Hacker News new | past | comments | ask | show | jobs | submit login
The price of the Manhattan Project (2013) (nuclearsecrecy.com)
103 points by sytelus on Dec 9, 2018 | hide | past | favorite | 67 comments

I have a strange take-away from this, in that the article proposes that The Manhattan Project’s underlying goal was to model the first group of prototypes as master copies for an assembly line, and that the true cost of the project was the production line, and not the first group of weapons used to test the scientific principle and ultimately bring the war to an end.

That makes sense, especially seeing the arsenal and industry grow as it did in the wake of the war.

The thing that leaves me curious is that despite the massive spending on the space program, we never see the same assembly line, or perhaps transit system, emerge from the spending.

With the space program, it would seem that all of the effort and financing really did get consumed by each moonshot, with no investment to back it up, and formulate repeatable process and the ready facility to continiously grow the number of people in orbit, and cycle rockets through their trajectories.

The nuclear energy/weapons complex and the space program seem to be comparable endeavors, but one paved the way for sustainable operations while the other produced a series of events and little else. It doesn’t seem plausible to just wave one’s hands and say that rockets are that expensive, so what gives?

The government program rockets aren't designed to be cost efficient, they're designed to buy as many votes as possible. To do that, they spread everything out a thin as possible, all the work across as many states as possible so that the elected officials from each of those states will vote for it.

This is why SpaceX, with their single factory, are something like an order of magnitude cheaper.

I think NASA spent a lot of money on figuring out what works and what doesn't work. They probably could have done it much cheaper but SpaceX certainly had a very good blueprint to start from and didn't have to figure out everything from scratch.

If this is as true as it sounds - holy cow that's so sad !

As someone who worked for NASA for 15 years I can tell you this is absolutely true.

Not really - lots of great, high paying engineering and manufacturing jobs came out of that program. Plus all the support that goes with it.

> [...] lots of great, high paying engineering and manufacturing jobs came out of that program.

That's a pretty bad excuse: leaving the same money with the taxpayers would have funded jobs that actually did something people wanted enough to pay with their own money.

There's also a strategic purpose: it's easier to hire ICBM designers if designing ICBMs isn't the sole career path for them.

Yeah, similar story for nuclear power. I don't think people understand that very few competent STEM undergrads would choose to go into the submarine force if they saw no job on the other side of the suffering necessary to get through prototype and multiple deployments. As it is I've heard something like 30 out of 190 per year group make it to command some 20 years later. So that's 160 people a year who have at least some thought that their skills can transfer to a civilian job.

It’s not purely an excuse. If somone makes 100k on one of these projects but would have made 50k if they did not exist then the economy is out 50k worth of labor. But, some of that delta between 50k>100k is payed back in additional taxes.

There is plenty of waste, but it’s significantly smaller than the direct costs suggest. Especially when such programs reduce unemployment as the economy is not out useful labor, and taxes collect a significant portion of the outlay back.

I think of it as the governmental make work program. It’s only been digestible, politically, when it’s tied to this sort of project. Without it, my home state would be severely affected.

> Without it, my home state would be severely affected.

New Mexico?


The economy isn’t “out 50k of labor”, the government is saving 50k that it could use to pay someone else who wants to work at wherever their headquarter is.

I am referring to the opportunity cost.

Assume Bob could do something and get paid 50k per year, government pays Bob 50k/year to stay at home. Bob makes 50k either way.

But, if Bob does not work for a year then the economy loses 50k/year worth of labor, that’s true if bob makes 40k staying at home or 200k staying at home. It’s also true if the government pays Bob to do something that creates zero value.

My takeaway is that people don't have a very good sense of scale on government spending.

Manhattan project: 30 billion (2012 dollars)

Apollo program: 83 billion (2005 dollars)

Shuttle program: 196 billion (2011 dollars)

2008 bailouts: 700 billion (2008 dollars)

Iraq war: over 1 trillion dollars

That's right, both the Iraq war and the 2008 bailouts each dwarf the entire space program. Yet which thing attracts more hand wringing about costs? Which things require years of tiresome political wrangling, and which happened at the snap of a finger?

Yearly US defense budget is increasing to 750 billion (2018 dollars).


Another good one: total money spent by the US on fusion research, ever: 37.6 billion (2005 dollars).

