Talk:Demographics of Eritrea

This article is rated B-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects:

Africa: Eritrea Top‑importance This article is within the scope of WikiProject Africa, a collaborative effort to improve the coverage of Africa on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.AfricaWikipedia:WikiProject AfricaTemplate:WikiProject AfricaAfrica articles Top This article has been rated as Top-importance on the project's importance scale. This article is supported by WikiProject Eritrea (assessed as Top-importance).

Text and/or other creative content from Ethnic groups of Eritrea was copied or moved into Demographics of Eritrea with this edit. The former page's history now serves to provide attribution for that content in the latter page, and it must not be deleted as long as the latter page exists.

The contentious topics procedure applies to this page. This page is related to the Horn of Africa (defined as including Ethiopia, Somalia, Eritrea, Djibouti, and adjoining areas if involved in related disputes), which has been designated as a contentious topic. Editors who repeatedly or seriously fail to adhere to the purpose of Wikipedia, any expected standards of behaviour, or any normal editorial process may be blocked or restricted by an administrator. Editors are advised to familiarise themselves with the contentious topics procedures before editing this page.

Demographics of Eritrea Describing estimates of the Eritrean population is contentious. The most recent discussions include: RfC: consensus for the inclusion of all relevant estimates.

Proposal: restore this version of the article and add to the graph (and the text) the Ministry of Information (Eritrea) datum of 3.56 million for 2002, from this archived URL, in a properly formatted, dated, archived reference.

If anyone has an objection to this proposal, please clearly state why you oppose the proposal, preferably with specific references, and give your reasons for objections, based on Wikipedia policy. If you have a reason why either the United Nations Department of Economic and Social Affairs Population Division or Statista, the two sources that give the low estimates, is an unreliable source, please state that reason. Boud (talk) 17:53, 2 March 2021 (UTC)[reply]

Object, to your new edit proposal. If the second government source (2002 estimates) is included it should be as a compliment to highlight consistency with the more recent six million estimates. Otherwise it don’t serve a purpose to cite a 19 year old estimate. The statista source is most likely based on the UN figures which is the least credible one. The estimates in the lead, info box and body should be based on:

1. CIA (2021 estimates), 6,14M [1]
2. COMESA (2019 estimates), 6,75M [2]
3. African Development Bank (2017 estimates), 5,5M [3]
4. WHO (2016 estimates), 5,2M [4]
5. Government source (2020), 6M [5]
6. Government source (2002), 3,5M to show consistency with other sources. [6]

The five sources are all independent of each other and provides different estimates close to 6M and they are also consistent with the then 3,5 (2002) government estimates. The UN source should be excluded since there exist strong support (sources) for estimates of around 6M. I will make a new draft adding the references and sources that have yet been included (WHO, COMESA, AFDB and Government sources). Leechjoel9 (talk) 21:41, 2 March 2021 (UTC)[reply]

• Thank you for agreeing that the Statista source is a second source for the low estimate of about 3.5 million in or around 2020.
• You have raised a new issue about Statista: you hypothesise that Statista data is based on the 2019 UN DESA data. I agree that that is likely. However, the same hypothesis applies to the other group of sources.
• You have presented no evidence for your four sources (CIA 2021, COMESA 2019, ADB 2017, WHO 2016) being independent from one another. There are two sources for the low value and (now) four sources for the high value. See the next point for the Eritrean Ministry of Information value.
• I don't understand why you again state Government source (2020), 6M [https://shabait.com/amp/2020/11/page/9/]. So I will state it in bold: the 2020 Eritrea Min of Information estimate that you have provided before and repeated here is a URL that does not state any estimate of the population of Eritrea. It does not state that the population of Eritrea is 6M. If you still insist that it does, then please quote the relevant part of the page, or explain where on the page the information is located.
• You have not provided an argument for excluding the 2019 UN DESA estimate, nor for excluding the Statista estimate. We do not know why UN DESA has a highly detailed, highly falsifiable set of data that gives lower estimates over many decades than other estimates. Stating that it's in a numerical minority neither explains why it's right nor why it's wrong. Boud (talk) 03:06, 3 March 2021 (UTC)[reply]

Comment See the section above (not in the proposal itself) for three websites that give 2021 estimates that show that the authors of these websites judge the UN DESA 2019 estimate to be accurate. This is only a weak argument in favour of the proposal, which is why I only give the details above. Boud (talk) 15:18, 3 March 2021 (UTC)[reply]

The sources are independent of each other since they vary in terms of estimates figures. They also come from various organisation engaging with Eritrea in number of areas i.e African Development Bank (Finance), COMESA-Common market for East and South Africa (Trade), WHO (Health), CIA (Intelligence), and then the government sources which cites 6M, all consistent with each other. You have lacked to reach consensus on this matter.Leechjoel9 (talk) 15:32, 3 March 2021 (UTC)[reply]

It's true that ADB, COMESA, WHO, and CIA are sufficiently independent organisations that the information they more or less agree on is likely to be a serious claim. But that doesn't answer the question. You have not given any reason why the UN DESA 2019 Revision is an unreliable source; you have only stated that in terms of websites of several major organisations, the UN DESA 2019 Revision estimate is in a minority. That's not an argument why UN DESA 2019 Revision is unreliable, it is only an argument that it's in a minority when we ignore the dates.

UN DESA has a research group that does demographic research based on sources, such as the Eritrea – Demographic and Health Survey 2002 (DHS 2002) of 8754 women in Eritrea. As far as I know, none of the other sources above (including Statista, Worldpopulationreview, Worldometers, Macrotrends) does their own demographic analyses. Some probably do some checking and some modelling, but it's quite likely the main influence on their data is the pre-2019-Revision UN DESA analyses. This would explain why they more or less agree with each other. (We have no evidence so far of why the UN DESA 2019 Revision drastically reduced its historical and current estimates of the Eritrean population compared to earlier estimates; but we know that that's what the UN DESA demographers have done.)

If you have a reason why the UN DESA 2019 Revision analysis is unreliable as demographic information, not why it is in a minority compared to non-demography-research organisations, then please state that reason. Boud (talk) 20:32, 3 March 2021 (UTC)[reply]

I have given you explanations per above along with the sources provided. It is a outlier, practically the only source showing this figure. The sources above provides great variations of estimates, are consistent with each other and are based on actors engaging with Eritrea & the country itself. As stated before it seems it’s possibly based on old estimate (~2002 as government source above). Leechjoel9 (talk) 21:08, 3 March 2021 (UTC)[reply]

It is a[n] outlier is not an argument for why UN DESA 2019 Revision is unreliable.

The UN source make broad generalisation estimates, it lets user print estimates up the to year 2050 The UN DESA Population Division is not just a website. It's an organisation of people that do demography research, and it provides a web interface and data files to the global community. You seem to be arguing that UN DESA 2019 Revision is unreliable because it extrapolates into the future. This is not an argument for why UN DESA is unreliable for 2020 or 2021. Demographers do their best to understand current data, and in some cases, such as this one, also predict future data, based on a series of hypothetical scenarios. So this is not an argument for UN DESA 2019 Revision to be unreliable for 2020 or 2021. Boud (talk) 21:33, 3 March 2021 (UTC)[reply]

In this case it becomes unreliable since practically everyone report other figures. You haven’t reached consensus for your proposition per WP:CON. Leechjoel9 (talk) 21:51, 3 March 2021 (UTC)[reply]

If you have an argument for why the UN DESA 2019 Revision is unreliable, then please present that argument, base on the fact that this type of knowledge is part of demography. Reliability is not a popularity contest. Boud (talk) 22:59, 3 March 2021 (UTC)[reply]

Object to Boyd’s Original Research. The consensus of sources agrees with a population in the 6 million. Many TPLF media outlets and blogs have been presenting a false narrative of Eritrea’s population. The agenda is to claim that Eritrea has less people than in 2002 therefore the Tigrinya population of Eritrea needs to be replaced by Tigrayans. This is a known part of the Abay Tigray Manifesto to invade and annex Eritrea as part of Tigrayan only Nation called GreatyTigray. Political agendas like these should not be allowed to confuse and try to spread Fake News via Wikipedia. The credible sources that state in 6 million should stay, anything that is clearly pushing some obscure agenda should be disregarded.Clownshking (talk) 23:15, 3 March 2021 (UTC)[reply]

@Clownshking: Do you have an argument why the UN DESA 2019 Revision is wrong? I am not the United Nations Department of Economic and Social Affairs Population Division. The TPLF does not control UN DESA. The TPLF does not control the Eritrean National Statistics Office. Please read the section below to understand why Eritrea "lost 1.8 million people" in terms of revisions of knowledge about the past. Boud (talk) 23:56, 3 March 2021 (UTC)[reply]

