Tunisia-Libya: Regulating an Open, Rural Border

The Ras Jedir border crossing lies at the northern tip of the border between Tunisia and Libya. The vast majority of the smuggling trade between Tunisia and Libya operates through this border crossing, and has done so for the last three decades. A variety of internationally produced consumer goods are shipped from Turkey or East Asia to the ports of Western Libya, packed on smaller cars, and brought into Tunisia through the crossing. Here, the goods are completely undeclared or mis-declared, with the avoidance of the full import tariff making up a significant section of the profit margins of local traders. The trade is routinized, and occurs in broad daylight. Hundreds of young men earn their livings as drivers between Western Libya and Southern Tunisia. As this usually occurs alongside a payment to the customs agents at the border, media reporting connects the procedure to corruption, painting the region as a “lawless, Wild-West-like town” (Prentis Reference Prentis2018).

Closer observation, however, reveals a fundamental regularity in the procedures at Ras Jedir. For decades, the trade at the border crossing, while officially illegal, has been regulated through an informal agreement between local traders, customs officers, and both the Tunisian and Libyan state apparatuses. This agreement regulates the types of goods that may pass informally through the crossing, their quantity, the means of transport, and the cost of this informal trade—meaning both the practice of under-reporting and the side payments made to border agents. This agreement has not stayed fixed—many of its terms have varied over recent years, based on shifting power balances between the respective actors. However, while the details of the agreement have fluctuated, its claim over the regulation of these features of the trade, has, with the exception of a few months after the 2011 revolution,^{Footnote 7} remained constant.

In order to illustrate, this is the agreement as it existed in April 2017:

• • Traders who were transporting goods with a value of less than 2000 Libyan dinar (USD 1507)^{Footnote 8} plus 150 litres of Libyan gasoline were not required to pay anything at the border crossing, neither the official tariff nor any standardised bribe or other informal fee.^{Footnote 9}

• • Traders who were transporting goods with a value of over 2000 Libyan dinar (USD 1507) would pay 1500 Libyan dinar (USD 1130) to an intermediary on the Libyan side, while 300 Tunisian dinar (USD 125) would be paid to the Tunisian customs officials.^{Footnote 10} While the money on the Libyan side would be divided between multiple groups that exercised some control over the border crossing, including the municipality of Zuwara, the money on the Tunisian side would partly go towards the tariff income of the state, while some would go directly to the private income of the customs agents.

• • Weapons and narcotics were excluded from the deal, and not allowed to be imported through the border crossing. Medicine is only allowed to be imported informally if paperwork by a doctor can be provided. All goods that were brought through had to pass through a scanner at the border crossing to ensure that they are in accordance with this rule.^{Footnote 11}

It is worth noting that this agreement represented one of the more “deregulated” versions of this institution. Its immediate predecessor, in effect from January until April, had set an upper limit per car of goods with the total value of 4000 Libyan dinar (USD 3013) plus 150 litres of gasoline,^{Footnote 12} and a later iteration again introduced a maximum limit of 5000 Libyan dinar (USD 3767).^{Footnote 13} While this agreement only prohibited the informal import of illicit products alongside medicine, agreements that had been in place before had placed additional restrictions on the types of goods that could be brought through the crossing without paying the formal tariffs, mainly to protect goods that politically connected businessmen in Tunis had an interest in protecting from competition through informal trade. This included almonds, pistachios, high-value electronics, and bananas.^{Footnote 14} Earlier agreements had also included restrictions as to the type of cars that could be used for informal trade through the border crossing—this boiled down to a prohibition against the use of large transporters, as the use of smaller cars helped to employ young men from the region, and made the practice less accessible for traders based in Tunis.

The agreement described here is unwritten, but knowledge of its details is widespread—a wide variety of traders in different locations and networks were able to recite it to me. However, an earlier version of this agreement was even written down—between January and April 2017, a text existed that had been produced through negotiations including municipalities from Western Libya, local smugglers from Tunisia, militias, security forces, and one member of the Tunisian National Assembly. A translation of the agreement and list of its signatories can be found in online appendix 2.

