Are the Arba'in pilgrimage numbers Iranian propaganda?

On the 'largest peaceful gatherings' and the difficulty of estimating

A ‘fact’ that I heard sometime in my teenage years, and that you might have heard too: the Hajj, the famous pilgrimage to Mecca that devout Muslims are religiously bound to go on once in their lives, is not even the biggest Islamic pilgrimage! There’s a bigger one in Iraq.

This was one of those ‘facts’ that made me go ‘huh, that’s interesting’ but I never actually incorporated into the rest of my knowledge; it just kind of hung out in my subconscious memory as a thing I’d heard once. Only recently did it bubble back to the surface, at which point I realised how weird it was. If there was another Muslim pilgrimage that was bigger than Hajj, how come I’d not heard of it?¹ How come it didn’t figure in my understanding of the religion—or, for that matter, my understanding of the world? How did it come to be that I heard of it once as a ‘fun fact’, and never again?

So: first to Google, which soon takes me to the Wikipedia ‘List of largest peaceful gatherings’. And sure enough, Hajj is not the largest regular Islamic pilgrimage, not even close. According to this list, in 2017, 30 million people attended the annual Arba'in pilgrimage, where Shia pilgrims travel to Karbala, Iraq; meanwhile, there are only about 2-3 million Hajj pilgrims on the busiest years.²

Wait a second. Thirty million people? That’s insane! That’s a huge number! And the 2017 Arba'in pilgrimage wasn’t a one-off, either: the Wikipedia list tells me there were 27 million pilgrims in 2015, 22 million in 2023, and 21 million in 2022. The only peaceful gatherings in human history with more attendees were a number of Kumbh Mela pilgrimages in the Ganges basin.

Now, more than 600 million people live in the Ganges basin, the majority of them Hindus, so a Hindu pilgrimage in the Ganges basin really is a plausible candidate for the title of ‘largest peaceful gathering in history’.³ But I had to ask myself, is it really possible that the second largest ever is a Shia pilgrimage in a second-tier Iraqi city? Can it even be bigger than Hajj?

First thought: there are about 2-ish million permanent residents of Mecca (I think),⁴ and its infrastructure has famously had trouble coping with the number of people who arrive every year for Hajj, with the Saudis having to begin a large and very controversial building programme. But there are only about 700,000 people in Karbala.⁵ If a city of two million struggles to support an extra two million extra people arriving all at once every year (which seems plausible), how on earth does a city of 700,000 people manage to cope with twenty million arrivals?⁶ Keep in mind, too, that Saudi Arabia is pretty politically and economically stable; Iraq has, um, not enjoyed the same level of stability over the last 20–30 years. Such stability is usually a prerequisite to consistently scaling up infrastructure.⁷

Second thought: going by the CIA World Factbook, there are probably around 25 million Iraqi Shia Muslims.⁸ So that means that some years, the number of people on this Shia pilgrimage in Iraq exceeded the number of Shia Muslims living in Iraq. Now, obviously people from elsewhere can and do travel to Iraq for a pilgrimage: the obvious source of Shia pilgrims is Iran. But then, there’s at most 90 million Shia Muslims in Iran (again going by the CIA Factbook). For the Wikipedia numbers to be true, the number of pilgrims in 2017 would have been equivalent to a quarter of all the Shia Muslims in Iran and Iraq combined having gone on the pilgrimage. And then a similar feat being repeated every single year.

Third thought: the Arba'in pilgrimage is specifically a Shia pilgrimage, and Sunni Muslims have no truck with it. Only about one in ten Muslims worldwide is a Shia Muslim; that’s maybe 200 million-ish people? many of whom don’t live anywhere close to Iraq. But the Hajj is a pilgrimage for all Muslims, Sunni or Shia—indeed, it’s a religious requirement for all Muslims that they go on Hajj at least once before they die! Hajj is just drawing from a much larger pool of possible pilgrims, and it’s religiously mandated; it would be surprising, to say the least, if the Arba'in pilgrimage was bigger.

