Technical Efficiency Analysis of Coffee Production

<b>Description of the data and file structure</b><i>Principal Investigator Contact Information</i>Name: Alemu Olika (MSC &amp; MA)Institution: Wollega University ^ Development Bank of Ethiopiaemail: [e-mail] / [e-mail]<i>Alternate Contact Information</i>Name: Gemechu Mulatu (PHD)Institution: Wollega UniversityEmail: [e-mail]<b>Dataset Overview</b>These data were collected from the coffee producers’ farmers in the study area. A few kebeles were chosen for the research. Before data collection, the following considerations were done:• The purpose &amp; the importance of the study were explained for the participants of the study. Then, the respondents were orally informed that they have the right to participate or not in the filling the questionnaire. Thus, the participants in the study were participating in the study by only filling out the questionnaire.• Oral communication was used to explain to the sample responders that the data-gathering procedures should not cause confusion &amp; harm participants. Clear and impartial preparation went into creating the questionnaire.This study did not use experimental subjects based on humans or animals. It was only a technical efficiency study of farmers’ coffee producing practices.The enumerators were trained in the data collection procedures. In the study, cross-sectional household data from the 2021 main harvest cropping seasons were used. Data for input (such as land, human labor, fertilizer, coffee plants, and herbicides) were used, and the output of coffee production was collected from a specified period of time. Data on input use and output were collected in local units and converted into standard units. In addition, primary data were collected by interviewing the selected coffee producers’ farmers and variables that cause variation in production efficiency, such as age, education, household size, extension contact, and gender. In addition, socioeconomic variables such as demographic data, credit access, livestock holdings, wealth indicators, and institutional data were collected. On the other hand, data related to coffee production trends, input supply, and extension services are gathered to clarify and support the analysis and interpretation of primary data.The questionnaire has been printed after it has been approved by the College of Business and Economics Research and Technology Transfer Associate Dean of Wollega University. The researcher personally visited the selected smallholder farmers at coffee bean collection and harvesting time and kindly encouraged them to fill out the questionnaire objectively without any biases.<b>Sources of Data</b>This data was prepared to study the technical in/efficiency of coffee production. Thus, as the primary data the data was collected from the selected farmers, those currently participate in coffee production. Therefore, it can desrcibed as both quantitative and qualitative data, as well as primary and secondary sources. The primary data were gathered using a structured questionnaire. In the collection of data, a structured questionnaire was developed and evaluated. The questionnaire was refined and modified based on the pre-test input. The primary data collection process was conducted by the enumerator, the district’s development agents, and the researcher. This data was also gathered from governmental and non-governmental institutions, published and unpublished documents, websites, and other relevant sources for analysis and descriptive purposes.<b>Dates of Data Collection</b>1. Primary data collection - 20212. Secondary data collection - 2021Approximately 31,610 farmers from 30 kebeles represented the district’s entire coffee-producing population. During the second phase, four kebeles belonging to the main coffee producers were purposively chosen from these kebeles because of their sizable coffee fields and the necessity of determining the districts’ most and least productive coffee-producing areas. There were 1108 people living in these four kebeles. The third stage was the random selection of 285 samples using the Kothari formula.Declaration of FundingThe authors did not receive support from any organization for the submitted work. No funding was received to assist with the preparation of this manuscript. No funding was received for conducting this study at all.<b>Human Subjects De-Identification Statement</b>To the greatest extent feasible, this data has undergone thorough de-identification. Names, addresses, email addresses, and other direct identifiers have all been permanently deleted, along with all other personally identifiable information (PII). Furthermore, even with the use of easily accessible information, the data has been processed to remove any chance of reasonably determining an individual’s identity.