Dilèm Aksyon Kolektif nan Matisan

Generatim discite cultus

(Learn the culture proper to each after its kind)

— Virgil, Georgics II

Stephen Biddle, Nonstate Warfare: the Military Methods of Guerrillas, Warlords, and Militias (Princeton: Princeton University Press, 2021)

By way of making this more than a merely armchair review, I will be discussing the developing situation of state failure in Haiti, which is providing a personally harrowing example of the phenomena theorized and studied in this book. NB: additional situation reports like the one I quote from below will appear at this OCHA webpage.

I. Increasingly Scale-Free Military Activity in the 21st Century

In this follow-up to 2004’s Military Power: Explaining Victory and Defeat in Modern Battle (also from Princeton), Stephen Biddle continues to elucidate the many ramifications of the one-to-many relationship which came to dominate the battlefield between the Napoleonic Wars and World War I. Over that century and in the decades that followed, individual-service weapons increased in rate of fire from a (very) few rounds per minute to ~10 rounds per second, in effective range from ~100 to >300 meters, and in accuracy from (optimistically) 10 to 1.5 milliradians. Say 2 ½ orders of magnitude improvement in RoF, half an order of magnitude in range, and one order of magnitude in accuracy; multiplying these together to create a sort of index of effectiveness, I get an overall change of 4 orders of magnitude, with stark implications for battlefield environments.

Where it was thereby feasible for Napoleonic armies to march directly toward a defensive position and overwhelm it before many of them could be cut down, by the early 20th century, a tiny fraction of surviving defenders (that is, after an ostensibly preparatory bombardment) could annihilate a far larger attacking force. This created enormous selection pressure on tactics and, during the final eighteen months of World War I, began to yield the “modern system,” summarized by Biddle in his earlier book as “cover, concealment, dispersion, small-unit independent maneuver, suppression, and combined arms integration.”

Nonstate Warfare‘s central point is that in the 21st century, the ever-growing capabilities of individual-service weapons (and, I might add, communication systems), besides continuing to move nation-state militaries away from the Napoleonic extreme, are giving nonstate military actors, even those in quite small groups, the option of moving away from the opposite, “Fabian” extreme of high concealment and low lethality—basically the occasional assassination—toward far more open, direct challenges to Westphalian-Weberian states.

“Hybrid warfare” (a term Biddle generally avoids) is not new; much of the action in Missouri, 1861-65, consisted of guerrilla activity not always readily distinguishable from smaller battles between regular forces. But what happened here was heavily determined by politics, demography, and geography, rather than hardware. What is happening around the world today is massively enabled by cheap fully-automatic rifles, even cheaper instantaneous communication beyond lines of sight, and—especially in the case I describe in detail below—high-density urban environments which strongly favor insurgents.

In the broadest sense, Biddle’s exposition is one of various independent variables of capability versus a single, overriding dependent variable of behavior of military actors. This could be visualized as multiple three-dimensional graphs, which he perhaps wisely eschews in favor of a single table (4.2) of predictions on page 105. For this review, I compress it into a single quadrant diagram of my own devising:

The classical relationship across a graph of capability-vs-behavior was linear and direct, as indicated by the pale red and green quadrant labels above: organizations with little firepower and minimal organization were forced into Fabian behavior, while those with great firepower and reliable institutional functionality automatically employed Napoleonic behavior. Only quadrants II and III were viable; quadrants I and IV were nonsensical. But the historically new ability of a single combatant to suppress or directly disable tens to hundreds of opponents, and the potential (and even newer) ability of an otherwise resource-poor group to tightly coordinate its activities, has both advanced the behavior of the formerly weak and forced adaptation on the part of the formerly strong. Everyone is moving toward the center of the graph, Biddle’s “midspectrum.”

