns that the programme’s emphasis on maize has discouraged production of other crops such as sorghum and cassava (Haantuba (p.51) et al., 2011). Coupled with these are wider concerns about poor monitoring and evaluation, expansion of the programme as regards both volume of subsidized inputs and the subsidy rate despite initial intentions to scale both back, annual budget over-runs and consequent growing programme cost—averaging nearly 40% of the Ministry of Agriculture’s budget from 2002 to 2009 (Civil Society for Poverty Reduction, 2005; Jorgensen and Loudjeva, 2005; Zambia Agricultural Consultative Forum, 2009; World Bank, 2010b). Nevertheless, the programme is estimated to yield an economic cost—benefit ratio that is greater than 1 (Jayne et al., 2007) but a large number of studies comment that it is crowding out alternative and higher return investments in longer term research or infrastructural programmes (Jorgensen and Loudjeva, 2005; Jayne et al., 2007; Minde et al., 2008; Bigsten and Tengstam, 2008; Govereh et al., 2009; Xu et al., 2009a).
Nigeria has implemented large-scale fertilizer subsidies since the 1970s with the broad objective of promoting agricultural productivity and, latterly at least, improving food security by making fertilizers more affordable and accessible to smallholder farmers. Subsidies have been marked by wide variation in rates and modalities over time, accompanied by substantial differences between states (Akande et al., 2011). There is, however, general consistency in a very active role of the state in fertilizer delivery and widespread reports of high diversion of fertilizer, such that smallholder farmers receive only 30% of subsidized fertilizers at subsidized prices (Nagy and Edun (2002) cited by Liverpool-Tasie et al. (2010b), Banful and Olayide (2010)), with generally late delivery of subsidized fertilizers and frequent reports of concerns about quality (Nagy and Edun (2002) cited by Liverpool-Tasie et al. (2010b), Kiger and Adodo (2010), Banful and Olayide (2010)). There is, however, very little information on the impacts of subsidies in terms of increased fertilizer use, production, food security, or poverty reduction (Mogues et al., 2008). Banful and Olayide (2010) and Banful et al. (2010) report that farmers and stakeholders consider unavailability of fertilizers, late delivery, and poor quality to be much greater constraints to fertilizer use by farmers than its expense (although access to credit is also cited as a major constraint (Banful and Olayide, 2010)). Liverpool-Tasie et al. (2010b) therefore recommend that much more emphasis should be given to stable policies that encourage the development of private sector suppliers. One approach to this is the use of fertilizer vouchers to deliver subsidies. Small pilots of this approach were championed by IFDC from 2004 (Gregory, 2006) and judged to be successful in demonstrating that a voucher-based system could improve farmer access (p.52) to fertilizers if implemented with the private sector. Progressively larger pilots were implemented, with large programmes in two states in 2009 and 2010 reaching 194,000 and 171,000 farmers respectively with sales of 29,800 and 16,397MT (Kiger and Adodo, 2010). Kiger and Adodo (2010) report that the programme has been very successful in improving farmers’ access to subsidized fertilizers, in reducing leakages, and in demonstrating and building private capacity in fertilizer supply, while recognizing continuing but surmountable challenges facing the programme. There is, however, no information on programme impacts in terms of increased fertilizer use, production, or food security, and a survey by Liverpool-Tasie et al. (2010a) find that farmers still reported late delivery and poor fertilizer quality. Their study also raises questions about beneficiary targeting—criteria seem to be very broad, with voucher distribution to members of farming or other groups, and there was some evidence of multiple receipt within households, but recipients tended to be poorer than non-recipients (Liverpool-Tasie, 2012). In this context, and with over 70% of farmers outside the subsidy programme buying fertilizer in the private market, Liverpool-Tasie (2012) do not find evidence that subsidy receipt depresses the probability of farmers buying unsubsidized fertilizer, and indeed find that for subsidy recipients who are buying unsubsidized fertilizers there is a significant crowding in effect, with receipt of a bag of subsidized fertilizer increasing unsubsidized purchases by 0.8 of a bag. This is attributed to the subsidy programme encouraging the establishment of a better private distribution network. Overall, there seems to be little doubt that the programme is a substantial improvement over the standard programmes being administered by the Federal and State governments.
