I have been mentioning this postulated attack on Bitcoin for months in various posts of mine. I figured it was time to give it a thread, so we can discuss it.
I think it is an economic attack, so I place it in the Economics forum. Also because I don't get good reception from Bitcoin developers when I try to post in the developers forums. Lets see if they ignore this thread or come post to refute it. I doubt they will.
Once Bitcoin's coin rewards decline to less than can pay for the miner's costs, e.g. <1% per annum debasement by 2033 and <0.2% by 2040, then transaction fees are supposed to fund miners. The following attack applies whether transactions are voluntary, variable, fixed, or mandatory-- it makes no difference.
But a cartel (e.g. Amazon.com) could for example harvest transactions from its vast network and keep them without forwarding them to other miners. Then put them on the blocks found by its own mining servers. This would starve the rest of the network of funding and eventually the cartel would be doing all the mining. They could even offer 0% transaction fees (even refund mandatory tx fees) to entice more of the masses to process through their servers.
That is the same as turning Bitcoin into a centralized currency, and thus eventually controlled by the government and thus back to fiat again.
Note this postulated attack wouldn't be possible for 20 years or so, so this is a long-term issue. The problem is if we wait, it will be too late to undo and revert, because we only get one chance to create a digital currency that the masses adopt. Once they adopt one, they will stay with that one due to inertia and network effects.
Thus I see Bitcoin is doomed and it is not a solution to anything long-term, although short-term it shows us what might be possible with decentralized currencies if we were to improve them.
http://hackingdistributed.com/2013/11/08/fairweather-mining/#comment-11263785531) declining block reward and constant gains from monopoly fees
I believe I am the first person to raise that in my
Bitcoin : The Digital Kill Switch article? I am naming it the
"transactions withholding attack" since it means not forwarding transactions in order to monopolize transaction fees, as coin rewards diminish.
Do you know of any prior art to mine? Do you know of any discussion on this attack other myself constantly mentioning it and no one seemingly willing to discuss it? (because the only solution I see is to change Bitcoin's diminishing coin rewards supply curve)
Other recent discussion:
https://bitcointalksearch.org/topic/m.3607709
Related. I have shown there is no economic advantage to a money supply that is constant:
https://bitcointalksearch.org/topic/m.3609132https://bitcointalksearch.org/topic/m.3607535