I like a good contrarian argument. So I was open to Ryan Avent’s post on Trump’s fiscal policy proposals (which combine massive tax cuts with big spending increases.)

On the demand side, permanent tax cuts for the very rich would be expected to have a multiplier far smaller than tax cuts for poorer workers or direct spending, but even so, new deficits of several percentage points of GDP could not help but raise demand (assuming the Fed did not react to the plan by raising interest rates sharply). It would be an inefficient stimulus, but it would be a stimulus, and given that the American economy is still close to the zero lower bound and is probably still operating with a fair amount of economic slack, such a stimulus should increase growth in real output for a year or two at least. In the short run, the economy would not crash (not as a result of Mr Trump’s tax cuts, anyway).

What about the longer run? Well, $34 trillion in new debt over the next two decades would be a lot, without question. It probably wouldn’t be a catastrophic amount, however. The Congressional Budget Office estimates that American GDP will be around $40 trillion in two decades. That suggests that Mr Trump would raise the ratio of gross debt to GDP by about 80 percentage points relative to what it would otherwise be. On top of the growth CBO already projects for American government debt over the next two decades, that would push up gross debt to GDP to about 220%.

That’s a lot! On the other hand, it implies that America would have a ratio of debt to GDP two decades from now that is 30 percentage points less than Japan’s government debt is right now.

In the end, Ryan isn’t trying to defend Trump’s proposals, but rather to make a different point:

The point here is not that Mr Trump’s plans are sensible. To be clear: they are ridiculous. It is to illustrate an uncomfortable truth, though, which is that the world economy right now is also a little ridiculous. Government bonds are not supposed to be doing what they’re doing. Countries are not supposed to be able to borrow like Japan is borrowing. So while playing to people’s gut instinct that fiscal sobriety is good is probably good politics, it might also be an obstacle to political efforts to respond in a sensible way to the weirdness of the economic moment: by making a positive case for deficit-financed public investment and for a change in central bank targets to something better suited to the moment, for example. Dangerous demagogues like Mr Trump thrive in conditions like those now afflicting the global economy. If the adults in the room remain unwilling to take macroeconomic challenges seriously, Mr Trump’s debts will be the least of our worries.

Notice that Ryan alludes to the fact that the US debt ratio is currently expected to rise to about 140% of GDP over the next few decades, even without Trump’s tax cuts. This trend has worried many economists, and there’s a rough consensus today that we need to get a better handle on entitlement spending. On the other hand, Japan’s gross debt is already 250% of GDP, and there are no obvious signs that this level of debt is unsustainable.

Ryan might be correct. It’s possible that the US might be able to dramatically increase the debt ratio over the next few decades, with no ill effects. Nonetheless, I think it would be very unwise to do so.

Let’s start with the demand stimulus. It’s very unlikely that major fiscal stimulus would boost demand at all—instead the Fed would offset the effects with higher interest rates. People often argue that monetary offset does not apply at low interest rates. That’s wrong. It might not apply if the current interest rate were above the level preferred by the Fed. But that has not been true for years. Ryan also mentions to supply-side effects, but you can get most of those with sensible tax reform, without blowing a hole in the budget. Lower marginal tax rates, close loopholes, and switch from taxing investment to taxing consumption.

My other concern is that we need to expect the unexpected. Can we really trust CBO projections about the future path of variables such as NGDP? For instance, I very much doubt that America’s NGDP will be $40 trillion in 2 decades. In 2005 and 2006, lots of Americans took out mortgages, secure in the knowledge that American housing prices don’t suddenly plunge by 30%–until they did. The Greek government ran up large debts, secure in the knowledge that modern European economies don’t see their NGDP crash by 25%–until it did. Once you’ve run up large debts, it’s too late if things don’t play out as you’d hoped.

Suppose our debt is 220% of GDP, and interest rates return to 5%. Now we’ll be spending 11% of GDP on just interest payments. To put that number in perspective, it’s about three times what we spend on the military. Do I think interest rates will rise to 5%? No, I do not. But back in 2006, when interest rates were 5%, I had no idea that they were about to go to zero.

A massive national debt is playing Russian roulette with the economy. Japan is quite sensibly trying to reduce its budget deficit, by raising consumption taxes. Or at least they were doing so, until the Abe government decided to splurge on fiscal stimulus, partly in response to the advice of Western academics. Japan should have adopted NGDP targeting, done whatever it takes to hit their target, and simultaneously raised consumption taxes even further (or cut spending.)

Of course Ryan is not a fan of Trump’s plan, but I would even oppose his more modest proposals. I don’t see any benefit in further raising our debt, as we move into an era where entitlement spending will put increasing pressure on the budget. If there are infrastructure projects that make sense, then by all means do them, hopefully with as much private sector participation as possible. Maybe we could privatize our airports, the FAA and Amtrak. But we should pay for any government projects with higher taxes on consumption—preferably the consumption of more affluent people like me.