This user is heavily involved in pushing a TPLF POV by using multiple socks! There needs to be an investigationClownshking (talk) 23:45, 2 March 2021 (UTC)[reply]

Evidence of Boud talking about another new user as if he knows him: https://en.wikipedia.org/w/index.php?title=Talk%3ATimeline_of_the_Tigray_War&type=revision&diff=1009921851&oldid=1004690756

Boud = The People’s Front of Judea user

Boud wrote this in the Plain English section of the Talkpage: Timeline of Tigray War “The source says nothing about a military threat by the TPLF against Eritrea in the recent (2020) epoch. The Peoples Front of Judea, who reverted your revert, might be able to explain this better if my explanation is unclear.“ Clownshking (talk) 02:11, 3 March 2021 (UTC)[reply]

The above comment is a personal attack to which I have responded here. This is not the page to continue this discussion. Boud (talk) 03:29, 3 March 2021 (UTC)[reply]

The UN DESA Population Division 2019 Revision provides these population times series for Eritrea and elsewhere in this file, calculated by the demographers who work there. The last two decades of data for Eritrea are the following, for the total population in thousands:

2000    2292.413
2001    2374.721
2002    2481.059
2003    2600.972
2004    2719.809
2005    2826.653
2006    2918.209
2007    2996.54
2008    3062.782
2009    3119.92
2010    3170.437
2011    3213.969
2012    3250.104
2013    3281.453
2014    3311.444
2015    3342.818
2016    3376.558
2017    3412.894
2018    3452.797
2019    3497.117
2020    3546.427

According to the Ministry of National Development (Eritrea), as quoted by the the Eritrean National Statistics Office (NSO), the 2010 population estimate for Eritrea is given in this NSO pdf file, pdf page 31, numbered page 3. This official source from Eritrea states, No population census has ever been carried out in Eritrea. However, based on a population count by the Ministry of Local Government and NSO estimates, the total resident population of Eritrea was about 3.2 million as of 2010 (MND, 2010). This is consistent with the UN DESA 2019 Revision, since 3170.437 rounds to 3200. This document from 2010 presents the results of the Population and Health Survey (EPHS2010) that is one of the documents used for Eritrea in the UN DESA 2019 Revision.

For the explanation of how Eritrea seemed to "lose 1.8 million people" in the UN DESA 2019 Revision, please go to:

• https://population.un.org/wpp/Download/Standard/Population/ ;
• at the words "release note" in "On 28 August 2019 a minor technical correction ... see release note for further details" you will see this release-note-v1 pdf file with release notes for the 2019 Revision;
• on page 2 there is the Eritrea update explanation: The population of Eritrea in 2019 is 3.5 million, which is about 1.8 million (34.1 per cent) less than the previous estimate from the 2017 revision. The decrease is due to the availability of new official population estimates for several years (population count in 2000, official estimates up to 2018) that contribute to lower the size of the population in the recent years, as well as to revised past estimates since 1950. I put 2017 revision in bold because this means "the data and time series that were published by UN DESA in 2017". The 2019 revision is a new set of data, for all years from 1950 to 2020.

So now we have an official UN DESA 2019 Revision explanation for why many websites still have the higher estimates: they are quite likely still based on the UN DESA 2017 Revision and have not bothered to update to the latest data yet. Boud (talk) 23:52, 3 March 2021 (UTC)[reply]

You haven’t gained consensus for your proposal. This is trying impose your views on this article. It has been addressed to you above. Leechjoel9 (talk) 20:14, 4 March 2021 (UTC)[reply]

I disagree with Boud using this one source to change the info in the article when there are multiple reliable credible and official sources which agree with the figure of 6 million.Clownshking (talk) 21:35, 4 March 2021 (UTC)[reply]

See below for the issue of consensus. Boud (talk) 17:10, 5 March 2021 (UTC)[reply]

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

Consensus is clear for option 2, that is, the inclusion of all relevant estimates. Dispute does exist, but overall the position amongst both frequent editors of this article and unattached ones suggests that the uncertainty and discrepancy involved in Eritrean population estimates needs to be discussed in the article, including the addition of the 3.5 million estimate. This is a pretty long discussion with a lot of metaphorical ink spilled, some of which has been about the exact wording of what option 2 would look like rather than about whether it's acceptable, and some of which has been passionate advocates of option 0. For better or worse, option 0 has not received community support, while option 2 has. Vaticidalprophet 19:01, 19 April 2021 (UTC)[reply]

---

In Demographics of Eritrea, should the UN DESA 2019 Revision estimate of 3.5 million for the current population of Eritrea (Option 0) be excluded completely; (Option 1) be the only estimate included; or (Option 2) included together (NPOVed) with the estimates of 6 million? RfC extended by Boud (talk) 22:10, 7 April 2021 (UTC), originally opened by Boud (talk) 17:01, 5 March 2021 (UTC)[reply]

Overview with sources: Eritrea has never had a census (pdf page 31), but has had random population samples interviewed (same reference, 2010, Eritrea National Statistics Office + partners). In 2019, the UN DESA Population Division 2019 Revision made a major change to its time series of Eritrean population estimates, with an explanation in the Release Note v1 (p2) of why the 2019 estimate is 1.8 million lower than the value in the UN DESA 2017 Revision. The Release Note v1 also briefly explains why the whole time series, from 1950 to 2020, was revised in the UN DESA 2019 Revision estimates in this xlsx file.

UN DESA publishes not only estimates of past populations, but also projects to the future, through to 2100. So older publications of data by UN DESA (Revisions) can be used to "predict" today's Eritrean population count. Several websites seem to be using the UN DESA 2017 Revision or maybe much older projections when they state the "current" (2020 or 2021) Eritrean population, but generally give no details of where they get their data from. These websites state about 6 million for the 2020 or 2021 population of Eritrea. Other websites state the 3.5 million estimate, and mostly refer to UN DESA as a source.

The 2010 Eritrea National Statistics Office (NSO) for 2010 is 3.2 million (pdf page 31), attributed to the Eritrean Ministry of Local Government and the NSO; the UN DESA 2019 Revision estimate for 2010 is 3.170437 million. The Eritrean National Statistics Office and UN DESA 2019 Revision agree on the 2010 value.

In the article Demographics of Eritrea, should the UN DESA 2019 Revision estimate of 3.5 million for 2020:

• Option 0: be excluded completely?
• Option 1: be included, with no mention of the roughly 6 million estimates?
• Option 2: be included together with the estimates of 6 million under WP:NPOV in both the lead and body, and with a brief discussion of Release Note v1 of UN DESA 2019 Revision?

Please state the best option and justify it with reasons. Boud (talk) 17:08, 5 March 2021 (UTC)[reply]

• Object to it being included in the lead, info box and in body. This has been explained to Boud several times in all of the users proposals, in all the discussion above and in Eritrea talk page. The consensus has for several years been 5-6 million. The estimates in the lead, info box and body should be based on:

1. CIA (2021 estimates), 6,14M [7]
2. COMESA (2019 estimates), 6,75M [8]
3. African Development Bank (2017 estimates), 5,5M [9]
4. WHO (2016 estimates), 5,2M [10]
5. Government source (2020), 6M [11]
6. Government source (2002), 3,5M to show consistency with other sources. [12]

The five sources are all independent of each other and they vary in terms of estimate figures in the 6M and they are also consistent with the then 3,5M (2002) government estimates.

