The Parallel Spice Market of Arrakis, 10160–10191 AG: Smuggler Tally-Books, Prices Against the CHOAM Allocation, and the Guild's Spice-Paid Tolerance

1. Introduction

Entry 212 of the tally-book that the CHOAM Directorate catalogues as TB-11 records a consignment of four stores units, 6.0 tonnes of melange, bought from a Fremen supplier at a staging field in 10174 AG for 0.29 of the Directorate's allocation price for that year and sold off-world at 1.41 times it. A Guild surface-to-orbit manifest of the following week lists a cargo of the same mark and a mass within two per cent. Its consideration line reads 0.74 tonnes, passage in kind, which is 12.3% of the lot. No tariff line in solaris appears.

We write c. 10250 AG, some five decades after the Jihad of Muad'Dib, when the Directorate's interdiction files and an extract of the Guild's Arrakis manifests can be read side by side. That smuggler bands were active on Arrakis under the Harkonnen concession is not in doubt: Gurney Halleck, an Atreides officer, found refuge with one after the Harkonnen attack of 10191 AG, and Fremen traded water and spice with such bands. What has been missing is a measurement of how much spice moved, at what price against the allocation, and why a trade that could leave the planet only in Guild hulls was allowed to continue.

The Guild supplies the frame. It holds the monopoly on interstellar transport, so every tonne that left Arrakis left in Guild carriage, and it kept no weather satellites over the planet, a refusal the Fremen secured by paying it in spice. A carrier paid in melange for what it chose not to observe could as easily be paid in melange for what it chose not to inspect. We test that analogy against ledgers rather than assume it. The reconstruction below is our own.

2. Sources and Method

Three series carry the study. The first is the Directorate's annual concession returns for Arrakis, 32 returns for 10160–10191 AG, giving tonnes delivered into the allocation and the allocation price (CHOAM Directorate, 10160–10191 AG). The second is the interdiction series, 41 actions in 10163–10191 AG, in which investigators impounded 19 tally-books from 12 bands (CHOAM Directorate, 10163–10191 AG). The books hold 1,126 dated entries; 874 are consignment sales with a mass and a price, and 702 of those fall in 10166–10191 AG. The third is an extract of the Guild's Arrakis surface-to-orbit manifests, 436 manifests for the same years that carry a passage consideration in kind (Spacing Guild, 10166–10191 AG), read with the Guild's tariff schedule for stores cargo (Spacing Guild, 10150–10191 AG).

Each tally entry gives date, supplier, route and consignment marks, mass, and purchase and sale prices; we convert both prices to ratios of the allocation price in the year of sale. Matching to manifests is by consignment mark, with a date within 21 days and a mass within 5%, and yields 126 matched consignments. Every matched manifest carries an in-kind line and none also carries a tariff line in solaris. Because the bands kept books for their own accounts and the Guild's clerks wrote manifests for passage, we treat the lists as independent captures and estimate the population with a two-list capture–recapture estimator with a small-sample correction. Loosening the rule to 45 days and 10% gives 141 matches; tightening it to 7 days and 2% gives 104.

Volume and ratio intervals come from a simulation over 20,000 draws; other intervals use a normal approximation. The price regression uses 26 annual means on the log scale, with no first-order autocorrelation in its residuals (lag-1 coefficient −0.05).

Access shaped the record. The first author's office sits inside the Directorate, which had reason to read impounded books as shrinkage, and the third author's belongs to the Guild's banking arm, which had reason to read its in-kind lines as ordinary pricing. Consignee names were masked, so matching rests on mark, date and mass. Staging-field placement draws on unpublished blow-district sheets of the second author's office.

3. Volumes and Routes

The two-list estimate for 10166–10191 AG is 2,418 consignments (95% CI 2,100–2,740). Mean lot mass in the 702 entries is 3.14 tonnes (95% CI 3.01–3.26), giving roughly 290 tonnes a year (95% interval 250–330). The Directorate's returns give a mean delivery of about 2,800 tonnes a year over the same years, ranging from 2,320 to 3,300. The parallel market therefore moved about 10% of the official volume (95% interval 9.0–11.9%). The matching rule moves this: the loosest and strictest rules give roughly 260 and 350 tonnes a year, 9% and 13% of the official delivery.

Three kinds of supplier appear in the 702 entries: Fremen, 428 (61%); harvest crews selling from their own district's output, 190 (27%); and 84 (12%) with no legible mark. Fremen entries record water as well as spice passing between bands and sietches; we count only the spice. If lot masses are similar across suppliers, which we cannot check, about 80 tonnes a year, 2.8% of the official delivery, were diverted from harvest before reaching the returns. The rest never entered them.

Route marks sort the surface legs into three (Table 1). Every route crosses deep desert, where the Guild kept no weather satellites. Interdictions fell where the traffic did: of 38 on surface routes, 17, 14 and 7 fell on the three routes against 15.8, 12.6 and 9.6 expected from traffic shares (χ²(2) = 0.94, p = .63). Three more occurred at the port-field gate. None occurred after a consignment had been loaded into Guild carriage.

Lot sizes show how the trade fitted itself to Guild freight: 103 of the 126 matched consignments (82%) were exactly one, two or four stores units, the 1.5, 3.0 and 6.0 tonne containers the manifests use.