II. Contexts for Haiti and Martissant

The island of Hispaniola/Kiskeya experienced a population crash after Columbian contact; this is often described as the result of violent conflict, but as elsewhere in the Americas, was almost entirely an epidemiological phenomenon. The Spanish largely abandoned the island in the mid-1500s, and the natives having all but vanished, it reverted to jungle. After a period of buccaneer activity, especially around Tortuga, the French obtained formal possession in the late 1600s and ramped up a massively extractive colony consisting mostly of sugarcane plantations worked by African slave labor. Haiti became an immensely valuable possession, responsible for a majority of the value of all imports to France, such that after their defeat in the Seven Years’ War, the French preferred to yield Quebec to the British, it having the smaller GDP. The population of Saint-Domingue, closely approximating Haiti’s present boundaries, was nearly seven-eighths slave by the late 1700s, concentrating 450,000 recently transported Africans or their immediate descendants in the 9,000 km ² of the country flat enough for intensive cultivation. A third of the entire Atlantic slave trade fed Saint-Domingue; the only environment in the US approximating its concentration and demographics was Lowcountry South Carolina.

The resulting state of what we would now call self-organized criticality, made even more acute by the outbreak of the French Revolution, culminated in the plotting and initiation of the world’s only successful large-scale slave rebellion, which began in August of 1791. A dozen years of bloody chaos ensued, in which one-fifth of the population may have perished. The ultimate and immensely strategic accomplishment of the Haitian Revolution was the defeat of the Saint-Domingue expedition, commanded by Napoleon’s brother-in-law, which had been intended to reconquer the island—and then proceed to garrison La Louisiane, which would have firmly established the western border of the US at the Mississippi River. Instead, Haitian independence led quickly to the Louisiana Purchase, at 2.1 million km ² the single greatest episode of Westward Expansion in American history. It also led to the first curbs on the Atlantic slave trade, even ardent pro-slavery politicians in the US having taken heed of the deadly consequences of overconcentration of slave labor.

For American baby boomers, Haiti was synonymous with Duvalierist tyranny until well into our adulthood. Factionalism and general desuetude after Baby Doc’s removal in 1986, eventuating the departure of Jean-Bertrand Aristide in early 2004, caused the introduction of MINUSTAH (Mission des Nations Unies pour la stabilisation en Haïti) to suppress armed gangs, especially in Cité Soleil, a notorious shanty town on Baie de Port-au-Prince due west of Aéroport International Toussaint Louverture. Even generations-old armored personnel carriers from Third World armies like the ones I occasionally saw parked in the median along Route des Rails in the Arcachon neighborhood of Carrefour were sufficient to intimidate the gangs. But a Nepalese MINUSTAH unit based on the Rivière Meye a few kilometers south of Mirebalais introduced cholera (a known Asian variant, V. cholerae serogroup O1, serotype Ogawa, biotype El Tor) into the Artibonite basin in October of 2010, killing thousands of people. MINUSTAH was replaced by the much smaller MINUJUSTH (Mission des Nations unies pour l’appui à la Justice en Haïti) 7 years later; I have not yet attempted a calculation to determine whether MINUSTAH was responsible for net lives saved vs lost over the period 2004-17, but I note that the sort of lives it potentially saved included those of foreign mission workers, ie mine, while the sort of lives it certainly took were mainly those of deeply isolated rural Haitians many hours’ travel time from medical facilities.

It was Haiti’s geological context, however, which brought it fully into the American consciousness in early 2010. After nearly a quarter-millennium of relative quiescence, at 4:53 PM local time on Tuesday 12 January of that year, the Enriquillo-Plantain Garden fault zone ruptured 10 km southeast of Léogâne in a 7.0 M_w (≈800 TJ, around half a megaton of TNT equivalent) tranblemanntè, followed by numerous aftershocks, mostly westward of the mainshock hypocenter. However intermittently lurid the video of earlier unrest might have been, the images of massive destruction in Port-au-Prince were absolutely spectacular, and Americans opened their wallets; over half the US population made charitable contributions to Haitian relief in the immediate aftermath.

