Mining is already too centralized

Mark Friedenbach writes: "Bitcoin is already centralized. Do you realize that a cabal of a half-dozen people (I'm not talking about the developers) have the power, even if they have yet to exercise it, to arbitrarily control bitcoin? That this power also rests with anyone who controls the networks used by this cabal, which is presently confined to a small number of datacenters? That if they were in the US all it would take is a couple of national security letters for full control over the bitcoin network? I assure you the Chinese government has much stronger strings to pull.

The story of bitcoin over the last two years has been struggling hard to keep bitcoin decentralized in step with wider usage. In this effort we are floundering -- bitcoin scales far better today than it did in early 2013 (at which time it couldn't have even supported today's usage), but centralization pressures have been growing faster still."

Greg Maxwell writes: "So far the mining ecosystem has become incredibly centralized over time. I believe I am the only remaining committer who mines, and only a few of the regular contributors to Bitcoin Core do. Many participants have never mined or only did back in 2010/2011... we've basically ignored the mining ecosystem, and this has had devastating effects, causing a latent undermining of the security model: hacking a dozen or so computers--operated under totally unknown and probably not strong security policies--could compromise the network at least at the tip... Rightfully we should be regarding this an an emergency, and probably should have been have since 2011. This doesn't bode well for our ability to respond if a larger blocksize goes poorly. In kicking the can with the trivial change to just bump the size, are we making an implicit decision to go down a path that has a conclusion we don't want?"

Owners of mining pools have lots of power, since only a handful of them that control a vast majority of the hash power. They could coordinate and double spend, or censor transactions. Or a hacker could do one of those things if enough of them got hacked. Or, governments could compromise them and start censoring transactions.

Counterarguments

 * Although this raises the likelihood of a short term attack, these pool operators could not control Bitcoin for long, because power ultimately resides with users. If they started double spending or censoring transactions, miners would move to other pools. If miners who owned their own hardware controlled a majority of hash power and were behaving in a way that Bitcoin users objected to, there would be some sort of fork to reduce the power of these miners. As an extreme example, a hard fork could change Bitcoin's hash function and completely destroy the hardware value of whoever was controlling the network. It would be very painful for many in the Bitcoin community, but the threat of this provides a check on the power of miners.
 * It's unclear what exactly has caused existing mining centralization, which would be useful to know when trying to prevent further centralization. The argument that large blocks give miners a significant advantage due to block propagation latency could use more data. If individuals with mining equipment are joining centralized mining pools because it's the easiest option and not because running a full node costs too much, then we shouldn't worry as much about larger block sizes driving more miners to centralized pools or resulting in more mining farms.