A new census is being released in Saudi Arabia for the first time since 2010, having been carried out by the General Authority for Statistics (GASTAT) in 2022.
Total population: 32.2 million
Saudis: 18.8 million people (58.4 percent)
Non-Saudis: 13.4 million (41.6 percent)
Median age (Saudi citizens): 25.5 (vs. 34 for non-Saudis)
Under 30 (Saudi citizens): 63% of the Saudi population
Buried in the announcements, however, was a slight revision to previous population statistics. In an Arab News story, for example, from GASTAT President Fahad al-Dossari:
“For that reason, we decided to use 2022 as a base year for us to do the backcasting – an international practice whenever there is a new census – as an exercise to the previous census. “We looked at 2022, and we backcasted based on the multiple sources of data that helped us estimate the changes and numbers of the census, especially from 2010 to today.” The backcasted 2010 data shows that the total population of Saudi Arabia has increased by 8.2 million (34.2 percent) since 2010; the number of Saudis by 4.8 million (33.8 percent); and the number of non-Saudis by 3.5 million (34.7 percent).
The GASTAT website for the 2010 census now warns:
This is not the first such revision. In a 2008 International Journal of Middle East Studies research note, Onn Winckler mentions that population numbers in the first two Saudi censuses (1974 and 1992) were regarded by some as “highly inflated,” and that numbers for the 2004 census came in substantially lower than previous population estimates – around 500,000 fewer Saudi individuals than the estimate for 2003. Winckler rejoiced that the new statistics “had probably solved the half-century debate on the Saudi citizenry number,” and used the new statistics to estimate that the 1992 census was probably off by about 700,000 in counting the population of Saudi nationals – around 6%.
Likewise, when GASTAT moved to measure the size of the labor market using government administrative data on workers (rather than inferring it from regular surveys of the workforce) some 2 million citizens dropped out of the workforce between Q3 and Q4 of 2016.
Revisions to the 2010 Census
Still, the extent of the latest “changes” is significant. The 2010 census gave Saudi Arabia’s citizen population as 18,776,510 and its non-citizen population as 8,459,646 – with Saudi citizens making up slightly less than 70% of the population. GASTAT now implies that more accurate numbers were around 14 million Saudi citizens (an overcount of around 34%) and nearly 10 million non-citizens (an undercount of around 14.5%). The 2004 numbers are almost certainly off as well; a very crude estimate (just assuming the same rate of population growth as shown in the 2004 and 2010 census years) suggests that Saudi citizens were overcounted by about 25% in 2004. Comparing the new census data to the last available mid-year population estimate shows how under-counting non-citizens and over-counting citizens continued to throw off population estimates through 2021.
All of the old population estimates on the GASTAT website have now been moved to a section called “Archived Data”: https://www.stats.gov.sa/en/1204. GASTAT released a full methodology for its revised 2010-2022 figures on the census website here: https://portal.saudicensus.sa/static-assets/media/content/20230529_Back-Casting%20Methodology.pdf?crafterSite=gastat-portal
Earlier reported numbers versus revised numbers for Saudi citizens. 2004 numbers are my own rough approximation.
Earlier reported numbers versus revised numbers for Non-Saudis.
Looking at the estimated population by region suggests that the over-count affected the number of Saudis recorded in much of the country, albeit especially the more mountainous Southern regions (where presumably some communities were harder to visit or had family members residing in Riyadh much of the time). Still, Saudis in almost all regions were overcounted by at least 30%.
[Update June 5] A few friends noted that the age composition of the Kingdom changed slightly between the last available population estimate and the 2022 Census. GASTAT seems to have had more accurate data for recent births (as well as deaths among younger Saudis) presumable due to better administrative record-keeping and closer monitoring of individuals while in the schooling system. Accordingly, the mid-year 2021 population estimate tracks the 2022 census quite closely for ages 0-19, drops off for individuals 20 and older (born before 2002), and then considerable for many of those aged 45 and older (those born in 1977 or earlier).
