How many have died in Syria?

The Syrian Observatory for Human Rights (one of many bodies attempting to report death tolls in Syria), has released a report that claims that some 115,000 have been killed in Syria since the beginning of the war ( Among other things, the summary by Mark Seibel for McClatchy notes that “pro-Assad forces continue to make up the largest group of the dead – around 41 percent, followed by civilians, who may be on either side, or neither. Rebel forces make up the smallest fraction of the dead, just under 21 percent. It is simply incorrect when Western officials imply that the Syrian government has killed more than 100,000 of its own people.”

It certainly would be unusual if in any war all deaths were directly caused by the regime in power. I doubt that ever happened. It is not usually the way one judges overall responsibility for deaths in war. But there is far more to delve into in relation to these figures.

First, let’s assume we take the report by the Syrian Observatory at face value, and accept that some 40 percent of the dead are either regular Syrian troops or pro-regime paramilitaries. It’s a sad fact to contemplate, but it is true that when a regime wages war against its people and they fight back, regime troops get killed. I know that may sound quite remarkable, but in fact Syria is not the first time that ever happened. Rather a lot of South Vietnamese troops were killed during that war for example (indeed, lots of army and police were already being killed in the late 1950s by NLF …. “terrorism,” long before the mass intervention of either the US or the northern Vietnamese regular forces.

As for calling the regime troops “pro-Assad forces,” or even more, the heading of the Observatory’s June report – that 43 percent of the dead were “Assad-backers” ( – that seems an entirely unwarranted way to describe regular army troops that haven’t yet deserted. Unfortunately, in such wars, many of those killed are on the front lines still “doing their job,” for reasons such as to earn their daily bread, for example.

Second, again if we accept this report, that some 40,000 non-combatant civilians have been killed: who are most of these? To simply claim that civilians “may be on either side” seems to obscure the issue. Is it likely that most might be people massacred by long-range missiles, being bombed from the sky by military aircraft and helicopter gunships, tanks, cluster bombs, long range artillery, and the various enormous massacres from Houla last year to Banyas in May this year? Or are most likely to be the victims of the smaller number of smaller scale massacres committed by Al Qaida or rogue rebel elements using light arms?

The answer is obvious; indeed, the qualitative difference in scale between the systematic crimes of a massively-armed, entrenched state apparatus, and the non-systematic crimes of rogue elements of a guerilla army, is kind of the whole point of the distinction some of us among the left are drawing in this case – as in every other case I’ve ever been aware of. The fact that some on the left don’t make this distinction only in the case of Syria is the curious issue.

What did we say, for example, when during the Second Intifada the kill counts continually showed a Palestinian majority of dead (just as here there is a “Sunni” or “anti-Assad” majority in these figures), but Israeli deaths still hovered around a quarter of the total, nearly 1000 of 4000 killed in 2000-2004 ( And indeed, in contrast to the Assad side here, the majority of these Israeli deaths were of civilians (about 650 civilians and about 300 “security” personnel)? Did we say therefore that “both sides commit atrocities, they’re both bad?” Or does this kind of liberalism apply uniquely to Syria?

But in any case, we also need to examine the organisation itself and the fact that other reporting organisations exist, including the UN which has its own people on the ground. The well-known fact is that the Observatory is, as the New York Times explained, “virtually a one-man band” run by Rami Abdul Rahman, who fled Syria 13 years ago and who “operates out of a semidetached red-brick house on an ordinary residential street in this drab industrial city, using the simplest, cheapest Internet technology available” ( He claims to rely on some 200 personal informants on the ground in Syria. I don’t say this to disparage the obviously dedicated work of the Abdul Rahman, but 200 in a population of 24 million is in fact thin, and even he has reportedly said that the real figures could well be double.

