Sticky Trajectory for Wage Inflation

Simple microeconomics suggests that price falls in response to a surplus.   In labor markets, the price of labor, the wage, should drop in response to a surplus of labor.  

The Market Monetarist explanation of the Great Recession is that nominal GDP fell to a much lower growth path while the price level fell to only a slightly lower growth path.   This resulted in a large down shift in real expenditure on output, to which firms responded by shifting their production to a much lower growth path.   Since this lower growth path of production required less labor, they reduced employment as well.

Market Monetarists don't deny that "supply-side" factors might have reduced potential output a bit, or that workers may have decided they don't want to work quite so much.   But as a matter of fact, we think these were not very significant.   And so, it would seem that on a micro level, the typical market should be in surplus.    

Of course, spending on output began to grow in 2009, as did production and employment.   These factors would tend to reduce the surpluses, but a growing eligible population and rising productivity are tending to increase the surpluses.   In fact, the typical Market Monetarist interpretation is that the growth in production and employment is at best just keeping up with growing capacity.   The underlying surplus remains little changed.

But how is that consistent with basic microeconomics?   If that is true, prices and wages should have both decreased enough so that real expenditure would rise enough to match the productive capacity of the economy.   The surpluses should close through lower prices.  Lower product prices for product markets and lower wages for labor markets.     If nominal GDP is actually kept along the new lower growth path (rather than allowed to drop further,) the decrease in output prices raises real expenditures and so real sales.   The lower wages allow costs to fall enough so that this doesn't generate losses for the firms.     The firms expand production and employment to match the increase in real sales.

Now, let's suppose that firms refuse to actually cut wages.   Wages are sticky, or even rigid, in a downward direction.   Then, the surpluses of labor should result in wages remaining constant.   Prices might fall a bit, but since wages won't fall, firms cut production and employment along with the rising real wages.   Once spending begins to rise, prices should rise along with production and employment.  

Of course, what happened since 2008 is that wages have continued to rise though the rate of increase has decreased.    And while there was one quarter's worth of actual deflation in output prices, generally prices have continued to rise, again, at a slightly lower rate.   Both price and wages are on slightly lower growth paths.  

And so, if the Market Monetarist approach is correct, prices and wages have both continued to increase despite surpluses of output and labor.    Market Monetarists (or most of them) realize that we need to do more than explain that wages won't fall, but rather why the trajectory of wages and prices is sticky.   Why do prices and wages continue to increase in the face of surpluses of output and labor?

In my view, the problem is fundamentally one of "insiders" and "outsiders."    Long run success for a firm depends on having a reputation for being a good employer.   Such a reputation requires that current workers be protected whenever possible.    Managers should be servant leaders, promoting the good of their team.

No existing firm, however, need take responsibility for new entrants into the labor force.  And so, problems develop.

Suppose that population is increasing one percent a year, and the typical career lasts 50 years.   In equilibrium, each year 2 percent of the labor force retires and new hires are 3 percent of the labor force.    Spending on output is growing 5 percent, with employment growing one percent, productivity growing 2 percent, and inflation is 2 percent.      Employers provide compensation plans that provide a cost of living raise and rewards for productivity, so wages increase 4 percent per year.

Spending on output drops to a 15 percent lower growth path.   Employers, seeking to protect their reputations, keep their compensation program unchanged, raising wages 4 percent.   To keep up with rising costs, they continue to raise prices 2 percent.   With productivity rising 2 percent, they can pay 4 percent more and have unit costs rise 2 percent.    

With the significant drop in spending on output and continued price and wage increases, the real volume of sales drop significantly.   The firms don't lay off their employees, since that would hurt their reputation as well.   But they quit hiring new workers.    As workers retire, total employment shrinks by attrition, 2 percent a year.   After 4 years, this would cause employment to fall approximately 8 percent.   The gap between employment and the trend of employment would be approximately 12 percent.

Of course, not all firms can protect all of their employees.    With a severe and rapid drop in sales, some firms may have to restrict or even stop pay increases or layoff employees.   A sharp increase in separations should be expected when spending on output is absolutely dropping.   And perhaps some prices might fall and the rate of increase in prices and wages might slow a bit.   But after biting that bullet, the firms will want to stop laying off employees and return to raising the pay of existing employees according to compensation programs aimed at maintaining their long run reputation as a good employer.

