Janet Yellen on monetary offset

Benn Steil and Steve directed me to an interesting comment by Janet Yellen during her testimony:

. . . certainly that [austerity] has been a headwind on the economy and something we’ve tried to offset . . .

It’s around the 1:03 mark.  Just to be clear I’m not claiming she’s a market monetarist.  She went on to caution that:

.  .  . our tools to do so are not perfect.

All along I’ve made several arguments:

1.  The degree of offset will depend on how well Fed officials do their job.  It will be time-varying.

2.  What matters is the expected multiplier, not the actual multiplier.

3.  The expected multiplier will be zero if the Fed is doing its job.

4.  You need to think about fiscal policy as a systematic regime, not as unpredictable shifts in policy.

5.  Fed officials will tend to deny monetary offset, partly because it sounds like sabotage (when policy is becoming more expansionary) and partly due to cognitive illusions.  Monetary offset doesn’t seem like monetary offset.  You might well get a different answer from a Fed official depending on how you asked the question:

a.  If Congress did more would you react by doing less, offsetting the stimulus effect?

b.  If Congress enacted austerity would you do your best to offset the drag on the economy?

These are identical questions, but I’d wager that many Fed officials would answer “no” to the first and “yes” to the second.

PS.  Kudos to Senator Mark Warner.  Right after he asked the question (framed as his support of monetary offset), he asked a question about cutting IOR to give banks an incentive to move the money out into the economy, citing Denmark. Yellen was noncommittal.

I often hammer Congressmen for their inept questions, so Warner deserves praise for two great questions.

PPS.  I have a post over at Free Exchange



23 Responses to “Janet Yellen on monetary offset”

  1. Gravatar of Ashok Rao Ashok Rao
    15. November 2013 at 12:41

    There needs to be more discussion about penalties on excess reserves. At this point the monetary base is so much larger than what markets expect it will be. Rather, the Fed needs to signal it is cool with a permanently larger price level. Penalty on excess reserves best way to do this.

    There are only two points against this that I hear mostly from the Alphaville team. Whatever instability it causes as interest-bearing reserves are substitutes for short-term treasuries can be offset by a complimentary sale or taper of Treasuries. If the Fed believes a certain level of liquidity is important for a banking system it can require that level, pay a small bonus after, and start charging penalties still after.

    Paul Krugman writes that the 0.25% on excess reserves is too small to do anything. I agree. But the zero lower bound does not bind reserves. So the question is not 0.25% – 0.0%. It is 0.25% – (-n%). And suddenly this policy seems a lot more contractionary than it is.

  2. Gravatar of Mark A. Sadowski Mark A. Sadowski
    15. November 2013 at 12:57

    Off Topic.

    This is in response to a comment by PeterP on Granger causality tests involving the monetary base.


    “Sadowski’s stuff is of course nonsense. If you have a 100% jump in the monetary base in late 2008 then any test will show that it “Granger caused” anything that came afterwards. So you can say: “the monetary base Granger causes the iPad”. He is either mathematically illiterate or thinks the Sumner readers are stupid (not a bad assumption). Of course, such an abrupt event is not a good moment to check causality of unique events that came afterwards.”

    A time series X is said to Granger-cause Y if it can be shown that lagged values of X collectively provide statistically significant information about future values of Y. There is simply no coherent way that values of X immediately preceding the range of observations could cause the lagged values of X in the range of observations to spuriously provide statistical information about future values of Y.

    There are lots of things that the US monetary base does not Granger cause over the period since 2008, many of them things that people claim that QE has had an effect on. In particular the monetary base does not Granger cause the price of gold, copper, crude oil, or emerging market equities. To claim that simply because there was a near doubling in the monetary base between August and December 2008 that this would lead to spuriously positive Granger causality results with anything that occurred after it is quite frankly silly.

    “Funny enough, if he dared to look at inflation, he would conclude that the monetary base expansion Granger causes… deflation, it happened in Sweden and Japan:



    The example of Sweden during the 1990s is not very relevant for a couple of reasons. First of all it wasn’t QE in the conventional sense of the central bank making large scale asset purchases. It is better described as Credit Easing (CE). The Swedish Central Bank lent heavily to the financial sector due to the 1990s Swedish financial crisis. Secondly the Riksbank repo rate was extremely high throughout the period of monetary base expansion (1994-1997) reaching 8.9% from May 1995 through October 1996:


    This was done mainly in order to defend the krona’s exchange rate. Sweden in 1994-1997 is simply not an example of a central bank doing QE because it was at the zero lower bound in interest rates.

    In the case of Japan there is in fact published research showing that reserve balances Granger caused core CPI during their original QE from 2001-2006. See Table 2 on Page 29:


    “People looked at Granger causation from the base, there is none:”



    Carpenter et al’s Granger causality results can be found in Table 1. What they actually show is that reserve balances Granger causes required reserves, reservable deposits and large time deposits, but not loans and managed liabilities. The fact that reserve balances didn’t Granger cause loans isn’t surprising since the time period in question is 1990-2007, which completely precedes the US QE.

    Komaromi’s results can be found in Table 1 on Page 36. What they show is that the monetary base does not Granger cause M1 or M2, but M1 and M2 Granger cause the monetary base. But again, the period in question is 1998-2007 which completely precedes the US QE.

    It’s not surprising that the monetary base does not Granger cause broad money during a period of time when the main policy instrument is the fed funds rate, something which I implicitly referred to in several comments in the Money Illusion comment thread to which these comments are replying.


    Now, let me comment on the econometrics employed in these two Granger causality results.

    Carpenter et al. did their test on logged and first differenced data, which evidently was done to address the issue of nonstationarity. In the words of the master econometrician Dave Giles, “this is wrong”. Granger causality tests should be done on level data otherwise the results will be inconsistent. This is in fact addressed by James Hamilton in “Time Series” on pages 651-3. I’m very surprised that nobody has commented on the fact that Carpenter et al’s Granger causality results are inconsistent before now.

    Komaromi did his test on level data and first-differenced data. He reports the results for the level data and states that the results for the first-differenced data are similar. But there is no evidence that the issue of nonstationarity was dealt with in the published results for the level data. Thus it is very likely that they are also statistically biased.

    All of my Granger causality tests have been done in levels using a VAR technique developed by Toda and Yamamato that addresses the issue of nonstationarity by adding additional lagged terms as exogenous variables. Thus my results are not subject to the same kinds of econometric problems that the Granger causality results contained in Carpenter et al and Komaromi almost certainly have.

  3. Gravatar of Wawawa Wawawa
    15. November 2013 at 13:15

    I wonder how many people are directed to this blog by your “mass media” pieces. Do you ever analyze hit counts after pieces such a this? Claims like “economic theory does not predict that lower interest rates will encourage more investment, just as economic theory does not predict that low prices will be associated with higher consumption” and “monetary policy has been highly contractionary since 2008” are sure to throw people for a loop and result in finding the blog look for explainations to reconcile conventional wisdom. Anyways, nice article.

  4. Gravatar of jknarr jknarr
    15. November 2013 at 13:31

    Scott, on the IOR – money into the economy question, I do continue to wonder about the vector between reserve formation and economic activity if rates are at zero and lending supply/demand is muted.

    Expectations are not enough.

    For instance, diamonds are fairly common in the earth’s crust — I even expect that they will remain this way. Yet DeBeers locks most of the diamonds up, keeping them out of circulation. The price of diamonds remains high because supply is limited in practice, not theory – the vector is missing.

    The Fed system operates reserves like a monopoly, similar to DeBeers. We know that reserves are cheap, but the money that you or I can use is simply not in circulation. For example, there’s been a new mine discovered — reserves — but it remains all locked up. So the “price” of money does not fall/NGDP rise.

  5. Gravatar of Tom M. Tom M.
    15. November 2013 at 14:36

    I think the problem for non-economists like me is that it seems a lot easier to see how fiscal policy “works”.

    The economy is running below potential with a lot of unemployment so you deficit spend to weatherize homes and businesses which puts people to work immediately and has the added benefit of increasing energy efficiency for the long term. That’s fiscal policy and is straight forward and easy to understand.

    On the other hand, you have Quantitative Easing which involves a central bank buying assets like mortgage backed securities. In layman’s terms, how is that supposed to reduce unemployment and bring unemployment down? Explain it so the average Joe can understand it.

  6. Gravatar of Tom M. Tom M.
    15. November 2013 at 14:42

    I meant to write how is that supposed to reduce unemployment and bring the economy back to full potential? Fridays…

  7. Gravatar of ssumner ssumner
    15. November 2013 at 16:55

    Ashok, Krugman is wrong about the IOR. You can’t judge it like an ordinary shift in the fed funds rate.

    The increase in RRs in 1937 only raised rates by 25 basis points.

    Wawawa, I don’t think there is a big impact on the number of hits.

    jknarr, That analogy only gets you so far. Money is much more complex than diamonds.

    Tom, You asked:

    “Explain it so the average Joe can understand it.”

    I’m afraid that’s impossible. There are some things the average Joe simply cannot understand. One is Heisenberg Uncertainty, another is the transmission mechanism for money at the zero bound. They are both too difficult.

    For above average people I’d recommend my short course on money, in the right column.

    For the average Joe the best you can do is explain that QE increases spending on goods and services, which boosts employment. But he’ll never be able to understand why QE increases spending on goods and services.

  8. Gravatar of lxdr1f7 lxdr1f7
    15. November 2013 at 18:10

    “What matters is the expected multiplier”

    Would transferring all the excess reserves into peoples accounts increase the expected multiplier? Isn’t the transmission mechanism causing the multiplier to decline?

  9. Gravatar of benjamin cole benjamin cole
    15. November 2013 at 19:05

    Sen. Mark Warner is an interesting story…arrived at the Senate a lifetime ago regarded as featherweight who had married well…butt of jokes…got serious good record…and now asking the right question about IOR…a rather esoteric topic…

  10. Gravatar of Ralph Musgrave Ralph Musgrave
    16. November 2013 at 01:00

    “What matters is the expected multiplier, not the actual multiplier.” What, so households and firms do detailed calculations as to what they think the multiplier is? This is straight out of la-la land.

    95% of households and entrepreneurs haven’t the faintest idea what the multiplier is.

  11. Gravatar of Michael Michael
    16. November 2013 at 05:32

    Ralph wrote:

    “95% of households and entrepreneurs haven’t the faintest idea what the multiplier is.”

    Irrelevant. 95% of households understand layoffs vs. job growth, strong vs. weak sales, availability of credit and expectations of all of the above. That’s all that they need to understand and have expectations about.

  12. Gravatar of benjamin cole benjamin cole
    16. November 2013 at 06:14

    Charlie Plosser rhapsodizing about deflation and saying the Fed should have single mandate, inflation-fighting of course…see blogs Historinhas and The Grumpy Economist…it is like Japan and the last five years never happened…when did macroeconomic policymaking become a subset of the inflation-fixated and tight-money cults?

  13. Gravatar of TravisV TravisV
    16. November 2013 at 07:05

    Prof. Sumner,

    In a recent post, you wrote: “food additives are pretty safe.”


    But what do you think about trans fats?


  14. Gravatar of Geoff Geoff
    16. November 2013 at 07:32

    “For the average Joe the best you can do is explain that QE increases spending on goods and services, which boosts employment. But he’ll never be able to understand why QE increases spending on goods and services.”

    No wonder the “average Joe” could know more. He’s being “educated” by those who are making erroneous claims such as the one above. Sumner is the average Joe.

    The argument that inflation raises spending on final goods and services is one thing. But it is a non sequitur to assert that this “boost employment”. At the individual firm level, yes, this argument is true, but that’s not the context under discussion here. The context here is the economy.

    When transitioning the context from the individual firm to the economy, there is an “offset” mechanism at work that requires us to reverse the order of argument. In the economy as a whole, final spending on goods and services comes at a trade-off to spending on labor. Indeed, every type of spending, for example spending on food, comes with the trade-off of a lowered spending on every other thing that is not food.

    “Demand for commodities is not demand for labor” – John Stuart Mill.

    This very useful lesson seems to continue to be misunderstood on this blog. What it says is that in the aggregate, none of us are helping the unemployed by spending money on our own consumption, or spending money on acquiring capital goods for ourselves. We only “benefit” the labor class by purchasing labor.

    Of course what is almost certainly imagined by Sumner in the claim that increasing spending on final goods “boosts” employment”, is the notion that after sellers of final goods receive higher revenues, they will then turn around and use a portion of those additional funds to spend more on labor. Yet this isn’t in fact a claim that higher spending on final goods “boosts employment”. It is a claim that higher spending on labor boost employment. (Of course even that is not accurate, because a higher spending on labor might be accompanied by a higher price of labor, and thus no change to labor hours purchased, and thus no “boost to employment”).

    So why not advocate for more spending on labor directly, which is the only way to actually “boost employment” if it does this at all? Is it because there is no reason for employers to do this unless they expect higher future revenues? But then why not advocate for employers to pay a lower price for labor, and why not advocate for the laboring class to accept lower prices? Surely with a low enough price of labor, the existing demand for labor would tend to clear the market.

    Sticky wages is it? And here is where the whole market monetarist world collapses. What they are advocating is exacerbating wage stickiness. Surely higher spending on final goods cannot be presented as the cure given sticky wages. That would be like presenting alcohol as the cure for headaches given hangovers continue to take place.

  15. Gravatar of Mike Sax Mike Sax
    16. November 2013 at 08:04

    “The argument that inflation raises spending on final goods and services is one thing. But it is a non sequitur to assert that this “boost employment”. At the individual firm level, yes, this argument is true, but that’s not the context under discussion here. The context here is the economy.”

    Geoff you believe in the macroeconomy? I thought you don’t believe in aggregates and that all that matters is the individual level. As a microeconomis you have a focus beyond the individual firm, consumer, or worker?

  16. Gravatar of paul Einzig paul Einzig
    16. November 2013 at 09:19

    Off topic:

    Scott, a counterfactual:

    Fed chairperson is an hereditary title and you are tutor to the 15 yr old Dauphin. What is he reading now?

  17. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    16. November 2013 at 09:41

    Kevin Erdmann has a pretty funny takedown of Andrew Huszar’s recent WSJ piece of self-promotion;


    You would think that Mr. Huszar would have been familiar with these spreads when he was managing the Federal Reserve’s $1.25 trillion agency mortgage-backed security purchase program. But, if you picture the Fed as the Interest Rate Wizard of Oz, it must be difficult to construct a narrative about interest rates when they regularly move in the opposite position from the Fed’s claimed target.

    Mr. Huszar is not the first person to float this idea that the Fed has been flooding the banks with cash, and that the banks have been pocketing the cash instead of performing their patriotic duty by loaning it out to good Americans. But, leaving empirical evidence aside, how do these people think the world works?

    Have the banks always had this monopoly power over mortgage profits? Why did they choose now to pocket extra cash?

    There must be 50 banks within 50 miles of my house that I could shop for a mortgage. Do these people think the local Credit Union is somehow part of a shadowy cabal of tuxedoed fat-cats secretly setting excessive mortgage rates? Are banks run by underwear gnomes who are raking in billions by NOT issuing mortgages, even while the Fed begs them to?

  18. Gravatar of Mark A. Sadowski Mark A. Sadowski
    16. November 2013 at 10:56

    Patrick R. Sullivan,
    That was very amusing. And just to piggyback on that, I came across the following:


    “…Here is Mr. Huzsar on CNBC, where he appears to backtrack a bit. And for good reason: Melissa Lee dismantles him immediately with facts that contradict his argument. Most of his discourse is a babbling brook of incoherence. What is the man saying? What is his point?…”


    I rarely encourage people to watch videos, but this is a must see.

  19. Gravatar of Jon Jon
    16. November 2013 at 11:49

    Does the zlb matter?

    If the fed buys tbills above par their interest rate will be negative. This is true even if the fed buys in the secondary market as the primary dealers will willingly pay above par to sell to the greater fool…

    This mechanism seems quite powerful because the Feds injection will be permanent.

  20. Gravatar of Geoff Geoff
    16. November 2013 at 12:27

    Mike Sax:

    “Geoff you believe in the macroeconomy? I thought you don’t believe in aggregates and that all that matters is the individual level. As a microeconomis you have a focus beyond the individual firm, consumer, or worker?”

    No, it’s just taking all individual firms together at once in forming arguments about the individual form in the abstract that purportedly show something true about all individuals firms taken together.

    Yes, all that matters is the individual level. But that is NOT the same thing as saying that in every analysis we can only consider one single individual firm and ignore all the rest. On the contrary, we can take into account more than one individual firm, recognizing that when more than one firm is considered, our arguments about individual firms changes.

    This is not the same thing as saying that aggregate concepts such as “aggregate demand” have any reality apart from the individual’s spending.

  21. Gravatar of ssumner ssumner
    16. November 2013 at 17:21

    lxdr, I don’t understand the question.

    Ralph, The only thing out of la la land is your understanding of macro. I made no assumptions at all about the public’s understanding of the multiplier. I was talking about the policymaker’s expectation.

    Travis, No opinion.

    Paul, Perhaps Irving Fisher and the other interwar economists.

    Patrick and Mark, Those are good takedowns.

    Jon, I don’t understand the question.

  22. Gravatar of Ray Lopez Ray Lopez
    13. January 2015 at 04:30

    From the Free Exchange Economist comment section, this makes sense: “Yes, this is just an exercise sophistry and jargon – he [Sumner] defines ‘tight money’ as any circumstance that produces both low growth and low reported inflation, no matter how much money has been created and how bloated a CB’s balance sheet has become. By this reasoning, there’s nothing but lack of money printing that keeps us all from being prosperous all the time. Academics really are irresponsibly dangerous, aren’t they? Does this piece actually have anything useful to offer?”

  23. Gravatar of Ben J Ben J
    13. January 2015 at 06:06

    Glad to see your terrible reading comprehension transcends both space and time Ray

Leave a Reply