The March jobs report

The headline numbers were more of the same, but there are a couple of other data points that refute some recent hypotheses that have been floating around:

1.  Average hourly earnings fell from $24.31 to $24.30.  That’s probably not statistically significant, but it does slightly undercut a recent argument that people have been making.  Hourly wages have been trending up at about 2% per year since 2009. If we are reaching full employment, and/or the Fed’s 2% inflation target, you might expect a modest rise toward a more normal 3%, or a bit higher.  Some observers thought an upswing was starting last month, when the 12-month increase rose to 2.19%.  With this number the 12-month increase is back to 2.06%.  Because Janet Yellen focuses on wage growth, this is a slightly dovish signal for monetary policy. They need to do more to hit their 2% inflation target. BTW, nominal wages are the correct measure of “inflation” for monetary policy purposes.  We care about disequilibrium in the labor market, not the rental market or the oil market.  Good to see Yellen focused on that variable.

2.  Ed Lazear is a very fine economist, but recently wrote a long WSJ opinion piece that is almost entirely based on a simple misconception.

The job-equivalence number is computed simply by taking the total decline in hours and dividing by the average workweek. For example, if the average worker was employed for 34.4 hours and total hours worked declined by 344 hours, the 344 hours would be the equivalent of losing 10 workers’ worth of labor. Thus, although the U.S. economy added about 900,000 jobs since September, the shortened workweek is equivalent to losing about one million jobs during this same period. The difference between the loss of the equivalent of one million jobs and the gain of 900,000 new jobs yields a net effect of the equivalent of 100,000 lost jobs.

.  .  .

What accounts for the declining average workweek? In some instances””but not this one””a minor drop could be the result of a statistical fluke caused by rounding. Because the Bureau of Labor Statistics only reports hours to the nearest 1/10th, a small movement, say, to 34.449 hours from 34.450 hours, would be reported as a reduction in hours worked to 34.4 from 34.5, vastly overstating the loss in worked time. But the six-month decline in the workweek, to 34.2 from 34.5 hours, cannot be the consequence of a rounding error.

Was it the harsh winter in much of the United States? One problem with that explanation is that the numbers are already seasonally adjusted.

Seasonal adjustments reflect an average winter, not a harsh winter.  Thus a “harsh winter” is a quite reasonable explanation for the fall in average weekly hours. Because Lazear misinterpreted the nature of seasonal adjustments, he went on to search for other possible reasons for the decline in average weekly hours, such as the disincentive effects of Obamacare.

Now it looks like the harsh winter did explain the drop in hours—the March numbers bounced back to 34.5, essentially where they have been since 2006, except for a brief dip in 2008-09.  The WSJ needs a follow-up piece on average weekly hours. Consider this my one word submission:




55 Responses to “The March jobs report”

  1. Gravatar of TravisV TravisV
    4. April 2014 at 05:48

    S&P 500 is up 0.17% so far.

    By the way, everyone should visit Bonnie Carr’s blog:

    She’s a market monetarist, and, more importantly, a HUGE inspiration!

  2. Gravatar of Major_Freedom Major_Freedom
    4. April 2014 at 08:03

    Inflammable means flammable? What a country!

  3. Gravatar of benjamin cole benjamin cole
    4. April 2014 at 08:09

    Economics is just politics in drag. Ed Lazear joins the fray, poorly garmented, too. He should check out Krugman or Cochrane for dressing tips.

  4. Gravatar of TravisV TravisV
    4. April 2014 at 08:15

    “We May Finally Be On The Verge Of A Seismic Shift In The Labor Market That Everyone Has Been Waiting For”

    “workers may finally be returning to the job market because the economy is improving and confidence in job prospects is coming back.

    The unemployment rate was unchanged at 6.7% in March, defying Wall Street’s consensus forecast for a tick down to 6.6%. However, many economists characterized the rate as staying elevated for “good reasons.”

    In other words, the unemployment rate remained elevated not because job gains weren’t solid “” at 192,000, the magnitude of the increase in payrolls was on the order of those that have dragged the unemployment rate down significantly over the past year “” but because labor force participation rose.”


    “If you are a dove on the FOMC, you can point to the rise in participation, steady unemployment rate and flat growth in earnings as a way to reiterate the idea that rate hikes remain in the distance,” says Neil Dutta, head of U.S. economics at Renaissance Macro.

  5. Gravatar of Morgan Warstler Morgan Warstler
    4. April 2014 at 08:23


    When you say nominal wages are the correct measure for inflation…

    Supposed nominal wages stay the same or grow by say 1%, BUT everyone consumes 5% more per year for their very slowly growing nominal wage.

    Say this happens for 5 years in a row.

    Why would you be sure that moving to 2% inflation (we’re not talking NGDP here) wouldn’t lead to less overall consumption?

    And isn’t whatever policy leads to the most consumption always the proper one?

    I’m talking about my invisible RDGP again…

  6. Gravatar of ssumner ssumner
    4. April 2014 at 08:29

    Morgan, Of course I’m not recommending 2% inflation. But I think 0% NGDP growth is unwise for sticky wages reasons.

  7. Gravatar of TravisV TravisV
    4. April 2014 at 08:40

    Morgan Warstler,

    George Selgin, Bill Woolsey and Prof. Sumner discussed what might happen with a 6% NGDP target vs. a 3% NGDP target in the comments section of this old post (January 2010):

  8. Gravatar of Morgan Warstler Morgan Warstler
    4. April 2014 at 11:51

    That’s my next question Scott:

    How much more consumption YOY would cause people to stop suffering the mental block to sticky wages.

    Say labor’s consumption was increasing 5, 10, 20% YOY, is there no amount of new consumption that makes people ok with flat or even diminishing wages?

    Normally, we think in terms of desperation, labor must eat, so UAW can pay new workers in Detroit half as much, but maybe it’s bc they are consuming a TON OF FREE SHIT, that makes everything else seem silly.

    Kids don’t want cars. Higher end mass fashion players are getting crushed bc kids don’t care about having $80 jeans anymore. Followers and views are coin of the realm, and they are free.