Fourteen months later I undertook the first of ~20 trips to Haiti over the next 8 ½ years; some of my impressions are recounted in Poukisa Mwen Te Ale An Ayiti. On my last visit I encountered … challenging second-order effects of the destruction of Venezuela’s economy by “Bolivarianism,” in the form of a severe energy crisis in the wake of curtailment of subsidized fuel deliveries under the PetroCaribe program; a somewhat mathematical account is at Ayiti Pa Nimewo Yo. Shortly thereafter, Haiti experienced a literally overnight 37% deflationary shock followed by what is shaping up to be 90% annual inflation. All this seems to have engendered something close to collapse of legitimate authority.

Around a third of Port-au-Prince’s territory [~50 km ² —JDM] is affected by the criminal activity and violence propagated by an estimated 95 armed gangs. Since 1 June, a significant upsurge in deadly clashes between these rival gangs in the metropolitan area, triggered by a reconfiguration of gang alliances and ongoing territorial disputes, continue to fuel widespread insecurity and displacement, with devastating consequences for the civilian population. The situation has worsened over the last five days and will likely continue to deteriorate in the coming weeks, as gangs are expected to fight back to regain territorial control, potentially triggering new population movements.

HAITI: Displacements due to gang violence in Port-au-Prince
OCHA Situation Report No. 3 (14-22 June 2021)

While I would not claim truly intimate knowledge of the affected areas, I have been to most of them, and some of them many times. I will explicate Martissant in detail due to its combination of extraordinarily high density and location astride a vital transport artery, Route Nationale #2, which connects Port-au-Prince with most of the Tiburon (southern) Peninsula.

Where the indigenous population had a density of ~5 km⁻ ² at Columbian contact, the island of Hispaniola’s current population density is approaching 300 km⁻ ², and Haiti’s has reached 400 km⁻ ². Times also having changed in other ways, Haiti’s GDP, and its GDP per capita, is now ~2% of Quebec’s. The crowding and poverty are at their most glaring in Martissant, the neighborhood immediately southwest of downtown Port-au-Prince. Geographically, it extends from just over a kilometer southwest of the Palais National westward 2.5 km along RN2, and from the Bay southward 4.5 km to the summit of Morne Saint-Laurent (884 m). Almost the entire population of the neighborhood, however, lives in the northern half of this area, in elevations ≤250 m, and that population is enormous, exceeding 250,000. The effective density is therefore ~50,000 km⁻ ², or one person for each twenty square meters. Even with narrow streets and two-story buildings, individual human beings are rarely as much as ten meters apart, and typical lines of sight, other than along RN2 (Boulevard Jean-Jacques Dessalines) itself, are even shorter.

(To bring this home, at the density of Martissant, the entire ~9.5M population of Chicagoland would fit in a semicircle centered on the Loop with its northern boundary in Uptown, its western boundary in Austin, and its southern boundary in Hyde Park.)

Doing a one-significant-figure calculation from the paragraph quoted above, we find that each armed gang controls, on average, 50 hectares. In the lower-elevation, populated part of Martissant this could mean noticeably affecting the activities of more than 20,000 people. There would be ~10 such gangs; the importance of RN2 strongly suggests that they would contend for stretches of it less than 300 meters in length, with their territories running north-south by more than 2 kilometers. Their familiarity with structures, narrow streets, and passageways, in combination with high rate-of-fire individual weaponry, seem certain to make them difficult to entirely dislodge, especially if their only opponents are Police Nationale d’Haïti.

III. Nonstate Actors in Haiti

Constraints on nonstate aggressors in Haiti do not, and seem unlikely to, come from the PNH. Gangs have attacked police stations, raided their armories, and killed policemen. I am separately informed via a long-time missions worker that some gangs have issued demands for ransom for captives over radio stations. A Haitian Georges Danton living in 2021 might well say: De l’impunité, encore de l’impunité, toujours de l’impunité et la Patrie sera détruite!

If the PNH can be, even briefly, overwhelmed and their arms—which include rifles, many of them fully automatic, and even some light and medium machine guns—confiscated by gangs, Haiti may become another ghastly case study like the ones in Nonstate Warfare (Croatia, Iraq, Israel/Lebanon, Somalia, and of course Vietnam).

