Our monetary system is becoming increasingly primitive

David Glasner has a good post on the recent Bank of England essay on money, but I’d like to respond to one of his comments:

What the authors mean by a “modern economy” is unclear, but presumably when they speak about the money created in a modern economy they are referring to the fact that the money held by the non-bank public has increasingly been held in the form of deposits rather than currency or coins (either tokens or precious metals). Thus, Scott Sumner’s complaint that the authors’ usage of “modern” flies in the face of the huge increase in the ratio of base money to broad money is off-target. The relevant ratio is that between currency and the stock of some measure of broad money held by the public, which is not the same as the ratio of base money to the stock of broad money.

Actually my claim also applies to currency.  Here is the data for 1929 and today:

January 1929: Currency = $3.828b  M1 = $26.109b  M2 = $55.119b

Today:  Currency = $1239b    M1 = $2793b    M2 = $11137b

I realize that “everyone knows” that the modern economy relies more on bank deposits and less on cash than in the old days, but I’m afraid that’s just one of the many things “everyone knows” that is not true.  Note that the C/M1 ratio would have increased even if you had only looked at currency held within the US (believed to be as much as half of the total currency in circulation.)  It’s unclear what would have happened to the C/M2 ratio.



25 Responses to “Our monetary system is becoming increasingly primitive”

  1. Gravatar of Dustin Dustin
    22. March 2014 at 10:34

    Is it possible that period in 1929 was anamolous and D Glasner is referring to a long term average trend? His stuff is very well thought out and deliberate, I’m inclined to think he wouldn’t make such an erroneous claim – perhaps misstated. I’ll have do some querying in FRED later.

    And might ‘modern’ mean a break from gold/silver backed currencies?

  2. Gravatar of ssumner ssumner
    22. March 2014 at 11:09

    Dustin, David didn’t say my argument would be wrong if I had used currency data, he said currency data is what I should have used. Currency use increased sharply in the 1930s and 1940s, then fell sharply for many decades, and has been rising again in recent decades (as a share of GDP.) Even in 1914 the ratios were lower than today, albeit not as dramatically so as in 1929.

  3. Gravatar of Dustin Dustin
    22. March 2014 at 11:55

    Ok ok. I had to re-read this and his post a couple of times as the meaning terms aren’t immediately clear to me. So I understand:

    base money = currency + reserves

    You used base money in the numerator and David says you should’ve used currency.

    And if I understand correctly, he says this because the ‘reserve’ component of the monetary base is not present in M1/2/3 denominator. Therefore when using reserves in the numerator, the direct ratio of currency-to-deposits cannot be isolated. If I am way off here, please don’t let this detract from your pleasant spring-ish New England Saturday.

  4. Gravatar of Major_Freedom Major_Freedom
    22. March 2014 at 12:13

    Don’t think the portion of the monetary base held at the Fed due to IOR should be counted in the ratio(s) of monetary base to monetary aggregate(s).

    The large increase in the ratio during the 1930s and 1940s was obviously due to inflation financing a big portion of WWII.

    Then it gradually declined until 1980, and then there was a period of a very small gradual increase until 2008, when there was a huge increase in the ratio due the Fed bailing out the banks.

    This is a rough estimate of monetary base to monetary aggregates dating back to 1929. I used personal incomes as a proxy for aggregate money since personal income data goes back to 1929 whereas M2 and M3 do not:


  5. Gravatar of ssumner ssumner
    22. March 2014 at 13:06

    Dustin, I think he just meant that currency doesn’t seem “modern” while reserves do.

  6. Gravatar of benjamin cole benjamin cole
    22. March 2014 at 15:27

    Fascinating post…and the amount of currency in circulation per capita is astounding…
    There might be a larger “underground” economy than suspected…certainly in Los Angeles I thought that could be true…
    What, if any, are the ramifications for monetary policy of such large amounts of cash in circulation?

  7. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. March 2014 at 16:53

    Here’s the currency multipliers (the money multiplier excluding reserve balances) from 1907 to present for various measures of money supply:


    M3 is in blue, M2 is in red, M1 is in green and MZM is in purple. Note that the more or less continuous measures from 1907 through the late 1960s are the Friedman and Schwartz series. The other continuous measures from 1959 to present are the Federal Reserve series.