Like many countries in Africa, Tanzania has implemented a variety of input subsidy programmes over the years. Following the withdrawal of subsidies in the 1990s, fertilizer transport subsidies were introduced in 2003 with the objective of facilitating fertilizer use in remote areas (Minot, 2009). The programme subsidized transport costs and fixed margins and prices for farmers, with government management of stocks and transport. Although this appeared to lead to some increases in fertilizer use, there were problems with the heavy involvement of government in managing distribution, late fertilizer delivery due to budgetary processes, ineffective price controls, and leakage from target beneficiaries (to local unsubsidized markets and also to neighbouring countries), with the latter problems associated with lack of skill and financial capacity among agro-dealers (Minot, 2009; World Bank, 2009). (p.53)
In view of these problems, the government started piloting a ‘smart subsidy’ programme from 2007. This grew into the National Agricultural Input Voucher Scheme (NAIVS), with increasing pilots in 2008 and the development of a very large programme with World Bank support implemented from 2009 to 2012 (World Bank, 2009).
The objectives of the NAIVS are to facilitate fertilizer use in high-potential areas, to offset rising international fertilizer costs, to reduce food prices by stimulating production, and to stimulate expansion and increased capacity in the private input supply system. The main features of the programme, summarized by Minot (2009), include the use of vouchers for food crop inputs (fertilizer and maize and rice seed) and distribution to targeted beneficiaries with complementary support to help them improve the efficiency of input use and to expand input suppliers’ financial and skills capacity. The programme is also intended to first of all scale-up and then scale-down, with a maximum of three years access by each beneficiary and termination in 2012 (although extension of the programme to other regions is under consideration). Vouchers, covering 50% of full input prices, are redeemed by farmers at agro-dealers who then redeem them through the large branch network of the National Microfinance Bank. Vouchers and inputs are targeted to high-potential regions according to the number of maize and rice farmers per region and then targeted to full-time resident maize or rice farmers with less than one ha and the ability to part finance input purchases. Among eligible farmers, priority should be given to female-headed households and those not using improved inputs in the previous five years.
It has not been possible to source evaluation studies of the programme since 2010, but the World Bank (2009) identifies benefits and challenges from the pilot programmes. Major benefits of the targeted subsidy mechanisms over the previous transport subsidy include easier monitoring of input distribution and impacts, wider coverage and improved input affordability for farmers, and mutually beneficial links and strengthened relations with and between the National Microfinance Bank and agro-dealers. There are, however, also challenges identified from the NAIVS 2008/9 pilot where there was a need for more human and financial field resources for implementation and technical support (for the Government and National Microfinance Bank), there were insufficient inputs for all eligible farmers and late delivery of inputs to farmers, there were late bank payments to agro-dealers and late government payments to the bank, and agro-dealers lacked sufficient working capital. Minot (2009) commends the programme design for its targeting approach and use of vouchers, intention to scale-down after 2012, and complementary support to the fertilizer subsidy (in terms of a public awareness campaign, capacity building and certification for agro-dealers, seed sector support, an emphasis on integrated soil fertility management, and investment in monitoring and (p.54) evaluation). However he also questions whether it will be politically feasible to scale the programme down after 2012, and if the targeting will be effective.
Pan and Christiaensen (2012) and Patel (2011) report household survey findings in 2008 and 2009 that provide some insights into the targeting and timing of input delivery. In a sample survey in the Kilimanjaro district, Pan and Christiaensen (2012) report some difficulties with targeting in 2008. The major observation is that there are conflicts between potential targeting objectives, with an emphasis on economic efficiency in increasing production requiring targeting outcomes that favour those households with the highest marginal return to fertilizer use on fertilizer that they would not have used without subsidy receipt. These may not be the same households as those that should be targeted to reduce poverty and inequity. However, examination of actual targeting outcomes shows that elected village officials received about 60% of the distributed vouchers, and this ‘substantially reduces the targeting performance...At the margin, village elected officials are usually not the more efficient input users’ (Pan and Christiaensen, 2012: p. 1627).
Patel (2011) finds that cost was the most commonly cited reason for farmers not using improved inputs (cited by 69% of households not using seeds, and by 36% of households not using fertilizers), but lack of awareness was also a significant impediment. Lack of fertilizer availability was cited as problem by only 6% of farmers. They also find that the targeting criteria were not consistently followed but do not suggest that vouchers are going to the wrong farmers—contrary to Pan and Christiaensen (2012) they find no evidence of elite capture. While beneficiaries tended to be wealthier than non-beneficiaries, they note that the programme is not intended to be a pro-poor programme and targets ‘middle-level farmers’ most able to make good use of the inputs. They recognize, however, that this may lead to problems of displacement of unsubsidized sales. Poorer and female-headed households, which the programme also aims to prioritize, are often unable to finance the purchase cost even of the subsidized inputs. They find that in terms of programme implementation, NAIVS operations generally adhered to the guidelines, although anecdotal reports for subsequent seasons are less optimistic. However survey estimates of beneficiary coverage were much lower than indicated by MAFC’s programme data, which estimate coverage above 80% in some districts. They make initial estimates of incremental production gains from input use of 147% for maize and 35% for paddy, but are not clear that this is high enough for unsubsidized input application to be profitable, raising questions about the economic returns to the programme.
