Determinants of technical efficiency of coffee production in Jimma Zone, Southwest Ethiopia

Abstract

Agriculture has been and continues to assume center stage in the economic policy of Ethiopia. Coffee is one of the most vital sources of income for many coffee growers and continues to be still the leading export commodity in the national economy of the country. Despite the economic use, the productivity and quality of the coffee are unsatisfactory. The adoption of coffee yield-enhancing technical efficiency is key to improving coffee productivity and quality. Therefore, this study aims to investigate determinants that influence the technical efficiency of coffee production in Jimma Zone, Southwest Ethiopia. The primary data was collected among 398 coffee growers in research locations during the 2020/21 season. Descriptive statistics and econometric methods were developed for the data analysis. The estimated average value of technical, allocative, and economic efficiencies was 82.63%, 78.35%, and 74.65% respectively, which shows the existence of inefficiency in coffee production in the study area. The findings of OLS regression indicated that technical inefficiency is affected by age, sex, education status, landholding, livestock holding, credit uses, the extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted. Coffee yield technical efficiency was associated with a significantly higher coffee yield and per capita annual income of coffee cultivators. Concerned bodies should give important attention to coffee yield enhancing technical efficiency which is base for improving production. The summary of this coffee production technical efficiency by policymakers and plan designers could bring better enhancement to the coffee cultivator in the study area.

1. Introduction

Agricultural development is one of the foremost powerful tools to finish extreme poorness, boost shared prosperity and guarantee food security [1]. Developing countries and Africa account for massive shares of value, employment, and exports. It takes the biggest share of the economies of most Sub-Saharan African countries, which contributes between 15 and 60% of their GDP and provides employment for quite two-thirds of their population. However, agriculture in the Social Security Administration remains dominated by the husbandman and subsistence sector [2]. Therefore, Agricultural policy in Africa is vital for the agricultural development of Africa's countries to extend the production and productivity of smallholder farmers. However, they have not achieved the required goals. The major agricultural production in Africa targeted solely improvement inputs and technology, therefore, additional efficiency analysis is required such as technical efficiency.

Agriculture is the bone of the Ethiopian economy [3]. The Ethiopian Government has created different efforts to attain food security and scale back poorness at the family level by increasing production and productivity. However, Ethiopian Agriculture is dominated by a subsistence and smallholder-oriented system [4]. The Ethiopian economy depends heavily on agriculture; a sector persistently played a leading part in employment creation, poverty and food insecurity alleviation, and export revenues. The world development indicators [5] show that the sector contributes to more than 70% of the total population and it is the central driver of Ethiopian economic growth. Consequently, the sector currently contributes over 36.3% of the national GDP, 90% of export revenue, 70% of foreign exchange earnings, 70% of raw materials to country industries, and 85% of the labor force [6]. In Ethiopia agriculture dominates a large portion of the population, income, foreign exchange, and job creation [7]. It is characterized by the developed economic policy of agricultural development lead industrialization. Hence, the production and efficiency levels of agriculture in Ethiopia are generally below the world's mean due to poor attention given to the sector, leads an unsatisfactory yield than the world's mean. Smallholder farmers in agriculture with the same inputs are farming different per hectare yields due to inefficient utilization of the resource, backward technologies, poor cultivating techniques, and infrastructure, lack of access to credit and extension services, and inappropriate introduction and implementation of the yield-enhancing technical efficiency in Ethiopia. The sector plays a central role in economic policy, poverty alleviation, and food availability to smallholder farmers. Poverty alleviation and food security can be mainly achieved by agricultural growth. Agricultural development is assumed to be the means for improving Millennium Development Goals by 2015, despite the truth that agriculture creates a spillover effect on the industry, service, and tourism sectors [8]. Consequently, agriculture has been the most important sector of the country's development lead industrialization strategy for many years [9].

Coffee is the second most worldwide traded cash crop commodity in terms of volume and with its importance next to crude oil and the most valuable traded agricultural cash crop commodity in history of the world. In the world more than 2.25 billion cups of coffee are consumed by consumer every day. Moreover, the most crucial cash crop coffee is world widely traded commodity with a predicted export of 18.4 billion dollars in 2014/15 [10]. Specifically, more than 90% of coffee cash crop production takes place in the least developed countries, whereas, consumption takes place mostly in industrialized economies. In the world, most crucial countries such as Brazil (32%–34%), Viet Nam (12%–13%), and Colombia (8%–9%) together produced more than 55% of the world's coffee cash crop up to recent years. In 2010 world total coffee cash crop cultivating sector creates about 26 million people's employment in 52 coffee-producing countries [11,12]. In the world around 125 million people depend on coffee for their livelihoods [10]. Specifically, coffee is not only a crucial contributor to foreign exchange earnings but it also contributes for a sizable proportion of tax revenues and gross domestic product of many countries. During the 2005–2010 coffee cropping period, the average total export earnings share of eight coffee-producing countries exceeded 10% [11]. Coffee is very crucial in global trade and it contributes very important worth to the national economies of many least-developed countries.

Arabica and Robusta have mostly planted coffee species in Ethiopia. Specifically, Ethiopia is assumed to be the birthplace of coffee in Arabica. The country is the center of the origin, diversification, and dissemination of coffee plants which is good potential to increase coffee productivity and quality. Ethiopia is the leading Arabica coffee producer in Africa, the fifth coffee producer in the world, and the tenth coffee export country in the world. The mean annual coffee yield is 229,351.3 tons and the mean production of coffee is 0.71 tons/ha. The country has very huge potential to improve coffee yield endowed with suitable elevation, good temperature, recommended soil fertility, quality of planting materials, and sufficient rainfall in coffee cultivating belts of the country [13]. Despite its important contribution to the Ethiopian economy, the coffee yield sector is facing various challenges specifically, low productivity and quality [14]. The low status of coffee production is the result of several demographic, environmental, socio-economic, institutional, and input or farm management challenges. Hence, the inability of the grower to adopt good agronomic practices such as weeding, mulching, pruning, use of a variety of coffee seeds, fertilizers, and soil erosion control can also threaten coffee production [[14], [15], [16]]. With the good potential for improving coffee production, the mean/ha yield remains unsatisfactory at 0.71 tons/ha. Coffee in Ethiopia accounts for 15% of the country's yield directly or indirectly deriving their well-being from the coffee cash crop. Coffee plays a central role in the country's economy and accounts for more than 25% of the national GDP, 40% of total export earnings, 60% of agricultural export, 10% of government revenue, and 25% of the total population of the country. Worldwide Ethiopia is not only a cultivator and exporter of coffee but also the second largest consumer in the coffee cultivating countries next to Brazil [11]. The coffee price and share of the market influence coffee productivity and quality since low price and poor market share reduces coffee growers' yield and efficiency.

