Friday, December 14, 2012

A history of the pound sterling's medium-of-account

Shillings issued during Queen Elizabeth's reign

There are plenty of rumours that Mark Carney will implement some sort of NGDP targeting regime when he arrives at Threadneedle Street. If so, this will mark the seventh medium-of-account used to define the pound sterling since the pound's establishment in the early part of the last millennium. This storied list of media-of-account includes silver, silver/gold, gold, the US dollar, the Deutsche Mark, CPI, and perhaps NGDP.

First, some definitions. The pound sterling is a unit-of-account. Think of it as a word, a unit, or a brand name. The unit-of-account is generally defined in terms of some other good. This other good is called the medium-of-account. Some quantity x of the medium-of-account equals the unit-of-account. (See this older discussion of the definition of the word medium-of-account.)

1. Silver

The pound's first medium-of-account was silver.  A pound sterling was defined as 5,400 grains of 92.5% fine silver. We don't use grain measurements much these days, but a grain was legally defined as the weight of a grain seed from the middle of an ear of barley. So whatever weight of silver equated to 5,400 grain seeds defined the pound sterling.

The pound's 5,400 grains of silver was subdivided into a smaller unit of account, the shilling. Twenty shillings made a pound, each shilling equal to 270 grains. Over the centuries, monarchs redefined the unit-of-account by increasing the amount of shillings in each pound. For instance, Henry V divided the pound unit into 30 shillings, not 20, while Henry VII increased the amount of shillings in a pound to 40. This allowed the monarchy to issue more shilling coins from the same 5,400 grains of silver. By Queen Elizabeth I's time, a shilling only had 93 grains of silver, down from 270 grains a few centuries before. This meant that instead of coining just 20 shillings from 5,400 grains of silver, Elizabeth could issue 62 shillings from that amount.

2. Silver & gold - bimetallism

Gold coins called "guineas" were issued in 1663. Each guinea containing 118.6 grains of pure gold. While the value of the guinea was allowed to float relative to silver, this policy changed in 1696 when the monarchy declared one guinea equal to 22 shillings. Since 5,400 grains of silver was defined as 62 shillings, and 22 shillings was now defined as 118.6 grains of gold, the shilling was now dually-defined. Enter bimetallism, a system in which both silver and gold were the medium-of-account. The dual definition would be slightly modified by Sir Isaac Newton so that a guinea equaled 21.5 shilling in 1698 an 21 shillings in 1717.

3. Gold

In 1816, the bimetallic standard officially ended. The pound continued to be defined in terms of a quantity of gold grains and silver grains, but silver was confined to serving as the medium-of-account on payments below two pounds. For all practical purposes, gold had taken over the task of serving as the pound's medium-of-account. From 1816 to 1931, the pound would be defined as 113 grains of pure gold.

4. US dollar

After going off the gold standard in 1931, the pound had no publicly-disclosed medium-of-account until 1940, when it was redefined as US$4.03. While the USD served faithfully as the pound's medium-of-account, the specific amount of USD used in this definition changed three times over the next decades. In 1949 the pound dropped to $2.80 and in 1967 to $2.40. After Nixon closed the gold window in 1971, the ensuing Smithsonian Agreement redefined the pound upwards to $2.6057. This definition would only last for a few months when in June 1972 it became impossible to defend that rate. The dollar ceased to be the pound's medium-of-account.

5. Deutsche mark

In October 1990, John Major entered the European Exchange Rate Mechanism by defining the pound as 2.95 deutsche marks. As a result, the deutsche mark was now the pound's medium-of-account. The Bank of England was allowed to let the pound diverge from this underlying definition by a band of +/-6.5%, but the pound fell out of this band in September 1992 due to massive speculation by the likes of George Soros.

6. CPI

Since September 1992, the pound unit-of-account has been defined in terms of the consumer price index (CPI). In short, the medium-of-account is now the CPI basket, and an ever-shrinking basket at that. For the first few years, the pound was defined such that it bought a basket that declined in size by 1-4% each year. After 1997, the rate of decline was made more precise, 2.5% each year. The Bank of England is held accountable should the pound-denominated liabilities it issues fail to fall in the line with the ever shrinking medium-of-account.

7. NGDP?

If a switch is made to NGDP targeting, then the pound's medium-of-account will be updated from a variably-sized CPI basket to a varying NGDP basket. A pound sterling will be equal to a trillionth (or so) of UK nominal output. One could do so even more formally by adopting an NGDP futures market. Here is Scott Sumner describing such a scheme as "analogous to a gold standard regime, but with NGDP futures contracts replacing a fixed weight of gold as the medium of account."

You'll notice there are plenty of large gaps in the above history where the pound had either no public definition or was undefined altogether. Perhaps it's not necessary to always have a medium-of-account. Changes in the medium-of-account tend to be acrimonious and attract intense public attention. The bimetallism debates defined the 1896 US election, as evinced by the famous cross of gold speech. The drive to adopt NGDP as a medium-of-account seems no less controversial, at least if the debate  in the blogosphere is any sign.


  1. Nice post. The pedant in me feels obliged to point out that the RPIX index was the nominal anchor between 1992 and 2003, a slightly different index to the CPI
    used from 2003.

    I find the "gaps" very interesting, particularly the 1980s. There was even a nice hint that HM Treasury considered NGDP as a nominal anchor in the late 80s, see third PDF page 11 here:

    1. Excellent. This post is all about being pedantic so I'm happy that you're joining in. [].

      The gaps interest me too. Is it fair to say that there was no medium-of-account during these gaps, or is it simply that the MOA wasn't made public? You give the example of a shadow NGDP target in the 1980s. Perhaps certain high level bureaucrats had an idea in their heads for how the unit was to be defined, and the rest of us were all left guessing?

      This article is interesting too... it talks about the gap between 1931-1940. Speculators tried to ferret out the medium of account by piecing together the EEA's short term actions.

  2. Nіce poѕt. I learn something totаlly new аnd challenging on
    ѕites I stumbleupon everyԁay. It's always useful to read through content from other authors and use something from their sites.

    Also visit my webpage: payday loans