Determinants of teff row planting technology adoption on small farms yield in North Shewa zone, Amhara region, Ethiopia

Abstract

Adoption of agricultural technology is an important avenue to increase agricultural productivity and thus improve the food security and livelihood of societies. To achieve this, it is quite important to understand the determinants of the adoption of new agricultural technologies. This study examined the determinants of teff row planting technology adoption on small farms’ yields in the North Shewa Zone, Amhara Region, Ethiopia. The study applied the quantitative research method. A sample of 796 agricultural households was considered, and the logit model was used to estimate the factors affecting the application of determinants of teff row planting technology adoption on small farms’ yield. According to the study, plot distance, active household member, number of oxen, extension service, access to credit, distance from market, marital status of household head, and distance from main road are the main factors influencing teff row planting technology adoption on small farm yield in North Shewa Zone, Amhara Region, Ethiopia. The findings suggests that expanding credit access, extension services, and market access to farm households are important to improving the application of teff row planting technology.

Keywords:

• Determinants
• row planting
• adoption
• logit model
• North Shewa Zone

PUBLIC INTEREST STATEMENT

Agriculture is the main source of the Ethiopian economy. However, the adoption rate of modern agricultural technologies is very low. Therefore, this study intends to examine the determinants of teff row planting technology adoption on small farms’ yields. The results showed plot distance, active household members, number of oxen, extension service, access to credit, distance from market, marital status, and distance from the main road are the main determinants of teff row planting technology adoption in the North Shewa Zone.

1. Introduction

Agriculture has been and continues to be at the center of the economic policy of many less developed countries (LDCs) in general and Sub-Saharan Africa (SSA) in particular. Consequently, growth in the agricultural sector has been critical to achieving poverty reduction and income growth, which creates spillover to the remaining sectors (World Bank, 2014). However, the production and productivity of the agricultural sector in Sub-Saharan Africa (SSA) and much of the developing world is generally low due to poor technological adoption (Abraham et al., 2014).

According to Diriba (2018), agriculture is by far the largest sector of Ethiopia’s economy, serving as the basis for the country’s food security and a source of livelihood for over 80% of its people. It accounts for about 34.1% of the Gross Domestic Product (GDP), employs 79% of the population, accounts for 79% of foreign earnings, and is the major source of raw materials and capital for investment and the market. Teff is Ethiopia’s most important staple crop and it has the largest value in terms of both production and consumption in Ethiopia, and the value of the commercial surplus of teff is second only to coffee (Minten et al., 2013). In 2018/19 production year in Ethiopia, a total of 54,034,790.51 quintal of teff was produced by nearly 6.8 million farmers (CSA, 2019; Mariyono, 2019a). According to CSA (2017), teff accounts for 3,017,914 (24%) hectares of the grain area, followed by maize (17% and sorghum (15%). Teff accounts for 1.664 tons per hector. The Amhara and Oromia regions are the two major regions of teff producers in the country. Collectively, the two regions account for 85.5% of the teff area and 87.8% of the teff production. Different reasons, including a traditional cultivation method, have contributed to teff’s low productivity and quality. Given the scarcity of suitable arable land, it becomes largely difficult to meet the increasing needs of the rapidly growing population through expansion of the area under cultivation (Vandercasteelen et al., 2013).

In Ethiopia, the broadcasting method of teff planting used by the farmers is one of the main reasons why teff productivity is low (Negussie, 2022). It was also argued that the broadcasting method of teff planting reduces productivity due to uneven distribution of seed and increased competition between teff plants for nutrients, water, and light (Fufa et al., 2011). Adoption of modern agricultural technologies is believed to improve the income of smallholder farmers through enhanced agricultural productivity. Improving the agricultural productivity of farmers requires developing and disseminating cost-effective agricultural technologies. Accordingly, increasing agricultural production, reducing poverty, and meeting the demand for food without irreversible degradation of the natural resource base is possible (Asfaw & Shiferaw, 2010). The results from previous studies showed that the productivity of teff was 14.8 quintals per hectare and 20.1 quintals per hectare for the broadcasting and row planting of teff, respectively (Tesfaye et al., 2015).