<b>Description of the data and file structure</b>The primary raw data from the chosen study area was entered into an Excel spreadsheet called “Raw_Data_s_of_Technical_efficiency_analysis_of_coffee_production” and shared to this data repository. Additionally, Description_and_Meanings_of_Variables: in this there is one dependent variable called total output and two types of explanatory variables. We used this 1st numbering (first independent variables) output as dependent variables and the rest five variables (landcoff = land where the matured coffee covered, laborcoff = labor force for coffee production, Coffplant = matured coffee plant those starting to give a coffee bean, orgfert = organic fertilizer and herbicides) as explanatory variables to measure the elasticity of coffee production in the study area.The second ((second numbering) second variables) set of 12 explanatory variables (ageofhhh = age of household head, sexofhhh = sex of household head, educofhhh =educational level of household head, hhsize = size of household, tlu =tropical livestock unit, offincome = off/non-farm income, totcultland = total cultivable land, totlandfrg = total land fragmentation, avrgplotdist = average plot distance, extcontact = extension contact service, train = training to farmer, credit = credit service for farmer) are factors that may contribute to technical inefficiency of coffee producers in the study area. The Description_and_Meanings_of_Variables xls sheet has a thorough description of each and describe unit of measurement applicable in the study was explained.<b>Word document uploaded as: Questionnairs_for_Technical_Efficiency_Analysis_of_Coffee_Production</b>These are the word documents that we utilized to get respondents’ raw data. It comprises the six components of the raw data collected from the respondents. Part I: General information about sample farmers; Part II: Economic information; Part III: General information about coffee farming; Part IV: Fertilizer &amp; Chemicals (Herbicides); Part V: Extension service and training; and Part VI: Credit service<b>Word document uploaded as: Sources_of_data_Sampling_Technique__and_Sample_Size</b>These documents provide data sources, sampling methodology, and sample size calculations. In general, to explain where the data were collected, the selection of sample household heads, and the calculation of sample size.<b>Figures:</b>Figure 1: Graph of Input-oriented measures for technical, allocative and economic efficienciesFigure 2: Graph Technical Allocative and Economic efficiency through output oriented measurementFigure 3: Sketch of Conceptual Framework of the StudyFigure 4: Map of the Study Area; location of the kebeles where the data was collectedFigure 5: Skewedness of Farmers Technical EfficiencyFiles and variablesFile: Questionnairs_for_Technical_Efficiency_Analysis_of_Coffee_Production.docx<b>Description:</b> The structured questionnaire designed to collect the primary data from respondent.File: Raw_Data_s_of_Technical_efficiency_analysis_of_coffee_production.xlsx<b>Description:</b> It contains the collected data in a suitable form for Stata software. In these there is one dependent variable and two sets of independent variables.<i>Variables</i>totoutput = total output, which is employed in the study as a dependent variable. The amount of coffee yield obtained in one season of the production year.(landcoff = land where the matured coffee covered, laborcoff = labor force for coffee production, Coffplant = matured coffee plant those starting to give a coffee bean, orgfert = organic fertilizer and herbicides) as independent variables to measure the elasticity of coffee production in the study area.The second variables set of 12 idependent variables (ageofhhh = age of household head, sexofhhh = sex of household head, educofhhh =educational level of household head, hhsize = size of household, tlu =tropical livestock unit, offincome = off/non-farm income, totcultland = total cultivable land, totlandfrg = total land fragmentation, avrgplotdist = average plot distance, extcontact = extension contact service, train = training to farmer, credit = credit service for farmer) are factors that may contribute to technical inefficiency of coffee producers in the study area.File: Figure*1 *Graph_of_Input_Oriented_measures_for_Technical_Allocative_and_Economic_Efficiency111.tiff<b>Description:</b> The graph of Input-oriented measures of a farmer’s efficiency that use two inputs (Z1 and Z2) to produce a single output under the assumption of continuous return to scale are depicted in Figure 1 Z1 and Z2, the two inputs, are displayed on the vertical and horizontal