Of course, under normal conditions, many firms fail and lay off employees.   Existing firms that are doing well and new firms expand employment.   For the economy as a whole, if there are few new firms and the relatively successful firms are not doing very well, and so don't expand employment, the drop off in total employment can be very rapid.   Some firms just have a pause in new hiring.   Other firms that would have been successful and hiring, just replace the employees that leave.   And some firms are only replacing some of the employees that leave, and still others are cutting back total employment as fast as possible without actually laying off employees.     With the sort of massive turnover normal in the U.S. economy, rapid decreases in total employment are possible by just limiting new hires.

With the drop off in new hires, there are plenty of workers who would like to have jobs.    Many economists emphasize that it may take some time for these workers to recognize that they cannot find jobs with wages consistent with the previous trend in wages.   Only gradually will they reduce their reservation wages.    But after many years, surely they will accept lower wages.

Of course, labor markets hardly ever involve employees posting offer prices, and then firms choosing to take them.   It is rather that employers offer employment.    And while successful economists may focus on the negotiating process they have experienced, for many new entrants to the labor force, employment is more of a take or leave it "bid" by the employer.     This would surely be true in a buyers' market where total employment is contracting, or at least, growing much more slowly that the number of new entrants into the labor force.

No, what is supposed to happen is that the entrepreneurs in charge of the firms will  change their compensation programs, paying new employees less.   The firms are supposed to recognize that there is a buyers' market and that the unemployed workers have a lower reservation wage.   While this only slightly lowers average wage statistics and average costs, it does reduce marginal costs.   The firms should expand production and employment.

And, sure enough, wages and prices are rising a bit more slowly, and production and employment are growing.    Suppose that spending on output begins growing 5 percent (remaining on a 15 percent lower growth path.)    Firms adjust their compensation packages and start new employees at lower wages, and so inflation is only  1.5 percent.   Real output grows 3.5 percent rather than 3 percent.   How long does it take to recover from a 13 percent shortfall in real GDP?   Decades?

What is wrong?   It is the next step.    Suppose unemployed workers and employees all provided offer prices for their labor, and employers just took it.   The unemployed workers reduce their offer prices, and the employers take them, and some of the existing employees become the unemployed workers unless they also lower their offer prices.    Of course, labor markets work nothing like this.

Now, suppose instead that when employers realize that they can hire new employees at a lower wage, they go to existing employees and explain that they can be replaced at a lower wage.   They also point out to the employees that if they quit, it is difficult to find new jobs, and so they really have no choice but to just take the pay cut.   By taking advantage of their workers, they cut wages and costs rapidly.    Rather than small decreases in marginal costs, there are large decreases in marginal costs.   If firms are competitive, output prices fall.   If nominal GDP is kept on the new (lower) growth path rather than drops further, then real expenditures rise.   Firms expand production and employment.   Both wages and prices fall to the new, much lower growth path that is necessary for real expenditure to rise back to potential output.

So, why don't firms "take advantage" of their existing employees?   Why don't they force pay cuts on them?   There is a surplus of labor.    Their workers have no choice but t0 accept the pay cuts.   Why?  It is because they don't want to have a reputation of taking advantage of their employees.    

Now, if this situation was expected to be permanent, then perhaps the absolute pay cuts would occur.   But even in this scenario, where it is only new hires that take the pay cuts, a firm can maintain its reputation for protecting all existing employees and eventually the economy returns to full employment.   The situation is not permanent.

Even more so, if there is an expectation that spending on output will grow more quickly, generating a recovery more quickly, then any firm that took advantage of existing employees would have a difficult time recruiting good workers, or would have to pay workers more during "normal" times, and suffer a competitive disadvantage.  

More interestingly, if enough firms cut wages and lowered prices, then the other firms could explain to their employees that they have no choice but to do the same.   If all firms cut wages and prices when nominal GDP was low, then no firm suffers a competitive disadvantage during normal times.   (Perhaps all employers would be considered greedy and wicked.)   But if a few firms follow the policy of taking advantage of their employees, they might have fleetingly lower costs and increased profits during the recession, but they would suffer during "normal" times.

For Market Monetarists, the answer is easy.   Don't allow nominal GDP to fall to a lower growth path, and if it does, get it back up to the target growth path as soon as possible.  Allow firms to develop their compensation programs based upon an expectation that spending on output will grow at a slow, steady rate and then fulfill the expectation.

