Ethiopia is the largest teff producer in the world. In 2017, Teff accounts for 24% of the grain area, followed by maize 17% and sorghum 15% (Table 1). Amhara and Oromia are the two major regions, and collectively, the two regions account for 85.5% of the teff area and 87.8% of the teff production.

Teff is popular because it is a low risk crop [7]. Of the 12 million smallholder farmers in Ethiopia, 6.2 million grew the crop between 2004 and 2014 [8]. It can be harvested two to five months after sowing [10]. It is relatively resistant to many biotic and abiotic stresses [3]. Thus, it adapts to a range of growing conditions where major crops may fail [12]. However, the overall teff production in Ethiopia is at a rudimentary stage. Adoption of improved teff seeds is low, farm plots are fragmented, mechanization is almost absent, harvest loss is substantial, and public investment in research is lacking [3, 17]. To estimate inefficiency in teff production, Bachewe et al. (2015) use the average Relative Total Factor Productivity (RTFP). Their calculated RTFP value is 0.361, indicating the RTFP could be improved by 177% [8].

In fact, Ethiopia initiated teff research programs as early as late 1950s. The research programs focused on breeding teff varieties to enhance production [1, 17, 18]. With such efforts, 35% of teff producers adopted improved teff seeds in 2012, compared to 7% in 2002. Although increased over the decade, the limited access to and unaffordable price of the improved seeds still prevent smallholder farmers from widely adopting the seeds [17].

Changes in the types of cultivated teff are observed over time. A noticeable change is the increase in white-colored teff at the expense of red and mixed-colored teff. The white teff made up 69.6% of the teff grown in 2012, compared to 48.2% in 2002. On the other hand, the share of red teff declined from 36% to 19.7% during the same period [17]. The color of teff grains, white, mixed and red, mostly decides its grade for quality and price. The white commands the highest price and red the lowest. The very-white magna, a sub-type of the white, commands even a higher premium price [3]. The premiums for the white and magna teff partly come from the social preference for the white color, and Slaklkneckt (1997) described the white teff as follows [10]:

White teff is the preferred type but only grows in the highlands of Ethiopia. It requires the most rigorous growing conditions, has the mildest flavor, and is the most expensive form of teff. Just like white bread has been a status symbol in the United States, white teff is usually reserved for the wealthiest and most prestigious families in Ethiopia.

In line with the description above, more of the white teff tends to be cultivated and traded around Addis Ababa [17]. Addis Ababa is the capital city where many of the affluent reside, and the largest teff market of the country [7]. Besides the social preference for the white color, the popularity of the white comes from the introduction of the improved variety Quncho [17]. Quncho meaning the top in the local language is the major achievement of the teff breeding program in Ethiopia. The variety was derived from the combination of two traits, high yield and white seed. With its release in 2006, Quncho became the first improved variety broadly adopted by the Ethiopian farmers and contributed to 10% increase in teff productivity [1, 7].

The agronomic practices to cultivate teff have little changed, slowing improvement in teff productivity. One such practice is broadcasting for sowing. Broadcasting is a traditional method commonly used for teff sowing in Ethiopia due to its small grain size [17]. A teff grain measures 0.9-1.7 mm long and 0.7-1.0 mm wide, and 150 teff grains weigh equal to one wheat grain [15]. The current averaged seed rate in broadcasting is around 40-50 kg per ha [17]. This rate compares to 3-4 kg per ha, a suggested teff sowing rate in Australia [15]. Broadcasting scatters seeds unevenly in the field, and establishes irregular stands. It in turn makes a proper farming a challenge [13]. Instead, row planting is recommended as a good alternative to the broadcasting.

The grounds to support the row planting are that it reduces the amount of seeds per unit area, and competition between seedlings for water and nutrients while increasing efficiency in weeding [19]. Some studies show promising yield advantages and reductions in input costs with the row sowing method [13]. However, Vandercasteelen et al. (2016) argue otherwise [19]. The adoption of the row planting, in fact can be costlier than the broadcasting. It may require 30% more labor than the broadcasting due to lack of the mechanized row planter. This research further identifies there were no statistical differences between the two planting methods for the farmer profits and land productivity. Such different results imply a careful assessment with a choice of suitable teff sowing schemes [19].