Therefore, technical efficiency is important to indicate growers are efficient in the employ of the existing economic resource and the decision to conduct the new cultivating agricultural technologies [17]. Hence, some research concerning new and improved agricultural technology is highly focused on factors that influence agricultural technology adoption decisions, but it is not complementary to its determinants of technical efficiency implementation [[18], [19], [20]]. Therefore, it is very difficult to have a clear understanding of the adoption of improved new agricultural technology and its technical efficiency. The study was evaluated by Ref. [21], using a stochastic production frontier model on the topic of technical efficiency of the yield of growers in the Garawa district. According to his study, the average technical efficiency was 81.5%, and production was significantly affected by agricultural technology and its technical efficiency. Specifically, technical efficiency was significantly affected by demographic, environmental, socio-economic, institutional, and input variables. A similar study by Ref. [22], conducted using the stochastic production frontier model the findings indicated that different variables such as demographic, environmental, socio-economic, institutional, and input variables affect technical efficiency.

The improvement in agricultural yield is not only affected by yield-enhancing agricultural technology adoption decisions but also determined by the cultivator's technical efficiency [23]. Technical efficiency denotes the cultivator's capability to achieve the optimum output using the existing determinants of yield. Additionally, in an input-poor country like Ethiopia, where the rate of the agricultural yield-enhancing technology adoption decision is relatively low, technical efficiency enhancements in the agricultural sector are assumed to be more advantageous [24]. In the country, improving the total yield and productivity is a necessity and the most important concern in their plan and policies. Coffee yield and productivity can be enhanced by using recommended inputs, and advancements in technology and technical efficiency of coffee growers [17]. Improving technical efficiency in yield allows growers to improve their yield without any additional inputs and advanced yield technologies [25]. There is different literature focusing on factors affecting technical efficiency [18,21,26]. This empirical literature focused on the study of the technical efficiency of cereal crops and thus, with little attention to the analysis of the technical efficiency of coffee cash crop production in Ethiopia.

However, there is little research evidence regarding the potential of technical efficiency on coffee productivity [24,27]. In particular, the determinants of technical efficiency and the sources of technical inefficiency of small-scale coffee growers have not been explored in detail. Their studies specifically, technical efficiency is significant in coffee productivity. Different technical efficiency studies were undertaken in less developed countries and many parts of Ethiopia [17,18,[21], [22], [23],28,29]. However, most of these studies were limited in dealing with identifying the factors affecting the level of technical efficiency, a measure of coffee production, and sources of technical inefficiency of coffee cultivators. To this end, the current study was conducted to investigate determinants that influence the technical efficiency of coffee production in Jimma Zone, Southwest Ethiopia. Specifically, the objectives of the study were to investigate factors affecting the level of technical efficiency and identify sources of technical inefficiency of coffee growers in the study area.

The study estimation strategy was guided by the conceptual framework. This conceptual framework was developed and modified based on the empirical literature [30]. The conceptual framework conducted in (Fig. 1), indicated that coffee growers' characteristics such as age, sex, educational status, size of family, family labor, and off-farm activity; institutional factors like access to credit, infrastructure, and extension; farm-level characteristics like landholding and ownership of land; input variables like livestock holding, fertilizer, seed, and variety of coffee planted are some crucial determinants that influencing coffee cash crop yield technical efficiency. Technical efficiency is very key in terms of increasing coffee grower yield. The developed conceptual framework indicated that important determinants and their relationships with each other are expected to influence the technical efficiency of the coffee cash crop. The conceptual framework is presented in (Fig. 1).

Fig. 1.

Conceptual framework.

2. Research methodology

2.1. Description of the study area

This study was developed in the Jimma Zone, located in the Oromia regional state of Ethiopia. Jimma Zone is laid 333 km to the Southwest of Addis Ababa, which is the capital city of Ethiopia. Jimma Zone is the oldest Zone in Ethiopian history. According to the [31] total population of the Jimma Zone is 2,773,730 (100%), of which 1,382,460 are male and 1,391,270 are female, and 2,432,562 of the population are rural dwellers. The Jimma Zone is located between 7°13′ to 8°56′N latitude and 35°49 to 38°38′E longitudes. The total land elevation in the Jimma Zone ranges from 880 m asl to 3,340 m asl. The average annual rainfall varies from 1,200 mm to 2,500 mm and has a mean annual temperature lies between 20 °C and 25 °C. Agro-ecologically, the Jimma zone is classified into three zones: highland (35%), midland (47%), and lowland (18%). The largest part of the Jimma Zone is midland and the Zone is suitable for agricultural production. The total area of the Jimma Zone is 1.1 million hectares, of which 0.1 million hectares of land is covered by coffee cash crops. Coffee is a vital contributor to the socio-economic well-being of coffee growers in Jimma Zone. The majority population of the Jimma Zone is young coffee growers. The Zone was a suitable Zone for cultivating coffee for many reasons. Firstly, the Zone with a high potential for coffee yields. Secondly, the technical efficiency application has been expanded and implemented for coffee production. Thirdly, widely applicable extensions and recommendations on coffee growing technical efficiency were conducted in the Jimma Zone.

