Archive for February 2012


The Bank of England’s inflation report

Commenter 123 sent me the February BOE inflation report, with this very interesting graph:



















I suspect the NGDP instability contributed to fluctuations in RGDP growth, but it also looks like Britain has serious supply-side problems. Remember that the BOE inflation target is 2%.

I have a question for my Keynesian readers.  Again and again the report refers to changes in real output as changes in “demand.”  I’m pretty sure that’s Keynesian terminology, but it makes absolutely no sense to me.  Here’s an example:

The pace of four-quarter global demand growth slowed in 2011. But, within that, there was some divergence in growth rates across different regions. And business surveys suggest that activity may have picked up a little in some countries at the beginning of this year (Chart 2.8). Against the backdrop of subdued global demand growth, world trade rose only modestly in the year to November 2011.

Chart 2.8 shows output trends, not AD.  They also refer to real domestic final sales (C+I+G) as “domestic demand.”  I find this terminology bizarre, but maybe I’m missing something.  For instance, suppose the AD curve was stable, and the AS curve shifted to the right, producing real economic growth.  The BOE would apparently call that real growth an increase in “demand” even though it was caused by an increase in supply.  It seems to me that causes needless confusion, and is likely to lead one to overemphasize the importance of demand shocks in driving RGDP growth.

I prefer to call NGDP “demand.”  In that case a rise in AS doesn’t have any impact on demand, but rather increases both supply and real GDP.

PS. There is a lot of confusion about the impact of supply-side problems on economic growth.  All countries including Switzerland and Congo grow at the same rate in the long run.  It’s all about levels.  Efficient countries have a per capita GDP closer to the richest (non-oil) country than less efficient countries.  Britain recently raised the size of its government as a share of GDP from 37% to 50%.  Some of that was cyclical, but not all.  This will slightly reduce Britain’s steady state GDP per capita relative to the richest country, but will not permanently reduce it’s growth rate.  It may currently be in transition, say from 70% of Singapore’s GDP/person to 63%.  If so, it will grow a total of 10% less than normal during that period of transition, which might last a decade.  Obviously I just pulled these numbers out of a hat, to illustrate the proper way to think about supply-side issues.

PPS.  The Economist estimates that in 2012 Britain’s GDP/person will be 69.5% of Singapore’s.  The US will be at 94.5%, not bad for a country with brain-dead policymakers.

All I can do is shake my head

Jared Pincin sent me this gem from Reuters.  I don’t know the author, but it really doesn’t matter as he is probably just expressing the “conventional wisdom.”

Seeking to make good on past threats in Congress to rein in the Federal Reserve’s powers, a prominent Republican lawmaker said on Thursday he will introduce legislation to focus the U.S. central bank on a single mandate to fight inflation and protect the dollar’s value.

.  .  .

The Fed’s aggressive actions to fight stubbornly high unemployment rates in recent years have been seen by some as straying into fiscal policy – traditionally the responsibility of Congress – and Brady’s bill aims to clear up any confusion by stripping the Fed of its jobs mandate.

That’s right folks, fiscal stimulus creates jobs and monetary stimulus creates inflation.  Milton Friedman must be rolling over in his grave.  I think you’d have to go back to the price controls of the 1970s to find a time when there was such mass stupidity among the chattering class.  Make no mistake about it, stupidity was the cause of the 2008-09 NGDP crash.  Even if the policymakers themselves weren’t stupid, they had to contend with a zeitgeist that was totally clueless about AS/AD.  Even Ben Bernanke can only do so much.

Bernanke favors NGDP targeting, he just doesn’t realize it

Here’s the conventional wisdom from Renee Haltom at the Richmond Fed:

Changing the Rules

The Fed’s policymaking committee discussed NGDP targeting at its November 2011 meeting and provided a hint that major changes are not on the table. “We are not contemplating at this time any radical change in framework,” Chairman Bernanke said after the meeting. “We are going to stay within the dual mandate approach that we’ve been using until this point.”