But there are constraints after all, powerful ones, among the independent “software” variables glancingly alluded to in my quadrant diagram above. As of early summer 2021, Haitian gangs seem vulnerable to severe limitations imposed by the very criteria Biddle discusses: lack of specialization, weak (intraorganizational) institutions, and limited stakes. Ironically, the PNH’s inability to fully suppress the gangs also removes a selection pressure that might induce organizational maturation. If a two-digit number of gang members with nothing more substantial than stolen M4 carbines and a few ammunition magazines apiece can set up roadblocks and carry out shakedowns and abductions while holding the police at bay, their main problems are 1) spot shortages of just about everything, made marginally worse by their own obnoxious activities 2) other gangs and 3) internal stresses, which are likely to eventually become significant.

If any environment in any human society today is going to engender lots of decision-making “prisoner’s dilemmas,” it’s urban Haiti. For Ayiti Pa Nimewo Yo, I calculated that a pedestrian in Port-au-Prince can routinely cover an area inhabited by 200,000 people. Given the inverse relationship between the number of people one can encounter and the probability of seeing any particular one of them, that’s going to make Axelrod‘s “discount parameter” for assessing the likelihood of future interaction a whole lot lower than it is in the Haitian countryside, where the corresponding number is at most 3,000, and sometimes in the hundreds. In any case, gangs are by definition collections of game-theoretic “defectors.” The minimum effective size of an armed gang in Martissant is probably in single digits, gang members being unlikely to score highly on the personality dimension of “agreeableness”—see the OCHA situation report’s mention of “a reconfiguration of gang alliances and ongoing territorial disputes.” Vocabulary word: fissiparous.

But what is a likely maximum size, and why? This is where I actually add some value to Biddle’s analysis by introducing the Dunbar Number and its implications. That number is often blandly given as 150, but as Robin Dunbar himself explains, there are various nuances:

Generally speaking, humans each have one to two special friends, five intimate friends, 15 best friends, 50 good friends, 150 “just” friends and 500 acquaintances. Our relationships form a series of expanding circles of increasing size and decreasing intensity and quality of the relationship. Not only do we see these circles in the structure of [modern, real-world] social networks and of huntergatherer communities—they’re also reflected in big data gleaned from Facebook posting and telephone call frequencies. It also turns out these layers are germane to the organizational structure of modern armies.

I suggest that Haitian gangs may well evolve toward these thresholds, and that—in the event of reaching or surpassing them, which is far from guaranteed—will acquire increasingly mature institutions, specialization, and eventually what Biddle calls “permissive” internal politics, less dependent on strictly personal loyalties. Recall that (again to one sig fig) ~100 gangs are thought to be operating in Port-au-Prince as of June 2021. The total strength of the PNH is ~10,000; aggregate gang strength could exceed it even without higher-order organization. A Haitian “Cyrus” living in 2021 might well exhort them with: Mwen di, tan kap vini an se pou nou … si ou ka konte!

It seems likely that most Haitian gangs are in the 15-50 size range and that few, if any, are above the 150 threshold. A functioning gang of 500 with specialized members and a commitment to territorial control, especially if it suppressed other criminal activity within that territory, would represent a significant development.

IV. Lessons Learned

Some relevant observations:

