Future of African Remittances: National Surveys 2010 - Ethiopia

Abstract
---------------------------

The Future of African Remittances (FAR) team conducted research on remittance flows to measure and understand the remittance process in sub-Saharan Africa. This ambitious and important research is initially focused on three countries in East Africa – Ethiopia, Kenya and Uganda.

In order to glean insights into the remittance process in the three designated countries, the World Bank designed a two-phase survey process. Phase 1 involved conducting a national survey in each of the three countries. The purpose of the first phase of research was to collect a large representative sample of the adult population in each country. The national surveys provide important baseline data about international remittance flows including: an estimate of the percent of the total adult population that regularly receives remittances, the average amount of each remittance received, most common methods of receipt and top sending countries. Additionally, through the analysis of the national survey results, World Bank was able to identify areas of each country that have high concentrations of international remittance recipients. This important piece of information guided Phase 2 of the research – surveys of remittance receivers in each country. Whereas the national surveys aimed to collect general data about the remittance process, the surveys of remittance recipients allowed for the collection of more detailed data about the remittance process itself, how remittances are used, the relationship between sender and receiver, and interest in various financial products.

The results of this research will not only provide estimates of total annual amounts of remittances for each country, but also will tell us the percentage of the population in each country that is involved in the international remittance process. Furthermore, it will offer insights as to the degree to which Ethiopians, Kenyans and Ugandans depend on international remittances and how the money is used, saved and/or invested. Results will also measure interest in financial products that, if utilized, can significantly impact the financial well-being of the population and the overall economic stability of each country.

Geographic coverage
---------------------------

National Coverage

Analysis unit
---------------------------

Households
Individuals

Universe
---------------------------

The Survey covered the adult population.

Kind of data
---------------------------

Sample survey data [ssd]

Sampling procedure
---------------------------

General:

The total samples were compiled utilizing multi-stage stratified random sampling through respondent selection. Multi-stage random sampling ensured that a random sample of adults was collected in each country. First, after stratifying the population of each country by region and population density, sampling points (SPs) were determined. SPs were then randomly selected within each stratum. At each SP, respondents were randomly selected to participate in the survey.

Phase 1:

The first phase consisted of national surveys of the adult population of each country. The three survey samples were designed to be representative of the adult populations of these three countries. World Bank coordinated and oversaw all aspects of the sampling and interviewing process. A team of local field experts was hired in each country to conduct the actual interviews. All interviewers were professionally trained and supervised by research personnel. In this phase of the research, a total of 2042 Ethiopian adults were interviewed.

Phase 2:

Once the national surveys were completed, the results were analyzed to determine the areas of concentration of the remittance recipient population, after which the second phase of the project was conducted. This phase of the project included a targeted survey of the remittance recipient population of each of the three East African countries. Sampling Points were established based on the analysis of the national survey data and the identification of areas within each country that showed the highest concentrations of remittances received from relatives abroad in proportion to the sample size of all areas surveyed. Once again, local field experts were hired in each country to conduct the interviews, training and supervision of field operations. Languages of interviews were the same as those employed in Phase 1 and, again, all interviews were conducted in person using the PAPI method. A total of 400 interviews with regular international remittance recipients were completed in each country during August and September of 2010. The margin of error for all three surveys is approximately ±5 percentage points and the 95 percent level of confidence.

Detail:

The total sample was compiled utilizing multi-stage stratified random sampling through respondent selection. This sampling method enabled B&A to ensure that a representative random sample of Ethiopian adults was collected. There are three stages to this type of sampling methodology. First, after stratifying the Ethiopian population by region and population density, sampling points (SPs) were determined. SPs were then randomly selected within each stratum. In the second stage, using the random route method, dwellings were selected within each SP. The random route method involves selecting an address in each SP at random as a starting point. Each interviewer was given instructions to identify additional dwellings by taking alternate left and right turns and stopping at every Nth dwelling. The third and final stage involved selecting actual participants - for each selected dwelling, individual respondents were chosen using a Kish grid. In a Kish grid, prior to beginning the interview, the interviewer first asks for the ages and genders of every household member (only persons aged 18 or older were eligible for selection). The individual to be
interviewed was then chosen based on a random number in the grid.

