distance elasticity −1.97***
contiguity −1.855***
common language −1.759***
US MLAT suppression −2.539***
observations 353,983
cryptomarkets 10
drug categories 57
exporters 63
importers 88
classifier accuracy 98.4%
live markets 8
panel extended 2013–22
distance elasticity −1.97***
contiguity −1.855***
common language −1.759***
US MLAT suppression −2.539***
observations 353,983
cryptomarkets 10
drug categories 57
exporters 63
importers 88
classifier accuracy 98.4%
live markets 8
panel extended 2013–22
Vendor Profile
Active · Los Angeles · Verified
ryder mawby
illicit market analyst · cryptomarket intelligence · empirical economist
vendor ryder mawby
ships from Los Angeles, CA
category Cryptomarket Intelligence · Illicit Trade Analysis · OSINT
feedback 353,983 observations · 98.4% accuracy · distance β −1.97***
payment correspondence · collaboration · data
profile → about  ·  CV  ·  github
Working Paper
SSRN Working Paper · May 2026 · Prepared for Journal of International Economics
Anti-Gravity Exists: Evidence from Darknet Drug Markets
Hikari Archive (CMU) · Gwern Darknet Market Archive · CEPII Gravity Database · DOJ MLAT Coverage  ·  PPML · ppmlhdfe · country-pair clustered SE
Conventional gravity theory predicts that shared borders, common languages, and colonial ties facilitate trade. On darknet cryptomarkets, they behave antithetically — yielding a distance elasticity of −1.97 and sign inversions across all standard facilitation variables. The suppression operates entirely through the intensive margin and is consistent with a bilateral enforcement coordination mechanism.
−1.97***
Distance elasticity — exceeds the canonical licit-goods range of −0.9 to −1.5, consistent with added enforcement deterrence on international illegal shipments. Lower bound given advertised-destination measurement.
inverted
All facilitation variables carry the wrong sign. Contiguity (−1.855***), common official language (−1.759***), and colonial ties (−1.015**) suppress bilateral drug trade — the inverse of every licit-goods gravity study in the literature. Effect operates entirely through the intensive margin.
−2.539***
MLAT enforcement coordination mechanism. Routes with formal US mutual legal assistance treaty obligations generate substantially less conditional trade volume — independent of the language coefficient — directly implicating institutional enforcement coordination, not anonymity-seeking behavior.
98.4%
8-stage multilingual classification pipeline resolving vendor-entered listings into a 57-category pharmacological taxonomy across English, German, Dutch, Finnish, and French. Validated on 353,983 Hikari listings. Publicly available for any cryptomarket dataset.
Methods & Tools
PPML gravity · high-dimensional FE
Panel data · DiD · event study
Multilingual classification pipelines
Live cryptomarket data collection
Stata · R (httr/rvest, magick, tidyverse) · Python
Causal inference · IV
OSINT · darknet market monitoring
Illicit payment rail analysis