Surprising things can be true, of course and I could have tried to dig more into the specifics to try to verify it. (A thought might be: the Saudis have acted like monopolists when it comes to Hajj, restricting the number of people who travel to drive up costs and target especially wealthy pilgrims; maybe in a ‘perfectly competitive pilgrimage marketplace’ Hajj would be the biggest Islamic pilgrimage, but the ‘marketplace’ isn’t perfectly competitive.) But I had looked a bit further down the list, and I had a bigger question.

The list of largest gatherings is overwhelmingly dominated by Shia religious events, to a crazy degree. Of the events on the Wikipedia list with more than ten million attendants, about three-quarters are Shia pilgrimages, religious gatherings, or funerals of Shia leaders (Ayatollah Khomeini); and if you rule out the events that happened in India, the world’s most populous country where (obviously) major gatherings are pretty common, literally all of the remaining gathering on the list are Shia religious events.

Even if Shia Islam the most pilgrimage-y of all the major religions, it is relatively small? As I said, there are maybe 200 million Shia Muslims; compare that to Catholicism, which also is famously pilgrimage-y, and has easily in excess of a billion adherents.⁹ The largest Catholic event on the list is the most recent Translación in Manila, which is said to have drawn a crowd of about 8.1 million. That number is huge (though not entirely implausible). But this is the largest (peaceful) Catholic religious event in history! We’re being asked to believe that Shia Islam regularly puts on events you’ve never heard of, drawing an average of two-and-a-half times as many people as the largest ever Catholic gathering?

No. It simply cannot be true. And so, my alternative hypothesis: this is (possibly Iranian) propaganda.

See, if you look into it, it seems that the Arba'in pilgrimage was de facto banned for many years by Saddam Hussein.¹⁰ The 2003 pilgrimage after his overthrow was the first pilgrimage in decades, and it was by all accounts genuinely massive, with a huge influx of pilgrims from across Iraq and—importantly—Iran.

Was it ‘thirty million people’ massive? No, definitely not. But how big exactly is a tricky question. After all, there’s no way to just individually count the people who attend a given event, starting at one, then two, then three… until you get to ~several million. You have to extrapolate from limited data, make your best guesses about key variables, and maybe you’ll get something with one significant figure of accuracy if you’re lucky. If you have more sources of information (drone footage, economic data, etc.) you can reinforce an estimate, adjusting it from multiple angles for higher confidence; but we obviously don’t have a lot of high-quality sources of data for immediate-post-invasion Iraq.

If you want a sense of how many difficult questions there can be in estimating how many people attend an event, just look at the footnotes of this very post, where even my entirely-slapdash non-scholarly efforts have required loads of qualifications and explanation. If I had to pick a number for 2003, I might follow this guy who says two million: I’m not sure what his source is, but it seems around the right ballpark.

But very quickly after 2003, the sources for the pilgrimage’s size start to get … much less trustworthy. The numbers that are issued every year come from the ‘shrine authorities’ who control the Shia religious sites in Karbala—a remarkably slippery group to get a handle on. Admittedly, I don’t read Arabic, which limits my research abilities here, but it’s still remarkable how hard it is to figure out who these clerics actually are. But regardless, we can learn some things about them based on how they behave. This Oxford academic was apparently told that the numbers are ‘calculated by the shrine using AI technology’. This suggests that Shia clerics in a mid-sized Iraqi city have been at the frontier of AI applications for over a decade; or maybe —just maybe!—that those Shia clerics are willing to tell self-aggrandising lies to credulous foreigners when they can get away with it.¹¹

And notably, many sources try to bolster the reliability of clerics’ numbers by providing a trustworthy lower bound: there are ‘at least two million Iranians whose numbers are easier to calculate because they require documentation to enter Iraq’, wrote the Independent in 2017, and many other news orgs provide similar citations. If there really were two million Iranians going on the pilgrimage every year, then total annual numbers in excess of ten million would be much less shocking. But who does the actual ‘calculating’ of the number of Iranians on this pilgrimage? Why, the Iranian state, of course!¹² How kind of the Iranians, to step in with helpful numbers when Shia clerics elsewhere in the world need them.

So here’s my (evidence-free) guess at what’s going on. Shia authorities and the Iranian state saw the 2003 pilgrimage as a propaganda victory for Shia Islam and (as a result) Iranian interests. So they started releasing numbers every year, through the clerics who run the shrines in Karbala. Of course, clerics who run pilgrimage sites are not actually equipped with the skills to properly estimate attendance numbers: that takes a whole different set of skills. But they didn’t need to be, they just needed to pick a big number that sounded good and (crucially) hope that people would trust them.