Once the national survey was completed, B&A analyzed the results to determine the areas of concentration of the remittance recipient population, after which the second phase of the project was conducted. This phase included a targeted survey of the remittance recipient population in Ethiopia. Sampling Points were established based on B&A's analysis of the national survey data and the identification of areas of the country that showed concentrations of international remittance receivers in proportion to the sample size of all areas surveyed. Once again, local field experts were hired to conduct the interviews and B&A conducted all training and supervision of field operations. Interviews were conducted in Amharic and all interviews were conducted in person using the PAPI method. A total of 410 interviews with regular international remittance recipients were conducted in Ethiopia during August and September of 2010. The margin of error for the surveys is approximately ±5 percentage points and the 95 percent level of confidence.

Mode of data collection
---------------------------

Face-to-face [f2f]

Research instrument
---------------------------

Phase 1:

This survey consisted of 12 questions that were aimed at helping to identify some of the basic characteristics of the remittance recipient population in each country. Some of the variables included in this survey were – location, age, gender, amount of money received, method of receipt, origin of remittance, etc.

Phase 2:

The survey instrument for Phase 2 consisted of approximately 35 questions and included a number of variables aimed at obtaining greater detail about the remittance receiving process including costs, amounts received, information about the sender and the relationship between sender and receiver. Additionally, the survey measured interest in various financial products.

Response rate
---------------------------

Every effort was made to achieve the maximum possible coverage, taking cost, timing and other factors into account. A coverage rate of 97% was achieved in the national survey and the 3% of the country that was not covered consisted of areas that were either very remote (and difficult to travel to) or that had extremely small populations.

Sampling error estimates
---------------------------

The margin of error is approximately ±5 percentage points and the 95 percent level of confidence.

非洲汇款未来（FAR）团队对撒哈拉以南非洲的汇款流动进行了研究，旨在衡量和理解汇款流程。这一雄心勃勃且至关重要的研究最初聚焦于东非的三个国家——埃塞俄比亚、肯尼亚和乌干达。

为了深入了解这三个指定国家的汇款流程，世界银行设计了一个两阶段的调查流程。第一阶段包括在每个国家进行一次全国性调查。研究第一阶段的目的在于收集每个国家成年人口的代表性样本。全国性调查提供了关于国际汇款流动的重要基础数据，包括：估计接受汇款的成年人口占总成年人口的百分比、每次汇款的平均金额、最常用的接收方式以及主要汇款国。此外，通过对全国性调查结果的分析，世界银行能够识别出每个国家国际汇款接收人高度集中的地区。这一重要信息指导了研究的第二阶段——对每个国家的汇款接收人的调查。与旨在收集关于汇款流程的一般数据的全国性调查相比，汇款接收人的调查允许收集关于汇款流程本身、汇款的使用方式、发送者与接收者之间的关系以及各种金融产品兴趣的更详细数据。

该研究的结果不仅将提供每个国家年度汇款总额的估计，还将告诉我们每个国家参与国际汇款流程的人口百分比。此外，它将揭示埃塞俄比亚人、肯尼亚人和乌干达人依赖国际汇款的程度以及资金的用途、储蓄和/或投资情况。结果还将衡量对金融产品的兴趣，这些产品如果得到利用，将显著影响人口的经济福祉和每个国家的整体经济稳定。

地理覆盖范围
---------------------------
全国覆盖

分析单元
---------------------------
家庭
个人

总体
---------------------------
调查涵盖了成年人口。

数据类型
---------------------------
样本调查数据 [ssd]

