Resources > NumisWiki

New NumisWiki Article: Roman Coins - How Many Were Made?

(1/2) > >>

Joe Sermarini:
New NumisWiki Article: Roman| |Coins - How Many Were Made?
by Ken Pilon, November 27, 2020

Roman| |Coins - How Many Were Made?

A few comments:

I believe the estimate that 150 coins were struck per day per pair of dies is far too low.  

Because Roman coins were a store of value, many circulated far longer than 30 years. That average may apply to modern coins but not ancient coins.

Some Roman coins were demonetized after only a short period and did not circulate 30 years.

The base assumptions on coins per die per day  and die-life are pretty far off from most modern  research. 

Witschonke (2012) summed up the research as to the Republic era:  "various modern experiments replicating ancient conditions indicate that a striking rate of 4,000 to 5,000 coins per day is not unreasonable. [3] I have no independent view on the contentious question of the average number of coins struck by an ancient die, but if we accept a figure of c.20,000 coins per die, this gives us an average die life of 4-5 days, which agrees with Carter’s figure (1981,p. 193). ﬈us, taking a working year as 320 days (Buttrey 1976, p. 101), a single anvil would use up 65 to 80 dies per year, and strike 1.3 to 1.6 million coins."

In addition, since the main issue of money in the Republic was seemingly for legionary pay and expenses, and campaigns did not last nearly a full year, much of the coinage was needed in the first half of the year before standing armies became the norm.  So the rate of production had to be pretty high.   This also explains tolerance of so many poor strikes.

  These numbers also keep the mint organization/headcount relatively low as well, rather than the vast enterprise that would be needed at "150"

While an interesting thought experiment or starting point for discussion, almost every single 'fact" or assumption in this wiki article can be questioned and in fact should be questioned and expanded upon.

The question is do we treat this as a real wiki and start adding and/or editing, or treat it as an article by one author and therefore restrict our comments to a thread like this.

To take only one example: "Let’s assume the average coin lasted 30 years before it was lost or became too worn to use any more."  The problem is we can't make that assumption.  At all.  That logic might work with a relatively unchanging denominational system - for example US pennies that have basically been identical for over a century.  But in the Roman Empire hoard evidence shows that coin circulation came to abrupt ends when the denominational system changed significantly - based on metrological factors such as size, weight, fineness, etc.  So in, say AD 270 you have nothing but small base-metal antoniniani in circulation.  Almost nothing from before 265 was in circulation.  At other times 100 year old coins still circulated - say sestertii and denarii of Trajan during Septimius Severeu's reign.


Joe Sermarini:
I sent a notification of this thread to Ken. Hopefully he will comment.

The assumptions and numbers in the article seem wildly off base, as do the conclusions.

LRBs would have to factor heavily into any overall average stats, and number of dies for a given type can be huge. Try looking for a die match for a given type ... it can sometimes take years to find one for common types! If there were only 13 "die sets" per "variety", as the article suggests, then it would be unusual NOT to find a die match in a group of 5 coins ... (probabilities are counter-intuitive - you only need a group of 23 people to have a 50% chance of two of them sharing a birthday, despite there being 365 different birthdates).



[0] Message Index

[#] Next page

Go to full version