They also come from various organisation engaging with Eritrea in number of areas i.e African Development Bank (Finance), COMESA-Common market for East and South Africa (Trade), WHO (Health), CIA (Intelligence), and then country itself (Ministry of Information) cites 6M (2020), all the sources figures are also consistent with each other. The UN source is the outlier in this case and the only one that shows estimates in the 3,5M region and as heavily outnumbered it’s considered least credible, as pointed out numerous times. Leechjoel9 (talk)

• Comment on the current version of Leechjoel's statement: 5. Government source (2020), 6M [https://shabait.com/amp/2020/11/page/9/]. The 6 million quote is absent from the Ministry of Information URL provided (and archived). Boud (talk) 19:53, 5 March 2021 (UTC)[reply]
• Comment the WHO estimate is 5.0M (4.955M), not 5.2M. So, it's actually closer to the new DESA estimate than the COMESA estimate, for example. Also, in the "Government source (2020)" link, I can't find any reference to population. Is it in one of the articles on that page? BubbaJoe123456 (talk) 20:44, 5 March 2021 (UTC)[reply]

• OBJECT to inclusion of Bouds estimate of 3.5 million in 2021,Eritrea’s population count in 2002 was 3.5 million but now according to the reliable official sources as mentioned by Leech, the population is roughly 6 million. Bouds estimate is original research that is incorrect and not based in facts provided by the current sources.Clownshking (talk) 19:43, 5 March 2021 (UTC)[reply]
• Option 2 Given the underlying uncertainty in the number (lack of quality statistics), the existence of the DESA figure should be acknowledged in the article. We've got five relatively current sources: CIA (6.14), COMESA (6.72M), ADB (5.5M), WHO (4.96M), and DESA (3.55), and two historical gov't figures: 3.56M in 2002, and 3.2M in 2010). As noted above, I can't find the current gov't source in that link. The average of the relative current sources is 5.4M, with a range from 3.5-6.75M. If the population was 3.2M in 2010 (per the gov't stat), then most of those figures would imply growth rates of 6-8% annually, which seems really high. I'd propose text along on the lines of "The lack of a formal census creates challenges for estimating Eritrea's population; estimates range from 3.5M to 6.75M, with most clustering around 5-6M."BubbaJoe123456 (talk) 20:44, 5 March 2021 (UTC)[reply]

BubbaJoe123456 here is a new link [13] to the government 2020 6M citing. Also put in mind the WHO figures of 5M is five years ago, so in 2021 it would be more consistent with 6M and not the UN DESA figures. Here is a reference to 5,2M from the country department at WHO [14], from 2015. This seems to have been adjusted in the 2016 data set of WHO Global health observatory [15]. Leechjoel9 (talk) 02:49, 6 March 2021 (UTC)[reply]

That is not quite "the government", it is "Zeru Woldemichael Tesfazghi Chartered insurer, Chief Executive, National Insurance Corporation of Eritrea", and it's more of a personal essay than a government statistical office statement. He states 5.8 M, and he says, I was appointed General Manager ... I recognise we will make ... I have been a powerful advocate ... The source URL (archived) is also Eritrean. The date is unclear, but it refers to the COVID-19 pandemic, so it's recent enough for our purposes. Since no senior management in Eritrea is going to dare say something that could land him or her in prison and the 'Jesus Christ', 'helicopter' or 'number eight' position for a few weeks, it's quite likely that in some sense, this is an "official" value circulating in Eritrea right now, even if the National Statistical Office is not involved. Boud (talk) 16:27, 6 March 2021 (UTC) (minor fix Boud (talk) 16:30, 6 March 2021 (UTC))[reply]

Boud your last part of this post is quite telling about your antiEritrea POV. I say that because I assumed you were discussing this with an open mind. However the accusation you made mentioning alleged torture titles attributed to the Eritrean military. I can tell you are from the region specifically Tigray. Now the question is why is a Tigrayan like yourself so interested in changing the population of Eritrea from 6 million to 3.5 million in this article?Clownshking (talk) 17:04, 6 March 2021 (UTC)[reply]

The DESA number is completely wrong for the simple fact that Eritrea’s population since 1995 would’ve stayed at 3.5 million until 2021. The other sources are more accurate and based on factual data unless there is a bias against using government sources? If so, then the US census data should be disqualified? If not, then there is a bias against African governments.Clownshking (talk) 05:56, 6 March 2021 (UTC)[reply]

We have a government source (National Statistics Office, cited in the UNICEF report, saying the population was 3.2M in 2010, and another saying it was 3.5M in 2002. So, the gov't itself is saying that the population didn't grow at all, and maybe even shrank, from 1995 to 2010.BubbaJoe123456 (talk) 16:37, 6 March 2021 (UTC)[reply]

The UN DESA Revision 2019 data says 2204.227 thousand for 1995 and 3546.427 thousand in 2020. That's exp(log(3546.427/2204.227)/25) = 1.0192, i.e. about 1.9% growth per year on average from 1995 to 2020. That may sound low, but a significant effect in the Eritrean population is massive emigration, and we don't have numbers on the mortality from crimes against humanity that the Universal Periodic Review reports take place in Eritrea itself. If you look at pdf page 41 of the 2010 NSO report, you'll see that the population pyramid is severely distorted: there's a huge lack of men in the 20-50 age range. Biologically, the reproduction rate is constrained by the number of women, but sociologically, that would require polygamy, with men having children with at least 2-3 women partners each, around 2010-ish. If social preferences are against de facto polygamy or against single women raising children and keeping the sperm donors anonymous, then the lack of men will constrain population growth.

Keep in mind that according to UN DESA Revision 2019, there was a slight drop in the population from 1990 to 2000. See the graph of the UN DESA 2019 Revision data. This decade was just after the TPLF + EPLF overthrew the Ethiopian government; and Eritrea became independent in 1993. A big historical event can quite reasonably explain an unusual demographic event. Boud (talk)

After the declared independence (1991-1993) population was ~3,1M. After this the population grew coming years, and during 2002 the population was 3,5M which the Ministry of Information source above cites, not a local office. The population growth possibly followed the economic growth rate of the country, GDP more than doubled between 1995-2020. Some years the pop. growth has been more significant, so even if we entertain the idea of lower estimate in 2010 the population could still grown to WHO estimates (or one the other estimates) in this period of time. The best in this case is not to speculate but rely on the sources, which all besides UN DESA support a population in the 6M. To push for the single UN DESA source that is a outlier and is outnumbered should not even be considered as an option. Leechjoel9 (talk) 17:26, 6 March 2021 (UTC)

There is undue weight on the DESA source which completely disagrees with the majority of the credible sources. I think the Article with 6 million is NPOV as is and adding the DESA source would be pushing an obscure POV not based on consensused facts.Clownshking (talk) 18:06, 6 March 2021 (UTC)[reply]

The UN DESA Revision 2019 population for Eritrea in those three years are: 2.266356 million in 1991; 2.257593 million 1992; 2.238631 million in 1993. In other words, 2.27 million, 2.26 million, and 2.24 million. There was a slight loss of population during those three years according to the current best estimates by demographers. These are sourced values. See the file. The Eritrean National Statistics Office and the Ministry of Local Government are presumably national ministries . We don't know that, because nobody has so far created their articles. But "of Local Government" does not mean "local" in the sense of zoba or town or village. Boud (talk) 21:38, 6 March 2021 (UTC) (minor correction Boud (talk) 23:21, 6 March 2021 (UTC))[reply]

what is a zoba?Clownshking (talk) 21:49, 6 March 2021 (UTC)[reply]

It is a word that is used 378 380 times in the 2010 report by the Eritrean National Statistics Office. If you wish to understand something about the Demographics of Eritrea, then I suggest you have a browse through the 2010 report. The UN DESA 2019 Revision uses that report as one of its sources. For zoba, see Regions of Eritrea for details. Boud (talk) 22:06, 6 March 2021 (UTC) (correction: 380 times; two lines have the word twice Boud (talk) 22:20, 6 March 2021 (UTC))[reply]

• Option 2b: same as Option 2, but also including (in the body) discussion of and a graph of the 2010, 2012, 2015 and 2017 UN DESA Revisions together with the 2019 Revision. Here is a graphical justification of why we should include the UN DESA 2019 Revision, but also comment, using the sources, on the changes between the 2010, 2012, 2015, 2017 and 2019 UN DESA Revisions. This graph explains why all the sources cited above are compatible. The 2019 Revision is drastically lower than the earlier Revisions. But each of the sources such as COMESA, CIA, ADB are compatible with either estimates or projections using the older data files listed below through to 2020. For example, the CIA estimate is consistent with using UN DESA Revision 2015; COMESA is consistent with a linear extrapolation from the UN DESA Revision 2010.
  If you have difficulty with reading the files, then see libreoffice, export the files to csv format, and then grep the line containing "Eritrea". If you are using a computer that does not respect your freedom, then you may have to find other technical methods of reading the files.
  To summarise the changes between the Revisions: in 2012, the overall curve was raised; in 2015 and 2017 the curve was lowered; in 2019 the curves was dramatically lowered. We have a brief explanation in Release Note v1 (p2) of the 2019 Revision to explain why the drastic lowering occurred.
  One ironical aspect of this is that the UN DESA Revision 2019 is the first Revision that agrees with the Eritrean National Statistical Office 2010 estimate, which is the latest estimate of that sort that we have from Eritrea. In other words, UN DESA was quite slow to lower its time series to agree with the Eritrean NSO 2010 estimate, and appears finally to have accepted the Eritrean official results. Boud (talk) 01:08, 7 March 2021 (UTC)[reply]

Eritrea UN DESA Revisions 2010, 2012, 2015, 2017, 2019

Graphs are unavailable due to technical issues. There is more info on Phabricator and on MediaWiki.org.