It really puts into perspective how many things commonly held to be incredibly difficult, or expensive, are in fact simply not a priority. Almost any pie-in-the-sky technological project or public works project you can think of utterly pales next to the money that is casually spent on dubious causes like propping up broken financial systems and destabilizing foreign countries.

Maybe we should stop using "Manhattan Project" or "moonshot" to refer to huge, ambitious practically unfundable government works, when they're so cheap in the scheme of things. Maybe we should speak of "bailouts" instead.

Anecdotally, I hear a lot more complaints about the bailouts and the war than I do about anything related to NASA.

These programs did lead to an assembly line that created thousands of very powerful rockets and the infrastructure to launch them. They just weren’t used for space flight.

Those two projects were directly intertwined. The Titan missile is what took people into space in Project Gemini. The space program was, in my opinion, a front for the intercontinental ballistic missile program.

Not even a front, I'd say, that implies this was somewhat hidden. The first thing anyone thought about Sputnik was that it could as well be a bomb. Rockets reliable enough to fly celebrities to orbit were pretty direct advertising of each nation's capabilities. (By celebrities I just mean that they were famous before they flew, and thus losing one would have been impossible to cover up, thus increasing the signal.)

A handful of prototype atomic bombs isn’t enough to beat the Soviet Union in a war, but one rocket to the moon is enough to beat them in space.

  As long as it worked, who cares?
That statement is a declaration of assent that you find the outcome acceptable.

It doesn’t examine the manner of resource usage, and the returned benefits for each cost.

In one case, they built a car and a garage to fix many cars in.

In the other case, they built a car off some value and drove it off a cliff. For the price, could we not have also obtained a matching garage?

Why are you “quoting” something I didn’t write and then claiming it means I find the outcome acceptable?

> For example, one of the questions that people ask me again and again is how close the Germans were to getting an atomic bomb. The answer is, more or less, not very close at all. Why not? Because even if their scientific understanding was not too far away — which it was not, even though they were wrong about several things and behind on several others — they never came close to the stage that would be necessary to turn it into an industrial production program, as opposed to just a laboratory understanding. That sheer fact is much more important than whether Heisenberg fully understood the nature of chain reactions or anything like that.

I found this article to be very interesting. The point above in particular, for example, has parallels with the allied code-breaking efforts, which were similarly organized primarily as an industrial operation. (The Bletchley Park data base lists more that 10,000 people who were associated with work there during the war). As with the Manhattan project, much is written about key figures, like Turing and Flowers, whose contributions were of course crucial, but the true genius of the effort was the fact that the British, and later the Americans, created a large industrial organization for code breaking.

I found this conference talk about WWII being at its heart a "battle of the factories" enlightening on this point: https://www.youtube.com/watch?v=N6xLMUifbxQ#t=26m20s

Wikipedia thinks that the V2 project cost 50% more than the Manhattan project. If this is credible (and it roughly matches what I remember) then the Germans' problem wasn't that they were unable to (afford to) build factories fast enough.


I think what the article is saying is that there was an awful lot of technical knowledge between the scientific understanding (e.g. talking about what Heisenberg did or didn't know) and being able to build a factory. This knowledge the allies also didn't have at the beginning, of course.

Also, as the article points out, it's really hard to write exciting books about achieving incredibly tight tolerances in manufacturing.

And engineers tend to be less colorfully interesting than physicists. (No offense intended, I'd include myself in this characterization)

Keep in mind that the Germans didn't have access to many resources. The allies had Africa, the Middle East and South America to supply them with all they needed.

Indeed, there could have been some other critical shortage barring the way. (IIRC the soviet bomb-makers mined half their uranium in germany, but something else maybe.) But aggregate industrial capacity wasn't obviously a show-stopper, if they were able to fund another very high-tech high-cost project.

The industrial base, and the fact that production could continue unabated by bombings on the American continent for years, was a deciding factor in the war. As were the Russians repelling Hitler. My historical perspective on this has shifted massively since first learning about the war in school. It wasn't D-Day or the Battle of Britain or Midway. All of those battles were key wins but the war overall was really organized and executed by tens of thousands of typists and mid-level managers.

Yes. Eisenhower was a logistics guy. D-Day was a triumph of logistics. The Allies built up enough resources to force a win before attempting the invasion. They were willing to wait because of the failures of Dunkirk and Dieppe. Those operations didn't have enough resources behind them.

We're under fire! Yeah - under fire!! Those five thousand ships the Allies couldn't possibly have, as you say - they've got 'em! Yeah, they've got 'em!! - Major Werner Pluskat, June 6, 1944.

