Early in David Edgerton's history of technology, The Shock of the Old he discusses cost benefit analysis applied to some military weapons.

He argues, for example, that the US would have caused a lot more damage to Japan and might have ended the war earlier if they had used the budget for the Manhattan Project on more conventional bombers or other weapons (though he also admits that deterrence was later a significant benefit).

But one particular example of an apparently bad choice was particularly stark. The German V2 rocket cost far more than other, more useful, weapons. Worse, and very surprisingly he makes this claim:

The V2 'was a unique weapon', says its historian, Michael Neufeld, in that 'more people died producing it than died from being hit by it'.

This struck me as a fairly extreme claim. Is it true that its production killed more than its use and is the V2 unique in this statistic?

Note: since many weapons are developed but never used, the claim should apply only to weapons that have actually been used in war.

  • 10
    Consider asking this on the History site? – DJClayworth Oct 9 at 14:58
  • 15
    @DJClayworth If I wanted a discussion of the history of the V2 I would ask this on a history site. But I was interested in the specific statistical claim so it seemed OK here. – matt_black Oct 9 at 15:00
  • 42
    I'm not sure if "unique" is to be taken literally here. Surely there must be weapons programs that have never been used in combat but still involved deaths in production. ICBMs or thermonuclear weapons, for instance. – Nate Eldredge Oct 9 at 15:02
  • 18
    It's about finding the right community. The History site has lots of expert historians. This site does not. I suspect you are more likely to get speculation and hasty research on this site. – DJClayworth Oct 9 at 15:08
  • 6
    Certainly this is a valid skeptics question. It fits the profile. It's possible that this one will be beyond us, though. My suggestion would be to leave it be where it is, we see what we can do to answer it, and then if we simply can't come up with a decent answer in a reasonable amount of time, re-ask on the history site. – Ben Barden Oct 9 at 15:11

Q.1: This struck me as a fairly extreme claim. Is it true that the V2 production killed more than its use?

In all likelihood, yes. The director of the Mittelbau-Dora commemoration site is quoted with

According to Jens-Christian Wagner, director of the Mittelbau-Dora concentration camp memorial site, it "killed more prisoners in the production of the weapon than [other victims] in their deployment. This is unique; I don't think there was any other weapon that claimed so many lives during the production." WP: Aggregat 4

In that article some numbers are given as:

Between September 1943 and April 1945, 16,000 and 20,000 concentration camp prisoners and forced laborers, most of them twenty to forty years old, died in the Mittelbau-Dora camp complex, on liquidation or so-called evacuation transports, according to conservative estimates. Approximately 8,000 people lost their lives using the weapon, most of them in the London and Antwerp area.

And that is the standard view:

Because of the conditions, deaths in the tunnels between November 1943 and March 1944 numbered almost three thousand. Beyond this toll, there were also another three thousand “muslims” (Muselmänner), prisoners unable to work because of grave illness, injury, or psychological shock. ese men were sent on to Bergen-Belsen or Maidanek with little chance of survival. […]

The completion of the barracks with washrooms in the spring and early summer of 1944 and other accommodations, including a sick bay staffed by medical personnel, while not adequate, was a large improvement over the murderous conditions of the tunnels. The number of deaths fluctuated drastically, lowering when the conditions improved and rising when new transports of prisoners arrived from the East.

Of the more than 5,000 V-2 rockets assembled in Kohnstein by the prisoners of Mittelbau-Dora, 1,500 of them landed in London and Southeast England, killing more than 2,000 people. Another 1,500 V-1 and V-2 weapons were aimed at Antwerp, Belgium, killing almost 7,000. It would appear that more prisoners died in the concentration camp serving Mittelwerk and further development of the “wonder weapons” than did helpless civilians in Germany’s enemy countries.

Gretchen Schafft & Gerhard Zeidler: "Commemorating Hell. The Public Memory of Mittelbau-Dora", University of Illinois Press: Urbana, Chicago, and Springfield, 2011.

So it depends on your point of view of what you'd call "producing it". Most forced labourers did not die tightening a screw and slipping. Many died of sickness, hunger, malnutrition, maltreatment, overwork and total exhaustion etc.