Indeed, when his June report claimed 96,000 deaths (with the UN claiming 93,000), the Observatory noted the real figure could be around 130,000, and the UN said its figures excluded some 38 000 reported killings “because records – which require the victim’s full name and date and location of death – were incomplete”


The Syrian Observatory is not the only centre collecting death statistics. For example, the Syrian Network for Human Rights ( currently documents 83,598 deaths, of which 75,992 were civilians and 7,606 were rebel fighters (while not attempting to collect figures for regime casualties). These figures are slightly higher than those of the Violations Documentation Centre in Syria (, which has documented 55,786 civilians and 18,889 rebel fighters, along with 11,400 regime military deaths.

Now I don’t know which body is more accurate, and I reckon no-one knows that. But the big discrepancy is that both these reports show a much higher number, and proportion, of civilians killed than the Observatory, and the Observatory shows a much higher proportion of regime troops.

However, the Observatory’s very high figure for regime troops is also at odds with the Observatory itself.

In April, the Observatory released its report for the end of March ( According to this report, at that point 62,594 people had been killed in the conflict (the UN figure at that time was 70,000), and this included “at least 30,782 civilians, 15,283 troops, and 14,302 rebels.”

Thus, about 50 percent civilians, and 25 percent each for rebels and regime troops, a ratio of 2:1:1 civilians:rebels:regime troops.

The proportions of the 6005 actually killed in March were 2080 civilians, 2074 rebels and 1464 regime troops (with 387 “whose identities were impossible to verify”), thus the proportion of regime troops was even less than in the total figures. The ratio for the month was thus about 4:4:3 civilians:rebels:regime troops.

Yet in its report released in June, the Observatory’s figures had jumped to 96,431 (, including 35,479 civilians, 16,699 rebels and 41,793 regime troops and militia (the latter covering 24,617 troops and 17,031 militia, plus 145 Hezbollah). All of a sudden, the ratio had completely changed to around 7:3:8 civilians:rebels:regime troops.

The current report ( essentially continues that latest ratio, showing 40,146 civilian deaths, 23,707 rebel deaths and 47,206 (28804/18228/174) regime deaths, a ratio of around 8:5:9.

The increases between June and September were nearly 5000 civilians, 7000 rebels and 5500 regime troops, or a ratio of close to 1:1:1, especially if we exclude foreign fighters from rebel figures and Hezbollah from regime figures. Thus, this returns to figures similar to the monthly toll in the April report, and the total toll until then (the biggest increase since June was in rebel deaths and the smallest was in regime “militia” deaths).

Therefore, the single gigantic anomaly, in sharp contrast to all other total and monthly figures, was the alleged spike in regime deaths, from 15,283 in the April report to 41,793 in the June report.

Is this logical? It was during this period that the Assad army conquered Qusayr, and went on an offensive around the country, retaking a few other minor parts, in an operation that many observers imagined meant “Assad was winning.” It is possible that more troops died in that period. But most reports spoke of the increased role of the “militia” and the specific role of Hezbollah in Qusayr, usually noting the Assad preferred not to use his army too much due to concerns about loyalty. And in any case, civilians and rebels also suffered massively during these offensives, which can thus hardly explain such a spike.

The Observatory itself claims the new regime figures “were drawn up after it received thousands of names it hadn’t previously recorded from areas controlled by the Syrian government.” These new informants were allegedly in Latakia, among Alawite friends Rahman had known from school days, who had contact with people in the military, where Alawites have a large influence at the officer level. The regime itself doesn’t publish figures, but these informants allegedly had access to regime figures. No explanation has ever been given about how this anomalous group of figures were collected.

The reporter David Enders for McClatchy noted that in his trip to Damascus last year, “speaking to witnesses in the neighborhoods involved (ie, regime-controlled) required avoiding police and army checkpoints and an ever-present risk of raids by militia or government troops. The Syrian government has granted reporters some access to the country and has allowed video and photography of military funerals, but reporting without a government minder remains difficult.”

All in all, therefore, the VDC reports and figures seem the most likely to me, and their ratios in any case are not that different to those of the Observatory if we subtract the one highly anomalous spike in regime figures which contradict all other monthly and total reports by the organisation itself.