The thing about the bomb in particular, that stands out in this discussion, is that it was/is a fairly simple physical device. It had to be constructed to some very exact standards but the main thing was that getting uranium 238 was extremely difficult. Most of the ingredients were relatively simple - the major single, cost overall was K-25/Oakridge.

The process used was very crude, simple and inefficient so they needed to do it on a huge industrial scale. And only the US could do it on that scale.

You could call it "the dumb luck of nature" that something like an atomic bomb couldn't have been produced with, say, 19th century technology but that it could, barely, be produced by the full industrial might of a mid-20th century nation.

Which is to say that start-up producing amazing things in basements now certainly rely on the sum-total of previous industrial achievements, we don't how much some interesting bit of knowledge compared to those achievements could be leveraged.

I wish I could remember the exact quote, but in the diary of a German soldier fighting on in France he said something like "For every Krupp shell we sent to the Americans, they sent a hundred Detroit shells back."

In "The Rommel Papers," (which is incidentally a really splendid book that you should definitely read if you have any interest in tactics at all), Rommel himself claims that the ratio was in practice closer to 500:1 in the African theater.

Extend to many things.

If Wikipedia is to be believed, the US produced more M4 chassis (all vehicle types) in 1943-alone (21,231, admittedly a huge surge in yoy production) [1] than Germany produced of any single chassis type during the whole war (most numerous: Pz III at 15,747 total) [2].

[1] https://en.m.wikipedia.org/wiki/American_armored_fighting_ve...

[2] https://en.m.wikipedia.org/wiki/German_armored_fighting_vehi...

That most obvious savings generated by the Manhattan Project was probably the displacement of perceived need to invade Japan. It would be nice to see an estimate of the cost of that invasion as a comparison. But the primary cost of an invasion would have been the lives of Americans and Japanese, and it would be difficult to arrive at a consensus value for that. As part of that wider comparison the cost of the destruction in Hiroshima and Nagasaki should also be included on the Manhattan Project's side of the ledger.

In July of 1945, Szilard and a number of other scientists involved with the Manhattan Project, signed a petition asking that Japan be publicly informed of the terms of surrender and refused then before the bomb was dropped. Not only was that simple request ignored, the US Government tried to blacklist the scientists who signed that petition. Arguing that the bombs were necessary to prevent an invasion pretends that any other option was tried.


I feel like this splits hairs a bit, as the Potsdam Declaration[1] already had warned of utter destruction unless Japan surrendered. Our firebomb raids had already demonstrated this resolve and killed more civilians than the atomic bombs, in likely more horrific ways.

[1] https://en.wikipedia.org/wiki/Potsdam_Declaration

This is to one side of the point - many historians hold that the U.S. was keen to demonstrate the power of these bombs in order to cow the U.S.S.R., (and prove they were practical) even if it meant delaying the end of the war.

Firebombing was even more effective (but not economical) in killing Japanese civilians, as you note. However, firebombing would not have been as effective vs the Soviets (far less flimsy, compact wood construction) nor have the same effect vs ground forces: making a live demonstration of atomic bombs still useful intimidation, according to this view.

Besides cowing the USSR, I think that a valid war aim was fully cowing Japan.

One of the lessons of WWI was that military victory wasn't enough, too many Germans could tell themselves afterwards that they were stabbed in the back, etc. If the goal was to defeat Japanese fascism, then it was important that no Japanese could believe such things afterwards. And for this, a shocking weapon (one plane, one city) was effective in a way that (say) a starvation blockade would not have been.

One of the goals of the Potsdam Deceleration was to be intentionally vague about the terms of surrender, and the terms were never formally rejected. It doesn't fit the guidelines of the petition, and barely counts as an attempt at peace without bombing or invasion.

> One of the goals of the Potsdam Deceleration was to be intentionally vague about the terms of surrender

Do you have a cite that the ambiguity was intended to reduce the chance the Japanese would surrender? I don't know the history well.

I have the impression that, given the US's knowledge of its own military superiority, it would never have accepted anything besides total surrender and occupation. (This is not just vindictive; cease fires and truces are less effective for ensuring the enemy doesn't rise later.) Therefore, if the declaration was vague, this if anything would have been to encourage getting the Japanese to the bargaining table.

