For many reasons, such as stealing electricity and having access to chip manufacturers, there are a handful of miners in China who have the majority of the computing power for Bitcoin. They can mine bitcoins at a cost that is less than for everyone else.
This creates problems because there are updates to Bitcoin that are needed, such as allowing the Bitcoin network to process more than 7 transactions a second (for reference, Visa does 40,000 a second). Unfortunately, this small block size rate means people have to pay extra fees to prioritize their bitcoin transactions to happen in the next 20 minutes.
The miners get to keep these additional fees, so they are incentivized to keep the network slow and people paying more. This is why the Chinese miners have rejected any updates to improve Bitcoin transaction rate, meaning bitcoin is slower and more expensive for everyone but the miners extract more money from the network. It's tragedy of the commons.
This means that the people who get the most decision making power are incentivized to only make changes that help their investment. Any change that would help a majority of users, but harm these top users, is basically a non-starter.
This is incorrect and a common misunderstanding about Bitcoin. Hard forks require users to update, and it doesn't matter what the miners do. It also doesn't require a majority of users to upgrade, anyone who does upgrade will be on the new network, and anyone who does not upgrade will be on the old network.
Bitcoin (and other blockchains) are structured so that miners have as little power as possible. Pretty much the only thing miners can do in practice is choose to censor transactions, and that would be considered an attack. Networks have recourse like bricking all mining hardware, which typically acts as a sufficient deterrent to such attacks. Miners can also double-spend (a form of creative self-censorship), but the same network recourse applies.
In practice I don't think there are any examples of miners intentionally making a blockchain slower and more expensive. Miners pretty much always fit every possible transaction into every block, and any throughput restrictions are determined at the protocol level by protocol devs, not by miners.
This is why I reference the bitcoin block size debacle. Bitcoin Cash was created as a hard fork that miners ultimately did not accept and is now just a dwindling alt coin.
Segwit2x had 80% hashrate support at the time it was proposed. It also had the support of most of the exchanges and major centralized players in the space. And yet, Segwit2x did not succeed.
Miners don't control Bitcoin.
Block producers (miners) must find buyers for the blocks they produce, if they don't find buyers, they go bankrupt.
>changes to update the Bitcoin protocol need to be accepted by the miners
You've got it backwards, changes to Bitcoin need to be accepted by block consumers (node operators). If they don't demand those changes, blocks with those changes don't get produced.
Demand and supply are fundamental components to economic action. The steal man version of your argument is: "While consumers induce production, some consumers' demands might be flippant -- They signal they will only buy kosher bread but they'll accept an alternate good. Though production switching costs are practically zero in SHA256 PoW, and entry into production is non excludable, an adversary has enough funding to pay premiums to producers to forego market demand, for the good they produce, longer than consumers are willing to refrain from consumption -- inducing a consumer-demand shift."
In game theory and economics its not a dominating strategy which is indicative of the many failed attempts to cartelize SHA256 PoW
In a buyers market. With the small block size, bitcoin is a seller's (miner's) market and refusing to upgrade preserves their market power - hence the tragedy of the commons.
>You've got it backwards, changes to Bitcoin need to be accepted by block consumers (node operators).
The number of bitcoin nodes has been dropping for years. The rational actors who are incentivized to support the network are making the decisions right now by choosing which forks to support.
The produced good (SHA256 hashes and the transferable UTXO set of bitcoin nodes) is an excludable, rivalrous good. Hence it doesn't suffers from tragedy of the commons problems. This is the foundation of excludability in economics.
https://en.wikipedia.org/wiki/Excludability
"seller's" or "buyer's" "markets" are weak concepts that don't control which goods are produced. If an agent market sells or market buys that doesn't dictate which goods are produced.
Again, no amount of production nor no amount of consumption can get a consumer or producer to shift their consumption or production to a good they do not want.