Commentary
Saudi newspapers have carried little commentary on the census overall thus far (though I will update as it changes). Al-Riyadh ran a laudatory editorial on June 2 , noting that: “There is no doubt that the reliability of the [census] numbers is an important tool for analyzing and understanding demography and developing the country’s economic and social policies and plan.” No mention of the past discrepancy despite refetring to the 2022 Census as “as the most comprehensive and most accurate in the history of the Kingdom.”
Meshari al-Thaydi mostly repeated GASTAT’s press release for Asharq al-Aawsat, commenting only on perennial concerns over there being too many foreigners in the Kingdom, gingerly referring to the Crown Prince’s hope that eventually Saudi Arabia would be a 50-60 million person country with as little as “half” the population being Saudi citizens. (Update: even this last point was included in some of GASTAT’s press releases)
On Twitter, initial reactions often involved jokes about 1.2 million Saudis disappearing:
عمرك سمعت بدولة يطير منها مليون مواطن كل سنة 😂#تعداد_السعودية_2022 pic.twitter.com/qjWeAfZvgp — خط البلدة 🚐 (@saudibus222) June 1, 2023
Inevitably, the number of foreigners in the country came in for criticism as well, particularly the 2022 census indicated that Saudis were a lower share of the population (<60%) than the original 2010 census data (~70%) despite numerous departures of foreigners in recent years.
#تعداد_السعودية_2022 موضوع ان المقيمين زادو بحوالي ١٣ مليون هو طبيعي لان في مهن وأعمال مايشتغلها الا الأجانب خصوصا ان عندنا مشاريع تنموية ولازم فيها أعمال بناء وذي قائمة على الأجانب وكذلك التنظيف ولا ننسى ان في مهن قاعدة يتم توطينها والآخر تم توطينه pic.twitter.com/LcRsVHtWmm — Hosam (@hosam9570) May 31, 2023
GASTAT quickly put out a “correction” on social media addressing some of the controversy, noting (summarizing here):
There is no decrease in the number of Saudis between 2021 and 2022
The population actually increased by 1.4 million between 2021 and 2022
GASTAT had published revised 2010-2022 numbers
This is all standard practice for census agencies
30,000 Saudi field investigators took part in gathering census information
The number of Saudis increased by 34% between 2010 and 2022 (4.8 million)
#تنويه | توضيحات حول نتائج#تعداد_السعودية_2022 pic.twitter.com/sotRD8WWjn — محمد الدخيني (@mdukhainy) May 31, 2023
GASTAT also appears to have hired some influencers to talk up its professionalism on social media?
كيف ساهم الذكاء الاصطناعي الذي اعتمدت عليه الهيئة العامة للإحصاء @Stats_Saudi في تقليص نسبة الخطا الى ٥٪؟! وهذي النسبة تعتبر ضئيلة جدا مقارنة ببقية دول العالم؟ احنا نعيش في دولة تتحدث فعلا بلغة المستقبل 😃🫶🏻🇸🇦#تعداد_السعودية_2022 pic.twitter.com/ifkjEdHLBz — Amani A. Alajlan (@AmaniAAJ) May 31, 2023
…and possible reaching out to select influencers wondering where all the Saudis have gone.