There are different contributing factors to the low adoption level of new agricultural technology in less developed countries. Mariyono (2019b; 2019b; 2019c) stated that the farmers’ knowledge, market place, seed technology, farm site, the role of credit, and commercialization were important factors that have persuaded farmers to adopt technology in Indonesia. Suprehatin (2021) stated that farmer technology attributes or farm household characteristics, farm characteristics, and institutional factors are the main determinants of agricultural technology adoption in developing countries. Tamirat (2020) and Tefera (2018) stated that sex of household head, age of the household, farmer’s experience, total annual income, education of household head, family size of household head, holding of livestock, use of credit, extension services access, and attending training at a farmers’ training center are the determinants of smallholder teff farmer’s row planting technology adoption. Additionally, the educational status of the household head, land size, participation in training, membership in the association, the number of livestock owned in the tropical livestock unit, marital status, household labor, distance to the central market, extension contact, off/non-farm income, farm size, household size, access to mass media, and access to credit affected the farmer’s adoption decision of the row planting of teff production (Negussie, 2020) and (Ayal et al., 2018).

The contribution of this study to the existing literature is three-fold. First, some studies were conducted on the adoption of wheat row planting technology in different parts of the country, for example (Amare, 2018; Dinku & Beyene, 2019; Tamirat et al., 2016). However, this study is examining the determinants of teff row planting technology adoption on small farms in the North Shewa Zone. Second, previous studies have emphasized the impact of technology adoption on the income of a rural household in different parts of the country, for example (Negussie, 2020; Natnael, 2019; Yonas, 2006; Tesfaye et al., 2015). However, this study is examining the determinants of teff row planting technology adoption on small farms’ yield in the North Shewa Zone, Amhara Region, Ethiopia. Third, no research has been conducted so far in the study area in recent years at the zonal level about determinants of teff row planting technology adoption on small farms in North Shewa Zone, Amhara Region, Ethiopia. Therefore, the goal of this study was to fill these gaps and provide evidence on the determinants of teff row planting technology adoption on small farms’ yields in the North Shewa Zone, Amhara Region, Ethiopia.

2. Methodology

2.1. Description of the study area

North Shewa is one of the eleven administrative zones found in the Amhara Region of Ethiopia, which takes its name from the kingdom or former province of Shewa. The capital city of the zone, Debre Berhan, is about 130 kilometers north of Addis Ababa. The absolute location of the study area is found at latitude 80 43‘06“−100 43’ 58” N and longitude 380 39“50” −400 06“32”’ E. The zone is bordered on the south and the west by the Oromia Region; on the north by the South Wollo Zone; on the northeast by the Oromo Nation Special Zone; and on the east by the Afar Region. The total population of the North Shewa zone is estimated at 3.5 million inhabitants. The estimated terrain elevation of the area above sea level is 3009 meters. Teff is mainly produced in the areas that are centrally located in the country, with many woreda in the Amhara Region around Lake Tana and between Bahir Dar and Addis Ababa. At national level, the highly productive woreda are grouped in a single cluster and of the top 25 producing woreda, 15 are located in Amhara and the remaining ten are in Oromia. The study area, North Shewa Zone, is one of the major cereal crops, including Teff growing areas of Ethiopia. In the area, teff is the most important crop grown next to wheat production in terms of area coverage and volume of production (Warner et al., 2019). The geographical location of the zone is presented in Figure 1 as follows.

Figure 1. Geographical location of North Shewa zone; source: (Fikire & Zegeye, 2022).

2.2. Data sources and method of data collection

This study uses data collected from both primary and secondary sources. The study is based on household-level cross-sectional survey data collected from north Shewa zone farmers using a questionnaire. The questionnaire was designed to collect data about the demographic, economic, social, and institutional characteristics, and adoption practices of farm households. It was prepared in the form of closed and open questions. Secondary data was collected from documented and published sources like books, journal articles, conference proceedings, and reports from the North Shewa Agricultural Office.

