I can't speak for his overall NKVD management, but for the nuclear program (which he ran), there was a very distinctive "Beria school of management" (a joke phrase some colleagues and I like to use, not something anyone called it):

• Don't trust your spies — always get the information confirmed from two or three sources, always get it re-confirmed by your own scientists, always check the blind alleys they tell you that you don't need to check, etc. You can use espionage information but never naively.

• Don't trust your own scientists, either — they could be saboteurs or foot-draggers. Always have teams doing redundant work, use them as a check against each other (and the espionage information). If teams come up with data that contradicts each other, pit them against each other with subtle threats about what will happen if one is found to be a saboteur or foot-dragger.

• Keep your spies in the field highly productive with threats. If they don't turn up useful info, recall them back to Moscow, never to be heard of again. Spying isn't a game; it's a business, and they need to "always be closing."

• Every major player on the project gets both an understudy and a "bodyguard." The bodyguard is making sure they don't do anything screwy. The understudy is there in case you have to purge them.

• Heavy, heavy compartmentalization by site. People at site A don't need to know that site B even exists most of the time, much less where it is or what it is doing. Only a small number of people need to know the source of the espionage information (it is "Sovietized" rather rapidly). Secret work is done in closed cities — scientists don't mingle with the general population. Secrets stay secret this way.

• Keep up the driving, relentless schedule for success. Do things the conservative way the first time around. Make it clear that failure will be ruinous for the participants. Push, push, push for results.

Now some of these things might seem obvious or not. The handling of espionage information is not what most people in the West generally assume is what you would do — they don't just take the spy info and turn it into blueprints. In one sense, the Berian suspicion of the spy data was counterproductive, because they didn't exploit it for maximum speed at all. On the other hand, the experience of spying in World War II is that countries that overly trust their spies and moles tend to get double-crossed by them (Cf. Germany), and with technical matters it's always possible that even if you represented the American work correctly, the Americans themselves might just be wrong on some matters (and indeed in a few cases they were — the Americans concluded that centrifuges were not a useful form of uranium enrichment during the war; successful centrifuges were first developed in the Soviet program, in part because they ignored that conclusion). In its favor, this resulted in a Soviet nuclear program that never felt dependent on foreign "assistance" (voluntary or not) — being forced to repeatedly "reinvent the wheel" meant that they always knew exactly how the wheel worked. (There is evidence that countries that rely too much on external assistance tend to have far less capable and flexible programs as a result.)

Another counterproductive approach was the push for the spies in the field (which is to say, the professional NKVD/GRU/etc. spies, who should be distinguished from the "moles" at places like Los Alamos) to constantly be producing results under threat of death. This does generate a lot of material, but it turns out to also generate some exaggerations and false information, as historians have later found. Soviet spies in the US claimed to have recruited many more people than they actually did, as part of this pressure to perform, and it is clear that the Moscow NKVD sometimes had an incorrect picture of what was going on abroad. (On the other hand, the spies in the field did not falsify any of the intelligence info that I know of, just their own record of progress. So the actual intel was still what is was. One might argue there was more of it because of this drive.)

Lastly, it's worth noting that many of the Berian tactics here would be totally incompatible with a more free society (totally closed, secret cities, for example, would have stopped Manhattan Project leaks and espionage, but would not have been acceptable to Americans who are used to freedom of movement), and lead to a pretty oppressive work environment (being constantly threatened with your life is not a pleasant way to live). It is not obvious that this is the most efficient or fast way to get things done; the Soviet program cost about the same amount of resources as the Manhattan Project but did take somewhat longer to accomplish the same ends. There are reasons beyond management that come into play here (the Soviet Union was uranium-poor, and poorer in general), but I just want to make sure one doesn't come away with the impression that this is a "better" way to run a nuclear program — it is a "different" way to do it, and certainly not the worst from the perspective of end results, but it's not obviously the best, either.

How much of this is attributable to Beria? He was very hands-on in this, and a lot of research has shown that who you have at the top of your nuclear program (in the Beria or Groves-like roles, not the Oppenheimer or Kurchatov-like roles) matters a lot. Nuclear programs without hard-drivers who are almost monomaniacally focused on getting quick results tend to meander and take decades to get success (if they get it). And there needs to be a firm linkage between the hard-driver at the top and the top scientists, as there were with Groves and Oppenheimer, Beria and Kurchatov, so that this institutional push can be coupled with good technical decisions.

So in a sense, you would need someone like Beria running things to get those results. Could you have gotten different results with someone else? Depends on who else one imagined, I suppose. I don't know how much of the above is specifically about Beria and how much would be considered normal NKVD operating procedure. But the participants in the Soviet bomb program seem to credit a lot of the above to Beria specifically, for whatever that is worth.

For more information:

• Holloway, Stalin and the Bomb (best account of the Soviet nuclear program from the inside)

• Gordin, Red Cloud at Dawn (focuses on comparing the Soviet program and the Manhattan Project in terms of information use and organization)

• Hymans, Achieving Nuclear Ambitions (makes arguments about the importance of internal organization and strong institutions to the success of nuclear programs)