جمعني اتصال هاتفي مع مسؤول من الهيئة, ذكر أن المنهجية تغيرت, و أصبحت أكثر دقة.. بالتالي تم تصحيح الأرقام. — Abdullah AL-Khmais (@AZK_SA) May 31, 2023
حياك الله نسعد دوماً بتواصلكم، ونفيدكم انه بإمكانكم الاطلاع على حساب المتحدث الرسمي للهيئة بخصوص توضيحات حول إعلان نتائج تعداد السعودية 2022م https://t.co/bFAD0Mmj6l — الهيئة العامة للإحصاء (@Stats_Saudi) May 31, 2023
A few commentators went more in-depth. Fahad al-Bogami, an academic in information technology at King Saud University, posted a video of himself asking Fahad al-Dosari (head of GASTAT) where the “missing” Saudis were:
“Most people thought the estimate of the number of Saudis was ~20 million. Today you announced 18.8 million. There is a difference there, one that has generated a reaction on social media… Have we lost people, or were the previous numbers a mistake?” “We neither lost people nor is it a mistake. Whenever there is a census you try to be comprehensive. Use of (new) techniques gives us a better opportunity than before. The (old) techniques used in 2010 didn’t look the same as today. But we cannot release the numbers today, i.e. 2022, without going back to change what is in the past. What we see now – maybe some advice for all – look and you’ll see the number of Saudis hasn’t fallen. Instead, it increased – there’s an increase. You can’t compare and say ‘we expected 20 relative to 18.8’ – I don’t agree with this. Look at 2010 and what was the situation of Saudis? Non-Saudis are the one where there has been variation [increase and decrease]. And this is realistic – non-Saudis depend on the economic environment. There have been changes in the Kingdom such as Corona, etc. [Cut in the Video Clip] This is something where we’re being transparent with all. This is what we’ve done! If I’m a foreign investor, an international organization, I want to see – have you all applied the best measures? This is what guided us. [Missed next sentence] We didn’t bring something new. If there was something we could have done that was better, we would have done it. I was speaking to the statistical authority in Britain and elsewhere, and when I explained to him what we were doing they said ‘You’re lucky.’ They said you have a foundation that you can build on [in Saudi Arabia].” Source: https://twitter.com/FahdAlbogami/status/1663880513205960704
Economic advisor Eid al-Eid appeared on Al-Ekhbariyya television station to discuss the results on June 1, faulting the reliance on “home visits” to collect data for the “discrepancies” between the 2010 and 2022 censuses. Per Eid, this might have thrown off the count due to Saudis owning homes in multiple regions of the Kingdom, being double- or even triple-counted as a result. By contrast, he perceived the newer count as more accurate due to a greater reliance on government apps such as Absher to collect data. Still, he described the population numbers as “very weak” given the developmental goals of the Kingdom. “The government must find solutions… to increase the number of births or speed up marriage.”
المستشار الاقتصادي "عيد العيد": أطالب بإيجاد حلول فورية لانخفاض أعداد الأسر وقلة المواليد بعد صدور إحصائية #تعداد_السعودية_2022. pic.twitter.com/rDFAtNJ5hC — هاشتاق السعودية (@HashKSA) June 1, 2023
Economic consultant Abdullah Alnasser also has a comprehensive thread discussing the differences between post-2010 population estimates and the new data, while also discussing GASTAT’s methodology.
بحسب مؤتمر الهيئة فإن عدد السعوديين كان 14 مليون في 2010 وهو رقم مخالف لعدد التعداد السكاني في ذلك العام والذي بنيت عليه كل تنبؤات النمو في اخر عقد زمني! بل مخالف لآخر مسح سكاني (لم اطلع بعد ع كامل التقرير) – وهذا يعني أن معدل الولادة تقريبا 24 لكل 1000 سعودي، ويعد رقماً عاليا! https://t.co/boOguOpc9D pic.twitter.com/xtUJUFsDHu — Abdullah Alnasser (@AbdullahLN) May 31, 2023
In Conclusion (for Now)
There is a sizeable literature out there on the reliability of statistics from governments, particularly from non-democracies, with a heavy focus on China. (See, for example, Jeremy Wallace’s excellent book from 2022 on the subject, Seeking Truth and Hiding Facts). Still, this literature tends to focus on strategic reasons for fudging the truth–less so on revealing massive discrepancies in the past. GASTAT’s explanation does have some face validity – why reveal a massive error in the past unless it’s about demonstrating ability to collect better data in the present?
Still, without open discussion of past census practices and independent analysis of the census results by Saudi scholars, it is hard to know if GASTAT has “solved” the problems of census-taking in Saudi Arabia any better than they did in 2004 or 2020. Hopefully, for the sake of economic and policy planning in Saudi Arabia, the next census isn’t too long in the making. There’s no telling exactly how the past statistics might have shaped policies in the Kingdom for a decade, or what an overreliance on select projections or statistics might do to policymaking at present.