Wikipedia seems consistent with this:

> A major aspect relating to the Potsdam Declaration was that it was intended to be ambiguous. It is not clear from the document itself whether a Japanese government was to remain under Allied occupation or whether the occupation would be run by a foreign military government. In the same manner, it was not clear whether after the end of the occupation Japan was to include any territory other than the four main Japanese islands...This ambiguity was intentional on the part of the U.S. government in order to allow the Allies a free hand in running the affairs of Japan afterwards.

One can certainly criticize the US for not accepting conditional/partial surrender as a way to avoid unnecessary deaths, but the ambiguity of the declaration doesn't seem like the issue.

>Do you have a cite that the ambiguity was intended to reduce the chance the Japanese would surrender? I don't know the history well.

If fairly clearly shows that they were not listening to the requests made in the Szilard petition, which is what I am claiming. They did not publicly offer the terms of surrender or wait for a refusal.

Oh, OK, sure. It sounds like the Szilard petition wanted to incentivize Japan to surrender by conceding to them some post-war self-determination. The US would not accept this -- which, I as I said, might have been immoral -- so, to me, the ambiguity of the surrender terms in the Potsdam Declaration is not really surprising or damning. If we can trust Wikipedia, Japan declined surrender even after the first bomb on Hiroshima in the hopes of striking better terms with the USSR.

For others, here are the relevant excerpts from the Szilard petition, which is quite short and worth reading in full:

> We feel, however, that such attacks on Japan could not be justified, at least not unless the terms which will be imposed after the war on Japan were made public in detail and Japan were given an opportunity to surrender....If such public announcement gave assurance to the Japanese that they could look forward to a life devoted to peaceful pursuits in their homeland and if Japan still refused to surrender our nation might then, in certain circumstances, find itself forced to resort to the use of atomic bombs....we, the undersigned, respectfully petition: first, that you exercise your power as Commander-in-Chief, to rule that the United States shall not resort to the use of atomic bombs in this war unless the terms which will be imposed upon Japan have been made public in detail and Japan knowing these terms has refused to surrender; ...


>It sounds like the Szilard petition wanted to incentivize Japan to surrender by conceding to them some post-war self-determination.

This is not the case, the petition was to prevent the US from dropping a bomb unless Japan had been given the terms of surrender and refused them. This did not happen, and thus the situation can't be boiled down to "nuke or invasion" as there was a third option was not tried. I'm not making any further arguments, my positions on them are extreme and I don't want to try to convince others of them.

I mean it was war. Like it or not you don’t go into war negotiations with a limp hand - it’s a weak negotiating point. So I don’t think the US needed to make a hat in hand offer for it to be considered valid.

In accounting for the costs of counterfactuals I think the question is less what is possible and more what is likely. The outsized response you describe to the simple request is evidence of the strong political will for more aggressive options.

I suspect that US military leaders would be much more hesitant to spend the lives of untold thousands of US soldiers, and a prolonged firebombing campaign would have been the most likely non-nuclear continuation. Whether that would have saved lives or cost more is impossible to guess

People make the mistake that the Japanese government actually CARED about the suffering of its people. Japanese cities were already being bombed daily and Japan was starving. The war had been lost for years.

John McCloy proposed explicitly including a description of the bomb in surrender terms too.

'Perceived' is correct, as the other option was conditional surrender of Japan. But it sounds much nicer, and is better for morale, to say the bombs or an invasion were the only possibilities.





John Toland et al, in contrast, see the Russian invasion of Manchuria as the final straw that forced Japan to surrender; Japan continued fighting after the bombs were dropped, but the U.S. did not immediately have more bombs to drop.

I read about this recently here:


Fascinating how much sense it makes. Likely the truth is somewhere in-between.

Hell to Pay: Operation Downfall and the Invasion of Japan by Giangreco is a comparison of the American plans for the invasion of the Home Islands with the Japanese plans for their defense. It talks through the cost in live of such an invasion. Though it doesn't explicitly speak to the value of the Manhattan project, it argues for it indirectly by explaining the horrifying brutality that would have been that invasion.

I find it curious that if you add up the very first column of figures in the article -- cost in current dollars -- you find that it comes to way over 40 billion dollars, not 30.

Points taken about industrial might, and Germany's increasing industrial output throughout the war, despite all the (apparently ineffectual) bombing, though. Maybe leave off the sums?

The entries on the table for K-25 through S-50 are included in the Oak ridge total.

It'd be interesting to see a cost comparison to the USSR's program around the same time, leading up to the development of the first Soviet nuclear bomb.