2.2. Sampling technique

For the study, multi-stage sampling methods were developed to select sample coffee growers. In the first stage: the Jimma Zone was purposely selected based on agro ecology, the potential of coffee yield, and the application and introduction of technical efficiency of the coffee cash crop. Jimma Zone is better in terms of coffee production, application of coffee yield enhanced new agricultural technology, and practices coffee yield technical efficiency than the remaining Zone in the Oromia regional state. This is a rationale and an important point of why the research location was selected regarding coffee yield and influencing factors. In the second stage: among the total coffee-growing districts in the Jimma Zone, five districts namely Manna, Limmu Saka, Limmu Kosa, Gera, and Gomma were randomly selected based on the coffee cash crop yield. Thirdly, the total number of coffee growers in the yield year 2020/21 was identified. The total number of coffee cultivators in the five selected districts was 116,466 which are stratified by employing technical efficiency status. The sample size was determined based on the formula given by Ref. [32]. Accordingly, a total of 398 coffee cash crop growers were selected for the field survey data during the 2020/21 cropping season. The coffee grower is an adopter of technical efficiency with innovation from initial that coffee grower becomes aware innovation to the emblements to apply technical efficiency.

$$\mathrm{n}=\frac{\mathrm{N}}{1+\mathrm{N}\left(\mathrm{e}2\right)}=\frac{\mathrm{116,466}}{1+\mathrm{116,466}\left(0.{05}^2\right)}=398$$

A total number of 398 coffee growers were selected from each stratum using proportionate selecting procedures.

$${\mathrm{n}}_{\mathrm{i}}=\frac{\mathrm{N}\mathrm{i}}{\mathrm{N}}\left(\mathrm{n}\right)$$

Where, n_i is the total number of selected samples from each ith selected districts; N_i is the total number of headed households from ith selected districts; N is the total number of headed households in the selected districts; e is an acceptable error margin, and n is a total sample size. Finally, a total number of 398 coffee growers were selected from five districts by employing a simple random sampling method in the (Table 1).

Table 1.

Sample of coffee grower based on the level of technical efficiency.

Selected District	Total number of the coffee grower (Ni) from each ith district	A total sample size of the coffee grower (ni) from each ith district
Manna	30,540	104
Limmu Saka	14, 804	51
Limmu Kosa	24,893	85
Gera	16,229	55
Gomma	30,000	103
Total	116,466	398

Note: n_i = total sample size of the coffee cultivator from each ith district (i = 1, 2, 3, 4, 5); N_i = total number of the coffee cultivator from each ith district (i = 1, 2, 3, 4, 5).

2.3. Types and sources of data

In this study, descriptive and econometric data analyses were developed. Primary and secondary data sets as well as both qualitative and quantitative primary data were developed for the study. The primary data sets were collected including coffee grower environmental, demographic, institutional, and inputs characteristics and adoption decisions of technical efficiency. Before the field study, the instrument was rigorously reviewed and necessary changes were made. The questionnaires were administered in 398 coffee grower-headed households in Jimma Zone, Southwest Ethiopia. The structural questionnaires' employed were prepared to contain questions on coffee outputs, prices of coffee yield, quantities inputs, and all environmental, demographic, and institutional factors that influence the coffee grower's technical efficiency. Both open and close-ended questionnaires were conducted to achieve all objectives of the study. Primary data was prepared from February to June 2020/21 coffee growing seasons. Suitable and reliable persons were contacted to respond to the questionnaire that has a better understanding of the study. Classically, the questionnaires were distributed and collected at a later date after completion. The supplementary data such as secondary data sets were collected from published and unpublished sources, agricultural and rural development offices, the internet, empirical literature, rural coffee cultivators, and non – cultivators. The study was conducted through cross-sectional field survey data of the 2020/21 main coffee growing season.

The study was undertaken after the approval of the institutional research ethical review committee of the college of business and economics, Jimma University. Official permission to conduct the study was requested via an official letter from the college of business and economics, Jimma University, and granted by local authorities of the selected districts. All respondents were given detailed information about the objective and purpose of the study and verbal consent was obtained from each respondent before the interview. All the information acquired through the study was kept confidential and run according to the ethical guidelines.

2.4. Data analysis

The data for the study were analyzed by using both descriptive and econometrics data analysis. Descriptive analysis is identifying coffee growers' environmental, demographic, institutional, and input characteristics. For the descriptive analysis frequency, percentages, averages, standard deviation, maximum values, minimum values, t-test, and χ2 were employed. Particularly, this study employs χ2 tests for examining relations between coffee-growing technical efficiency and qualitative determinants of technical efficiency. Additionally, a t-test should be employed for assessing associations between coffee growing technical efficiency and quantitative factors affecting technical efficiency. Furthermore, this study developed econometric methods to evaluate in-depth analysis. This study develops a stochastic production frontier model to examine factors influencing the coffee grower technical efficiency among coffee cultivating farmers [33]. As the study [33] independently proposed the stochastic production frontier model in the following equation form (Eq. (1)):

$$Y_i=f\left(X_{ij};\beta \right){\mathit{exp}}^{\left(v_i-{\mu }_i\right)}$$ (1)

where, $Y_i$ is the potential production level of the ith firm; $f\left(X_i;\beta \right)$ is a suitable function; $X_{ij}$ is the Vector of actual jth inputs used by the ith firm; β is a vector of parameters to be estimated, and $v_i$ is random variability in the production that cannot be influenced by the firm, and ${\mu }_i$ is a deviation from maximum potential output attributable to technical inefficiency of ith coffee grower or a non-random error term associated with the farm-specific factors which contribute to the ith the farm does not attain maximum efficiency. The symmetric error term $v_i$ captures the stochastic effects outside the coffee grower's control. The error term ${\mu }_i$ is a one-sided μ ≥ 0 efficiency component that captures technical inefficiency. The one-sided error can follow such distributions as half-normal, exponential, and gamma [33]. The two components $v_i$ and ${\mu }_i$ are also assumed to be independent of each other.

