The Mining Centralization Myth – Part 1

hqdefaultIn this 2-part article, I have decided to address some of the very commonly spread myths in Bitcoin space, namely that of mining centralization, and its effects on the BTC price valuation.  At the end of reading this I hope that you will have a better understanding of the complicated topic of decentralization in terms of economic factors, and also how everything is perfectly reflected in the BTC price.  Also, I hope that you would have a new found appreciation that BTC price is going to continue to fluctuate wildly and even may go to zero, under certain certain scenarios.

But first, to the often repeated, and universally unsubstantiated claim:  That Bitcoin is suffering from miner centralization.  First off, I want to stave off all the thoughts that the proponents of this opinion are thinking now: “You must not have heard of this thing called the Great Firewall of China!”, “You must not understand what the propagation delays are and its affect on orphan rate!”, “You must not know about the mining relay network!”, or my favourite, “Maybe you haven’t heard, but they have this corrupt oppressive government over in China!”.   Rest assured, I know, I have heard all the arguments, and I have well connected business associates in China.

The fact of the matter is, with each and every one of the people I talk to who have this view, when we get to the crux of the issue, it becomes a matter of “Protecting what little decentralization we have left”.   But when I ask them how they define decentralization, and how do they quantify it exactly, I generally get blank stares.  (virtual stares at least, I couldn’t confirm that jaws were dropping in disbelief on the other side of the monitors, but I could sense it.)  Firstly, we must assume we are talking about decentralization in terms of the ability of the network to resist corruption, and not just the physical definition of having a lot of computers in many different places.  Thus what we really mean when we speak of decentralization is a cumulative general notion of network resiliency or security.  So with this, a common quoted situation that should seem obvious to anyone: the more nodes spread out over a bigger topology (a fancy computer science term analogous to geography), spanning many political boundaries, is “better” that a few nodes in one place.  Obviously.  Right. But are you sure?  Yes? Okay good, then why is it better in terms of security of the network?  This is where I lose most people.  They know that more nodes are more secure, but they cannot say why, other than that it is “harder” to compromise the network, or to co-opt it to do dishonest things.  “Better”, “harder”…. great terms to use at dinner table discussions, but sadly not concrete enough to convince the professional skeptics and the critically minded.

When you think on this question long enough, or if you have an economic or financial background, you may already know the answer.  The reason why the first situation is more secure is because it costs much more to compromise the network, where it costs relatively little in the latter situation.  What the important point to note is, it is not the number of nodes that matter, it is the cumulative cost of compromising more than 51% of them that matter.

That cost can come in the form of bribery, incentive, or threat of violence, it doesn’t matter, so long as someone is willing to pay it.  The other important point to note is that the more liquid money that these nodes have themselves, the more they are resistant to external coercion.  To illustrate, simply think about how much it takes to bribe your local highway patrolman (someone with little money) vs bribing a congressman, or the President of the United States.  So security depends on the cost or the amount of money you need to PAY to usurp the network.  This money needs to be an external cost to the system, i.e. bribes in the network currency don’t really work, because you are paying actors to sabotage the system itself, therefore devaluing their bribe*.

If you have followed this line of reasoning, then it will be very clear to you that the current mining situation in Bitcoin isn’t all that bad.  In fact, it is as decentralized as it currently needs to be.  Could it be better? Perhaps, but in order for it to be more decentralized, money needs to be fed into the economy to pay for it.  This is why small block proponents want a fee market to develop so that people can pay more for transactions, in order to incentivize more decentralization or at the least to prevent it from getting worse.  Trouble is, this is a red herring.  Any money artificially thrown into the market will be absorbed as profits for the current miners and not significant enough to stimulate new miners to join the industry. This is because most of the value in mining comes from the block reward and only about <1% from user fees.  Students of monetary theory may realize a stark similarity of this strategy to that of Keynesian economics, where spending money to stimulate an economy artificially have been known to have long term negative effects.

So the next time somebody says that Bitcoin mining is ‘broken’ and how having 100,000 individual nodes is much more decentralized than all mining power being concentrated into 10 companies in China, then you know how to respond.  The correct answer is a question: is the 100,000 nodes run by 16-year olds in their basement?  And how large are these 10 mining companies in terms of capital invested, what are their profit margins, and projected future revenues?  (and which is cheaper to buy off?) The answer to which, will determine whether or not their idea of decentralization is based on theory or actual economic reality.

In part 2, I will discuss the price components of Bitcoin and how mining security contributes a major part to it.

*incidentally, this is why Proof of Stake will never work.

Tip me with ChangeTip!

Please donate!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.