In thinking about the economic outlook, I try to stay focused on underlying fundamentals because they determine the outlook for the economy over the medium run, the time horizon over which monetary policy can affect the economy. In my view, the underlying fundamentals supporting the U.S. economic expansion remain sound. These include accommodative monetary policy, household balance sheets that have improved greatly since the recession, continued progress in the labor market, a more resilient banking system, and low oil prices. There are risks around all forecasts, but my modal forecast has been that over the next two years the U.S. economy will continue to expand at a pace slightly above its longer-run trend, which I estimate to be about 2 percent; that the unemployment rate will remain slightly under its longer-run level, which I estimate to be about 5 percent; and that inflation will continue to gradually return to the Federal Reserve's 2 percent target.

It is quite reasonable that we will get close to the 2 per cent target on the current path that we are on. The economy is improving but at a relatively slow rate. The inflation rate has been improving but at a very slow rate as well. So I think we are on the right path. It is not obvious to me right now that we are necessarily going to overshoot.

But one of the things that I highlighted in that New Hampshire talk is that [there are] both benefits and costs for waiting. The benefits tend to be accruing to the labour market and hopefully getting labour force participation to improve...

But there are also some potential costs. We talked about one, which is financial stability. A second one is that in the past we have tended to overshoot the natural rate of employment and when we overshoot significantly on the natural rate of unemployment it takes a while for the inflationary pressures to pick up but we are not so good at slowing down the economy and getting right at the NAIRU.

So if you look at the unemployment rate curve you tend to blow through the natural rate and when you start tightening you tend to get a recession shading. So it tends to be much more abrupt than we were hoping. Which means we are not like a thermostat where you can easily calibrate the difference between 70 degrees and 68 degrees . . . When we start tightening we tend to get more of a reaction in the economy than we are hoping for.

So it makes me a little tentative to hope we would overshoot significantly, because I am not so sure we are good at fine-tuning the economy. History would say we haven’t. Ideally what you would see is long periods where we are around the estimates of full employment. That is not actually what you see.

I think there are three elements to my concern. One was I thought that our description of the economy was overly pessimistic, that our language we used was not as optimistic as it might have been.

And I also think that the signal we send about 2013 signaled that we were pretty pessimistic for some time to come. I thought that was not appropriate.

I didn`t think the data justified it at this point. But even if we did justify that in terms of the outlook, I was concerned that now was not the time to make a policy decision.

The data had been very volatile. The stock market had been very volatile, and that it was inappropriate to react on a very short-term basis.  Monetary policy needs to be focused on the intermediate to longer term.

And I thought we were overreacting to the events. So I would prefer us to sort of sit tight for a little longer, let the data come in to get a better handle on the outlook and forecast.

We have been very careful to calibrate our actions so as to accommodate the needs of credit markets and the economy, not political imperatives. We are well aware that some of our balance sheet additions, designed to pull markets and the economy from the edge, have raised a few eyebrows (like the $1.25 trillion in mortgage-backed securities we have pledged to purchase if necessary—although it has unquestionably driven mortgage rates to historic lows). And while it is not unusual for the System Open Market Account to buy Treasuries along the yield curve, the FOMC’s decision to purchase $300 billion in U.S. Treasuries—a decision made to improve the tone in the private credit markets—has been viewed by some as skating a little too close to the edge of political accommodation.

I can tell you that the FOMC is well aware of the doubts being voiced about its intentions. I can also tell you that nobody I know on the committee wants to maintain our current posture for any longer and to any greater degree than is minimally necessary to restore the efficacy of the credit markets and buttress economic recovery without inflationary consequences. Indeed, as I speak, we are studying ways to unwind our balance sheet in a timely way.
...
[T]here are concerns that the Federal Reserve will be politicized. For example, some have called for increased congressional involvement in the selection of Federal Reserve policymakers and a reduced role for member banks. I trust that the Congress will resist this initiative and not upset the careful federation that has for so long balanced the interests of Main Street with those of Washington, just as we at the Federal Reserve must resist the urgings of some to accommodate the short-term financing needs of the Treasury.