• By way of making an obligatory nod to Strauss and Howe, there’s plenty of incompetence to go around during a Crisis Era. The relevant features of this one include dubiously effective foreign aid, domestic Haitian political gridlock, “Bolivarian” lunacy in Venezuela, and of course American defocusing under an Administration led by a cognitively impaired President and a deeply unserious Vice-President. As for the UN, I do not expect it to last the decade, not that neglect of state failure in a single country physically smaller than Maryland and with fewer people than Ohio is going to bring it down; the current disaster is more of a falling-barometer reading than a storm surge.
• Comparisons will tempt some beyond their ability to resist, but really, Ayiti ≠ Etazini, not by a long way. As I have mentioned in earlier commentary, Haiti has no tradition of political compromise and has experienced at least 32 coups since its founding. But it also has no tradition of an armed citizenry; the 1987 Constitution nominally in effect definitely does not allow for such a possibility, and I strongly doubt that the newly-proposed one does either (although I have not yet read it in English translation), leaving most people vulnerable to victimization in the absence of effective—that is, sufficiently intimidating—policing, which can itself be deeply problematic. In the absence of some cultural shift, the effective alternatives may be possibly oppressive PNH vs certainly oppressive, and viciously competitive, gang leaders … or outright warlords if the scaling challenges noted by Dunbar are overcome. So no, it’s not Portland, or CHAZ, or Biden saying something silly about F-15s and nukes. American leftists are doing a lot of IF/F signaling and LARPing; in the past month alone, Haitian gangs have rendered nearly a tenth of the entire population of the second-largest metro area in the Caribbean unable to access essential services.
• Actual defensive tactics would require, to borrow a term, community organizing. Their effectiveness would be greatly enhanced in more suitable physical environments, those with abundant “prospect and refuge” and longer lines of sight. This does not leave much hope for Martissant unless portions of it are cleared of structures, which implies even more disruption to the inhabitants than is occurring now. It wouldn’t take that many armed guards to secure a sizeable area, but they’d have to be intelligently deployed, as I have observed at gas stations, where one guard is conspicuous, while another waits behind concealment to get the drop on anybody messing with the first one. (Their weapons are usually pistol-grip shotguns like the Mossberg 500.) The PNH would be well advised to surround its commissariats with most of a hectare of unobstructed, level pavement and construct new, windowless buildings with gun ports in the walls. And probably get some drones. Where the money to do any such thing will come from is … nonobvious.
• The North American short-term missions biz will all but shut down in the absence of a great improvement to the current situation. State and the CDC both have Haiti at Level 4—Do Not Travel. That’s going to scare people off all by itself (and make travel insurance difficult if not impossible to get). The physical courage of mission volunteers is, well, actually unpredictable. They’re as skittish as you would expect of any pampered suburban American or Canadian—until suddenly they’re not. The immediate problem for someone like me is that I’d be worth more, possibly much more, as a hostage than as a Random Benign White Guy. My presence could easily become actively harmful. Nor do I foresee the subculture in question adopting armed escorts as the norm. So they’ll stay away.

Things I’d like to see prediction markets for:

• Reverting to type as a project manager, any apparent easing of the “triple constraint” for Haitian gangs, that is, weapon capability and territory size (scope), revenue and manpower (cost), and time horizon (schedule). I suggest above that existing gangs have a two-digit number of members armed with automatic rifles of 300-meter effective range and a three-digit number of rounds apiece; they control areas of tens of hectares and are thinking no more than a few days ahead. Gangs of many hundreds, in possession of weaponry (IEDs or RPGs) capable of stopping at least light armored vehicles, perhaps driving “technicals” of their own, specializing and coordinating their activities to control square kilometers and effectively police the lives of a hundred thousand people for months at a time, using higher elevations for surveillance and maybe even conducting drone reconnaissance, would be a qualitative as well as quantitative leap. Also as noted above, the “software” challenges of internal politics and organization are far greater barriers to this development than hardware acquisition.
• On the positive side, for Haiti, adaptation through the use of facilities other than those of Terminal Varreux in Port-au-Prince to get vitally needed supplies, especially food, to the majority of the entire population potentially cut off by gang activity in the capital, through which every major highway in the country is routed. There are, at least notionally, alternatives in Les Cayes, Petit-Goâve, and Jacmel in the south, and Cap-Haïtien, Gonaïves, and Saint-Marc in the north. For passenger transport, as I well know, it is possible to fly PAP-JAK, and several airlines fly in to CAP.
• While I believe there is relatively little overlap, qualitatively or quantitatively, between the weaknesses in civil society affecting Haiti and the US, both have obvious problems. Serious threats in the US are endogenous and localized, as in the corruption/cooperation between city officials and gangs in Chicago (which is thankfully sui generis), or for a more conceptual phenomenon, badly-managed electricity grids in California and Texas. But a cultural contagion of sabotage and supply-chain disruption, whatever its origin and purpose, would act like, and be perceived by most of the country as, an exogenous threat. There are not, for example, all that many high-capacity bridges over the Missouri and Mississippi Rivers; and to a first approximation, the lights are being kept on in this country by coal trains emerging from the Powder River Basin in Wyoming on a single rail line following the North Platte River. With a third of a billion people, we probably have at least a few hundred thousand who would imagine themselves to be saving the planet (or, more narrowly, somehow thwarting their political opponents) by cutting those links, or blowing up pipelines, or even just attacking a neighborhood electric-power substation … or, to pick a topical example, shooting a cop. The single metric to monitor, however, is the same, ultimate, and utterly unambiguous effect of unanswered gang activity in Haiti: IDPs, internally displaced persons. That number is, to my knowledge, zero in the US. If it’s, say, a couple of million a couple of years from now, well, I’ll have been wrong about that Ayiti ≠ Etazini thing.