Morocco–Melilla: Regulating an Open, Urban Border

Smuggling in North Africa is not limited to the rural periphery. The Spanish enclave of Melilla is highly urbanised, covering less than five square miles of territory, bordering the Mediterranean in the North, and Morocco in the South. Due to concerns about migration, its 6.8-mile border has been heavily securitised in recent years. And yet it is estimated that at least half of the goods that arrive in the port of Melilla, virtually tax-free, are brought into Morocco informally, bypassing any official trade or tariff regulations (Castro and Alonso 2014). The most common goods traded this way are textiles and foodstuffs, as well as a variety of European consumer goods. While some scholars have described the procedures around Melilla as “a space of lawlessness” (McMurray Reference McMurray2001, 123), the situation is in fact similar to that in Ben Guerdane—upon closer observations, clear rules and regulations emerge.

The Beni Ensar crossing is named after the Moroccan suburb that borders it. It contains three lanes going into Morocco: one for cars, one for normal walking traffic, and one for smugglers on foot. Late at night, Moroccan drivers start lining up in a long queue of cars in the Moroccan town of Beni Ensar, to be let into Melilla early in the morning, and drive to one of the countless warehouses just on the other side of the border crossing. They fill their cars with clothes, tires, foodstuffs, or other imported products, and return to Morocco, where the goods are being sold on markets across the country. As the cars pass back into Morocco, their imports are let through unregistered and untaxed in exchange for a bribe to the Moroccan customs officers. Here, too, bribes are relatively standardised—they are dependent on the value of the goods, and typically range between 50 and 200 Moroccan dirham (USD 5 –20). The cars are inspected by Moroccan customs officers, allowing some estimation of the goods value (and hence the level of the bribe), but also the enforcement of some restrictions on goods that are not allowed to be traded this way—alcohol, for example, or medicine cannot be traded in this fashion. While there is no official restriction on how many cars can engage in this trade, they are limited through standardised times during which the trade may occur—usually between 6:00 a.m. and 1:00 p.m., after which the trade shuts down and the traders load off their goods and get in line for the next day.

Trading on foot is even more common. Two lanes are open for foot traffic. One is for tourists and visitors, and can also be used by small-time traders carrying one bag—their crossing with a small number of goods is legal. For all other traders, there is a special lane. A member of the Spanish guarda civil directs people into the correct lane. The specialised lane for traders is significantly busier, and around half a dozen members of the Spanish police are typically involved in getting traders to line up, and walk through the narrow fenced-off path, through an iron turnstile, and then past the Moroccan customs officers and into Morocco. The payments to Moroccan customs officers, in exchange for not applying the law with respect to the goods being imported, can be differentiated into three categories.

The first group are independent smugglers with very small amounts of goods, often elderly women from the Oriental region who are selling their products to local stores. While they used to pay small bribes to Moroccan customs officials in order to pass, custom officials stopped systematically demanding bribes from them in 2010 and 2011, and have not done so ever since. There are still occasions in which bribes are demanded or goods are confiscated, especially when they are carrying enough to fall into the second category, but the systematic taking of bribes from very small-scale traders seems to have stopped—I will return to this point shortly.

The second group are smugglers who also work independently but are carrying larger amounts of goods. They typically pay a bribe to the Moroccan customs officers, which, similar to trade in cars, is very roughly proportional to the quantity and value of their goods, and usually ranges between 50 and 100 Moroccan dirham (USD 5–10).

The third group are smugglers who work for a wholesaler, a “boss,” and are usually supplied with a “ticket,” which carries their name, the name of their boss, and the goods that they are transporting. As they pass the Moroccan customs officers, they use the ticket to identify themselves as employees of a particular wholesaler. They would hence not have to pay a bribe, as the bribe to the customs officers would be part of a larger, regular arrangement between the wholesaler and the customs officials. The traders would then deliver their goods to a transporter working for the same boss on the other side of the border, with their ticket again identifying them as employees of a certain wholesaler, and specifying the goods that they have to deliver, as well as the payment that they are entitled to receive. For the goods carried by hand, similar restrictions apply as to the ones transported by cars—medicine and alcohol, even sanitary wipes that contain alcohol, are not let through. After Morocco imposed restrictions on the use of plastic bags in convenience stores for environmental reasons, traders reported that these bags also became prohibited.^{Footnote 15} Similar to the trade via cars, trade on foot is typically restricted to the time between 6:00 a.m. and 1:00 p.m. Although they receive no payments from the traders, much of the organisation of the trade is done by the Spanish police at the border, which is primarily made up of riot police from mainland Spain, which are brought into Melilla in two-week rotations. If conflicts break out at the Beni Ensar crossing, policemen draw on the mediation of an informal cross-border trader, whom they have given a police pin to mark her status, and preferential treatment at the border crossing as payment.