    How are we sure that if wages aren’t moving up, it isn’t bc demand (relative to the history that caused $ Illusion to become a thing) for higher wages has fallen?

  9. Gravatar of TravisV TravisV
    4. April 2014 at 13:29

    Marcus Nunes wrote a new post with AWESOME graphs:

    Bill McBride’s conventional wisdom is so complicated while the market monetarist story makes so much more sense……

  10. Gravatar of Major_Freedom Major_Freedom
    4. April 2014 at 15:38


    Every one of those charts are completely consistent the theory that employment declines cause declines in NGDP, ceteris paribus. Investment declines first which does not immediately result in a decline of NGDP, and lower investment causes lower wage payments and profit income ceteris paribus, and lower wage payments and profit income causes lower spending on final goods ceteris paribus.

    These charts do not exclusively show or verify or prove that the steep decline in NGDP caused the decline in wage payments.

    If Nunes wants to show NGDP changes cause wage payment changes, then he won’t be able to use these charts.

  11. Gravatar of benjamin cole benjamin cole
    4. April 2014 at 16:04

    Excellent point about labor force, Travis.

  12. Gravatar of Tom Brown Tom Brown
    4. April 2014 at 16:11

    O/T: Can someone please explain the logic here to me, I don’t think I quite it:

    I’ve never really understood that post completely, but you’ll notice that both Rowe and Sumner (Sumner in the comments) make a similar point.


    How low do you want the inflation target to go? How big do you want the central bank to be? How big do you want the fluctuations in size of the central bank to be?

    Do you want a central bank that sometimes needs to own all the government bonds, all the commercial bonds, all the shares, all the farmland, all the houses…to keep inflation (or NGDP) on target?

    You probably don’t. (Unless you are some sort of extreme socialist who wants the government-owned central bank to own everything.)


    Regarding the “extreme socialists,” I like to sometimes tease conservatives who want really low inflation by pointing out that that they are advocating socialism.


  13. Gravatar of ssumner ssumner
    4. April 2014 at 16:15

    Tom, The lower the inflation rate the higher the demand for base money. The higher the demand for base money the higher the ratio of the base to GDP. The higher the ratio of the base to GDP the higher the ratio of the Fed’s assets to GDP.

  14. Gravatar of Tom Brown Tom Brown
    4. April 2014 at 16:17

    Scott, thanks. That makes sense. So when Miles Kimball talks about a plan for a 0% inflation target, how does he avoid that problem? Is it because he gets rid of cash too and thus the CB can decrease demand for deposits (all deposits) by instituting a negative interest rate?

  15. Gravatar of ssumner ssumner
    4. April 2014 at 16:25

    Tom, Even at zero percent inflation the demand for cash is manageable–it’s reserve demand that can get really really large.

  16. Gravatar of Tom Brown Tom Brown
    4. April 2014 at 16:28

    Ah… OK, so does that mean negative IOR then? (to keep the Fed small at a 0% I.T.)?

  17. Gravatar of Tom Brown Tom Brown
    4. April 2014 at 16:33

    Basically, what would Miles do to avoid a huge Fed?

  18. Gravatar of Morgan Warstler Morgan Warstler
    4. April 2014 at 17:04

    Travis, Marcus, Scott…

    Those charts are great, I mean nearly mind meld break thru bc they get to the VERY THING I keep talking about.

    DIGITAL DEFLATION is not being measured correctly.

    Look at the productivity chart:

    According to it, the productivity gains of 2001 are best by far, then 1982, then 2007 and 1990.

    Marcus even points to 2001’s massive effect on NGDP:

    “Left to be explained is the identical rise in NGDP in the 1990 and 2001 cycles associated with very different employment gains on the two occasions. An explanation can be gleaned from the stellar behavior of productivity in the 2001 cycle.”


    It just is. I can do with 2 coders what I could do 20 back then. The smart phone has done exponentially more than the computer for productivity. AirBnB, Uber, Drones, not Jet Fighters, the list is forever. The joke amongst guys who’ve been here since day one online, is that all the ideas we had in 1999, are now all billion dollar plays proving we were right back then, but the tech couldn’t do all the stuff we dreamed up when we needed 20 coders instead of 2.

    This is the point: SOMEONE IS LYING or SOMONE IS TOO DUMB to be allowed to measure productivity gains.

    If that Purple line (2007) in Productivity was correctly drawn it would be way above the Green line of 2001, it would put 2001 to shame.

    And if my story is correct:

    We have even more stellar UNMEASURED productivity which gives us the super negative effect on employment.

    And my question is, in light of this Scott, HOW CAN WE BE SURE, that if we pushed the NGDP higher (via inflation), that it wouldn’t have brought our now correctly drawn awesome productivity line down?

  19. Gravatar of TravisV TravisV
    4. April 2014 at 17:07

    George Soros advocated 5% NGDP targeting for the Eurozone in 2012:

    And now he’s talking about investing in European banks and real estate:


  20. Gravatar of TravisV TravisV
    4. April 2014 at 17:26

    Morgan Warstler,

    Right now, we have huge excess capacity. Therefore, there is a strong positive relationship between NGDP growth and productivity growth.

    There are huge economies of scale. Just think of the U.S. economy as a factory. If demand increases and factory capacity utilization increases from 80% to 95%, then factory owners achieve a far higher return on their fixed investment (a sunk cost).

    That’s the explanation for this graph:

    Here’s some more great stuff on technology and excess capacity:

  21. Gravatar of TravisV TravisV
    4. April 2014 at 18:02

    Re: technology and unemployment,

    Marcus Nunes created a key graph here:

    Beginning in 1996, U.S. productivity started surging thanks to new technology. From 1996 to 2000, real growth exploded to an average of 4.4%. As real growth accelerated, the Fed should have reduced inflation to stop NGDP growth from overheating. But they were too slow to act and NGDP stayed well above the trend line for several years.