axes, respectively. KK’ is an isoquant of a completely effective company. Each point on this isoquant represents technically efficient manufacturing. Let’s say a company is operating at point X in Figure 1, producing at the same rate as the fully efficient farmers.File: Figure_2_Technical_Allocative_and_Economic_efficiency_through_output_oriented_measurement<b>Description:</b> The Graph The output changes that a company may accomplish with the same amount of inputs are the main emphasis of output-oriented measurements of efficiency. The idea of outcome-oriented Figure 2 may be used to show the efficiency measures of a company that uses one input to produce two outputs (let’s say Y1 and Y2). The horizontal and vertical axes, respectively, indicate the two outputs, Y1 and Y2. The production possibility curve, or SS<i>, displays many combinations of two outputs (Y1 and Y2) that can be generated from a certain level of input (Y1). An effective technique from a technical standpoint is the SS</i> production possibility curve. Technically efficient firms are those that are generating at this curve. Technically speaking, a company producing at point C is inefficient since it is below the production potential curve (SS*), which shows the maximum amount of possible output.File: Figure_5_Skewedness_of_Farmers_Technical_Efficiency<b>Description:</b> Stata output graph.File: Figure_3_Conceptual_Framework_of_the_Study<b>Description:</b> Conceptual framework of the study described as a web of connected ideas that, when taken as a whole, offer a thorough comprehension of a situation. Stated differently, it is a written or visual result that provides a narrative or graphic explanation of the primary subjects of study. The conceptual framework of this study is based on the new institutional economics’ institutional assessment and growth technique.<br>File: Figure_4_Map_of_the_Study_Area<b>Description:</b> The map of the area being studied shows the kebeles that make up the study geographic area. A few kebeles were chosen purposively because of their large coffee farms and the need to determine which areas of the district were most and least productive for producing coffee.File: Description_and_Meanings_of_Variables.xlsx<b>Description:</b> uploaded to explain the meanings of the data’s variables one by one. It includes an explanation of each of the two sets of independent variables in the data as well as the dependent variable (output).<b><i>Variables</i></b>totoutput = total output, dependent variable.set 1. Independent variables that measures the elasticity of coffee production in study areaset 2. Independent variables that may cause technical inefficiency of coffee producers in study area<b>Unit of measurement</b>1Quintal of coffee yield = 100kg1hectare land = 10000m^21day working hours = 8 hoursCredit = is the type of loan which will obtained from local credit service in Ethiopian birrThis data is Cross-Sectional DataThis Data is Collected 2022 Production yearTLU = Tropical Livestock unit ; Estimation of Livestock owned by farmers; (by tropical livestock unit conversation). It is an estimation amount. Because cattle can be die, sold etc…Sex = Sex of the household head a dummy variable. It coded with a value of 1and, 0.HH Size = It is not the entire family memberFile: Sources_of_data_Sampling_Technique__and_Sample_Size.docx<b>Description:</b> These documents provide data sources, sampling methodology, and sample size calculations. In general, to explain where the data were collected, the selection of sample household heads, and the calculation of sample size.Code/softwareIn this work, a cross-sectional dataset including 285 respondents was utilized for econometric research to estimate the combined frontier inefficiency model. The many factors influencing the productivity efficiency of coffee growers were estimated using the Stata 15.0 version software package (StataCorp LLC, 2017)Access informationOther publicly accessible locations of the data:-Data was derived from the following sources:Agricultural and Rural Development OfficeCoffee, Tea and spices Development OfficePrimary data was collected within a structured questionnaire by trained enumerators and researchers from the district’s development agency employees and from the selected farmers of the study area.Human subjects dataTo the greatest extent feasible, this data has undergone thorough de-identification. Names, addresses, email addresses, and other direct identifiers have all been permanently deleted, along with all other personally identifiable information (PII). Furthermore, even with the use of easily accessible information, the data has been processed to remove any chance of reasonably determining an individual’s identity.<br>

