Money That Was Never One Money

A money nobody minted

A Lübeck merchant who entered a debt in his ledger entered it in marks lübisch. The figure was exact. The thing it named did not exist: no coin anywhere carried the word mark, and no mint had yet struck one. When the debtor came to settle he arrived with a sack of small silver, and the two of them counted it out against a number that lived only in the book.

The counting was at least fixed. One mark was sixteen shillings, or forty-eight Witten, or a hundred and ninety-two pennies [CR], and a clerk in Wismar divided it the same way a clerk in Hamburg did. That is what a money of account is: a unit for keeping the books in, whether or not any coin carries its name. Most large trading regions had one, and nobody gave the arrangement a thought until the metal moved. It is a considerable convenience. Two men in different towns, holding coin from different mints, could agree a price and a court could hold them to it, without either of them producing a purse.

The metal was the part nobody fixed. A coin has a weight, and it has a fineness, which is the share of that weight that is really silver rather than the copper it is alloyed with. Both belonged to whoever ran the mint, and the right to run a mint was something the towns held, guarded and used. They used it: silver was periodically scarce and always expensive, and a council that struck a slightly lighter coin from the same bullion had found revenue without calling a tax. So the accounting stood still while the silver underneath it moved. A mark was forty-eight Witten by definition; a Witten was worth whatever the current Witten contained; and when the current Witten contained less, the definition did not notice. None of it required dishonesty from anyone. The unit was a name for a ratio, and the ratio was to a coin that kept changing.

The mark was eventually struck as a coin, in 1502, which is very late. By then the unit that the new piece was made to embody had been shedding metal since 1401, and the coin was minted to match what the unit had become rather than to restore what it had been.

What three towns agreed in 1379

On 9 February 1379, at Lübeck, the councils of Lübeck, Hamburg and Wismar agreed a recess about coin [CR]. Its politics have already been told. Its contents have not, and the contents are what made it last.

The three towns settled which denominations they would strike, what each would weigh, how fine it would be, and how the striking would be tested. The Witten came out at roughly 1.33 grams of metal, of which about 1.12 grams was silver [E]. Lüneburg entered the arrangement in 1381.

That is a short list and a real institution. A Hamburg merchant handed Wismar coin could know what he had without weighing it, and a mint that drifted from the standard could be shown the text it had drifted from. The testing clause is what makes the rest mean anything: a specification nobody checks is a wish, and these towns wrote down that somebody would check. Written standards of that kind are rare in the Hanseatic record. Most Hanseatic agreements were about privileges to be defended abroad; this one was about a thing to be made correctly at home.

Two facts then get run together, and the summaries usually run them. Rostock and Stralsund are commonly listed inside a four-town or wider union of 1379. They were not parties to the treaty of 1379, nor to Lüneburg's accession in 1381. What they did, later, was strike coin to a compatible standard, and other northern towns moved in and out of the arrangement on the same loose footing. Minting to somebody else's specification and sealing your council's name to it are different acts with different exits: the first can be abandoned in an afternoon when bullion gets expensive, and the second has to be renounced.

Who agreed is worth reading as politics. Lübeck, Hamburg, Wismar and then Lüneburg are the same core that had run the war chest against Denmark and taken the Stralsund terms off its council. That machinery was now pointed at a peacetime problem, and it produced the one durable written constitution those towns built in the period.

And it was regional. Cologne is not in it. Neither is Danzig, nor Riga, nor Reval; the Prussian towns and the Livonian towns ran their own coinage on their own standards and went on doing so for the rest of the period. A merchant sailing from Lübeck to Riga found a familiar council and a familiar court, and still had to change his money at the far end and pay for it. The agreement of 1379 was four towns at the western end of the Baltic settling what a Witten would weigh. No comparable standard was ever agreed across the whole Hanse.