    Eventually, the Fed realized they had been asleep at the wheel and they aggressively tightened monetary policy in early 2000. Unfortunately, they overdid it. They didn’t get NGDP to return to the trend line. Instead, it fell way, way below the trend line. As a result, we had a recession and high unemployment during most of GWB’s first term.

    A couple things to note: first, in the graph above, notice that the positive NGDP gap from 1996 to 2000 is way smaller than the negative NGDP gap from 2000 to 2006. Clearly, the Fed really hated high NGDP growth (and thus inflation)!

    Second, central banks overreact like this all the time, resulting in more unemployment than is necessary. Which leads many people to the (understandable) conclusion that “new technology means fewer jobs.”

    But it doesn’t have to be that way. If the Fed would just keep NGDP growth more stable, then the Luddites (discussed here: wouldn’t be correct anymore.

  22. Gravatar of Morgan Warstler Morgan Warstler
    4. April 2014 at 18:36


    We don’t have much excess capacity at all. Not in the things people want to buy. And not in the humans laying around.

    My point is we want less cars. That the factories want to make MORE of them? Uninteresting.

    We want cheap clothes. We have figured out the CHICKEN WINGS are the best part of the chicken.

    What people want is a $350 VR rig that they can pout on their head and consume UNLIMITED amounts of virtual consumption, all day, never go buy a book, or a cd, or dvd, whatever again…

    WHEN they do want to buy something, say Whiskey, they want to buy something hand crafted in small batches, that by DEFINITION cannot make “productivity gains” or it becomes inauthentic.

    What people want is smartphone that rewards the most social (free), not the people with expensive baubles.

    And ANYBODY who can develop full stack mobile apps for smartphones? Instantly has work and makes a pretty penny doing it. No excess capacity inthe tings people WANT.

    MEANWHILE, the people who don’t have jobs?

    Not much relative capacity in them at all!

    That’s why we need GI/CYB, bc the productivity gains that we make will ALL come from search matching low end skills to let the lower and middle class get the kind of personal touch services, that DO NOT scale that the upper middle and upper class can afford today.

    The same coffee buying experience, BUT fancy coffee in poor areas costs 60% less, because labor costs, 80% less.


    My larger point is still we can’t accurately measure “growth” bc the things we all want, the things we all demand, the things that have NO MORE CAPACITY AT ALL – they are all FREE.

    If 57 people can completely replace the entire globe’s SMS messaging system…

    If VR rigs can turn you happily into Neo in the Matrix, we have a fundamental shift in the polarity of Economics.

    Now you can say, well thats in the future! But that’s not an answer.

    If we have only 20 years until people consume 1/3 as many atoms, and NOBODY has any debt, and NOBODY WANTS anything that comes out of a factory, and they ALL want to buy small batch, locally grown, handmade authentic calories.

    The idea of what is a “productivity gain” has to fundamentally change.

    Because every year is more invisible RGDP, so RDGP starts to look outright negative, while everyone is clamoring to plug in a drop out the current economic text books.

  23. Gravatar of Major_Freedom Major_Freedom
    4. April 2014 at 18:51


    “Right now, we have huge excess capacity. Therefore, there is a strong positive relationship between NGDP growth and productivity growth.”

    That is a non sequitur.

    But more importantly, there is no rational justification in lumping together private production directed to consumer preferences, and coercive government redirection of capital leading to production of weapons of war and the domestic police state apparatus, and then claiming the world is a better place simply because you observed a temporal increase in the aggregate supply of “stuff”.

    “Beginning in 1996, U.S. productivity started surging thanks to new technology. From 1996 to 2000, real growth exploded to an average of 4.4%. As real growth accelerated, the Fed should have reduced inflation to stop NGDP growth from overheating. But they were too slow to act and NGDP stayed well above the trend line for several years.”

    “Eventually, the Fed realized they had been asleep at the wheel and they aggressively tightened monetary policy in early 2000. Unfortunately, they overdid it. They didn’t get NGDP to return to the trend line. Instead, it fell way, way below the trend line. As a result, we had a recession and high unemployment during most of GWB’s first term.”

    There was nowhere near enough deflation to expose the widespread malinvestment. The Fed reinflated too much yet again (below trend NGDP can still be too inflationary from a market standpoint).

    Precisely because of the significant inflation up to 2000, a recession was inevitable. Instead of allowing enough corrections to take place, the Fed decided to reinflate to avoid the short term pain. As a result, they blew up an even bigger economic bubble, which was composed in part by a housing and real estate bubble.

    The Fed was again far too loose post 2008. Once again, instead of allowing enough corrections to be made, the Fed reinflated, and blew up an even bigger economic bubble. The continued bubble formation (increase) since 2008 is even worse than the bubble formation and partial correction prior.

    You cannot eliminate malinvestment with more of what caused the malinvestment in the first place.

  24. Gravatar of TravisV TravisV
    4. April 2014 at 18:55

    Morgan Warstler,

    We’ll agree to disagree. The minimum wage increases have reduced AS, no question about it. But I still think we have a lot of excess capacity. Particularly after reading stories like this:

    Bottom line: Ambitious tech start-ups should be rooting for a high NGDP growth target.

    If national income is grows rapidly, innovators will capture a larger and larger share of that spending.

    If national income grows slowly and the economy is less dynamic, then crony rentiers will have an easier time guarding their oversized share of national income.

  25. Gravatar of TravisV TravisV
    4. April 2014 at 19:29

    I’m very curious whether small-cap tech stocks have followed this same pattern since 2008:

  26. Gravatar of Morgan Warstler Morgan Warstler
    4. April 2014 at 19:44

    Travis, I’m not sure it’s agree to disagree yet.

    The graphs clearly show that someone measured productivity and said 2007 was worse than 2001.

    And as far as I, or anyone who knows technology, am concerned that’s either a lie, or an giant idiotic mistake.

    If you or Scott or anybody think it’s not mistake, please tell me why.

    IF however you can grok that Tyler Cowen has been PROVEN WRONG, people have CHOSEN smartphones over toilets and planes and cars, and that means this is the BIGGEST BEST NON-STAGNATING PERIOD EVER IN HUMAN HISTORY.

