









































Pa
ge

 
1



Pa
ge

 
38

American Journal of  Smart 
Technology and Solutions (AJSTS)

Determinants of  Teff  Production in North Showa Zone, Central Highlands of  Ethiopia 
Tadele Anagaw Zewdu1*

Volume 2 Issue 1, Year 2023
https://journals.e-palli.com/home/index.php/ajsts

Article Information ABSTRACT

Received: February 28, 2023

Accepted: March 16, 2023

Published: March 18, 2023

Using primary data collected from 115 households of  North Showa Zone, Ethiopia, the 
study examined determinants of  Teff  Production. In order to investigate the effect of  
each predictor variable on the household teff  production level a bivariate analysis was 
performed. Among the econometric method of  analysis, a logistic regression model was 
fitted to analyze the potential variables affecting household teff  production level in the 
study area. The result of  the descriptive analysis revealed that about 85 (73.90%) were 
teff  producers while 30 (26.10%) of  the households were found to be non-producer. 
Moreover, the logistic regression model estimates that among the nine variables included 
in the logistic model, eight of  them were significant at different probability level (1, 5, and 
10). These are the education of  household head, total cultivated land for teff  production 
per hectare, number of  oxen, technology adoption, access of  extension services received 
by households, sex, fertilizer application and kind of  teff  variety used. Finally, improving 
land quality, creating awareness towards importance of  farm technology adoption, and 
providing frequent extension service were recommended.

Keywords
Determinants, Logit,Teff, Teff  
Production, Ethiopia

1 Salale Universitys, Fiche 245, Ethiopia 
* Corresponding author’s e-mail: tadeleanagaw21@gmail.com

INTRODUCTION
Ethiopian agriculture is virtually small-scale, subsistence-
oriented, and depends on rainfall (Anderson, 2007). 
It is also kicked off, with high population pressure and 
traditional farming systems have caused ecosystem 
degradation in the form of  soil erosion and declining 
soil fertility and erratic climate are the challenges to 
production. Additionally, the current smallholder farming 
systems are undergoing a reverse transformation in which 
farm sizes are declining, few farmers are moving out of  
agriculture, and instead are diversifying into non-farm 
activities from a small farm base (ATA, 2013). 
On the other hand, agriculture is the main source of  
livelihood for a large proportion of  the population, 
especially for the people residing in the rural areas 
(Schmidt and Kedir, 2009).    Teff  is the major staple food 
crop to most Ethiopian people living in the highlands, 
comprising more than 65% of  the population. However, 
the national average yield of  teff  is very low and 1.4 
tons per hectare and the development of  high-yielding 
cultivars would be very beneficial (CSA, 2013).  Hence, 
the need for improved crop varieties that are high yielding 
and with the capacity to survive in such a degraded and 
risk-prone environment is important (Spielman, 2008). 
There is still a question of  yield stagnation due to the 
low yield potential of  the existing teff  varieties and 
other determinants (Tareke et al., 2008).  Hence, the 
main question of  this study was to identify the major 
determinants of  teff  production in Grar-Jarso district 
and provide relevant information to the concerned body. 
Data source and sampling
Two-stage probability sampling technique was employed 
to select the sample of  farm households. In the first stage, 
sample kebeles1 were selected randomly. Then, in the 
second stage, using the list of  farm households living in 

each of  the selected kebeles as a sampling frame, sample 
households were selected randomly using a probability 
proportional to size sampling technique. Hence, this 
study is based on a primary data collected from a sample 
of  randomly selected 115 smallholder farm households. 
The data was collected through face-to-face interview 
using semi-structured questionnaire.
Estimation Strategy
The logit model was the appropriate econometric model 
to identify the determinants of  teff  production in the 
study area.  This model was chosen; it has an advantage 
in revealing the relative influence of  the probability of  
teff  production through different input utilization. Logit 
model, which has a discrete part, is appropriate which 
handles the probability of  the extent of  production 
in a proper way. Logit model which helps to test the 
determinants of  teff  production can mathematically be 
specified as follows:
pi= E (Y=1 such that; xi = β0+βi xi....... (1)
Where; Y=1 implies the given farmer participates in teff  
production
 xi= a vector of  independent variable
 β0= The constant term
 βi= i= 1, 2 ...n.  are the coefficient of  independent 
variable to be estimated
1 the lower class of  district  

Where; zi=β0+βi xi   ifpi, is the probability of  being 
producer and (1-pi) the probability  of  being a non 
producer of  teff

https://journals.e-palli.com/home/index.php/ajsts
mailto:tadeleanagaw21%40gmail.com?subject=


