August 22

One of the salient qualities of the internet is the short attention span. Huge amounts of information in small bites abound, but a sustained argument is rare. I am going to see if I can maintain focus on a single question over an extended period.

As with my other projects this one will need an armature to provide a structure. So I am going to attempt a through examination of the question: "What is the supply of dollars?"

I choose this question for a number of reasons. In the first place, there is some basis for thinking that this question does have a quantitative answer. Second, it is a question whose answer has some real consequences and the impact of these consequences is not limited to citizens of the United States. And, lastly, because finding the answer will be a non-trivial exercise.

I say that the exercise is non-trivial because I have already taken a pass at it and came away empty-handed. First there is a problem of definition. I expressed the question as "what is the supply of dollars" for a reason. The usual term,"money supply" implies a series of subsidiary definitions: M1, M2, ...M(n) which I may not want to adopt. I chose "dollar" instead of "currency" because I think that is the more interesting question. I will, however, start by examining how much currency is in circulation because it appears to be a simpler question. At the least a one dollar bill is definitely a dollar. One hundred pennies is likewise a dollar. It should be possible to determine how many of these items have been manufactured by the US Treasury, less the amount physically destroyed and arrive at a total. Even this simple first step has a few questionable aspects: can counterfeiting be considered negligible? Within the US, I would think so but elsewhere, I am not as sure. The United States issues a limited amount of gold and silver coins that are sold at prices based on the value of the metal in them. Are these currency? Is there enough of them to matter?

That's a start. Stay tuned.

August 23

Yesterday I said I was going maintain focus on a single topic and I set out to determine how much US currency is in circulation.

The Federal Reserve publishes some data pages which gives: "Currency outside U.S. Treasury, Federal Reserve Banks and the vaults of depository institutions." as being $775.9 billion for the week ending Aug 11, 2008.
That's a nice number. Is there any way to audit it?

The Federal Reserve of San Francisco has site with a chart that gives a total of 783.1 billion dollars in paper currency in circulation as of Dec 31 2006.
The main Federal Reserve site cited above gives Dec 2006 currency as 748.9 billion. Since this would include coinage, it should be larger than the figures given for paper currency alone. But it's smaller and the difference is 34.2 billion or about 5%. Since one is a USA Today type chart and the other is an official publication of the Federal Reserve I would assume that the 748.9 billion is the more trustworthy but it does raise the issue of how accurate these numbers are.

What I really want to find are production stats for the Treasury's printing presses since this is a number which can be known with very good accuracy. I am sure that someone at the Treasury knows exactly how much currency was printed last week. Inquiring minds want to know.

August 24

The story so far: we are trying maintain focus on a single topic for an extended period. The chosen topic is: how much US currency is in circulation?

I found exactly what I want. Sort of, For coins,annual totals for three years. Not a real big data set but her it is. This is from the us mint's web page where they give figures for "Circulating Coin Production (in millions) January through December"
The raw figures are below. Since they give the figures in number of coins, I have to convert each into dollars which, this being Sunday, I'll do tomorrow.
2007 total: 14440.65 M
$.01:7401.20 M $.05:1197.84 M $.10:2089.50 M $.25:2796.64 M $.50:4.80 M $1.00:950.67M
2008 total: 5877.16 M $.01:3098.80 M$.05:330.96 M $.10:530.00 M$.25:1554.80 M $.5:03.40 M$1.00:359.2 M
2006 total:15517.50 M
8234.00 M 1502.40 M 2828.00 M 2941.00 M 4.40 M 7.70 M
August 25
Here's the data from yesterday in cleaner form. The raw data is here

 
year 2006   	
8234.00 M pennies =       $82,340,000  in pennies
1502.40 M nickels =       $75,120,000  in nickels
2828.00 M dimes =        $282,800,000  in dimes
2941.00 M quarters =     $735,250,000  in quarters
   4.40 M half dollars =   $2,200,000  in half dollars
   7.70 M dollars =        $7,700,000  in dollars

             total = $1,185,410,000

 
year 2007   	
7401.20 M pennies =      $74,012,000 in pennies
1197.84 M nickels =     $ 59,892,000 in nickels
2089.50 M dimes =      $ 208,950,000 in dimes
2796.64 M quarters =   $ 699,160,000 in quarters
   4.80 M half dollars =  $2,400,000 in half dollars
 950.67 M dollars =     $950,670,000 in dollars

             total = $1,995,084,000

 
year 2008   	
3098.80 M pennies =      30,988,000 in pennies
 330.96 M nickels =      16,548,000 in nickels
 530.00 M dimes =        53,000,000 in dimes
1554.80 M quarters =    388,700,000 in quarters
   3.40 M half dollars =  1,700,000 in half dollars
  359.2 M dollars =     359,200,000 in dollars

             total = $850,136,000

The federal reserve gives seasonally adjusted currency for comparable periods as follows: Jan 2006 729.4 billion Dec 2006 749.6 billion Jan 2007 750.3 billion Dec 2007 758.7 billion Jan 2008 757.8 billion Aug 2008 775.9 billion
August 26

Maybe it's time to explain a little about what I'm up to here.