    Tyler’s PROOF, has been proven wrong. His reasoning has been proven faulty. His new work is a mea culpa to my crowd.


    So IF that graph is wrong, if 2007 is far higher than any other line, then we have not only a new set of facts, we have proof that the MEASURING STICK is BROKEN.

    And I don’t know how we jump to the higher or lower NGDPLT discussion, until we have sussed out the productivity graph.

  27. Gravatar of TravisV TravisV
    4. April 2014 at 20:32

    Morgan Warstler,

    You might be able quibble with government productivity measurements. However, you can’t argue with stock prices (powerful forward-looking indicators).

    Consider David Glasner’s research:

    What did the behavior of stock prices tell us? From 2003 to 2008, we had (just about) full employment. Thus, the relationship between NGDP growth expectations and productivity growth expectations was small.

    But then, from 2008 through 2012, we had HUGE excess capacity, resulting in a strong positive relationship between NGDP growth expectations and productivity growth expectations.

    On 9/13/2012, when Bernanke surprised everyone with the size of QE3, tech start-ups instantly revised their expectations of future demand, inflation and productivity growth upward. How do we know? Stock prices and TIPS spreads surged.

    And I think the introduction of the Evans Rule in December 2012 might have had an even bigger positive impact on productivity growth!

    Think about the relationship between productivity growth and stock prices in the 1920’s and the 1990’s…..

  28. Gravatar of Morgan Warstler Morgan Warstler
    5. April 2014 at 00:09

    “Tech start-ups instantly revised their expectations of future demand?”

    Tech start ups did no such thing. Maybe the stock market did. Tech startups already expect nearly unlimited demand of their free product, but only if everybody gets it (network effects). The value of a piece of software for users is relative to the # of users using the software.

    Valuations of tech startups increased bc the money supply increased. They have to sell LESS EQUITY to buy the same amount of money:

    But the “relative” shift of power between the publicly traded vs. the startups, We can’t say whether QE helped or hurt that.

    And that is the productivity measure that matters.

    The Q I am struggling with is: I’m 100% sure 2007 is most productive time. And today even more than 2007.

    But I view Monetary stance as a secondary force here. It can help or hurt marginally, but it CANNOT actually change the fact that 2007 is most productive time.

    My Q is: HOW do we know that 9/13/2012 made 9B people on the planet MORE LIKELY to quit using SMS and start using What’sApp?

    Because THAT is the productivity gain that matters.

    And maybe if everyone was even more desperate to save a $, the world would have adopted a What’sApp (the money saver) two years earlier.

    That’s the alternative history that Scott is always talking about, he likes to say “things could be worse” – and I’m saying “things could be better”.

    But since we know that graph is wrong… it seems like Economists have some explaining to do about their assumptions, about their weights and measures.

    That graph looks like the proof the Global Warming scientists have been juking their stats. When it’s proven they are juking them, how does that effect their credibility?

  29. Gravatar of W. Peden W. Peden
    5. April 2014 at 01:28

    I can’t remember if this was circulated when it occured-

    Notable professor advocating NGDP targeting in London academic circles.

  30. Gravatar of TravisV TravisV
    5. April 2014 at 04:56

    Morgan Warstler,

    Part of the appeal of Sumner’s story is that it’s pretty darn simple. I feel like you’re throwing other stuff into the mix that overcomplicates it.

    I can vaguely see this notion that “Depression is good! It cuts the crony capitalist rentiers down to size and makes them less powerful.”

    I don’t think there’s much to it, though. Crony capitalist rentiers remained very powerful during Japan’s depression.

    And it was far better to have been a tech start-up in the early 1920’s than in the early 1930’s.

    The economy is not a zero-sum game. Tech start-ups should not root for depression. They should favor capacity utilization and high employment along with all the rest of us.

    And to reiterate, the behavior of stock prices during the Great Depression and this recession indicate that there are economies of scale.

    If we move from 50% capacity utilization to 90% capacity utilization, then productivity growth should accelerate.

  31. Gravatar of Morgan Warstler Morgan Warstler
    5. April 2014 at 05:51

    Any alt.universe that gives the world Skype, What’sApp, a TB of free cloud storage, every movie, song, classroom for free, a free Smartphone and VR rig as a civil right and half as many public employees, that’s the optimal correct Monetary stance.

    If we currently have 2-3% Real growth happening today and not being measured, bc free and network effects conflicts with the idea of money illusion and atomic economy… the latter gets out of the way.

    I’m not sure tho, that MM grok and believe that there is major unmeasured productivity driven growth happening.

    Right now it’s just me and some tech guys saying it. It’s just Summer and co mis-diagnosing whats going on and saying things like we need negative interest rates to have full employment, so sadly we need more fiscal spending.

    But if MM learn and know it is real, they come to believe in it like they believe in NGDPLT, I’m not sure that it doesn’t mean right now have 4% NGDPLT and have a 20 years path towards 2% NGDPLT.

    It all starts with that line on productivity graph, and admitting Tyler’s GS is as wrong as goldbugs in 1930’s.

    Ultimately, that free smartphone world I keep talking about – it not only does happen, but it’s the world where we no longer have paper money.

    And not using MP to race towards that future is morally wrong, since that’s the economy we all agree works mostly easily, right?

  32. Gravatar of TravisV TravisV
    5. April 2014 at 06:13

    Morgan Warstler,

    I don’t think Sumner is mis-diagnosing what’s going on. I think there really is something big to his story. Why? Data like this:

    The data tell us that stocks hated inflation from 1968 to 1995. But then, as inflation kept falling, stocks began loving inflation.

    When I look at data like that, I’m inclined to think that downward nominal wage rigidity matters A LOT.

    Do you have an alternative explanation to Sumner’s and Glasner’s for the relationship between stocks and NGDP growth expectations?

