CAMBRIDGE – The low rate of inflation in the United States is a puzzle, especially to economists who focus on the relationship between inflation and changes in the monetary base. After all, in the past, increases and decreases in the growth rate of the monetary base (currency in circulation plus commercial banks’ reserves held at the central bank) produced – or at least were accompanied by – rises and falls in the inflation rate. And, because the monetary base is controlled directly by the central bank, and is not created by commercial banks, many believe that it is the best measure of the impact of monetary policy.

For example, the US monetary base rose at an annual rate of 9% from 1985 to 1995, and then slowed to 6% in the next decade. This decelerating monetary growth was accompanied by a slowdown in the pace of inflation. The consumer price index (CPI) rose at a 3.5% rate from 1985 to 1995, and then slowed to just 2.5% in the decade to 2005.

But then the link between the monetary base and the rate of inflation was severed. From 2005 to 2015, the monetary base soared at an annual rate of 17.8%, whereas the CPI increased at an annual rate of just 1.9%.

To explain this abrupt and radical change requires examining more closely the relationship between the monetary base and inflation, and understanding the changing role of the reserves that commercial banks hold at the Federal Reserve.

When banks make loans, they create deposits for borrowers, who draw on these funds to make purchases. That generally transfers the deposits from the lending bank to another bank.

Banks are required by law to maintain reserves at the Fed in proportion to the checkable deposits on their books. So an increase in reserves allows commercial banks to create more of such deposits. That means they can make more loans, giving borrowers more funds to spend. The increased spending leads to higher employment, an increase in capacity utilization, and, eventually, upward pressure on wages and prices.

To increase commercial banks’ reserves, the Fed historically used open-market operations, buying Treasury bills from them. The banks exchanged an interest-paying Treasury bill for a reserve deposit at the Fed that historically did not earn any interest. That made sense only if the bank used the reserves to back up expanded lending and deposits.

A bank that that did not need the additional reserves could of course lend them to another bank that did, earning interest at the federal funds rate on that interbank loan. Essentially all of the increased reserves ended up being "used” to support increased commercial lending.

All of this changed in 2008, when a legislative reform allowed the Fed to pay interest on excess reserves. The commercial banks could sell Treasury bills and longer-term bonds to the Fed, receive reserves in exchange, and earn a small but very safe return on those reserves.

That gave the Fed the ability in 2010 to begin its massive monthly purchases of long-term bonds and mortgage-backed securities. This quantitative easing (QE) allowed the Fed to drive down long-term interest rates directly, leading to a rise in the stock market and to a recovery in prices of owner-occupied homes. The resulting rise in household wealth boosted consumer spending and revived residential construction. And businesses responded to this by stepping up the pace of investment.

Although a link between the Fed’s creation of reserves and the subsequent increase in spending remained, its magnitude changed dramatically. The Fed increased its securities holdings from less than $1 trillion in 2007 to more than $4 trillion today. But, rather than being used to facilitate increased commercial bank lending and deposits, the additional reserves created in this process were held at the Fed – simply the by-product of the effort, via QE, to drive down long-term interest rates and increase household wealth.

That brings us back to the apparent puzzle of low inflation. The overall CPI is actually slightly lower now than it was a year ago, implying a negative inflation rate. A major reason is the decline in gasoline and other energy prices. The energy component of the CPI fell over the last 12 months by 19%. The so-called "core” CPI, which excludes volatile energy and food prices, rose (though only by 1.8%).

Moreover, the dollar’s appreciation relative to other currencies has reduced import costs, putting competitive pressure on domestic firms to reduce prices. That is clearly reflected in the difference between the -0.2% annual inflation rate for goods and the 2.5% rate for services (over the past 12 months).

Nonetheless, inflation will head higher in the year ahead. Labor markets have tightened significantly, with the overall unemployment rate down to 5.4%. The unemployment rate among those who have been unemployed for less than six months – a key indicator of inflation pressure – is down to 3.8%. And the unemployment rate among college graduates is just 2.7%.

As a result, total compensation per hour is rising more rapidly, with the annual rate increasing to 3.1% in the first quarter of 2015, from 2.5% in 2014 as a whole and 1.1% in 2013. These higher wage costs are not showing up yet in overall inflation because of the countervailing impact of energy prices and import costs. But, as these temporary influences fall away in the coming year, overall price inflation will begin to increase more rapidly.

Indeed, the inflation risk is on the upside, especially if the Fed sticks to its plan to keep its real short-term interest rate negative until the end of 2016 and to raise it to one percentage point only by the end of 2017. If inflation does rise faster than the Fed expects, it may be forced to increase interest rates rapidly, with adverse effects on financial markets and potentially on the broader economy.

The writer is Professor of Economics at Harvard University

© Project Syndicate