From a rational strategists point of view that seems like an idiotic approach. Within Nazi ideology that views those lives not only as expandable but to be exterminated eventually anyway, if not through labour than by other means, this would make perfect sense. A Nazi would conclude that a V2 was an extremely effective weapon, killing only enemies and that already when still manufacturing it.

Given that not only the one concentration camp was involved in manufacturing the rocket, the death-toll for the rocket on German soil is likely even higher than most estimates:

In total, around 60,000 prisoners passed through the Mittelbau camps between August 1943 and March 1945. The precise number of people killed is impossible to determine. The SS files counted around 12,000 dead. In addition, an unknown number of unregistered prisoners died or were murdered in the camps. Around 5,000 sick and dying were sent in early 1944 and in March 1945 to Lublin and Bergen-Belsen.

WP: Mittelbau-Dora concentration camp

Q.2: Is the V2 unique in this statistic?

Difficult to answer unequivocally without further definition. The V2 is fairly unique in a number of categories. Depending on point of view: yes and no.

Given that the V2 killed more prisoners – or people working for Germany – than enemy civilians one might rephrase the question to whether any other weapon killed more of its own people in some kind of broad interpretation of friendly fire?

The answer to that might be found in Amerithrax attacks. American made weapons grade anthrax was never used against official enemies, but against

Stevens, Bob - photo editor at American Media Inc, dies of inhalation anthrax, October 5, 2001
Curseen, Joseph Jr. - DC area postal worker, dies of inhalation anthrax, October 22, 2001
Morris, Thomas Jr. - DC postal worker, dies of inhalation anthrax, October 21, 2001
Nguyen, Kathy - employee at Manhattan hospital, dies of inhalation anthrax, October 31, 2001
Lundgren, Ottilie - Connecticut woman, dies of inhalation anthrax, November 22, 2001

Also compare that to Anthrax: full list of cases.

Similarly, it is believed that the United Kingdom never used nerve gas in a war. But manufacturing it seems to have caused up to 41 deaths and many more injuries:

The Ministry of Defence is reopening a 30-year-old inquiry into a series of deaths and serious illnesses at a chemical weapons factory in Cornwall amid accusations that the original report was the subject of a high-level cover-up.

The new analysis could show that, for decades, civil servants fed false figures to ministers who then rehashed old answers to a succession of local MPs, she said. "The 41 deaths have always seemed a much higher than average death rate to me. I found it extraordinary the MoD has always alleged it isn't."

Robert Mendick: "Deaths Inquiry at Nerve Gas Plant", Independent, 7 May 2000.

Do we all remember the use of smallpox in war by the Soviet Union?

[…] The smallpox formulation—400 gr. of which was exploded on the island—"got her" and she became infected. After returning home to Aralsk, she infected several people including children. All of them died. I suspected the reason for this and called the Chief of General Staff of Ministry of Defense and requested to forbid the stop of the Alma-Ata-Moscow train in Aralsk. As a result, the epidemic around the country was prevented. I called [future Soviet General Secretary Yuri] Andropov, who at that time was Chief of KGB, and informed him of the exclusive recipe of smallpox obtained on Vozrazhdenie Island.
WP: Aral smallpox incident

Making use of smallpox in war by the Soviet Union death count zero against a really undisclosed number of casualties, three officially.

The use of anthrax again, this time by the Soviet Union, is also zero, but the numbers killed by it on friendly soil are around 100. (WP: Sverdlovsk anthrax leak, 1979.)

If we go for an individual weapon alone then the atomic bomb also makes the list. A death toll of zero from the first explosion stands against the death of Harry Daghlian and Louis Slotin.

The V2 may stand out as among the weapon with the highest death toll, according to the criteria, and having been even envisioned and almost designed to kill workers in the process of manufacturing it. But producing weapons is inherently dangerous. Producing them without using them as originally intended should automatically make them contenders for the list. The V2 is in one way certainly not unique. Its older brother, the fastest Volkswagen at the time, the V1 was also built by forced labour, but had after a while quite low death toll around the intended targets, as those bombs had abysmal accuracy and the countermeasures improved greatly.

A total of 6876 long-range weapons struck English soil. 8938 people were killed and 24 504 injured. Belgium was hit by 8661 long-range weapons. Here 6448 people were killed and 22 524 injured.
Dieter Hölsken: "Die V-Waffen. Entwicklung und Einsatzgrundsätze", Militärgeschichtliche Mitteilungen; Freiburg Bd. 0, Ausg. 2, (Jan 1, 1985): 95. p 116.