Teff growers till soil 4.4 times on average per production cycle. The rationale behind the frequent tillage is that unbroken soil prevents proper germination of small teff seeds [17]. On the contrary to this common perception, Gebretsadik et al. (2009) show plowing frequency did not affect teff yields in their field experiment [20].

Application of agro-chemicals has risen in teff production. The share of teff producers using herbicide grew from 31% to 62.9% in 2012, and the most frequently applied herbicide was 2-4-D to control broadleaf weeds in a surveyed area. This increase in herbicide use may well be expected as weeding is highly labor-intensive, yet critical for teff productivity [17].

For fertilizer, teff accounts for 54% of the total fertilizer applied in the grain production of Ethiopia, and for the largest fertilized area among cereals. In 2011, 981,000 ha of teff was fertilized whereas maize and wheat were 565,000 ha and 556,000 ha, respectively [3, 21]. In terms of production costs, chemical fertilizer made up 20.2% of the total cost, only behind the land rent 21.4% [22]. It may seem counterintuitive that teff producers would use more fertilizer in this low-yielding cereal (Table 1). However, the increasing teff market price could justify fertilizer application. In addition, due to its long shelf life, Ethiopian farmers attach an intrinsic value to teff [21].

As in the white teff production, more fertilizer is applied around Addis Ababa, the largest teff market. The proximity to this market incentivizes teff growers to maximize yields by increasing fertilizer application [17]. Although the cost of chemical fertilizers can be three times higher around the capital city than other remote areas, the yields and output values per ha are high enough to offset the cost [7, 22]. Nonetheless, anecdotal claims report that distribution, access and affordability are still serious issues with fertilizer. An underlying reason for these issues is the dominance of the government in the agricultural input supplies; the government agencies are the sole fertilizer importer and distributor. Its dominance limits private sector’s participation and competition for an efficient fertilizer supply chain [22].

Household characteristics influence teff production [18]. Those characteristics include the gender of household head, number of dependents in the household, level of labor supply, age structure, and involvement in the extension service. Female-headed households are negatively associated with teff production due to relative shortage in labor and capital resources, compared to the male-headed households. Households with a large number of family members are positively associated with teff production due to high labor availability. A U-shaped relationship between a producer’s age and probability of producing teff implies learning would be required for a full engagement in teff production. A household’s involvement in the extension service seems to increase the likelihood of producing teff by acquiring necessary knowledge and skills [18].

The word ‘teff’ comes from the Ethiopian word ‘teffa’, which means ‘lost’ because of its minute grain size [15]. Teff grains are manually harvested by sickles and threshed with ox tramping on them [3]. Crymes (2015) estimates 25-30% of teff would be lost before and after harvest, and lodging may contribute to the yield loss up to 30%. The high losses along the production processes can reduce the available quantity of teff by up to 50% [10].

To better estimate the loss, Amentae et al. (2016) conducted a survey in one of the largest teff production areas in central Ethiopia [14]. The results show the farm-level loss was the highest or 8.18% of teff produced, followed by 3.58% at a consumer level, 2.85% at a retailer level, and 1.67% at a wholesale level. When the farm-level loss was dissected, threshing accounted for 2.91%, transporting from farm to home 2.62%, harvesting 1.87%, transporting from home to market 0.28%, and storage 0.5% [14].

Ineffective tools and little mechanization appear a main contributing factor to the large loss; mechanical tools are developed for maize in Ethiopia, but not for other cereal crops including teff. Poor road conditions and a traditional transporting means such as donkeys are also identified a cause for loss. Inappropriate storage conditions might be a source of loss, yet do not appear to be a major concern [13, 14].

Amentae et al. (2016) further relate teff losses to other characteristics: gender of household head, household size, output level, and distance to the nearest market [14]. Female-headed households were more prone to higher loss due to the labor-intensive activities required. Larger family sizes reduced loss with more labor available. Higher levels of output were associated with higher loss because of the handling difficulties. As the nearest market was farther, more loss was expected from spillage and poor handling [14].

