The subject often gets stated rather loosely, that Bitcoin miners mine Bitcoins. In reality, they mine blocks.
To understand the difference, one needs to understand a basic principle in accounting. If the only types of transaction that exists were such, that the amount of funds need to be withdrawn from one account, before they can be forwarded to another account, then the sum total of the balances of the accounts could never change. And when Bitcoin was starting up, there were initially zero Bitcoins in existence.
But with Bitcoin there exists another type of transaction, which forwards an amount to a mining pool, as a reward for the fact that they have solved the hash-difficulty associated with mining one block. This amount is also referred to as the “block reward”, is not taken from an existing balance, and is capped according to a consensus that surrounds Bitcoin, and that is shared by all the mining pools. I don’t know how many Bitcoins constituted the original block reward, but as I’m writing this, the block reward equals 12 Bitcoins.
Every 210,000 blocks, this reward gets halved, and in such a way that it gets rounded down to the nearest Bitcoin. The last time the reward was halved was in the year 2016, just before which it was at 25 Bitcoins, and the next time will be in the year 2020. Hence, in 2020 the reward will be reduced to 6 Bitcoins.
But this block reward serves more of a purpose, than just to act as a reward to the mining pool, for having solved the hash-problem of one block successfully. It also determines the sum-total of Bitcoins in existence. The mining pool is expected to spend at least some of this amount, thereby forwarding it to other parties in the network, and causing a positive amount of Bitcoins to exist between wallet-holders.
Therefore, when miners mine blocks, they are also mining Bitcoins.
But as I wrote above, in 2020 the reward will go down to 6, in 2024 it will go down to 3, in 2028 it will go down to 1, and in 2032, the block reward will go down to zero ! And so the question can exist, as to whether Bitcoin will stop working for that reason. And the answer would be No.
In addition to collecting a block reward for mining 1 block, it needs to be remembered that each block potentially records thousands of regular transactions, initiated by wallet-holders, and that miners collect transaction fees for these transactions.
Since the current block-size today is limited to 1MB, and since a so-called typical transaction only takes up 500 Bytes, it could be said that 1 Block can record up to 2000 of these transactions. Since the transaction fee today is approximately 2 millibits per kB, this means that in addition to the block reward, the miners could be receiving up to 2 Bitcoins per block in transaction fees.
Once the block reward has gone down to zero, the assumption everybody is making today is, that miners will continue to mine blocks, but only collect transaction fees. Further, these transaction fees are taken from the holdings of the people who initiated the transactions, but with the expectation that mining pools will also spend them again. This means that by default, the transaction fees will not get sucked out of the system. Those will continue to operate, on the premise that the total amount of Bitcoins in existence will not change.
And so in theory, wallet-holders should still be able to send each other Bitcoin-amounts. Ultimately, it’s difficult for anybody today to know with certainty, what the realities of the year 2032 will be. The entire technological landscape could be very different from what it is today, or it could be disappointingly similar. We don’t know.
But one thing we do know, is that if money-minded people devise a currency-scheme, or an investment-scheme, and if those people are also honest in how they do it, it is not a habit to invest a lot of money, and then to decide on one day to shut down the whole system, and to say, as of 2032 the system will be worthless, because an activity needs to take place beyond that point in time, to allow the investment to continue earning income, but that that activity will just become mathematically impossible, in one sudden year.
What we do know, is that at the time I’m writing this, more than 75% of the Bitcoins that will exist, have already been mined. These are ultimately just numbers and not discrete entities. We also know that the mining of blocks does not need to keep creating new Bitcoins, in order to earn income for the mining pools. The blocks contain transactions, and the pools receive transaction fees.
Now, we could take this analysis a bit further, and ask, ‘Is it fair for the block reward to get halved every 210,000 blocks?’ I suppose that depends on who it’s supposed to be fair to. By holding back the actual amount of Bitcoins, the people who designed the system were limiting supply, in the expectation that demand will continue to increase, which also increases the value of 1 Bitcoin, in terms of other existing currencies.
This will make it more expensive to ‘buy’ 1 Bitcoin, but also means that if the Bitcoin is being used as a currency, products and services will end up costing smaller fractions of a Bitcoin. At the same time, while some people have purchased Bitcoins as a kind of investment, the value of those will increase, in that the Bitcoin-price of anything else will decrease, so that those people are likely to be kept happy.
OTOH, There does not only exist the supply of actual Bitcoins. There also exist limitations in block size – which are currently capped at 1MB – which restrict the total number of transactions that can take place. Again assuming that each block can only edify 2000 transactions, and that only 1 block can be mined in 10 minutes, this means that only 2000 transactions per 600 seconds will be supported – while the true size in bytes of 1 transaction will vary – but these assumptions also imply that a maximum of ~3 transactions can take place per second, on average. This will place a serious obstacle in the way of Bitcoin actually becoming a currency. We would not see that the maximum number of transactions which can take place in ?Canadian dollars? would be capped at 3 per second, and so it should not be for Bitcoin. Unless, this was just supposed to be some sort of commodity, for the sake of people speculating in that commodity.
But then it also follows, that transaction fees will start to increase. It might be nice to say that a certain product will only cost 0.55 millibit. But when it also turns out that the fee for 1 transaction has become 1.00 millibit, because that transaction takes up 500 Bytes of a block, then the transaction fee has exceeded the value of the product or service, so that to buy this product or service becomes unviable. Therefore, a shortage in transactions does not have the same implication, as an increase in the value or cost, of a unit of a currency, or of a commodity.