抽样程序
---------------------------
一般：

通过受访者选择的多阶段分层随机抽样编制了总样本。多阶段随机抽样确保在每个国家收集了成年人的随机样本。首先，按地区和人口密度将每个国家的总人口进行分层，然后确定抽样点（SPs）。然后，在每个分层内随机选择SPs。在每个SP处，随机选择受访者参与调查。

第一阶段：

第一阶段包括对每个国家成年人口的全国性调查。这三个调查样本被设计成代表这三个国家的成年人口。世界银行协调并监督了抽样和访谈的所有方面。在每个国家雇佣了一支当地现场专家团队进行实际访谈。所有访谈员都接受了专业培训并由研究人员监督。在这一研究阶段，共访谈了2042名埃塞俄比亚成年人。

第二阶段：

一旦完成全国性调查，分析结果以确定汇款接收人人口集中的地区，随后进行了项目的第二阶段。这一阶段的项目包括对三个东非国家中每个国家的汇款接收人进行针对性调查。抽样点基于全国性调查数据分析和识别出每个国家接收来自国外亲属的汇款比例最高的地区。再次，在每个国家雇佣了当地现场专家进行访谈、培训和现场操作的监督。访谈语言与第一阶段相同，再次，所有访谈都是通过面对面（PAPI）方法进行的。2010年8月和9月，每个国家完成了400名定期国际汇款接收人的访谈。所有三个调查的误差范围约为±5个百分点，置信水平为95%。

详细：

通过受访者选择的多阶段分层随机抽样编制了总样本。这种抽样方法使B&A能够确保收集了埃塞俄比亚成年人的代表性随机样本。这种抽样方法有三个阶段。首先，按地区和人口密度将埃塞俄比亚总人口进行分层，然后确定抽样点（SPs）。然后，在每个分层内随机选择SPs。在第二阶段，使用随机路线法，在每个SP内选择住宅。随机路线法包括在每个SP内随机选择一个地址作为起点。每个访谈员都被指示通过交替左转和右转，并在每N个住宅处停下来来识别额外的住宅。第三和最后阶段涉及选择实际参与者——对于每个选定的住宅，使用Kish网格选择受访者。在Kish网格中，在开始访谈之前，访谈员首先询问每个家庭成员的年龄和性别（只有18岁或以上的人才有资格选择）。然后，根据网格中的随机数字选择要访谈的个人。

一旦完成全国性调查，B&A分析了结果以确定汇款接收人人口集中的地区，随后进行了项目的第二阶段。这一阶段包括对埃塞俄比亚汇款接收人进行针对性调查。抽样点基于B&A对全国性调查数据的分析以及识别出国家中接收国际汇款的集中地区。再次，在每个国家雇佣了当地现场专家进行访谈和现场操作的培训与监督。访谈在阿姆哈拉语进行，所有访谈都是通过面对面（PAPI）方法进行的。2010年8月和9月，在埃塞俄比亚进行了410次定期国际汇款接收人的访谈。调查的误差范围约为±5个百分点，置信水平为95%。

数据收集方式
---------------------------
面对面 [f2f]

研究工具
---------------------------
第一阶段：

该调查包括12个问题，旨在帮助识别每个国家汇款接收人的一些基本特征。该调查中包含的一些变量包括——位置、年龄、性别、收到的金额、接收方式、汇款来源等。

第二阶段：

第二阶段的调查工具包括大约35个问题，包括一些旨在获取关于汇款接收过程的更多详细信息的变量，包括成本、收到的金额、关于发送者的信息以及发送者与接收者之间的关系。此外，该调查还测量了对各种金融产品的兴趣。

响应率
---------------------------
尽一切努力实现最大可能的覆盖率，同时考虑成本、时间和其他因素。全国性调查的覆盖率为97%，未覆盖的3%的国家地区要么非常偏远（难以到达），要么人口极为稀少。

抽样误差估计
---------------------------
误差范围约为±5个百分点，置信水平为95%。