Bringing the effective death toll to 15386 and 47028 people injured by those weapons.

Hölsken calculates that in total 22384 V1 and 3170 V2 were fired at the enemy. Curiously 11 V2 were targeted on German soil. None of the V2 used managed to hit that target.

Since the German war machine depended heavily on slave labour that should also exterminate the workers, a look on further candidates that were produced and brought to the frontline without killing much people might yield a few more, like: Krummlauf, Schwerer Gustav (as Dora), V3, but hard numbers for comparison seem unavailable.

Globally, there might be another twist to definitions. Some items and implements are classified as "defensive weapons", like shields and walls, and while those are surely not covered by the definition sought after in the question, it looks like they pretty much all would qualify. Whether Chinese Wall, Limes or Maginot…


The death toll in manufacturing V1 and V2 weapons is for one concentration camp alone between 12000, officially registered by the SS in the camp, with a number more probably around 20000. That number does not include other deaths in the other camps and over 50 supplying factories, companies or "transport losses", whether by train or on death marches from the East.

Compared to 15386 losses of enemy civilians the claim holds up. As the claim over-focusses on the V2 it is unproblematic to deny its uniqueness. The V1 can be seen as equally responsible for more deaths in production and perhaps even less effective if loss of human life is seen as the goal.

  • 7
    The "treated blankets" attacks against the Native Americans suggest that use of smallpox as a weapon of war netted more than zero. Still, for the Russians maybe this was true. – Ben Barden Oct 9 at 18:05
  • 5
    @BenBarden If you have any evidence of small pox blankets actually being used as a weapon, I'd love to see it. The evidence I've seen amounts to one military figure suggesting that it might be a good tactic. Did it see any actual use? – Michael W. Oct 9 at 23:32
  • 3
    A very thorough answer but strays too far into discussion about things that were never used in actual wars. It would be an even better answer were it shorter. – matt_black Oct 10 at 9:34
  • 2
    Just a (to me) interested tidbit - “muslims” (Muselmänner) doesn't me literal Muslims; it was camp slang for concentration / death camp prisoners who were in poor health and spiraling down to almost certain death. – Amorphous Blob Oct 10 at 15:17
  • 5
    @matt_black I'm thinking that thoroughness is worth more than brevity for this stack. – Ben Barden Oct 10 at 15:59

Tools of War: History of Weapons in Modern Times by Syed Ramsey claims that there was a 2011 BBC documentary stating numbers of 9000 killed in attacks, with 12000 forced laborers and concentration camp prisoners killed in production.

There's also A BBC News Magazine article stating that 9000 number directly. It mentions a mountain-factory in Germany that had 60,000 slave laborers - making the "forced laborer" death toll at least plausible, given how the Nazis were running things. It also notes that the V2 would have been much more effective if it had not been shut down so thoroughly by allied bombing.

Finally, the Daily Mail website states both numbers directly... for what that's worth. They may just be following along with the oft-referenced BBC numbers, but they have a number of other statistics that are not so broadly spread, so possibly not.

On something like "historical data on the V2 Rocket" I really don't think you're going to get much better than the BBC. It seems at least likely that there were more lives lost among the workers during production than there were lost to the attacks themselves. There is some potential for confusion, however. The workers were in a concentration camp environment, and there are some fuzzy questions about whether they'd all count. (Does it count if they're summarily executed while working? What about if that execution is intended primarily to motivate their fellows?)

I do not have an answer to the other, and I'm not convinced that it is answerable if true. Proving a negative is really very difficult at the best of times, and proving a negative in a matter of historical footnote is much moreso. It also somewhat depends on your definition of "weapon" - specifically in how distinct it has to be in order to not just get grouped in with other weapons of the same type. If a very specific make and model of cannon had some horrible laboratory accident, only produced one instance, and then had that instance spiked after being fired but without managing to do much damage, I can totally imagine that it would fit the answer... but I doubt it would be easy to find.