The study conducted by Ref. [33], on coffee growers' technical efficiency was significant in production. This presents that users of coffee growing technical efficiency enhance production more than non–users. This study addresses evaluating the effect of coffee growers' technical efficiency on yield which is crucial in measuring coffee growers' food security. The use of coffee growers' technical efficiency and growers' food security are positively related. This indicates that any change in coffee growers' technical efficiency brings a change in growers' yields and food security. To assess factors affecting coffee grower technical efficiency among coffee cultivating farmers, the analysis model to be developed shall take the following form:

$$\ln \left(\mathrm{Y}\mathrm{i}\right)={\mathrm{X}}_{\mathrm{i}}\mathrm{B}\mathrm{i}+\mathrm{v}\mathrm{i}-\mathrm{\mu }\mathrm{i}$$ (2)

$$\mathrm{i}=1,\mathrm{2,3},4,\dots \dots \dots \dots \dots ,\mathrm{N}$$

where, ln is a natural logarithm of the coffee grower; i is an ith farm in the sample; Y_i is an observed coffee output of the ith sample farmer; β is a vector of unknown parameters to be estimated; X_i is a vector of covariates evaluating environmental, demographic, institutional, and inputs characteristics that are assumed to affect coffee grower farmer i (Table 2); ${\mathrm{u}}_{\mathrm{i}}$ is a stochastic term of the method which is assumed to be independently and identically distributed as ${\mathrm{u}}_{\mathrm{i}}\sim \mathrm{N}\left(0,{\mathrm{\sigma }}^2\right)$ The stochastic production frontier model is mostly conducted model to estimate coffee grower technical efficiency given by (Eq. (2)). Various equation forms have been conducted to evaluate the association between input and coffee output. Most of them are Cobb – Douglas and the trans-log equation. The current research was conducted by applying Cobb – Douglas production equation with the log-likelihood test. According to the [28,34,35], the coffee growing yield technical efficiency is interpreted in terms of observed output to the corresponding frontier output applying the given constant technology will take the below form (Eq. (3)):

$${\mathrm{T}\mathrm{E}}_{\mathrm{i}}=\frac{\mathrm{Y}\mathrm{i}}{{\mathrm{Y}\mathrm{i}}^{*}}=\frac{\mathrm{A}\mathrm{c}\mathrm{t}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}}{\mathrm{P}\mathrm{o}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}}=\frac{\exp \left(\mathrm{X}\mathrm{i}\mathrm{B}-\mathrm{u}\mathrm{i}\right)}{\exp \mathrm{X}\mathrm{i}\mathrm{B}}=\exp (-\mathrm{u}\mathrm{i})$$ (3)

technical efficiency takes the value on the interval (0, 1), where 1 reveals an optimum efficient farm. The yield gap of the ith coffee grower in coffee yield is the difference between potential yield and actual yield i.e., YG_i = PG_i + AY_i. Economic efficiency is the ratio of minimum conducted total production cost (C*) to actual total production cost (C) i.e., EE_i = C_i/C_i*. The allocative efficiency index can be evaluated as the ratio of economic efficiency to technical efficiency i.e., AE_i = EE_i/TE_i. To evaluate the determinants that affect the technical inefficiency of coffee production, the empirical estimation methods of data analysis were developed. The stochastic production frontier model is a crucial model to evaluate this effect on coffee production. The Cobb – Douglas yield-enhancing functional form employed is specified as follows [34,36,37]. Moreover, the Cobb – Douglas yield-enhancing function have the following (Eq. (4)).

$${\ln (\mathrm{Y}}_{\mathrm{i}})={\mathrm{B}}_{\mathrm{O}}+{\mathrm{B}}_1\ln \left(\mathrm{a}\mathrm{g}\mathrm{e}\right)+{\mathrm{B}}_2\ln \left(\mathrm{s}\mathrm{e}\mathrm{x}\right)+{\mathrm{B}}_3\ln \left(\mathrm{f}\mathrm{a}\mathrm{s}\right)+{\mathrm{B}}_4\ln \left(\mathrm{e}\mathrm{d}\mathrm{u}\right)+{\mathrm{B}}_5\ln \left(\mathrm{l}\mathrm{a}\mathrm{h}\right)+{\mathrm{B}}_6\ln \left(\mathrm{o}\mathrm{f}\mathrm{a}\right)+{\mathrm{B}}_7\ln \left(\mathrm{f}\mathrm{a}\mathrm{l}\right)+$$

$${\mathrm{B}}_8\ln \left(\mathrm{c}\mathrm{r}\mathrm{u}\right)+{\mathrm{B}}_9\ln \left(\mathrm{e}\mathrm{x}\mathrm{s}\right)+{\mathrm{B}}_{10}\ln \left(\mathrm{a}\mathrm{c}\mathrm{i}\right)+{\mathrm{B}}_{11}\ln \left(\mathrm{f}\mathrm{e}\mathrm{u}\right)+{\mathrm{B}}_{12}\ln \left(\mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\right)+{\mathrm{B}}_{13}\ln \left(\mathrm{l}\mathrm{i}\mathrm{o}\right)+$$

$${\mathrm{B}}_{14}\ln \left(\mathrm{l}\mathrm{a}\mathrm{o}\right)+{\mathrm{B}}_{15}\ln \left(\mathrm{v}\mathrm{c}\mathrm{p}\right)+\mathrm{v}\mathrm{i}-\mathrm{\mu }\mathrm{i}$$ (4)

Table 2.

Variables summary of technical inefficiency.

