By Ahmed Yusuf
Addis Abeba, September 03/2020 – A new survey released on August 25 by Afrobarometer, a pan-African research network, revealed that Ethiopians (61%) embrace a federal system as opposed to a unitary form of government. When it comes of the type of federalism, “49% prefer the current federalism delimitation; (Rural resident prefer this method); and 48% prefer geographic federalism (urban residents favor this form of demarcation); 37% prefer a unitary form of government.”
Furthermore, “Large majorities say that the Constitution should be amended – rather than replaced or maintained as it is –to reflect the needs of contemporary Ethiopia and that ordinary citizens should be consulted during the review process,” Afrobarometer said in a statement.
However, unlike the deliberate confusion created during reporting from various local media houses, the survey reveals the amendment was favored only to incorporate three progressive agendas: Designating additional working languages for the federal government (73%); Limiting the prime minister to serving a maximum of two terms (68%); and Establishing a Constitutional Court separate from the House of Federation, which is currently vested with the power to interpret the Constitution (55%).
The most consequential results of the survey indicate that 50% of those surveyed want to maintain the guarantee of Article 39 on the rights of nations, nationalities, and peoples to self-determination, secession, and establishment of their own regional state government; while only 43% of those surveyed support removing it.
Similarly, the survey result in another hotly contested topic showed that “Fewer than four in 10 Ethiopians (37%) want to remove the emblem at the center of the national flag, while a majority (52%) are opposed to its removal,” said the result of the survey.
The survey comes in the midst of heated debates on whether Ethiopia should keep the current multinational federalism, adopted in 1995; and whether the emblem at the center of the Ethiopian national flag should be removed.
Of these surveyed 49% say the federalism should continue to be delimited, based on “ethnicity”, a term used frequently in the survey albeit it never appears in the Constitution of the Federal Democratic Republic of Ethiopia (FDRE), whose preamble begins with: “We, the Nations, Nationalities and Peoples of Ethiopia.”
And the survey results further indicate that 69% of Ethiopians prefer the current constitution to be upheld with amendments recommended only on three areas which are “adding additional working languages, limiting the Prime Minister’s terms and establishing a constitutional court separate from the house of federation.” While 18% want the constitution untouched and a minority (11%) want it discarded completely.
Despite the globally acclaimed track-record Afrobarometer enjoys criticisms linger. One of them is its careless use of the term “ethnicity” while surveying a constitution that doesn’t use the word to describe itself.
The survey used the term “ethnic group” or “ethnicity” several times although it is not defined nor is it mentioned anywhere in the constitution. The appropriate term would have been “nations, nationalities and peoples” who, according to Article 47 sub. Article 2 of the constitution, have their own States.
The survey result mentioned that out of the 61% that prefer federalism, 49% want the federalism to be delimited based on “ethnicity”. But according to the words of the constitution, States shall be delimited on the basis of the settlement patterns, language, identity and consent of the people concerned. It does not state anywhere regions should be based on “ethnicity” avoiding the singularity of the term.
So the question remains, where did terms like “ethnic group” and “ethnicity” come from? Afrobarometer survey does not explain how it reached at a conclusion to use the terms, nor how it translated these terms into local languages?
Other criticisms were raised too, and have cast some doubts among skeptics. These include the loaded nature of the questionnaires and past criticisms on interviewers errors. The never released Afrobarometer’s 2013 survey on Ethiopia has also resurfaced among few critiques.
Dr. Mehari Taddele Maru, a Professor at the European University Institute in Florence, Italy, during his interview with Addis Standard said “Afrobarometer’s achievements in Africa are laudable. We need more data. We need more surveys like these.” He continued “But there are some challenges that need improvement. For example, some of the questions in the survey were leading to the extent that they suggest answers.”
He advises caution in approaching questions such as “Should Ethiopia’s constitution be amended?” “These loaded questions would have been avoided by qualifying the ‘changes or reforms’ as ‘major or minor changes or reforms’,” argues Dr. Mehari.
Regarding the question “Remove Article 39 on self-determination”, Dr Mehari suggests “People may not want Article 39 discarded in its entirety. Some may want only the phrase “right to succession” to be removed and they may want the remaining part intact. Had the Article 39 question been divided in this manner, we would have seen a different response.”
He continues to argue that statements such as “there is no need for changes or reforms” are suggestive. “The respondent is being given words that he might not have considered.”
“When you say the ‘the constitution should be amended to meet the needs of the current time’, it is a general statement that no one wants to disagree with.”
“When asking ‘the current Constitution should be discarded and replaced with a new Constitution’, it should have given multiple options to include different perspectives such as those who want to replace it constitutionally, those who want to replace it via referendum albeit unconstitutionally and those who want to replace it in an extra-constitutional way,” argues Dr. Mehari.
Dr. Mehari also indicated that using Urban-Rural variable was not relevant for questions that involve Article 39. “It may not be relevant for a policymaker to know Urbanites want Article 39 discarded more than Rural residents. Instead, the variable Educational Level would have been more relevant for explaining opinions regarding Article 39. However, Urban-Rural variable is relevant when it comes to the question of land ownership.”
Leila Demarest, Assistant Professor of African Politics at the Institute of Political Science, Leiden University, published a paper on 2017 titled “An Assessment of Interviewer Error in the Afrobarometer Project”.
She asserts that her analyses indicate that interviewer variances are “substantial in Afrobarometer data and require further attention”. “Interviewers often conduct interviews with poorly educated or illiterate respondents, which possibly increase their impact on how questions are understood and/or answered,” says Leila Demarest.