2.3. Sampling method and size

For this study, samples were drawn by using the multi-stage sampling method. First, four districts of the zone, namely Minjar Shenkora, Angolela Tera, Moretna Jiru, and Menz Gera, were selected purposively since these districts have high potential for agricultural practices and topographical similarity. According to the North Shewa zone administration office (2021) in the selected districts, there are a total of 117,149 households. Having this, Vogel (1986) and Malhotra (2012) suggested that for a largely homogenous population, if it is between 35,001 and 150,000, the researcher can select a maximum sample of 800 respondents. Following this, for this study, 800 sample households were drawn. Second, from the total Kebeles [1] of the selected districts, 30 Kebeles were randomly selected, and finally, simple random sampling was used to select each respondent from each selected Kebele. Due to missing information, four observations were dropped. Thus, the final sample size of the study is determined to be 796 farm households.

2.4. Methods of data analysis

This study employed both descriptive and econometric methods of data analysis. In this study, the STATA version 13 software package is used to analyze and estimate statistical and regression models.

2.4.1. Descriptive analysis

Descriptive statistics such as percentage, frequency, mean, and standard deviation were used. Descriptive statistics help to gain a better understanding of the demographic, socio-economic, and institutional characteristics of the farm households and to look at the teff row planting technology adoption in the study area.

2.4.2 Econometric analysis

Adoption is measured in terms of the probability used by smallholder farmers. The probability of technology adoption refers to whether the household head adopts teff row planting technology adoption in the 2021 production season. In the econometric method, a logit regression is employed to identify the determinants of teff row planting technology adoption in the study area.

2.5. Model specification

In this study, the adoption of row planting technology is the dependent variable. The response variable is binary, taking values of one if the farmer adopts teff row planting and zero if the farmer does not adopt row planting. However, the independent variables are both continuous and discrete. This study estimates the logit model, which models the teff row planting technology as a function of socioeconomic and institutional characteristic factors, to identify determinants for teff row planting technology adoption. Following Gujarati (2004), the binary logistic model is expressed as:

(1) ${\mathrm{L}}_{\mathrm{I}}=\mathrm{l}\mathrm{n}\left(\frac{\mathrm{P}\left({\mathrm{Y}}_{\mathrm{i}}=1/{\mathrm{X}}_{\mathrm{i}}\right)}{1-\mathrm{P}\left({\mathrm{Y}}_{\mathrm{i}}=1/{\mathrm{X}}_{\mathrm{i}}\right)}\right)={\mathrm{Z}}_{\mathrm{i}}=\alpha +{\beta }_{\mathrm{i}}{\mathrm{X}}_{\mathrm{i}}+{\epsilon }_{\mathrm{i}}$(1)

Where: represents the dependent variable (adoption of row planting); represents the intercept of the model; represents the unknown coefficients to be estimated; represents a vector of explanatory variables that can affect the dependent variable; and represents the disturbance terms of the model. The coefficients need to be adjusted to be marginal effects in the case of the logit model. In other words, the marginal effect, which gives the partial derivatives indicating the change in the probability of the dependent variable relative to a unit change in one of the independent variables, needs to be computed. As the relationship between the explanatory variable and the absolute probabilities is non-linear, marginal effects vary according to the choice of vector Xs and, consequently, they will vary among individuals according to the point of evaluation. By differentiating the logistic model, we find the marginal effects of the explanatory variables on the probabilities:

$\frac{\mathrm{\Delta }\mathit{\text{pij}}}{\mathrm{\Delta }\mathit{\text{xi}}}=\mathit{\text{pij}}(\beta j-\beta i^{\prime })$

Where

(2) $\beta i^{\prime \prime }={\sum }_j\mathit{\text{pij}}\beta j$(2)

The interpretation of the logit model is relative to the reference or base category group is difficult, even if this study used non-adopter as a base category. The coefficients need to be adjusted to be marginal effects in the case of the logit model. For continuous variables, the marginal effect is the probability change in response to a unit change in the value of the independent variable at the mean value. For dummy variables, the marginal effect is computed as the difference in probabilities of the dependent variable between the group with a designated value of one and the base category. Furthermore, it should be noted that the signs of the beta (β) coefficients are not necessarily the same as those of the marginal effects.

