The warning, and what it leaves out
"Deflation is dangerous" is one of the most reliably repeated statements in economics. It is taught, quoted in central bank speeches, and used to justify a permanent positive inflation target.
It is also a sentence about two different phenomena that happen to share a name, and it delivers one verdict on both. Separating them is the whole argument, and it turns out the honest version does not end where the Bitcoin version usually ends.
Two mechanisms, one word
Productivity deflation happens when goods get cheaper because they cost less to produce. Better tooling, better logistics, more competition. Prices fall, real incomes rise, and nothing is wrong. This has been happening continuously inside the consumer basket for decades in computing, lighting, textiles and long-haul travel, and nobody has ever called for a bailout of the electronics sector because televisions got cheaper.
Credit-collapse deflation is the opposite in origin. Bank credit is money, and when lending stops and defaults rise the money supply contracts. Prices fall not because there is more to buy but because there is less to buy it with. The 1930s are the reference case.
Both are recorded as a negative number in the price index. They arrive from opposite directions and they mean opposite things about the state of the economy. A policy framework that reads only the sign of the number cannot tell them apart, and that is a real problem rather than a rhetorical one.
The strongest case against me
The serious objection is not "people will stop spending because prices might fall next month". That version is easy to knock over, which is presumably why it gets attacked so often. The serious objection is Irving Fisher's, and it is a mechanism.
In The Debt-Deflation Theory of Great Depressions (Econometrica, 1933), Fisher set out a loop. Falling prices raise the real value of existing debts. Debtors sell assets to service them. Distress selling pushes prices down further. The real burden rises again. Each turn makes the next turn worse, and because it feeds on itself, a small initial fall can produce a large collapse.
Add the constraint that policy hits: nominal interest rates cannot be pushed far below zero, so when prices are falling, real rates rise exactly when you would want them to fall. The central bank runs out of the tool at the moment it needs it most.
That is not a straw man and it is not a lie told to protect bankers. If leverage in an economy is high enough, mild deflation genuinely is destabilising, and anyone arguing for sound money who has not engaged with Fisher is arguing with the easy version.
Why the objection concedes the point
Read Fisher's sequence again and notice where it starts. It does not start with falling prices. It starts with over-indebtedness. Deflation is the trigger that lights the fuse; the explosive is the debt that was already there.
Which means the standard defence of a positive inflation target is not really the sentence it appears to be. "We must have inflation because deflation is dangerous" is, unpacked, "we must have inflation because the system carries more debt than it could service at stable prices". That is a statement about the balance sheet, not about the price level, and it is a much less comfortable thing to say out loud.
It also has an uncomfortable corollary. If persistent inflation is what keeps a levered system solvent, then the policy does not merely tolerate the leverage. It underwrites it, and underwriting something reliably produces more of it.
What the arithmetic does to a borrower
The transfer is easy to see, so it is worth doing exactly rather than by feel.
Take a 100,000 dollar debt, no interest, repaid in a single payment ten years out. At five percent annual inflation, the purchasing power surrendered at repayment is 100,000 divided by 1.05 to the tenth, or about 61,400 dollars in today's money. The borrower repays roughly 61 percent of what they took.
Run it at two percent annual deflation instead. The repayment is 100,000 divided by 0.98 to the tenth, or about 122,400 dollars in today's money. The borrower repays roughly double what they took, on a loan whose stated interest rate was zero.
Nothing changed except the sign on the price level. That single arithmetic fact is why the two camps in this argument are not really disagreeing about economics. They are on opposite sides of a transfer.
For context, the US target that produces the first column is two percent, and it is defined in terms of personal consumption expenditures prices rather than the CPI (Federal Reserve). The stated engineering reason for a positive target rather than zero is exactly the Fisher problem: it buys room above the zero bound and errs away from the spiral. That is a coherent reason. It is also, unavoidably, a standing transfer from people who hold currency to people who owe it, and the largest single borrower in most currencies is the government that defines the target.
Malinvestment, using our own industry as the example
The Austrian claim is that cheap credit does not just move money around, it funds the wrong things: projects that look viable at a manipulated cost of capital and are not viable at a real one.
The usual illustrations are somebody else's, which makes them easy to accept and easy to dismiss. So take the cleanest recent example from inside this industry. Between 2020 and 2022, bitcoin-adjacent lending grew a credit structure with very little visible collateral discipline, and when the price fell it unwound in weeks rather than years. We covered Celsius halting withdrawals and Three Arrows Capital being ordered into liquidation, and what both showed is that the entities failed on leverage, not on the price.
That case cuts against the tribal version of the argument, which is why it is the useful one. No central bank set the rates in that market. The malinvestment was produced by leverage, opacity and the ordinary human appetite for yield, and a hard-money denominator did not prevent any of it. The lesson transfers in both directions: leverage is the variable, and it does not need a central bank to become dangerous.
Where the gold standard actually went
The historical bit is usually told as a morality play, and the mechanics are more mundane. Under a metallic standard the money supply grew roughly with the supply of the metal, which limited discretion and also removed the ability to respond to a banking panic. Governments suspended convertibility in wars and restored it after, until they stopped restoring it: on 15 August 1971 the United States ended the dollar's convertibility into gold, which brought the Bretton Woods system to a close (Federal Reserve History).
The trade made then was explicit. Discretion was gained and the constraint was given up. Everything since has been an argument about whether that was a good exchange, and the answer depends entirely on how much you trust the discretion.
What Bitcoin actually is here, precisely
It is worth being exact, because the loose version of this claim is wrong in a way that matters.
Bitcoin's supply is fixed, not shrinking. A fixed supply is not the same thing as deflation. Prices measured in bitcoin fall only if demand for it grows faster than the economy denominated in it, and that is a fact about demand, not a property of the protocol. Calling bitcoin "deflationary by design" overstates what the code does.
What the code does do is remove the discretion. There is no committee that can decide the target should be three percent this year, and there is no lender of last resort. That second half is a genuine tradeoff rather than a footnote: the absence of a backstop is precisely what makes the supply schedule credible, and it also means a credit system built on top of bitcoin would have nobody to call in a panic. The 2022 unwind above is what that looks like in miniature, and it was allowed to run to completion, which is either the point or the problem depending on where you sit.
Where that leaves the argument
Falling prices are a symptom, and the same symptom shows up in a healthy economy getting more productive and in a levered one coming apart. The variable that decides which is how much of the economy owes money it could not service if prices stopped rising.
That is a choice about leverage, and it is made continuously, by lenders and borrowers and by the policy that prices their risk. Framing it as a choice about the price index is what allows the leverage to keep growing while everyone argues about the wrong number.