I’m very sympathetic to Gracy Olmstead and Connor Friedersdorf in their arguments against getting rid of cash. I think the utopian arguments for a cashless society are kind of silly, and the privacy implications of getting rid of cash are arguably not trivial.

However: I also think the dystopian implications are kind of silly, and that by and large the privacy ship has sailed. Right now, all of our most important financial transactions are electronic; only criminal enterprises do any significant business in cash. And the examples that Friedersdorf and Olmstead use to prove the continued utility of cash frequently prove the opposite.

Already, you have to be a serious obsessive to avoid leaving an electronic footprint to be read by Big Data – that’s the lesson of Janet Vertesi’s story. ATMs and credit cards already make it trivially easy to spend more than you have in your wallet at any given moment. As for the utility of cash for the poor, 90% of African-American adults own a cell phone, as do 84% of those who earn less than $30k/yr; by contrast, over 20% of African-American households are unbanked, and an even higher percentage of low-earners. The existing financial infrastructure is abandoning the poor, making reliance on cash more and more expensive for them, even as mobile technology is penetrating more fully.

Getting rid of cash is not going to eliminate crime or the black market, and it’s not going to abolish recessions. Nor is the development of an effective electronic wallet going to suddenly put poor people on a level playing field with the wealthy in terms of the daily costs they face executing simple transactions. Rather, this is another one of those situations where we’ll have to keep moving just to stay in place.

These are reasons to hope that electronic alternatives to cash continue to develop, and make cash more and more obsolete. But that’s no reason to abolish cash entirely. The main reason to abolish cash entirely is a technical one, related to the economics of demographic decline.

In a world where we expect robust economic growth, interest rates will be positive, and the yield curve will be upward-sloping – that is to say, a lender who lends for 10 years gets a higher rate than a lender who lends for 1 year. If expectations for economic growth drop to low levels, or even to negative levels, then long-term rates drop as well.

Now, when demographic growth is positive, both real and nominal growth will generally be positive. But when demographic growth is negative, the only way real growth will be positive is if productivity growth is high enough to compensate for the “burden” of negative demographic growth. And this is rarely likely to be the case for countries on the technological frontier.

So for those countries experiencing demographic decline, the rational expectation is for real growth to be zero or negative over the long term. Which should translate into low or negative real interest rates. And, in the absence of high inflation expectations, low or negative nominal interest rates.

Which is where we run into a problem. Because, when long interest rates are low or negative, there’s no incentive to invest in longer-dated assets. It makes more sense to hoard cash. When that’s a cyclical event, we call it a recession.

What is cash? Cash can be thought of as a bearer bond with no maturity date and an interest rate of zero. In what we think of as a “normal” world of positive nominal interest rates, cash is a poor investment; almost anything else has a higher expected yield. But in a world of low or negative nominal long rates, cash will be a persistently attractive investment vehicle. Which is to say: investment in actual productive assets will be persistently unattractive. Which is a real problem for capitalism.

What’s the solution? One solution would be higher inflation expectations. This would drive nominal rates up, and make cash less attractive. But there are several problems with this solution. First of all, once inflation gets going, it may be hard to keep it from accelerating. This is emphatically not our problem today – we’re continually teetering on the edge of deflation – but if we’re imagining a world where we’re relying on inflation to keep us out of persistent recession, we’re also imagining a world in which we have to worry about an inflationary spiral.

But perhaps a bigger problem is: how do you engineer higher inflation expectations? The main mechanism central banks have is to purchase financial assets – which they have been doing, in very large volumes, without appreciably causing a rise in inflation. Now, maybe the problem is one of psychology – maybe they just need to say that they will keep buying assets of all kinds until inflation starts to tick up. But there’s a chicken-and-egg problem: expectations of institutional behavior don’t change easily. What fundamentally changes them is proof that they have changed. When Paul Volcker allowed unemployment to go through the roof to kill inflation, that proved that the Fed was genuinely serious about fighting inflation – and expectations changed, permanently. Expectations that the Fed is serious about creating inflation won’t become manifest until we actually get inflation, and the Fed declines to combat it.

Meanwhile, the zero-bound set by cash creates problems for the Fed in engineering actual inflation. The Fed can drive very short-term rates down by making short-term loans to financial intermediaries – injecting cash into the system. This should spur the expectation of higher economic activity, which in turn should spur higher long-term rates – an upward-sloping yield curve. Which is what you want to get out of a recession. But short-term rates can’t go below zero because they are competing with cash. So the Fed has had to make longer and longer-dated loans – this is “quantitative easing.” But QE is a paradoxical strategy, whereby the Fed has to buy long-dated assets in order to drive their price down. Is it surprising that it has had only middling success?

Get rid of cash, and the problem goes away. Nominal rates can now go below zero, because there is no competing instrument that carries a zero nominal interest rate. And if nominal short-rates can go below zero, then you can engineer an upward-sloping curve even if long rates are low or zero – reflecting low long-term growth expectations.

I want to stress: I’m not saying that central banks will be able to wish away recessions if cash were eliminated. I’m saying that central banks have a hard time fighting their way out of situations where long rates are close to zero. In a world of positive demographic growth, very low long-term rates are a sign of something seriously wrong. In a world of negative demographic growth, very low long-term rates will be normal. If we don’t want high unemployment to also be normal, we need better tools for fighting unemployment when long rates are very low. Which may mean that we need to get rid of cash.

Again, it’s not that we need to get rid of cash to usher in utopia. We may need to get rid of cash just to get back to adequate management of monetary policy. We may need to run faster just to stay in place.

The United States will not be the first country to go cashless; we’ll probably be among the last, along with the UK. Singapore will probably go first, and Japan will probably be the first large country to make the transition, for a variety of reasons – the advanced progress of demographic decline, lack of comfort with immigration as a stop-gap solution, comfort with technology, less concern about privacy, less attachment to the artifact of physical currency. So we’ll have time to observe how the transition goes elsewhere before contemplating it ourselves.

And we may not follow the same route. We may keep cash for a long time for sentimental reasons. But we are already living in a world in which cash means something very different than it did a generation ago. Reliance on cash is already not a practical way to protect privacy or to reduce our reliance on banks. So if we want a more decentralized future, we’ll likely have to think about achieving that within the context of cashlessness.