Adoption and Diffusion of Sustainable Intensification Practices for Maize-Legume Production in Ethiopia: A Panel Data Analysis

Abstract

This paper documents the adoption of improved maize technologies by smallholder farmers in the maize–legume based farming systems of Sustainable Intensification of Maize and Legumes Cropping Systems for Eastern and Southern Africa (SIMLESA) intervention districts representing two maize growing environments: the high rainfall, long season maize production environment and the medium to short season, low rainfall maize growing environments in Ethiopia. The report is based on a panel data set collected from two household surveys in 9 districts. While 898 randomly selected households were interviewed during the first round survey conducted during 2011, the second round survey conducted in 2013 managed to re-visit 874 households with a success rate of 97.3%. Results are based on descriptive analysis augmented with some econometrics. Analysis of household characteristics indicates that the majority of sample households are male-headed with large family size. Male household heads are older and better educated than their female counterparts. Irrespective of gender of the household head, most of the household heads are in the active age category (average age of 41 years) suggesting that they are capable of contributing t labor and managerial decisions that determine the livelihood strategies perused. Household endowments often referred to as assets also determine the livelihood strategies pursued by farm households. Examination of asset ownership indicates that land, livestock, farm implements, and household utensils are the most important assets of economic significance to farm households. On the average, a typical household owned 2.4 ha and 2.2 ha in 2010 and 2013, respectively, with female-headed households owning statistically less land compared to male-headed households. Social capital often expressed in mutual trust and connectedness helps to access information, enforce contracts, and cope with shocks and vulnerability. In this study, about 72% of households in 2010 and virtually all of the surveyed households in 2013 were members, at least, in one local farmer group. Level of participation in saving and credit associations has more than doubled in 2013 from the base year of 2010 suggesting the increased role of such networks in the economic and social arena of households. In terms of crop production, maize cultivation dominates legume crops. During the two periods, while absolute area allocated to maize has declined from 1.0 ha per household to 0.8 ha, the area share allocated to maize has increased from 54% to 56% in the years 2010 and 2013, respectively, suggesting the relative importance of maize in in the households’ food security portfolio. Sustainable agricultural practices (SAIPs) promoted in the maize-legume based farming systems of the study areas include high yielding maize varieties (OPVs and hybrids), maize-legume intercropping, legume rotations, inorganic fertilizer, farmyard manure, minimum or zero tillage and physical soil and water conservation (SWC) measures such as stone, soil and terrace boxes. Among the 8 SAIPs promoted in the study areas, improved maize varieties and inorganic fertilizers are found to be widely adopted in both years across all districts. Of the two most widely used SAIPs, the level of adoption of improved varieties is the highest with about 84% and 88% during the years 2010 and 2013, respectively. Similarly, the level of adoption of inorganic fertilizer remained almost at the same level at about 72% in both years, suggesting the use of improved varieties and inorganic fertilizers are very well rooted among smallholder farmers in all districts. The level of use of other SAIPs, however, remained fairly at low levels. And yet, adoption of legume rotations, legumemaize intercropping as well as soil and water conservation practices increased from the 2010 level of 17%, 10% and 21% to 19%, 16% and 27% in the year 2013 registering significantly substantial increases over the 2010 level. Least in the adoption scale are the use of mulching and planting trees around boundaries of maize fields. The mean number of SAIPs adopted by a typical maize grower also increased from about two in the year 2010 to 3 in 2013 suggesting adoption of sustainable agricultural intensification practices has improved over the years. Gender disaggregated adoption of SAIPs revealed that male-headed households adopted on overage about three SAIPs compared to their female-headed counterparts. In addition, the level of adoption of improved maize varieties, inorganic fertilizer, and maize-legume intercropping and soil conservation practices is significantly higher among male-headed households than among female-headed households. The level of use of farmyard manure, maize-legume rotations, and stubble-mulch, however, appears to be similar between the two groups of households. The data further revealed that the level of use of minimum tillage practices not only low but also did not differ by gender of the household head suggesting a lot remains to be done to demonstrate the technical efficiency as well as the economic benefits of minimum/conservation tillage to smallholder farmers. Further analysis of the adoption and diffusion of maize commercial varieties indicate level of use varied considerably across districts but not over periods. The proportion of households using non-recycled improved seeds increased marginally from about 70.8% in the year 2010 to about 72.8% in the year 2013. Yet, femaleheaded households used more recycled hybrid seeds than male-headed counterparts. Across districts, four hybrid maize varieties namely BH 540, BH 660, BH 543 and Shone dominate maize production. Of these hybrid varieties, BH540 and BH660 remain to be the two most important commercial maize varieties dominantly grown in moisture-stress and in high rainfall and long season maize growing areas, respectively. About 50% of the maize growers rely on a single variety, BH-540, suggesting reliance on a single improved variety would have far-reaching consequences on maize production and productivity. Intensity of improved maize adoption, measured as share of maize area under freshly purchased hybrids or OPVs recycled at most for three years, indicate that overall intensity of improved maize adoption has not changed much over the two periods with about 72% and 70% for the years 2010 and 2013, respectively. Intensity of hybrid maize adoption, however, increased slightly from 63% to 66% registering about 3 %age points over the two periods. The study further indicated that among the nine districts, three districts namely, Hawassa, Misrak Badawacho, and Shalla had the largest share of improved maize area covering about 93%, 89% and 82% of the maize area during the 2010 production year and about 94%, 79% and 86% of the total maize area during the 2013 production year, respectively. Intensity of improved maize is lowest in Pawe and Dugda Districts covering about 20% and 54% in 2010 and 33% and 48% of the maize area during 2013, respectively. Among the other sustainable agricultural intensification