Sources: UN DESA Revisions:

• UN DESA Revision 2010
• UN DESA Revision 2012
• UN DESA Revision 2015
• UN DESA Revision 2017
• UN DESA Revision 2019

To push for UN DESA source despite all other sources showing ~6M is trying to impose this view. We got to have some balance and should not be biased and favour this source. Consensus is 6M and should stay.Leechjoel9 (talk) 12:03, 7 March 2021 (UTC)[reply]

Clarification needed: @Leechjoel9: Please focus on the specific point instead of attacking me for "trying to impose this view". This part of the discussion is about Option 2b. Please avoid ambiguity in referring to the UN DESA Revisions. Do you have any argument against discussing and presenting the four new (in the Wikipedia discussion) sources: UN DESA Revision 2010, UN DESA Revision 2012, UN DESA Revision 2015, or UN DESA Revision 2017 in this Wikipedia article? You clearly wish to exclude UN DESA Revision 2019 (for no reason apart from "it's in the minority in my list"), and you appear to wish to exclude the Eritrean Population Health Survey 2010 (called EPHS2010 or PHS2010 in various places) made by the Eritrean National Statistics Office. (1) The 2010, 2012, 2015 and 2017 UN DESA Revisions are generally compatible with all the high projected estimates for 2020/2021; should they be excluded from the article according to you? (2) If yes, then why? (3) If no, then what source do you propose we use for the full history of the total Eritrean population from 1950 to 2020? Boud (talk) 12:55, 7 March 2021 (UTC)[reply]

We have always focused on the specific, as now. We have had this discussion for a long period and the point was not about targeting anyone but just pointing out the lack of support for the source. It has been explained to you in this extensive discussion here (multiple sections above) and in the Eritrea article. Eritrea became independent 1991, a chart with data prior to this is not relevant (N/A) and chart is not needed to explain this. Leechjoel9 (talk) 16:52, 7 March 2021 (UTC)[reply]

The territory of Eritrea existed prior to independence. The demographics of Eritrea would be more meaningful if they showed how the population of the territory developed prior to independence. This article currently has a lot of detail about ethnic categorisations - these cultural identities didn't suddenly come into existence in 1991. However, we could start the graph from 1991 from when the Provisional Government of Eritrea came into existence. The question of what year to start the data from is reasonably open to debate.

However, while you say We have always focused on the specific, your phrase the lack of support for the source is again vague and ignores the multiple sources that we have. There are now five different UN DESA Revisions that we have links for in this Wikipedia discussion, so we have five different UN DESA Revisions (demographic sources) and one NSO 2010 (PHS) demographic source. You cannot write "the source" for six demographic sources. You have not answered questions (1), (2) or (3). You seem to hint that your answer to (1) is "No". So in that case, please answer question (2). Why should we exclude the 2010, 2012, 2015, and 2017 UN DESA Revisions that are broadly consistent with the numbers on websites that are not demography research institutions (these are the ones that you consider "reliable")? Boud (talk) 20:29, 7 March 2021 (UTC)[reply]

I have many times described my views. Also the UN DESA is a single source and not five sources. The estimates of UN DESA is heavily outnumbered and contested therefore the population estimates for Eritrea should not be based on their numbers. We have six credible independent sources showing estimates in the 6M, also all consistent which each other. This is and all of the reasons given above is reasons to base the estimates on the other sources. Leechjoel9 (talk) 09:07, 8 March 2021 (UTC)[reply]

wikipedia:RSUW there is undue weight been given to the UN DESA source also Boud is doing Original research using the data in UN DESA source which disagrees completely with consensus of reliable sources. I don’t know How many ways this conclusion can be stated to the RfC question. The answer is NO the UN DESA source cannot be added or given weight in this article. How do we officially close this RfC, we need an admin?Clownshking (talk) 23:05, 8 March 2021 (UTC)[reply]

• Option 2 - include both estimates. (arrived via WP:AE thread) we have to mention the fact that the UN population estimate is so different from others (such as the CIA estimate) but there's no reason to emphasize it over other estimates. power~enwiki (π, ν) 21:56, 15 March 2021 (UTC)[reply]

Shouldn’t the latest data ie 2021 be used instead of data’s from 2010,2019?Facttell (talk) 22:37, 15 March 2021 (UTC)[reply]

• I lean towards option 2. (Came here from WP:NPOVN.) Certainly the UN DESA estimate should be included in the body, and I think it's probably worth including in the lead too, as the UN is a prominent reliable source. Of course we should make clear that this estimate does not represent the majority view among reliable sources, per WP:WEIGHT. —Mx. Granger (talk · contribs) 16:48, 16 March 2021 (UTC)[reply]

The UN DESA is the estimates is a outlier and the only minority source. Such views should not be presented as a majority view per undue weight WP:RSUW. If it needs to be included it should be given minor attention in the Body “Population” section with a row that at least mentions the UN DESA has a contested estimate of 3,5M for 2019. But as pointed out before, the majority view along other arguments provided puts the estimates around 6M. So the estimates in the lead, info box and body should be based on the other sources, but the UN DESA estimates could be mentioned briefly in the body section. Leechjoel9 (talk) 22:04, 17 March 2021 (UTC)[reply]

• Option 2 As per mentioned in above. Ali Ahwazi (talk) 08:40, 28 March 2021 (UTC)[reply]
• Option 2 per the comments made above.Sea Ane (talk) 10:20, 10 April 2021 (UTC)[reply]
• Option 2. I agree with reasons provided by BubbaJoe123456 - Idealigic (talk) 13:31, 13 April 2021 (UTC)[reply]
• Option 2, this is murky enough that it is best to list all the estimates and let the reader figure it out if they want to.VikingDrummer (talk) 07:59, 18 April 2021 (UTC)[reply]

Discussion

This subsection is for wider discussion or questions that would make individual replies to the RfC too complicated; add Option 0 or Option 1 or Option 2 (or similar) and your reasoning in the above subsection, not here. Boud (talk) 23:54, 15 March 2021 (UTC)[reply]

• @Facttell: No, we cannot use the latest (observational) data for 2021 because it does not exist (as far as any of us knows). The most recent survey (not a full census) that we know about is also the biggest that we know about, which is the 2010 Population and Health Survey (PHS) by the Eritrean National Statistics Office of 34,423 households. Demographers (not me!) make models of the population, taking into account lots of data such as birth rates, death rates, emigration and immigration rates, and make estimates of years in between the "measurements", and projections into the future. Boud (talk) 23:54, 15 March 2021 (UTC)[reply]

What is total population of Eritrea for 2020 according to your source?Facttell (talk) 01:59, 16 March 2021 (UTC)[reply]

I assume you're asking about projections from the full population histories of all the 2010 to 2019 UN DESA Revisions. Those Revisions did make projections; I haven't searched for the archive files of those projections (except for the 2019 Revision with a short extrapolation), but they're quite likely available. If, instead of finding those projections earlier than 2019, we extrapolate all the earlier UN DESA Revisions linearly for the 2010, 2012, 2015, 2017 Revisions and use the stated 2019 Revision for 2020, then we get projections (my guess by eye from the graph above) for the year 2020: 2010 Revision: 6.8 M; 2012 Revision: 7.5 M; 2015 Revision: 5.8 M; 2017 Revision: 5.3 M; and stated value in the .xls file for the 2019 Revision: 3.55 M. So the COMESA projection for the year 2019 of 6.75 M is quite reasonably consistent with UN DESA 2010 Revision, and the CIA projection for the year July 2021 of 6.15 M appears consistent with UN DESA 2015 Revision. So, for example, including the COMESA and CIA values along with the UN DESA 2019 Revision value would effectively give equal "due weight" to (1) the UN DESA 2010 Revision/COMESA 2019, (2) the UN DESA 2015 Revision/CIA 2021, and (3) the UN DESA 2019 Revision. The simplest explanation is that COMESA 2019 and CIA 2021 are badly out of date, and still based on the old Revisions. However, whether this reasonable explanation is enough to justify Option 1 or not depends on whether the interpretation counts as original research. Option 0, with the same interpretation, effectively says that only the older projections for 2020 are valid. Terminology: the year of a Revision is the year when the research was done, not the year of the population. Boud (talk) 03:03, 16 March 2021 (UTC) (no need for extrapolations; see below Boud (talk) 00:53, 17 March 2021 (UTC))[reply]

So what you are stating is that your Source kept projecting but not providing a verified estimate that agrees with the other sources? Can you provide a total population estimate clearly stated by your source, for example on the CIA website for Eritrea they give a total of some where in th 6.5 million estimate? I’m sorry but you’re source requires computation to come to a total number.Facttell (talk) 03:17, 16 March 2021 (UTC)[reply]

Clarification: In the terminology of demographic statistics, population estimates are for the past and population projections are for the future (I may have messed up this distinction earlier). Until there is a full census, none of these values could really be described as "verified". The CIA value for July 2021 is clearly a projection.