Central bankers don’t take changes to the conduct of policy lightly. All central banks face the temptation to boost growth for temporary gain at the expense of longer-run price stability. To convince the public that monetary policy won’t give in to that temptation “” to therefore maintain credibility and keep inflation anchored “” many central banks stick to consistent “rules,” either explicit or implicit, to effectively tie their own hands.

Radical change?  Bernanke has recently emphasized that the Fed takes its dual mandate seriously.  He indicated that they put equal weight on both sides of the mandate.  I recall he responded to a conservative senator with something to the effect that if Congress wasn’t happy with their policy they should give them a different mandate.  And of course the Fed influences employment by affecting RGDP growth.  I wouldn’t say the Fed puts equal weight on inflation and RGDP growth, but Bernanke’s comments about the dual mandate suggest their policy is pretty close to NGDP targeting.

And yet Haltom is right that a switch to NGDP targeting is generally viewed as a major change in Fed policy.  That got me wondering if it would be possible to do a policy that looked very much like current policy, and yet was almost identical to NGDP targeting.  Given Bernanke’s insistence that both sides of the mandate (inflation and employment) are treated equally by the Fed, it ought not be too difficult.  I’d like to hear your suggestions, but here’s one idea.    The Fed announces that it will fulfill its dual mandate as follows.

1.  It will aim for 2% core PCE inflation every single year.

2.  It will aim for RGDP growth equal to the average RGDP growth rate of the previous 20 years.

This policy will tend to produce roughly 2% headline inflation over any long period of time.  Importantly, it will aim for 2% core inflation even in the short run.  The RGDP target is aimed at the Fed’s employment target.  High employment is more likely to occur in an environment where RGDP growth and inflation are relatively stable.  Indeed my preference would be for the Fed to “target the forecast”, that is, set policy such that it is expected to hit this target.

Now let me anticipate objections:

1.  No, this isn’t trying to hit two targets at once, it is a single target; NGDP growth equal to the 20 year moving average rate of RGDP growth plus 2% inflation.  That’s a single number, a single target.

2.  It answers one popular objection to NGDP targeting, the fear that the price level would no longer be anchored.  This still anchors inflation at 2% over the long run.   The math is kind of complicated, but the average inflation rate would approach 2% as the time period approached infinity.  Thus if trend growth fell from 3% to 2%, as may be occurring right now, there would be a transition period when inflation would temporarily exceed the 2% goal.  But then it would asymptotically fall back to 2%.  And the reverse would occur when the trend rate of RGDP growth rose.  When trend growth is stable then inflation should be stable at 2%.

In contrast, under NGDP targeting a fall in trend RGDP growth from 3% to 2% would leave inflation permanently higher.

The more I think about this idea the more I like it.  The 2% inflation target is “locked in” to this system, not left to chance.  It’s an explicit 2% inflation target.  The system gives equal weight to inflation and employment fluctuations.  The Fed is trying to keep inflation at 2%, and trying to stabilize employment.  This gives it a definite target, no more of the ambiguity that drives both liberals and conservatives nuts when they are trying to evaluate Fed policy.  We can finally hold them accountable.  We can say; “We’ve instructed them to do X, now let’s see how close they came to doing X.”  As it is, it’s hard to hold them accountable, because their interpretation of the mandate is so vague.

Consider Haltom’s second paragraph quoted above.  I think she expresses the conventional wisdom, but I’m not sure that’s true.  Ever since Lehman failed inflation has averaged just over 1%.  This is quite odd, given the Fed’s dual mandate.  If you think about it, even 2% inflation would have violated the dual mandate, as it would have implicitly given zero weight to jobs.  If inflation had been 2% over the past three years, the Fed should have been trying to raise inflation above 2% to reduce unemployment.  Does anyone serious believe they would have, given that even at slightly over 1% inflation it was really hard to get them to explicitly try to raise inflation?