2.6. Description, measurement, and expected sign of variables

Table 1 shows the description, measurement, and expected signs of the variables.

Table 1. Measurement, description of variables, and expected sign

Name of variable and symbol	Description and measurement of variables	Expected sign
Teff row planting technology adoption	It is a categorical variable = 0 for non-adopters, 1= for adopters of teff row planting technology	Dependent variable
Sex	It is a dummy variable which can be (male = 1 and 0= otherwise).	+
(ii)Age of household head	Defined as the number of years from the date of birth to the day of the survey interview date in the full year. The continuous variable measured in year	+
(iii)Marital status	It is a categorical variable 0= if single,1= married	+
(iv)Family size	One or more persons occupy a housing unit and have the same eating arrangement. Measurement in number.	+
(v)Education	It is a categorical variable. 0= No formal education, 1= Literate	+
(vi)Cultivated land Size	It is a continuous variable measured in hectares	+
(vii)Off-Farm participation	It is a dummy variable. 1=Yes, if an individual participates in an off-farm activity, 0= No if an individual does not participate in an off-farm activity.	+
(viii)Livestock Unit	It is a continuous variable measured in number	+
(ix)Distance from Market	It is a continuous variable measured in kilometers.	-
(x)Access to credit	It is a dummy variable; representing 1= if a household had credit access and 0= otherwise.	+
(xi)Extension service	It is a dummy variable; representing 1= if a household had access to extension service and 0= otherwise.	+
(xii)Total land size	It is a continuous variable measured in hectares	+
(xiii)Active household	It is a continuous variable, measured in the number of the person whose age is between 18 and 65 years.	+
(xiv)Distance from the main road	It is a continuous variable measured in kilometers.	-
(xv)Saving	It is a continuous variable measured in Ethiopian Birr	+
(xvi)Agricultural experience	It is a continuous variable, measured in the number of years.	+
(xvii)Agricultural cooperative	It is a dummy variable, 0= if farmers are not a member of the cooperative, 1 = are a member of farmers.	+
(xviii)Plot distance	It is a continuous variable measured in kilometers starting from home to plot area.	-
(xix)Access to transportation	It is a dummy variable that can be (yes = 1 and 0= No).	+
(xx)Number of oxen	It is a continuous variable, which is measured in number.	+

3. Result and discussion

This chapter presents the results of the descriptive and logistic regression analyses of the study. The first part provides the descriptive statistics, while the second part presents a regression analysis where an examination of the results based on the equation specified in the model specification section is presented.

3.1. Descriptive statistics

Table 2 presents the summary information on the categorical independent variables of the studied households. The information in the table above shows that out of a total of 796 households in the sample, 79 households (9.92%) are headed by women and 717 (90.08%) are headed by men. Regarding the educational level of the household, out of 796 respondents, 309 households (38.82%) have not attended formal education and 487 (61.18%) of household heads are literate. Regarding the marital status of households, out of 796 respondents, 698 people (87.69%) are married and 98 (12.31%) households are single. On the subject of off-farm participation, out of a total of 796 respondents in the sample, 670 (38.07%) engaged in off-farm economic activities, while 493 (61.93%) of the rest of the sample did not. Looking at the membership of farm cooperatives, out of a total of 796 respondents in the sample, 670 (84.17%) are cooperative members, while 126 (15.83%) are not members. In addition, the table shows that 425 (53.39%) sample households have access to credit, while 371 (46.61%) sample households do not have access to credit. Regarding extension service 743 (93.34%) sample households have access to extension service, while 53 (6.66%) sample households do not have access to extension service. Last but not least, about 402 (50.50%) of sample respondents have access to transportation, whereas 394 (49.50%) of sample respondents do not have access to transportation services.

