A is actually the problem right now, not B. There are some things you can buy with bitcoin, but not much. This can change over time.

Reason B is that the Federal Reserve backs the US dollar and will help prevent inflation. This is, essentially, a forward-looking argument — they WILL prevent inflation. There is some historic evidence, since the government has often (though not always) done this in the past, but the crux of the B argument is forward-looking. It’s also self-fulfilling. The Federal Reserve backing & inflation prevention make for a good reason to believe precisely BECAUSE lots of people care about preventing inflation and promises to buy them back.

B, for bitcoin, is actually its most compelling component, and one you didn’t mention. And it is not technological, it is economic: that the rate of release of new currency is fixed by algorithm, and not able to be manipulated. This is very, very important to a large number of people. This is, after all, why we have a Federal Reserve in the first place: to prevent political meddling. Except many people believe it didn’t completely work, and bitcoin provides a better method of preventing the political manipulation of the money supply.

This is huge. It’s more than enough to give many people a “reason to believe,” which is all B really is — a reason to believe this currency has economic benefits over another.

The economic breakthrough is the predictable, algorithmic growth of the money supply, unable to be meddled with by politicians. That is real, that is new, and that is a very very big deal to many, many people.

(Update: I used the word “consistent” instead of predictable originally. I have modified it, thanks to Rob C Grant pointing out that it’s not actually consistent. The point remains the same.)

The question, then, is that a positive innovation (as in, it actually will make a difference in how money works) or a normative one (as in, people just want it to)? This is a very valid question, but one, I think, the positive answer to is still unknown, since we’ve not really tried it before, have we? Definitely worth exploring, and a concrete economic — not technological — reason why many people are excited about it.

This is, after all, nothing more than a build-a-better-mousetrap version of one of the reasons we invented a Federal Reserve in the first place.

If nothing else, bitcoin could provide incentive for governments to not tamper with the money supply.

One other aspect: DeLong’s gold analogy is flawed. Gold doesn’t quite work the way DeLong describes, and everyone knows it. The mining technology component is more or less the same for gold and bitcoin — mining technology will get more expensive. Where DeLong’s incorrect is in thinking as if there is a bottomless supply of gold. This is, of course, physically impossible, barring medieval alchemy, and everyone knows it. Estimates vary on the amount of gold left in the ground, but they all fall within a pretty narrow range.

The only difference between bitcoin and gold**, then, is that a level of uncertainty exists for the finite amount of gold left in the ground, and zero uncertainty exists for bitcoin. This aspect of gold is common knowledge and inherently factored into the price, right now, just as it is with bitcoin. Yet still gold makes for a viable currency.

To say this couldn’t be the case for bitcoin, then, would be to claim that you need a level of uncertainty >0 regarding a total currency supply to be viable*, which is actually a fairly interesting claim, and worth exploring. But to my knowledge, this hasn’t actually been proven.

* I suppose one could argue that you might alternatively need a level of uncertainty on the rate of increase of mining costs, but it seems to me both gold and bitcoin share this.

** Gold, as DeLong points out, can also be used to make jewelry. This is true, but Bitcoin can be used to buy a pizza, and there are many many people out there who would rather have a pizza than a necklace. I kid.