  33. Gravatar of John Becker John Becker
    5. April 2014 at 06:14


    Shouldn’t wage growth be more flexible than what you seem to be recommending? For instance, if NGDP grows 5% and wages stay flat for the year, real wages have fallen and you would expect the labor market to get tighter. If the Fed were to ease more aggressively in this case, wouldn’t that be completely counterproductive to the goal of helping the economy?

  34. Gravatar of Michael Byrnes Michael Byrnes
    5. April 2014 at 06:21

    Morgan Warstler wrote:

    I don’t think this has any relevance for monetary policy targets.

    What you are describing is a real issue, not a nominal one. Free stuff is free whether a central bank’s target is 5% NGDPLT, 3% NGDPLT, a 2% flexible inflation target, or even a gold standard. Productivity gains mean that REAL prices fall, not necessarily nominal prices.

    NGDPLT is about supply the quantity of money that the public wants to hold. Maybe productivty gains and more and more free stuff will reduce the quantity of money the public wants to hold – in which case the supply of base money would need to fall simply to maintain the target.

  35. Gravatar of Mark A. Sadowski Mark A. Sadowski
    5. April 2014 at 06:57

    Off Topic.

    Thanks to the Freedom of Information Act (FOIA), Joseph Gagnon has gotten hold of the package of 21 memos on monetary policy at the zero bound that were prepared by Fed staff just before the December 2008 FOMC meeting:

    April 4, 2014

    What Were They Thinking? The Fed on the Brink of Zero
    By Joseph Gagnon

    “…The primary tool the Fed had to achieve its federal funds rate target at that time was the interest rate on bank reserves at the Fed, which was 1 percent in early December 2008. Almost all participants wanted to lower this rate and a few argued for a rate of zero percent. But most wanted a rate slightly above zero and they settled on 0.25 percent, which has remained constant to this day. The primary reason for a rate slightly above zero was concern that banks and money market funds needed to earn a spread between deposit and lending rates in order to pay their operating costs and they would find it difficult to impose negative interest rates on their depositors. In addition, there was some concern about reduced liquidity and disruptions in the bond markets that were already evident at low interest rates. The possible harm from further disruptions in banking and securities markets was judged to outweigh any small macroeconomic benefit from an even lower interest rate.

    The FOMC did not discuss the possibility of a negative interest rate on bank reserves, but it is widely agreed that a significantly negative interest is not feasible because banks would convert their reserve balances to paper currency. A lingering puzzle is why the Fed never lowered interest on reserves to zero in subsequent years, when financial strains had diminished and depositors and market participants had gotten used to the low rate environment, but standard macroeconomic models imply that the benefits of such a small decline would have been correspondingly small…”

    The part about converting reserve balances to currency caught my eye. He seems to be referring to his own memo, “Purchases of Conventional SOMA Assets”, which he prepared with Spence Hilton. On page 124 it says the following:

    “Yet another potential channel operates through the impact of expanded ER on bank deposit rates. The Federal Reserve pays for its acquisitions of assets by crediting the accounts of the sellers’ banks at the Federal Reserve. These banks experience an exogenous increase in deposit liabilities which is initially accompanied by an equal increase in their holdings of ER.5 The empirical results of Frame, Hancock, and Passmore (2007) suggest that when a bank receives an unexpected surge of deposits, it tends to reduce other liabilities, especially managed liabilities such as borrowed federal funds. In normal times, the federal funds rate is higher than most short-term deposit rates, so this response lowers total costs of the bank. However, if the federal funds rate were zero and short-term deposit rates were positive, the inflow of deposits would likely encourage the bank to lower its deposit rates and shrink its total deposits. As depositors moved to competing banks, deposit rates would fall throughout the banking system.6 To the extent that this process reduces banks’ overall cost of funds, competition would likely push down their lending rates, thus helping to stimulate economic activity.7”

    Footnote 6 reads:

    “6 Indeed, the only way the banking system can disgorge an aggregate increase in ER is by lowering deposit rates far enough to induce households and businesses to hold more currency. However, the elasticity of currency demand is believed to be very low, so we ignore this effect.”

    Recall that the endogenous money enthusiasts were constantly harping on the fact that “banks don’t lend out reserves”, and consequently negative interest on reserves would have no effect at all on reserve balances? Well the Fed seems to have thought that if you lower interest on reserves enough, this would have led to depositors holding more currency.

    Joseph Gagnon also says the following in his blog post:

    “The FOMC discussion shows that there was little appetite for a dramatic push to increase inflation expectations, with some participants expressing doubt that the Fed could raise expectations substantially through statements about its intentions without any additional actions. But there was also an acknowledgment that the Fed had not been as clear as it could have been about what inflation rate it aimed to achieve. Speeches and other published materials seemed to show a comfort zone for inflation with a lower end around 1 to 1.5 percent and an upper end at 2 percent. One of the background memos assumed an inflation goal of 1.75 percent. Participants did not agree on a common inflation goal at this meeting.

    The FOMC eventually did adopt a common goal of 2 percent inflation at the January 2012 meeting, which is at the high end of the narrow range that had been widely perceived by the markets. In recent years some economists have made the case for a temporary or permanent inflation target as high as 4 or 5 percent, but Fed officials have made it clear that such a change has not been taken seriously within the FOMC.”

    I’ve made the argument based on FOMC three-year forecasts for core PCEPI that the FOMC’s implicit inflation target is about 1.6%. So it’s interesting that if you turn to page 173 you’ll find the following:

    “This note examines the quantitative magnitude of the stimulus from conditional commitments about the path of the funds rate and from nonconventional policy measures (such as those discussed in Notes 16 and 17) as well as from fiscal policy measures. All simulations use the FRB/US model and assume perfect-foresight/model-consistent expectations in financial markets; the latter assumption implies that expectations of future actions drive asset prices. The simulations are based on the October Greenbook projection. We assume that the Committee’s underlying preferences are to stabilize inflation at 1¾ percent and the unemployment rate at the NAIRU (4¾ percent) over the longer run…”

    So the sumulation assumes the FOMC’s inflation preferences are 1.75%, not far above my estimate of 1.6%.