Initially, they started modestly, inflating the numbers by a few million at a time. But they began to notice that nobody ever doubted them. So they got more aggressive: by 2010 they were saying 10 million, and then in 2014 they tried 20 million. And still, nobody doubted them. The Wikipedia page for the pilgrimage got a section entitled ‘Among the largest annual gatherings’ (!), and prominent news sites (relying on Wikipedia) started putting the pilgrimage on lists of ‘the ten largest gatherings in human history’. They started inviting Oxford academics, and telling them they used cutting-edge AI to generate accurate numbers, (and presumably pissing themselves laughing at the idiot westerners afterwards,) and still nobody doubted them. So they said ‘fuck it’ and declared that they’d hit 30 million in 2017. And still, nobody doubted them.

And so, they went even further, and started inflating the estimates of a whole load of other Shia religious events. Go through the Wikipedia largest peaceful gatherings list: with the exception of the Ayatollah’s funeral, all the Shia religious events are from the last fifteen or so years. And yet there are so many Shia religious events on the list. They absolutely dominate the list! The hypothesis ‘the numbers on the list reflect probably-somewhat-inaccurate but basically unbiased attempts at figuring out the largest events’ does not explain this fact; the hypothesis ‘half these data points are Iranian propaganda’ absolutely does.

If you can read Arabic or (indeed) Persian, you might actually be able to investigate and find evidence for or against this hypothesis, rather than the pure speculation given above. In the spirit of openness, I present the strongest evidence against my own hypothesis that I’ve been able to find: this 2008 press release from Coalition forces, who obviously didn’t have any interest in spreading Iranian propaganda, saying that nine million people showed up to celebrate Arba'in. But then, would Coalition forces necessarily have bothered questioning the ‘official’ local estimate, especially when (as evidenced in the press release) they could spin it for their own purposes?

Whether this is deliberate propaganda or not, this is a sad example of the low quality of information in the public sphere. ‘The people who run a local religious shrine’ simply cannot be reliable estimators of the number of pilgrims who arrive: estimating that kind of thing requires specific skill and knowledge, even when there actually is high-quality and consistent footage of the event, something that is notably lacking when it comes to the Arba'in pilgrimage.¹³ Yet, if you can put a number out into the world with a source attached to it that isn’t obviously untrustworthy, and you can manage to be the only number available, the media will pick it up and run with it, preferring something over nothing. And if you can get it onto Wikipedia, even better—journalists’ only instinct to plausibility-check stuff now is to look at Wikipedia, and (indeed) even ‘prestige’ outlets like those linked above sometimes decide to just cannibalise a Wikipedia article and release it as if it were one of their own.

Indeed, I would guess that many of the other events on the Wikipedia list, especially the Hindu pilgrimages, are subject to the same phenomenon: the numbers are bullshit—but if you’re the only one who has a number, then bullshit or not journalists will use it and people will trust it. This lesson applies more generally.

At the end of all that, my best-guess answers to the concrete and specific questions here:

Was the Arba'in pilgrimage in 2003 bigger than Hajj 2003? That seems eminently likely to me. 2003 was just another year for Hajj, no particular importance; but Arba'in 2003 was the first after the overthrow of Saddam! I’m sure it drew lots of Shia Muslims from all over for the special occasion.

Was the Arba'in pilgrimage in 2024 bigger than Hajj 2024? Eh, not impossible, but I think it’s unlikely. The main consideration in favour of this conclusion is that the Saudis are definitely, to a significant degree, exploiting their monopoly over Hajj by making it available to fewer tourists and thereby driving prices up. But the sheer number of Sunni Muslims compared to Shia Muslims, the central religious importance of Hajj for both Sunni and Shia Muslims, and the logistical / infrastructure questions make me think that probably Hajj is bigger. I could definitely be convinced the other way by substantive evidence, though.

Are 20–30 million people going on the Arba'in pilgrimage every year? Absolutely fucking not. There is zero reason to believe this.