It would be an apples and oranges comparison to some extent. The Soviets were able to skip some expensive research and production steps that the Americans went through by exploiting intelligence gained from espionage [0].

[0] https://en.wikipedia.org/wiki/Soviet_atomic_bomb_project

I'd be interested in that as well. The US program was expensive in part because they were trying to do it as quickly as possible, didn't know exactly what a finished bomb would look like and basically pursued every possible method for isotope separation at massive scale all at once.

A program that just pursued plutonium production and separation to produce implosion devices would presumably be a lot cheaper (though still massively expensive in absolute terms, obv).

> The US program was expensive in part because they were trying to do it as quickly as possible,

The US also was trying to just do it, whereas the USSR had the benefit of stolen research.

Well, thanks to the likes of Julius and Ethel Rosenberg the Russians were no doubt able to save a helluva lot of time and money.

This is known by everybody who's read about the Manhattan Project. The hard problem was separating uranium isotopes. The Hiroshima bomb was a nuclear IED. It could be built today by a good auto racing shop if they had the enriched uranium. Even machining uranium isn't that hard. There's a how-to guide from Union Carbide's tools unit. It takes some precautions, but not remote manipulators or anything like what's needed for plutonium.

I think the Manhattan Project is the exclamation point of what really decided the ultimate victors of WWII:

Industrial capacity and efficiency.

What’s interesting is that German industrial capacity, despite being bombed 24/7(US daylight and British nighttime raids often in excess of 1,000 bombers) were able to continue producing warfighting equipment in increasing record numbers often with the use of slave labour.

Despite increased German industrial production of warfighting equipment despite being bombed, the US strategic industrial depth ensured ultimate victory rather than tactical prowess.

What’s interesting in Stan McChrystal’s recent book “Team of Teams” is his mention that while 20th century conventional conflicts were about industrial capacity and efficiency, the 21st century will likely be about adaptability.

You may be interested in reading about the offset strategies employed by the US DOD. In the 60s the US military conducted a still classified analysis called Project Oregon Trail. Included (apparently) was a wargames exercise which completely undermined Eisenhower's doctrine of tactical nuclear weapons to fight a land war. Oregon Trail supposedly concluded that 20,000 tactical nukes would be required to win in Europe.

The result of Oregon Trail appears to be what led to the second offset strategy, which emphasizes intel, stealth, and precision guidance tech. But it also provided information by way of a two part study by a panel of historians. The historians studied 100 asymmetric wars throughout history, which lead to this key insight: if a government can figure out the reason people support the insurgents, and offer an alternative, then they would likely win. If they cannot find the reason, or are unable to win the hearts and minds of the people, they will likely lose. The more military force is substituted for politics, the less likely the government would be to win. We saw this outcome in Vietnam (and have to a certain degree seen the same in the Middle East).

I believe this insight is what has lead to the US' third offset strategy, which is deterrence of war through overwhelming technological advantage.

The point to all this being that industrial capacity and efficiency is just one cog in the machine. The bomb was a technological advantage. You don't need much industrial capacity or efficiency to make one. Inevitably, the best way to win a war is to convince the other side of your cause, and then they'll just stop fighting.

To put things into perspective, the B-29 bomber project cost around $3.2 billion dollars in 1945 dollars, half as much again as the $1.9 billion atomic weapon.

The bomb project was expensive because they were accelerating perhaps 30 years of normal research and development into 4 years. They did not know what would work, so they made two different types of separation plants, and two different types of bombs. All the clean room technology needed for the factories needed to be developed from scratch.

With the wisdom of hindsight, the "race to the bomb" was a waste, but at the time it seemed prudent. Then again war is waste, war is about laying waste.

They tried four different types of separation: centrifugation, thermal diffusion, gaseous diffusion, and electromagnetic separation. Centrifugation proved unworkable at the time and the remaining processes were used in concert (taking the slightly enriched products of one into the feed of another) to produce HEU.

The Manhattan Project was so expensive largely because it was a shotgun approach to development. They tried 6 different ways to produce 3 different possible bomb fuels and two different bomb designs. Ultimately the most promising methods didn't work (centrifugation) but this didn't cause the program to halt or have to rethink things, it kept chugging on everything else running in parallel and ultimately succeeded with producing 3 different ways of making HEU and one for making Plutonium plus a bomb design that worked for each of the two fuels.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact