Rarely is the question asked: Why should we want to balance the budget? Deficit hawks tend to talk in general terms about not wanting to pass down a huge debt on to our children, or needing to avoid a debt crisis, but what this actually means is often left underspecified.

The actual concern is much more banal: The hawks are worried about interest rates. High levels of debt, the theory goes, suggest that the country holding that debt might not be able to pay it back. That risk leads bond purchasers to demand higher interest rates. And since government debt is generally the safest asset you can buy, that in turn means interest rates for everything from mortgages to credit cards to business loans shoot up. That means businesses are less likely to borrow money to invest in big projects and purchases, and consumers are less likely to borrow money to buy houses, cars and so on. This phenomenon is known as "crowding out," and it's quite bad for the economy.

But there's a big assumption embedded in that story: that increased government debt leads to higher interest rates. It seems intuitive enough, and many economists believe it. Just this past month, Morgan Stanley's David Greenlaw, UC-San Diego's James Hamilton, Deutsche Bank's Peter Hooper and Columbia's Frederic Mishkin put out a widely read paper claiming to show that the U.S.'s current debt path sets it up for high interest rates. The blue line below is their prediction, and it looks pretty grisly:

If bond yields really do creep up to 10 percent or more, that'll lead to the crowding out dynamics that deficit hawks fear. Greenlaw et. al. get the above result by extrapolating from a sample of 20 countries on which they ran data analyses to get a sense of how debt levels affected interest rates. But Matt O'Brien of the Atlantic notes that there is a huge problem with this sample. Here are the countries plotted out below, with each country represented four times (one for each year from 2008 to 2011):

Apart from a few outliers on the right, at first glance this appears to show a clear positive relationship between debt and interest rates: When debt goes up, so do borrowing costs. But O'Brien notes that much of the sample is made up of Eurozone countries, including countries like Portugal, Ireland, Italy, Greece and Spain (PIIGS) that have seen particular trouble. That really matters. Normal countries by definition never have to default on their debt. That's because they control their own currency and so can, if worst comes to worst, just start printing money to pay off creditors. That may not be the best approach (more on that later), but it's a real option.

By contrast, Eurozone countries don't control their own currency, and don't have that option. That means they face a much higher risk of default, which should in turn lead to investors demanding higher interest rates out of them. Countries like the U.S. that control their own currencies don't have that problem. And sure enough, when O'Brien ran the same analysis without Eurozone countries, he found literally no relationship between debt levels and interest rates. That's hardly dispositive, but that's not the point. "Don't pay attention to the evidence," O'Brien concludes. "Pay attention to the lack of evidence."

And indeed, there are a number of countries that make trouble for the conventional view that high government debt necessarily increases interest rates. Singapore and the U.K. both face low borrowing costs despite big debt loads, but the real standout here is Japan. Japan has a debt to GDP ratio of about 218 percent, which according to the CIA World Factbook is the highest government debt burden in the world. And yet, as of last Thursday, its nominal interest rate on long-term debt was a measly 0.67 percent, or about a third of the U.S. long-term interest rate (which is itself ridiculously low).

So what gives? Why is Japan able to have so much debt and such low interest rates? The technical term is "financial repression," or the government using regulatory or other tools to ensure that lenders keep buying government bonds and thus make debt more manageable. As economist Noah Smith explains here, the Ministry of Finance in Japan requires banks to hold a certain number of "safe assets," which in practice means Japanese government debt. That has the unfortunate side effect of making returns on savings in those banks very, very low for Japanese households. It's unclear why they're willing to accept this, aside from country-specific reasons like the low return on stocks in Japan in recent decades and a strong sense of national loyalty. But the effect of this policy is to keep interest rates very low even as debt is high.

That's not good for Japan, really. But could it work for the U.S.? Arguably, it already is. The U.S. is pursuing a number of policies that push interest rates down. The Fed, for one thing, is pushing down long-term interest rates through quantitative easing and short-run rates by keeping the federal funds rate extremely low (and in fact negative, in inflation-adjusted terms). The Basel III accord and Dodd-Frank both push bank liquidity requirements up, which mean that U.S. banks are pushed toward safe assets like, oh, I don't know, U.S. treasury debt.

Economists Carmen Reinhart (now at the Kennedy School) and Belen Sbrancia (at the University of Maryland) have argued that the U.S. pursued financial repression from 1945 to 1990, and that it amounted to a 3 to 4 percent of GDP reduction in debt every year. "For the United States and the United Kingdom, the annual liquidation of debt via negative real interest rates amounted to 3 to 4 percent of GDP on average per year," they write. "Such annual deficit reduction quickly accumulates (even without any compounding) to a 30-40 percent of GDP debt reduction in the course of a decade." If true, that means it played a very big role in reducing the U.S. debt load. And if the economic performance of the U.S. in those years is any indication, the ill effects weren't that huge.

"Financial repression" is a loaded term, but the fact remains that real interest rates in the United States are very low at the moment, in part due to government policies. Does that mean that we don't have to worry about the debt-to-GDP ratio growing out of control? Probably. As the economist Jamie Galbraith explains here, when real interest rates are negative, as they were from 1945 to 1980 and have been since the late 1990s, the debt-to-GDP ratio will continue to rise as long as there's a deficit, but it will approach a limit rather than starting to grow exponentially. Galbraith hardly speaks for most economists, but the equation he's using here comes courtesy of Willem Buiter, chief economist of Citigroup and a fairly mainstream guy. Here, for example, is what happens to debt-to-GDP if we have a real interest rate of -1 percent, a deficit of 5 percent of GDP, and economic growth of 3 percent a year:

The debt ratio keeps growing, but at some point it tops out. If Galbraith is right that negative real interest rates are the natural state of things in the United States — and with the exception of the 1980s, history seems to bear him out on that — then we don't have to worry about ever-rising debt-to-GDP ratios. If he's right, then ever-rising debt ratios in such a scenario are literally impossible.

So U.S. policymakers have a choice, if they want to avoid high interest rates. They can force interest rates low through Fed policy or financial regulations, as Japan has done and as the U.S. did for much of the last century. Or they can reduce the deficit. And the evidence that the former will work is arguably stronger.