And how about the BOJ, or the ECB under Trichet?  Do they seem like central banks just chomping at the bit to print money, but help back by inflation targets?  Japan’s had mild deflation for nearly 2 decades.  And Trichet bragged that he drove inflation to a level far below that of the Bundesbank in the midst of the greatest debt crisis in European history.

As I look at world history, it looks to me like the major central banks were constantly trying to boost growth until about the early 1980s, and since then have constantly been obsessed with reducing inflation.

I don’t know why these multi-decade long mood swings affect our central banks, but it needs to stop.  NGDP targeting would be ideal, producing a perfectly calm ocean surface over which the ship of free market capitalism could sail.  But my “20 year average of RGDP growth plus 2% inflation” idea is almost as good.  It would produce very long and graceful swells in the economy’s surface, barely detectable to most people.  Even a tsunami is hardly noticeable in the middle of the ocean.  A 20 year moving RGDP average plus 2% inflation isn’t much different from a fixed NGDP target, but looks much more like a dual mandate featuring 2% inflation targeting, and it also anchors inflation in the long run.  What’s not to like Mr. Bernanke?

PS.  I’ve ignored the level targeting vs. growth rate targeting issue, as I think we first need to get a consensus on what type of target.  This proposal could be amended to go either way, although I obviously prefer level targeting.

Why sticky wages matter

Based on reaction to my previous sticky-wage post, it might be useful to consider how sticky wages operate at the industry level.  For simplicity, assume a unit elastic demand for cars (nothing crucial will hinge on that assumption—it makes the math easier.)

Now suppose that Americans decide they want to spend 10% more on cars in 2012 than 2011.  Because demand is unit elastic, the demand curve will shift right (and upward) by 10%.  How will automakers react to this increase in demand?  You can construct plausible scenarios where there is no impact on unit sales.  Suppose we were at full employment, and both wages and prices had been trending upward at a 10% rate (like 1979.)  In that case car companies might react to the extra nominal demand by merely selling the same number of cars at a 10% higher price.

But suppose the situation is more like America circa 2012.  Lots of slack in the economy.  Wage contracts reflecting very low expected NGDP growth.  Nominal wages are highly sticky.  In my view under current conditions the automakers would respond to 10% more demand for cars by selling lots more (say 7% or 8% more) and raising prices only slightly (say 2% or 3%.)  And I think sticky wages are one reason why this sort of “nominal spending shock” in the auto industry would boost output, although not the only reason. 

The evidence that nominal wages are sticky in the short run is so overwhelming that I’d like to take it as a given.  Instead, let’s consider various possibilities for how car prices would react to more demand:

1.  Suppose the auto industry is perfectly competitive and sticky wages are a big part of costs (including wages for the making of components.)  With sticky wages the MC curve will shift upward by less than the demand curve, and equilibrium output will rise.  Prices will rise somewhat, and the pseudo “real wage,” i.e. wages/car prices, will fall.  This is the classic case of wage stickiness leading to counter-cyclical real wages, as companies move along a stable labor demand curve.

2.  Suppose the auto industry is monopolistically competitive.  In that case prices might be very sticky.  It’s possible that both wages and prices don’t change, or both might rise by something like 2%.  Again a 10% nominal spending shock will cause much higher real output in the car industry, but in this case the “real wage” in terms of auto prices will be roughly unchanged, and this will seem to refute the sticky-wage model of the business cycle.  In fact, all it will really show is that the perfect competition model doesn’t fit the auto industry.  Back in 1921 when the US economy was somewhat closer to perfect competition, real wages rose sharply when NGDP fell.

The bottom line is that the sticky wage model that relies on W/NGDP is almost always a good explanation of fluctuations in employment.  Because nominal wages per hour are sticky, when nominal spending falls sharply we can expect hours worked to fall.  This is one of the most reliable regularities in macroeconomics.  But we can’t go much beyond that.  Theories that try to explain fluctuations in terms of real wages (W/P) will work well in some settings, and not others.  It might work in perfect competition, but not monopolistic competition.  It will depend on whether prices are also sticky, and on how you measure things like housing prices.  The official CPI numbers show housing prices up 7.5% in 5 years, whereas Case-Shiller has them down 32% over 5 years.  Housing is a third of the CPI, so it makes a big difference which housing numbers you use.  On the other hand NGDP is more cut and dried—just the dollar value of expenditure on final goods–no assumptions about prices are necessary.

