Transforming Staple Food Value Chains in Africa: The Case of Teff in Ethiopia

Abstract

We study changes in the last decade in the teff value chain, Ethiopia’s most important staple food crop by area and value. Upstream, there is increasing adoption of modern inputs and new varieties – especially by those farmers living close to cities – leading to higher land productivity. Mid- and downstream, we find improved processing costs and increasing willingness-to-pay for convenience and quality, as illustrated by the emergence of one-stop retail shops and the rise of more expensive teff varieties. Because of the large numbers of teff producers and consumers, this transformation has important implications on the country’s food security.

Notes

1. The importance of teff area-wise has increased absolutely but stayed relatively the same over time. In 2003, teff made up 20 per cent of all cultivated area as well while maize was making up 18 per cent.

2. The median producer price for mixed teff in the 2011/12 was 8.5 Birr/kg, significantly higher than wheat (6.45 Birr/kg), maize (4.1 Birr/kg), and sorghum (4.9 Birr/kg).

3. Kebeles are the second administrative level for the city under a given sub-city (recently kebeles have been re-organised to woredas with slight changes in geographical coverage).

4. This is in the perception of farmers, as it is often not clear after a couple of generations if varieties are improved or not.

5. The Central Statistical Agency of Ethiopia regularly collects data on the agricultural sector in their agricultural sample survey. While there is little information on different types of modern input use, they have, however, data on fertiliser use on teff fields. Comparing the results of 2003/04 with those of 2012/13 for the same five zones of our sample shows an increase in fertiliser use per hectare by 54 per cent over this period. These increases are in line with the results of our sample.

6. There is, however, an increasing perception with richer customers that red teff has major health benefits (presumably because of higher iron content). 65 per cent of the retailers ‘agreed’ or ‘strongly agreed’ with the statement that ‘red teff is increasingly been bought by rich consumers that are concerned about health’.

7. Using the lowess command in Stata.

8. While the recommended rate of fertiliser use in teff production is 100 kg of DAP and 100 kg of urea per hectare (Kenea, Ayele, & Negatu, 2001), a minority of farmers do this in practice.

9. Quncho is a combination of the magna quality DZ-01-196 and the white quality DZ-01-974. It has seemingly been successful because it combines the preferred magna characteristics of the DZ-01-196 and better yield performance of DZ-01-974 (Fufa et al., 2011). The quick spread of quncho is remarkable given that the variety was only released recently. The first farmers in our survey zone only adopted quncho in 2010, three years prior to the survey. It is found that while more than 40 per cent of the teff area is planted with quncho for close-by areas, this drops to almost zero for the areas farthest out.

10. In the Online Appendix, we test to what extent there has been a change in marketing margins over the period considered. We find that there have been significant decreases in processing costs and that the share of farmers in final retail prices has gone up but we do not find that there was a reduction in real marketing costs. While improved road infrastructure and communication technology have likely put downward pressure on margins, it seems that the increasing labour and fuel costs over the period studied might have offset these changes.

11. The changes that are noted at the farm level are seemingly not confined to teff producers only when Comtrade data as well as national data collected by the Central Statistical Agency are analysed. Over the period 2004–2014, it is estimated that total chemical fertiliser consumption increased by 144 per cent and that the share of cereal farmers that applied chemical fertilisers increased from 46 per cent in 2004/05 to 76 per cent in 2013/14. Other chemical use, such as the use of pesticides and herbicides by farmers, increased significantly as well. For example, Comtrade data show that Ethiopia’s quantities and values of herbicides imported tripled over the period 2004 to 2013 (from 4.7 million USD in 2004 to 15.5 million USD in 2013). More improved varieties for the major cereals were released in the period 2000–2011 than in the 30 years before (Bachewe, Berhane, Minten, & Taffesse, 2015) and while there are still problems in distribution of improved seeds, its use – while still relatively low – doubled during this period as shown by CSA statistics, from 10.1 per cent of cereal producers in 2004 to 21.5 per cent in 2014.

12. This is seemingly explained by a number of reasons. Attention to teff by Ethiopian agricultural researchers was discouraged locally during the Derg period in the 1970s and 1980s (Tareke, 2011). Internationally, teff does not fit well within the agricultural research priorities of the international agricultural research institutes of the CGIAR (Consultative Group for International Agricultural Research). The CGIAR institutes typically work closely with the NARS (National Agricultural Research System) in the development of better varieties and technologies, and have seen significant successes over time (Spielman & Pandya-Lorch, 2009). However, CGIAR research focuses on crops grown in a large number of countries. Consequently to date, teff has been the subject of almost no research by the CGIAR.

13. While there has been a significant debate on the exact estimates of rates of returns to agricultural R&D, they are overall evaluated to be high and to seemingly justify increased investments in this area (Hurley, Rao, & Pardey, 2014; Pardey, Alston, & Chan‐Kang, 2013). Using simplified assumptions, the quncho experience illustrates such positive benefits as well. Flaherty, Kelemework, and Kelemu (2010) estimate that in 2008 Ethiopia invested 70 million USD in agricultural R&D. Assuming, generously, that 10 million USD would have been spent annually on teff research, and that the quncho development was achieved after breeding investments for five years, this leads to a cumulative investment of 50 million USD. The results in this paper show that productivity of quncho – mostly responsible for the ‘improved seed’ coefficient in Table 3 – is on average 10 per cent higher than for traditional varieties. Assuming that half of the producers in Ethiopia would adopt quncho and that prices remain unchanged, that would lead to a sustained yearly benefit of 80 million USD. The internal rate of return for that investment would amount to 160 per cent (spread over a 20 year period). Admittedly, no costs for extension efforts are included in this calculation. However, even if the investments were quadrupled to 200 million USD to accommodate this, the rate of return would still be high at 40 per cent.

14. For example, Bachewe, Koru, and Taffesse (in press) estimate, using data from a large household survey collected in high-potential areas in the 2010/11 main agricultural season and relying on data envelopment analysis, that there is significant scope for productivity increases in teff if existing efficiencies could be improved. They estimate that teff yields and output could increase by approximately 50 to 190 per cent.

Additional information

Funding

The authors would like to acknowledge funding from the Ethiopia Strategy Support Program, funded by USAID and DFID.