It may seem that I am reinventing the wheel since there are enormous quantities of economic data available already. Specifically, since the treasury releases figures for currency in circulation from 1959 to the present, why bother?

Part of the answer is that I have a primitive mind and I like my data very raw. I understand that, in today's world, actual cash is a minor component of the flow of money and it is steadily becoming less relevant. On the other hand, physical currency can be measured with a fair degree of accuracy. The figures for coin production can be measured to an accuracy of better that one per cent.

The Fed's figures for currency in circulation on the other hand, have to have a number of assumptions and adjustments embedded in the final number. It is possible that all these adjustments result in a more accurate picture of the state of the world.

It is also possible that some of these assumptions are false, or even actively misleading. It would not be the first time.

Before I am finished I will take a close look at the Fed's currency figures and the more expansive definitions of money as well. But first I want as good a baseline as possible to start with. It is possible to know how much money is rolling off the Treasury's printing presses. After that, I may have to guess or start making assumptions about things but I would like to know what I can know first.

Or know that I can't know the simplest quantitative basics of the subject. That would be interesting too.

August 27

Okay I found a sample of what I am looking for in the Treasury Department's annual report where it says:
"During fiscal year 2007, the Bureau of Engraving and Printing (BEP) produced approximately 38 million notes a day with a face value of approximately $750 million."

So:
$750,000,000 x 260 working days per year = $195,000,000,000 for 2007.

Unfortunately these are approximate numbers. On the other hand, the same report states:
"Currency shipment discrepancies per million notes 0 .01%."
So while the precision is not all I could wish, the accuracy is very good.

I'm kind of pressed for time today so I'll do the citations for the above tomorrow.

August 28

Here's the citation for the currency production figures I quoted yesterday.

I think I've done enough googling to convince myself that I'm not going to answer my question very quickly this way. Time to try something else: Asking nicely.

Posted on the Bureau of Engraving and Printing Contact Us page:

Do you publish currency production figures? That is: how many $5 bills were printed in 2006. in 2005? How many $10s? etc. Thank You, Dave Driscoll

We'll see what that produces.

August 29

I've received no reply to yesterday's email to the BEP yet. It won't surprise me if I never do.

Meanwhile, it occurred to me that people who collect currency have an interest in knowing the rarity of an issue of currency and, therefore, the quantity originally produced.
I did a quick check around the web and couldn't find anything relevant. I suspect that the information, if available, is not free. So my next step is to check the libraries and book stores.
I'll let you know what I find.

August 31

I checked the local library and the local bookstores to see if I could find any publications connected to numismatics that had any data on paper currency production. Couldn't find anything useful. I did buy a copy of the Blackbook price guide to United States Paper money. That gave me some pointers that lead to lots of web links to various currency related sites. I checked some of them but still no joy. That lead me to foreign exchange trading and I checked there but it seems as though paper money is so minor a contributor to foreign exchange rates that no one cares about it.

I think it's time for me to shelve this line of inquiry. I'd still like to get the production numbers because:
They are something that can be known with a high degree of precision.
It would be difficult to fake.
And they should have a fairly stable relationship to the broader supply of money.

But there comes a time for even the most desperately obsessive paranoiac to recognize that they are trying to drive a dead horse down a blind alley.

Tomorrow I will summarize what I have so far and move on to the aggregate currency data.

September 1

year	currency 1	paper	coin
2008 2	775,900,000,000	no data	850,136,000
2007	785,700,000,000	195,000,000,000 3	1,995,670,000
2006	749,600,000,000	783,100,000,000 4	1,185,410,000

Notes:
1. currency defined as:"Currency outside U.S. Treasury, Federal Reserve Banks and the vaults of depository institutions."
2. 2008 figures are year to date for mid-August.
3. My calculation is 750million per day x 260 working days per year. If I use 365 days instead, I get $273,750,000,000.
4. From the chart at the bottom of the SF Fed Reserve web site. My calculation is:9+1.5+10.5+16+119.2+62.8+564.1 billion dollars.
5. I have no data for coin or paper production prior to 2006.