    Some of the memos that Gagnon doesn’t refer to also look fascinating.

    For example, there’s a memo on Japanese fiscal policy called “Japanese Fiscal Policy: A Bridge to Nowhere?” that shows that Japanese fiscal policy was poorly coordinated with Japanese monetary policy (they did almost all their fiscal stimulus before they hit the zero lower bound), and accomplished little other than dramatically increasing public debt. The associated graphs are useful for comparing the era of fiscal stimulus (1990s) with the era of QE (2001-6).

    The memo called “Effects of Very Low Policy Rates on Money Market Funds” also looks to be worth reading, especially in light of the Fed’s new reverse repo facility.

  36. Gravatar of Morgan Warstler Morgan Warstler
    5. April 2014 at 07:00


    Free stuff becomes free because investors go from owning debt to equity.

    “Software eating the world” literally means “there will be no debt.”

    What I’m struggling with is: Can NGDPLT set too high keep us from converting to a equity based system.

    And I get that I’m talking about real vs. nominal.

    Bt if real prices are falling faster than we are admitting (see that productivity graph that screws up 2007), or they could FALL EVEN FASTER – then that’s the optimal MP.

    And I’m committed tot he idea of full employment – but it ought to be done via GI/CYB.

    And I’m committed to using NGDPLT so that the economy doesn’t get us into 2008 territory when we get Obamacare and $1T is wasted fiscal stimulus.

    But since very few seem to grok yet the way productivity real growth is being mis-measured…

    We ought to.

    Right now we think NGDP is 4.5% and RGDP is 2.5%+ of it, and inflation is 2% or less.


    Well if RGDP is really 5.5% bc 3% of the growth / productivity gains are not being measured, it has a meaningful effect on the calculation we use to what it takes to get to full employment using MP, no?

    “Maybe productivty gains and more and more free stuff will reduce the quantity of money the public wants to hold – in which case the supply of base money would need to fall simply to maintain the target.”

    I think I agree with is, I just can’t figure out how NGDPLT is used to reduce the monetary base, without lowering the NGDP target over time.

  37. Gravatar of Philippe Philippe
    5. April 2014 at 08:35


    “Software eating the world” literally means “there will be no debt.”

    What does that mean? Please explain, thanks.

  38. Gravatar of ssumner ssumner
    5. April 2014 at 08:53

    Tom, Ask Miles, but he has advocated negative IOR.

    W. Peden, That’s good to see.

    John, You can argue that stable hourly wage growth is even more stabilizing than stable NGDP growth. But I see practical problems with targeting wages.

    Mark, I just did a post on that. I’ll add an update with a link to your comment.

  39. Gravatar of Peter N Peter N
    5. April 2014 at 09:30

    “elasticity of currency demand is believed to be very low”

    So how large is the elasticity of currency demand?

  40. Gravatar of Morgan Warstler Morgan Warstler
    5. April 2014 at 13:02

    Software businesses, web companies, etc. They are debtless things.

    Atoms are the reason previous businesses that dealt in atoms did so by leveraging debt…. the atoms were collateral.

    Digital businesses have basically nothing to leverage. Stop paying the power bill and the thing is turned off. Stop paying the coders and in 1 year the code is old and worthless.

    As such digital businesses work in debtless systems. You sell equity until you make profits. Once you have profits, you sell more equity to scale – since cost of scale is nothing compared to profits, since you already know you have a profitable thing.

    The WORST thing that can happen is get your equity diluted, so you don’t want to sell to much too soon.

    EVERYTHING is becoming software, so massive Brick and Mortar businesses built around atomic structures in basically every sector of economy except energy is becoming software.

    Software will close 90% of the college campuses, and the the companies that run them will have no debt, and the student that attend them, will have no debt.

  41. Gravatar of Philippe Philippe
    5. April 2014 at 15:42


    software startups might be able to fund themselves purely by selling equity, but that’s hardly the case for most businesses. And this free digital stuff – it’s usually only free because advertisers pay for the information that it generates, so that they can then sell more real stuff. If there weren’t the sales of real stuff to finance the production of free stuff, much less of the free stuff would get produced.

  42. Gravatar of Morgan Warstler Morgan Warstler
    6. April 2014 at 06:02

    Phillippe, when I say free, I mean cheap.

    I give you multiple examples on this site all the time, its not about advertising.

    “Software will close 90% of the college campuses, and the the companies that run them will have no debt, and the student that attend them, will have no debt.”

    I am not joking, we are all going to be putting on VR rigs in living in the metaverse (think Neo in that pod in the Matrix, except people all want to be in one and it’s not dystopian Skynet ugly). Eventually you basically only consume calories and electricity.

    That is the long run path.

    My point is:

    1. we should move there as fast as possible.
    2. MP can help our hurt our moving there.
    3. Slowly down moving there = being less productive.
    4. Moving there faster = making productivity gains.

    So software eats the world = long run productivity gains = you are weaning VR rig all day long.

  43. Gravatar of Philippe Philippe
    6. April 2014 at 07:03

    “Eventually you basically only consume calories and electricity.”

    Of course it’s important whether people have a real choice, or whether they have no choice but to consume only calories and electricity.

    What you’re describing could turn out to be a form of ‘digital poverty’, where people can’t afford real things and only have access to cheap virtual things. In this scenario real things become a luxury only enjoyed by the wealthy. Real college campuses cater to the upper classes, everyone else has to jack in to virtual reality campuses. That is pretty dystopian.

    I don’t see how wearing a virtual reality rig = long run productivity gains….?

  44. Gravatar of Morgan Warstler Morgan Warstler
    6. April 2014 at 07:47


    It’s a choice, and the choice is to plug in and consume VR. It doesn’t happen bc people have to do t, it happens bc they get to do it.

    Look, it’s hard to mentally cheer this future, bc it seems so bleak, but that’s only because you haven’t spent a Friday night with your friends hunting hot men (or women) in loin clothes with laser blasters on a planet moon.

    VR is a productivity gain, bc you still have the meeting, attend the class etc. You just don’t have to go anywhere. Why waste all that electricity and all those calories to have a sub-optimal experience.

    MY POINT IS in a system of ever better “conservation” of atoms and electricity is where technology leads.

    You UNDERSTAND THIS. Even with some concerns or caveats, you get the larger storyline path, you grok it is happening, and if it is a free choice, it is a happy future.

    Now then, think about monetary policy in a world of equity ownership and no debt.

    How does it get us from here to there/

    This is the question I’m wrestling with.

  45. Gravatar of Philippe Philippe
    6. April 2014 at 08:23

    your story doesn’t seem to make much sense..

    1. people will spend all their time in virtual reality, because they want to.

    2. most things (?) in virtual reality will be provided for free.

    3. people won’t consume much more than food and electricity.

    4. therefore there will be no debt.

    5. therefore we need NGDPLT targeting…?

    It sounds a bit silly.

  46. Gravatar of Philippe Philippe
    6. April 2014 at 08:27

    just to inject a tiny dose of reality into this scenario, what about people’s homes? Will they still need to have large mortgages to pay for their homes? Or will everyone be living in tiny pods?

    How do you pay for your mortgage, your calories and your electricity if everything that you make and consume (apart from calories and electricity) is given away for free?

  47. Gravatar of Morgan Warstler Morgan Warstler
    7. April 2014 at 08:52


    Start here —————————————> End there

    Today ——————————————> Cheery Matrix

    What do you think happens to people’s homes? THEY GET SMALLER.

    Now again, this is choice based, so it’s not like you have to go gently into Virtual Reality.

    And certainly economies, work, and consumption, exists in virtual space.

    Again, I’m not talking in absolutely, I’m actually explaining what’s going on in these amorphous phrases guys like Larry Summers use “secular stagnation”

    We are headed towards consuming less atoms and instead going digital.

    This means a shift from debt to equity.

    I just want to try and frame the discussion of the level of NGDPLT around the assumption of this shift, bc I’m not sure, that different levels don’t speed up or slow down this shift.

  48. Gravatar of Philippe Philippe
    7. April 2014 at 13:18


    “This means a shift from debt to equity”

    If people are going to be spending less in future, because so much is given away for free in the Cheery Matrix where everyone spends most of their time, then people’s incomes will also shrink.. This would suggest that if there was a reduction in total debt, there would also be a reduction in total equity… no?

  49. Gravatar of Morgan Warstler Morgan Warstler
    7. April 2014 at 14:53

    Phillipe, this is WHY we all favor NGDPLT.

    It ensures that we can take a bat to sticky wages via productivity gains, WITHOUT suffering the pain of deflation.

    Let me give you an example: Say that 20 years from now, atomic college attendance is down 50%. This means less professors staff etc. The remaining ones are teaching virtually and making MORE money.

    But what about all those people who used to be professors?!?

    NGDPLT lets us say who knows, who cares?

    Because overall there is no possible danger of having a recession where NGDP falls.

    Better would be nominal wage targeting, but NGDPLT is a close approximation.

    THE QUESTION I’M ASKING is just whether NGDPLT in 20 years should be 4% or 2%.

    In either case, we still get to say who knows who cares, bc there is ZERO chance that the average wage of the average worker will fall NOMINALLY.

    But 2% might get us more creative destruction, more progress, more consumption (real growth), than 4%.

    I’m not positive about this, but I still haven’t gotten a real analysis from anyone.

  50. Gravatar of Major_Freedom Major_Freedom
    7. April 2014 at 15:40


    “Because overall there is no possible danger of having a recession where NGDP falls.”

    Not true dude. Recessions are encapsulated by real phenomena, and typically caused by monetary factors. But once the real economy is distorted from inflation, there is absolutely no possible way of avoiding the consequences. The only thing that could be done is both delay and exacerbate the recession to the future. You will never eliminate a recession caused by inflation.

    How long can it be postponed? That is a historical question. But make no mistake, with NGDPLT, precisely because it demands for even more monetary inflation (at the onset of a recession) than previous price targeting rule inflation, and accelerating besides, what this means is that if the central bank is relentlessly pursuing NGDPLT, then the recession will manifest itself eventually as hyperinflation, since as distortions accumulate to infinity, it requires an infinite rate of money growth to postpone to the future.

    A choice to continue accelerating money growth, or allow corrections to occur, is what happened in Australia, incidentally. Money supply growth grew to a peak of over 20% annually by 2008, after years of accelerating growth. The RBA decided to allow (some) corrections to occur, and unemployment to rise. If instead they kept attempting to reverse the growing deflationary forces, and prop up NGDP against the market, then that 20% rate of growth would have reached 25%, 50%, 75%, 100%, and so on, and then the recession would be even more acute than it actually was.

    Money is not neutral. It is not neutral in the short term, the medium term, nor the long term. The long term is a function of successive short terms.

    Right now our economy isn’t just the exact same as it would have otherwise been in a free market of money, with the only difference being the height of the price level. That is a preposterous belief about economic activity. In truth, entire new industries can arise due to inflation that never would have otherwise arisen. For example, because we live in a world with many governments who impose their own fiat systems on each population, the monetary system is so completely warped that billions of dollars and tens of thousands of individuals are employed in the finance sector whose sole economic function is to attempt to safeguard against currency volatilities due to political whim. White blood cells attracted to cancer, turning into callouses. The world would not have looked like this if we had a free market in money, that is, money subject to market forces. Money today is almost entirely chaotic. Virtually no market signals in money anywhere in the world.

    Thousands more economists who do nothing all day but study the best ways for central banks to rule.

    Indeed, the bulk of Wall Street hedge funds are likely an outcome of fiat money.

    That there is malinvestment caused by inflation is irrefutable, by the way. More inflation cannot solve it.

    NGDPLT is a pipe-dream. A new God to replace the old God.

  51. Gravatar of TravisV TravisV
    7. April 2014 at 18:28

    Morgan Warstler,

    Why would there be substantially more tech innovation at long-term NGDP growth of 2% than at 5%?