practices (SAIPs) promoted, the level of inorganic fertilizer adoption is the second highest with about 73% and 72% of households using on maize during the 2010 and 2013 cropping seasons, respectively, suggesting level of use remained relatively stable over years. Adoption rates, however, varied considerably from one district to the other. The data also showed that, adoption rates are higher for DAP with 72% and 69% of households whereas the corresponding figure for Urea is 53% and 54% for 2010 and 2013 cropping seasons, respectively. As expected, inorganic fertilizer use is higher in Gubuesyo, Bako Tibe, and Hawassa Zuria districts, endowed with relatively high and less erratic rainfall, with over 92% compared to districts receiving low and erratic rainfall. As expected, rate of use of Urea is the lowest in the moisture stress districts of Dugda and Adami Tulu with less than 15% of households adopting. The sheer number or proportion of households using fertilizer is unlikely to depict the true level of inorganic fertilizer adoption on maize unless complemented with figures indicating the intensity of use per unit of maize area. On the aggregate, intensity of fertilizer use on maize has significantly increased from the 2010 level of 97.3 kg/ha to 118.7 kg/ha in the year 2013. Interestingly, much of this increment is attributed to higher intensity of Urea than DAP. While the rate of Urea use increased from the 2010 level of 41.3 kg/ha to 57.3 kg/ha in 2013 production year registering 44% increment, the intensity of DAP use increased from 56 kg/ha to 61.4 kg/ha during the same period registering only 10.9% increment. Further scrutiny of the data indicated that intensity of fertilizer adoption varied not only by year but also by district. Of the nine districts, inorganic fertilizer intensity of use is highest in Bako Tibe followed by Gubuesyo and Hawass Zuria Districts with an average rate of 233 kg/ha, 211 kg/ha and 202 kg/ha, respectively. The lowest rate of use is registered in Dugda and Adami Tulu districts with an average rate of 26.0 kg/ha and 31.3 kg/ha, respectively. Further analysis of the intensity of inorganic fertilizer use by gender revealed while female-headed households, on average, applied 60 kg/ha, male households used about 113 kg/ha on their maize crop suggesting female-headed households applied significantly lower rates of fertilizer than their male counter parts. The observed discrepancy in the intensity of fertilizer use is more pronounced in the use of Urea than DAP. On overage, female-headed households applied 26 kg/ha of Urea and 34 kg/ha of DAP whereas male-headed households used 52 kg/ha of Urea and 62 kg/ha DAP suggesting the gap is much wider in the use of Urea than in DAP. Econometric analysis of determinants of improved maize technology adoption revealed that gender; education level of the household head; access to extension; number of oxen owned; and participation in rotating credit associations (locally known as Ekub) had statistically significant and positive impact. While age of household head, participation in off-farm activities and distance to market exerted a statistically significant and negative influence. Location showed a differential effect on hybrid maize adoption. Similarly, the econometric analysis on inorganic fertilizer adoption revealed that education, oxen ownership, and membership in micro finance institutions had a significant and positive impact on intensity of inorganic fertilizer use. Age of the household head, participation in off-farm activities, land weighted by adult equivalent and distance to market negatively and significantly influenced both adoption and intensity of inorganic fertilizer use. Empirical results of the multinomial logit model exploring the factors influencing the interdependence of hybrid maize varieties and inorganic fertilizer adoption indicate that, compared to the base category of using non-hybrid maize varieties without inorganic fertilizer, extension contact has been the only cross cutting significant factor enhancing the use of both hybrid and non-hybrid maize varieties. Family size, education, oxen ownership, and reliance on relatives (social capital) positively determine the combined use of inorganic fertilizer with maize varieties. Age, offfarm factors, and distance to market negatively influenced the use of hybrid maize with inorganic fertilizer whereas credit enhances their combined use. Gender and maize area share of good soil negatively influenced the combined use of none-hybrid maize and inorganic fertilizer. Land per adult equivalent is the only significant variable that impacted the use of hybrid maize without inorganic fertilizer. The multinomial model results, therefore, revealed different sets of factors with different magnitude influence the adoption pattern of hybrid maize and inorganic fertilizer adoption, providing evidence that the decision to use hybrid maize varieties is influenced by the decision to use inorganic fertilizers and vice versa. The survey allows evaluating the impact of improved maize adoption on maize production and productivity and household food security. Maize productivity is basically yield per unit area measured in kilograms per hectare. Food security, on the other hand, was a self-assessment of the food production and consumption situation of households into four distinct groups, namely chronic food security, transitory food security, self-sufficient and surplus producers. Households reporting food shortages throughout the year and a few months in a year are grouped under chronic and transitory food security categories while households claiming to have ends meet and produce surplus for the market are grouped under self-sufficient and surplus producers. Analysis of the welfare impact of adopting improved maize technologies in terms of productivity and income gain suggested that both maize productivity and profitability have increased in year 2013 compared to 2010. Significant maize yield differences have been observed between adopters and non-adopters, male and female-headed households, moisture abundant and moisture stress locations. The study further indicated that productivity differences between hybrid and OPVs as well as between non-recycled and recycled maize seeds are significant in favor of the former in the series. The gross margin analysis, for both years, showed that the use of DAP and Urea took the lion’s share of costs that vary. Across districts, female-headed households generated significantly lower gross margin than male-headed households. Furthermore, analysis of the relationship between self-declared level of food security and maize productivity, suggests that households in the food secure group not only enjoyed higher productivity but also produced more maize grain per household than households in the food insecure group. The data further revealed that food secure households cultivated significantly higher maize area than the food insecure group. However, maize area for both household categories declined significantly in 2013 from the base year. The results suggest that perceived food security and productivity are interrelated warranting further investigation.