A belief shared by every major economist, every peer-review published research paper on the subject, as well as more than a century of real-world data.
EVERY economist? EVERY peer-review published research paper?
But, to correct a rather clumsy distortion of my statement, it is not "the privileged" who should get such items, but rather "those willing and able to pay for them". In a free market, graphics cards, like Ferrari sports cars, go to those who have worked hard for the necessary funds, and wish to use a portion of them to this end. In the neo-communist egalitarian dystopia you envision, they indeed go only to "the privileged" -- high government officials, and those they favor.
Six of one, half-dozen of the other. And, yes, from other posts, you've made your opinion/aversion to communism/socialism/market-regulation quite clear. But your analogy doesn't fit here.
What you describe is not what happens in the GPU/scalper market. What happens is that SOME few manage to snag them before the scalpers and miners do. Neither the scalpers nor the miners benefit this system at all. That's something you've still not proven.
(snip irrelevant, self-disproving portion)
Your attempts to misstate my remarks are quite persistent. My actual statement was that the equilibrium price from the manufacturer would be lower due to several reasons: the primary one being the fact that manufacturers service the entire market curve, whereas scalpers serve only the upper portion. Deadweight losses involved in these particular scalper transactions are a factor, but a fairly minor one; I'd estimate 5% or less of the transaction. Regarding scalping/quasi-black market transactions in general, however, deadweight losses can be much higher. Research data on such sales in authoritarian regimes regularly see deadweight losses of 90% or more of the transaction amount.
Authoritarian regimes are not relevant to this conversation at all, ergo, I won't bother asking for proof of this cherry-picked research.
I have misstated nothing. You said:
Incorrect. The point you fail to grasp is that the price manufacturers need to set to balance supply and demand is lower than that extractable via scalping, as manufacturers service both ends of a "bell curve" gaussian distribution, whereas scalpers service only one. Scalper transactions additionally have a certain amount of deadweight loss which is avoidable in a direct transaction. Thus, the price manufacturers would be required to charge to balance supply and demand would be significantly lower .
That bolded conclusion is what you have failed to prove. You make the underlined claim, unsupported, which, while technically true, is true to a very insignificant amount. You then state the bolded-italicized conclusion, which is only a restatement of the underlined claim, with the word "significantly" dropped in there.
If manufacturers increase the price as you suggest, they are not servicing both ends of the bell curve... they are, like the scalper, closing off the low end. That is the consequence of an increased price.
The only thing you can possibly claim is that manufacturers will cut off less from the low end than scalpers will. The difference in how much the manufacturer doing what you suggest versus the scalper, will be small, because anything lower will re-introduce the scalpers.
That is, to requote what you said above:
Deadweight losses involved in these particular scalper transactions are a factor, but a fairly minor one; I'd estimate 5% or less of the transaction.
So, if the scalper's deadweight is ≤ 5%, then if the manufacturer brings the price to less than 95% of the scalper's selling price, then it is again worth it for the scalper to get involved.