However, depending on the study areas and contexts, results of the loss vary. For instance, teff loss along the teff value chain to Addis Ababa appears negligible [23]. The estimated loss during harvest was 1.8%, followed by 0.2% at a retail level and 0.1% while transporting it. This study concludes the overall loss could range from 2.2% to 3.3% along the teff value chain to Addis Ababa. It challenges the conventional perception that staple supply chains in developing countries are susceptible to high loss. These different findings with the teff loss may indicate that some necessary investment has been put in place to minimize teff loss around the largest teff market, Addis Ababa. Or, post-harvest handling is more conservative in Ethiopia than in developed countries where cosmetic qualities are often of critical importance [23].

Efficient farming practices, favorable household characteristics and loss management are essential to improve teff production, but not an end by itself [22]. A functional value chain, market and supportive policy should coexist to enhance the teff industry.

Ethiopia has yet to develop an efficient teff value chain and marketing scheme. Its value chain is often described as unsophisticated or untraceable [7, 14]. Currently, little evidence exists for modernized teff trading and retailing practices. For instance, the role of credit is minor, most of the transactions are on the cash basis, and standardization of teff grading is virtually absent [7].

Channels through which teff passes need to be identified to understand the value chain. For instance, Gebremedhin et al. (2007) show the most important channel for teff producers appeared to be a producer-wholesaler in their surveyed area [18]. On average, 65% of teff producers sold teff to wholesalers, 31% to retailers, and 4% to rural assemblers or directly to consumers [18].

Urgessa (2011) identifies eight different channels for teff marketing, based on the teff flow in volume (Table 2). Of the channels, Channel 3 carried the largest volume followed by Channel 4. Urgessa (2011) further characterizes this teff market as oligopoly; the four largest teff traders controled 65% of the total teff volume, expressed as CR4 or market concentration ratio [22].

Minten et al. (2016 a) conclude the major teff value chain in Ethiopia was relatively short. The study area represents 42% of the national commercial surplus of teff, and supplies over 90% of teff to Addis Ababa. Along the value chain, there are three critical players between teff producer and Urban Retailer (UR): Farmer Traders (FT) operating at a village level, Rural Traders (RT) at a regional level, and Urban Traders (UT) at an urban level. This value chain ranges from zero node where URs purchase directly from teff growers, one node passing through one of the FT, RT or UT, to five nodes FT-FT-RT-RT-UT. The study shows the value chain with two trade nodes or fewer is dominant with 85% of the cases, thus concludes the value chain to the largest teff market is fairly short [7].

Since 2008, the price of teff has increased, widening the price gap between teff and other crops such as wheat and maize [3]. From 2015 through 2017, the average year-on-year price of teff increased 9% or 145 Birr (Ethiopian currency equivalent to USD 5.3) per 100 kg in the surplus market. In January 2017, the price of 100 kg teff was 1750 Birr or USD 64.1, which was twice than that of wheat. The price of teff tends to be individually negotiated and likely varies from one transaction to another. But, the traders in Addis Ababa play a key role in determining a teff price for the major production areas. Referring to the brokers in Addis Ababa, traders in the regional markets set their prices [3].

Minten et al. (2016 a) investigated a teff price structure in detail [7]. One notable result is that teff growers obtained 79.4% of the final retail price on average, ranging from 78% up to 84%. Another study by Urgessa (2011) reveals a similar result; the teff producers took 78.7% of the consumer price while the assemblers, wholesalers and retailers shared the rest of the price. Further, the teff producers collected the highest net profit followed by the urban retailers, farmer traders, regional retailers, urban assemblers, and regional wholesalers in the descending order [22].

The high profit share of the producers appears contradictory to the common observation about producers in developing countries, who may not take a substantial share of the price [7]. Nonetheless, teff producers might not benefit from the price structure when profit per unit of time to produce teff is considered. As the crop is highly labor-intensive with little mechanization, approximately 141 person-days are needed per ha per production cycle. Taken this into account, an average reward per person-day amounts to USD 2.25 per day [7]. It is slightly above the World Bank’s extreme poverty threshold, or USD 1.90 per day for low income countries. The high labor requirement to produce teff also imposes negative impacts on Return On Investment (ROI). Elias et al. (2017) estimate the ROI of teff as 3.88%, which is very low compared to ROI of wheat or 31.6% [24]. In addition, traders often cheat farmers in weighing. The case is particularly intense at a harvesting season when farmers have to sell in bulk. Such practices can prevent farmers from being rightly paid [22].