Table 2. Percentage distribution on a categorical independent variable

Variables	Category	Freq.	Percent
Sex	Female	79	9.92
	Male	717	90.08
Education	No formal education	309	38.82
	Literate	487	61.18
Marital Status	Married	698	87.69
	Single	98	12.31
Saving	No	181	22.74
	Yes	615	77.26
Off-farm participation	No	303	38.07
	Yes	493	61.93
Farm cooperative member	No	126	15.83
	Yes	670	84.17
credit	No	371	46.61
	Yes	425	53.39
Extension service	No	53	6.66
	Yes	743	93.34
Access to transport	No	394	49.50
	Yes	402	50.50

Table 3 shows that the mean age of the household is 43.136 years old with a standard deviation of 11.026; the minimum age of the respondents in the sample is 19 years old; and the maximum age is 80 years old. In terms of family size, the mean family size is 4.98 with a standard deviation of 1.992, the minimum family size is 1, and the maximum family size is 14. Regarding the total land size of the household, the average arable land area is 1.619 hectares, with a standard deviation of 0.758 and a minimum of 0.25 and a maximum of 5.5 ha of land. Besides, the cultivated land size of the household head on average is 1.581, with a standard deviation of 1.101 and a minimum of 0.25, and a maximum of 10 hectares of land. The mean distance to the market is 9.892 kilometers, with a standard deviation of 9.148 and a minimum of 0.01 and a maximum of 70 kilometers from the market. The mean number of oxen is 1.997 with a standard deviation of 1.03 and a minimum of 1 and a maximum of 10 oxen. Compared with tropical livestock units, the mean is 6.263 units with a standard deviation of 4.32, a minimum of 1, and a maximum of 44.6 livestock units. The mean number of active household members is 3.369 with a standard deviation of 2.271 and has a minimum of 0 and a maximum of 14 members. The mean plot distance is 5.528 kilometers with a standard deviation of 6.638, a minimum of 0.01 and a maximum of 50 kilometers.

Table 3. Summary statistics for continuous variables

Variable	Mean	Std. Dev.	Min	Max
Age of household head	43.136	11.026	19	80
Family size	4.98	1.992	1	14
Total land size	1.619	.758	.25	5.5
Cultivated land size	1.581	1.101	.25	10
Distance from market	9.892	9.148	.01	70
Number of oxen	1.997	1.03	1	10
Tropical livestock	6.263	4.32	1	44.6
Active household member	3.369	2.271	0	14
Plot distance	5.528	6.638	00.01	50

Table 4 shows that 511 (64.20%) farmers were non-adopters of teff row planting technology, whereas 285 (35%) households were adopters of teff row planting technology in the North Shewa zone.

Table 4. Adoption of teff row planting technology distribution of sample respondents

Teff row planting technology adoption	Freq.	Percent
Non-adopter	511	64.20
Adopter	285	35.80

Source: own survey (2021)

3.2. Econometrics analysis

3.2.1. Diagnostic tests

3.2.1.1. Multicollinearity test

Before going into the basic steps of regression and model interpretation, it is imperative to check whether the data set suffers from multicollinearity. Therefore, the study tests for multicollinearity for categorical and continuous explanatory variables by using the test of contingency coefficient and variance inflation factors. For continuous explanatory variables, the mean-variance inflation was found to be less than 10. Therefore, table 5 shows that the data is free from the problem of multicollinearity.

Table 5. Variance inflation factor

Variables	VIF	1/VIF
Number of oxen	1.397	.716
Tropical livestock	1.328	.753
Cultivated land size	1.196	.836
Active household member	1.184	.844
Family size	1.156	.865
Distance from market	1.117	.895
Plot distance	1.089	.918
Total land size	1.013	.987
Mean VIF	1.185

Table 6 shows that the contingency coefficient, or correlation coefficient, between all categorical explanatory variables, in which the result was found to be lower than 0.85. Therefore, we can conclude that the data is free from multicollinearity problems in both categorical and continuous explanatory variables.

