The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« What's The Fed Up To? | Main | Turmoil At The EU Summit »

June 16, 2005

What's The Fed Up To? Part 2

In my previous post I noted that (a) for the period from June 2003, when the FOMC lowered the funds rate to 1 percent, through June 2004, when they began the current string of rate increases, the federal funds rate was somewhat lower than would be predicted by a Taylor rule describing "typical" Fed behavior on the basis of realized inflation and deviations of output from it's "potential"; and (b) that departure has disappeared as a result of FOMC rate decisions since that time.  Here's a version of the graph I showed in that earlier post, isolating the period since 2000.


In commenting on the Thoma/Ritholz Econblog debate, Max Sawicky has this to say:

Mark Thoma defends the Fed's "inflation-targeting" in today's WSJ econoblog debate. Mark is liberal, well-intentioned, and knows a lot more about monetary policy than me, but I can't get the past the fact that the labor market is anemic and the most likely relief can only come from the Fed.

Max's position seems to be that the FOMC's focus on price stability has driven it to ignore the real side of the economy.  In his part of the Econblog discussion, Mark does point out that price stability and output stability are not, within the boundaries of what monetary policy can achieve, necessarily at odds.  But even beyond that, the vision of a Fed that is impervious to real weakness in the economy just doesn't seem justified by the facts.  Here's a picture of what the Taylor rule prescription would be if we assumed the Committee decided to ignore the output gap completely:


So, it seems that even given falling rates of inflation over the period the FOMC was expressing concern about disinflationary pressures, the funds rate chosen was lower yet as a result of GDP falling short of its long-run potential.

Does that mean, as some have suggested, the central bank has downgraded it's concerns about the evolution of inflation?  Don't think so.  Here's a picture similar to the one above, assuming that the FOMC has been reacting to output gaps, but not inflation:


Put the two pictures together, and what do you get?  A monetary authority with a dual mandate, that faithfully -- and consistently -- pursues both.

A few technical notes for the geeks:

-- The estimated Taylor rule in these two posts is "inertial" -- it includes one lag of the funds rate, thus incorporating an interest-rate smoothing motive.  The empirical specification also includes  a first-order moving-average error term (although that really doesn't play a very large role).  The output gap is measured relative to Congressional Budget Office estimates of potential.

-- No, I did not use real-time data.  Yes, I know that is an important issue.

-- Although I do say that these Taylor rule experiments suggest the FOMC behaves "consistently," note that I did not say optimally.  As we all know, that is much trickier issue.

June 16, 2005 | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference What's The Fed Up To? Part 2 :

» INNOCENT QUESTIONS from MaxSpeak, You Listen!
Readable Rightie Dave Altig defends Yoda from our atavistic attacks with some dazzling graphics of Taylor rule simulations. Since Yoda refuses to answer these sorts of questions, maybe David will: Based on the most recent data, 1. How big is... [Read More]

Tracked on Jun 17, 2005 11:26:36 AM


The lagged funds rate fits the funds rate very nicely.

Posted by: Gerard MacDonell | June 16, 2005 at 11:14 AM

Great stuff! Thanks.

Posted by: CalculatedRisk | June 16, 2005 at 11:33 AM

I have a version of the Taylor rule that uses unemployment rather then the output gap so I can do it monthly -- normally the unemployment rate and the output gap as a percent of real gdp are virtually identical.

But my version of the Taylor rule goes all the way back to 1958 and fits extremely well from 1958 to 2000 except under Volcker.

But it implies that the Fed should have started raising rates in late 2002 and that fed funds should now be 5%

Posted by: spencer | June 16, 2005 at 12:32 PM

Don't want to belabor the technical, but since you mentioned it I am curious -- why did you use the MA(1) specification in your model estimation? And what is "real-time data"?

Great post!

Posted by: Mike | June 16, 2005 at 01:12 PM


Let's add the sentence to Max's statement that you left out in your post.