S. No	Variable name	Variable type	Variable description and its measurement	Expected sign
	Dependent variable
	Technical inefficiency	Continuous	Stochastic yield frontier model
	Independent variable
1	Age	Continuous	In years	–
2	Sex	Dummy	If 1 = Male and 0 = Female	–
3	Size of family	Continuous	In numbers	–
4	Education status	Categorical dummy	During the year of school	–
5	Landholding	Continuous	In hectares	–
6	Off-farm activity	Continuous	In Birr	–
7	Family labor	Continuous	If 1 = Yes and 0 = otherwise	–
8	Credit use	Dummy	If 1 = Yes and 0 = otherwise	–
9	Extension service	Dummy	If 1 = Yes and 0 = otherwise	–
10	Access to infrastructure	Dummy	If 1 = Having and 0 = otherwise	–
11	Fertility use	Continuous	Kilograms/hectares	–
12	Seed	Continuous	Kilograms/hectares	–
13	Livestock ownerships	Continuous	TLU	–
14	Land ownership	Dummy	If 1 = landownership and 0 otherwise	–
15	Variety of coffee planted	Dummy	If 1 = using high-yielding variety and 0 otherwise	–

Source: Authors' hypothesis 2020/21

Output is the total yield of coffee cultivated measured in kg/ha. β is unknown yield equation parameters, $v_i$ is the disturbance error term, independently distributed as N (0, σv2), and μi is a non-negative random variable, identically distributed as N (μ, σμ2). The stochastic cost frontier model is formulated as the (Eq. (5)):

$$\ln \left(\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\right)={\mathrm{\alpha }}_{\mathrm{O}}+{\mathrm{\alpha }}_1\ln \left(\mathrm{a}\mathrm{g}\mathrm{e}\right)+{\mathrm{\alpha }}_2\ln \left(\mathrm{s}\mathrm{e}\mathrm{x}\right)+{\mathrm{\alpha }}_3\ln \left(\mathrm{f}\mathrm{a}\mathrm{s}\right)+{\mathrm{\alpha }}_4\ln \left(\mathrm{e}\mathrm{d}\mathrm{u}\right)+{\mathrm{\alpha }}_5\ln \left(\mathrm{l}\mathrm{a}\mathrm{h}\right)+{\mathrm{\alpha }}_6\ln \left(\mathrm{o}\mathrm{f}\mathrm{a}\right)+$$

$${\mathrm{\alpha }}_7\ln \left(\mathrm{f}\mathrm{a}\mathrm{l}\right)+{\mathrm{\alpha }}_8\ln \left(\mathrm{c}\mathrm{r}\mathrm{u}\right)+{\mathrm{\alpha }}_9\ln \left(\mathrm{e}\mathrm{x}\mathrm{s}\right)+{\mathrm{\alpha }}_{10}\ln \left(\mathrm{a}\mathrm{c}\mathrm{i}\right)+{\mathrm{\alpha }}_{11}\ln \left(\mathrm{f}\mathrm{e}\mathrm{u}\right)+{\mathrm{\alpha }}_{12}\ln \left(\mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\right)+{\mathrm{\alpha }}_{13}\ln \left(\mathrm{l}\mathrm{i}\mathrm{o}\right)$$

$$+{\mathrm{\alpha }}_{14}\ln \left(\mathrm{l}\mathrm{a}\mathrm{o}\right)+{\mathrm{\alpha }}_{15}\ln \left(\mathrm{v}\mathrm{c}\mathrm{p}\right)+\mathrm{v}\mathrm{i}-\mathrm{\mu }\mathrm{i}$$ (5)

Cost is the TC of inputs spent to produce coffee, the seed is TC of coffee seed, feu is TC of fertilizer use, lah is TC of the rental value of land, fal is TC of the available labor force, vcp is TC of variety of coffee planted, and α is unknown cost equation parameters. Determinants of inefficiency are evaluated by applying Ordinary Least Squares. This function was developed to measure the key determinates that affected the technical inefficiency of coffee growers in the study area. The inefficiency equation is specified as following (Eq. (6)):

$$\mathrm{\mu }\mathrm{i}={\mathrm{\alpha }}_{\mathrm{O}}+{\mathrm{\alpha }}_1\mathrm{a}\mathrm{g}\mathrm{e}+{\mathrm{\alpha }}_2\mathrm{s}\mathrm{e}\mathrm{x}+{\mathrm{\alpha }}_3\mathrm{f}\mathrm{a}\mathrm{s}+{\mathrm{\alpha }}_4\mathrm{e}\mathrm{d}\mathrm{u}+{\mathrm{\alpha }}_5\mathrm{l}\mathrm{a}\mathrm{h}+{\mathrm{\alpha }}_6\mathrm{o}\mathrm{f}\mathrm{a}+{\mathrm{\alpha }}_7\mathrm{f}\mathrm{a}\mathrm{l}+{\mathrm{\alpha }}_8\mathrm{c}\mathrm{r}\mathrm{u}+{\mathrm{\alpha }}_9\mathrm{e}\mathrm{x}\mathrm{s}+{\mathrm{\alpha }}_{10}\mathrm{a}\mathrm{c}\mathrm{i}+{\mathrm{\alpha }}_{11}\mathrm{f}\mathrm{e}\mathrm{u}$$

$$+{\mathrm{\alpha }}_{12}\mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}+{\mathrm{\alpha }}_{13}\mathrm{l}\mathrm{i}\mathrm{o}+{\mathrm{\alpha }}_{14}\mathrm{l}\mathrm{a}\mathrm{o}+{\mathrm{\alpha }}_{15}\mathrm{v}\mathrm{c}\mathrm{p}+{\mathrm{\xi }}_{\mathrm{i}}$$ (6)

where, i is the ith coffee growers; μ_i is a technical inefficiency score; δ_i is a vector of the parameter to be estimated; ${\mathrm{\xi }}_{\mathrm{i}}$ is an error term; age, size of family, land holding, off-farm activity, available family labor, fertilizer use, seed, and livestock ownerships are continuous explanatory variables, while sex, access to credit, extension use, access to infrastructure, land ownership, and variety of coffee planted are dummy explanatory variables and education is a categorical dummy variable for the study. The above-listed explanatory variables are environmental, demographic, institutional, and input characteristics that are intended to evaluate the technical inefficiency of ith coffee growers. For the study dependent variable is the technical inefficiency score and all the above demographic, environmental, institutional, and inputs variables are explanatory variables. The explanatory variables summary of technical inefficiencies such as variable name, type, description, measurement, and expectation are presented below (Table 2).