Table 6. Pairwise correlations matrix for categorical variable

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) sex	1.000
(2) education	0.051	1.000
(3) marital status	−0.34*	0.024	1.000
(4) saving	0.071*	0.134*	−0.13*	1.000
(5) off-farm participation	0.043	0.094*	−0.045	0.161*	1.000
(6) agriculture experience	0.073*	−0.24*	0.011	−0.041	−0.057	1.000
(7) farm cooperative	0.075*	−0.014	−0.07*	0.315*	0.064	0.147*	1.000
(8) credit	0.052	0.008	−0.12*	0.082*	0.082*	0.010	0.050	1.000
(9) extension	−0.004	0.021	−0.049	0.072*	0.019	−0.049	0.022	0.013	1.000
(10) access to transport	0.020	0.016	0.023	−0.13*	0.013	−0.017	−0.08*	0.040	0.215*	1.000

3.2.1.2. Goodness of fit

Table 7 shows the results from the generalized Hosmer-Lemeshow and pseudo-R-squared statistic tests, which are used to test whether the model is fit or not. The generalized-Lemeshow statistic assesses whether or not the observed events match the predicted events by sub-grouping the probabilities estimated from the data. The null hypothesis is that the differences between the observed and predicted events are insignificant, so the model is correct or fitted. The result of the pseudo-R-squared statistics shows that the higher pseudo-squared values for the fit models compared to the intercept models indicates that the fitted full models better predict the outcomes of the dependent variables, and the predictors are effective in modeling the different outcomes of the determinants of row-planting technology adoption. Therefore, Table 7 presents the results of the goodness of fit and it supports that the model is fit or correct from all results.

Table 7. Hosmer-Lemeshow test

Logistic model for row planting, goodness-of-fit test
Number of observations	796
Number of groups	5
Hosmer-Lemeshow chi2(3)	3.02
Prob > chi2	0.3885

3.2.2. Determinants of teff row planting technology adoption on small farms yield

3.2.2.1. Logit estimation results and interpretation

Table 8 shows logistic regression and estimation results with marginal effect, the following variables are the significant determinants of teff row planting technology adoption on small farms’ yields.

Table 8. Logistic regression and estimation result with marginal effect non-adopter as base level

Row planting adoption	Mfx(dy/dx)	Coef.	Std.Err.	t-value	p-value	Sig
Distance from main road	−0.017	−.078	.019	−3.90	0.000	***
Sex	0.034	.157	.299	0.45	0.654
Age of household head	0.002	.009	.013	0.69	0.491
Education	0.011	.048	.126	0.37	0.715
Marital status	0.060	.268	.108	1.90	0.057	*
Family size	0.005	.024	.057	0.43	0.668
saving	0.045	.197	.288	0.84	0.403
Off-farm participation	0.053	.236	.242	1.24	0.217
Agricultural experience	0.000	.001	.012	0.11	0.914
Farm cooperative member	0.041	.186	.228	0.68	0.497
Total land size	0.029	.13	.102	1.13	0.261
Cultivated land size	0.022	.101	.14	0.80	0.426
Distance from market	−0.018	−.079	.015	−5.05	0.000	***
Access to credit	0.081	.366	.13	1.96	0.05	*
Extension service	0.169	.705	.773	1.84	0.065	*
Number of Oxen	0.079	.352	.174	2.88	0.004	***
Tropical livestock unit	0.000	.00836	.028	0.01	0.998
Active household member	0.019	.087	.044	1.83	0.067	*
Plot distance	−0.009	−.038	.016	−2.48	0.013	**
Access to transport	0.061	.272	.164	1.27	0.205
Constant	0.017	1.002	2.191	1.25	0.213
Mean dependent var = 0.642			SD dependent var = 0.480
Pseudo r-squared = 0.269			Number of obs = 796
Chi-square = 163.901			Prob > chi2 = 0.000
Akaike crit. (AIC) = 800.637			Bayesian crit. (BIC) = 898.909

Note: *, ** and*** are level of significance at 10%, 5 % and 1 % respectively

The model has Prob >chi2 = 0.0000, and Pseudo R2 = 0.269. Pseudo R2 tells us our model as a whole fit significantly better than an empty model (i.e., a model with no predictors).

The model has Prob >chi2 = 0.0000, and Pseudo R2 = 0.269. Pseudo R2 tells us our model as a whole fit significantly better than an empty model (i.e., a model with no predictors).