# 数据与文件结构说明
## 项目负责人联系信息
姓名：阿莱穆·奥利卡（理学硕士与文学硕士）
所属机构：沃莱加大学（Wollega University）、埃塞俄比亚开发银行（Development Bank of Ethiopia）
电子邮箱：[e-mail] / [e-mail]

## 备选联系人信息
姓名：盖梅丘·穆拉图（哲学博士）
所属机构：沃莱加大学（Wollega University）
电子邮箱：[e-mail]

# 数据集概述
本数据集采集自研究区域内的咖啡种植农户。本研究选取了若干凯贝莱（kebele，埃塞俄比亚基层行政单位）开展调研。数据采集前，已完成以下准备工作：
1. 向研究参与者阐明本研究的目的与重要性，随后以口头形式告知受访者，其有权选择是否参与问卷填写。本研究的参与者均为自愿填写问卷的农户。
2. 以口头沟通的方式向样本受访者说明，数据采集流程不会造成困惑或损害受访者权益。问卷的编制过程严谨公正、条理清晰。

本研究未使用人类或动物实验对象，仅针对农户咖啡种植实践开展技术效率分析。数据采集员已接受数据采集流程培训。本研究采用2021年主收获季的截面农户调查数据。

本研究采集了土地、人力劳动、肥料、咖啡植株、除草剂等投入要素数据，以及特定周期内的咖啡产出数据。投入使用量与产出数据均以当地单位采集后转换为标准单位。此外，通过访谈选定的咖啡种植农户，采集了生产效率差异相关变量数据，包括年龄、受教育程度、家庭规模、农技推广接触情况与性别等；同时采集了人口统计数据、信贷获取情况、牲畜存栏量、财富指标与制度相关数据等社会经济变量。另一方面，还收集了咖啡生产趋势、投入品供应与农技推广服务相关数据，以辅助对一手数据的分析与解读。

本问卷已通过沃莱加大学商务与经济学院研究与技术转移副院长审核并印刷。研究者亲自前往选定的小农种植户所在的咖啡豆采收与收获时段，诚恳鼓励他们客观无偏地填写问卷。

# 数据来源
本数据用于研究咖啡生产的技术非效率问题。作为一手数据，其采集自当前参与咖啡种植的选定农户。因此，本数据集兼具定量与定性特征，同时涵盖一手与二手数据来源。

一手数据通过结构化问卷采集。数据采集前，已编制并评估结构化问卷，并根据预调查反馈对问卷进行优化与修改。一手数据采集工作由数据采集员、辖区发展专员与研究者共同完成。本数据同时从政府与非政府机构、已发表与未发表文献、网站及其他相关来源采集，用于分析与描述性研究。

# 数据采集日期
1. 一手数据采集：2021年
2. 二手数据采集：2021年

该辖区共有约31610名咖啡种植农户，分布于30个凯贝莱。第二阶段，从上述凯贝莱中有意选取4个以咖啡种植为主的凯贝莱，因其咖啡种植面积广阔，且需明确辖区内咖啡生产效率最高与最低的区域。这4个凯贝莱共有人口1108人。第三阶段，采用科塔里公式（Kothari formula）随机选取285个样本。

# 资金声明
本研究提交的工作未获得任何组织的资助。本文稿的撰写未获得任何资金支持。本研究全程未获取任何研究经费。

# 人类受试者去标识声明
本数据已在最大可行范围内进行了全面去标识处理。姓名、地址、电子邮箱及其他直接标识符，以及所有其他个人可识别信息（Personally Identifiable Information, PII）均已永久删除。此外，即便结合公开可得信息，本数据也已进行处理，以消除通过合理手段识别特定个人身份的可能性。