  • 1
    Good points: Also define "laboral accident" in a prone to be bombarded and full of volatiles environment – jean Oct 9 at 15:53
  • 1
    If it made itself a bombing target it was an effective weapon. – Joshua Oct 9 at 17:22

I have a very clear alternative example of another weapon system that killed more people in production than use. The H. L. Hunley, the first submarine to successfully sink a surface ship. In testing and trials the submarine sunk twice.

On August 29th [1863], the Hunley was moored at Fort Johnson, preparing to depart for its first attack on the blockade when it suddenly sank at the dock...taking five of her crew down to their deaths

On October 15th [1863], Horace Hunley scheduled a demonstration of his boat in Charleston Harbor. He announced his vessel would dive beneath the CSS Indian Chief and surface on the other side. Once the submarine disappeared beneath the waves, it was not seen again for weeks.

In total 13 people died in these preparations. When the Hunley actually did complete its only combat operation, "the spar torpedo detonated and the explosion blew a hole in the ship. The Housatonic sank in less than five minutes, causing the death of 5 of its 155 crewmen."

The Hunley was lost after the attack, and the Confederacy did not pursue submarine warfare again. To sum up, 13 people died developing the weapon, and it killed 5 people.

New contributor
lazarusL is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
  • 1
    You might want to put some years onto your dates... I had a vague memory that this occurred during the American Civil War but it wasn't until the last paragraph, when you mentioned the Confederacy, that I was sure. – Oscar Bravo Oct 12 at 14:48
  • 2
    I like this example; 1) weapon used in a war: ✓ 2) caused enemy deaths: ✓ 3) caused more friendly deaths: ✓ – matt_black Oct 12 at 16:10
  • @matt_black: and was probably worth the deaths, if not the lost of equipment, in exchange for sinking an enemy ship. – jmoreno Oct 12 at 23:57

Definitely no, lots of weapons systems were never used in combat

One glaring example would be B-58 Hustler. This bomber was operational in USAF, fortunately never used in war (as it was designed as nuclear bomber), but unfortunately killed some people while it was tested by company test pilot and flight engineer. So people died while producing it. B-58 has been retired a long time ago, so chances it would participate in some war are almost non-existent.

Another bomber that participated in war but didn't directly kill anyone was B-47 Stratojet. There were lot of accidents involving this plane, including deadly accidents with factory test pilots. It did participate in Vietnam war, but only as reconnaissance and ELINT aircraft. It is also retired so its production did kill more people then its use .

  • 3
    I did specify that the meaningful claim was limited to weapons that were actually used in a war. – matt_black Oct 10 at 17:25
  • @matt_black Well, it was used in Cold War :) I would try to find some recce platforms that were used in Vietnam – rs.29 Oct 10 at 17:27
  • @matt_black Check B-47 – rs.29 Oct 10 at 19:08

You could regard the Burma Railway as a weapon of war, albeit a logistical weapon rather than a direct offensive weapon. This was a railway that the Japanese constructed in World War Two to supply their forces in Burma directly from Thailand. It is estimated that around 100,000 forced laborers and prisoners of war died during the construction. It is impossible to quantify how many Allied deaths can be attributed to munitions shipped on the railway, but Japanese forces were not very efficient at killing Allied soldiers, with the total number of Allied military deaths in the entire 4-year Burma Campaign being around 100,000 as well. Unless every Allied death was directly attributable to the supplies shipped on the railway (implausible to say the least, especially since the railroad wasn't completed until the war was half over), this would be a clear case of a weapon which killed more during its production than in its use (if you count construction as a form of production. I would say that it is. If not, the prisoners who died in constructing the underground factory where the V2s were made would have to be carefully subtracted from the V2 figures).

Most non-lethal weapons would probably qualify, as would any weapons never actually used (e.g. ICBMs) as long as at least one person had a fatal accident during production.

Note that the primary military impact of the V2 was not death toll, and it was likely not expected to be - compared to the amount of explosives a single bomber could drop a lot more than the 975 kg of the V2 warhead.

But the V2 produced terror. It was almost impossible to defend against with the air defense systems of the time, and the time between warning and impact was much shorter than for a bombing raid.

Some of the information contained in this post requires additional references. Please edit to add citations to reliable sources that support the assertions made here. Unsourced material may be disputed or deleted.