Pa
ge

 
39

https://journals.e-palli.com/home/index.php/ajsts

Am. J. Smart. Technol. Solutions 2(1) 38-42, 2023

Therefore, we can write this equation as;

Laterally,  pi/(  1-pi )  is the odds ratio of  producer farmers 
with the ratio of  the probability that a given farmer can 
participate in production to the probability that the 
farmer who will be participating in production.  Then, if  
we take the natural logarithm of  equation (e), we obtain;

If  the disturbance term Ui is taken in to account, the logit 
model becomes

Where β0 is an intercept and β1, β2, … βn are slopes of  
the equation in the model, and X is the vector of  relevant 
farmer  characteristics.
Consequently, Li , which is the log of  odds ratio, is called 
logit or logit model (Gujarati, 2004).  Hence, the above 
Logit Model is employed to estimate the effect of  the 
hypothesized explanatory variables on the teff  production 
decision of  farmers. 

RESULTS AND DISCUSSIONS
This analysis is based on data obtained from the 
questionnaire survey. The questionnaires of  115 households 
had been examined for incorrectness and missing data 
were grouped (classified) into two groups, namely, teff  
producer and none producer groups. The data presented in 
the following part explains the distinction between the two 
groups of  households. This section briefly presents the 
major determinants of  teff  production, the relationship of  
predictor variables with the household outcome variable, 
and the econometric model analysis in the study area. 

Descriptive statistics 
The study found that among 115 sample households, the 
number of  male-headed households and female headed 
households are found to be 82 and 33 in number and 
covered 71.3% and 28.70%, respectively. Out of  85 
producer households, 70 were male and 15 were female 
and out of  30 non-producer household, 12 were male 
and 18 were female.  Out of  115 sample households, the 
number of  illiterate household heads, grade 1-8, grade 
9-10, and grade 11-12 household heads are found to be 
47%, 20%, 18.3% and 14.8% respectively. Moreover out 
of  30 producer households, that 29 which covers 53.7% 
are illiterate, 1 which covers 4.3% are  grade1-8, 0 which 
covers 0% are grades 9-10 and 0 which covers 0% are 
grades 11-12.
From the total 115 sample households, 30 were 
technology adopters. From that, 27 which covered 90% 
were both adopters and producers but 3 which cover 
10% were adopters but non producers. And out of  the 
total sample household, 37 respondents were moderate 
adopters of  the technology. From that 36 which covered 
97.3% were both adopter and producers, but out of  
this 1 which covered adopters, 1 which covers 2.7% was 
adopter but not producer. From the total 115 households, 
70 respondents were getting extension service and both 
of  them are producers and there is no teff  producer that 
got extension access (see Table1 below). 
Out of  the total, 93.8% of  the respondents were growing 
local teff  variety which is recycled from year to year 
and those farmers who used the improved teff  variety 
were producers. The finding indicates a significant 
difference in teff  variety utilization between producer 
and non-producer groups at the 1 percent probability 
level of  significance. Consistently, out of  the total 115 

Table 1: Summary of  descriptive statistics for dummy/categorical explanatory variables
     Variable Catagories Producer % Non-producer  % x2 (p-value)
Sex Male 

Female 
85.4  
45.5  

14.6 
54.57           

19.43***

Level of  educational Illiterate
Grade 1-8
Grade 9-10
Grade 11-12       

46.3 
95.7
100
100

53.7
4.3
 0
0

40.409***

Technology adoption Adopter 
Moderatly adopter
Low adopter 
Non-adopter  

90
97.3
100
3.7

10
2.7
0
96.3

90.957***

Access to extension 
services

Yes 
No 

100
33.3

066.7 63.37***

Teff  variety Local
Improved 
Both 

  93.8
  100
  100

6.2
0
0

1.053***

Fertilizer application Yes    
No

98.8
17.1

1.2
82.9

84.095***

Note: *** Significant at 1% probability level, Source: own survey results 2021

https://journals.e-palli.com/home/index.php/ajsts


Pa
ge

 
40

https://journals.e-palli.com/home/index.php/ajsts

Am. J. Smart. Technol. Solutions 2(1) 38-42, 2023

households, 98.8% of  smallholder farm households were 
used fertilizer for their cropping purpose. The finding 
indicates a significant difference in fertilizer application 
between producer and non-producer groups at the 1 
percent probability level of  significance.
The land holding of  all sample households ranges from 
0 hectares to 5 hectares. The mean land sizes of  teff  
producer and non-producer households were 2.0176 
hectares and 0.3167 hectares, respectively. On average, 
the mean dependency ratios were 56.88% and 60.09% for 
teff  producers, non-producers, respectively. Moreover, on 

average 1.930 numbers of  oxen were used by household 
teff  producers. This means producer households had 
approximately 2 oxen on average and households who 
did not produce teff  were not having ox.  The study 
indicates that the average farm experience of  producer 
respondents in the study area was 9.86 years and there 
is a significance difference in teff  production experience 
between producer and non-producer respondents at 1% 
level of  significance showing that producer respondents 
have better teff  production experience than non-producer 
respondents in the study area (see Table 2 below).

