Blockchain = Powerful Tool for Keynesian Monetary Policy

[Roger_Murdock]

Suppose the central bank controls 51% of hashing power and wants to achieve a stimulus equivalent to a 3% increase in inflation.

Simply demand a txn fee equal to 3% of coin-age (with age measured in years). This is just like instantaneously increasing the inflation rate by 3%. Take the fee proceeds and hand them out to banks. Voila. You have stimulated spending. Once the economy recovers, the txn fee can be lowered again. The inflation rate of every single block is completely at the bank's discretion. This is 100% impossible with the tools available today.

Now people might try and horde (the bank can't steal my coins... I'll just wait till they lower the fee and spend then). Ha, this would not work at all. The bank has a historical record of its inflation policy and would bill these guys. The fact that you haven't paid yet won't help them. The seignorage will be deducted from their account whenever they finally decide to spend.

Why would you pay the central bank's exorbitant transaction fees when there are other miners willing to process your transactions for less? I suppose the bank could attempt to use its 51+% hash power to orphan blocks mined by miners not playing by the bank's rules, but couldn't the rest of us simply retaliate? If the bank doesn't play by OUR rules, their blocks aren't recognized. They might have a lot of hashing power, but so what?  They opted not to use that power to gain 51% control of Bitcoin; they instead chose 100% control of Fedcoin (which seems like a strange choice). If that scenario occurs, I think I'll stick with Bitcoin.

[fergalish]

I see some merit in what cunicula has to say. However, I don't think any central bank could abruptly gain 51% (and thus, obviously, 100%) of hashing power - it would cause an instantaneous backlash, and there would be an immediate hard blockchain fork with the bank's servers being blockaded (caveat - what if the bank connects over TOR?).  However, in the far future of bitcoin, when many people depend on the blockchain for the salaries and groceries, then a central bank that managed to gain 51% surreptitiously could quite easily begin to levy a tax. They'd have to walk a fine line between rejecting untaxed blocks mined by others, and not arousing too much suspicion. And it would be easier as bitcoin became more accepted as a mainstream currency.

I have to say, the development of ASICs has me plenty worried. Hypothesize if you will a scenario where bitcoin mining is restricted to licensed entities only. Well, everyone can still have their GPUs for gaming, and even FPGAs have legitimate uses. But ASICs that only know how to SHA256(SHA256()) could have only one plausible, and therefore illegal, use. No-one could easily compete on mining, and since building ASICs requires big factories and investment, it would be easy enough to shut down unlicensed manufacturers. Think, the Treasury would be replaced by an ASIC factory.  I would be *very* happy to see the devs change the mining protocol to be ASIC unfriendly, if such a thing is possible; even better would be to make it FPGA and GPU unfriendly, and bring back CPU mining only -- only then will mining be truly fully distributed.

Of course, if this scenario came to pass, you'd still have other miners mining away, probably on an underground black-market blockchain - like I said before: "the blackchain":

...[if one entity gains 51%] you'd end up with an 'official' blockchain, where gov't salaries would be paid, and tax [as tx fees] would have to be paid in that chain.  And [there would also be] a black market chain, obviously illegal - the 'blackchain' 

[Rassah]

But ASICs that only know how to SHA256(SHA256()) could have only one plausible, and therefore illegal, use. No-one could easily compete on mining, and since building ASICs requires big factories and investment, it would be easy enough to shut down unlicensed manufacturers. Think, the Treasury would be replaced by an ASIC factory. 

Illegal where? Which country's treasury? I think you're falling into the old "Bitcoin only exists within one country, and thus falls under one set of laws" fallacy.

[davout]

I am kind of confused

No shit

[mcgravier]

Please continue this discussion as long as possible, it makes me soooo haappy 

[bitCooper]

I think this hypothetical scenario is unrealistic. The cost of obtaining 51% control is so high that such an investment would be highly risky, given price volatility, especially if the plan is to implement a controversial monopoly which could cause the price to collapse.

As long as the government continues to control its own fiat currency, there's no need for it to hijack bitcoin in the near future. If some misguided policy folks thought it would be a good idea after bitcoin became hugely popular, then there would be a public outcry which would be a costly risk to any politician.

