This archived content originally appeared at, the predecessor site to, and does not carry the sponsorship of the Cato Institute.

…And my Own Attempt at an Answer

Earlier today I posed to Market Monetarists the following questions: "If a substantial share of today's high unemployment really is due to a lack of spending, what sort of wage-expectations pattern is informing this outcome?…Can it really be the case that NGDP (and equilibrium wage rate) expectations continue to race ahead of reality, even when that reality involves what would normally be considered a perfectly respectable, if not excessive, growth rate of overall spending? How can this be?"

When I posed these questions I had no answer in mind; on the contrary I doubted that they could be answered without appealing to some extremely odd labor market behavior.

Having done a bit more thinking since, I now believe I have a better grasp of why the combination of an NGDP growth revival and more modest wage inflation hasn't sufficed to eliminate cyclical (demand-shortage based) unemployment. My further reflections make me more inclined to see merit in Market Monetarists' arguments for more accommodative monetary policy. But they also leave me as puzzled as before regarding the expectations-formation processes driving observed patterns of wage-rate adjustment. More specifically, they bring to light a degree of wage-inflation inertia that seems, on its face, difficult to square with the usual assumption that market participants behave rationally.

Previously I observed that it has been about two years since NGDP recovered its pre-Lehman's level, and that it has been growing at between 4 and 5 percent ever since. I also noted that the rate of increase in hourly wages has fallen persistently since the crash, and is now half what it was before then. So, why haven't these facts together added up to the elimination of cyclical unemployment?

My partial answer to those questions is one best captured by the following graph, showing index values (with 1/1/2003=100) for hourly wages and NGDP since 2005:

Here one can clearly see how, while NGDP plummeted, hourly wages kept right on increasing, albeit at an ever declining rate. Allowing for compounding, this difference sufficed to create a gap between wage and NGDP levels far exceeding its pre-bust counterpart, and large enough to have been only slightly reduced by subsequent, reasonably robust NGDP growth, notwithstanding the slowed growth of wages.

The puzzle is, of course, why wages have kept on rising at all, despite high unemployment. Had they stopped increasing altogether at the onset of the NGDP crunch, wages and total spending might have recovered their old relative positions about two years ago. That, presumably, would have been too much to hope for. But if it is unreasonable to expect wage inflation to stop on a dime, is it not equally perplexing that it should lunge ahead like an ocean liner might, despite having its engines put to a full stop?

ADDENDUM (July 9, 2012): It turns out that the graph I concocted appears after all to have misled me (I confess I was apparently to willing to be convinced that there might be a case for further NGDP stimulus after all); playing around with it further (by using the same scale for both plots and letting 1/1/2005=100), I come up with another that actually reinforces the position I took in my original post:

Tom Dougherty, below and at TheMoneyIllusion, gets a similar result using a newer and more comprehensive hourly compensation index.

In light of these further findings, I'm back to my original question: has the Fed, despite the still high level of unemployment, already done all that it ought to do in the way of monetary easing despite the still high level of unemployment?

  • Rob R.

    Very perplexing.

    Assuming that it is a given that the demand for labor shifted to the left since 2008 , then the supply of labor curve must be almost vertical and actually have itself shifted upwards in expectation of higher wage since then. That's the only way I can think of to explain this chart.

    Based on this I guess we have to hope that MMist are right and that if we increase NGDP back to trend that the wage index will continue on its trend (and not move upwards with NGDP) and the employment gap may be closed.

    But why is the supply curve for labor vertical ?

    • George Selgin

      The importance of "pinning down" the story of equilibrium wage expectations-formation rests precisely on the fact that, unless we do so, we cannot be certain as to how a change in NGDP will influence those expectations. Hence we also cannot be certain that the change in NGDP will have the desired effect on U.

  • Lorenzo

    I suggest that disaggregation by industry/sector would be informative. Private sector earnings, for example, seem to be roughly tracking CPI.

    And shifts in ratings of hiring are the main way employment moves around across a business cycle.

    So folk are not leaving and employers are paying to keep the people they have, while the insider/outsider split is increasing, as it does when you get prolonged unemployment.

    And then there's the public sector …

  • Lee Kelly

    The inertia is perplexing. Sticky wage-expectations appear to frustrating a return to monetary equilibrium. Looking at the chart, the effect seems much larger than I had expected. Did the financial crisis just expedite the U.S. economy's shift to a new, higher, natural rate of employment?