Table 3: The maximum likelihood analysis of  logit model
Variable Coefficients Std. Err Sign. Level
Sex 0.055 0.032 0.084*
Farm land size 0.058 0.016  0.000***
Number of  oxen 0.142 0.021 0.000***
Teff  variety -0.056 0.014 0.000***
Education 0.03 0.013 0.021**
Technology adoption 0.045 0.017 0.007***
Fertilizer application   0.245 0.046 0.000***
Dependency ratio -0.031   0 .104 0.762
Extension service 0.072   0 .037 0.051*
FEXPTPPY 0.004 0.004 0.215
Constant  0.184 0.077 0.019
Dependent variable = Level of  teff  production R2=0.9421 Number of  observation =115
*** Significant at less than 1% probability level; ** Significant at less than 5% probability level; * Significant at less than 10% 
probability level. Source: Model output, 2021

Table 2: Summary of  descriptive statistics for continuous explanatory variables
    Variable Producer  Non-producer t-value

Mean Mean Total mean
Total farm land size 2.02 0.32 1.57 10.037***
Dependency ratio 0.57 0.6 0.58 1.276
Number  oxen 1.93 0.03 1.96 17.765***
Teff   production experience 9.86 0.77 7.49 2.909***
Note: refer to 1% significance level, Source: Own survey results, 2021

Determinants of  Households’ Teff  Production 
Before entering the variables in to the model, the 
multicollinearity problems were checked in terms of  variance 
inflation factor (VIF) for continuous and contingency 
coefficients for dummy variables, respectively. After testing 
the degree of  association of  independent variables, all 
explanatory variables were used for estimation. Binary logit 
model was applied to identify the major determinants of  
teff  production among hypothesized explanatory variables 
that are assumed to have an influence on the household’s 
level of  teff  production by using a statistical package 
known as STATA version 15 (see Table 3 below).
Based on the model result, a possible explanation for 
each significant independent variable is given as follows.
Sex of  household head: logit model analysis showed 
that there is a positive relationship between the sex of  
households and teff  production level at a probability 

level of  10%. It indicates that male-headed households 
produced more teff  than female-headed households. As 
the involvement of  male-headed households increased by 
one unit, the level of  teff  production increased by 0.055 
on average. Total cultivated land for teff  production per 
hectare: In line with our expectation, farm land size for 
teff  production is found to positively affect the level of  
teff  production at 1% level of  significance. It shows that 
households with more farm land size are more likely 
to produce teff  than those with a small land size. This 
is possible because when the farmland owned by the 
household is more, the level of  production and income 
become higher and eventually the amount or yield of  
teff  production increases. The result shows that as the 
cultivated land increased by one hectare, the level of  teff  
production to be produced increased by 0.058 units on 
average.

https://journals.e-palli.com/home/index.php/ajsts


Pa
ge

 
41

https://journals.e-palli.com/home/index.php/ajsts

Am. J. Smart. Technol. Solutions 2(1) 38-42, 2023

Number of  oxen 
This variable also has a positive effect on the level of  
teff  production at the probability of  1% significant 
level. This means farmers who own more quantity of  
oxen produced more output of  teff  than others. This is 
because; oxen ownership would help farmers to carry out 
agricultural operations like ploughing, sowing, and others 
on time that would improve productivity. The analysis 
shows that on average, as the number of  oxen increased 
by one unit, the amount of  teff  to be produced increases 
by 0.142 units. A previous study by (Gebremedhin et al., 
2007) found a similar result. 

Education level of  household 
The level of  education has a positive influence on the 
level of  teff  production at 5% significant level. It indicates 
that households led by non-literate heads are less likely to 
understand the modern farming technologies provided 
to them through any media (extension workers, radio, 
etc) than literate household heads. It affect production 
positively since it makes household’s to have ability to 
take good and well-informed production on teff. And 
the model shows that as the households’ education 
level increased by one grade level, the amount of  teff  
to be produced increased by 0.030 units. It is consistent 
with the study found by Amaza et al. (2006) and other 
literatures; the higher the educational level of  household 
head, the more teff  is expected to be produced.