As regards programme implementation, the Ministry of Agriculture Food Security and Cooperatives (2012) provides information on programme (p.55) activities from 2008/9 to 2011/12: 3.5 million households are identified as eligible for voucher receipt, with actual recipients rising from 730,000 households in 2008/9 to 1.5, 2.0, and 1.8 million households respectively in the following three years. However the 2011/12 recipients excludes households who started receiving vouchers in 2008/9 and therefore ‘graduated’ after three years in the programme. This should give a total of some 2.5 million direct beneficiary households over the three years, each receiving three vouchers per year (two for fertilizer and one for maize or rice seed).
Activities designed to support private input supply development and more effective input use by farmers have included agro-dealer training, matching grants to guarantee agro-dealer loans from input supply companies, development of seed systems, and research and (to a lesser extent) extension on Integrated Soil Fertility Management and input use. Although some of these activities have been delayed for various reasons, significant progress has been made on others. Just under 4,000 agro-dealers have been trained and 23 agro-dealer associations formed; 2,335 agro dealers participated in the 2010/11 programme and 2,010 in 2011/12. The drop out of some agro-dealers in 2011/12 was due to discouragement from late payments in 2010/11, late delivery of the 2011/12 vouchers, phasing out of the matching grant for input supplier loans, and exclusion of a small number of agro-dealers following poor performance in 2010/11. Approximately 1,800 agro-dealers who received training did not participate in the programme in 2011/12. Some planned agro-dealer training and support activities were not implemented due to procurement process problems.
A full evaluation of the impacts of the programme on beneficiary and non-beneficiary households, on input access and use, and on the input supply system should be very informative.
In 2008 Rwanda responded to rapid rises in food and fertilizer prices by introducing a novel fertilizer subsidy system that involved (a) a general subsidy to fertilizers imported by the government and sold under auction to private companies obliged to sell these fertilizers at ceiling retail prices and (b) vouchers allowing farmers to buy a proportion of these at 50% of ceiling prices (Morris et al., 2009). Limited information available in 2009 suggested that both parts of the subsidy (through the ceiling price and vouchers) were broadly successful in raising productivity and in encouraging growth among private sector input suppliers (Morris et al., 2009). (p.56)
3.7. Mali and Senegal
Kelly et al. (2011) and Druilhe and Barreiro-Hurlé (2012) summarize key features of programmes in Mali and Senegal, noting that in both cases information is limited due to a lack of formal monitoring and evaluation processes. The programmes started in 2008 and were largely funded from domestic resources (though in Mali donors funded just under 40% of total costs in 2008/9 and 2009/10, but a number of donors then withdrew subsequent funding due to lack of transparency and reported ‘leakages’). Both programmes aimed to boost production and yields of staple food crops (for example, rice, maize, and wheat in Mali) and (in Mali) of cotton, to lower urban rice prices and to compensate farmers for high input prices. In both countries the primary objective was to promote food security, while in Mali there was an additional objective to promote exports. Programmes provided a 25 to 50% universal subsidy on fertilizers and seeds to all farmers cultivating target crops. Subsidies were provided for imports by private companies under tender and, in Mali, were administered using effectively a voucher system (a ‘caution technique’) through producer organizations and through agro-dealers associated with importers. In both countries fertilizer deliveries suffered from delays and importers suffered from late payments—although they benefited from growth in import volumes with, it appears, limited losses from displacement of unsubsidized sales. The lack of any apparent impact on rice prices, which remained high, calls into question both production statistics and programme benefits (Kelly et al., 2011).
3.8. Millennium Villages
The Millennium Villages Project (MVP) established integrated projects in selected villages to demonstrate the substantial changes that are possible with significant investments in health, agriculture, and community development across Africa. A major part of this is the provision of subsidized agricultural inputs (seed and fertilizer). Monitoring and evaluation systems are in place. This approach has invested in relatively small-scale, localized input subsidy programmes with much wider but generally unrealized objectives of national scaling-up.