A century of getting lighter

The first cut with a firm year on it is 1401. After it, one mark lübisch in the books stood for 46 grams of fine silver. Then 40 grams in 1411. Then 28 in 1424, 25 in 1433, and 20 in 1461 [E]. Reducing the silver in the coin struck under an accounting unit, so that the same name in the ledger stands for less metal, is debasement, and that sequence is what a century of it looks like from the ledger's side.

Five dates, with long flat ground between them. The standard set in 1424 held until 1433; the one set in 1433 held until 1461. Between reductions nothing moved at all, and a merchant who learned the weight of a Witten as a young man could reckon on that figure for decades and never be wrong. The mark did not slide downward year by year. It was cut, left alone, and cut again, each time by a council that had decided to cut it, usually because the difference was worth having.

At 1461 the evidence divides. Two measurements survive for that year and they give different numbers. The series above is coin content: the fine silver actually embodied in the accounting mark, 20 grams. The other reading comes from the mint's own counter. Lübeck's mint bought bullion to strike from, and the price it paid, about ten marks lübisch for a standard weight of fine silver, implies about 23.4 grams behind each mark [E].

They do not contradict each other, because they are not measuring the same thing. What a mint pays for metal has to sit above what it puts back out as coin, or the mint cannot pay for itself; the gap is seigniorage, what the mint keeps between the bullion it buys and the coin it issues. There is nothing scandalous in that. It is how the moneyers were paid and how the town took its cut, and a mint operating any other way would have closed. So 20 grams and 23.4 grams bracket 1461 from below and from above. Neither is averaged into the other here, and no value between them is quoted or drawn anywhere on this page, because no value between them is known.

One further reading carries the series to its end, and it has a different provenance from everything before it. By 1502 one mark lübisch stood for about 18 grams of fine silver [E]. 1502 is the year the mark was first struck as a coin, and 18 grams is the mint standard of that coin: thirteen marks to a fixed weight of fine silver. Every value in front of it is a scholarly reconstruction of what a name in a ledger stood for. This one is a specification a mint undertook to meet, in metal, on a coin that survives in cabinets. That is why the figure below sets it apart.

Where the scholarship disagrees, and why it matters

One value has been missing from all of that, and the rest is measured against it: where the century starts.

Two reconstructions exist, and they do not agree. The compilers of the Riksbank's historical monetary statistics for Sweden, who worked out a mark lübisch series because Swedish accounts were partly kept in it, put the late-fourteenth-century mark at 50 to 55 grams of fine silver [E]. Oliver Volckart, working from the Hamburg chamber accounts that carry the prices he studies, puts it at 61.53 grams for the years 1365 to 1374 [E].

That is a wide gap to find at the foot of a series, and it invites the obvious objection. If two careful reconstructions are that far apart about the starting height, why believe any of the rest of it?

Because they disagree about the height and agree about the shape. The two sets of values converge as the century goes on, and by 1461 they are on the same figure; the whole of the disagreement is banked at the start. Choosing between them changes how large the total fall was. It changes nothing else. The sequence of cuts, the flat ground between them, and the loss of more than half the metal all survive either base year intact.

Neither figure is anything anybody wrote down at the time. No clerk in Lübeck recorded the fine-silver equivalent of an accounting unit; the unit was what he measured other things with. Both values are reconstructions, worked backwards out of mint standards, surviving coin and the prices in somebody's account book, and both carry the estimate label for that reason.

Neither is preferred here and neither is called wrong. They were built out of different archives for different jobs. The Riksbank volume needed a Lübeck standard consistent with the Swedish payment system it was documenting. Volckart needed a silver equivalent internally consistent with the Hamburg accounts his prices come out of, because a price series and a deflator drawn from different sources disagree for reasons that have nothing to do with prices. Each is the better choice for the work it was built for, and neither was built to settle this.

Drawn to scale beside the reductions themselves, the distance between the two readings turns out to be taller than most of the individual drops the series records. That is the band at the left of the figure above.

Credit, forbidden and used

At the Peterhof in Novgorod, lending was against the rules, and the reason was not squeamishness about interest. A German merchant who could not pay was a German problem. Under the law everyone traded by, a foreign community stood surety for its own members, so a single default could bring seizure, arrest or a closed compound down on all of them. Forbidding credit was risk containment written as a rule of the house.