The routinisation of this process is even more striking at the Barrio Chino crossing in the southwest of the city, used almost exclusively by large-scale networks traders and their transporters. Barrio Chino has three lanes that lead into Morocco, all of them for foot traffic. Nothing signifies their purpose for illegal trade more clearly than the signs over the entrances to the border crossing: on top of the entrance there are two silhouettes depicting ‘portadores’: one man, and one woman, carrying a large bundle of goods.

Here, too, there are time restrictions for informal trade, typically between 7:00 a.m. and 11:00 a.m.^{Footnote 16} Transporters, so-called “portadores” or “hamala,” wait in lines, under sheet-metal shelters on a large space near the crossing. Trucks then bring goods from the nearby warehouses, pre-packed in bundles wrapped in plastic sheets, and painted with numbers marking them as the property of a particular owner. Organised by the Spanish police, transporters pick up the bundles and carry, push, and roll them up the small passageway to the border crossing. Three metal turnstiles present the key bottleneck of the border—the police let traders approach them in small groups, and local intermediaries, employed by the larger traders,^{Footnote 17} stand at the turnstiles to facilitate a smooth flow of carriers through them. Similar to the Beni Ensar crossing, there are payments to the Moroccan customs officials, although traders who work for one of the bigger wholesalers do not pay bribes, which are organised centrally by the wholesaler.

On the Moroccan side, transporters are ushered to larger vans, in which the bundles are stacked and transported across Morocco. Contrary to Beni Ensar, there seems to be less scrutiny of the content of the bundles at the moment they cross the border. This likely relates to the absence of small independent traders—the traders operating here, as well as their goods, are well known to Moroccan customs and security services. Still, it appears that some goods that cannot be brought through Beni Ensar, such as alcohol, are being traded in Barrio Chino. As in Tunisia, large elements of the regulation observed here are impersonal: anyone with a Nador residency card (that allows the free movement to Melilla without a passport) can work as a transporter, and no particular connections to customs officers are required. Even for people outside Nador, a residency card is easily acquired—transporters from outside the area told me they were able to purchase them for around 2500Dh (USD 250)—many of the “hamala” are not originally from Morocco’s northeast.

Morocco—Algeria: Regulating a Closed Border

The land border between Morocco and Algeria has been closed since 1994, leading to the absence of border crossings as natural points of regulated smuggling as in the two cases described earlier. And yet in the decades following the closure, a variety of points were established along the border that fulfilled the same function. The trade in gasoline, which for many years dominated the informal cross-border trade, is instructive. Gasoline would be collected at Algerian gas stations and then re-filled into jerry-cans in rural farms near the border. They would then be brought by Algerian traders to pre-agreed meeting points along the border. There would usually be a set of about ten potential locations, the one that would be used would be pre-arranged on a day-to-day basis. At these points, the gasoline would be sold wholesale to Moroccan “first buyers” who would transfer it to depots in rural Morocco where it would typically be re-sold to “second buyers” who would transport the gasoline across the country. Both Moroccan and Algerian soldiers were involved in the organisation of these exchanges, with the Algerian soldiers usually playing the dominant role. The soldiers would not only observe the procedure, but also act as mediators in the case of conflict between the traders, and as enforcers of a quiet and orderly exchange. Both Moroccan and Algerian soldiers would receive a payment from the traders. The level of these payments would typically be fixed; for example, Moroccan soldiers would receive 5 dirham for every 30-litre canister that a trader was buying that night.