Projections (published): The medium variant projections from UN DESA Population Division are available at the URLs in the graph above, in the second table in each .xls file. So we get medium projections for the year 2020: 2010 Revision: 6.848 M; 2012 Revision: 7.727 M; 2015 Revision: 5.892 M; 2017 Revision: 5.432 M; and value in the main table of the 2019 Revision: 3.546 M. The COMESA values (projections?) for years 2018 and 2019 are 6.64 M and 6.72 M and the UN DESA 2010 Revision medium values for 2018 and 2019 are 6.544 M and 6.696 M (almost identical). The CIA projection for July 2021 of 6.147398 M (not 6.5 M) is just slightly higher than the UN DESA 2015 Revision medium projection for 2021 of 6.029 M. So the simplest explanation is that COMESA 2019 is from the UN DESA 2010 Revision and CIA 2021 is from the 2015 Revision, both making tiny corrections. The other websites, which state raw values prior to 2019, should be expected to be out of date compared to the UN DESA 2019 Revision, since they didn't yet have access to the new Revision.

Anyway, now you have the values clearly stated by the UN DESA Revisions. Please see libreoffice if you are unfamiliar with reading .xls or .xlsx files. So from the UN DESA Revisions 2010–2019, the 2020 projection varies in the range 3.5–7.7 M, starting high, getting higher, and then dropping. Boud (talk) 00:53, 17 March 2021 (UTC) (clarify Boud (talk) 01:13, 17 March 2021 (UTC))[reply]

Having to go through xls files to find what UN DESA source is stating the total population of Eritrea was in 2020 is requiring computation thus becomes original research. The other sources clearly state the total number for 2020. Projections that were made in the past should not be added, this page needs an estimate such as the CIA source shows. Facttell (talk) 01:48, 17 March 2021 (UTC)[reply]

Reading a file in a very standard format, readable with free-licensed software (and proprietary software, I assume), is not going through a computation. The COVID-19 pandemic pages are full of numbers from online files, and many of them are archived on Wayback Machine or Archive Today. The topic of this article is demography for a particular country. The results of demographic research are typically provided in numerical tables, graphs and files in standard formats. If you re-read my previous comment for terminology, you will see that the CIA does not have an estimate, it has a projection for July 2021, most likely based on the UN DESA 2015 Revision. If you reject projections for 2020 made prior to 2020, then you reject most (probably all) of the values that we know about.

If you feel that you understand enough about the demographics presented in the sources to make a decision and justify it, then you should add Option 0 or Option 1 or Option 2 in bold, together with your justification, in the list above. This will make it easier for an uninvolved person to close the RfC when there are enough !votes with reasons and a near-consensus based on Wikipedia policy emerges. Boud (talk) 03:01, 17 March 2021 (UTC)[reply]

@Facttell: If you look in the whole talk section there is a ongoing dispute and an RFC further up the page. I alongside others have argued to exclude Bouds UN DESA source (3,5 M estimate) since it contradicts sources of CIA, African Development Bank, WHO, COMESA and Eritrean ministry of Information, all of them citing or presenting 6M estimate.Leechjoel9 (talk) 07:15, 17 March 2021 (UTC)[reply]

@Facttell: To quote from WP:NOTVOTE: "it is 'not the vote' that matters, but the reasoning behind the !vote that is important". Boud (talk) 12:07, 17 March 2021 (UTC)[reply]

@Leechjoel9: I see the difference. If the Eritrean MoI is giving their estimate then that is as good as a census if the other sources match. The UN DESA seems to be the outlier. But I am not completely sure what the actual estimate is? For estimates from all sources for 2020?Facttell (talk) 16:02, 17 March 2021 (UTC)[reply]

• Procedural note Facttell is a block-evading account and has been blocked accordingly.-- Jezebel's Ponyo^{bons mots} 20:30, 17 March 2021 (UTC)[reply]

• COMMENT: Have not had time to read everthing involved in the RfC but I think I have relevant point.
  

Do we have any reason to believe that any of the other sources for the Eritrean population ..

1. CIA (2021 estimates), 6,14M [16]
2. COMESA (2019 estimates), 6,75M [17]
3. African Development Bank (2017 estimates), 5,5M [18]
4. WHO (2016 estimates), 5,2M [19]
5. Government source (2020), 6M [20]
6. Government source (2002), 3,5M to show consistency with other sources. [21]

... did their own research (surveys presumably) to obtain their estimate of Eritrea's population as the United Nations Department of Economic and Social Affairs (UN DESA) did?
Do we have any reason to believe that any other reputable NGO or any other body has done research to get a population estimate?
If not, it seems reasonable to conclude those sources are drawing from the earlier UN DESA estimate(s) (mostly or most likely 2017) that the UN DESA now believes is defective. Am I right?
If so it would seem option 1 is the best and at the very least option 2. (NOTE: Editor is a volunteer for the WP:Feedback Request Service which randomly selects volunteers to give feedback to WP:RfC) --Louis P. Boog (talk) 15:19, 19 March 2021 (UTC)[reply]

@Louis P. Boog: To your question Am I right?, my general impression is yes. Here is one example which is consistent with that: The current "WHO 2016" estimate says Total population | 2016 | 4 955 000 and the UN DESA 2017 Revision projection for the years 2015, 2016, 2017 says 4 847, 4 955, 5 069 (in csv format) for the medium variant, in thousands. So in this case, WHO 2016 exactly matches the 2016 value of UN DESA 2017 Revision (medium variant). BubbaJoe123456 pointed out above (twice) that WHO does not state '5.2M'. Boud (talk) 21:33, 19 March 2021 (UTC)[reply]

A second example: the ADB estimate of 5.5 million is consistent with the UN DESA 2015 Revision (medium variant projection), which gives 5 228, 5 352, 5 482 thousand for 2015, 2016, 2017. So: WHO 2016 is consistent to 4 decimal places with the 2016 value of UN DESA 2017 Revision (medium variant projection); and ADB 2017 is consistent to both significant decimal places with the 2017 value of UN DESA 2015 Revision (medium variant projection). Boud (talk) 22:02, 19 March 2021 (UTC)[reply]

We cannot just make up assumptions on how COMESA, African Development Bank, CIA collect their data. For instance COMESA reported a much higher estimate for 2019 (6,75M), nowhere near UN DESA. The CIA estimate is not consistent with UN DESA and they have not revised their estimate for 2020 or 2021. If the UN DESA estimate was so accurate why did they have the need to revise their estimate and lower the estimate with several millions? UN DESA is not the census taker of any country and all organisation do not rely on their data for statistics. The estimates in the near 6M is also supported by the Eritrean Ministry of Information (2020). At the moment no organisation has made the revision according to UN DESA, if they and the Ministry of Information change their estimates then this wouldn’t be an issue. But for now the UN DESA is contested and outnumbered in terms of sources so only right thing to do is briefly mentioning it in the body but the overall consensus in the article (lead, body, info box) should be a estimate in the 6M. Leechjoel9 (talk) 15:28, 21 March 2021 (UTC)[reply]

Comparison of sources

UN DESA Pop Div WPP estimates/medium variant projections versus other sources

Source	In-depth demographic research report available[a]	Year in which Eritrea has this population[b][c]
		2010	2015	2016	2017	2018	2019	2020	2021
UN DESA 2010	Yes	5 253.676	6 077	6 236	6 391	6 544	6 696	6 848	7 001
COMESA 2018 2019	No					6 640	6 720
UN DESA 2012	Yes	5 741.159	6 738	6 937	7 135	7 331	7 529	7 727	7 927
UN DESA 2015	Yes	4 689.664	5 228	5 352	5 482	5 617	5 754	5 892	6 029
ADB 2017	No				5 500
Eritrean Insurance manager 2020[d]	No							5 800
CIA 2021	No								6 147.398
UN DESA 2017	Yes	4 390.84	4 847	4 955	5 069	5 188	5 310	5 432	5 555
WHO 2016	No			4 955
UN DESA 2019	Yes	3 170.437	3 342.818	3 376.558	3 412.894	3 452.797	3 497.117[e]	3 546.427	3 601
Eritrean Population and Health Survey 2010[f]	Yes	3 200
^ In-depth demographic research report available means that an organisation has published a detailed online report of its demographic research including sources, method and results. ^ Bold values are used to aid in comparison with the non-US-DESA values. ^ Italics are used to identify 2010, which is non-consecutive with the other years. ^ Leechjoel9 calls this Eritrean Ministry of Information (2020), since it was republished on the Ministry of Information website shabait.com. ^ The reason why 3.497 million is about 1.8 million less than 5.310 million is briefly explained in the 2019 Release Note. ^ In-depth demographic study by the Eritrean National Statistics Office together with international partners. Page 31 of pdf: "No population census has ever been carried out in Eritrea. However, based on a population count by the Ministry of Local Government and NSO estimates, the total resident population of Eritrea was about 3.2 million as of 2010 ([Ministry of National Development], 2010)."