Output and Employment? What Puzzle?

Nominal GDP passed it previous 2008 peak long ago.   It is nearly 8 percent above that peak.   Real GDP has also passed its 2008 peak, having risen nearly 2 percent!   On the other hand, employment has not reached its previous peak.   It is still 2.76 percent below that peak.

The "problem" is that productivity has increased.   The recovery in output has been generated without there being a matching increase in employment.

Right?

During the Great Moderation, real GDP increased by 110 percent and employment only increased 37 percent.      Thank goodness that real GDP rose more than employment, resulting in real GDP per worker rising 53 percent.

Real GDP is 13 percent below trend and employment is 9 percent below trend.   For both to return to trend, real GDP would need to grow more than employment.   In fact, real GDP would need to grow nearly 18 percent and employment about 11 percent.

NGDP and Wages

Nominal GDP divided by the wage rate (multiplied by 2000 to get an annual wage,) is approximately 13 percent below the trend of the Great Moderation.   (Since the GDP deflator and hourly production and nonsupervisory wages are both between 2 and 3 percent below trend, it should be no surprise that this ratio is below trend almost exactly how much real GDP is below trend.)

If nominal wages had remained at $18 as they were when nominal GDP peaked in 2008, the ratio would have increased by 8 percent.   Which is, of course, exactly how much nominal GDP has risen over its previous peak.   The ratio of nominal GDP to wages would still be about 5 percent below its trend growth path.

Market Monetarist Diagrams

The traditional Market Monetarist three:

Nominal GDP:

Real GDP:

Employment:

Keep on telling yourself that this just means that people decided they didn't feel like working as much starting in 2008, so production fell, and so did spending on output.

And what about prices and wages?

and wages:

And now for some new diagrams.   Nominal GDP divided by wages (hourly x 2000)

And since there is this odd notion that a sudden increase in productivity has caused lower employment, here is real GDP per worker:

This last diagram looks more consistent with the view that potential output has fallen a bit  rather than the notion that we are so much more productive that we don't need as many workers.   (Which I think is pretty much inconsistent with fundamental economic principles like scarcity and opportunity cost and instead entirely consistent with naive noneconomists views about the fundamental scarcity of jobs.)

Manufacturing NGDP?

Tyler Cowen argued :

My worry is that some Market Monetarists speak of ngdp as if it is some block of stuff, handed down from on high (of course in the past our central banks have not been targeting ngdp). It’s as if ngdp determines the size of the room, and a carpenter is then asked to build a house within that room. If the room is too small, a large house cannot be built. Or, if you are not given enough clay, you cannot build a very large sculpture. Along these lines, if the growth path of ngdp is not robust enough, the economy cannot do well.
I get nervous at how ngdp lumps together real and nominal in one variable, and I get nervous at how the passive voice is applied to ngdp.

Interesting, but not quite correct.   Market Monetarists recognize that a "bigger house" can be built in the room.  But rather it requires that it have a lower price.   As for the passive voice, the Market Monetarist approach is that people sometimes choose to increase their money holdings beyond the the amount by which the Fed and the private sector choose to increase the quantity of money.  Some of the people can and do increase their money holdings (as David Beckworth is fond of showing with charts showing a large increase in actual holdings of safe assets,) but it had the side effect of reducing spending on output.   Other people, whose incomes are lower because of this (and really, for many, this is relative to what they would have earned, rather than what they were earning while in high school or college four years ago,) are holding less money than they would.

What is passive about that?   People choose to accumulate money holdings and continue to hold high money balances.   And this has consequences for the stream of spending on output.

Cowen continues:

My framing is different. My framing is that the private sector can manufacture its own ngdp. It can do so by trade and it can do so by credit and of course velocity is endogenous to the available gains from trade. Most of the major central banks are, today, not obsessed with snuffing out recovery and increases in real output.
To say “ngdp is low,” or “ngdp is on a low growth path,” or “ngdp is below trend,” and so on — be very careful! Those claims do not necessarily have causal force. Arguably they are simply repeating, in a new and somewhat different language, the point that the private sector has not seen fit to engage in more trade, credit creation, velocity acceleration, and so on. Formally speaking, the claims are not wrong, but I don’t find them useful as an explanation for why economic growth or recovery, at some point in time, is slow. It is one way of repeating or re-expressing the slowness of economic growth, albeit with some transforms applied to the vocabulary of variables.

At first pass, from a Market Monetarist perspective, the way the private sector can increase nominal GDP--spending more on output--is by reducing the demand to hold money.   Hold less money and spend it.    Cowen's argument appears to be that people want to hold money rather than spend it.   Low nominal GDP is just another way of saying that people don't have anything they want to spend money on.     If they had something worth spending their money on, they would do it, and nominal GDP would rise.

From the Market Monetarist perspective, most of those who want to spend are those who are currently unemployed.   They would like to contribute to production, earn income, and then use that income to purchase the products they are going to help produce.    The problem would be that if all of those people did work, earn, and buy products, they would also want to hold more money.   This would increase the demand to hold money beyond the amount that exists.

Note, however, that implicit in this Market Monetarist argument is the notion that the demand to hold money is positively related to real income.    If the unemployed people start working, they will produce more, earn more, and buy more output, but they will hold more money.

The other side of that coin is that when real income falls, the demand to hold money falls as well.    This generates what I call the "Yeager effect."    If we imagine that people just decide to transact less, and so produce less output, then this reduces the demand to hold money.   If the quantity of money doesn't fall, the excess money is spent, raising demand for output back up its initial value.   If people really want to produce less, the result in inflationary.   At the higher price level, the real quantity of money falls, to match the lower demand.

Yeager used this argument to explain why real coordination failures don't result in lower real output and employment.   If we imagine that people really want to work and consume, but somehow, a coordination failure keeps them from doing that, then real output falls, the real demand for money falls, and given the quantity of money, an excess supply of money is generated, which raises the demand for output and labor.  In other words, versions of the "multiplier" where lower income generates lower consumption and so lower income, have no force.

Of course, if the demand to hold money rises or the quantity of money falls, then spending on output falls, and unless prices and wages fall enough so that real expenditures are maintained, then output and employment fall.   In particular, if people don't want to spend, and instead choose to accumulate larger money balances, that is an increase in the demand to hold money.   Unless the quantity of money rises to match that increase in demand, then spending on output falls.

But suppose that the demand for money is not related to real income.   If people choose to transact less, produce less, work less, then they earn less and spend less.   There is less "nominal GDP."     While real output is lower, this does not reduce the real demand to hold money.    There is no excess supply of money, and so no tendency for demand for output to be maintained.   In this particular scenario, this seems quite desirable.   And, if people chose to work more, produce more, and spend more, then spending on output would rise.   The higher real output would not raise the demand to hold money, and so, even without an increase in the quantity of money, spending on output and real output would rise.

However, we don't live in such a world.   All evidence suggests that the demand to hold money is positively related to real income--maybe not quite proportional--but close.

And more importantly, when people choose to hold money rather than spend, even if they have nothing they want to buy, this isn't the same thing as choosing not to produce and sell.    Nominal GDP level targeting works very well to solve the problem of  people choosing to work and not spend.   It solves the problem of an excess demand for money.

Suppose some individual is willing to work, produce and sell, not because they have some product they want to buy now, but rather because some good or service might appear in the future that they will want to buy.  This is called saving.    Usually, this is beneficial to others.   There are people who are willing to pay in order to dissave, as well as people who want to invest--use resources to be able to produce consumer goods in the future.    But if saving supply increases enough, the price that coordinates saving and investment could turn negative.   That means that producing now and waiting to consume in the future is not beneficial to the rest of society.    People might want to save, but they would have to compensate others to use their products now in exchange for being willing to provide consumer goods in the future that the savers just might want to buy.

But with "money" having a zero nominal interest rate, people can always save by accumulating money.   The nominal interest rate can be no lower than the cost of storing currency.     When nominal interest rates are very low, issuing currency is not profitable, and so excess demands for money can develop.   But now, people working, producing, but not spending not only are not creating a benefit for others on the margin, they are causing harm.   Others sell less, produce less, earn less, and so, are willing to hold less money, freeing up money balances for those who want to work and earn now and only maybe consume in the future.

Nominal GDP targeting solves this problem with a monetary regime that creates enough money to accommodate the added demand to hold money so that spending on output continues to grow at a slow steady pace.     (Well, it could also work to reduce the demand to hold money by generating a lower nominal and real yield on money balances.)

Generally, this implies that people who want to continue working and producing and have nothing they want to buy must be matched by firms and households that want to spend on output now.   However, the lower interest rates, perhaps even to negative levels, might convince people who can think of no output they want to buy to choose to work less.

And that is a disadvantage of nominal GDP targeting.   Not that negative real or nominal interest rates might cause people to work less.   It is rather that if people choose to work less, produce less, and spend less for any reason, then nominal GDP targeting causes undesirable disruption.    Like a fixed quantity of money or a gold standard, it results in inflation--both higher nominal incomes for those who continue to work as well as higher prices of products.  

Imagine we lived in a world where most people are self-sufficient.    They produce goods and then consume the goods themselves.    They produce all of their routine consumer goods and services themselves.   But from time to time, they make exchanges of luxuries.   A painting is exchanged for a song.   A short story is exchanged for a sculpture.  

Suppose we use a "song" standard, and calculate the "song" value of these barter exchanges.   Nominal GDP can be calculated as the sum of the "song" values of each barter of artwork.   Why forces these exchanges into a "room?"   Why not just let exchanges happen when people want to make them?   Why not let "nominal GDP" fluctuate according to desired exchanges?

But we don't live in that world.   We live in a market economic system where most people mostly sell goods for money and then use the money to buy the goods and services they need to live--to pay the bills.   Nominal GDP targeting is not perfection.   At least, I don't think of it that way.    It is the least bad monetary regime for the market economic systems that really exist.

Eli Dourado Replies to his Critics

Eli Dourado wrote some more about QE3.

The problem is that even if this story is true, we are probably, again, out of the short run. NGDP is almost 10 percent higher now than it was at the pre-crash peak. The number of people employed, even with population growth, is still below the pre-crash peak. Even assuming that insider nominal wages are totally inflexible, nominal output per worker has grown fast enough that insider real wages have probably adjusted. Furthermore, in five years, a non-trivial fraction of insiders retire or change jobs.

This inspired me to review the numbers again--calculate how far nominal GDP, real GDP, the price level (GDP deflator) and wages (production and nonsupervisory hourly,) are from the trends of the Great Moderation.    I also took a look at the changes from the previous peak.

For example, nominal GDP is $15.6 trillion.    It peaked in the second quarter of 2008 at $14.4 trillion.   Nominal GDP is 8.14 percent above its previous peak.  (I guess that is close to 10 percent, and nominal GDP is 9.19 percent greater than it was in the fourth quarter of 2007, when employment peaked and the recession officially began.)

Of course, most Market Monetarists don't focus much on where nominal GDP is now compared to the peak.   It is currently 15.2 percent below the growth path of the Great Moderation.   It would have to grow approximately 23 percent to get back to that the old trend in one year.

What about the price level?   Using the GDP deflator, the price level is now 115.   When nominal GDP peaked in 2008, it was 108.   The price level has increased 6.5 percent.    The price level has increased less than nominal GDP.  

The price level  is currently 2.4 percent below its trend growth path from the Great Moderation and would need to grow 4.27 percent to return to that trend in one year.

Real GDP is $13,546 billion.   When nominal GDP peaked in 2008, real GDP was $13,300 billion.   Real GDP has increased by 1.76 percent over the last 4 years.  Not much.

Suppose the price level is sticky, even stuck, downward, but had stopped rising when it was 108 back in 2008.  The current nominal GDP would be consistent with real expenditures and real GDP of  $14,400 billion.   If real GDP had increased to that amount, it would have risen approximately 8 percent.   That seems like quite a bit,  but it would still leave it well below trend.

Real GDP is currently 13 percent below the trend of the Great Moderation.   If it had increased 8 percent over the last 4 years, it would still be 7.61 percent below trend.   Absolutely no price increases would have fixed only slightly half of the problem after four years.   For real GDP to return to its trend for the Great Moderation, it would need to grow 17.75 percent.

What about wages?    When nominal GDP peaked in 2008, the hourly production and nonsupervisory wage was $18.   It is now $19.72.   It has increased 9.5 percent--even more than prices.   However, wages are now 2.5 percent below the trend of the Great Moderation, about the same as the GDP price deflator.   For wages to return to the trend over the next year, the increase would need to be a little more than 5 percent.