Another crucial fixed-point within the trade were gates in the border fence on the Moroccan side that the Moroccan military operated in order to organise and regulate cross-border informal trade. A variety of these points existed around the border near Oujda, along with specific times during which they were opened. One gate was opened every night between 7:00 p.m. and 2:00 a.m., another one between 8:00 p.m. and midnight. During this time, Moroccan soldiers would let traders pass across the border and back into Morocco. These crossings did not appear to be operated in coordination with the Algerian soldiers, and continued to operate even when Algeria started to crack down on the trade and increase its own border security. Moroccan soldiers would screen and inspect traded goods as they entered back into the country—at least in the case of new traders. It appears that illicit goods were not traded through these gates; dominant goods here were gasoline, cigarettes, fabrics, and household items. Prior to 2010-2011, soldiers would collect bribes from traders of all goods. In 2010-2011, in parallel to similar changes at the Melilla border crossings, this procedure appears to have changed: small scale-traders, especially small-scale traders of gasoline, were not asked to pay bribes anymore by the Moroccan soldiers, and could pass through the gates for free. Bribes for larger quantities and higher profit margins remained.

Regulation Outside the Nodes

The previous sections have outlined crucial nodes along the borders of North Africa that regulate their porosity, even for illegal trade. However, while the majority of illegally traded goods in the region pass through these nodes, smuggling also exists outside of them. Opting out of trading through the channels mentioned here usually involves less predictability, and requires larger capital, usually in the form of more powerful off-road vehicles or the ability to bribe higher officers in the military. Operating outside of the institutional structures I have described above hence is primarily attractive to those trading in illicit goods that are excluded from the institutionalised smuggling through the nodes discussed earlier, and those trading in large quantities of goods with a high profit margin, who can either negotiate a better deal outside of these nodes, or who are operating profitably enough to be able to finance the risks involved. Between Tunisia and Libya, traders move goods—high-value licit goods, as well as cigarettes, arms, alcohol, and narcotics across the difficult territory along the border, in what locals call the “contra route.” Similarly, on the border between Algeria and Morocco, smugglers try their luck outside the “doors”—many of them moving cannabis-based products grown in Morocco. Regulation along these routes, clearly, is much less visible than along the nodes discussed earlier. Still, there are strong indications that even here, regulation is not entirely absent.

First, the existence of nodes that provide more attractive regulatory conditions for the illegal trade of certain goods influences the type of goods that are traded on the contra route. Microwaves, carpets, or children’s clothes would not be traded illegally outside of the more normalised nodes, which would offer a vastly superior cost-risk trade-off, especially for small-scale smugglers.

Second, in some regions outside of these nodes, relations between smugglers and security forces approximate a cat-and-mouse game, which still contains rules that structure, control, and regulate interactions. Traders, security forces and politicians across these case studies report the existence of “security arrangements” in which smugglers, even as they are trying to evade security forces as they are crossing the border, will still report suspicious activities along the border and the presence of unknown individuals back to the security forces. While some of these agreements seem to have developed out of clientalist relationships with security forces, others seem to have grown historically out of similar agreements with pastoral communities in the borderlands. There is an incentive for smugglers to engage in these activities: increased perception by security forces that the part of the border that they operate in is unmonitored might lead to higher security-force presence there, increasing the cost of smuggling. Even beyond these arrangements, there appears to be a routinisation to these “border games” (Andreas Reference Andreas2009) that defies common characterisations of their interaction as unstructured or “under the radar of the state.” An anecdote serves to illustrate this.

In March of 2017 I was sitting in a roadside café in a small village near the border in Southern Tunisia. I was talking to the only two other customers: a friend who was working nearby, and a local smuggling boss, who has around ten young men working for him. On most nights, they drive across the desert to Libya, and return with a variety of goods, some licit, and some illicit. Most of the traders do not own their own cars, they lease expensive SUVs for one principal purpose: to outrun the local customs agents. “Our cars are much faster than theirs,” the man brags. It is easy for us to verify this from the little café: while many of the smugglers cars are parked next to the café, the customs agents’ cars are parked next to their small headquarter on the other side of the street. While smugglers and customs agents play cat-and-mouse at night, they are sitting right across the street from each other during the day. They know each other. “It’s like a game.”^{Footnote 18} If they get caught, they negotiate a bribe. If they make it through uncaught, they are home free. Once they are home, as the sun rises, their goods are no longer up for scrutiny. The likelihood and price of getting caught is a part of the calculations of the smugglers. Their expected income is subject to a repeatedly played game that every player understands. When one of the customs agents refuses to take bribes, the smuggler recounts, they know his superior and can arrange for him to be transferred elsewhere. And the game continues.