To clarify the relation between the sources, here's a table. Feel free to correct errors in the above table (add a signature below to help show that you modified it).

It looks like COMESA is using very old (2010) projections; ADB, the insurance manager ("Government source 2020") and CIA are using fairly old (2015) projections; WHO is using a moderately old (2017) projection exactly; the 2019 projection matches the Eritrean Population and Health Survey 2010 exactly; and nobody is using the highest UN DESA projection (7.9 million for 2021 from the 2012 Revision). Boud (talk) 17:23, 25 March 2021 (UTC)[reply]

This is making original research per WP:NOR. You left out for instance CIA sources that have estimate data for Eritrea dating back many years. I saw other errors as well, as mentioned before we can’t assume what methodologies that has/hasn’t been used or conclude that they are based on old estimate or on old UN DESA projections. All the organisations currently supports estimates in the 6M and so does the cited sources from the MOI of Eritrea. Leechjoel9 (talk) 22:15, 25 March 2021 (UTC)[reply]

The table does not "conclude" on the relations between the sources; the aim is to make it easier for an uninvolved person looking through this RfC to see the relations between the sources, and whether or not some or all should be excluded from the article, under Option 0, Option 1, or Option 2. Since you mention speculation, I do tend to think that speculating that the CIA does its own demographic modelling is a reasonable speculation; however, since the sources and method are secret, this is not even OR, it's pure speculation. In contrast, we know that UN DESA and the Eritrean Population and Health Survey of 2010 have reported on their sources and methods. Boud (talk) 22:58, 25 March 2021 (UTC)[reply]

• This is a very long RfC to sort through. It seems like the answer, whatever it is, should be much simpler. How about something like, "Sources disagree as to the current population of Eritrea, with some proposing numbers as low as XXX,XXX and others as high as XXX,XXX. Eritrea has never conducted an official government census." and just leave it at that? I don't think our readers really care about all of the details, and if they did, they wouldn't expect Wikipedia to give it all to them. I think we should admit that there are conflicting headcounts, give the range, cite some sources, and call it a day. And end the RfC. A loose necktie (talk) 07:03, 12 April 2021 (UTC)[reply]

  @A loose necktie: This comment looks very much like Option 2. Boud (talk) 22:48, 12 April 2021 (UTC)[reply]

• There seems to be a pretty broad consensus (universal among those who haven't been involved in the article, and a majority of those who have been, for Option 2. Quite a while back, I had suggested something like what A loose necktie recommends, and I still think it's the best path to take. Describe the uncertainty, and lay out the range of estimates.BubbaJoe123456 (talk) 23:53, 12 April 2021 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Do we have any sources apart from WPP2019_MIGR_F02_NET_NUMBER_OF_MIGRANTS.xlsx (UN DESA Revision 2019) for the historical migration data for Eritrea? Here's the graph from the UN DESA Revision 2019 data.

This was already in the article, since there is no reason to censor migration information from Demographics of Eritrea, but the content was deleted by Leechjoel9. Boud (talk) 23:18, 6 March 2021 (UTC)[reply]

Eritrean migration (immigrants minus emigrants; UN DESA Revisions 2010 to 2019)

Graphs are unavailable due to technical issues. There is more info on Phabricator and on MediaWiki.org.

Negative numbers indicate more emigration than immigration; positive numbers indicate more immigration. Sources: United Nations Department of Economic and Social Affairs (UN DESA) Net Immigration:

• 2010
• 2012
• 2015
• (not found yet: 2017)
• 2019

I've updated the graph with the 2010, 2012 and 2015 data. I haven't yet found the 2017 data. I have not added data from other sources, since nobody has proposed other sources.

The 2010 and 2012 are almost identical. The general changes are much less dramatic than the 2017 to 2019 overall population drop that matches the 2010 PHS. The biggest change I see is in the 2005-2010 net migration, which from 2012 to 2015 switched from net immigration to net emigration, and then in 2019 switched to a bit weaker net emigration. Anyway, it seems to me that the NPOV choice would be to show all the curves, especially in the context of the overall population drop.

@BubbaJoe123456 and Leechjoel9: Are there any objections to adding a migration subsection, with this graph and some appropriate text? As in the COVID-19 pandemic articles, graphs are a very useful addition to increase article quality; raw tables of numbers are more difficult to make sense of intuitively.

Clarification: I do not propose modifying or removing the one-line CIA datum that unfortunately is ambiguous, because we don't know if net migration in that one-line subsection means net immigration or net emigration, and it's a normalised value, which makes it highly uncertain because of the uncertainty in the total population, so less useful for the reader than an estimate of the un-normalised number. (The word "normalised" in this context means "divided by the total population".) Boud (talk) 23:12, 9 March 2021 (UTC)[reply]

 Done Boud (talk) 21:51, 10 March 2021 (UTC)[reply]

Eritrea population is roughly 6 million as the official reliable sources state. 3.5 million is an outdated incorrectly calculated number.Clownshking (talk) 18:10, 7 March 2021 (UTC)[reply]

On the contrary: 6 million is the old extrapolation for 2021, while 3.55 million is the most recent estimate for 2020. Please look at the graph in Option 2b of the RfC. See graph for an explanation of what a plot or graph is in this context. Boud (talk) 12:01, 8 March 2021 (UTC)[reply]

This original research which is against WP rules. Clownshking (talk) 17:18, 8 March 2021 (UTC)[reply]

Have a look at the COVID-19 pandemic articles. You will see many graphs. Boud (talk) 20:08, 8 March 2021 (UTC)[reply]

Current situation: In the current version of the article, we have several data tables (except for the first line of text about the total population) that give references without archived source files or html sources. The URLs are those of the UN DESA Revision 201x data releases (multiple files for each data release) listed above (where x is one of 0, 2, 5 or 7). They URLs mostly have the title "World Population Prospects 201x", and some have dates of access. The URLs clearly show that these are UN DESA Revisions.

Specifically (details below), we have these tables:

1. proportion of children – current usage: 2010 (checked)
2. Vital statistics section (births/deaths/natural change) – current usage: 2010 or 2012 (not yet checked)
3. Total Fertility Rate (TFR) – current usage: 2012 or 2015 (not yet checked)
4. Life expectancy – current usage: 2017 (not yet checked)

Question A: Since the UN DESA Revision 2019 is considered invalid by two users above, which of the UN DESA 201x Revisions (we can call these "World Population Prospects 201x", with x = 0, 2, 5, 7 or 9; these are just different names for the same thing) should we use for each of these four tables?

Question B: By those people who oppose the 2019 Revision, how should we explain our reason for continuing with one of the old data sets (2010, 2012, 2015 or 2017) and refusing the 2019 Revision, e.g. "The UN DESA 2019 Revision is unreliable because it disagrees with the older Revisions." ?

Clarification: a specific year 201x of a Revision refers to the whole data set over many years that was released in year 201x.