What the merchants did underneath the rule was another matter. Goods were priced in Russian money and matched against goods priced the same way. Obligations running in opposite directions were netted off instead of being settled coin for coin, so that wax owed one way and cloth owed the other cancelled on the books and nothing crossed the yard. Payment was deferred, and where deferral itself was forbidden it was dressed as something else. None of this was barter in the sense of exchange without prices: every bale in the compound had a price in somebody's money, and the rule shaped how credit was done rather than whether it was done.

The rule survives better than the practice. What comes down from the compound is overwhelmingly the record of things going wrong, because quarrels and appeals are what got written to Lübeck and kept; a merchant who netted an obligation with a man he trusted and sailed home generated no document at all.

Bergen ran the same instrument in the opposite direction. German merchants there advanced provisions, grain and equipment to households along the northern Norwegian coast against fish that had not been caught yet. The household ate through the winter on the merchant's flour; the merchant took delivery of stockfish the following season and set it against what was owed. The arrangement was older than the formal establishment at Bryggen, and it thickened as the Bergen trade concentrated in German hands.

It also worked as a wall. A Dutch skipper arriving with cash could buy fish at the going price. He could not buy fish that was already owed to somebody else, and a great deal of it was. Competitors without a network of debtors were competing for whatever the network had left over, which is a subtler barrier than a privilege and harder to revoke.

And the men working under all these prohibitions were writing bills of exchange, which are written orders to pay a sum in another place and usually in another money. Hildebrand Veckinchusen's surviving correspondence is full of them, alongside partnerships that reached Bruges and Venice. He also failed. What his papers show going wrong is overextension, stock he could not turn into money, and obligations falling due before the money to meet them came in. That is a merchant's failure of a particular kind, and it is available only to a man using the instruments: nobody defaults on a maturity mismatch by trading out of a strongbox. Hanseatic merchants are often described as financially old-fashioned, and Veckinchusen went under in a way that requires the modern equipment.

What a common coin actually bought

So what did having a common coin actually buy the towns that had one?

Boerner and Volckart went at the question through exchange rates. They assembled 4,156 yearly observations of the spread between the buying and the selling rate for money changed between two towns, across about 650 city pairs, from 1352 to 1562 [D]. The spread is the price of changing money. A changer quotes one rate to buy your coin and a worse one to sell you his, and the difference is what the service costs; it widens when he is unsure what your coin contains, when he has to hold a stock of it, and when the two towns' money is hard to compare. It is also one of the few medieval quantities that survives thickly enough to count, which is why it is the question that got answered.

Across all pairs the mean spread is 2.27 per cent. Across pairs where both towns belonged to the Wendish coinage union it is 1.16 per cent, on 41 observations. Their estimated effect of a shared coinage, holding the rest of the model constant, is about −0.907 percentage points [D]. A dummy for Hanseatic membership as such comes out negative too, and it is not robust: it survives one specification and not the others.

Three things sit on that finding. The Wendish cell is thin, and 41 observations are carrying a coefficient. Most of what the dataset counts is exchange quotations rather than bills or loans, so it measures what it cost to change money and not what it cost to finance a voyage; financing cost is the thing nobody has been able to reconstruct. And the direction of cause is not clean, because towns whose trade was already entangled were the towns most likely to agree a coinage in the first place, which means a union effect can be reading integration that was there before the union. The authors say all three themselves, which is a point in the result's favour and does not widen it.

There is a comparison readers arriving from the popular accounts will want, and it cannot be supplied. Whether a Hanseatic merchant paid more for his money than a Fugger or an Italian house is a good question with no dataset behind it. Comparable financing costs for the two sides have not been assembled, and a percentage produced here would be invented. What has been measured is the exchange spread, and what it says is that the coinage union moved it and Hanseatic membership did not.

What that implies about the League is a separate question, and it is taken up in the walkthrough on the Wendish monetary union.