3. Results and discussions

3.1. Descriptive analysis

The primary objective of this research is to investigate technical efficiency and evaluate the determinants that affect it among coffee cultivators in the selected district of Jimma Zone (Table 3). indicates the descriptive summary statistics of the coffee cultivator by type of coffee growing yield of technical efficiency (i.e., achieving technical efficiency status). Out of a total of 398 (100%), about 235 (59%) of the coffee cultivators' are a technically inefficient method of growing coffee cash crops, which was relatively larger than those who did 163 (41%) during the 2020/21 cultivating season.

Table 3.

Sample coffee growers by technical efficiency status.

Technical efficiency status	Frequency	Percent	Cumm. percent
Technical inefficiency	235	59	59
Technical efficiency	163	41	100
Total	387	100

Source: Computed from own survey data 2020/21

According to the sample respondents of coffee growers, there is high coffee cash crop technical inefficiency due to low interest to grow coffee, topography not suitable of planted land due to shortage of availably family labor, poor access to infrastructure, low credit access, weak fertilizer distribution, and logging water. To achieve coffee cash crop technical efficiency applying the modern agricultural technique of coffee crop is essential. The time-consuming practice of the cultivating method of coffee cash crop was among the major reasons found to face coffee yield technical inefficiency. Additionally, some of them mention that government needs to consider distributing of achieving coffee yield technical efficiency machines to substitute the labor force with machines and to save time for coffee growing.

According to (Table 4), the study reveals the descriptive summary statistics means and standard deviations for a major independent variable by achievement status. This study conducted t-values and χ2 values indicate the evaluation of averages of these explanatory variables across the technical efficiency and inefficiency categories of the coffee grower. According to the summary statistics, the majority of coffee growers were headed by males (76.63%), growers relatively older (56.21 years average age), literate (59.97% of whom are above primary education), on the average size of family (5.74), having own 12.11 TLU of livestock, and on average grow 2.72 ha of land. This finding is similar to the findings of [38]. As mentioned, (69.25%) of the coffee growers are extension service users, (56.12%) of the coffee growers had used credit and (77.52%) followed available for family labor. As to coffee growers, 10,410.47 Ethiopian Birr per year of non-farm activity had on average, on average (76.49%) of fertilizer utilized and (68.64%) of coffee growers had infrastructure on average. As presented in (Table 4), on average 80.50 kg/ha of seeds were utilized followed by on average (of 59.55%) ownership of land. This result is consistent with the results of [39].

Table 4.

Coffee growers (means) by achieving technical inefficiency status.

Variables	Total Sample	Efficiency	Inefficiency	Comparison	P-value
Technical inefficiency Indicators
Coffee crop yield (Quintal/year)	16.56	18.77	12.59	−10.76	0.000***
Household Characteristics
Sex of coffee grower (1 = male)	76.36%	78.91%	28.46%	0.61	0.760
Extension service (1 = yes)	69.25%	89.18%	54.32%	27.54	0.000***
Credit use (1 = yes)	56.12%	71.71%	46.13%	17.96	0.000***
Family labor (1 = yes)	77.52%	82.54%	73.79%	6.00	0.312
Land ownership (1 = yes)	59.55%	62.47%	58.78%	0.65	0.856
Educational status (1 = literate)	59.97%	96.06%	32.63%	37.47	0.000***
Age coffee grower (years)	56.21	52.27	59.05	30.98	0.622
Family-size coffee grower (number)	5.74	6.45	5.29	48.20	0.000***
Landholding of the coffee grower (ha)	2.72	2.86	2.53	34.67	0.000***
Livestock ownerships (TLU)	12.11	12.69	11.18	50.76	0.000***
Off–farm activity (Birr)	10,410.47	11,913.83	9114.15	3757.86	0.000***
Fertilizer use	76.49	89.29	59.23	33.35	0.000***
Access to infrastructure	68.64%	88.63%	55.14%	0.55	0.000***
Seed	80.50	74.79	82.54	38.83	0.889
Variety of coffee planted	70.75%	90.74%	67.25	0.53	0.000***
Total observations	398 (100%)	235 (59%)	163 (41%)

Source: Computed from own survey data 2020/21. t-values developed to measure continuous independent variables; Pearson's χ2 values developed to measure categorical dummy and dummy independent variables.

Additionally, crucial significant variations were addressed among technical efficiency and inefficiency in terms of coffee grower characteristics. Accordingly, coffee growers with technical efficiency had better educational status than those with technical inefficiency. This suggests that the education status of coffee growers might be negatively related to technical inefficiency, which is the similar result with the study results of [28]. Similarly, coffee growers with technical efficiency had significantly larger family sizes than coffee growers with technical inefficiency. This result is consistent with the results of [21]. Besides, efficient growers produced larger landholding and had more livestock units than their inefficient counterparts showing the crucial of coffee grower asset ownership in the adoption of technical efficiency decisions. Similarly, with study results of [17]. Technically efficient and inefficient groups significantly varied in terms of access to the infrastructure whereas efficient had better access to infrastructure than their inefficient counterparts. The technically efficient coffee grower had better on average in terms of extension services, credit use, landholding, and land ownership than counterparts. Similarly, with studies results of [29,40]. Regarding off-farm activity and fertilizer use, efficient coffee growers on average are better than their inefficient counterparts, which is in line with study results [40]. Furtherly, there were no significant differences in terms of age and sex of coffee growers suggesting a lack of relation between technically efficient and inefficient. However, there was no variation in terms of availability of labor, land ownership, and seed between technically efficient and inefficient. A coffee grower's technical inefficiency is negatively related to education status, size of family, producing land holding, ownership of livestock, extension and credit use, infrastructure, off-farm activity, fertilizer utilization, and variety of coffee planted. Any better variation in the above-listed explanatory variable brings a better variation of the technical efficiency of coffee cash crop production than inefficiency. This result is similar to the results of [21,24,28].

