Q. David Bowers
In summary, here are the official Mint figures for calendar year coinage (again) and, in parentheses, my estimates of the approximate actual mintage of coins bearing those dates.
1794 Mint figure: 1,758 (delivered); Author's estimate: No change
1795 Mint figure: 203,033; Author's estimate: 390,000
1796 Mint figure: 72,290; Author's estimate: 75,000
1797 Mint figure: 7,776; Author's estimate: 60,000
1798 Mint figure: 327,536; Author's estimate: 235,000
1799 Mint figure: 423,515; Author's estimate: 395,000
1800 Mint figure: 220,920; Author's estimate: 100,000
1801 Mint figure: 54,454; Author's estimate: 35,000
1802 Mint figure: 41,650; Author's estimate: 80,000
1803 Mint figure: 85,634; Author's estimate: 60,000
The years 1797, 1798, and 1800 in particular are interesting relative to the use of dies in year(s) other than indicated on the dies.
While estimating the number of coins originally minted and/or distributed by the government is an interesting exercise, the figures obtained tell only part of the story concerning the availability of various early dollar dates today. The numismatic market has its own history, and is a valuable source of information.
Gathering useful data concerning the rarity of coins within the coin collecting fraternity is a task more difficult than it might seem at first. In each. of the succeeding sections of the book, I give information pertaining to data gathering, which for each series shares a common philosophy, but there are individual differences. However, nowhere is the task more uncertain and difficult than with early silver dollars. The main dilemma concerns die varieties. It would have been easy had the object been to evaluate the rarity of early dollars by dates or by major types.
Over the years, early silver dollars 1794-1803 have been offered in numerous dealers' advertisements and price lists. However, for the vast majority of such offerings no attributions have been given to either Haseltine or Bolender numbers. Rather, most are listed by date and basic description, such as "1795 Flowing Hair" or "1798 Small Eagle reverse," or "1802/1 overdate." While a compilation of such offerings is useful to determine the relative rarity of a given major variety or type, it is of little use in making finer distinctions.
As I have noted, sometimes a listing can be repeated over and over. If a dealer offers an attributed coin such as a 1795 Flowing Hair dollar, Bolender-1 [BB-21], in EF grade, in five successive catalogues or five successive retail advertisements, does this mean that he has just one coin and has not been able to sell it, or has he had five different? The answer is unknown. The presumption is that he had just one piece, but for a common issue he may have had several.
Auction data are useful to an extent. Probably the best advantage to be secured from auction information is the relative rarity of a given issue at different grade levels. If 20 VF-20 specimens of a specific die variety are offered over a five-year span of time, even if some of the auction appearances represent duplication, it is still reasonable to assume that in VF-20 grade that particular variety is not too difficult to acquire. If over a span of decades, the highest specimen of a given variety auctioned is VF-30, then it is reasonable to assume that should an EF-40 coin appear, it definitely would be a Condition Census (in the top six known specimens) item. More about the Condition Census later.
The main flaw with auction data is that the numbers tend to make rarer coins seem more available than they really are, and to make common varieties seem less plentiful than their true proportion in the real population.
The reason for this is that numismatists building collections have sought just one example of each date, date and major type, or variety, as the case may be. They have not wanted duplicates. Thus, going far back into history, a representative collection by dates and major varieties would typically contain a 1794 dollar, 1795 Flowing Hair dollar, 1795 Draped Bust dollar, and so on. In actuality, among 1795 silver dollars, the Draped Bust variety is many times more elusive than the Flowing Hair type. If relative rarity were to be shown, a collection that has one 1795 Draped Bust dollar should have several of the Flowing Hair style. However, no collector, unless he or she is acquiring coins by minute varieties, wants more than just one of each. Because of this, Flowing Hair dollars and Draped Bust dollars dated 1795 seem, per the auction records, to be closer together in availability than they actually are.
To expand the illustration, consider the Lincoln cent series minted from 1909 to date. Over the years, just about every collection of Lincoln cents ever offered has had a 1909-S V.D.B., one of the lowest mintage issues in the series. The typical collection has also included a 1959 Philadelphia Lincoln cent, a coin made to the extent of over 610 million pieces. Today, for every Uncirculated 19098 V.D.B. cent in existence there are many thousands of 1959 Lincoln cents. However, if you Were to compile auction data, you would find that this great difference in availability would not be reflected. Many other examples could likewise be given.
The ideal way to determine the rarity of silver dollars dated 1794-1803 would be to have had access to the stocks and auction consignments of J.W. Haseltine, the Chapman brothers, Edward Cogan, W. Elliot Woodward, Ed. Frossard, S.K. Harzfeld, and other nineteenth-century dealers, and to have kept track of the die varieties they handled, making careful notes about scratches, toning, edge bumps, etc., So that the reappearance of a given specimen could be identified. In the present century, the stocks of B. Max Mehl, Thomas L. Elder, John Zug, Numismatic Gallery (Abner Kreisberg and Abe Kosoff), Stack's, New Netherlands, Heritage, Superior, and many others could also be audited.
However, this is not possible. If it had been, perhaps part of the fun of collecting and studying early dollars would be lost. After all, there is something fascinating about the unknown, the quintessential, the mysterious.
In most numismatic series; among the best sources for information today are the population reports published by the various grading services, with NGC and PCGS being the most widely known. While resubmitting the same coins has caused some inaccuracies, particularly among pieces that are more expensive and/or in higher grades, in general the proportions remain correct, in my opinion.
In the analyses of specific dollar die varieties to follow I make limited use of NGC and PCGS data. Unfortunately-very unfortunately these- two fine grading services do not attribute coins by die varieties, so we have no idea what particular varieties have been certified. That information must be obtained elsewhere, such as from a dealer or auctioneer who has attributed a group of dollars earlier certified by these services.