Details:

1. Population section: "The proportion of children below the age of 15 in 2010 ... ref name="WPP 2010" ... http://esa.un.org/unpd/wpp/index.htm|title=World Population Prospects, the 2012 Revision|access-date=31 May 2015"
   The first column "1 141 000 ... 5 254 000" shows that, as claimed, the content is clearly the UN DESA 2010 Revision data: 1141.3, 1236.248, 1423.599, 1619.301, ... 5253.676 (even though the title says 2012).
   The discussion then mentions the UN DESA 2012 Revision but the table is not modified. current usage: 2010
2. Vital statistics section, table of births/deaths/natural change: "Registration of vital events in Eritrea is incomplete. The Population Department of the United Nations..." – refers to the reference called "WPP 2010", but as before, the title of ref="WPP 2010" says it's for 2012. I have not checked the numbers, so they could be either 2010 or 2012. current usage: 2010 or 2012
3. "Total Fertility Rate (TFR)..." – I have not checked this, but the claim is that this is from "MEASURE DHS: Demographic and Health Surveys|access-date=31 May 2015" – as stated by UN DESA Revision 2019 there are both the the 1995 and 2002 DHS's, but UN DESA Revision 2019 also uses 2010 PHS, 2015-2016 ELFS and some other sources and reasoning and calculations. "30 May 2015" would seem to imply either the 2012 or 2015 UN DESA Revsion. current usage: 2012 or 2015
4. "Life expectancy" – "World Population Prospects - Population Division - United Nations|website=esa.un.org|access-date=2018-08-26" – I have not checked this, but the access-date seems to imply that we have the 2017 data. current usage: 2017 (apparently)

Boud (talk) 11:47, 8 March 2021 (UTC) (update to clarify that there is only one non-banned user who objects Boud (talk) 19:34, 9 March 2021 (UTC))[reply]

@BubbaJoe123456 and Leechjoel9: I see no answers to Question A and Question B. If there are no objections or serious proposals for which year's UN DESA Revision 201x data we should use (2010, 2012, 2015, 2017, or 2019), then I will update the tables and the text of these sections to the 2019 data. If there is an objection to the 2019 data, then we must have a proper sentence with a reference to justify to the reader why we reject the most recent data. We currently have a mix and match and the references are not archived, which forces any reader trying to see if the article matches the sources do an excessive amount of unnecessary work. Boud (talk) 19:45, 9 March 2021 (UTC)[reply]

For any material that relies on DESA data, there's no good rationale for not using the most recent revision. BubbaJoe123456 (talk) 20:05, 9 March 2021 (UTC)[reply]

There is no consensus for using the UN DESA at all at the moment. The only thing that needs to be updated are the figures for 2021. I have proposed to use the source of the CIA which is 6,14M which is a among the low estimate in the 6M range. Also some of the data tables in the articles are based on CIA data. The data with the old chart should be removed from the article. Leechjoel9 (talk) 20:25, 9 March 2021 (UTC)[reply]

Leechjoel9 As discussed previously, the CIA WF data is actually at the higher end of the range of estimates of total population (even excluding the 3.5M DESA estimate). So, is your suggestion to eliminate everything related to DESA? BubbaJoe123456 (talk) 20:38, 9 March 2021 (UTC)[reply]

@Leechjoel9: you wrote some of the data tables in the articles are based on CIA data. I do not see any tables in the current version that claim to be based on CIA data. Are you proposing that we delete all the tables in the article? Boud (talk) 21:36, 9 March 2021 (UTC)[reply]

The CIA (2021) is the latest estimate. So for instance a increase with ~1M from WHO (2016) which is the lowest one (ex.UN DESA) would be consistent with CIA (2021). It would only take an annual increase in pop. with ~4%, which is quite logical and close to the annual growth rate of the Eritrean population. But if we leave out some speculation and try choose the estimate in the middle (UN DESA ex.) it would be somewhere in between COMESA 6,75M (2019) and WHO 5M (2016). This would put the estimates around 5,8M which also the government source cites (2020). So CIA 6,14M (2021) is not far from this estimate and it gives us an actual presentable figure and source for 2021. The UN DESA is the only minority and should be excluded from the main parts (lead, info box, body) of the article due to undue weight per WP:RSUW. The estimates should be based on majority compatible facts and sources which puts the estimates in the ~6M. Boud, I was referring to the Demographics data which is from CIA, so that could stay. Only table that in this case would be removed is the population table (UN Data). Leechjoel9 (talk) 23:18, 9 March 2021 (UTC)[reply]

Thank you for confirming that there are no data tables using CIA data, and for confirming that you are only proposing to delete one table, not all of them. Since there are no objections to updating the Vital statistics section tables, I have done so. Boud (talk) 02:50, 10 March 2021 (UTC)[reply]

Just to use words carefully: the CIA 2021 estimate of 6.147398 million is not the latest estimate, it's the furthest-in-the-future estimate, excluding the UN DESA projections. July 2021 is in four months' time. Boud (talk) 02:50, 10 March 2021 (UTC)[reply]

There is CIA data on the article (Demographics) although not in table format. The current 5,7M in the article should be updated according to the new CIA 2021 estimate. Leechjoel9 (talk) 15:58, 10 March 2021 (UTC)[reply]

So in this section, we still have the unresolved question: should we remove the population table because Leechjoel9 objects to the UN DESA 2019 Revision based on a majority-of-websites argument for the projected current estimate of the total population? (This table includes the age distribution: children/mid-life/seniors: "Population aged 0–14 (%) Population aged 15–64 (%) Population aged 65+ (%)")

My answer to this question is 'no', we should not remove it, we should update it. I see no reason to hide the history of the population and age-distribution of Eritrea, both pre- and post-independence. The majority-of-websites argument does not apply in this case because we have old histories that are compatible with the majority-of-websites: the 2010, 2012, 2015 and 2017 histories are broadly compatible with the majority-of-websites-for-the-current-population (5 to 7 million for 2019/2021). So the majority-of-websites argument (using Leechjoel9's choice of which websites to include in the "vote") favours either the 2010, 2012, 2015 or 2017 UN DESA Revisions. However, BubbaJoe123456's opinion (with which I agree) is For any material that relies on DESA data, there's no good rationale for not using the most recent revision. So if we don't have any alternative data sets that someone can claim are reliable, the UN DESA 2019 Revision is the one that makes sense. We can NPOV it including the majority-of-websites-depending-on-how-you-count-them argument, or using words similar to what BubbaJoe123456 proposed above.

In any case, in this subsection of the talk page, the question is about the full history of the Eritrean population and its age distribution (as in the out-of-date table we currently have), not just the current total value. Boud (talk) 02:50, 10 March 2021 (UTC) (clarify Boud (talk) 11:11, 10 March 2021 (UTC))[reply]

The UN table (Pop. History) in the article also becomes an issue if we were to update it according to revised UN DESA source, since this minority source contradicts the majority sources that puts the population in the ~6M and not in the estimate of rev.UN DESA. So the right thing do in this case is removing the table. Secondly Eritrea became independent 1991, it’s incorrect to have such table dating back to 1950 when Eritrea was a region of Ethiopia and not an independent state, this data shouldn’t be applicable. Leechjoel9 (talk) 15:58, 10 March 2021 (UTC)[reply]

This an article about demography, not about the nature of a state (formally organised country). The people of Eritrea didn't land there from Mars in 1991.

So in terms of people discussing we would appear to have a majority of 2 versus 1 to update the table rather than delete it; I'm not sure if BubbaJoe123456 has an opinion on ignoring where the Eritrean people come from, i.e. pre-1991 data? However, in Wikipedia, we don't make decisions by pure majoritarianism, so we cannot just say that the 2/3 opinion is sufficient (it could easily switch depending on who else might join in). For the moment it seems that you (Leechjoel9) continue to insist that your majoritarian argument invalidates the 2019 Revision. So this point appears to be linked to the total population point, and probably we have to wait until more uninvolved editors look at the arguments presented so far and see if a consensus can emerge some time in the future. Boud (talk) 16:50, 10 March 2021 (UTC)[reply]

Eritrea was regarded as Ethiopia before its independence in 1991, data before that would be considered as Ethiopian population statistics which makes it a legitimate question. However if we we focus on the estimates It’s just not about majority views but also all other arguments provided to you in all talk sections above i.e undue weight WP:RSUW, consistent estimates of various sources, and estimates from organisation engaging with Eritrea and Eritrea itself. In this case it’s clear that it also would be pushing for a minority view and presenting it as a majority view. There are 2 supporting and 2 objecting. The consensus we have right now is current estimates. The issue now seems to be if the population history table should be excluded or not since we are not basing estimates on the UN DESA source. Leechjoel9 (talk) 07:40, 11 March 2021 (UTC)[reply]

See history of Eritrea for the history of Eritrea, e.g. "The boundaries of the present-day Eritrea nation state were established during the Scramble for Africa. In 1869[50] or '70, the then ruling Sultan of Raheita sold lands surrounding the Bay of Assab to the Rubattino Shipping Company.[51]" Ideally we should present the population evolution since around 1870: if we had the data from a reliable source.