3.2. Econometric results

According to the maximum likelihood estimation of Cobb – Douglas coffee production function, out of fifteen explanatory variables in function eleven variables such as age, sex, education status, landholding, livestock holding, credit uses, extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted affect coffee yield among coffee growers. Among these significant explanatory variables like age, education status, livestock holding, credit uses, extension uses, land ownerships, seed, and variety of coffee planted influence coffee yield at a 1% probability significance level. Coffee yield is also affected by sex, landholding, and off-farm activity at a 5% probability significance level. The yield elasticity concerning age, sex, education status, landholding, livestock holding, credit uses, the extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted indicate that as these variables increase, the coffee yield will enhance. Moreover, sex is significant at a 5% probability level. Specifically, this result presented that the male coffee grower is more associated with coffee production than the female counterparts in the study area. Coffee grower increases the age of cultivating coffee, education status, livestock holding, credit uses, extension uses, land ownerships, seed, and variety of coffee planted on average for the yield of coffee by 1%, they can enhance the level of coffee yield by 9.65%, 28.35%, 25.83%, 29.44%, 27.17%, 18.41%, 23.76%, and 27.55% respectively. This finding is in line with the findings of [28,29,40], who conducted their studies on technical efficiency. This indicated that there is good potential for coffee yield in the study area.

The stochastic production frontier model assesses the factors affecting coffee growers' decisions to the technical inefficiency suggested in (Table 5). The goodness fit of the coffee grower concerning predictive inefficiency of technical inefficiency was high with 339 (85.17%) of the 398 (100%) coffee cultivator sample included in the model perfectly predicted.

Table 5.

Maximum-likelihood estimates of determinants of technical inefficiency (n = 398).

Variable	Robust Coef.	SEa	Z	P >|Z|	dy/dxc
Age of coffee grower	−0.123***	0.014	−1.71	0.001	0.0965
Sex of coffee grower	−0.126**	0.236	−0.15	0.031	0.0844
Size of family	−0.218	0.143	−1.21	0.307	0.0621
Education status	−0.462***	0.150	−1.13	0.006	0.2835
Landholding	−0.637**	0.324	−1.43	0.043	0.1754
Livestock holding	−0.341***	0.106	−1.76	0.005	0.2583
Credit uses	−0.543***	0.215	−1.87	0.003	0.2944
Family labor	−0.425	0.263	−0.81	0.703	0.2416
Extension uses	−0.130***	0.382	−1.72	0.007	0.2717
Off–farm activity	−0.268**	0.101	−1.18	0.021	0.1934
Land ownership	−0.115***	0.406	−1.53	0.004	0.1841
Fertilizer uses	−0.374	0.247	−1.63	0.176	0.1342
Access to infrastructure	−0.651	0.414	−1.18	0.628	0.2567
Seed	−0.403***	0.226	−1.48	0.000	0.2376
Variety of coffee planted	−0.524***	0.563	−1.76	0.002	0.2755
Constant term	3.364***	1.415	1.31	0.000	–

Source: Computed from own survey data 2020/21; Number of observations = 398; LR chi2 (15) = 59.57; Probability > chi2 = 0.0000; Log likelihood = −78.347; Pseudo R2 = 0.437; ***, **and * are 1%, 5%, and 10% statistically significant levels respectively

(Table 5) suggests that from a total of fifteen explanatory variables eleven explanatory variables such as age, sex, education status, landholding, livestock holding, credit uses, extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted were found to have significant relation with the level of technical inefficiency of coffee growing. Particularly, age was revealed to have a strong negative relationship with the technical inefficiency of coffee. Specifically, citrus paribus, an extra year of coffee grower age is expected to be found in a 9.65% decrease in the probability of teff coffee technical inefficiency at (P < 0.01). Furthermore, coffee growers who are on average 10 years older are suggested to be 96.5% less likely to coffee cash crop yield technical inefficiency than their counterparts, the variable is scientifically determining coffee yield technical inefficiency. This result is in line with the results of [41,42]. The main reason for this is coffee growers more skill full at elder age due to cumulative growing experiences. The ability, physical capacity, information, knowledge, and skills increase at an elder age of coffee producers.

The regression result was found to have variables such as age, sex, education status of the coffee grower, landholding of the coffee grower, livestock holding, off-farm activity, credit uses, extension uses, land ownerships, seed, and variety of coffee planted had all expected to have negative relations with coffee yield technical inefficiency and significantly influence coffee yield technical inefficiency. The marginal effect of all these significant variables ranges between 8.44% and 29.44% on average, other factors remain constant. Furthermore, age, sex, education status, landholding, livestock holding, credit uses, extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted were respective relations with a 9.65%, 8.44%, 28.35%, 17.54%, 25.83%, 29.44%, 27.17%, 19.34%, 18.41%, 23.76%, and 27.55% lower probability of coffee yield technical inefficiency on average, cetris paribus. Specifically, coffee growers with educational status, credit use, extension use, and variety of coffee planted were expected found to be 28.35%, 29.44%, 27.17%, and 27.55% respectively lower probability of coffee yield technical inefficiency than their counterparts. Credit is a key element of the coffee cultivating system in terms of satisfying coffee growers' needs by solving liquidity and working capital problems. Coffee growers who get more credit at a given cultivating season are expected to have less technical inefficiency than their counterparts. On the other hand, the family size of the coffee grower, availability of family labor, fertilizer use, and access to infrastructure didn't have any relation to coffee yield technical inefficiency in the study area. The results presented in both descriptive results in (Table 4) and econometrics results in (Table 5) were the availability of family labor statistically insignificant variable. In addition, in the descriptive portage of coffee grower age, sex, availability of family labor, land ownership, and the seed of the coffee cultivator was no significant correlation with coffee yield technical inefficiency. The results from econometrics presented that the family size of the coffee grower; availability of family labor, fertilizer use, and access to infrastructure were statistically insignificant variables. This finding is consistent with the findings of [21,26,43], conducted a study on technical efficiency.