---

## 数据与文件结构说明
本研究区域采集的原始一手数据已录入名为《Raw_Data_s_of_Technical_efficiency_analysis_of_coffee_production》的Excel电子表格，并上传至本数据集仓库。此外，《Description_and_Meanings_of_Variables》表格包含1个因变量与两类解释变量：
1. 第一类自变量：以咖啡产出作为因变量，其余5个变量（landcoff：成熟咖啡种植用地面积；laborcoff：咖啡生产投入劳动力；Coffplant：已进入挂果期的成熟咖啡植株数；orgfert：有机肥与除草剂）作为解释变量，用于测算研究区域内咖啡生产的弹性。
2. 第二类自变量：共12个，分别为ageofhhh（户主年龄）、sexofhhh（户主性别）、educofhhh（户主受教育程度）、hhsize（家庭规模）、tlu（热带牲畜单位，Tropical Livestock Unit, TLU）、offincome（非农/副业收入）、totcultland（总可耕地面积）、totlandfrg（总土地碎片化程度）、avrgplotdist（地块平均距离）、extcontact（农技推广接触服务）、train（农户培训情况）、credit（农户信贷服务），上述变量均为可能影响研究区域内咖啡种植户技术非效率的因素。《Description_and_Meanings_of_Variables》表格对每个变量及研究中采用的计量单位进行了详细说明。

### 上传的Word文档
1. **《Questionnairs_for_Technical_Efficiency_Analysis_of_Coffee_Production》**：该文档为用于采集受访者原始数据的工具，包含从受访者处收集的六部分原始数据：
- 第一部分：样本农户基本信息；
- 第二部分：经济状况信息；
- 第三部分：咖啡种植基本信息；
- 第四部分：肥料与化学品（除草剂）使用情况；
- 第五部分：农技推广服务与培训情况；
- 第六部分：信贷服务情况。
2. **《Sources_of_data_Sampling_Technique__and_Sample_Size》**：该文档说明了数据来源、抽样方法与样本量计算过程，整体阐述了数据采集地点、样本农户户主的选取方式及样本量的计算逻辑。

### 图表
1. 图1：技术效率、配置效率与经济效率的投入导向测度图
2. 图2：基于产出导向测度的技术效率、配置效率与经济效率图
3. 图3：本研究概念框架示意图
4. 图4：研究区域地图：数据采集所在凯贝莱的地理位置
5. 图5：农户技术效率分布偏态图

### 文件与变量说明
#### 文件1：Questionnairs_for_Technical_Efficiency_Analysis_of_Coffee_Production.docx
> 说明：用于从受访者处采集一手数据的结构化问卷。

#### 文件2：Raw_Data_s_of_Technical_efficiency_analysis_of_coffee_production.xlsx
> 说明：包含已整理为适用于Stata软件格式的采集数据，包含1个因变量与两类自变量。
> 变量说明：
> - totoutput：总产出，为本研究采用的因变量，指一个生产季节内获取的咖啡产量。
> - 第一组自变量：用于测算研究区域内咖啡生产弹性的变量，包括landcoff（成熟咖啡种植用地面积）、laborcoff（咖啡生产投入劳动力）、Coffplant（已进入挂果期的成熟咖啡植株数）、orgfert（有机肥与除草剂）。
> - 第二组自变量：共12个，为可能影响研究区域内咖啡种植户技术非效率的因素，包括ageofhhh（户主年龄）、sexofhhh（户主性别）、educofhhh（户主受教育程度）、hhsize（家庭规模）、tlu（热带牲畜单位）、offincome（非农/副业收入）、totcultland（总可耕地面积）、totlandfrg（总土地碎片化程度）、avrgplotdist（地块平均距离）、extcontact（农技推广接触服务）、train（农户培训情况）、credit（农户信贷服务）。

#### 文件3：Figure*1 *Graph_of_Input_Oriented_measures_for_Technical_Allocative_and_Economic_Efficiency111.tiff
> 说明：图1展示了在规模报酬不变（Constant Returns to Scale）假设下，使用两种投入要素（Z1与Z2）生产单一产出的农户效率投入导向测度图。其中Z1与Z2两种投入要素分别对应纵轴与横轴。KK’为完全有效率企业的等产量曲线（Isoquant），该曲线上的每一点均代表技术有效的生产。假设某企业在图1中的X点进行生产，其产出水平与完全有效率农户一致。