Fertilizer application 
The results showed that fertilizer applications may affect 
teff  production positively at a probability level of  1%. It 
shows that if  fertilizer is available in the right amount and 
time, the level of  teff  production would be improved. The 
model analysis shows that as the utilization of  fertilizer 
increased by one unit, the level of  teff  production to be 
produced could be increased by 0.245 units on average. 
This result is similar to the study conducted by (Dickinson 
et al. 1990). 

Technology adoption
This variable is found to have a positive influence on teff  
production level at a probability level of  less than 1%. This 
means farmers who adopt farm technology like fertilizer, 
raw planting are more likely to produce more teff  than 
farmers who do not adopt it. The analysis shows that as 
farmers’ technology adoption level increased by one unit, 
the amount of  teff  to be produced would be increased by 
0.045 units on average.

Access to extension services
Access to extension services received by households 
has a significant positive association with level teff  
production status at a probability level of  10%. The 
positive relationship implies that when households get 
an extension service, the probability of  the household 
to produce teff  would be increased. On average, as 
producers’ access to the extension service increased by 

one unit, the production level of  teff  would be increased 
by 0.072 units. This result is consistent with the study 
found by (Babatunde, 2007).
 
CONCLUSION AND RECOMMENDATIONS
Identifying the major determinants of  smallholder 
farmers’ teff  production was the main purpose of  this 
study. To achieve this, primary data were collected from 
115 smallholder farm households. Descriptive statistics 
were used to explain the different socio-economic 
characteristics of  the sample households and inferential 
statistics were used to test the dummy and continuous 
variables. Logit regression model was used to identify 
the major determinants of  smallholder farmers’ teff  
production level.  The finding shows that the majorities 
(73.90% of  the sample households) were teff  producers 
and small numbers of  households (26.10% of  the sample 
households) were nonproducers of  teff. This indicates 
that more than 50% of  smallholder farmers are teff  
producers in the study area. The results of  the logistic 
regression model indicated that eight out of  ten variables, 
namely, sex of  the household, education of  household 
head, total cultivated land for teff  production per hectare, 
number of  oxen, technology adoption, and access of  
extension services received by households, fertilizer 
application and improved teff  variety used were found 
to be a major determinants of  household teff  production 
in the study area. Therefore, stakeholders should be 
considering these variables when smallholder farmers 
produce teff. 

REFERENCES 
Anderson, J. R. (2007). Agricultural advisory services of  

background paper for the world development report. 
World Bank, Washington, D.C., USA.

ATA (Agricultural Transformation Agency) (2013). 
New Teff  Technologies Demonstration Trials Draft 
Report. Addis Ababa, Ethiopia.

CSA (Central Statistical Agency). (2013). Agricultural 
Sample Survey: Area and Production of  Major Crops, 
Meher Season, 1, Addis Ababa, Ethiopia. 

Schmidt, E., & Kedir, M. (2009). Urbanization and spatial 
connectivity in Ethiopia: Urban growth analysis using 
GIS.

Spielman, D. J., D. K. Mekonnen, and D. Alemu. 
(2008). Seed, Fertilizer, and Agricultural Extension 
in Ethiopia. In Food and Agriculture in Ethiopia: 
Progress and Policy Challenges, edited by Paul 
Dorosh and Shahidur Rashid, Philadelphia: University 
of  Pennsylvania Press, 84–122.

Tareke Berhe and Nigusse Zena (2008). Results in a 
trial of  system of  teff  intensification:  Debre zeit, 
Ethiopia.

Gujarathi, Damodar M. (2004). Basic Econometrics. 
McGraw-hill.

Gebremedhin, Berhanu, Dirk Hoekstra, and Samson 
Jemaneh. (2007). Heading towards commercialization: 
The case of  live animal marketing in Ethiopia. IPMS 

https://journals.e-palli.com/home/index.php/ajsts


Pa
ge

 
42

https://journals.e-palli.com/home/index.php/ajsts

Am. J. Smart. Technol. Solutions 2(1) 38-42, 2023

Working Paper. 
Amaza, P. S., Umeh, J. C., Helsen, J., & Adejobi, A. O. 

(2006). Determinants and measurements of  food 
insecurity in Nigeria: some empirical policy guide 
(No. 1004-2016-78541).

Verstraete, Michel M., Bernard Pinty, and Robert 
E. Dickinson. (1990). A physical model of  the 
bidirectional reflectance of  vegetation canopies: 

Journal of  Geophysical Research: Atmospheres, 95, 11755-
11765.

Babatunde, R. O., O. A. Omotesho, and O. S. Sholotan. 
(2007). Socio-economic characteristics and food 
security status of  farming households in Kwara State, 
North-Central Nigeria. Pakistan Journal of  Nutrition 
6(1), 49-58.

https://journals.e-palli.com/home/index.php/ajsts