V. Flesh and Blood

She, and thousands like her, are why I hope I’m right.

Early in my first trip down, nine months earlier, the team had been tasked with delivering some supplies from the Eglise Méthodiste d’Haïti guest house in Pétion-Ville to its church in Carrefour. In combination with the ride through, among other places, Martissant, this induced in me a strong desire to depart, and I didn’t care where to, as long as it was less filthy, chaotic, and terribly overcrowded. (And indeed, rural Haiti, while obviously an L1 environment, is far safer.) I am not an agoraphobe, but the sense of impending doom was nerve-wracking. After what seemed like several hours, but was actually about 20 minutes, we got back into the van and headed on west.

This time we were driving back east to Pétion-Ville from Petit-Goâve after a week of working on a librairie (that is, a bookstore) for the church there. She was temporary kitchen staff at the maison du pasteur in the EMH compound in the La Hatte neighborhood. Haiti does not lack for beautiful young women, but I was amused by how tongue-tied young men seemed to become in her presence.

We dropped her off in western Carrefour, near where Route des Rails forks north from Boulevard Jean-Jacques Dessalines. The last we saw of her, she was crossing the street, probably to catch a ride home in a tap-tap, among swarms of people, filthy runoff, and mounds of rotting garbage.

Every Y chromosome in my body wanted me to: 1) scream; 2) jump out of the van; 3) grab her; 4) throw her back in; and especially 5) have us take her literally anywhere else in the world. And there was nothing I could do—except pray, I suppose, which I fervently did. For many months afterward, whatever I encountered or endured on a Haiti trip, “no fear upon that way was so present as the fear for what might befall her.”

She survived; I saw her again on a trip in ’13. Once again we gave her a ride back from Petit-Goâve, though we dropped her off somewhere a bit less frightening, close to where she lived, and where she knew people. But in 2021, who knows? Multiply her by tens of thousands, and that is Port-au-Prince today.

---

It is not so much that order is worth fighting for as that chaos is worth fighting against. Ideas of what constitutes order differ too greatly, even within American society; see Fischer’s Albion’s Seed. And it is no accident that politically-correct movements are “anti-fascist” or “anti-racist,” or that CAGW alarmists now prefer the term “climate chaos.” We all know chaos when we see it. Large-scale breakdowns in social order going back to collapsing Mesopotamian city-states have shaped most societies and driven responses ranging from massive centralization to anarchic dispersal. I lean toward decentralization myself—but the great challenge of our time may be finding the institutional balance Biddle discusses in his concluding chapter; from page 309:

… of course, none of these postures can convey very high effectiveness against all possible future opponents at the same time—each sacrifices something in order to enable its strengths in other respects. In fact, the sacrifices posed by the medium-weight posture actually yield a lower peak capability than any of the other alternatives considered …

So let us beware optimization and strive toward an admittedly suboptimal medium. Someday I will write an essay, inspired by Václav Havel and Virginia Postrel, and title it The Glamour of the Glamourless.