#### 文件4：Figure_2_Technical_Allocative_and_Economic_efficiency_through_output_oriented_measurement
> 说明：产出导向的效率测度重点关注企业在投入要素不变的情况下可实现的产出变化。图2可用于展示使用一种投入要素生产两种产出（例如Y1与Y2）的企业的效率测度情况。横轴与纵轴分别对应两种产出Y1与Y2。生产可能性曲线（Production Possibility Curve）SS*展示了在特定投入水平下可生产的两种产出（Y1与Y2）的多种组合。SS*生产可能性曲线为技术上有效的生产方式，位于该曲线上的企业均为技术有效率企业。位于生产可能性曲线（SS*）下方的C点生产企业属于技术非效率，因其未达到潜在最大产出水平。

#### 文件5：Figure_5_Skewedness_of_Farmers_Technical_Efficiency
> 说明：Stata软件输出的统计图。

#### 文件6：Figure_3_Conceptual_Framework_of_the_Study
> 说明：本研究的概念框架被定义为一套相互关联的理念集合，整体上可实现对研究主题的全面理解。换言之，其为以文字或可视化形式呈现的、对核心研究主题的叙事或图形解释。本研究的概念框架基于新制度经济学的制度评估与增长分析方法。

#### 文件7：Figure_4_Map_of_the_Study_Area
> 说明：本研究区域的地图展示了构成研究地理范围的凯贝莱。本研究有意选取了若干凯贝莱，因其拥有广阔的咖啡种植面积，且需明确辖区内咖啡生产效率最高与最低的区域。

#### 文件8：Description_and_Meanings_of_Variables.xlsx
> 说明：上传用于逐一解释数据中各变量的含义，包含对数据中两类自变量与因变量（产出）的详细说明。
> 变量：
> - totoutput：总产出，因变量。
> - 第一组：用于测算研究区域内咖啡生产弹性的自变量。
> - 第二组：可能导致研究区域内咖啡种植户技术非效率的自变量。
>
> 计量单位：
> 1 公担（Quintal）咖啡产量 = 100千克
> 1 公顷（hectare）土地 = 10000平方米
> 1 工作日工时 = 8小时
> 信贷：指从当地信贷服务机构获取的、以埃塞俄比亚比尔（Ethiopian Birr）计价的贷款。
>
> 其他说明：
> 本数据集为截面数据，采集于2022年生产年度。
> TLU（热带牲畜单位）：对农户拥有的牲畜进行估算的单位（通过热带牲畜单位转换法），为估算值，因牲畜可能出现死亡、出售等情况。
> 性别：户主性别，为虚拟变量（Dummy Variable），编码为1与0。
> 家庭规模：指家庭常住人口数量，并非指所有家庭成员。

#### 文件9：Sources_of_data_Sampling_Technique__and_Sample_Size.docx
> 说明：该文档说明了数据来源、抽样方法与样本量计算过程，整体阐述了数据采集地点、样本农户户主的选取方式及样本量的计算逻辑。

# 代码与软件
本研究采用包含285个受访者的截面数据集开展计量研究，以估计复合前沿非效率模型。采用Stata 15.0版本软件包（StataCorp LLC, 2017年）估计影响咖啡种植户生产效率的各类因素。

# 数据获取途径
本数据的其他公开获取来源如下：
- 数据来源于以下机构：农业与农村发展办公室、咖啡、茶叶与香料发展办公室
- 一手数据由经过培训的数据采集员与研究者通过结构化问卷采集，数据采集员来自辖区发展机构员工，且受访对象为研究区域内选定的咖啡种植农户。

# 人类受试者数据声明
本数据已在最大可行范围内进行了全面去标识处理。姓名、地址、电子邮箱及其他直接标识符，以及所有其他个人可识别信息（PII）均已永久删除。此外，即便结合公开可得信息，本数据也已进行处理，以消除通过合理手段识别特定个人身份的可能性。