With a global slowdown in full force, the Federal Reserve along with all central banks in the world are grappling with economic conditions for which there are few relevant examples. To deal with this, central banks must rely on economic models. Although imperfect, models offer the only sensible way to organize information about the economy. The most common models employed today are called New Keynesian models. Developed in the 1980s, these models synthesized two earlier approaches to understanding economic cycles.

New Keynesian models combine John Maynard Keynes many observations about the Great Depression with the traditional explanations of an economy that is largely self-correcting in the face of typical economic shocks. This has been a powerful innovation in our understanding of economic fluctuations, allowing economists to explain how small frictions and bubbles can add up to big swings in economic activity around the globe.

This sort of modeling approach differs from most other social sciences where competing schools of thought rely on strong arguments to swing opinion. In economics, what matters is almost wholly the quality with which these models fit the observed data. Indeed, almost nothing else matters.

Today, one of Keynes principal observations about the Great Depression is proving most vexing. It is the condition under which an infusion of money into a banking system does not reduce interest rates enough to increase lending. The old school name for this was the liquidity trap, but it is today referred to as the zero lower bound, as in interest rates cannot drop below zero. This makes sense—with interest rates below zero, lenders would be paying folks to borrow.

The liquidity trap, or ZLB, offers a huge risk to the global economy because it may eviscerate our most effective policy tools against an economic downturn. I use the word ‘may’ carefully here because there is reasonable disagreement as to the effectiveness of policies such as quantitative easing to boost the economy by pushing real interest rates into negative territory.

Indeed, there are currently many negative interest rates when adjusting for inflation. Many of our federal securities have yields lower than inflation, and if you park your money in a savings account you might well be earning a negative real interest rate. This is because core inflation over the past year was 2.2 percent, which is well above today’s typical savings account rate.

Zero or negative interest rates are a signal of huge underlying problems for obvious reasons. If businesses are unwilling to borrow money for free (or less), then that means they do not expect a positive return on that borrowing. The clear implication is that their expectations about the future are dismal. This is why it is called a liquidity trap, because there is no clear policy escape.

Central banks now are beginning to act against this liquidity trap, and the Fed’s increase in rates last December offer a modest buffer on this ZLB rate. Still, as much of the world slides into recession it is unclear whether public policy has much more to offer as a remedy.