{The debate over rules and descretion} is always on our minds. And for me, in the present circumstance, it arises in how you calibrate your concern about growth. Real interest rates have to fall when growth slows, but you want to communicate that you're not using your discretion to just shift your attention from one objective to the other. It's important for people to know that we're maintaining a continual focus on inflation. It's just that the state of the real economy justifies lowered growth now. You don't want it to be interpreted as we're throwing inflation overboard for a few months while we worry about growth. So it's letting people know that we're following what we view as a time-consistent strategy to commit to keeping inflation low, and that we're not just acting on a purely discretionary basis.
...
The Taylor Rule has proved incredibly useful as a very convenient way of summarizing algebraically and approximating in a fairly impressively good way, the way central banks behave. But keep in mind that Taylor Rule is a function of just things you can observe. Actual realized inflation recently. And what it leaves out are a lot of things policy makers can observe. What their forecast is, what's going on in oil markets, a lot of other data that go into our thinking, and that can affect our sense of how the economy's going to evolve. And those things, in turn, affect our policy setting. That doesn't mean we throw the Taylor Rule overboard. On the contrary. It's something that you want to deviate from only with good reason and only being very careful. And the farther you get away from it, the more you have to kind of check yourself as to whether you're really on solid ground or not.

As always, my view on economic growth and inflation emphasizes longer-run conditions. I could point to numerous past episodes of either faster or slower growth for a few quarters that we now ignore because long-run developments dominated the outcome and indeed dominate our current assessment of these periods. In assessing short-run developments, it is also essential to keep in mind that forecasts have standard errors. Over a four-quarter horizon, a GDP forecast has a standard error of about 1.5 percentage points and an inflation forecast has a standard error of about 0.5 percentage points. We know also that data are often revised. Finally, monetary policy cannot affect near-term conditions anyway. Thus, a focus on medium- and long-term fundamentals is always appropriate.

A few minutes ago I made the point that monetary policy cannot influence the economy’s growth or employment potential over the long run. The fact is that economists have had a very difficult time identifying any reliable and exploitable link between monetary policy actions and real output or employment in the short run either. Policymakers have neither the knowledge nor the tools to manage aggregate demand with the timing and precision necessary to neutralize the impact of unexpected shocks on output or employment.

Nonetheless, it is the economy’s best interest for policy to respond to conditions in a manner that is consistent with the goal of price stability. So, for example, if stronger productivity growth enables the economy to sustain higher output growth, then the market will demand a higher level of interest rates…

…Now, I recognize that the policy approach I am advocating here is difficult to implement. We often don’t observe economic shocks in a timely way, nor are we able to precisely measure the level of interest rates the economy is seeking in response to such shocks. Consequently, monetary policy should not be overly sensitive to short-run fluctuations. Thus, keen judgment is called for, and a little luck doesn’t hurt either.

But I believe the principle behind the approach is sound. Indeed, it is simply an elaboration of my first principle: the central bank should always set interest rates consistent with the goal of maintaining long-run price stability so as to foster maximum economic growth and employment. Moreover, trying to use monetary policy to stabilize output and employment in the short run — can actually do more harm than good.  A look back at the 1970s underscores the point quite dramatically. When the economy was hit with a series of oil price shocks, the Fed responded with stimulative policies intended to maintain output and employment growth. These policies largely failed, generating excessive inflation even as the unemployment rate rose and the economy weakened. The Fed was relying on a short-run relationship between economic activity and inflation that is simply not reliable.

Monetary policy must be forward-looking because policy influences inflation with long lags.  Generally speaking, a change in the federal funds rate may take an estimated 12 to 18 months to affect inflation measures. 

The existence of lags in monetary policy has two important implications.  First, the Federal Reserve should only respond to high current inflation to the extent that it is expected to be highly persistent.  if inflation pressures are seen to be temporary and policy is currently restrictive, maintaining the current policy stance may be consistent with a reduction in inflation over time.  Second, given the existence of policy lags, the actions that the Federal Reserve took over the past year in raising the federal funds rate have not yet had their full effects on the economy or inflation.