Employment?  When nominal GDP peaked in 2008, employment was 146 million.   It is now 142 million.   It  is 2.5 percent lower.  

That is pretty bad, of course,   But for most Market Monetarists, the problem isn't that it hasn't reached its previous peak (which was six months before nominal GDP peaked, though not significantly higher,) the problem is the shortfall from trend.  Employment is 9.25 percent below the growth path of the Great Moderation.   It would need to grow more than 11 percent to return to trend in a year.

Most Market Monetarists are struck by these figures--nominal GDP is 15 percent below trend, while prices and wages are about 2.5 percent below trend.   Real GDP is 13 percent below trend and employment is 9.5 percent below trend.

Given these large numbers, is comparing the current level of nominal GDP to its previous peak, and similarly for real GDP and employment, noting that real GDP has passed its previous peak by less than 2 percent while employment is 2.5 percent short--that important?  These are much smaller numbers.  Sure, it is a puzzle, but 2 and 2.5 is nothing close to 15, 13 and 9.

I calculated the ratio of nominal GDP to nominal wages.   I multiplied the hourly wage by 2000 to get an annual full time equivalent.    This ratio is 1.31 percent lower than it was when nominal GDP peaked in 2008.  The ratio is about 13 percent below trend.  

I also looked at the ratio of real GDP to employment.    That measure of productivity has increased by 4.4 percent since nominal GDP peaked in 2008.   Interestingly, it is  4.2 percent below its growth path from the Great Moderation.    To return to trend, it would need to grow 5.8 percent over the next year.

Like other Market Monetarists, I admit that it possible that potential output has fallen to a lower growth path and is now 13 percent below the trend of the Great Moderation.   The 2.5 decrease in prices and wages could have been sufficient to limit the decrease in real expenditures to the reduction in potential output.

It could be.   But I doubt it.

Sumnerian Secession!

Monetary policy in the NGDP Empire is dominated by the Island of Speculativa.   They have decided that the price of gold should grow 2 percent a year.   This results in the index number rising 2 percent per year.   They have an inflation target!

Meanwhile, over in Autarkia, the farmers are generally increasing their productivity.   Corn production is rising.   But when the weather is bad, corn production can grow more slowly or even fall.   Further, the substitution effect is greater than the income effect, so when the weather forecast is bad, the farmers plant less corn.   Employment is lower and so the production of corn is extra low when the forecast is right and weather is bad as was expected.

With the price of gold and so the price index going up 2 percent a year, when employment is low and the production of corn is extra low, nominal GDP grows more slowly and sometimes shrinks.   It almost appears as if slow growth of nominal GDP or decreases in nominal GDP are causing reduced production and employment.

Of course, nominal GDP is just the product of two unrelated things, and so, causes nothing.   Weather and expectations of the weather are causing decreases in employment and real output.

But suppose the Sumneran Party becomes influential in Autarkia and proposes that the quantity of money be expanded so that the price of gold rises more than the magic 2 percent.   The idea is that if index number rises faster, then nominal GDP rises faster.   They have a confused notion that the more rapid increase in nominal GDP will convince the farmers to plant more corn (despite the expectations of bad weather) or even more crazily, assume that it will change the weather.

Well, they can't convince the monetary authority over in Speculativa to create money faster to get gold prices rising faster.   No, they leave the union and become the independent Kingdom of Autarkia.     They issue their own fiat currency.   They no longer worry about the price of gold.  No, they take the quantity of money and divide it by the quantity of corn produced.   They call the quotient, the "price" of corn.   Following the long settled practice in the other island, they calculate an index number based upon the first price calculated.   Then they multiply the quantity of corn by the index, and call it Nominal GDP for Autarkia.   (If you skip the index number step, nominal GDP is just the quantity of money.  And even with the step, it is just an odd way of describing the same thing.)

Well, next time the farmers expect bad weather, the farmers plant less corn and when the weather is really bad, corn production is very low.   The Sumnerans do raise the quantity of money so nominal GDP growth is stabilized.   But the reduction in employment and production is the same as it ever was.

Because, of course, nominal GDP is just the product of real output and a price index.   Right?

P.S.   Maybe in Autarka, but not in the real world where people sell goods for money and buy goods with money.