    Since World War II, we’ve seen that various countries have a remarkable ability to thrive at many different rates of inflation (as long as it’s 1.5% or higher).

  52. Gravatar of TravisV TravisV
    7. April 2014 at 18:43

    Morgan Warstler,

    In fact, I think the vast majority of economists believe that a country with long-run inflation of 6% wouldn’t grow (and innovate) any faster if its long-run inflation were 2% instead. In the long-run, the rate of real (and thus productivity) growth is determined by other factors.

    David Glasner and I believe that a country with long-run inflation of 0% will grow slower than a country with long-run inflation of 2%:

    However, I think our view is rather rare among economists….

  53. Gravatar of Morgan Warstler Morgan Warstler
    8. April 2014 at 05:01


    Again, it’s all just one thing for me: DIGITAL DEFLATION.

    I’m not saying the historical # is wrong.

    I’m saying the current # is wrong.

    I’m saying we have a MORE REAL GROWTH than we have ever had in the US.

    Again, Tyler Cowen has been proven wrong. His GS work was wrong. The Internet / Smartphone is better than any previous human invention, the other silly little inventions Tyler thinks about like planes and toilets and cars and vaccines, etc – all that ANALOG stuff has been proven to be less valuable to humans – they want the smartphones more!

    Ok, so if today is the best fastest growth economy ever.

    Whether or not there are jobs for everyone is NOT INTERESTING. A host of serious next gen economists agree with me on this, you Travis agree with me on this…


    If the very thing that makes Tyler miss the importance of smartphones is the very thing that keeps us from “seeing” the real growth that is occurring.

    Say there is an extra 3% RGDP that isn’t being counted… how would we see it?

    Well we couldn’t measure prices, those are nominal, and we see them going up by 2% YOY.

    We measure the GDP and we see it at $17.7T or whatever.

    We then subtract 2% inflation.

    But if the measure of GDP is done by counting nominally then WHERE IN HERE:

    Where do we see the value of the modern digital age? Where do we see productivity gains as growth?

    We know consumption is higher today – people get MORE video, more songs, more written words but not more electricity, more calories for less money.

    We know how the end of $1T in college debt ends. We can see it coming.

    We know how cars get shared, and drones fly planes and fight wars, and VR is coming, and all that stuff, the most important stuff ever created, it all basically happens for a great big nominal discount over today.

    The ONLY way it happens is if it is a nominal discount over today.

    Why would 6% long run inflation vs 2% matter?

    Because in a digital world, prices fall. Real price gravity in digital space vs. atomic s like Jupiter compared to earth.

    And right now you are still thinking historical inflation in atomic space – about Amazon vs. Retailers, not your kid has a bike made out of nanotech carbon and plastic that simply gets melted and pulverized down and remade in your town when he grows 4″ – and having that bike recast costs $10 in today’s money. And the design of the bike costs nothing, some kid in India made it last year in 30 hours during a free online 3D modeling class.

    Greenspan used to say the weight of GDP was shrinking. What if he hadn’t even hit the tipping point? And DISTANCE of moving atoms is the real cost curve we’re going to end.

    I can be convinced 6% inflation is workable…

    But not by people who don’t scream how wrong the current measurements of the economy are.

  54. Gravatar of TravisV TravisV
    8. April 2014 at 05:35

    Morgan Warstler,

    No question amazing tech innovation is happening. But still, the human mind is the critical element driving the economy. Human psychology still matters. If anything, I think people resist nominal wage cuts even more than they did in the early 1990’s when Truman Bewley made his investigations.

    Nominal GDP certainly misses a lot of what’s actually going on. No question about it. But it’s still an excellent predictor of the unemployment rate. In late 2008, NGDP crashed, therefore it’s not surprising that employment crashed. Wages are sticky.

    Since then, NGDP has grown very slowly from the bottom at merely 4% per year. A recovery has happened but only very slowly. There has been a buyer’s market for labor, but it takes workers a long time to adjust their demands for wage growth downward to adjust to that new reality.

    There is every reason to believe that if the Fed had done a better job and grown NGDP at the rate it grew during the Reagan / Volcker recovery, then our employment / population ratio would be far higher today. As Sumner has pointed out in his National Affairs article, “During the first six quarters of the “Volcker recovery,” starting at the end of 1982, NGDP rose at 11%, RGDP rose at 7.7%, and (therefore) inflation was about 3.3%.”!

  55. Gravatar of Morgan Warstler Morgan Warstler
    8. April 2014 at 09:42


    The point of NGDPLT isn’t to create jobs.

    The point is to shave off the bottoms and tops of economic booms and busts, to stabilize the price level, to give everyone a long term perfect view of the size of the economy in X # of years, to quell the growth of bad government.

    “Full” Employment has nothing to do with this.

    Imagine that 50% of the people in US were mentally handicapped, would we count them in the labor force?

    Imagine we were heartless and all thought people earning even $1 a day was plenty, and we cared not a whit if the lived in street, and ate trash, etc. And 90% have jobs, is that “full” employment’?


    What you are talking about isn’t full employment, you are talking about he idea of a “living wage” – that somehow most people are going to be able to afford 2 bedroom, heat, hot water, A/C, cable TV, broadband, mobile broadband, education, and healthcare.

    And the % of people who can deliver value to others enough that they can afford to buy all those things on their own… is decreasing. Largely bc we have ever-expanding ideas about what is acceptable. 50mbps, not 2. 4G, not 3G. MRIs, not Xrays. Etc.

    So having “full” employment is really a function of deciding / admitting we are gong to give people these things, and simply requiring them to work for whatever they can earn to put in the kitty.

    So PLEASE just for this discussion let go of the concern about the employed. FORGET THEM.

    Now just think about how as quickly as possible we adopt digital deflation, moving from atomic economy to digital one.

    Again, pay no attention to who loses a job etc.

    The thing I’m trying to get to is just since we have full employment no matter what the NGDPLT, it seems likely to me a lower NGDPLT encourages this conversion.

Leave a Reply