What is more realistic is that a government could hijack bitcoin in order to introduce massive volatility and cause panic selling, in an effort to undermine crypto currencies. However, we have already seen bitcoin survive a massive crash, so I don't know if that would work anyway.

[Mike Hearn]

You need to be really careful with terminology when explaining that attack. It isn't "equivalent to a 3% increase in inflation" in any way except that it encourages spending. You'd do better to phrase your argument like,   "Governments use inflation to encourage spending. They could also encourage spending by charging tax on movements of old coins".

I think this is an interesting academic exercise. I don't believe it has any impact on the real world.

As noted by others, this wouldn't be implementable because if a government wanted to try this, there would absolutely be some kind of warning that it was coming. Users would then start bouncing transactions around to keep their apparent age low. It'd be impossible to keep such a plan secret, if only because governments and central banks are set up to be at least a little bit transparent. Meeting notes from the UK MPC for example, are publicly available:

   http://www.bankofengland.co.uk/publications/minutes/Pages/mpc/default.aspx

Anyway, if a government decided it monetary policy was more important than Bitcoin it would simply ban usage of Bitcoin and force everyone back to state issued currency. I don't believe there would be any reasonable argument for doing so, because I believe in a purely Bitcoin based economy there would be no business cycles and thus no need for monetary policy.

If you think about recessions in the recent past (the last 30-40 years or so), how many of them were caused by something fundamental like a natural disaster or disease? None of them. How many were caused by malfunctions of the financial system? All of them.

If you build a more stable financial system the need for centralized economic planning to try and stimulate spending would go away. There'd be no justification for it any more.

[fergalish]

But ASICs [would be] illegal...

Illegal where?

One answer: in your country!  Alternative answer: in any country. The attack I outline could be carried out by any country and its citizens could therefore be subjected to perfect financial regulation. Other countries might seek to join that blockchain, or they might have one of their own.

I think this hypothetical scenario is unrealistic. The cost of obtaining 51% control is so high that such an investment would be highly risky,

I disagree. Current network hash rate is ~ 25TH/s. BFL says they'll give you 1.5TH/s for $30k in a single box. You'd need only about 20 of them to gain > 50% of the hash rate, that's $600k. Six hundred thousand dollars! So let's suppose someone tries to set up this attack in the future when there are already, let's say, 10000 such boxes already running and connected. You'd still only need about $300 million (ignoring price reductions due to efficiencies of scale, and improved technology) - that'd still be small change to a central bank that's printing, what, $40 billion per month? Suppose there were 1 million such boxes running - that'd cost $30 billion, 8 months of QE3. Let me emphasize, the attack would have to be surreptitious so as not to raise suspicion. ASICs coming online would be the *perfect* moment to implement the attack because nobody will be worried at sudden massive increases in hashes-per-sec.

ASICs will be great for bitcoin if they become truly distributed, but it should be obvious to all that they'll never be as uniformly distributed as CPUs. Any argument on that point?

You are wrong, because you assume anyone would use a currency with a 51% weakness already in progress, and because your formula doesn't account for the initial investment in hardware required to achieve that 51%. You are only accounting for operating expense with that $212 million.

If civil servants get paid in a particular blockchain, then they'll use it. And if the police understand that their bitcoins become more valuable by beating on 'blackchain' exponents, then you can be sure they'll do just that.

[Rotsor]

It isn't "equivalent to a 3% increase in inflation" in any way except that it encourages spending.

But it is! If attack is successful, then it's equivalent to (transparent) inflation in every way except the position of decimal point.

IUsers would then start bouncing transactions around to keep their apparent age low.

Fail. Do you know losing 3% twice is the same as losing 5.91% once?

[bitCooper]

...

Anyway, if a government decided it monetary policy was more important than Bitcoin it would simply ban usage of Bitcoin and force everyone back to state issued currency. I don't believe there would be any reasonable argument for doing so, because I believe in a purely Bitcoin based economy there would be no business cycles and thus no need for monetary policy.

If you think about recessions in the recent past (the last 30-40 years or so), how many of them were caused by something fundamental like a natural disaster or disease? None of them. How many were caused by malfunctions of the financial system? All of them.

If you build a more stable financial system the need for centralized economic planning to try and stimulate spending would go away. There'd be no justification for it any more.

