With all the talk about increasing inflation, slowing output growth, and stagflation in recent days and how monetary policy ought to respond, I thought it might be useful to repost a graph showing estimates of the long the lag between changes in policy and changes in macroeconomic conditions.
The estimates shown in the graphs imply that the rate hikes at the beginning of the current set of increases, those that began in late June 2004, are still working their way through the economy. For example, the peak effect on inflation takes two years to be felt, so the largest effect on inflation from the very first rate hike in June 2004 won't be realized for another month. The biggest effect on output today comes from policy implemented a year and a half ago:
Rerun of part of post from from June 16, 2005:
...A paper in the February 2005 issue of the Journal of Political Economy by Lawrence J. Christiano and Martin Eichenbaum of Northwestern University, the NBER, and Federal Reserve Bank of Chicago, and Charles L. Evans of the Federal Reserve Bank of Chicago entitled “Nominal Rigidities and the Dynamic Effects of a Shock to Monetary Policy” provides evidence on this issue:
Nominal Rigidities and the Dynamic Effects of a Shock to Monetary Policy, JPE [subscription link only]: This paper seeks to understand the observed inertial behavior of inflation and persistence in aggregate quantities. To this end, we formulate and estimate a dynamic, general equilibrium model that incorporates staggered wage and price contracts … the model does a very good job of accounting quantitatively for the estimated response of the U.S. economy to a policy shock. … A key finding of the analysis is that stickiness in nominal wages is crucial for the model's performance. Stickiness in prices plays a relatively small role.
Here’s Figure 1 from the paper displaying how output, inflation, and other variables respond after a shock to the federal funds rate:
Model- and VAR-based impulse responses. Solid lines are benchmark model impulse responses; solid lines with plus signs are VAR-based impulse responses. Grey areas are 95 percent confidence intervals about VAR-based estimates. Units on the horizontal axis are quarters. An asterisk indicates the period of policy shock. The vertical axis units are deviations from the unshocked path. Inflation, money growth, and the interest rate are given in annualized percentage points (APR); other variables are given in percentages.
For those of you who aren't used to reading graphs like these the exercise is fairly simple. Start with an economy at equilibrium, then hit it with a single shock, in this case a federal funds rate shock, and see how the economy responds. The diagrams show two models, one is an estimated model (called a VAR model in the diagram) with a very flexible form that allows it to match actual data fairly well. The other is a simulated theoretical model called the benchmark theoretical version (benchmark because other version are investigated with varying degrees of price and wage stickiness, different policy rules, and different assumptions regarding price setting behavior). We are mostly interested in the VAR results for this discussion, but what the diagram shows is that the benchmark theoretical model does a fairly good job of matching actual patterns in U.S. data.
Let’s now turn to the question about lags in the response of inflation to changes in the federal funds rate. Here’s their summary of the results:
The results suggest that after an expansionary monetary policy shock,
- output, consumption, and investment respond in a hump-shaped fashion, peaking after about one and a half years and returning to preshock levels after about three years;
- inflation responds in a hump-shaped fashion, peaking after about two years;
- the interest rate falls for roughly one year;
- real profits, real wages, and labor productivity rise; and
- the growth rate of money rises immediately.
To me, it’s always surprising how long the effects of a monetary shock last; the peak effect on inflation takes two years to unfold and the effect on output peaks after a year and a half and takes three years to fully unwind. Let me add one more note. In his spare time, when he’s not blogging, David Altig also looks into these issues (from the citations in the paper):
Altig, David, Lawrence J. Christiano, Martin Eichenbaum, and Jesper Linde. 2003. "The Role of Monetary Policy in the Propagation of Technology Shocks." Manuscript, Northwestern Univ. An updated version of the paper is available as a Cleveland Fed Working Paper "Firm-Specific Capital, Nominal Rigidities, and the Business Cycle"
These are estimates and thus subject to uncertainty, but the point is that the lags are fairly long by these and most other estimates. Also, remember that there are data lags (e.g. GDP is only available with a one quarter lag and is subject to substantial revision after that), recognition lags (does one or two quarters of weakness require a response, or is it a temporary aberration?), and implementation lags (this is short for monetary policy) on top of the lags that exist once the policy is actually implemented. Thus, there is a delay between changes in conditions and changes in policy since data must be collected, analyzed, and then policy must be formulated, followed by another delay once the policy is actually implemented. This is why expected future conditions are so important in policy discussions. Incoming data are important because they help to shape these forecasts of future conditions.
Update: Thomas Hoenig, president of the Kansas City Fed is also thinking about policy lags:
Hoenig Says Fed Must Guard Against 'Overshooting' on Rates, by Greg Ip, WSJ: A Federal Reserve policy maker said the central bank has to be mindful of the lagged impact of previous increases in interest rates and the risks of "overshooting" with future rate increases.
Thomas Hoenig, president of the Federal Reserve Bank of Kansas City, said while recent inflation readings have been a bit higher than he anticipated, they result from prior accommodative monetary policy and he expects inflation and economic growth to moderate in response to the most recent rate increases.
"One of the things I've learned is … that monetary policy works with a lag, but it's hard to appreciate that when you're in the midst of the cycle," Mr. Hoenig said in an interview Friday with The Wall Street Journal.
"That's really our challenge right now: to be careful in judging where we are in the cycle as we have removed accommodation in the past, and [ask], is that enough, or do you need to go more?"
He added, "I'm still very much opposed to allowing inflation pressures to get out… But you don't want to be so dogmatic that you're not taking into account … how lagged [monetary policy's] effects are." ...
In the interview, Mr. Hoenig said economic growth has been a bit stronger and core inflation a bit higher than he expected. But he said both are a result of prior accommodative monetary policy, while the impact of the latest increases in rates will be felt in the coming year. ...
"I thought, given the higher energy prices and strength of the economy, we would see some pass-through from the total [consumer price index] into the core and we've seen that. It is a little stronger than I thought." But he said if he's right in his forecast that the economy will slow towards its potential growth rate starting in the current quarter, "some of that [inflation] pressure should come off. To what degree, I don't know yet. We have to continue to look at the data as it flows through."
He added, "Anywhere from six, nine [months] to a year from now is very much where our actions are going to have their greatest influence." ...