Distance from the main road is affects adoption negatively, in which the improvement of the likelihood of teff row planting is significant at the 1% significance level. The marginal effect indicates that, when the distance from the main road is increased by one kilometer, the probability of adopting row planting decreases by 1.7% compared with non-adopters, keeping all other variables constant. The proximity of farmers to all-weather roads is essential for timely input delivery and output disposal. As a result, investing in better road infrastructure is critical for promoting adoption and welfare gain. This result is consistent with (Beshir et al., 2012). In addition to distance to the main road, distance to the market also significantly affects adoption of teff row planting technology.

Distance to market negatively affects the likelihood of adoption of teff row planting and it is significant at a 1% significance level. The negative association indicates that as distance to the market decreases, the likelihood of adoption of row planting increases. The marginal effect indicates that, when the distance from the market is increased by one kilometer, the probability of adopting row planting decreases by 1.8 percent, compared with non-adopters, keeping all other variables constant. This is because distance will matter when the technologies are away from the farm area, which creates distribution problems and adds an extra cost to purchasing agricultural inputs. The result points out that through improving access to market points, there is a potential to increase the probability of technology adoption for food crops in the country. This result is consistent with (Vandercasteelen et al., 2013) and (Mariyono, 2019c).

Plot distance also affects adoption negatively and it is significant at a 5% significance level. This implies that as distance to the market decreases, the likelihood of adoption of row planting increases. The marginal effect indicates that, if the plot distance is increased by one kilometer, the probability of adopting row planting decreases by 0.9 percent, compared with non-users, keeping all other variables constant. The farther the plot is from the homestead, the less likely it will be utilized for agricultural inputs. The finding is in line with the findings of (Menale et al., 2012). Further to the above distance variables, institutional factors like access to credit and extension service also affects teff row planting technology adoption significantly.

Household credit access affects the adoption decision positively and significantly at the 10% significance level. This means that the more farmers have access to credit facilities, the more they will be motivated to adopt teff row planting. Indeed, when farmers have enough credit, they can purchase inputs when needed and in desired quantities. The results also show that the marginal effect on access to credit is 8.1%, keeping other variables constant. Small farm household heads who have the opportunity of getting credit for agricultural inputs participate more than those who have no access. The probable reason for the positive result is that credit use is one way of improving financial constraints for purchasing row planting machines from private owner farmers has to improve labor constraints in the study areas. Furthermore, access to credit leads to increased agricultural productivity. This study is consistent with (Tamirat, 2020) and (Mariyono, 2019c).

The extension service has both positive and significant significance at a 10% level. The positive association indicated that as farmers visited with extension agents, the likelihood of adopting row planting increased. The marginal effect shows that as the frequency of extended visits increases, the probability of applying row planting by 16.9%, compared to those who did not apply. All other variables remain unchanged. The main reasons for possible factors in farmers’ decisions to participate in row planting technology and their level of production are that farmers receive several services from extension services, including technical services on their production. This study is consistent with (Tamirat, 2020; Negussie, 2020; and Ayal et al., 2018).

Demographic variables like marital status of the household head and active household members also affect adoption significantly. The result from the estimation indicates that the coefficient for marital status is positive and significantly affects the improvement of the likelihood of teff row planting at the 10% significance level. The positive sign indicates that married farmers are more likely to adopt teff row planting technology than single-households. The marginal effect indicates that, if the household is married, the probability of applying teff row planting increases by 6% compared to singles, all else being equal. This result is in line with (Negussie, 2020; and Ayal et al., 2018).

Active household members have a positive influence on the application of row planting technology at a 10% significance level. This means that as the number of active household member’s increases, their decision to apply row planting technology is becoming more and more important. If an active household member increases by one person, the probability of applying row planting increases by 1.9% in the study area. The result is expected since family labor is the major source of the labor force in the rural area. Hence, those households who have access to more family labor are likely to participate in the row planting of Teff. The reason for this positive effect was that row planting was labor-intensive and, hence, its availability could increase the area under cultivation. This suggests that labor is among the critical variables in influencing the decisions of households to participate in row planting. This study is consistent with (Yonas, 2006) and (Tesfaye et al, 2015).