Segmented Porosity

What has emerged from this mapping is that smuggling in North Africa is a densely regulated activity. Informal institutions at border crossings, meeting points along the border, police checkpoints, and markets serve to structure, normalize, and control the trade. They are able to affect the number of goods that come into the country illegally through physical bottlenecks like turn-styles, through regulations on the size of the vessels in which goods may be transported, and time limits on when goods can be traded. Most smugglers trading through Ras Jedir and Melilla are not spending their days evading their police—they are spending their days standing in line. These institutions are able to affect the cost of the illegal import of these goods, through institutionalised bribery arrangements. Crucially, they can differentiate between different types of goods. As the earlier sections have demonstrated, across all cases studied here, different regulatory nodes exist for different goods. A pattern emerges that separates most licit goods from illicit goods: the path that contraband tomatoes—or textiles, or TV sets for that matter—would take across the region’s borders is not available to terrorists aiming to smuggle weapons, or to networks smuggling narcotics. Returning foreign fighters or those on travel blacklists would have serious difficulties passing through these nodes—and even face difficulties outside of them. This is not to say that the smuggling of illicit goods does not occur in the region—it does, and poses serious problems. But its origins do not lie in a general, unregulated porosity of the region’s borderlands, which can be used by smugglers of gasoline and guns alike. Instead, we must note that the porosity of North Africa’s borders, even where illegal trade occurs, is regulated and, in its regulation, segmented.

Observations in the global literature on smuggling suggest that in the prevalence of routinized informal regulatory structures, North Africa does not represent an unusual case. Scholars of smuggling in Sub-Saharan Africa have long conceptualised its interaction with state authority as a process of rule-making and re-making, describing complex arrangements of informal and hybrid regulations (Nugent Reference Nugent2003; Titeca and De Herdt 2010,; Raeymaekers Reference Raeymaekers2012,; Meagher Reference Meagher2014). In her study of the relationship between the Taliban and local business communities in Afghanistan, Ahmad (Reference Ahmad2017) notes how simple and predictable informal tax arrangements contributed to smuggling networks’ support for the movement. Herbert (Reference Herbert2018) highlights the crucial role that informal regulatory arrangements played in organising Mexico’s narcotics trade, and traces the eruption of violence as the power dynamics underlying these arrangements shifted. Tagliacozzo (Reference Tagliacozzo2009) in South East Asia, and Van Schendel (Reference Schendel1993) on the border between India and Pakistan, observe structured informal regulatory institutions with the participation of local state agents.

Characterising Informal Institutions

Despite their normalisation and visibility, these institutions regulating illegal cross-border trade must be classified as informal, as they remain in direct contradiction to the formal law—communicated through the “officially sanctioned channels.” While state actors play a crucial role in their maintenance, they do not do so in an official capacity. At the same time, their systematisation differentiates them from individualised instances of corruption, as the interaction is based upon commonly understood and coordinated rules of the game. This brief survey of informal institutions in the regulation of smuggling in North Africa highlights some contrasts between the informal institutions described here and common characterisations of informal institutions in modern mainstream institutional scholarship. Two points in particular stand out.

First, informal institutions can provide impersonal regulation. As discussed earlier, one of the traditional theoretical distinctions between formal and informal institutions in mainstream political economy rests on the assumption that informal institutions are inherently personalistic. The classic characterisations of informal institutions as small-scale, inefficient, and unwritten ultimately follow from this primary assumption. The analysis presented here does not support this assumption. The informal regulation of smuggling at the Ras Jedir border crossing, at the Melilla border crossings, and at the gates alongside the Algerian border with Morocco are all largely impersonal. They require certain characteristics of the people involved in them, such as Tunisian passports or a residency card for Nador, but this is typical for impersonal institutions. They are not tied to particular personal identities—surely, family connections help in the smuggling business, but they also do in the formal sector.