Nominal GDP shocks aren’t the only factor that explains business cycles, but I believe they explain the majority of the cyclical fluctuations.  Because of sticky wages, a stable path for NGDP would make RGDP growth more stable, and would also make financial crises less severe.

Sticky wages and NGDP shocks

Arnold Kling points to labor cost data that he sees as being inconsistent with sticky wage models of the recession.  I have a different view of macro than most economists, for instance I’ve often argued that inflation data is basically meaningless, and should be removed from all macro analysis (except perhaps “dorm room” discussions of long run changes in living standards.)  Instead NGDP is the nominal variable of interest, and hours worked is the real variable that best picks up the business cycle.  Kling’s post looks at the ratio of prices to unit labor costs, a variable that plays no role in my business cycle analysis.

In my view the ratio of wages to NGDP per capita (NGDPPC) is the best way of picking up nominal shocks that might throw the labor market out of equilibrium.  Normally these two variables will grow at about the same rate, although there can be gradual changes in the ratio of W/NGDPPC if the share of income going to capital increases, or if the share of total compensation going to non-wage benefits increases.

The St. Louis Fred has wage data only for the period since 2006.  Here’s the change in NGDP, NGDPPC, nominal hourly wages, and W/NGDPPC over the past 5 years:

Time period:        NGDP growth   NGDPPC growth  Wage growth  W/NGDPPC growth

2006:2-2007:2        4.85%                 3.95%                  3.72%               -0.23%

2007:2-2008:2        3.14%                 2.24%                  2.92%                0.68%

2008:2-2009:2       -3.90%               -4.80%                  2.93%                 7.73%

2009:2-2010:2        4.43%                3.53%                  1.94%                -1.61%

2010:2-2011:2       3.77%                2.87%                   1.99%                -0.68%

This is a very limited sample, but I bet if you went back to 1990 you’d see a similar pattern.  During normal times, nominal wages grow at roughly the same rate as NGDPPC, maybe a bit less for the reasons mentioned above.  Then from mid-2008 to mid-2009 wages soared relative to NGDPPC.  Why did this occur?  In my view it happened because the Fed let NGDPPC fall 9% below trend.  Nominal wage growth is quite sticky in the short run, so a sudden large change in NGDPPC will usually change the ratio of wages to NGDPPPC, which can be seen as the funds available to pay salaries.  If those funds drop sharply, and the hourly wage is stable, the number of hours worked will fall sharply.  If the Fed had engineered 3% or 4% NGDPPC growth in 2008-09, nominal wages would still have risen by about 3%, and there would have been far less unemployment.

Since 2009 wages have risen more slowly that NGDPPC, but at current rates it will take years to restore equilibrium in the labor market.  One might ask why nominal wages don’t fall quickly to restore equilibrium.  Paul Krugman presented data a few months back showing workers rarely receive nominal pay cuts.  So in “hard times” you have some workers getting zero raises, and others (in healthy parts of the economy) getting 3% or 4% raises.  The average wage increase falls to slightly under 2%.

The last 5 years fit this version of the sticky wage model almost perfectly.  If there are other versions of the sticky wage model (perhaps using W/P) that don’t fit the data very well, then perhaps we should consider discarding those non-useful models.

PS.  Readers may notice that I estimated NGDPPC simply by subtracting recent US population growth (0.9%) from the NGDP growth numbers.  Normally I don’t even talk about NGDPPC, as NGDP is close enough.  But for this exercise I thought the per capita numbers would make it easier to see the intuition.  NGDP shocks are like a game of musical chairs.  Remove 9% of NGDP relative to trend, and you’ll have 9,000,000 unemployed workers sitting on the floor.  It’s that simple.