“Within the Afrobarometer survey design, interviewers clearly play crucial roles. They are responsible for the sampling of households and individuals, establishing contact and achieving cooperation, and administering questionnaires in face-to-face interviews without computer aides. Interviewer-related errors in Afrobarometer data can therefore be substantial.”
“During the interview, interviewers can influence respondents by not following interview protocols, performing on-the-spot translations which lead to different interpretations, or by eliciting certain responses because of observable traits,” she adds.
She recommends a “detailed monitoring of the interviewer process and more quality control.” She suggests the use of “Computer Assisted Personal Interviewing (CAPI) to reduce mistakes”, “following recommendations from local partners during interviewer hiring process and rigorous training of interviewers.”
Why the 2013 Survey was not made public?
The first round of Afrobarometer survey in Ethiopia took place in 2013. One of the key findings of this survey was that 81% of the respondents believe Ethiopia was “either a complete democracy or a democracy with only minor problems”. Nevertheless, Afrobarometer did not come out publicly with this survey as it did for the second round (2020). Addis Standard asked what had happened.
“On the 2013 survey analysis, we had an outlier data.” said Mulu Teka, National Investigator, ABCON – Research & Consulting PLC. “A majority of our respondents believed we represented the government although at the start of the interview they were clearly informed who we are and what our objective was,” added Mulu. “After introductions and all questions are finalized, our interviewers ask them from where we came or who we represented. And the response we got was that we came from the government,” says Mulu. And this, according to the National Investigator, led Afrobarometer to rethink the whole project. For this reason, he said “round six and round seven surveys were skipped in Ethiopia.”
On May 2016, Afrobarometer released a working paper to explain the challenge it encountered. The working paper was titled “Ethiopians’ views of democratic government: Fear, ignorance, or unique understanding of democracy?”.
The paper read “Ethiopians’ highly positive assessment of political and economic developments in their country, according to the report, is “marked by a syndrome of ‘uncritical citizenship’ and a distinctively instrumental and paternalistic understanding of ‘democracy’. Other contributing factors include the country’s low level of development, especially with respect to education and communications; its long-standing one-party dominance and low levels of political freedom; and significant political fear and suspicion of the interview environment. Because of the idiosyncratic way in which Ethiopians understand democracy, extreme caution must be exercised in attempting to compare any responses to democracy questions from Ethiopia with those from other African countries. Analysts are advised to use the Ethiopia data set only in a stand-alone setting or to limit their comparative analysis to items that do not use the ‘d-word’.” The paper was co-authored by Robert Mattes, a professor of political studies at the University of Cape Town, South Africa, and Mulu Teka, national investigator for Afrobarometer in Ethiopia and Director of ABCON – Research & Consulting PLC, Ethiopia.
The Methodology of the Afrobarometer Survey
The Afrobarometer team in Ethiopia, led by ABCON – Research & Consulting PLC, interviewed 2,400 adult Ethiopians in December 2019 and January 2020. Sample came from 9 regional states, two city Administrations, Addis Abeba & Dire Dawa, and further disaggregated by urban & rural areas in proportion to their share in the national population.
According to Afrobarometer, a sample of this size yields country-level results with a margin of error of +/-2 percentage points at a 95% confidence level. This means, “if we take the sample of 2,400 a hundred times, there is 95% probability that the result is within the 40% +/- 2% (between 38% and 42%)” explained Afrobarometer team during the press release on August 25, 2020.
“The methodology and sample size is the same for both rounds [i.e. 2013/14 and 2019/20]” told Mulu Teka, National Investigator, ABCON – Research & Consulting PLC, told Addis Standard. He added “We were fortunate that the Central Statistics Authority (CSA) had prepared a new digital map and enumeration area for the upcoming census which had an updated estimated population size.” said Mulu.
“A sampling frame was obtained from the CSA. And together with the CSA team, we randomly identified 300 clusters or enumeration areas in all regions”, said Mulu. “It is important to understand how we came to 300. Since our sample size was 2,400, we divided it by 8 interviews per cluster and we get 300 enumeration areas”, he added. Afrobarometer uses clustered, stratified, multistage probability sample design. It is said the sample is representative of all citizens of voting age. At every stage of sampling, Afrobarometer says, it applies random selection method and applies sampling with probability proportional to population size.
According to Mulu Teka, the interviewers/enumerators for the survey research have a minimum of bachelor degrees and they are proficient in the local languages.
Who funds the Afrobarometer?
The 2nd round survey for Ethiopia was carried out with support from Freedom House, a U.S.-based, U.S. government-funded non-profit non-governmental organization that conducts research and advocacy on democracy, political freedom, and human rights.
Funding for its previous rounds of surveys primarily came from the U.S. State Department, National Endowment for Democracy, UK Department for International Development (DFID), World Bank, Bill & Melinda Gates Foundation, Duke University China Research Center, Transparency International, Institute for Security Studies, Royal Danish Ministry of Foreign Affairs (RDMFA), Danish International Development Agency (DANIDA), and Canadian International Development Agency (CIDA).
For the 8th round in other African countries, the Afrobarometer received financial support from Sweden, Mo Ibrahim Foundation, Open Society Foundations, William and Flora Hewlett Foundation, U.S. Agency for International Development (USAID) via the U.S. Institute of Peace.
Some of the globally renowned groups that use Afrobarometer to produce global and regional governance indicators are The Economist Intelligence Unit, the United Nations Development Program, the World Bank. And in 2016, Afrobarometer data became the first public attitude survey data included in the Ibrahim Index of Africa Governance.
It is expected that Afrobarometer will hold discussions and debates with its stakeholders in the near future. “It will also conduct follow up surveys in Ethiopia before embarking on Round 9 survey,” says Afrobarometer’s Ethiopia representative and director of ABCON – Research & Consulting PLC. AS