Max: "...I can't get the past the fact that the labor market is anemic and the most likely relief can only come from the Fed. Why should we care about five percent inflation if we can get more employment?"

Now let's go back to a question I raised on a previous post earlier this week (which you didn't answer, BTW) and compare our "official" unemployment to our broader unemployment/underemployment. (Germany - 11.8%; U.S. - 5.1% or U.S. - 8.6%. Germany provides 52 weeks of unemployment coverage, down from 18 months and 24 months of previous coverage. If qualified, U.S. citizens receive 26 weeks.)

Total U.S. employment: 141.475 million

Total U.S. civilian labor force: 149.122 million

Official U.S. unemployment: 5.1% in May 2005; 7.647 million individuals; unemployed 27 weeks and over - 1 in 5 unemployed individuals; 1.5294 million individuals

Total multiple jobholders - percent of total employed: 5.2% in May 2005; over 7 million individuals

Avg. workweek for production or nonsupervisory workers on private nonfarm payrolls: 33.8 hours

Avg. manufacturing workweek: 40.4 hours; avg. overtime at 4.4 hours

Avg. hourly earnings, total private: $16.03 per hour

Avg. weekly earnings, total private: $541.81 per week

Avg. annual earnings, total private: $28,174.12 per year

Table A-12 U-6 Total unemployment/underemployment: 8.6% in May 2005; 12.824 million individuals

Back to Max's missing question: "Why should we care about five percent inflation if we can get more employment?"

Dave: "Max's position seems to be that the FOMC's focus on price stability has driven it to ignore the real side of the economy."

Dave...more: "Mark [Thoma] does point out that price stability and output stability are not, within the boundaries of what monetary policy can achieve, necessarily at odds. But even beyond that, the vision of a Fed that is impervious to real weakness in the economy just doesn't seem justified by the facts."

Movie Guy's response to Max's question: The Fed doesn't focus on unemployment, underemployment, or the vast absence of any health or pension participation benefits for millions of workers unable to capture full qualifying workweek/workhours employment on a weekly or consistent basis. The Fed has a history of being more than willing to sacrifice additional unemployment/underemployment in lieu of "unacceptable" rates of inflation. Period.


1. What element(s) of the U.S. Government should Max turn to in order to seek solutions for his question?

2. Why does the U.S. Government run around waving the Flag at our 5.1% "official" unemployment percentage when it is very clear that it is a considerable distortion of the employment vs. unemployment/underemployment situation in the United States of America? Particularly with regard to health care coverage among lower income workers counted in the work force. And double counting of part-time employment by citizens.

3. Why is the chief spokesman of the Federal Reserve, the Government element responsible for monetary policy, actively and repeatedly promoting the reduction of Government fully or partially funded (including citizen funded flows) social services programs? Why not stick with monetary policy, and leave fiscal policy to the United States Congress and Administration?

4. If the Fed is going to play an activist role in fiscal policy decisions regarding citizen social programs, then perhaps Max has a point. He might as well ask the Fed to explain why it shouldn't undertake an active role in addressing employment support over inflation concerns. But we know full well that that will never happen.

Try the Congress, Max. Good luck, too.

Perhaps the Congress should start examining some of the negative effects of corporate supply side economics driving existing U.S. trade policies.





Posted by: Movie Guy | June 17, 2005 at 01:06 AM

I should have said "sacrifice additional employment" in the above post. Or said "add to unemployment/underemployment"...

Oh, well.

Posted by: Movie Guy | June 17, 2005 at 01:14 AM

Mike --

The MA term was included only to improve the fit and deal with some serial correlation in the residual of this specification. (The regression coefficients are biased without the correction.) It does helps in a statistical sense, but the pictures really look almost the same if I exclude the term.

By real-time data I mean the data that exists at the time a particular federal funds rate choice is made. What you see in the picture is estimated on final data -- the information we have after all revisions have been made. (The exception is the last data points in the picture, which are based on information I had yesterday.)