Second, informal institutions can provide regulation that contains third-party enforcement. Some of the logic behind the assumption that informal institutions are incapable of operating impersonally lies in the expectation that they lack third-party enforcement, that they are self-enforcing, and that those adhering to them are required to avoid persecution when they stand in opposition to the formal law, hence requiring them to operate on a basis of trust. It is worth noting that once more, in the context of the institutions observed here, these expectations do not hold: police, customs agents, and soldiers, albeit unofficially and illegally, systematically act as enforcement agents for these institutions and tolerate illegal activities within the realm of these institutions. While in some cases they appear as corrupt patrons of individual smugglers, in others we observe them enforcing informal rules, either collecting no bribe at all, or collecting a fixed bribe level that has been agreed as part of the institution itself. Crucially, they create interactions that are predictable, and necessitate neither trust nor evasion.

Given that they are impersonal and contain third- party enforcement, it is unsurprising that we also find some of these informal institutions are not necessarily always unwritten: both the institutions at the Ras Jedir crossing and those in Melilla at times included documents that were not only written, but also stamped and signed. It is hence also unsurprising that these institutions are able to operate on a large scale, regulating the activities of thousands of traders, involving values worth millions of dinars and dirham every day. They affect networks that operate not just across borders, but globally—it is common to find young traders in Southern Tunisia or Northern Morocco who have visited China, and remain in close contact with their suppliers there.

The set of informal institutions in political science is wide and diverse, and shifting an argument from one group of informal institutions to the whole set is not always appropriate. However, where personalisation and lack of third-party enforcement have been assumed as characterising features of all informal institutions, these empirical contradictions provide a serious challenge. The observations that informal institutions can operate impersonally and with third-party enforcement can be extended to noting that informal institutions do not necessarily need to be “stateless” in order to be informal. Even when they are in direct contradiction to the rules communicated through the “officially sanctioned channels,” they can still be supported and enforced, informally, by these same channels. While this will not surprise most observers of the politics of regulation, this has been under-appreciated as a systematic starting point in modern institutionalism. This extends the argument of the literature on forbearance on the purposeful non-enforcement of formal institutional regulation to the purposeful enforcement of informal institutional regulation.

None of this is to say that the two categories collapse into each other, even if distinctions such as personalisation and third-party enforcement are removed. The fact that they are created, communicated, and enforced outside the officially sanctioned channels still remains analytically meaningful, and should be retained as the primary substantive feature of informal institutions. All further features, however, cannot be deduced from the institutional type alone, or even the relationship between formal and informal institutions. This is because they are conditional upon the political environment: an analysis of the politics of informal institutions is needed.

The Politics of Informal Institutions

Understanding informal institutions necessarily requires an analysis of the power structures that produce and sustain informality. This includes the motivation, organisation, and power of those affected by the institutional environment as well as the relationship of the state to both levels of institutional regulation, and whether it is directly engaged in non-formal regulation, either through forbearance or through actively shaping informal regulation. Three central questions should be at the heart of a political analysis of informal institutions: How do regulatory actors benefit from the regulation’s informality? What political relationships emerge from this? And how do they remain in equilibrium? In the following, I will discuss these questions, drawing on the empirical context of this paper to demonstrate the insight they can generate into informal institutions and the regulation of illegal economies.

First, why do states participate in informal regulation instead of either repressing or formalising it? And why do members of the security forces participate in the maintenance of informal regulatory institutions rather than maximising bribe levels or crack down at will? Theoretically, a variety of mechanisms could be involved in keeping this power in check, including fear of retribution, capacity constraints, political settlements, or a shared moral economy of appropriateness. These cannot be assumed to have uniform effects on the politics of the institutional regulation, and should form the starting point of a political analysis of informal regulation.

According to interviews with politicians, bureaucrats, and enforcement agents in both countries, concerns about social unrest appear to be a crucial factor in the empirical context described here. “It absorbs the anger and the unemployment,” a high-level bureaucrat in Oujda summarises, “we are buying social peace.”^{Footnote 19} The toleration of some smuggling is framed as an informal subsidy to the borderlands, which have historically seen little formal state investment. This provides a good framework for thinking about the exclusion of illicit goods from these regulatory institutions, as well as the way in which many of these institutions impose maximum quantities that can be smuggled at any crossing at one time, thereby increasing its labour-intensity.