Finally, number of oxen the household owns is positive and significant at the 1% significance level. The positive association indicates that as farmers have more oxen, the possibility of adopting row planting technology increases. The marginal effect indicates that as farmers have more oxen, the likelihood of adopting row planting practices increases by 7.9%; all other variables remain unchanged. That is, having a large number of oxen allows him or her to participate in row planting of the teff crop. It could also indicate that adopters have better access to financial sources through the sale of livestock, which could be used to purchase farm inputs, such as seed and fertilizer, and livestock used to minimize risk. The main reasons are that household heads that have many tropical livestock unit will have a high income and they will use their oxen for plowing, so it is easy for them to participate. This study is consistent with (Tamirat, 2020) and (Tefera, 2018).

4. Limitation and area of further research

This study used cross-sectional data and examined the determinants of teff row planting technology adoption on small farm yields; it is limited to showing the time effect of adopting row planting technology on small farm yields. Future research is recommended to examine the impact of teff row planting technology adoption on small farm yields by using other data sets that can show time changes like panel data.

5. Conclusion and recommendation

5.1. Conclusion

Agriculture is the basis for the Ethiopian economy, whereas severe poverty, low agricultural productivity and food insecurity are the main challenges for the rural farmers in our country. Therefore, increasing agricultural production and productivity through the use of modern agricultural technology is considered a major solution to reduce poverty and increase food security. The main objective of the study is to examine the determinants of teff row planting technology adoption on small farms yielding in the North Shewa zone. Multistage sampling techniques and procedures were used to collect the data from 796 households. Descriptive statistics and econometric methods of data analysis were employed. The Logit model was used to identify the determinants of Teff row planting technology adoption on small farms’ yields. As a result of the study, it was discovered that plot distance, active household member, number of oxen, extension service, access to credit, distance from market, marital status of household head, and distance from main road are the determinants of teff row planting technology adoption on small farm yield in the study area. The distance variables (plot distance, distance from the market and distance from the main road) negatively affect the adoption of teff row planting. Whereas access to credit, high number of active household members, extension service and married marital status affect adoption positively.

5.2 Recommendation

This study recommends policies that aim to encourage and expand the adoption of new or improved agricultural technologies to enhance agricultural productivity and, in return, reduce poverty in the country. These policies includes expanding access to credit for poor farmers, road infrastructure development for distant areas and improved extension services especially for farmers with low level of education.

Specifically, this study suggests that zonal and regional governments need to strengthen policy interventions and expand access to credit and agricultural extension services. The credit has to be a long term credit with affordable low interest rate, in addition to governmental financial institutions applying policies that encourage private financial institutions to supply credit for adopting farmers is important. The credit facilities should target at poor farmers, particularly those who have not adopted row planting technology, but have the potential to adopt, due to a lack of operating capital. Extension services should be improved significantly. This is very essential to providing information mostly for illiterate farmers about the use and benefits of adoption of teff row planting technology. The extension service helps the farmers to acquire knowledge and skills about the distance between teff plants, characteristics and application of inputs. Further to this, it introduces them to prepare local instruments for the application of teff row planting. With regard to distance from the market, the government should work on making the market for agricultural inputs or outputs near to the farm households through infrastructure development and improvement in the supply chain of new agricultural inputs.

Acknowledgments

We would like to thank the editor and anonymous reviewers for their supportive comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors received no direct funding for this research.

Notes on contributors

Abebaw Hailu Fikire

Abebaw Hailu Fikire is Assistant professor at Debre Berhan University, College of Business and Economics, Department of Economics, Debre Berhan, Ethiopia. He earned a BA degree in Economics from Debre Berhan University and an MA degree in Development Economics from Hawassa University. His current research interests include poverty, food security, urban housing, technology adoption, and rural development.

Anteneh Bizualem Asefa

Anteneh Bizualem Assefa is a lecturer at Debre Berhan University. His research interest includes macroeconomics policy analysis and international trade