  • Paul Marks

    Long term mass involuntary unemployment has got nothing to do with "lack of spending"(AKO "lack of effective demand").

    Long term mass involuntary unemplouyment can ONLY be caused by something preventing the labour market clearing – i.e. something preventing the price of labour (WAGES) adjusting to changing labour market conditions.

    Printing more money (or creating more money by book keeping tricks) will not reduce unemployment in the long term if the labour market is broken (for example by government regulations that allow unions to monopolise the supply of labour) – all that will happen, in the longer term, is that wages will go up and the mass unemployment will return.

    Even Keynes understood this – as is clear from the introduction to the German edition of his "General Theory…."

    He understood that the reason that mass unemployment had been eliminated in Germany but remained in the United States (in spite of both governments indulging in a orgy of print-and-spend) was that, in Germany, the government no longer backed the institutional restrictions on the labour market – indeed the government had forceably acted to break those institutional restrictions.

    When on May 2nd 1933 the union officials came back from their day off for May Day events they found their offices occupied by Storm Troopers. No free market person could support such government interventionism – but Keynes seems to have had no real problem with what he openly called "totalitarianism".

    A free market solution would not have involved Storm Troopers (or World War II style wage controls – the real reason that unemployment collapsed in WWII, real "black market" prices rose more than wages so the LABOUR MARKET CLEARED). A free market solution would have been like 1921.

    Simply allowing the market place to operate (in the teeth of a credit money bubble bursting) – without govenrment support for wage rates (as both Hoover and Franklin Roosevelt did, in the name of "supporting demand") OR government attacking unions and imposing legal wage controls (as the National Socialists did in the 1930s and the American and British governments did during World War II).

    Odd that the economists (such as the late Lord Keynes) who claim that artifically "stimulating demand" will fix mass unemployment are quite happy to use what they openly call "totalitiarian" methods to clear a labour (a wage) market when the "stimulating demand" policy DOES NOT WORK in the long term (and so much for "in the long term we are all dead").

    Rather than Storm Troopers and government wage controls – would it be so bad to just repeal the government regulations (the welfare laws, the mininum wage laws, the pro union regulations) that actually cause "wages to be sticky downwards" and allow the market to work?

    "But we do not want wage rates to have to fall".

    No problem……

    Then do not create a vast credit bubble (such as the one that Benjamin Strong of the New York Federal Reserve helped create in the late 1920s) in the first place.

    If you do not artificially increase wage rates (in a credit-money "boom") they will not have to dramatically fall, after the "bust", later.

  • Paul Marks

    1921 is a classic refutation of Keynesianism – so (or course) it is ignored.

    A massive credit-money bubble bust (much to the astonishment of Irving Fisher – who assumed, as the "price level" was stable things must be basically O.K., Fisher learnt nothing and made the same mistake in 1929).

    So what does government do?

    Does it (as then Commerce Sec Herbert "The Forgotten Progressive" Hoover suggests) follow a policy of internventionism in the labor market (to prevent wage rates falling) and push government spending on various schemes?

    Well these "Progressive" policies had to wait till after the 1929 bust (then President Hoover also eventually followed a tax-the-rich policy).

    What actually happened in 1921 was the government allowed the market to clear (i.e. wages to fall) and cut its own spending. Not "cut the increase" – actual cuts in base line government spending (by some 25%).

    According to Keynesian (and some other Schools of economic thought) doctrine the bust should have turned into a depression.

    In reality the economy went into recovery in six months.

    No wonder that so called "empirical" economists do not actually like such empirical examples.

  • Tom Dougherty

    Hi George,

    I have posted the following comment on Scott Sumner’s blog regarding your graph comparing the index of NDGP to wages. I hope you don’t mind me reposting my comment (with slight revisions) here.

    I have only skimmed the comments, but I did not see anyone questioning why one would use the Average Hourly Earnings of Production Workers for the wage index. Since 2006 BLS has produced the Average Hourly Earnings of All Employeees which has more comprehensive coverage than production workers only. Using the production workers index you are basically looking at the manufacturing sector only and throwing out every other sector of the economy. The service industry is a very important aspect of the economy and should not be discarded, hence, the All Employee wage index should be used. The only problem with the All Employee index is that it has only been produced since 2006, but this should be sufficient for our purposes.