    Currently the currency multipliers for MZM, M2 and M1 are about 10.0, 9.0 and 2.2 respectively.

    The M2 currency multiplier was higher from 1915-17, 1922-31, 1959-94 and from August 2007 through January 2011 (or roughly 50 years out of 107). The M1 currency multiplier was always higher prior to 1998 on data going back to 1914 (or roughly 84 years out of 100).

    The peak value for the M2 currency multiplier was 13.29 in October 1986. The peak value for the M1 currency multiplier was 7.38 in October 1929. The peak value for any currency multiplier was 17.12 for M3 in February 1987.

    If a low currency multiplier is considered primitive, then the monetary system is more primitive now than it was in 1915.

  8. Gravatar of dannyb2b dannyb2b
    22. March 2014 at 21:34

    I agree with MM on most stuff. Why cant the fed create base without removing treasuries or MBS from circulation? Why the need to interfere with another market to increase the base? Can’t the fed just expand the supply of base (when targeting NGDP or inflation) to the public and not concern itself with other assets?

    Wouldn’t it be beneficial to increase the functionality of base by allowing any entity to electronically deposit and transact in it like depository banks? If the fed can simply expand base straight to the public then there is no excessive dependance on banks for monetary policy to function effectively.

  9. Gravatar of ssumner ssumner
    23. March 2014 at 04:55

    Ben, It’s not really a problem for policy, as long as they accommodate the demand for hoarding.

    Mark, Thanks, that data pretty much confirms what I thought.

    Danny, I’m not quite sure what you are proposing. How would the Fed get the base money into circulation? And I don’t see how it distorts the Treasury market to do QE. It’s just the government buying back it’s debt. Would a budget surplus distort the market?

  10. Gravatar of dannyb2b dannyb2b
    23. March 2014 at 05:55


    “I’m not quite sure what you are proposing. How would the Fed get the base money into circulation?”

    Several posibilities. The fed can issue online reserve accounts to general public for example.

    “And I don’t see how it distorts the Treasury market to do QE. It’s just the government buying back it’s debt.”

    There needs to be a separation of powers or independence between the CB and the executive. Its a conflict of interest for the fed to purchase treasuries. If the fed needs to expand then just expand base.

  11. Gravatar of Dustin Dustin
    23. March 2014 at 06:53

    Thanks for that graph, Mark.

    The current multiplier for M2 and M3 appear to be quite close to the average of the past 100 yrs (though it has been hugely variable), while the multiplier for M1 has been much less variable and definitely trending downward since ~1930.

  12. Gravatar of David Glasner David Glasner
    23. March 2014 at 09:15

    Scott, Touche. However, I would point out — I did a post about this a long time ago, but I can’t remember the title or the date as I type this — that a huge amount of US currency is held abroad as the safest asset available to people who have no access — or are otherwise reluctant to make use of — the modern financial system. So although I accept your correction about what the numbers actually show, I don’t necessarily agree with your characterization of what the numbers signify.

  13. Gravatar of ssumner ssumner
    23. March 2014 at 14:18

    Danny, That doesn’t answer my question. What does “issue” mean? Give the money away?

    David, As I said, even accounting for cash overseas the C/M1 ratio has definitely risen. The C/M2 ratio is less clear.

    PS. My dissertation studied hoarding of currency.

  14. Gravatar of dannyb2b dannyb2b
    23. March 2014 at 14:35


    Yes the fed gives the money away while performing ngdplt or inflation targeting.

  15. Gravatar of dannyb2b dannyb2b
    23. March 2014 at 15:06


    I know that if the fed gives money away it has no mechanism to contract it. It doesn’t need to though because reducing base will adversely affect rgdp and ngdp because of sticky prices.

    The fed could just stabilize the supply of base if inflation is excessive. Btw there hasnt been any time when inflation is too high and the base is stable that I know of. If inflation is persisting with a stable money supply then the source of inflation is not monetary and the onus is on the government to address it with its toolbox (regulatory, investment, etc…)

  16. Gravatar of Major_Freedom Major_Freedom
    23. March 2014 at 15:49

    “What does “issue” mean? Give the money away?”