4. Prices Against the Official Allocation

The tally-books give two prices per consignment, as ratios to the allocation price. At the staging field bands paid on average 0.31 (95% CI 0.30–0.32; SD 0.12). Sold off-world, the same material fetched 1.38 (95% CI 1.36–1.41; SD 0.30). The gross spread of 1.07 allocation-price units per unit of mass is large, before caravan, aircraft, interdiction and Guild costs.

Who paid above the allocation? Earlier work by one of us shows that allocation contracts were rationed by quota, so that buyers outside the contract system, and holders whose quota fell short, could obtain melange from the Combine only by waiting (Brisemont, 10231 AG). They faced a price of 1.38 or no spice.

If the parallel market served demand the allocation did not, its premium should rise when official deliveries fall. Across the 26 annual means, the log price ratio fell by 0.59 per unit increase in the log of official delivery relative to its mean (95% CI −0.83 to −0.35, p < .001; R² = 0.52): a 10% shortfall goes with a price ratio about 6% higher.

Valued at the delivered price, the 290 tonnes are worth about 14% of the official delivery at the allocation price (95% interval 12.5–16.4%). Because these buyers could not obtain official supply at 1.00, shadow sales displaced little official trade, except the diverted harvest output.

5. The Guild Leg: Consideration and Protection

Every consignment that left Arrakis left in Guild carriage, and the manifests show what the carrier took. On the 126 matched consignments the in-kind passage charge averaged 12.4% of cargo mass (SD 2.8 percentage points; 95% CI 11.9–12.9). The Guild's tariff for stores cargo is 7.9% of declared value, levied in solaris. Valued at the allocation price, the in-kind charge is 1.57 times the tariff a declared lot of the same mass would have borne (95% CI 1.51–1.63), an excess of 4.5 points. Applied to the shadow volume, receipts came to about 36 tonnes a year (95% interval 31–41), 1.3% of the official delivery.

The charge behaved like a schedule. It did not vary with lot mass (+0.4 points per doubling; 95% CI −0.3 to +1.1) and did not drift across the window (+0.02 points a year; 95% CI −0.05 to +0.09). The Guild's clerks entered it on the same manifest form as their ordinary charges, under a line that reads consideration. Ad hoc exactions would be unlikely to leave so regular a trace.

Protection came with the charge. Investigators acted on the surface and, three times, at the port-field gate, never against a lot in Guild carriage. With none of 41 interdictions on that leg, the upper 95% bound on its share is about 7%.

Earlier work on the Combine's shareholding places the Guild among its silent partners, with a profit-participation line of about 17% of dividend flow (range 12–22%). The in-kind charge appears in no such line: it is a second channel to the same recipient, outside the dividend ledgers, and paid in the commodity the Navigators cannot do without. Earlier work on the heighliner drive argues that the Navigator is a structural component of the fold and that melange dependence would survive a relaxed proscription, so a carrier so placed may value spice above the solaris it would otherwise collect.

6. Discussion

The parallel market was neither marginal nor improvised. At about one tonne in ten by mass and 14% by value, it was large enough for the Directorate's returns to understate the planet's flows, and it kept regular lots, three standing routes and one price structure for 26 years. Bands that bought at 0.31 and sold at 1.38 worked a spread that the allocation's rationing created, and the premium's response to official shortfalls shows a market behaving as the allocation's overflow.

Tolerance was priced. A flat in-kind rate, entered by the Guild's own clerks and set 1.57 times above the tariff for a declared lot, together with a carriage leg the Directorate never reached, separates pricing from a failure to inspect, which would predict none of them; a carrier that could see its cargo and set its terms would predict all three.

This extends the precedent of the satellites. The Fremen secured an unobserved sky over Arrakis by paying the Guild in spice; the bands, we find, secured an uninspected hold by the same means. The manifests say nothing of weather, so the link remains an analogy.

The 4.5-point surcharge need not price risk alone. Earlier work by one of us treats the Guild as a monopoly carrier able to discriminate by cargo and shipper (Iles, 10241 AG), and a surcharge flat across lot sizes and years fits a price chosen by the seller.

7. Limitations of the Record

Impounded tally-books are those that investigators found, and the 12 bands may not represent the trade. The Guild extract holds only manifests with an in-kind line, so consignments carried on other terms lie outside both lists. The estimator assumes independent capture; if the bands that kept books were also those the Guild favoured, the lists are positively dependent and the true count exceeds 2,418. We read the estimate as nearer a floor than a ceiling.

Prices are ratios to the allocation price, which hides any change in the solaris level, and the books carry no cost lines, so band profit cannot be computed. Navigator consumption of melange appears in no available series, so we cannot say how much the 36 tonnes a year mattered to the Guild's own supply. The record stops in 10191 AG.

These limits bound the figures and leave the findings standing. The parallel market was a measurable flow of about 290 tonnes a year, a tenth of the allocation by mass and about 14% by value, supplied mostly by the Fremen and priced at 1.38 times the allocation. The Guild captured a rent from it in spice: 12.4% of each lot, 1.57 times its own tariff, taken on the one leg of the journey that the Directorate never reached.