    The link compares the Average Hourly Earnings of All Employees index and the GDP index:

    Using the All Employee wage index compared to the GDP index one can see that the gap has nearly disappeared. I would say that wages have completely adjusted especially if you extrapolate the GDP data to 2nd quarter. Given that wages have now completely adjusted, it would seem that George is right that additional stimulus to the growth in nominal spending would have little impact reducing unemployment.

    • George Selgin

      Thanks for this, Tom. I settled on the more limited statistic because I thought it best to use a longer series so as to be able to normalize around a date (2003) that could reasonably be considered to be between cycles. Your chart makes me once again wonder whether there's no longer any basis for calling for further monetary stimulus, and as such seems to take the discussion back to where I left it in my first post. I will of course log on to MoneyIllusion to see what's going on there.

      • Rob R.

        Yes, this chart makes the point from yesterday's post very clear – and as it includes a wider range of workers it seems more relevant to the issue too. I hope Scott Sumner will comment on this data – as I think yesterday's chart let the MMist off the hook slightly.

  • Greg Ransom

    A large portion of the decline of employment has come in the government sector, where government unions have chosen to sustain wage and benefit increases (for relatively more senior employees) while letting go more junior employees.

    Take a look across California in all of the various government worker sectors, education, police and fire, etc.

    College administrators in California also continue to bring home ever more massive salaries. You can google that as well.

    In my area teachers continue to get wage & benefit increases and younger teachers continue to get red slips.

    And in all sectors, a lot of the "wage" increase is in the benefits portion of wages, healthcare, etc.

    • Paul Marks

      That is the way "collective bargaining" tends to work Greg.

      For all their screaming and nashing of teeth over unemployment, union bosses and activists much prefer higher unemployment to reductions in wages and benefits (and establishment "economists" rush to agree with them – by claiming that cuts in wages and benefits would "reduce demand" and be "bad for the economy").

      It is more seen in the "government sector" than in the "private sector" because, presently, governement workers tend to be more unionized than private employees are.

      Sane employers (private or government) should have nothing to do with "collective bargaining" – but government regulations tend to push it (for example it has been pushed in Britain since the Disraeli Act of 1875 – pushed even more strongly since the Act of 1906).

      In the United States if a majority of employees ask for collective bargaining (or, if the NLRB has its way, a majority of employees in a particular section ask for it) then an employer HAS TO grant it – even though it may well meaning destroying the enterprise in the long (or not so long) term.

      In government work collective barganining has the effect of leading to bankruptcy – as wages and (particularly) benefits (pensions and so on) become unsupportable for taxpayers (at this point government union activists scream "we are taxpayers to" thus showing a total lack of grasp upon reality, it is much as if a mugger stole your wallet and then handed you back a single ten Dollar note out of all the money you had in the wallet, and then wept and stamped his feet over this "ten Dollar TAX" he has had to pay).

      In practice things are not quite as bad as they would seem from the above.

      In "right to work" States people (whether in private or government employment) do not tend to join unions (although many private employees resist joining unions even in non right to work States) – so unions tend to be far less powerful and do far less damage.

      One of Milton Friedman's greatest (non monetary economics) errors was to oppose Right To Work laws. Technically he was quite CORRECT – in a free market if an employer wants to to demand that all employees join a union.

      However, the present situation is NOT a free market – the pro union regulations (such as forbidding "Yellow Dog" non union contracts) give unions unnatural power to undermine a free market (and help, along with government regulations and welfare programs, produce mass unemployment and bankruptcy). Milton Friedman was well aware of this – so his opposition to Right To Work laws is hard to explain.

      In the future States that have Right To Work laws are likely to be undermined by Federal regulations just as States that have relatively sound finances are likely to undermined to bailout bankrupt States (such as New York and California – both of which are likely to be bankrupt by 2014, with the economic collapse of 2013 bringing backruptcy much quicker than is presently exected) and by the effect of the FEDERAL GOVERNMENT.

      The Federal goernmnet is essentially a Black Hole in the economic universe – its fiscal (and regulation) position is basically hopeless.

      A relatively sane State should (logically) undertake secession from a Federal government whose monetary system is a vast credit bubble, whose regulations will strangle productive enterprise and whose fiscal position (unfunded liabilities being vastly MORE than the 16 TRILLION official nation debt) is hopeless.

      However American history (for example the Civil War of 1861-1865) will lead to any State that seeks to break away (no longer be bound to a corpse) being branded as "racist".