While I generally agree that a stricter, more predictable money supply would reduce the boom and bust cycles, I don't think we will ever eliminate business cycles completely. It seems to me that "easy money" exacerbates bubbles like the dot com and real estate bubbles, but you can still get over-investment without the government's "help."

[cbeast]

It is pretty clear that the blockchain gives central banks unprecedented control over monetary policy. All they have to do is command 51% of hashing power. An absolutely negligible investment for such a well-capitalized institution.

Remember the Boston Tea Party? These servers will be easy to locate and isolate from the bitcoin network. Block their IPs, and if necessary, permanently.

[Mike Hearn]

But it is! If attack is successful, then it's equivalent to (transparent) inflation in every way except the position of decimal point.

What is "transparent inflation"?

The proposed mechanism wouldn't cause prices to rise. You're just sucking money out of one part of the system and re-allocating it to a different part. It's a tax that redistributes money, taking it from the market and putting it into the hands of banks. It doesn't cause prices to rise any more than income tax does.

But I don't think there'd be any justification for that. If you assume things like the distributed bond markets get implemented then individuals would probably invest money anyway when good opportunities presented themselves. The assumption behind forcibly re-allocating value to banks is that banks will lend whereas individuals wouldn't. The only plausible explanations for why that would be are asymmetry of opportunity (which can be rectified with better technology) or that the banks would make worse or riskier lending decisions than actually make sense. Hardly a great outcome.

Fail. Do you know losing 3% twice is the same as losing 5.91% once?

My point was about the systems introduction. There has to be a point before this tax is applied and after it. If people learn about the tax before it is switched on, all the coins will end up with basically similar ages that won't change much and it amounts to a flat tax on savings. This is not demurrage nor inflation.

At any rate, I think this kind of argumentation is a poor use of time. It boils down to "if something external actor with great power forces the rules of the system to change, the rules of the system will change" which isn't a particularly interesting discussion. So what?

[bitCooper]

I think this hypothetical scenario is unrealistic. The cost of obtaining 51% control is so high that such an investment would be highly risky,

I disagree. Current network hash rate is ~ 25TH/s. BFL says they'll give you 1.5TH/s for $30k in a single box. You'd need only about 20 of them to gain > 50% of the hash rate, that's $600k. Six hundred thousand dollars! So let's suppose someone tries to set up this attack in the future when there are already, let's say, 10000 such boxes already running and connected. You'd still only need about $300 million (ignoring price reductions due to efficiencies of scale, and improved technology) - that'd still be small change to a central bank that's printing, what, $40 billion per month? Suppose there were 1 million such boxes running - that'd cost $30 billion, 8 months of QE3. Let me emphasize, the attack would have to be surreptitious so as not to raise suspicion. ASICs coming online would be the *perfect* moment to implement the attack because nobody will be worried at sudden massive increases in hashes-per-sec.

ASICs will be great for bitcoin if they become truly distributed, but it should be obvious to all that they'll never be as uniformly distributed as CPUs. Any argument on that point?

If civil servants get paid in a particular blockchain, then they'll use it. And if the police understand that their bitcoins become more valuable by beating on 'blackchain' exponents, then you can be sure they'll do just that.

My point is that while some groups could hijack bitcoin, if the price collapsed to $1 or less, the investment would become pointless from a financial and influence point of view.

I have no idea how evenly distributed asics would be. They don't yet exist. It all depends on the price of bitcoins. If the price continues to rise, more and more people will be able to afford high end mining gear, ASIC or otherwise. If the price crashes, then ASIC manufacturers may go bankrupt and people won't care as much about hash rate anyway.

[cunicula]

Anyway, if a government decided it monetary policy was more important than Bitcoin it would simply ban usage of Bitcoin and force everyone back to state issued currency. I don't believe there would be any reasonable argument for doing so, because I believe in a purely Bitcoin based economy there would be no business cycles and thus no need for monetary policy.

You are saying that the government would simply ban bitcoin. Does that make any sense to you at all? If you were in the treasury or a central bank is that the solution you would recommend? Did you read the ECB report? What was the ECB concerned about in order of importance (reread the document if you are confused)

1) Financial Disruption Stemming from Bitcoin Collapse (essentially bitcoin is like mortgage backed securities. There is a bubble and if it grows to big it causes problems when it pops)

2) Loss of control over Monetary Policy

Do you think banning bitcoin sounds like a good solution if the Central Bank's main concern is number 1? No that is a ham-fisted approach. You don't intentionally cause the problem you are most concerned about.