The results of the efficiency found that the average value of technical, allocative, and economic efficiencies was 82.63%, 78.35%, and 74.65% respectively. According to (Table 6) average technical efficiency of 82.63% ranges from a minimum of 56.78% to a maximum of 98.59%, while an average allocative efficiency also ranges from a minimum of 49.35% to a maximum of 96.28% with an average of 78.34%. Finally, economic efficiency ranges from a minimum of 49.73% to a maximum of 89.46% with an average of 74.65%. The results found that coffee growers are relatively better in technical efficiency than allocative and economic efficiencies. As the summary statistics of efficiency, the result of technical efficiency presented that the coffee yield of coffee growers can be enhanced on average by about 30.48% if better evaluations are taken to enhance the level of efficiency of the coffee grower. These presented average coffee cultivators were to achieve the technical efficiency level of their most efficient counterparts. Average coffee growers could achieve a 16.18% cost savings. e., 1–8263/9859*100, and technical inefficient cost savings of coffee cultivators is 42.40%i.e., 1–5678/9859*100. This result is similar to the results of [29,44]. Coffee growers achieve better cost savings of technical efficiency than allocative and economic efficiencies.

Table 6.

Summary statistics of efficiency score of coffee growers.

Types of efficiency	Mean	Std. Dev.	Min.	Max.
Technical efficiency	0.8263	0.34	0.5678	0.9859
Allocative efficiency	0.7834	0.32	0.4935	0.9628
Economic efficiency	0.7465	0.08	0.4973	0.8946

Source: Computed from own survey data 2020/21

3.3. Limitations and future research directions

As with all studies, this study has its limitations. Firstly, this study may be limited in scope and depth as it only considered the technical efficiency of coffee productivity. Therefore, future studies should consider the technical efficiency of the other different cash crop productivity. Secondly, the study only was in Jimma Zone, Southwest Ethiopia. Hence, this study was unable to incorporate quantitative data from other areas in Ethiopia. As a result, future scholars are advised to conduct this research by expanding the scope of the study area and conducting comparisons among different countries. Thirdly, Jimma Zone only due to quantitative data collection challenges during the first phase of the COVID-19 pandemic. Fourthly, the limitation was from the interaction of knowledge, strategy, and promoting research for technical efficiency of coffee productivity. Finally, to boost the technical efficiency of coffee productivity, optimizing the implementation action is very crucial.

4. Conclusions and policy implications

The agricultural sector is crucial in deriving sustainable economic development by enhancing productivity and efficiency in yield to alleviate poverty and food insecurity. The agriculture in the Jimma Zone is characterized by low coffee yield and technical efficiency. Technical efficiency is a key component to enhancing coffee yield. The general objective of this study aims to investigate determinants that influence the productivity and technical efficiency of coffee yield in Jimma Zone, Southwest Ethiopia. For this study, cross-sectional field survey data among 398 coffee growers during the 2020/21 coffee growing season were collected. This study used both primary and secondary data, as well as both qualitative and quantitative primary data sets. For the data analysis in this study, descriptive and econometric methods of data analysis were developed. In the data analysis, descriptive and Cobb-Douglas production function was conducted to investigate coffee productivity, a stochastic production frontier model was investigated to assess the technical efficiency of coffee, and OLS regression was developed to assess determinants influencing inefficiencies status. The finding evaluated that productivity was affected by age, sex, education status, landholding, livestock holding, credit uses, extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted. Results revealed technical efficiency was associated with significant improvements in household food security as reflected in significantly increased household per capita income. The estimated average value of technical, allocative, and economic efficiencies was 82.63%, 78.35%, and 74.65% respectively. Moreover, key coffee cultivators' characteristics such as age, sex, education status, landholding, livestock holding, credit uses, extension uses, off-farm activity, land ownerships, seed, and variety of coffee planted were found to be important factors affecting technical inefficiency in the study area.

Given these findings, several implications could emerge from the analysis upon which important suggestions could be made as key recommendations. Application of technical efficiency of coffee yield is relatively low in the Jimma Zone; coffee grower who adopted the technical efficiency should generally improve their welfare and farm productivity. Consequently, technical efficiency could be considered among the components of the agricultural improvement package implemented by local policymakers and actors as part of improving farmers' livelihoods in the study Zone. In particular, promoting technical efficiency practices in the Jimma Zone could help to achieve significant welfare and productivity gains thereby leading to better living standards among coffee-producing farm households in the study area. More importantly, the study results presenting the crucial factors underlying coffee growers' decision of reducing technical inefficiency should serve as a key input in designing a plan and making policies. For instance, educations have a strong association with the adoption of technical efficiency in coffee yield. To this end, strengthening rural coffee farmers' awareness/knowledge among farm households deserves attention for promoting the adoption of technical efficiency. This is besides the additional positive adoption-enhancing influence arising from access to extension services a separate effect from that attributable to better education. Consequently; extension programs could be focused on the less educated coffee grower through facilitating special training and technical support to improve the adoption rate of technical efficiency of coffee yield. Improved access to and provision of credits and extension services could also help achieve similar goals. To this end, the use of agricultural extension needs to consider recommended and improved agronomic practices. Extension use is particularly crucial in terms of improving the adoption of technical efficiency practices, which can, in turn, enhance coffee yield and subsequent improvements in household welfare. More specifically, appropriate inputs of coffee cultivation are vital to improving the coffee cultivator's technical efficiency. Hence, strengthening inputs like livestock holding, seed, and variety of coffee planted are very crucial recommended components of coffee cultivators which enhance the grower's technical efficiency. Concern bodies would create a conducive environment in education status, livestock holding, credit uses, extension uses, off-farm activity, seed use, and variety of coffee planted, thus, helping to enhance technical efficiency in coffee yield. This study summarizes the applications of technical efficiency by designing a plan and making policies that would bring appropriate improvement to coffee growers' income. Specifically, enhancing coffee yield technical efficiency is a key option to improve coffee growers' yield and income. Therefore, the agricultural sector and concerned bodies would give important attention to coffee production and its technical efficiency decision, which is a key indicator to alleviating poverty and food security.

Author contribution statement

Negese Tamirat: Conceived and designed the experiments; Performed the experiments; Contributed reagents, materials, analysis tools; Wrote the paper.

Sanait Tadele: Performed the experiments; Contributed reagents, materials, analysis tools; Wrote the paper.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

Data will be made available on request.

Declaration of competing interest

We declare no competing interests.