      Therefore the only logical POLITICAL hope is the victory of "Mitt" Romney in November.

      It is very doubtful if this man will change his (RINO) habits of a lifetime and become a dedicated free market person seeking to roll back government – but at least Mr Romney is not a Frankfurt School cultural Marxist who actually WANTS to destroy civil society (Barack Obama is and does).

      Also if a "President Romney" was a failure (the most likely possibility) States would have less of a difficult time in leaving the Union.

      After all with a rich white Republican as President it would be hard to brand a State government that tried to leave the Union "racist" (or other such).

      As for the national debt – it has mostly been built up to fund blatantly unconstitutional spending (Federal government spending in violation of the Tenth Amendment) and so should be repudiated by any State leaving the Union.

      So should unconstitutional (and financially unsupportable) entitlement programs such as Social Security, Medicare, Medicaid and SCHIP – with more moderate transitional programs being set up by States to give some support to those people currently dependent on Federal government programs and too old and poor to support themselves (although civil society action – by family, local voluntary communities, churches and so on will be vital in the terrible times ahead).

      As for currency – some States are already exploring what would be acceptable in terms of the payment of State and local taxes (no suprise that gold and silver are top of the list).

      "Paul you are totally wrong – Mitt Romney and a new Republican House and Senate will save the United States after they come into office on January 20th 2013".

      Well I certainly hope (for the good of the world) that this view is correct and that a wise "Mitt" Romney and a Congress in which (by some wonderful election results) neither Democrats or RINOs have any power, manage to guide the United States in the terrible times that are to come.

      The times of economic collapse that are already "baked into the cake" by monetary and fiscal policy judgements that have occured over DECADES.

      However, on the off chance that a President Romney and a Repulican Congress do not prove to be Supermen….. it might be wise to plan for possible secession.

      Of course if Comrade Barack wins things will be almost automatic.

      The long schemed for take over by the totalitarians will proceed (what to do to make Congress nonrelevant has already been worked out – and worked out in great detail).

      Then the choice will be to fight (fight with a very uncertain hope of victory – as any resistance to totalitarianism will be branded "racist") or to live as slaves of the state.

      At least live till the mass starvation starts. And with the sort of policies that the Comrades would follow (designed to do such things as cripple private farming) the msas starvation would not take many years to arrive.

      All rather depressing – perhaps the "bitter almonds alternative" (as practiced, for example, by that convicted trader in the court room) is the way to go.

      I state the obvious (for the "why should we care what happens to the United States, we live in…" crowd) if the United States falls to totalitarianism the rest of the West can not stand (certainly not Britain – which is on verge of collapse already).

      Even a break up of the United States would have terrible effects around the world – but it would be less bad than a world in which the United States had become totalitarian.

  • Tom Dougherty


    I posted a reply to Scott Sumner regarding his comment that the difference in the size of the gap between the graph I created and your graph was at issue. I think I have addressed that issue in the comments below and am reposting here in case you missed it:

    George’s graph does have a bigger gap than my graph but this is mainly a scaling issue. Also, mainly because of the scaling it looks as though the gap in my graph is closing faster than in George’s graph. If you notice on George’s graph, the GDP index increases by seven for every five the average hourly earnings graph increases. The GDP index is the right side scale and AHE is the left side scale. And on George’s graph the gap is actually widening between the GDP index and the AHE index.

    Reproducing George’s graph and putting both indexes on the same scale with the same scaling (not the 7 to 5 ratio) you can see the widening of the indexes:

    His graph is widening with GDP on the top and my graph the gap is narrowing with GDP on the bottom. Both are telling the same story that the GDP index is growing faster than the average hourly wage index for production workers or all employees. My version of George’s graph is simply easier to read and interpret. The choice of AHE is causing some of the difference too.

    Since the second half of 2009, GDP has increased by 11.1%, and per capita GDP has increased by 9.0%, AHE for production workers has increase by 5.6% and AHE for all employees has increase by 5.2%. I present these numbers just to quantify how much gap has changed since the second half of 2009. And to addresses your point of the difference between GDP and per capita GDP.

    • George Selgin

      I much appreciate your improvements on my attempts to get to the bottom of the question I raised in my first post, Tom.

      On his own site Marcus Nunes offers some critical observations on my own analysis, some of which would seem to apply as well to your efforts. I have responded to him there, but would welcome your own reaction to his remarks.