What would a sensible gov't would do? For example, a government that wants financial stability, revenue from seignorage, control over the monetary system, tax enforcement, etc.. Well, how about regulating bitcoin? How do you regulate bitcoin? You mine of course. Then you can have control over the monetary policy, seignorage revenues, control money laundering, tax evasion, etc.

This policy is theoretically more effective at manipulating the monetary system than any tool the gov't currently has at its disposal. Keynes, debating Gisell, dismissed such tools because they were infeasible, even if better in theory. You think the government wouldn't want this new technology? Why not?

Sure, the loonies here talk about collapse and flight to the secret underground blockchain and other nonsense. News for you: If bitcoin were widely adopted, the majority of users would not be libertarian crazies. They would abide by regulations and continue to use the bitcoin because it is convenient and backed by the support of government. Sure the libertarian crazy could go do their thing. No one would care about that. Just like they don't care about it now.

I agree this debate is almost certainly purely academic. Why? Because bitcoin almost certainly will never be widely used. However, if it became widely used, the government would try to regulate quite narrowly. The most obvious way of doing that is as I describe. This is actually a good thing.

[Rotsor]

There are multiple states in the world. Cunicula, do you think they will all cooperate to launch the 51% attack?

[Rassah]

If a 51% attack is launched, and a government gets enough control of the currency to dictate where it goes, how much transactions cost, and how much of the currency is created, then how will Bitcoin be any different from the current central bank controlled fiat currency system? I think a government with your goals would be way more effective if it just slapped a unique id on every dollar, and tracked all the transactions from the central ledger it already has. End result will be the same, and it won't cost a ton in mining gear. Especially since the ONLY purpose of mining is to secure the system in a DISTRIBUTED manner. If the ledger is already centralized, there is no need for mining.

Though, I suspect the people would be a bit upset if they found out their government is planning on tracking every single dollar they spend. In much the same way Bitcoin users will be upset if they found out some entity is trying to gain 51% mining power. Also, you are still misunderstanding how Bitcoin works. 51% doesn't give you the control you claim it does. At most, the 51% holding government will be burning a ton of energy while rejecting almost all transactions and transaction fees and creating orphaned blocks, while the remaining pools will continue earning fees and creating blocks the general population would agree to be valid.

[fergalish]

Also this: http://gavintech.blogspot.com.ar/2012/05/neutralizing-51-attack.html
Basically, it says that a crude proof-of-stake can quickly and easily be bolted upon the protocol if and when the attack comes.

This is interesting, and it seems like a valid defence, though I can spot two weaknesses. First is that having more bitcoin-days is "easily" accomplished by having more than half the bitcoins. So you'd need 51% of hashing and 51% of bitcoins. Second is more subtle. If people get paid (e.g. civil servants) in the govt-approved blockchain, then they'll have to use that chain to spend their money. If there are lots of civil servants, then the govt blockchain will be correspondingly more important.

Remember the Boston Tea Party? These servers will be easy to locate and isolate from the bitcoin network. Block their IPs, and if necessary, permanently.

They'll switch to TOR. Is there any way to block a miner from participating over tor that doesn't involve compromising tor itself?

I think a government with your goals would be way more effective if it just slapped a unique id on every dollar, and tracked all the transactions from the central ledger it already has.

Not possible with paper money. VERY possible with a blockchain.

At most, the 51% holding government will be burning a ton of energy while rejecting almost all transactions and transaction fees and creating orphaned blocks, while the remaining pools will continue earning fees and creating blocks the general population would agree to be valid.

Depends proportionately on how many people depend on the govt-blockchain.


Let's pose the problem another way. Suppose tomorrow the US government issued a statement to the effect that they really like the idea behind bitcoin and would be commencing work on a new government-mined blockchain, complete with a new genesis block; obviously with some different mining and block verification procedures. Police, military, doctors, teachers, ALL employees in ALL branches of government would be paid in that new blockchain. How do you see that situation evolving?  Now ask yourself, how is that situation any different from executing a surreptitious 51% attack on the current blockchain?