To give an example, at the time the May 3 decision was made, the last official estimate of 1st quarter GDP growth was 3.1 percent. Since then, it has been revised up to 3.5 percent, and it seems that it will be revised yet again in a couple of weeks, maybe to something near 4 percent. A real-time data set would use the 3.1 percent number.

There is some evidence that using real-time data makes a difference to Taylor Rule specifications - Athanasios Orphanides at the Board of Governors is a leading proponent of that view. It's an issue that deserves to be treated seriously. (FYI, an extensive real-time data set is maintained, and provided to the public, by the Philadelphia Fed.)

Posted by: Dave Altig | June 17, 2005 at 07:50 AM

Spencer -

Thanks for sharing your estimates with me. However, I personally do not like to use the unemployment rate in these calculations. The equilibrium -- or if you like "potential" -- level of unemployment is a tricky animal, and I just don't trust our ability to provide a sensible mapping from that to potential GDP (also tricky enough a concept as it is).

MG --

To Max's question: "Why should we care about five percent inflation if we can get more employment?" My answer would be that you can't get more employment with 5 percent inflation. In fact, over the long haul you would get less.

Here's what else I think -- I am not all convinced that we aren't nearer to the equilibrium unemployment rate than you or Max seem to believe. I have been taken with the number of reports I hear from employers suggesting that skilled labor is relatively scarce, while unskilled labor is plentiful. That suggests to me some structural issues in labor markets that may take some time to work through. And the more I look at it and think about it, the more the late 90s look like the aberration.

Posted by: Dave Altig | June 17, 2005 at 08:09 AM

Three questions:

Based on the most recent data, how big is the output gap?

What is the minimum acceptable or safe rate of unemployment?

What is the maximum acceptable or safe employment to population ratio?


Posted by: Miracle Max | June 17, 2005 at 11:12 AM


My focus goes far beyond basic and perhaps questionable considerations of equilibrium unemployment, the standard Phillips curve, and the non-accelerating-inflation rate of unemployment (NAIRU).

I have focused on the potential and known impacts of existing trade policy on U.S. domestic employment, wages, productivity share relationships, and employee benefit reductions (corporate and Government).

Current OECD thinking appears to endorse the potential for further western world GDP growth via broader free trade. Fine, but let's look at the pattern with domestic U.S. wages, corporate benefit reductions, and Government attempts to back out of social service programs currently incorporated into Law. Not a pretty picture for American blue-collar and service workers below the supervisory level.

Similarly, various U.S. policies embrace the "overwhelming merits" of supply side economics targeted toward producers. Again, fine, but worker share of the effects of such policies continues to decline as a share of growth. Profits gaining substantially, workers compensation lagging at a faster rate. We might call it a conundrum, but we know what is happening.

It offers little comfort to hear that U.S. unemployment is only 5.1% when a broader explanation reveals that worker income gains are not keeping pace on a proportional share basis with corporate profit gains. Obviously, worker gains generally will continue to slide in those industries affected by outsourcing, and foreign provider good and services. I say 'obviously' because it's obvious to me.

If one of the primary functions of Government and the Federal Reserve is to attempt to maintain inflation at low levels, then it's probable that the Government and perhaps the Fed will continue to promote Globalization, higher levels of imports, and corporate outsourcing. But wait..the Fed is calling for a reduction in consumer purchases of import goods and services.

As consumers and workers potentially fall into the very long-term trap of lower standards of living due to an accelerated expansion of trade policy based on corporate supply side economic-based decisionmaking at Government and corporate levels, who in Government is going to step to the microphone later on and explain away U.S. policies that embraced such 'beneficial' concepts which didn't result in net gross and disposable income benefits to American consumers and workers? The same people, including Fed spokesmen, who are saluting such initiatives at the present time?

The issue is net benefits, not illusionary asset valuations based on growing levels of personal debt including housing. Or simply cheaper goods. Net benefits include wage and salary gains commensurate with productivity gains. Well, that's the way it used to be...

Posted by: Movie Guy | June 17, 2005 at 12:15 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad