Matt Yglesias explains the Fed

Matt Yglesias has a new post that explains the Fed in 20 questions.  He gets off to a good start:

1. What’s the Federal Reserve System? It’s America’s central bank.

But then starts to struggle a bit:

2. What’s a central bank?  A central bank is many things. But what makes it a “central bank” is this. When banks take deposits they lend most of that money back out. As a regulatory matter, however, they are not allowed to lend all of it out. They have to hold some of the money back as reserves. Those reserves are placed on deposit with the Fed. Hence “Federal Reserve.” It’s central in the sense that it stands at the center of the banking system; the bank where banks do their banking.

3. Why’s that important? It’s actually not that important.

Let’s look at bank reserves in late 1914, when the Fed was getting up and running:

Date               Vault Cash       Deposits at Fed           Reserves

Sept. 1914   $1.662 billion         $0              $1.662 billion 

Oct. 1914     $1.642 billion             $0                $1.642 billion

Nov.  1914   $1.452 billion     $0.228 billion        $1.680 billion

Dec. 1914    $1.352 billion     $0.244 billion        $1.595 billion

When the Fed was created it simply moved a small portion of reserves that were held in large denomination currency (up to $100,000 bills) into an account at the Fed.  This was done to make interbank transfers easier.  It reduces armored car robberies.  That’s all. It’s a big mistake to make bank deposits at the Fed the centerpiece of what makes a central bank.  Instead, here’s what Matt should have said:

2. What’s a central bank?  A central bank is many things. But what makes it a “central bank” is this. The central bank has a monopoly on the production of the monetary base, which is almost always the medium of account, or at least one of two media of account.

3. Why’s that important?  If it is the only medium of account (as under fiat money), then the monopoly on the production of base money, combined with the ability to produce base money at near zero cost, gives the Fed almost unlimited ability to influence the value of money.

4.  But why is that so important?  Because most wages, prices and debt contracts are denominated in money terms.  If you change the value of base money, it influences all other nominal values in the economy. Even more importantly, many wages and prices are sticky, so it also influences all sorts of real variables, such as real GDP and employment, at least in the short run.

Matt Yglesias continues:

5. So what’s monetary policy? That’s a bit of a metaphysical issue. In a sense, “monetary policy” is just a word for “stuff central banks do to try to keep growth on track and inflation in check” and doesn’t refer to anything in particular. In practice, there are two main things the Fed does to try to do monetary policy””it manipulates interest rates and it tries to shape expectations about the future.

6. How does the Fed manipulate interest rates? Traditionally the overwhelming focus of monetary policy has been moving the “federal funds rate” up and down. That’s the rate that’s been at-or-near zero since the crisis began.

This disappointed me.  Matt’s much brighter than I am, and has studied philosophy.  He’s very good at logic. But unless I’m mistaken, he never answers the question of HOW the central bank manipulates interest rates.  He seems to say the Fed manipulates interest rates by manipulating a specific interest rate. The fed funds rate is their target, but how do they hit the target?  It’s not an administered rate like the discount rate; it’s set in the interbank loan market, between private firms.  How does the Fed control this rate?

7. What’s the federal funds rate? It’s the interest rate at which banks can lend reserves to each other overnight to make up for short-term funding gaps.

8. Is that really so important? Sort of. In a sense this is the foundational interest rate upon which lots of other interest rates are based. A higher federal funds rate propagates through the economy resulting in higher mortgage rates, higher auto loan rates, higher small business rates, etc. reducing investment and durable goods purchases and slowing the economy down. In another sense, though, the federal funds rate is important because it’s a symbol that shapes expectations.

But a symbol of what? It seems to me that Matt is desperately trying to avoid getting down into the gutter with us market monetarists, and talking about the monetary base.  Smart macroeconomists don’t discuss the base; it’s too low class.  They talk about interest rates.  And if they are especially smart they talk about the future expected path of interest rates.  But we still don’t know how the Fed controls interest rates!

Here’s how I would answer question 8:

8. Is that really so important?  Not in and of itself.  It’s an epiphenomenon that occurs when the Fed adjusts the supply or demand for base money.  (I haven’t mentioned demand yet, but the Fed also influences that variable.)  However when the Fed changes the supply or demand for base money it does have a short run effect on the Fed funds rate. In addition, the Fed usually targets the fed funds rate. If you combine those two facts, changes in the fed funds rate target provide an indicator as to whether monetary policy is getting more expansionary or contractionary. But only in the very short run, because the long-run effect of monetary policy on interest rates is different from the short run effect.

If the Fed cuts the Fed funds rate target, it suggests that the Fed would like to see faster NGDP growth. To achieve faster growth, the Fed will increase the current and future expected level in the monetary base. This will cause the Fed funds rate to decline right now. The effect on longer-term interest rates is ambiguous. But regardless of whether longer-term interest rates rise or fall, the effect on nominal GDP growth is expansionary.

PS.  You might argue that bank deposits at the Fed were a small share of the base in 1914, but in modern central banking they are very important.  But that’s only true since 2008.  As late as 2007 bank deposits at the Fed were only about 1% of the monetary base, an even smaller share than in 1915 (when it was about 4%.)  The US banking system was more cash-oriented in 2007 than in 1915.  In 1915 we had a more “sophisticated” monetary base than in 2007.

There’s no “metaphysics” at all.  As late as 2007, “monetary policy” was 99% printing currency and minting coins.  All the rest was the “effects” of monetary policy.



37 Responses to “Matt Yglesias explains the Fed”

  1. Gravatar of Ryan Ryan
    10. October 2013 at 07:21

    What, no talk about cleaning up asset portfolios/collateral. Seems to me that has been the Feds primary objective since 2009ish.

  2. Gravatar of Gregor Bush Gregor Bush
    10. October 2013 at 07:40

    I understand why people who are heavily invested in the Woodwordian NK model want to expunge money from macroeconomic discourse. But I don’t understand why liberal non-economists share the same impluse for a radical interest-rate-centric understanding of monetary policy. Are they just following thier NK economist overlords? Or is there something about admitting that “money matters” that they instinctively feel undermines thier worldview?

  3. Gravatar of jknarr jknarr
    10. October 2013 at 08:23

    There’s a longstanding effort to keep the public bamboozled on the interest rate question. Lest we throw it down the memory hole, recall that Yglesias’s credibility is damaged as a message-coordinating left-wing JournoLister.

    Tight money is the health of the one-party state. The emphasis on interest rates is an attempt to keep monetary base policy on the sidelines, and to keep fiscal policy front-and-center: easy debt rather than easy base money.

    The Fed ought to be taught in US high school alongside the three branches of your government. It’s not, and this ought to tell you something. There’s a reason that Yglesias has a vacuum of ignorance to rush into and (poorly, with misdirection) define what the Fed is and does.

  4. Gravatar of Doug M Doug M
    10. October 2013 at 08:56

    Why is the Fed funds rate important.

    It is the “risk free rate.
    “Risk free” – 1 day term, 102% collateralized by Treasury securities.
    It is the banks cost of capital.
    The differential, or “spread”, between the banks cost of capital and the banks lending rates represents the banks economic incentives to lend.

    When a bank lends it creates money.

  5. Gravatar of Philippe Philippe
    10. October 2013 at 09:16

    “But we still don’t know how the Fed controls interest rates!”

    it also changes the interest it pays on reserves, without changing the quantity of reserves, right?

  6. Gravatar of JP Koning JP Koning
    10. October 2013 at 09:34

    Yglesias’s post is just awful.

    Scott: “As late as 2007, “monetary policy” was 99% printing currency and minting coins.”

    I agree with most of your post. But in the above quote you seem to want to diminish the important of reserves relative to cash. However, when the Fed conducts open market purchases, it creates reserves, not cash. Reserve injections, not cash injections, are what simultaneously push asset prices higher, the fed funds lower, and expand NGDP.

    That reserves are usually smaller in proportion to cash shouldn’t detract from their importance, for the same reason that the small size of the base relative to GDP doesn’t detract from the base’s importance.

  7. Gravatar of TravisV TravisV
    10. October 2013 at 09:55

    Prof. Sumner,

    Felix Salmon wrote this and I thought it was interesting. Any chance that it could happen?

    “What I’d like to see from Yellen is less of an attempt to artificially move markets by saying the right words at the right time, and more of an attempt to be honest and clear about the full range of opinions on the FOMC. Where Bernanke always just attempted to get across a single consensus view, Yellen should instead be more open about the full spectrum of opinions on the FOMC, and how that spectrum ultimately ended up being reduced to a consensus about what to do and say.

    We live in a world where both the legislative and the judicial branches of government are racked with very open dissent “” and yet where the Fed likes to pretend that it somehow manages to always rise above such things. It doesn’t; it can’t; it shouldn’t even really aspire to doing so. The most effective communication is honest communication…..”

  8. Gravatar of Saturos Saturos
    10. October 2013 at 10:27

    I’m sorry, but Scott is wrong as well. Even if we accept the view that the medium of account role of money is what’s really important for business cycles, there remains the fact that the medium of exchange is a necessary role for money which the economy depends on. And most of the exchange media we use depend for their existence on Federal Reserve deposits. As Nick Rowe says, “What makes a central bank? Asymmetric redeemability and the will to act as one.” (Now central banks may *matter* for recessions and inflation because their money is also the medium of account, but that’s a separate matter.)

  9. Gravatar of Saturos Saturos
    10. October 2013 at 10:28

    The media is determined to give Yellen a triple mandate:

  10. Gravatar of OhMy OhMy
    10. October 2013 at 11:11

    You struggle too:

    “3. Why’s that important? If it is the only medium of account (as under fiat money), then the monopoly on the production of base money, combined with the ability to produce base money at near zero cost, gives the Fed almost unlimited ability to influence the value of money.”

    That is false. The Fed swaps two government IOUs so it has an ability to set the relative price, but that is all. Central banks can change interest rates, not create inflation. For that you need to change the total number of govt IOUs and that is what the Treasury does by deficit spending.

  11. Gravatar of RyGuy Sanchez RyGuy Sanchez
    10. October 2013 at 11:47


    If inflation is created by changing the amount of govt IOUs, why has Japan experienced deflation since 1990? By your theory shouldn’t their massive amount of govt debt created some kind of lift off zero?

  12. Gravatar of Don Geddis Don Geddis
    10. October 2013 at 12:08

    OhMy: according to your theory, if a fiscal government always ran a balanced budget, then it would not be possible for that economy to experience inflation. Yet surely that’s obviously historically false. (And for that matter, there’s the other side like Japan, with decades of lots of increase in government debt, yet essentially no observed inflation.)

    Your theories don’t seem to match observed reality. Perhaps you might want to reconsider them?

  13. Gravatar of Jared Jared
    10. October 2013 at 12:18


    How would you respond to Sandra Krieger, the head of domestic reserve management and discount operations at the NY Fed, who says: “The conventional, textbook view is that the Trading Desk buys and sells securities in response to policy easings and tightenings. From the Desk’s perspective, however, the supply-demand balance is primarily a function of the demand for required balances, which is almost completely insensitive to small changes in policy. Consequently, any change in the Committee’s target has virtually no effect on excess supply or demand in the funds market”?

    It seems as if her account of monetary operations is contrary to yours. In your account, the Desk manipulates the amount of monetary base, which causes the fed funds rate to change. In her account, the Desk is always adjusting the monetary base in order to MAINTAIN the current rate. When the FOMC announces a new rate, the Desk continues to supply only the amount of base demanded (no more, no less), and this amount is “almost completely insenstive to small changes in policy”. Therefore, changes in the monetary base often yield no change in rates, and, conversely, changes in rates can occur without changes in the monetary base.

  14. Gravatar of jknarr jknarr
    10. October 2013 at 12:46

    Jared — Krieger is right under most fed funds scenarios, just as, most of the time open market operation excess reserves are immediately shunted into required reserves or currency.

    But, the volume of open market operations required at 5% fed funds is very different than that required at 1% or 0.25% (as is the willingness to hold excess reserves is different). Below see that as Fed Funds approaches zero, the monetary base as a proportion of NGDP accelerates nonlinearly.

  15. Gravatar of JP Koning JP Koning
    10. October 2013 at 13:00

    Jared, I don’t know what Scott’s reply is, but mine would be that the Fed needn’t actually inject more reserves to reduce the overnight rate — it need only threaten to. If the rate doesn’t move upon the Fed’s announcement, then it will carry through on its threat. The use of threats rather than mechanical reserve increases means that there is unlikely to be a clear inverse correlation between the quantity of base money and the overnight rate.

  16. Gravatar of Jared Jared
    10. October 2013 at 13:16

    jknarr – I agree with you, but isn’t the upshot of Krieger’s description that the increase in the base is the effect of the rate change, not the cause? Consitent with your link, at much lower rates more money/deposits are created, and therefore, more monetary base is demanded. The Trading Desk responds to this demand and supplies the needed base. I thought Scott was describing the cause and effect in reverse.

    Also, I’m not sure what you mean by “the willingness to hold excess reserves.” According to Krieger, in normal times the Desk only supplies the amount of reserves that are required, and drains any excess. Since 2008, the Desk has intentionally added an excess amount, with which banks don’t have a choice whether to hold. If the central bank doesn’t drain the excess, then banks are stuck with them.

  17. Gravatar of Jared Jared
    10. October 2013 at 13:26

    JP – That’s my understanding too. But see my reply to jknarr above. Are you guys saying that Krieger’s description is consistent with Scott’s, which I took to be the “textbook” view she was countering?

  18. Gravatar of jknarr jknarr
    10. October 2013 at 13:33

    Jared — I would not worry about funds supplied or demanded per se. They produce funds sufficient to target the rate, and somebody holds them at a given fed funds. The set level of fed funds is most of the time simply reflective of the economy’s underlying natural interest rate anyway.

    They only drain excess reserves to the extent that they need to keep fed funds at the level that they want. Banks never have a choice on the volume of absolute reserves: however, they can keep it in excess, required, (or currency) form. Right now, they choose excess, partially because of IOR.

  19. Gravatar of TravisV TravisV
    10. October 2013 at 14:05

    Wow, Krugman:

    “there are two fundamental truths here: Wall Street is largely responsible for the mess we’re in, and financial types have been consistently wrong “” not just failing to see the risks before the crisis, but in diagnosing what would come next. Above all, they took the position that bailing out the banks would pave the way to recovery more broadly, and it hasn’t.

    Meanwhile, sensible academic macroeconomics has, as I often point out, performed very well “” and Janet Yellen is very much in that camp.”

    You have got to be kidding me! The macroeconomics profession has been a gigantic disaster!!!!!

  20. Gravatar of Paul Paul
    10. October 2013 at 15:22

    Krugman was hedging his bets. He gave Yellen the mildest, barest possible endorsement over Summers. Now it turns out she’s the winner – she was his choice all along!

  21. Gravatar of TravisV TravisV
    10. October 2013 at 15:44


    “Stocks Just Had A Heart Attack”

    “Minutes ago, the New York Times reported that President Obama rejected the GOP’s latest offer to extend the debt ceiling.

    The story is developing…..”

  22. Gravatar of benjamin cole benjamin cole
    10. October 2013 at 16:42

    Gee, I thought the fact that a central bank can print money and buy assets was pretty important…and when an economy is at or near ZLB very important…the people who sell the assets then have immediate claim on output or can invest in other assets…this happens with or without forward guidance…

  23. Gravatar of ssumner ssumner
    10. October 2013 at 17:29

    Philippe, We do know how they set the IOR rate–it’s simply administered by the Fed. But the fed funds rate is not an administered rate.

    JP, The Fed doesn’t really create reserves or cash, they create base money. The market determines the split between reserves and cash. Just to clarify, prior to 2008 most bank reserves were vault cash, so I didn’t mean to suggest that reserves were only 1% of cash, rather deposits at the Fed were 1% of cash.

    Travis, He’s probably right.

    Saturos, You said;

    “And most of the exchange media we use depend for their existence on Federal Reserve deposits.”

    Nope. Prior to 1914 there was no such thing as deposits at the Fed, and the banking system did just fine in providing deposits. I’m fine with people claiming that bank deposits are really important. So is electricity. But the Fed doesn’t produce either, it produces the monetary base. Sure, you can’t have DDs without base money, but you can’t have electricity either.

    And your claim is not even true today, or at least in 2007. In 2007 most reserves were vault cash, not deposits at the Fed.

    OhMy, You said;

    “That is false. The Fed swaps two government IOUs so it has an ability to set the relative price, but that is all.”

    You don’t seem to realize that the price level is the relative price of base money. Also that base money is not an IOU.

    I currently have a $100 trillion dollar Zimbabwe currency note in my office. What does the Zimbabwe government owe me, since cash is an IOU?

    Jared, There is no conflict at all. When you target interest rates, the base is endogenous. I accept that. And changes in the reserve target do not usually require sudden shifts in the quantity of reserves. I accept that. But over the long run you must adjust the level of base money as required to hit the inflation target. I presume she would agree. And you do that by also adjusting the fed funds target. If the Fed tried to target inflation and interest rates without any change in the base, they would fail. But yes, in the ultra short run the Chuck Norris effect is usually enough.

    The trading desk doesn’t see the “big picture” and thus reverses cause and effect, as do many economists. It would look very different if the Fed announced targets in terms of the base.

  24. Gravatar of Reader223 Reader223
    10. October 2013 at 18:04

    Prof. Sumner,

    OT, but as an undergrad interested in your NGDP futures market I was wondering if you’d be able to write a post on how the futures market alone- that is, without the Fed using it to target NGDP- would be a net benefit to the economy. Seems pretty intuitive to me but it’d be nice to read a post that summarizes the pros and cons.

    If you’ve already done so, I’d appreciate it if you (or any other long time readers of this blog) could point me towards a link that gives an explanation. Thanks!

  25. Gravatar of JP Koning JP Koning
    10. October 2013 at 18:29

    Ok, scratch the word reserves. Although Fed deposits were just 1% of cash before 2008, I don’t think their importance should be downplayed. Open market purchases create deposits which banks simultaneously try to get rid, pushing up prices. I don’t see cash being an important part of the early transmission process. Only after prices have risen does cash start to play a role, since at a higher price level people demand higher cash balances. Banks reduce deposits at the Fed to accommodate this cash demand, but as long as the Fed compensates by creating a few more deposits via open market purchases then this shouldn’t have any macroeconomic effects.

    Do you think this is a good explanation of the process?

  26. Gravatar of rob rob
    11. October 2013 at 03:49

    Scott you really should work on a semi-popular audience book about monetary theory that clearly explains all of the mechanisms from start to finish. Most books especially those written for undergrads or popular audiences tend to get very squirrelly when they start talking about how the interest rate is influenced and how in turn that effects NGDP, especially in making it congruent with long run effects. To be honest I think it wouldn’t hurt a great deal of economists to read either.

  27. Gravatar of sdfc sdfc
    11. October 2013 at 04:25

    No. Banks create deposits. Rises in base money are a function of bank deposit creation.

  28. Gravatar of sdfc sdfc
    11. October 2013 at 04:25

    No Koning that is.

  29. Gravatar of Matt McOsker Matt McOsker
    11. October 2013 at 04:45

    Aside from adding to and draining reserves to hit the FFR, the Fed also can target via interest on reserves, which sets the FFR base target.