Actually, thinking about that scenario myself, it would seem to be a much more logical path for government to follow. Hah, I just shot myself down 

[bitCooper]

...
Let's pose the problem another way. Suppose tomorrow the US government issued a statement to the effect that they really like the idea behind bitcoin and would be commencing work on a new government-mined blockchain, complete with a new genesis block; obviously with some different mining and block verification procedures. Police, military, doctors, teachers, ALL employees in ALL branches of government would be paid in that new blockchain. How do you see that situation evolving?  Now ask yourself, how is that situation any different from executing a surreptitious 51% attack on the current blockchain?

Actually, thinking about that scenario myself, it would seem to be a much more logical path for government to follow. Hah, I just shot myself down 

I think you're exactly right. If the government found the *theory* of bitcoin to be better than paper fiat, it would be simplest to bootstrap govcoin with centralized components rather than try to hijack an existing crypto currency by brute force. Digital currency is a lot simpler if you replace the hashing power with centralized servers.

At that point, people wanting to avoid government inflation could trade govcoin for bitcoin after getting their paycheck.

Ultimately, if one or more governments decided to outlaw bitcoin, they could make things a lot harder for the community, but it would be very difficult to eradicate. Bitcoin won't be able to topple visa without government ambivalence or support.

[Rassah]

I think a government with your goals would be way more effective if it just slapped a unique id on every dollar, and tracked all the transactions from the central ledger it already has.

Not possible with paper money. VERY possible with a blockchain.

Dollars are already tagged with unique serial numbers. Just track who took out the paper, and who redeemed it, in the exact same centralized ledger that tracks all the other dollars. Same as with Casascius bitcoins.
Besides, blockchain IS just an accounting ledger. The only difference between it and the central bank ledger is who decides the rules that it must follow.

[cunicula]

Also this: http://gavintech.blogspot.com.ar/2012/05/neutralizing-51-attack.html
Basically, it says that a crude proof-of-stake can quickly and easily be bolted upon the protocol if and when the attack comes.

This is interesting, and it seems like a valid defence, though I can spot two weaknesses. First is that having more bitcoin-days is "easily" accomplished by having more than half the bitcoins. So you'd need 51% of hashing and 51% of bitcoins. Second is more subtle. If people get paid (e.g. civil servants) in the govt-approved blockchain, then they'll have to use that chain to spend their money. If there are lots of civil servants, then the govt blockchain will be correspondingly more important.

Yes, any proof-of-stake system of course helps a lot. In this context, the best thing about proof-of-stake, is that it forces the state/organization to provide owners with compensation for seizure of the blockchain (i.e. they have to buy out 51%). That more or less guarantees at least 51% of owners will be made better off by state/organizational control. By comparison, proof-of-work allows anyone to seize the blockchain without compensation.

The secondary advantage of proof-of-stake is that it makes seizure an order of magnitude more expensive. Consider the example I gave before. A lower bound on a proof-of-stake currency's market valuation is the discounted sum of txn fees associated with stake ownership.

Before we had:
It is worth pointing out that the investment is chump change and that the investment would be highly profitable. Say bitcoin replaces all card payments in the Euro area. That is about 54.8 billion Euros in 2008. Suppose that the average fee on bitcoin payments is 0.1%. Suppose that hardware depreciate at a 20% rate [slow for computer hardware], that labor costs are negligible [labor costs make monopoly easier], and that the interest rate is 5%. Assuming a competitive market, we have

capital rental costs=flow of bitcoin payments
(0.20+0.05)K=0.001*54.8 billion
K=$212 million.

Repeating this calculation using stake is difficult because the coin will also have uses as a medium of exchange and a store of value which add to its value. We can just ignore these and focus on fee collection.
capital rental costs=flow of bitcoin payments
(0.05)K=0.001*54.8 billion
K=$1.096 billion.

If we assume a lower interest rate (like in the economy today), then the difference is much more than five-fold. Suppose the real interest rate is 1%.
Then the attack costs are $260 million for PoW vs $5.480 billion for PoS, about a 20-fold increase.

It still won't stop a large state, but it would be a non-trivial obstacle for a private organization.