1. To be clear about what I mean by ‘biggest pilgrimage’, I mean specifically a single gathering with the most individual people. So, for example, across the whole year, many more people go on Umrah than go on Hajj; but that’s because Umrah isn’t really a single gathering, you can go pretty much any time of the year. There is no single week during which the number of people in Mecca for Umrah exceeds the number who are there during Hajj week.

2. Saudi officials said that 2023 was the ‘largest Hajj in history’ with around 2.5 million people; but then, in the previous decade they were citing even larger numbers, over 3 million in 2012. Obviously, you shouldn’t trust the Saudis, the exact numbers are probably unknowable; but even if we know for sure they’re lying at some point, 2–3 million seems accurate enough as a range for the biggest Hajj-es.

3. To be fair, I think there’s reason to doubt the numbers given on the Wikipedia page. More anon. But still, it’s not prima facie crazy that this would be the biggest gathering.

4. Trusting the Saudis to count population accurately is a fool’s errand, but Mecca at least has the advantage of not having an immense underclass of de facto enslaved foreigners who the state has every incentive not to count accurately. Oh, to be sure, there are definitely some de facto enslaved foreigners in Mecca, but the ban on non-Muslims entering means there just aren’t as many of them, so the numbers are less inaccurate. 2-ish million sounds about right to me.

5. Iraqi population numbers are, for very obvious reasons, even less reliable that Saudi ones. I’ve seen anywhere from 550,000 to >700,000. But regardless, it seems that everybody agrees that the city is definitely under a million people, less than half the size of Mecca.

6. They don’t necessarily all stay in Karbala (the way you have to stay in Mecca for a week on Hajj), but they do need to stay somewhere? And it seems that a large proportion of them do stay in Najaf, another city in Iraq with a similar population, and then walk to Karbala, Camino de Santiago–style.

7. Iraqi infrastructure is a stupidly complicated topic that I have zero qualifications to opine on. Do not take this as a well-informed take; this was literally me wondering aloud about the plausibility of some numbers with zero research, just background knowledge.

News reports on the Arba'in pilgrimage that bother to ask questions about infrastructure big up the voluntary provision of food by ‘charities, mosques, and devotional groups see to it that no traveller goes hungry’. I’m fucking sorry, but even armies can run into trouble sourcing and supplying enough food for twenty million people marching from one city to another. (Supply chain considerations are absolutely vital for military organisation for exactly this reason!) If you could show me evidence that a bunch of disconnected charities and mosques run by random religious people and do-gooders manage every single year without fail to (a) find enough food, (b) finance the purchase of it, and (c) supply it effectively to twenty million people who have suddenly arrived in a region noted for its instability where the biggest nearby city is ~700,000—well, I would renounce my Hayekianism instantly and move to Karbala, to throw myself at their feet and beg them to teach me how they have managed to solve the fundamental problems of human organisation better than any system ever seen before.

(The only way something like this could be true is that some maybe-quasi-autonomous arm of the Iraqi or, more insidiously, Iranian state is coordinating the efforts, disguising its contributions as ‘charities, mosques, and devotional groups’ for propaganda purposes—see below.)

8. Though see note 5: Iraqi population numbers are not very trustworthy.

9. The percentage of devout adherents will obviously be higher for Shia Islam than Catholicism, but probably not enough to offset a >5x base population difference?

10. Even this claim, which should be simple and straightforward, is hard to source beyond a shadow of a doubt. But it gets repeated a lot.

11. The utter credulousness of the rest of the academic’s report—apparently he completely swallowed the idea that Sunni Muslims and Christians go on this pilgrimage too (!!!), presumably fed to him to make the whole thing look like an anti-sectarian interfaith effort—might lend some additional credence to the latter hypothesis.

12. Sometimes ‘authorities in the two countries’, both Iran and Iraq, are cited jointly. But obviously (a) it’s not that hard to find an ‘authority’ somewhere in Iraq that’s getting its de facto marching orders from Tehran, and (b) the Iraqi state is famously bad at making the people who are within its borders legible to itself, so I’m not sure how it would even know. If I knew exactly which Iraqi authorities were providing these numbers I might be more willing to lend it credence.

13. There’s lots of low-down shots of a small handful of people, or slightly-higher shots of individual streets or squares, but nothing that could ever be used to construct an actual estimate.