On the consumer side, market uncertainty entails additional costs to verify quality of teff they purchase. Ethiopian consumers purchase teff at a cereal shop or a market, clean it at home, and get it milled at a mill [17]. Therefore, the final price consumers pay is likely higher than the retail price, given their additional time and costs [3]. However, mills around the capital city started making changes. The mills are becoming a one-stop shop; they began selling, cleaning and milling in a stream line for consumers [17]. Urban better-off consumers may prefer and be willing to pay premiums for the one-stop service for reduced search costs, guaranteed quality and increased convenience [3]. Working women highly demand the convenience because their opportunity costs are high from the multiple steps involved from purchasing teff to preparing injera at home [17].

Ethiopia grows more than 90% of the teff in the world [11]. Despite its largest production volume, the country is not capitalizing its own crop in the international market [3]. Instead, other countries are actively participating in teff production and marketing to capture the rising market [15].

Eritrea was once a major importer of Ethiopian teff. But, from 1998 to 2000, the war between the two countries severed their formal trading relations. Critically in 2006, the Ethiopian government imposed a ban on exporting unprocessed teff grain and flour. Since the imposition of the ban, the export volume of teff has declined [3].

The rationale behind the ban is to bring the domestic price of teff to an affordable level, and improve food security. A low domestic price of teff benefits consumers, especially the rural and urban poor. Removing the export ban would likely increase the local price of teff to a higher international level. It would hurt domestic consumers [13]. For teff producers, the export ban limits their exposure to price volatility in the international market, and discourages multinational companies to take over the local teff industry. Otherwise, their takeover would likely drive smallholder farmers out of teff business, and raise land conflicts, a similar phenomenon with quinoa in Bolivia. Additionally, exporting teff might weaken the nutritional status of Ethiopia. With teff less available and costlier, poorer Ethiopians could be forced to switch to cheaper yet less nutritious substitutes such as sorghum, barley or wheat as their staple [10].

Despite the standing export ban, the domestic price of teff remains high. The high price is attributable to the rapid increase in domestic demand from the growing population, income, and urbanization. A study shows income growth of 20, 30, or 40% would increase the urban teff demand 22, 33, or 44%, respectively [17]. Minten et al. (2016 b) also show the income elasticity in teff is higher in urban areas than rural, 1.2 for urban and 1.1 for rural.

Other factors holding the teff price high include alleged smuggling to neighboring countries and increasing injera export [3, 10]. Following the export ban of teff grain and flour, export volumes of injera increased to 2.5 million kg in 2012, or 270% increase from the 2008 level [3]. The United Arab Emirates (UAE) was the top destination of the injera exports, absorbing over 65% of the volumes. The UAE was followed by the USA, Bahrain and Sweden, each of which made up about 10% of the injera exports [3]. Over the last several years, many local companies entered the injera business to benefit from the growing market. The Mama Fresh, for instance is one of the largest local companies. According to a survey, the local companies export over 30,000 pieces of injera daily to Washington and New York in the USA where large Ethiopian communities exist [11].

Teff is the only cereal crop that Ethiopia has a comparative trade advantage. It can be grown in a large part of the country. The current lack of opportunities for international markets discourages teff producers and traders to improve yields, reduce post-harvest losses, add more values, and develop standardized teff grading systems [3]. The food security issue hardly justifies the present export ban; teff is not a staple for the most food insecure due to its relatively high price [13]. Thus, Ethiopia needs to reconsider the export ban in order to lead Ethiopian farmers to engage in the rise of teff popularity. Otherwise, the country would miss a good opportunity to enhance farmer income, livelihood as well as GDP. And once the global teff market is mature, Ethiopia would have to grapple with countries with advanced farming technologies, a large-scale production system, and efficient value chain so as to capture the share of the teff market [13].