  30. Gravatar of Jared Jared
    11. October 2013 at 06:30

    Scott – Thanks for that explanation. I think I can now reconcile your view with Krieger’s (and my own). But could you say more about that last line of your’s: “It would look very different if the Fed announced targets in terms of the base.” I think you’re right, that would look very different, but isn’t that very much what Volcker attempted early on in his tenure. And wasn’t it a practical mess/failure?

    jknarr – You said: “[Banks] can keep [reserves] in excess, required, (or currency) form. Right now, they choose excess, partially because of IOR”. I don’t think that’s right. For example, let’s assume a 10% reserve requirement, if a bank has $1 million of reservable deposits and $100,000 of reserves, that means all of those reserves are required reserves. If the bank increases their reserves to $150,000, then $100,000 are required and $50,000 are excess. There’s no choice in the matter. Even if the bank decides to hold some of that in vault cash, that doesn’t change anything since vault cash is counted as reserves (required or excess). And the bank can’t force depositors to hold more cash since they can always, and easily, redeposit the cash back at the bank.

    I don’t see how IOR influences the amount of excess reserves at all. IOR simply helps set the FFR as long as the central bank has decided to add and keep excess reserves in the system.

  31. Gravatar of Jon Jon
    11. October 2013 at 07:07

    Something no one discusses much: the largest banks are toeing the legal deposit limit (not reserve ratio, percentage of all bank deposits).

    So the largest snd healthiest banks: JP Morgan, WF, BoA are legally bared from expanding faster than than the industry.

    What impact does this have on the monetary transmission mechanism?

  32. Gravatar of ssumner ssumner
    11. October 2013 at 08:16

    Reader, A NGDP futures market would provide a real time measure of changes in expected growth in AD. That would be invaluable to policymakers trying to figure out how various shocks are impacting the economy. For instance, what does fear of default do to AD? The NGDP futures market would tell us.

    JP, You said;

    “I don’t see cash being an important part of the early transmission process.”

    The very early transmission process concerns asset prices. But it seems to me that applies to both cash and reserves. Take the easiest example, short term interest rates. If the Fed boosts the base by 5 billion dollars, the banks and the public must somehow be induced to hold the extra money. In the very short run, before prices change, that inducement comes from lower interest rates, which is a lower op. cost of holding base money. I agree that the reserve demand is probably more sensitive to rate changes than cash, at least in the short run. But also note that there is far more cash (normally). So the extra willingness to hold cash as a result of lower opportunity cost is not a trivial factor. Indeed cash demand shot up after 2008, when rates hit zero.

    There is a sense in which cash in endogenous, but its the same sense in which the entire base is endogenous.

    rob, I agree, but it’s hard to find the time. But I’ll get it done eventually.

    Matt, Yes, that changes the demand for base money.

    Jared, Volcker didn’t really target the base, but I agree with your point. If he had done so, it would have been a mess, as you say.

    I simply meant that if the Fed used the base at a target, markets would start to see the base as a causal factor, much like they now see interest rates. Instead markets pay little attention to base movements, because they are endogenous in the short run (while interest rates are being targeted), and perhaps in the long run too (if inflation is being targeted.)

    Jon, I’m afraid I don’t know much about that. Hopefully someone else can explain it.

  33. Gravatar of JP Koning JP Koning
    11. October 2013 at 10:52

    Scott, that makes sense to me.

  34. Gravatar of Reader223 Reader223
    11. October 2013 at 17:21

    Awesome, thanks for the quick summary. I’m excited for your book- I need to see how on Earth something so useful hasn’t been made yet.

  35. Gravatar of Rajat Rajat
    12. October 2013 at 15:28

    “If the Fed cuts the Fed funds rate target…This will cause the Fed funds rate to decline right now… But regardless of whether longer-term interest rates rise or fall, the effect on nominal GDP growth is expansionary.”

    Is that right? What if the Fed cuts the FFR but by less than the market expected? Then long term rates would fall but that would reflect a tightening of policy, which would hurt NGDP growth, won’t it?

  36. Gravatar of Jon Jon
    15. October 2013 at 06:47

    Scott agree, but that appears to be due to mood affiliation–rather like Krugman and fiscal stimulus.

  37. Gravatar of ssumner ssumner
    16. October 2013 at 10:14

    Rajat, Yes, but I was referring to cuts larger than expected.

Leave a Reply