And yet it would be technically feasible for states to set up formal regulations that mimic the informal ones: free trade zones, or geographically limited special trade permits. While the literature on forbearance has highlighted the advantages for politicians in the non-enforcement of certain laws, thinking about the power relations and rents created through informal regulation is equally productive. Informal regulation of illegal activities typically shifts regulatory power towards states’ executive and security apparatuses, often mapping upon the relative power of the latter, especially under authoritarianism. In the cases studied here, security services are able to extract significant revenues from the trade, and bureaucrats interviewed for this project in both countries have identified security services as obstacles to formalisation.

State involvement in informal regulation also facilitates the practice of forbearance in a sector that is security sensitive, such as smuggling, allowing the toleration of certain illegal activities alongside their close monitoring. In other cases, it allows for the toleration and regulation of commercial activities that the state has an interest in maintaining, but would be seen as violating national, cultural, or international norms. It can create a merchant class that is excluded from formal markets that have been captured by politically connected business networks and create barriers to entry that restrict informal rent streams to local networks or survivalist activities. In the context presented here, the restriction against large trucks at the Ras Jedir crossing presents such a barrier. Equally, if rules are unwritten, their communication can restrict access to selected social networks, even if they are impersonal. While changes in the institutions discussed here are usually communicated quickly through word of mouth and social media, the speed of access to new information would still present a barrier to actors entirely unfamiliar with the local context. Finally, there is good reason to believe that informal regulation is also easier and quicker to reverse, thereby creating a relationship of dependency between the beneficiaries of this economy and its regulators.

This suggests a second question: what forms of political relationships emerge from this kind of regulation? As discussed earlier, the classical characterisation of informal institutions as personalistic suggests clientelism as the principal form of interaction with the state (Bratton Reference Bratton2007). While clientelistic relationships are widespread, this analysis suggests that it is neither the only nor the natural political relationship to emerge out of informal institutional regulation. While the literature on forbearance has already widened the space for different forms of political relationships, the empirical cases here support the argument that no forms of political relationships observed in the context of formal regulation should be ruled out in an informal context.

The institutions I describe contain instances of individual corruption, but also of systematic and coordinated strategies across different organs and agencies of the state. The agreement on the procedures at Ras Jedir provides an example: its negotiation included the regional leadership of the Tunisian customs service, a member of parliament, and a civil society organisation founded by local smugglers that has met and negotiated with a variety of government officials, including cabinet ministers and the country’s president.^{Footnote 20} Coordinated procedures for the regulation of small-scale illegal trade across the cases discussed here that do not necessitate the payment of bribes point to a similar dynamic.

The toleration and informal regulation of much of North Africa’s smuggling economies trading in licit goods are best conceptualised as a club good^{Footnote 21} distributed among the population of the periphery in order to secure their acquiescence in the context of a development model that has concentrated formal rent streams in the political centre. For many smugglers interviewed in this project, the politics of their activity are not embedded in relationships of patronage with politicians or members of the security services, but wider agreements that have been negotiated with state representatives and interest groups, and through which smuggling is being framed as a local public good, if not a right. In Tunisia, Morocco, and Melilla, recent years have seen the emergence of formally recognised civil society organisations representing groups of smugglers in negotiations around informal regulatory institutions. In light of these considerations, it is unsurprising that the earlier sections have described cases where policemen and soldiers oversee the smuggling of some goods without collecting any bribes at all,^{Footnote 22} as their participation is not driven by corruption, but by their role in a larger strategy of informal regulation by the state.

At the same time, it is worth noting that the political relationships observed here are not normally distributed over different classes of goods. In parallel with the exclusion of illicit goods from impersonally regulated nodes, the club-good clientelism through impersonal informal regulation does not extend to illicit goods, which appear to be significantly more embedded in individual patron-client linkages. However, rather than assigning features to these goods, this pattern is best examined through the politics of this trade. Here, the networks trading in licit goods are more closely connected to labour-intensive survivalist activities and more palatable to local moral economies, making them more natural candidates for toleration and impersonal regulation by a state intending to utilise them as informal subsidies to borderland communities.

Instances when the informal regulation of smuggling activities in the region has broken down also highlight how deeply embedded it has become in the region’s wider distributional politics. In 2010, 2016, and 2018, disagreements over the procedure at Ras Jedir led to the crossing’s closure by Libyan authorities, crippling the local smuggling economy. In every case, large-scale protests, strikes, and riots broke out, as locals called upon the Tunisian state to either negotiate with the relevant actors to reach a new agreement over the regulation of smuggling, or provide alternative income streams to the region through development projects (Gallien Reference Gallien2018). Every instance ended with the reopening of the border crossing.