3.9. Overall lessons
We now consider wider lessons across the eight reviewed programmes against the major issues identified in Chapter 2 as important for subsidy programme evaluation. (p.57)
3.9.1. Programme objectives
Here, we consider how far the different possible programme objectives discussed in Chapter 2 are found in the different programmes and types of programme.
• Food security (household or national), input adoption, and producer welfare are found as objectives of all or almost all programmes (with variation as regards particular emphasis on poorer or food-insecure producers).
• Only three programmes explicitly recognize the potential for producer subsidies to benefit poor consumers, except subsistence producers. Although this is small, given the importance of consumer benefits in consideration of large-scale staple crop subsidies, it is an improvement over findings in Dorward (2009b) where none of the reviewed programmes appeared to consider consumer benefits at all. Nevertheless, there is no recognition of the potential role of subsidies in addressing the price-productivity tight rope, and only in the MVP is there a wider recognition of the potential role of subsidies in driving forward pro-poor growth: even here there is no explicit consideration of the mechanisms by which this may be achieved or of processes of structural change in the economy as a whole.4
• Input use efficiency and soil fertility replenishment are only explicitly considered as programme objectives in the Tanzania programmes.
• Input access and input supply system development are explicit objectives in all programmes except Mali and Senegal (and limited sources may mean that this has been overlooked), the Zambia Food Security Pack and the MVP.
• Political considerations were important for the Ghana programme and the Zambia Fertilizer Support Programme, but are not explicitly mentioned in the documentation on other programmes (though they are likely to have been important). In Nigeria the FVP’s implementation is driven by an interest in depoliticizing some aspects of the existing wider Federal and state subsidy programmes, which are highly politicized.
3.9.2. Design and implementation
As regards design and implementation features of the different programmes, there is broad commonality across the different programmes as regards
• the basic focus of subsidy systems on producers as major (and generally sole) direct subsidy beneficiaries;
• a primary focus on subsidizing inputs for staple food production (for subsistence production or for sale into domestic markets);
• very substantial subsidized input price reductions (of 50% or more for most programmes), consistent with measures to address both affordability and profitability constraints to input use;
• almost all programmes clearly rationing (or attempting to ration) the quantity of subsidized inputs to be received per household, with vouchers being a common (but not universal) means of achieving this; and
• use of private sector importers to provide basic fertilizer supplies.
There are differences across the programmes as regards
• scale, with some national programmes and others piloting potential national programmes;
• targeting, with differences as geographical targeting (in some programmes) and some programmes focusing on food-insecure/vulnerable households and others seeking to maximize production by focusing on less poor households;
• use of vouchers for targeting, rationing, and/or supply system development;
• private sector and farmer organization involvement (and nature of involvement) in distribution;
• recognition of the importance of gender awareness in targeting and entitlement and access;
• complementary policies, and their links to programme objectives; and
• graduation and exits (with the Tanzanian programme explicitly limiting farmer access to three years and with scaling-back aspirations for both the Zambian programmes, but little or no mention of exits or graduation in other programmes).
3.9.3. Programme outcomes
Different programme outcomes—or information gaps about particular outcomes—are closely related to the programme objectives. Thus, limited examples (p.59) of subsidies leading to output (food staple) price changes and the lack of information on labour demands and markets and longer term and wider welfare and growth impacts are not surprising. Similarly, the lack of information on soil fertility replenishment is consistent with the lack of emphasis on this in programme objectives. There are, however, other similarities in outcomes that cut across differences in programme objectives, notably very common (but not universal) problems with late input delivery in subsidy programmes and common (but again not universal) leakages. Both of these are important for programme impacts, irrespective of programme objectives. There is a lack of reliable information on displacement and on production and productivity impacts, although increases in input use are reported for most programmes.
Lack of reliable information on a number of topics is associated with a paucity of programmes with monitoring and evaluation systems. Consequently, there are few estimates of economic benefits although, as will be discussed in Chapter 9, such estimates are difficult to calculate and, once calculated, often difficult to compare.
There are, of course, also substantial differences across programmes, some of these related to differences in programme objectives, as noted above. Thus, differences in reporting of input supply system impacts are related to differences in interest in these impacts. However programmes with the intention of developing supply systems may actually undermine them, if poorly designed and implemented.