The undue weight argument would imply that the main weight should be on UN DESA Revision 2019, as the best quality demographic source. None of the other sources discussed on this talk page, apart from 1995 DHS, 2002 DHS, and 2010 PHS, are by institutes that do demographic research. consistent estimates of various sources, See the graphs above, and please try to understand them. The 2010, 2012, 2015 and 2017 UN DESA Revisions are broadly consistent with the 5–7 million estimate for 2020/2021. But they are now outdated, replaced by the 2019 Revision. estimates from organisation engaging with Eritrea and Eritrea itself. The UN DESA 2019 Revision is consistent with the most recent estimate made by an Eritrean demographic institution (in cooperation with international partners): the 2010 Population Health Survey (2010 PHS). The 2010 PHS point appears precisely on the UN DESA 2019 Revision curve. Please look at the curve. Please look at COVID-19 pandemic articles if you are not familiar with graphs and their usage in Wikipedia.

we are not basing estimates on the UN DESA source No, we do not currently have consensus among editors to exclude the most recent, most reliable demographic source: the UN DESA 2019 Revision together with (consistently) the 2010 PHS. I recommend that you avoid ambiguity by referring to the year of the UN DESA Revision in this discussion, because you are effectively arguing in favour of the 2010, 2012, 2015 and 2017 Revisions and against the 2019 Revision. Due weight does not mean excluding the highest quality, most recent demographic sources.

You are correct that we do not have editorial consensus to exclude the population history table. Boud (talk) 10:02, 11 March 2021 (UTC)[reply]

As other pointed out in discussions the UN DESA is not the census taker of any country, it’s clear that it’s the least credible, contested and disputed in the case of the Eritrea population estimates and a outlier estimate. The other sources are also of credible and of high quality. You can interpret it the you want but the undue weight and neutrality has been put in context to the matter we are discussing. I suggest you read WP:RSUW. Leechjoel9 (talk) 11:05, 11 March 2021 (UTC)[reply]

• As other pointed out in discussions the UN DESA is not the census taker of any country, This is pedantically true but misleading: the UN DESA 2010, 2012, 2015, 2017 and 2019 Revisions are, for Eritrea, mainly based on interviews of randomly selected samples of people (what we could call "micro-censuses"); in particular, the 2019 Revision is based on: 1995 DHS: 5054 women + 1114 men; 2002 DHS: 8754 women; 2010 PHS: 34423 households;
• it's clear that [the UN DESA is] the least credible, contested and disputed in the case of the Eritrea population estimates and a outlier estimate. An ambiguous argument is difficult to progress on. If we interpret this as UN DESA 2019 Revision, then we have no evidence for this being less credible than the earlier UN DESA Revisions. If we interpret it as the pre-2019 UN DESA Revisions, then the phrase is clearly wrong, because the 2010, 2012, 2015, and 2017 Revisions are fully compatible with the 6 million projections for 2020/2021;
• The other sources are also of credible .. quality They are credible as being consistent with the old UN DESA Revisions and the 1995 DHS and the 2020 DHS, but they are not demographic research sources.
• and of high quality. They come from well-known major organisations, but those organisations are not demographic research institutes, which is why detailed presentations of methods and sources are not provided. So no, we cannot say that these are high-quality sources in comparison to the demographic research institute sources.
• I suggest you read WP:RSUW. Done. So I suggest you look, at, for example, this part: Care must be taken to establish that corroboration is indeed independent, to avoid an invalid conclusion based on uncredited origination. We have no evidence at all that ADB, COMESA, or the CIA have done demographically valid interviews with big population samples in Eritrea without cooperating with the Eritrean National Statistics Office, and without publishing the results. We have no evidence that they are independent from 1995 DHS, 2002 DHS and the UN DESA 2010, 2012, 2015 and 2017 Revisions. So we have no evidence that they are independent from one another for the purposes under discussion. Boud (talk) 21:32, 11 March 2021 (UTC)[reply]

• ADB is for 2017 and WHO is for 2016. These are dates before the UN DESA 2019 Revision. So they are consistent with the pre-2019 UN DESA Revisions. Only the CIA (July 2021), COMESA (2019), and insurance manager who mentioned a value without saying where he got it from (2021) give the 6 to 6.75 million estimate. It's fully credible that these organisations (CIA, COMESA) have been slow to update to the more recent research. Boud (talk) 22:01, 11 March 2021 (UTC)[reply]

Arguments have been explained to you countless times in all sections you have created here, in other articles and in other discussions. The others sources are many in numbers, are credible sources, consistent with each other and from actors engaging with Eritrea in various fields and Eritrea itself. Basing the population history table on the only minority source rev.UN DESA would be contradicting all the knowledge of the undisputed majority sources that concludes Eritrea population estimates is in the 6M. Leechjoel9 (talk) 03:28, 12 March 2021 (UTC)[reply]

Any objections to removing the NPOV tag? While there is still quite a bit of tidying that could (should) be done, I think the issue now is of quality problems, not NPOV. Boud (talk) 22:41, 12 May 2021 (UTC)[reply]

I'm OK with it. Agreed that the article still needs work, but the issues aren't NPOV issues. BubbaJoe123456 (talk) 01:03, 13 May 2021 (UTC)[reply]

I’m ok with removing the tag. The changes on this article is not an issue. I think BubbaJoe123456 made good changes to this article and tried to present it as fairly as possible with a minor issue. Everybody has agreed that the various sources estimates between 3,5- 6,7M. However majority source do point to estimates in the 5-6 M. The estimates of 3,5M UN DESA should not weight as much as the other sources combined (in terms of presenting it) so it should be presented as a minority source. It also makes more correct, Eritrea do not have two populations (3,5) or (5-6M), one of the estimated is more accurate. For that reason I say we should mention both estimates but present the majority estimates (~6M I.e CIA) as main (primary) estimate in the info box and lead in the Eritrea article and in the lead In this article. Leechjoel9 (talk) 15:17, 13 May 2021 (UTC)[reply]

The RfC did not going into judging which sources are the more serious demographic sources. I'm not going to repeat the detailed arguments already presented in the RfC (and in other sections above). Boud (talk) 16:01, 13 May 2021 (UTC)[reply]

This particular section is about removing the NPOV tag on this article, given that the recent edits are consistent with the result of the RfC and Wikipedia policy on NPOV and WP:CENSOR. I object to the proposal to favour the majority-of-a-certain-way-of-listing-sources argument in this article; I propose that you open a new section here if you wish to re-introduce that despite the huge effort that went into explaining things earlier; I again recommend that you read demography to understand something about where demographic information comes from. I will remove the NPOV tag since currently there are no objections. Boud (talk) 16:35, 19 May 2021 (UTC)[reply]

I swapped out the infobox (which was for an ethnic group, not a country), for the country demographic infobox. A lot of fields are still to be populated, although most of the data is in the article. BubbaJoe123456 (talk) 11:16, 17 June 2021 (UTC)[reply]

Eritrean Tigrinya are not 25% Muslim.

Eritrean Tigrinyas to 90% Christian. The highlands of Eritrea where the Tigrinya tribe are a majority a predominantly Christian.

Tigrinya people have been Christians for almost 1700-1800 years and survived the ottman and Arab invasions in east Africa because of geographical situation in the Eritrean highlands making it impossible for the jihadist to enter invade and control the highlands

Eritrean Tigrinya are predominantly Christian. — Preceding unsigned comment added by 2003:CB:4F07:7800:A89E:B8F2:9E8F:8C5D (talk) 06:38, 12 November 2022 (UTC)[reply]

Template:Largest cities of Eritrea has been nominated for deletion. You are invited to comment on the discussion at the entry on the Templates for discussion page. --Triggerhippie4 (talk) 10:06, 6 December 2021 (UTC)[reply]

Eritrean Christian and Eritrean Tigrinya population is undercounted on the Wikipedia page of Eritrea.

Eritrean Tigrinyas are 55%-60% of the population of Eritrea. And the Eritrean diaspora is predominantly Eritrean Tigrinya as well.

Eritrean Tigrinyas are over 90% Christian and up to 10% Muslim. Towns of the highlands like Asmara Mendefera Adu Qwala Seneafe Adi keyih Dekemhare Are majority Christian city. There even more churches than mosques in this areas.

The places in the highlands were named ether by Etnic tiginyas or Christians like Zaada Christian/ white Christian a suburb of asmara which is Tigrinya village. The people of the area of Asmara are called as Hamasien who ethnic tigrinyas and of Christian faith.

The highlands of Eritrea is the most populous region of Eritrea.

And Wikipedia claims that there 1.8 million Eritrean tigres desite the Tigre tribe is only 30% of the population of Eritrea

Besides that Christian people also exist in the other Eritrean tribes like in the kunama tribe Bilen tribe and Tigre tribe who originally were Christians as well but were converted to Islam by Arab and ottman invasions in Eritrea and East Africa 2003:CB:4F07:7800:A89E:B8F2:9E8F:8C5D (talk) 06:47, 12 November 2022 (UTC)[reply]