This leads directly to a third question: how do informal institutional arrangements remain in equilibrium, and how do they change? More recent works in institutional theory describe how changes in the overall balance of power between groups affected by the regulation impact institutional design and performance (Khan Reference Khan2010). An analysis of institutional change needs to begin with considering the interests and “holding power” of formal and informal actors. However, based on the characterisation of informal institutions discussed here, this is not enough: it also needs to be extended to their informality itself, as different actors may have heterogeneous preferences over how impersonal, for example, an informal institution is.

The institutions I describe offer multiple examples of changing informal institutions. As described earlier, Moroccan soldiers on the border with Algeria stopped collecting bribes from smugglers operating in very low quantities at roughly the same time as customs officers at the border crossing with Melilla did the same. This occurred in late 2010, at a time when the Moroccan government had good reason to be concerned about protests and social pressure, and was framed locally as a subsidy to the poorest traders. Even more significant changes in 2011 occurred at the Ras Jedir border crossing in Southern Tunisia. As the capacity of security forces was weakened in the context of the 2011 revolution, bribe levels at the border crossing dropped to zero, before increasing again from 2014 onwards. It is noteworthy that informal traders, who had financially benefitted from the absence of bribes, formed an association and lobbied state representatives for the return to a fixed bribe level. Anticipating the power of security forces to rebound and fearing a crackdown, there was an interest in re-establishing stable and predictable, albeit costly, regulatory institutions which were in line with the overall balance of power (Gallien Reference Gallien2015). Similarly, a leading official in a relevant ministry described the arrests of multiple smugglers in the summer of 2017 not as an attempt to enforce the law, but to reinstate a “certain equilibrium:” “we need to get back to an acceptable level of smuggling for both the state, and the social peace. That is not an official goal, but that is the goal.”^{Footnote 23}

Similarly, we see examples of the characteristics of informal institutions changing as a result of changes in the balance of power. With the 2011 revolution in Tunisia, the influence of the Trabelsi family, which had used their connection to the presidency in the country to monopolise both formal and informal rent streams, collapsed (Freund, Nucifora, and Rijkers Reference Freund, Nucifora and Rijkers2014). As a result, the trade in selected licit goods that had previously been organised through patronage relationships, such as sport clothing, became regulated impersonally under the wider agreements for the Ras Jedir border crossing. With shifting preferences and shifting levels of trust among Libyan actors involved in the negotiation, we also see the agreement at the Ras Jedir crossing changing from being unwritten prior to 2017, to being written in the winter of 2017, and then unwritten again. Both cases also illustrate the importance of detailed analyses into the preferences and organisational structures of actors involved in informal trade networks—I will return to this point later.

Considering negotiations around the informal regulation of illegal economies provides an insightful perspective on the concept of state capacity. The literature on the distributive politics of enforcement already suggests that the non-enforcement of formal rules does not necessarily reflect on state capacity, identifying the inelasticity of enforcement to increased resources as an indicator of forbearance (Holland Reference Holland2016). Observing the negotiation of informal regulation can help to break down this relationship into its components, including the organisational structure on both sides and their respective motivations, and analyse the regulatory consequences of additional investments in enforcement.

In the context of smuggling networks, such an analysis is particularly relevant in considering the effects of new security infrastructure. Typically, these are imagined as “enhancements of state capacity,” through new institutions in a context of no regulation, or as a new formal institution in a context where they are competing with “subversive” informal institutions. However, if we recognise the real regulatory role of informal institutions in this context, new border security infrastructures are in fact interacting with pre-existing regulatory frameworks. Often, these are not immediately obvious to policy makers, especially from the perspective of the international community, and they risk disrupting informal regulatory arrangements, being bypassed through them, or subsumed into them. As informal trade with Morocco has been profitable for the Spanish enclave of Melilla, the construction of expensive new border fortifications, primarily aimed at deterring migration, have remained porous for smugglers. The effect of new interventions on illegal trade flows hence cannot be predicted without careful consideration of pre-existing regulations, as well as the political economy superstructure that creates and sustains them.