The Drought Tolerant Maize for Africa (DTMA) project was initiated with the aim of developing drought-tolerant maize varieties (DTMVs) with a potential yield of 1 t/ha under moderate drought conditions, increasing productivity under farmers’ conditions by 20-30%, and producing grain with an annual average value of US$160-200 million in drought-affected areas. The project was funded by the Bill & Melinda Gates Foundation and jointly implemented by the International Maize and Wheat Improvement Center (CIMMYT) and IITA in partnership with the national research and extension systems of 13 African countries.
Since it was launched in 2006, the DTMA project has documented successes in the development and dissemination of DTMVs in Africa. About 160 DTMVs were developed between 2007 and 2014. Most have been successfully disseminated to maize farmers in 13 African countries. This report provides some evidence on the productivity and welfare impacts of adoption of DTMVs in Nigeria, a country participating in the DTMA project.
Sampling and data collection
Data for this report came from a household survey conducted in Nigeria from November 2014 to February 2015. To ensure the selection of a representative sample of households, a multi-stage random sampling procedure was applied. First, based on the area of land devoted to maize production, the 36 States in Nigeria were divided into five homogenous subgroups; 18 of these States were randomly selected, accounting for about 62% of the total area of land devoted to maize production in the country. Then, following the recommendation of the National Bureau of Statistics (NBS) for a nationally representative data collection, 10% of the local government areas (LGAs) in each of the selected States and 5% of the total extension areas (EAs) per LGA were randomly selected. From the list of communities obtained from the NPC, two communities were randomly selected in each of the selected EAs. Finally from the households in each of the selected EAs, five farming households were randomly selected, resulting in a total of 2305 households. The survey questionnaire included information on socioeconomic characteristics of the households, household expenditure on food and non-food, output for maize and other notable crops, and income from various sources. The data was collected electronically using the “surveybe” software.
Measuring adoption and outcome variables
Adoption was measured based on whether or not the household was cultivating one DTMV or more in the 2014/2015 production season. The outcome variables are maize productivity as measured by maize yield (kg/ha), household welfare as measured by per capita food expenditure (N/year), and the Foster, Greer and Thorbecke (FGT) indices of poverty (head count index, poverty gap index ,and poverty gap-squared index). The head count index measures the rate of poverty, the proportion of people living below the poverty line, which was computed using the World Bank’s US$1.25 per-capita per day measured at Purchasing Power Parity (PPP)1. The poverty gap index measures the depth of poverty, which is the extent of income shortfall from the poverty line. The poverty gap-squared index
measures the severity of poverty that indicates the degree of income inequality among the poor themselves.
Adoption rate of DTMVs in Nigeria
About 52% of the sample maize farmers reported they had access to DTMVs through the DTMA project. The adoption rate among them was about 44%, suggesting that not every farmer who had access to the DTMV seeds actually planted DTMVs. Irrespective of access to seeds, the adoption rate was 24.5% in the study sample, which is quite appreciable if we remember that the first DTMVs are not even 10 years old (Fig. 1). For estimating impact of adoption, we used 24.5% as the effective adoption rate and employed an Endogenous Switching Regression (ESR).
Productivity and welfare impacts of DTMVs
Adoption of DTMVs in Nigeria has resulted in measurable gains in maize productivity and impacts on household welfare. Our result shows that maize yield has increased by 23% (268 kg/ha) as a result of the adoption.
Impacts by gender
Analysis of impacts by gender revealed that adoption of DTMVs has greater productivity gains on average among female-headed households (314kg/ha) than male-headed households (262 kg/ha). In addition, average productivity gains were much higher in dry savanna agroecological zones(Fig. 2). Despite about 314 kg/ha productivity gains for female-headed households average productivity gains for the whole sample were much lower (268 kg/ha) since the these households constituted only 10.45% of the total sample.
Impacts on per-capita food expenditure
Like productivity gains, changes in per-capita food consumption were also sizable as a result of the adoption of DTMVs. Our result shows that adoption has increased per-capita food consumption by N10,683 (about $35).
Impacts on per-capita food expenditure by gender
Analysis of impacts by gender revealed that as a result of adoption of DTMVs, male-headed households increased their per-capita food consumption by N11,303 and female-headed households by N5919. Note that this should not be interpreted as a gender-biased outcome. Since the initial level of per-capita food expenditure was higher among femaleheaded households in our DTMA sample; it is reasonable to see such large increases among male-headed households.
Impacts on per-capita total expenditure
Changes in per-capita total consumption expenditure similar to improvements in per-capita food expenditures,were also sizable as a result of the adoption of DTMVs. Our result shows that adoption has increased per-capita total consumption expenditure by N13,907 (about $46 per year).
Impacts on per-capita total expenditure by gender
Analysis of impacts by gender revealed that as a result of adoption of DTMVs, male-headed households increased their per-capita total consumption expenditure by N 14632 and female-headed households by N7384.
Impacts on poverty outcomes
Based on preliminary results from the endogenous switching regression approach using awareness about DTMV as an instrument, we estimated poverty outcomes for adopters. Using per-capita total consumption expenditure values and the World Bank’s $1.25 per-capita per day measured at Purchasing Power Parity (PPP), about 62% of adopters were observed to be below the poverty line. Had they not adopted DTMVs, our preliminary results suggest that the rate of poverty among adopters would have been about 83%, suggesting that the gain in maize productivity due to adoption had eventually led to about 21% point poverty reduction (Fig.3) among adopters of DTMV. When it comes to depth of poverty, adoption
of DTMVs yielded a 15.8% point reduction in depth of poverty for the same group. Furthermore, it helped to reduce the severity of poverty among the poor by 9.4% points for the same adopters. For the population of maize farmers as a whole, this translates into an average poverty reduction point of only 4.9% in 2014/2015. Looking at impacts by gender, we found that the poverty headcount rate declined by 23% points among male-headed households and 7% points among female-headed households. We reported a higher productivity gain among female-headed households in the previous section. However, while examining the poverty impacts of adoption by gender, we found higher effects for male-headed households. This result shows that higher productivity gains are accrued for more affluent (non-poor) female-headed households.
Estimating the total number of individuals lifted above the poverty line
The next challenge is then to calculate the total number of households lifted above the poverty line as a result of adoption. Based on adoption rates and predicted poverty reduction rates, the total number of households lifted above the poverty line as a result of DTMV adoption is estimated to be 2.68 million individuals (0.37 million households). This result reflects only direct effects—effects on households that adopted DTMV. Indirect effects on non-adopters and consumers through reductions in prices are not included.
The account provided some evidence on the productivity gains and welfare impacts of adoption of DTMVs in Nigeria. The report also provided heterogeneity impacts of adoption, focusing on gender and ecological zone. Results indicate that adoption of DTMVs resulted in a 268 kg/ha gain in maize productivity and N10683 per annum in welfare benefits as measured by the gain in annual per-capita food expenditure. In terms of poverty reduction, we found a 21% point reduction in poverty as a result of adoption. An estimated 370,000 households (equivalent to 2.7 million individuals) managed to move out of poverty as a result of adoption of DTMVs. Further estimated impacts on productivity gains suggest that the program was more beneficial to female-headed households. However, estimated results on poverty outcomes suggest that the project targeted better-off female headed households. Moving forward, there needs to be a more targeted intervention for addressing disadvantaged groups such as poor female-headed households. In particular, constraints related to access to DTMV seeds and other barriers associated with adoption have to be addressed.