3.10. Conclusions from recent experience
A number of observations from the limited programmes reviewed here warrant particular emphasis:
First, we reiterate a point made by Druilhe and Barreiro-Hurlé (2012) that the resurgence of agricultural input subsidy programmes in Africa is not a temporary phenomenon—they are attempting to address a real set of agricultural and development problems and their visibility and immediacy make them politically attractive. In this context, debates about their effectiveness and about ways to improve that effectiveness are healthy and should be welcomed. It is, however, important that such debates are based on thorough agronomic, economic, political, and administrative analysis of historic, current, and potential costs and achievements of subsidy programmes.
In this context our second observation, again supported by Druilhe and Barreiro-Hurlé (2012) is disappointing: it is notable how difficult it is to find comprehensive reviews of subsidy programmes, despite the substantial number of programmes that have been or are being implemented across Africa and the very substantial investments of public funds in these programmes. (p.60) There is a major need for country studies to document country experiences, using a comprehensive conceptual framework linking inputs, activities, outcomes, and wider impacts, as developed in the previous chapter.
Third, we note a continuing tendency for programmes to focus on production objectives and producer welfare, and to ignore the interests of consumers and the processes (and necessary conditions) for subsidy programmes to contribute to wider pro-poor economic growth. This is a critical omission, and is linked to the limited extent that the design and implementation of many programmes are integrated with complementary investments. Such integration is needed first for subsidy programmes to effectively deliver their stated objectives of incremental production, and then for them to contribute to the wider processes of pro-poor growth. Recognition of the importance of consumer price benefits and of the ‘price productivity tightrope’ is particularly important here. Druilhe and Barreiro-Hurlé (2012) also note the tendency to focus on producers, production objectives, and expansion of input access, and argue that there is insufficient attention paid to improved soil fertility and health, to development of private sector input supply, to complementary investments raising input productivity, to effective programme implementation (with more secure entitlement systems, better targeting, better monitoring and evaluation), and to phasing out and exits for input subsidies. However Druilhe and Barreiro-Hurlé (2012) themselves make little mention of programme benefits for consumers and for wider economic growth, and provide no discussion of farm level (as opposed to input supply) processes whereby subsidy delivery may lead to reduced need for and benefits from subsidies.
Fourth, and related to the previous two points, there appears in some programmes to be an unfortunate lack of interest in improving effectiveness and efficiency. This is evident from the limited monitoring, evaluation, and audit systems in some programmes, limited cost benefit and fiscal efficiency analysis, and limited attention to problems of late delivery, displacement, and leakage. Challenges from fiscal constraints, likely rising fertilizer prices, and the effects of climate change will make it even more important that in the future governments improve the efficiency and effectiveness of input subsidy programmes in both raising productivity and promoting wider pro-poor growth within and beyond agriculture.
Limited apparent interest in exits and graduation is also common. In the examples where this is not the case (Tanzania and Zambia) there appear to be difficulties in implementing this. This may be related to an apparent and indeed remarkable lack of attention to the question of why and how scaling back, graduation, and exit should and could occur. In the Tanzanian case, for example, farmers are to access subsidies for a maximum of three years. It is not clear how or why farmers will no long need access to the subsidy in a (p.61) fourth year, but this raises important questions about the processes of change needed for this (in farmers’ livelihoods and resources, in local economies, in input supply systems). Similar questions arise regarding the development of private sector input suppliers.
Two notable commonalities observed across programmes are (a) the lack or limited focus on replenishing soil fertility and (b) a strong (almost universal) prevalence of heavy subsidies (50% to 100% subsidy rates) on rationed inputs. This commonality occurs despite differences between programmes as regards first relative emphasis on improving national food security (and total input use and production) as against improving household food security (and helping food-insecure households) and second relative emphasis on supply system development. Political objectives and strong political influences on programmes are explicitly mentioned in only some of the programmes reviewed, but the scale of resources allocated to these programmes and their continued implementation suggests very strong political interest in and commitment to these programmes—even if the implications of this for programme design and implementation are not generally given very much emphasis.
When compared with the earlier review by Dorward (2009b), this review suggests that there is increased implementation of important aspects of smart subsidies, but there are still weaknesses in design and implementation, particularly late input delivery. There is also a continuing lack of emphasis on improving programme effectiveness and efficiency, limited attention to graduation processes, and inadequate attention to integration with complementary policies and programmes for improving achievements of both direct and indirect benefits of input subsidy programmes. The mixed record of input subsidies continues.
(2) Total budgeted subsidy cost was $25 million but only about $15 million was directly for the subsidy inputs and voucher costs (personal communication, Afua Branoah Banful).
(4) Other programmes may also implicitly consider that increased productivity and producer welfare may drive forward growth, but consideration of the food price, non-staple and non-farm production, and demand mechanisms is absent.