    According to the Dallas Fed President, “issue” currency means “gift.”

  17. Gravatar of David Glasner David Glasner
    23. March 2014 at 16:25

    I see. But why do you think that estimates of US currency held abroad are reliable? Given the motivations for holding currency, I would think that any estimates would be highly suspect. Anyway, if currency is being held mostly for the reasons I think it is, I don’t think it’s misleading to say that the monetary system is becoming more primitive based on the rising C/M ratio.

  18. Gravatar of Benjamin Cole Benjamin Cole
    23. March 2014 at 19:03

    OT but Not:

    Okay, so there is $3,400 in cash in circulation for every US citizen…You know, your typical family of four keeps $13k or so in the cupboards.

    Okay, so maybe half of this is overseas….

    Still, a question remains: If you asked a typical economist, what would it mean for the amount of fiat cash in circulation to quadruple into the thousands of dollars per capita, I think most would say “It would mean rampant inflation.”

    In part, that is because that is the answer that modern economists monomaniacally give to every question, but in part it would be a logical deduction.

    But we have not had more inflation, in fact we have lower inflation.

    What gives?

    Do we know less about what really causes inflation than we think?

  19. Gravatar of Major_Freedom Major_Freedom
    23. March 2014 at 19:23


    “Okay, so there is $3,400 in cash in circulation for every US citizen…You know, your typical family of four keeps $13k or so in the cupboards.

    Okay, so maybe half of this is overseas…”

    A good example of avoiding reasoning from means and averages.

    A few drug cartels can make up for 3 million homes not holding any cash in their cupboards and you’ll see the same average per household.

  20. Gravatar of Benjamin Cole Benjamin Cole
    23. March 2014 at 19:46


    The amount fiat money in circulation has tripled since 1996, to about $3,400 cash per resident in the United States.

    Prices are up about 40 percent in same time frame.

    Deduction: Radical increases in U.S. cash in circulation are not associated with radical increases in U.S. prices. It just ain’t happening.

    Major F: Yes, but then you must assume that drug cartels are absorbing all the new money, not a constant fraction of it.

    Why would drug cartels absorb more, or less, of fiat money from one decade to the next?

    Moreover, I do not think drug cartels are that big. Afghanistan is the globe’s dominant supplier of heroin (that happen under our tutelage, btw) Something glide 80 percent to 90 percent of heroin comes from Afghanie.

    But the crop is estimated at $60 billion or so. Seems like druggies and crooks would need some cash, but not $1.2 trillion….

    Maybe the Bureau of Engraving prints up hundreds of billions to give to the Mafia, the CIA, Russians and Star Chamber overlords?

  21. Gravatar of Mark A. Sadowski Mark A. Sadowski
    24. March 2014 at 04:57

    Benjamin Cole,
    Currency in circulation has risen from about 5.2% of GDP in 1996Q2 to 7.2% of GDP in 2013Q4:


    A lot of that increase can probably be explained by the decline in short term interest rates.

  22. Gravatar of TallDave TallDave
    24. March 2014 at 08:11

    Interesting, thanks for sharing.

  23. Gravatar of ssumner ssumner
    25. March 2014 at 05:37

    Danny, Given our large deficit it seems crazy to give money away. Why not use it to start paying off the national debt?

    David, Given that I did my dissertation on the subject, I think my own estimates are reliable! Seriously, I don’t think we know for sure, but we do have a ballpark estimate. We also know the C/NGDP ratios in other developed countries that probably do not have a lot of currency overseas.

  24. Gravatar of David Glasner David Glasner
    26. March 2014 at 09:35

    Scott, And how long has it been since you wrote your dissertation? So how do the C/NGDP ratios for other developed countries compare with the US ratio?

  25. Gravatar of Scott Sumner Scott Sumner
    26. March 2014 at 11:04

    David, 30 years.

    At that time the C/GDP ratios were higher than the US in lots of countries, including Germany and Japan, and lower in others, including Canada and Australia. The US rate was typical. That made me skeptical of how much of the mystery the “money overseas” could explain. As far as I know the global C/GDP ratio is surprisngly high, and unless Neil Armstrong brought cash to the moon, it’s all here on planet Earth.

Leave a Reply