  • 7
    Do you have a citation for this, as is required? Given the number of deaths that happen with something as allegedly nonlethal as a taser, it's not as clear cut as you think most likely. – Laurel Oct 9 at 19:29
  • 1
    That is a good point there, yes. – Tom Oct 9 at 21:03
  • The question already excludes weapons that have never been used. And as for the V2 being designed to produce terror it is no different to a large part of the conventional bombing campaign against Germany. But weapons that don't kill many are not that terrifying. – matt_black Oct 10 at 9:38
  • I missed the exclusion, my mistake. As for the terror - of course this is a side-effect of all weapons. The V2, however, was never produced in large enough numbers (about 5000 in total) to be a credible killer. The HE 111 could carry up to 2,500 kg of bombs per flight and was produced in a similar number. – Tom Oct 10 at 10:41
  • 2
    @matt_black, weapons which cause lifelong disabling injuries without killing (e.g. landmines) are arguably more terrifying. – Peter Taylor Oct 10 at 10:51

Another possibility would be the V-3, although I haven't seen a number for production deaths. It was built by slave labor, so it seems likely that there more than ten deaths. It was a gun with multiple chambers along the tube, each contributing propellant (hopefully) after the shell went past.

It was actually used to bombard Luxembourg City during and after the Ardennes Offensive (Battle of the Bulge), firing 183 rounds and killing ten people.

Mining uranium isn't good for your health.

It has been estimated that 500 to 600 of the thousands of uranium miners who worked between 1950 and 1990 died of lung cancer, that most of these deaths were associated with radon exposure, and that a similar number would die after 1990.42 A 2000 study of Navajo miners reports that there were 94 lung cancer deaths documented from 1969 to 1993, that 63 of these individuals were former uranium miners, and that uranium miners had a relative risk of 28.6 compared with controls.59 Frank Gilliland et al. point out that this appears to be a “unique example of exposure in a single occupation accounting for the majority of lung cancers in an entire population.


The single biggest case of deaths directly related to radon exposure in Canada happened as far back as the 1970s and 1980s, when at least 220 miners in Elliot Lake, Ontario died of lung cancer from years of exposure in the town’s uranium mines.

The Elliot Lake tragedy ultimately compelled the United Steelworkers union to go on strike in 1974. The result led to the appointment of the Ham Commission, which helped pass one of the most significant pieces of worker health and safety legislation in Ontario five years later with the Occupational Health and Safety Act.

Dangers of exposure

Under the terms of the Act, the Workers Safety Insurance Board of Ontario (WSIB) accepted the claims of the 220 miners based on the hazardous working conditions in the mines. The Steelworkers union believes the true number of deaths is actually much higher, based on its knowledge of those who were thought to have suffered from exposure, and the number of actual claims filed with the WSIB. Their plight helped sound the alarm about the dangers of the invisible and odourless gas, which quietly damages the lungs over years of exposure.


Specifically, extant literature on Grants miners has found that about 52.4% of the NMSMI identified miner population over 1957–1976 were unique.24 Using this data, I estimated that 38 754 unique miners worked in Grants between 1955 and 1990. Therefore, approximately (73 903–38 754) = 35 149 miners were the same miner being counted twice or more in the NMSMI data. The sensitivity of the final results to the choice of this point estimate is examined at the end of the paper.

Excess exposure-related lung cancer deaths were estimated at 550.5 miners for the period 1955–1990, or 134.9 between 1955 and 1970 and 415.6 between 1971 and 1990. In standardized terms, excess lung cancer mortality over the entire mining period was 2185.4 per 100 000 miners. This was approximately 1.4% of the unique Grants miner population who worked over 1955–1990, or about 2.2% of the unique miner population who worked >6 months over the same time period. Results are significantly lower than comparable studies on Colorado Plateau uranium miners, where the most recent excess lung cancer mortality estimate is 13 052.8 per 100 000 miners.52


Deaths due to accident, Emphysema, Tuberculosis, and Pneumoconiosis were also much higher than expected. here

When trying to come up with a total for the North American continent for deaths attributed to producing nuclear weapon feedstock, numbers like 20,000 get bandied about.

  • While uranium mining may be dangerous, it is several steps away from specific weapons (should we count coal mining deaths in the production cost of tanks?) And Nuclear weapons only use in war was in 1945 so related deaths occurring after that don't count. – matt_black Oct 12 at 16:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .