Somalia's evolving political market place: from famine and humanitarian crisis to permanent precarity

ABSTRACT Somalia has a long history of famine and humanitarian crisis. This article focuses on the years 2008–2020, during which governance and aid practices changed substantially and which include three crisis periods. The article examines whether and how governance analysed as a political marketplace can help explain Somalia's repeated humanitarian crises and the manipulation of response. We argue that between 2008 and 2011 the political marketplace was a violent competitive oligopoly which contributed to famine, but that from 2012 a more collusive, informal political compact resulted in a status quo which avoided violent conflict or famine in 2017 and which functioned to keep external resources coming in. At the same time, this political arrangement benefits from the maintenance of a large group of displaced people in permanent precarity as a source of aid and labour.

humanitarian crises and the manipulation of response.We argue that between  and  the political marketplace was a violent competitive oligopoly which contributed to famine, but that from  a more collusive, informal political compact resulted in a status quo which avoided violent conflict or famine in  and which functioned to keep external resources coming in.At the same time, this political arrangement benefits from the maintenance of a large group of displaced people in permanent precarity as a source of aid and labour.

I N T R O D U C T I O N
Somalia has a long history of famine and humanitarian crisis, from drought, conflict and refugee crises in the s and s, to the civil war and famine in , and to protracted conflict and humanitarian crises ever since (De Waal ; Maxwell & Majid ; Majid et al., ).The nature and severity of crisis have changed over time; with severe crisis in ,  and .These crises are not isolated events but the result of historical and structural inequalities in power relations between clans and struggles over control of resources.Production, trade, services and aid have long been controlled by a few powerful individuals or businesses, with benefits often allocated according to clan affiliation.Other groups have been marginalised, discriminated against and exploited, in particular the agro-pastoral populations in Bay and Bakool regions and the riverine farmers along the Juba and Shebelle rivers (African Rights ; Majid & McDowell ; Jaspars et al. ).These groups have shown persistently high levels of acute malnutrition and suffered the most severe famines, including in  and  (Young & Marshak ; Jaspars ).
While some people in Somalia have been persistently vulnerable to crisis and famine, governance, aid practices and the nature of crisis have changed over time.The s was a period of warlords, looting and large-scale diversion of food aid.The s saw the establishment of different governments and the rise of the Union of Islamic Courts (ICU) as well as changes from food aid to cash transfers and nutrition interventions.The year  was the first attempt at a national government since the fall of President Siad Barre in , but the Transitional National Government (TNG) was unable to govern (Menkhaus ; De Waal ).Dominated by interests within the Hawiye (one of the dominant clans), it soon faced resistance.This was followed by a new Transitional Federal Government (TFG) in , which allied with Ethiopia but which faced opposition from a Mogadishu group of clans, militia leaders, civic groups and the ICU, ultimately leading to a resumption of conflict.It was not until  that a broadly accepted Federal Government of Somalia (FGS) was established.The ICU came to prominence from  and briefly held much of southern and central Somalia.The group Al-Shabaab, initially a militia under the Courts' control, became the principal Islamic extremist  S U S A N N E J A S P A R S E T A L . group after the defeat of the Courts in December , and has controlled most of southern and central Somalia thereafter.
This article examines whether and how governance analysed as a political marketplace can help explain Somalia's humanitarian crises and response.The political marketplace is a system of governance where transactions or deals to buy political loyalty dominate institutions, laws and regulations.In other words a system of governance in which political loyalty is bought or sold (De Waal ).The basic premise is that political outcomes are driven by material transactions and/or violence rather than formal institutions or the rule of law.While the idea of governance as political marketplace is not new, its nature varies across countries and changes over time.Among the earliest articulations of politics as a marketplace was Andreski's (), who defined kleptocracy as the trade in political allegiances in accordance with the laws of supply and demand, typically conjoined with gangsterism.Almost all subsequent debate has focused on kleptocracy as politically organised corruption.In this article, we consider the political marketplace as a form of governance in which political loyalties are bought to gain or maintain political aspirations or power, rather than simply being about corruption and personal gain.Over time, a price principle evolved from earlier barter transactions.Political rents, or finance, may be gained from land or natural resources, from being able to assert sovereignty, from external patronage, as well as from using or threatening violence (De Waal : ).Furthermore, the political marketplace has become liberalised, dollarised and internationalised.In Somalia, President Siad Barre's regime in the s and s, can be called a rentier kleptocracy based on corruption, bribery and militarised patronage (based on superpower rents).When he was unable to regulate private political funds, and political contests remained unresolved, his government collapsed (De Waal : ).
During the civil war of the s, Somalia's political marketplace turned into a competitive oligopoly between armed groups, who financed their operations through looting and extortion, involving violent land-grabbing and aid manipulation.With the establishment of a notional government in the early s, politico-military-business groups competed for power.In addition to humanitarian aid contracts, offers of future political positions were a new source of political finance and bargaining (Majid et al. ).The political market also became globalised: with oil exploration, new forms of international aid, security cooperation (counter-terrorism following the War on Terror), political (pay-offs for cooperation), and international crime (human and arms trafficking, piracy), all providing new forms of rentierism (De Waal : ).Since , the FGS has been positioned as a superior actor vis-à-vis its federal member states and within a system designed to distributelargely externalrents broadly enough to buy in everyone (except for those designated as terrorists).This led to elements of collusion within the marketplace to maintain this income stream (De Waal : ).Over this recent period, between  and , Al-Shabaab arguably moved from an initial ideological phasebased on Islamist and anti-Western principles, to one mainly focused on self-preservation and increased taxation.
This article covers the period from - and from -.Covering three famines or humanitarian crises, several changes in government, and the rise of Al-Shabaab, it provides plenty of scope for analysing why crises unfold differently at different times.Our analytical framework links political economy, political finance and famine.In line with the political marketplace toolkit (De Waal et al. ), we start with identifying critical junctures and thus times of substantial political change.Our first period, -, is associated with attempts at centralised government, ongoing violent conflict and two periods of humanitarian crisis and famine.The second period, -, marks a distinct change because of the establishment of an internationally recognised federal government and federal member states, associated with changes in international aid and investment.For each period, we start with an overview of key political and economic developments (including key actors), followed by sources of political finance, and use this to analyse the nature of the political marketplace (including political alliances) and how it links to humanitarian crisis and famine in that period.
Geographically, we focus on southern and central Somalia (including Mogadishu) because that is where most famines and humanitarian crises have occurred.Outside of Mogadishu, this includes Bay and Bakool regions, and the areas along the Juba and Shebelle rivers; inhabited by historically marginalised populations such as the Rahanwein and Somali Bantu.The majority of the displaced in Mogadishu (and other urban centres) are from these areas and populations.The reasons for their vulnerability are discussed throughout the paper.Much of southern and central Somalia has been controlled by Al-Shabaab since .Since  key towns such as Afgoye, Baidoa and Kismayo have come under government control, creating a series of government-held city-states surrounded by rural areas held by Al-Shabaab.For each period, we consider the role of Al-Shabaab within the political marketplace, and in particular how its actions, as well as international policies, relate to famine and humanitarian crisis.
We argue that from -, the political marketplace operated as a competitive oligopoly, and that the associated violence contributed to the  famine.From , it acquired elements of collusion, when an informal political compact was created in which political actors benefited from the development of a status quo which avoids violent conflict or famine, and keeps external resources coming in.It also includes the maintenance of a large group of displaced people as a source of external aid and casual labour.While a severe famine in  was averted, the post- political marketplace is associated with a state of precarity for some.Many displaced people have sold, or been driven off, their land, and live a precarious and uncertain existence as a flexible source of casual and exploitable labour.When we started writing this article, in , the compact was holding but with the possibility of a return to a more competitive political marketplace.Aid organisations once again identified a risk of famine in .


The article partly draws on previous research reported in Jaspars et al. (), which explored recent changes in the political economy of food with changes in governance and in aid practices.Fieldwork was done in in Nairobi, Mogadishu and Baidoa, with in total  interviews.This included qualitative semi-structured interviews with long-term aid workers (), businessmen () (in transport, trade, money transfer), government officials () and IDP representatives ().Given the challenges of operating in Somalia, and issues of security, access and trust, we started by interviewing people with whom we had established relationships of trust over  or more years.They would in turn suggest others, and we continued until information could be triangulated from multiple sources.The first set of interviews was done from  April to  May  with aid workers in Nairobi, many of whom had been involved with aid in Somalia since the s.This helped identify themes and interviewees for the second stage with government officials, traders, transporters, IDPs as well as aid workers.Fieldwork in Somalia was carried out from  June to  July .Half of those interviewed in stage  were Somali and all were Somali in stage .The reliability of data collection was maximised by having a clear analytical framework, and validity by triangulating data collection (asking the same question from different sources, interviews and reports).For reasons of security and confidentiality, all interviews have been anonymised and given a number according to each category (aidworker -, businessman -, etc.).Using these empirical findings, this article presents a new analysis on political marketplace and famine links, also drawing on relevant literature.It builds on the longterm experience of the authors in conducting research in Somalia; including on food aid targeting and distribution (Jaspars ; Jaspars & Maxwell ) and on famine (Maxwell & Majid ).One of us (Guhad Adan) is Kenyan Somali and has participated in a number of research projects on famine and aid in Somalia.
The article starts with a section on the situation of displaced populations in Mogadishu and Baidoa, and their precarious economic, political and social condition.This is followed by an analysis of, first, the political marketplace and famine between  and , and second, the political marketplace and the prevention of severe famine from  and .The next section analyses persistent marginalisation and the functions of displacement, including how and why IDPs became a business opportunity and a key component of the collusive oligopoly.The article ends with a discussion of the period from  onwards, which is characterised by increased displacement and later the risk of famine (in ) and conclusions on how an analysis of the political marketplace can add to our understanding of famine and humanitarian crisis in Somalia.

D I S P L
In the villages we come from, we Somali Bantu or Jareer are being wiped out because whenever there is drought or floods, Abgaal take advantage and buy all our land and literally we will have no place.(IDP representative   Int.) IDPs are used as a human commodity.I think they were forcibly displaced, then taken advantage of in towns, then [formed to create] a cheap labour force.(Government officer   Int.)By the late s and early s, IDPs in Somalia had become a business opportunity: for local camp managers or gatekeepers, for local authorities, for land-owners, business (e.g. in construction, commercial farming, portering and more), and for aid organisations.The number of displaced increased dramatically between - from  to  million (UN OCHA ).This increase happened in particular in  and numbers remained high in subsequent years.In , the total number of displaced was  million (IDMC ), which represents almost % of the population of Somalia (Jelle et al. ).More than , people have been newly displaced since early  (UN OCHA ).Most displaced populations belong to the politically marginalised clans of Bay and Bakool regions and along the Shebelle and Juba rivers.These populations have been historically excluded from systems of governance, services and economic opportunities; they have been subjected to land-grabs andduring conflictlooting, destruction and displacement.They have also regularly been excluded from humanitarian assistance or receive much less than aid organisations plan (Jaspars & Maxwell ; Thomas & Opiyo ).When displaced, their main source of income is labour (Jaspars & Maxwell ; Bakonyi ; Thomas & Opiyo, ).For these reasons, they have historically also been the most vulnerable to famine.The famine of  was most severe among these groups, and again these populations were food-insecure and displaced in , ,  and at the time of writing.
The reasons for displacement are usually reported to be conflict or drought.A number of other factors are also important, however, in particular that displacement offers benefits for a range of actors.The first are gatekeepers for camps of IDPs.Camps in Somalia do not offer UN protection or aid managementthey consist of hundreds of scattered settlements in urban areas, on private land, that are managed by camp leaders, as they are sometimes called, or gatekeepers.Gatekeepers provide IDPs with access to land and try to connect them with aid organisations, and as such provide an essential service (UN Security Council ; Bakonyi ).They also, however, tax displaced populations on any aid received.In our interviews, IDPs generally reported this to be about %.Gatekeepers also benefit from building and service provision contracts (UN Security Council ; Bakonyi ).At district and national level too, huge benefits could be gained by obtaining contracts for aid transport and aid money transfer, and from having a role in its distribution (Jaspars et al. ).The more IDPs in camps, therefore, the more money there is to be made.As one aid worker in Mogadishu commented: 'I saw a huge increase in gatekeepers from  to .They were picking people up in buses.It became an industry … part of the urban economy' (Aid worker   Int.).According to Human Rights Watch (), gatekeeper-ship itself can be bought or sold.A related factor is that IDP settlement increases the value of the land, as access to services such as healthcare and water improves.Combined with an influx in investment and construction, following the establishment of the Federal Government in , the value of land increased even more dramatically.This has resulted in multiple evictions from displaced camps.UN OCHA () reported , evictions between  and .In , most IDPs have been evicted many times (Jelle et al. ).A humanitarian rent economy has developed which allows land-owners and camp leaders to capitalise on aid flows (Bakonyi ).For displaced populations, evictions result in loss of assets and social networks, and physical and mental health issues (Jelle et al. ).
In addition to benefits from attracting and taxing aid, and increasing the value of land, IDPs also present a source of flexible and exploitable labour.The increase in investment and development from  onwards has resulted in a construction boom and a need for labourfor which IDPs offer a cheap and plentiful source.The mid-s also saw an increase in commercial agriculture and export, in particular sesame and lemon, due to increased demand from the Middle East, India and China (Jaspars et al. : ).This increase has been associated with the acquisition and consolidation of land by more powerful groups or, conversely, the sale by marginalised groups (IDP representatives - ; Aid workers -  Int.).Displaced representatives said they sold their land to powerful clans in the area before leaving, in particular in times of drought.Land sales also create labour, including agricultural labour.One aid worker informed us that the displaced in Dollo and Luuq work as labourers on farms or as porters in town (Aid worker   Int.).In Mogadishu, portering, housework or construction appear most common.In other words, whether by aid contracts or taxation, increased land value or cheap labour, IDPs have become an important business opportunity.The remainder of this article examines whether and how changes in the political marketplace can explain this situation of maintaining precarity but not famine.How did a country which has received large volumes of international aid since the s create and maintain  million people in a precarious and exploitable condition?
The years from  to  are characterised by active conflict, displacement, high volumes of humanitarian assistance, and famine.Since before the civil war, a small elite has controlled wealth and power in Somalia, but the nature of these power dynamics has changed over time.From , the changes in governance, aid and foreign investment have led to changing elite resource capture, competition and collusion.This section analyses the political marketplace by examining political and economic change, elements of the political marketplace and the reported causes of famine.The political market was defined by a struggle for power between businessmen (from powerful clans), Al-Shabaab and government, with the government as the relatively minor player.It follows an earlier period in the s in which Somalia's business class had become politically and economically prominent (Menkhaus ), and when the resources of UN Operation in Somalia (UNOSOM) were a key means of business creation and expansion.By , businesspeople essentially controlled telecommunications, remittances, transport and trade, and had considerable political clout.

Political developments: the rise of Islamist groups and violent conflict
This period saw a number of attempts to establish national or federal governments, but also the rise of Islamist movements.The TFG was established in  with Ethiopian and US support but ultimately had little power or control over territory.Big business played a key role in the establishment of the TFG, as they wanted an end to destructive war but not a strong state.Their cross-clan alliances (in order to operate nationally) were key to the elite bargain which enabled the TFG (Hagmann & Steputtat ; Menkhaus : ).Business was also key in the establishment of ICU control in Mogadishu in ; as it involved an arrangement of business figures, clan leaders and Islamic sheikhs.When the TFG proved unable to govern, due to competing interests and internal dysfunction, the ICU brought about a sense of stability and rule of law albeit along very strict Islamist lines.It lowered the cost for businesses, as Islamists operating across clan-lines were more cost-effective than previous strategies of paying for security (Ahmad ).It led to the reopening of Mogadishu port.Similarly, Kismayo port opened and became a source of revenue for Al-Shabaab from - (Hoffman et al. ).De Waal (: ) called the role of the ICU a non-rentier way of regulating political violence.
ICU control of Mogadishu, however, immediately led to an Ethiopian offensive in  resulting in conflict between the TFG and its backers, and the ICU and the Al-Shabaab militia, involving extreme violence.Around , people were displaced from Mogadishu to the north-west of the city in the direction of Afgoy (also known as the Afgoye corridor), leading to a humanitarian crisis and large-scale humanitarian response.Following the defeat of ICU, the more militant Al-Shabaab became the main Islamist force, controlling much of southern and central Somalia by  (including the towns).At the same time, the war against the ICU may ultimately have saved the TFG (Menkhaus : ).It enabled the government to move back into Mogadishu, and the threat of Al-Shabaab yielded a wealth of counterinsurgency funds for the government.This period also coincides with Al-Shabaab's ideological phase, fiercely Islamist and anti-Western, and supporting the historically marginalised population groups in disputes over land.It established a court system widely recognised as more effective than that of the government.It also, however, taxed farmers heavily and banned Western aid (see below).

Aid and political budgets: the importance of aid and logistics contracts
Aid flows increased following the establishement of the TFG in , the Djibouti peace agreement in  (between the TFG and the Alliance for the Reliberation of Somalia, which included ICU and other opposition leaders), and in response to the humanitarian crisis that same year.The US, EU and the UN all stepped up engagement, as they wanted the TFG to succeed (Menkhaus : ).The combination of a new system of government and an influx of resources meant that MPs could put a price on their votes.Both the first TFG President, Abdulahi Yusuf, and the second, Sheikh Sharif (from ), had the necessary business interests to buy votes and could attract external resources to maintain political loyalty.Hagmann () uses the term 'extraversion' to describe this kind of dependent relationship with the external world to appropriate resources and authority.With government based on clan representation, clan will be used to attract resources and trump policy and humanitarian concerns.
Humanitarian aid, in particular food aid, was big business during this period and in turn financed politics.The best example of this is the  scandal of three Somali contractors who colluded with implementing partners and World Food Programme (WFP) staff to divert food aid (see UN Security Council ).In response to the - humanitarian crisis, WFP's operation had changed from . million beneficiaries to . millionnearly a quarter of the population of Somalia (Nicholson et al. ).The three contractors received WFP transport contracts worth $ million.'It turned them into oligarchs' (Aid worker   Int.).Although the UN report was controversial at the time 'the UN Monitoring Group confirmed what everyone already knew: large amounts of food aid had indeed been going astray in South Central Somalia, although the level of diversion came as something of a shock' (Maxwell & Majid : ).As much as half of all food aid to that part of the country was being diverted (ibid.).An earlier study had also shown collusion between businessmen, implementing agencies and WFP, and can be considered as a precursor to the larger-scale collusion and diversion in  (Jaspars & Maxwell ).  Politically, Sheikh Sharif was closely connected with one of the three big WFP contractors, who had a wealth of business interests, including construction and charcoal exports as well as food aid transport (Jaspars & Maxwell ; Hoffmann et al. : ).The later (and current) President Hassan Sheikh was for some time a WFP contractor himself (Jaspars & Maxwell : ).Logistics contracts, in this period in particular for humanitarian assistance, have been an important source of political finance (Elder ).WFP's logistics contractors used the wealth gained from food aid transport as a direct source of political finance.More generally, competition over 'logistics rents' within a small elite has been an ongoing source of instability in Somalia's governance (Elder ).
Sources of political finance expanded with the  famine response.In addition to US, UK and UN aid and security sector support, Turkey became a key donor from  andin addition to humanitarian aidprovided direct government budget support until  (Jorgic & Karadeniz ).The African Union Peace Mission in Somalia (AMISOM), mandated by the African Union Peace and Security Council in , and Ethiopia also provided revenue as part of the counter-insurgency operation against Al-Shabaab's control of southern and central Somalia.Only Dollo (on the Ethiopian border) and Mogadishu were held by the government, leading to a concentration of aid and displaced populations in these towns.This in turn created local strongmen.Mogadishu was notorious for the strength of its District Commissioners (DC), many of whom controlled and profited from access to IDPs (often through links to 'gatekeepers') and aid.The DC of Dolo, Abdirashid 'Janan', was one of the most notorious strongmen, and later became Security Minister of Jubaland (Bader ).
Much aid in response to the  famine was in the form of cash transfers.Food aid decreased because of the WFP scandal described above, because the US proscribed Al-Shabaab as a terrorist organisation in  and because Al-Shabaab itself banned US food aid in areas under its control from .CARE pulled out of south-central Somalia in  and WFP in .Like food aid, cash could also be used as a source of political finance; by influencing who received contracts at the local level (as before), who received aid, and from gatekeeping and 'taxing' displaced populations (Human Rights Watch ;

Humanitarian crisis and famine
Somalia experienced a severe humanitarian crisis in  and a famine in .In -, a number of factors came together to cause humanitarian crisis.In addition to the Ethiopian invasion, battles with the ICU and the displacement of , people, parts of Somalia experienced drought and the entire country faced high food prices as a result of the  global food crisis (Menkhaus ).Somalia was considered the world's worst humanitarian crisis, with . million displaced and . million people in need of emergency assistance (Menkhaus ; Maxwell & Majid : ).Difficulties in aid delivery contributed to humanitarian crisis too.Delivering aid was difficult because, first, TFG administrative and security measures hindered aid provision to the displaced, who were considered an enemy or supporters of Al-Shabaab.Second, uncontrolled TFG security forces and criminal gangs set up about  checkpoints.Third, the US designation of Al-Shabaab as a terrorist organisation made it difficult for Western aid agencies to distribute food in areas under their control (Menkhaus : ).Fourth, aid diversion through collusion between businessmen, implementing agencies and WFP, as indicated above.


The agro-pastoral population of Bay and Bakool and the riverine populations along the Juba and Shebelle rivers were among the most affected, and according to FSNAU had the highest prevalences of acute malnutrition between -.These groups received little of the aid intended for them, when western aid could still go into rural Al-Shabaab controlled areas (Jaspars & Maxwell ) as well as when they were displaced, because aid coming into the country was controlled and diverted by business-political elites.
The famine in  was more severe than the  crisis.Conflict between the TFG and Al-Shabaab continued.Internal leadership struggles weakened Al-Shabaab and made negotiating access more difficult, and a major AMISOM offensive in February , drove Al-Shabaab out of Mogadishu (Maxwell & Majid ).In addition, drought was severe and global food prices were once again high.Obstacles to aid delivery were again considerable.Al-Shabaab limiting access for humanitarian organisations and US anti-terrorist legislation delayed response and reduced food aid.In , Al-Shabaab banned food aid from its territory, and placed restrictions on population movement.WFP left because of the scandal (see above) and security risks.This hugely limited the scope and capacity for responding to the crisis.When famine was declared in July , more than  million people were affected (Maxwell & Fitzpatrick ).With food aid banned, and limited access, aid organisations switched to cash transfers, but even then the hardest hit areasmostly in Al-Shabaab territorycould not be reached To conclude, in  and in , the political marketplace was a violent competitive oligopoly, in which food aid was a key business opportunity and source of political finance.The resources and nature of the political marketplace contributed to famine and humanitarian crisis because the struggle for resources and political power led to violent conflict and large-scale displacement, and large-scale diversion of aid as a source of political finance.This was aided by the concentration of displaced and aid flows in a limited number of cities. Al-Shabaab contributed indirectly to displacement by banning aid in their areas, as did US anti-terrorist legislation, and by taxing populations under its control.Both contributed to displacement, aid concentration in towns, and famine in southern and central Somalia.
This period is characterised by an internationally recognised federal government, and the establishment of member state governments, increased external funding and investment, rapid humanitarian response in  but also a large increase in internal displacement.From , the political marketplace changed from a competitive oligopoly based on resource extraction (e.g. aid and land by force and diversion) into a more collusive oligopoly based on an informal political compact between key actors.Al-Shabaab is located in opposition to the federal system (and arguably led to its creation) but can also increasingly be seen as more ambiguous, sometimes colluding and sometimes contesting the FGS (Majid et al. ), as this section will show.With the introduction of cash transfers (replacing food aid), the role of aid in the political marketplace became more dispersed, continuing the trend of creating local strongmen.Overall, however, humanitarian aid became less important as a source of political finance when development aid, investment, and consequently other business opportunities, proliferated.This section discusses each of these changes.

Political and economic developments; federal government and expansion of aid and business
The year  marks the beginning of an internationally recognised Federal Government of Somalia (FGS) with Hassan Sheikh Mohamud as President.In political market terms, the federal formula establishes the basic regulatory structure for the market as an oligopoly, with the FGS as the superior actor and the Federal Member States (FMS) as subordinate actors.This period saw more finance, more actors and more aid.In addition to foreign aid, the return of diaspora actors increased investment substantially (Hagmann et al. ).Business opportunities expanded: from money transfer, aid and trade; to mobile money, construction (property) and cash crops, all of which continue to be controlled by a powerful but expanding and sometimes new elite (see below).Al-Shabaab remains a major source of violence against FGS and AMISOM (the AU Mission in Somalia), and to people it sees as collaborators, including coercion for payment of taxes and attacks on targets associated with the FGS and its allies.
Between  and , it was driven out of most towns in southern and central Somalia, but it extended its taxation capacity over much of the economy (Hiraal Institute ).During this period, Al-Shabaab appears to have shifted from supporting marginalised groups and production, to self-preservation and a survival policy of increased taxation.It not only increased taxes on harvest, humanitarian aid and checkpoint fees from people in territories it controls (UN Security Council ), but also from businesses.It collected fees for trade and movement of goods through their areas (which traders often prefer because it is safer than through government areas), for cash-crop production, import and export of goods from Al-Shabaab to government areas, and in ports (interview with food trader, Baidoa,  June ).Other fees in government-held areas, including those for hotels, restaurants and aid organisations means that it was essentially running a protection racket.Ahmad et al. () point out that the group also continues to provide some services, such as the courts, as part of a more predictable governance, frequently out-competing the FGS.At the same time, the threat of Al-Shabaab maintains funding for anti-terrorist activities.
International support and investment was symbolised by the Somali New Deal Compact between donors and the Somali government () and later the New Partnership for Somalia () (De Waal ).Turkey continued its influence with contracts for port and airport operations, including the expansion of Mogadishu International Airport (MIA) and the resulting concentration of resources there.The establishment of the 'green zone' for aid and political actors at MIA massively increased the presence of private military and security actors (Norman ).According to government figures, Turkey committed US$ million in aid to Somalia between  and  (Pitel ) and bilateral trade grew from US$ million in  to US$ million in  (Hoffmann et al. : ).The US, EU and AMISOM remained important, with the UAE and Qatar coming in as new actors (opening missions in  and  respectively).The importance of development, security and diplomatic aid was far greater than humanitarian assistance during this period (Majid et al. ).
Construction, oil and cash crops became key business opportunities.The diaspora and the expansion of mobile money have fueled a property boom, particularly in Mogadishu (Hoffmann et al. ).The three big WFP contractors have switched to similar lucrative ventures post- too, including AMISOM and UN contracts for construction, accommodation, transport, port management or fuel storage and supply (Aid workers - ; Businessmen - ; Government officer   Int.).These contracts are key sources of political finance in the post- period, and a source of ongoing competition between elites (Elder ).Food aid often gave businessmen the capital to invest in new more profitable ventures.Another new business opportunity is cash crops.In the last - years, cash crops have been revived as a source of revenue, mainly sesame and dried lemon.This too is controlled by a limited number of businesses, in particular their export, including out of Al-Shabaab areas (which is expensive) (Aid workers - ; Government officials  and   Int.).In fact, a limited number of big companies appear to invest in whatever is profitable at the time, ranging from cash crops and trade to construction, and need to be able to shift operations quickly.Often the same companies are involved in aid, production, trade and politics (Jaspars et al. ).Al-Shabaab is intimately involved as it continues to contribute to displacement, and taxes all of this (and ensures payment with the threat of violence).

Political finance and budgets; a more diffuse political market
Politics-business links to external finance (state or international aid) remained a key source of political budgets for those in government areas (Hagmann ).Being an internationally recognised government did not, however, increase its ability to tax people, as people have little trust in government to provide services As a result, government (federal and local) and alligned militia demand a plethora of informal taxes and fees and politicians need to maintain a source of political finance and patronage to maintain loyalty.Business actors and connections have become increasingly central to political campaigns (Hoffman et al. : ), in particular because of the need for widespread vote-buying (Menkhaus : ).Like Sheikh Sharif, President Hassan Sheikh had a number of business connections, including companies involved in oil exploration and storage, port re-construction and operations (Hoffman et al. : -).The highest political offices benefited from a booming economy in Mogadishu: real estate and servicing private contracts (ibid.).Companies secured access to contracts by placing close allies in government positions and by proposing projects to please donors (ibid.).
Overall, however, a more diffuse or dispersed political marketplace developed from  onwards because of the creation of the FMS and because of the new modalities of cash transfer programming.Cash programming has led simultaneously to a dispersal of power and resources by involving more small retailers, as well as to its centralisation in the form of the money transfer and communications company Hormuud.Compared with central transport contracting, WFP's food vouchers support a large number of small retailers (over  in ).Possibilities for collusion and diversion have continued to exist but at a lower administrative level, including between district authorities, WFP staff and NGOs for contracts (Businessmen , -  Int.).Diversion of aid and or gatekeeping continues (Aid workers  and  ; IDP representative  and   Int.; see also Majid & Harmer ).District authorities can gain power from selecting retailers and through their involvement in selecting beneficiaries.At the same time, the development of mobile money systems, including for aid, has contributed to the rise of Hormuud as one of the most powerful actors in Somalia.Since , Hormuud has been the key agent for aid agencies' mobile cash transfers, and it has invested in construction, electricity, agriculture, food processing and exports, including a monopoly on sugar imports through Mogadishu (Aid workers , -, as well as Hormuud itself, Businessman   Int.).It dominates the Chamber of Commerce (Hoffman et al. : ).This means Hormuud influences much of Somalia's food production and trade, as well as controlling communication and mobile money transfer (including as aid).Hormuud differs from the logistics firms in that it has aligned more closely with the state (in the interest of economic growth and security) whereas logistics contractors remain embroiled in 'dirty' domestic politics based on bribery and corruption (Elder : ).In addition, the telecoms industry is vital to the functioning of Al-Shabaab and vice versa.Hormuud operates extensively in Al-Shabaab areas, and thus provides a substantial proportion of Hormuud's revenue.This also means that Al-Shabaab can indirectly exert pressure on the government via Hormuud.

Prevention of catastrophic famine?
Despite competition for resources and political finance, the period after  saw the emergence of an informal political compact between government, business and aid actors.According to Menkhaus (), the informal rules that govern this collusive oligopoly might include: clan elites have a monopoly on resources in states where they are dominant; restrained political violence and some commitment to rule of law to keep foreign aid and investment coming in; the need to include the diaspora; and that clan-partnerships may be necessary for business.The threat of Al-Shabaab, its role in external support for the federal system and the influx of external resources, are key in maintaining this informal political compact (Menkhaus ).Al-Shabaab itself benefits from the expansion of resources by being able to tax any aid, trade or production.Money-transfer agencies are part of the compact, but in contrast to the food aid contractors before, their strategy appears to be to buy political compliance (rather than loyalties or positions) so that their business can continue to expand.While humanitarian aid became less important as a source of political finance, we argue that a political marketplace has developed over time which involves, and may even have come to depend on, maintaining large numbers of displaced in urban areas.This political compact may have contributed to preventing a severe famine in .Evaluations concluded that famine was averted because of a fast and coordinated response (see for example ICG ; Daniels & Anderson ; DuBois et al. ; Maunder et al. ; Clayton ) but we argue that the changingmore collusivepolitical marketplace was an added factor.
In August , an estimated . million people were in need of humanitarian assistance with . million people severely food-insecure.As before, the epicentre was southern and central Somalia.Around , people were newly displaced.Donors responded within months and, unlike in , the Somali government, diaspora and private sectoras well as international organisationsresponded quickly (Dubois et al. ) and in a coordinated manner (Clayton et al. ).However, there were other key differences with .By , Somalia had a broadly accepted federal government based on an elite bargain or informal political compact.The political marketplace existed within this federal arrangement and was more diffuse (thus less violent), with more actors, more sources of finance and more money; turning it into a collusive oligopoly in which powerful actors (including government, business and aid organisations) maintain a set of rules which restrains political violence in order to keep external resources coming in.Large-scale violence and famine would risk upsetting a status quo in which all elite actors benefit.Federal government and member states receive aid and other external investment, business can bid for a range of lucrative contracts, construction, urban development and commercial farming expands, and Al-Shabaab is able to tax them.Al-Shabaab benefits from the status quoor compactthrough its extensive taxation abilities (on production, trade and aid) and the increased flow of resources from  onwards.It can be seen as part of the political marketplace because the threat it poses maintains external support for the federal system and conditions in areas under its control maintain displacement and aid flowsthus maintaining sources of political finance for government-business actors.Its control over all aspects of the food chain arguably makes it an equal or more important actor in causing or preventing famine, than the government (or international organisations).Within this compact, the marginalisation and exploitation of certain groups has continued, including their displacement to urban areas.Aid provision in urban areas, and poor coverage in rural areas meant aid continued to act as a pull factor (Dubois et al. : ).By late , WFP had been unable to reach , extremely food-insecure peoplemostly in rural areas controlled by Al-Shabaab (Daniels & Anderson ; Maunder et al. ).Despite an influx of aid in urban areas, IDPs from poorly connected clans were excluded or marginalised (Dubois et al. : ).As discussed above, manipulation of aid in terms of diversion by local authorities and taxation by gatekeepers, for those that did receive aid, continued.Hence the need for IDPs to engage in labour opportunities and become part of the expanding urban and commercial farming economy.The exploitation and displacement of these groups has become a business opportunity and part of the political compact or political marketplace.It requires part of the populations to be kept in a state of permanent precarity, to function as a source of cheap and exploitable labour (for example on commercial farms, for construction, or as porters) and to keep aid coming in.It is striking that although severe famine is said to have been prevented, internal displacement has massively expanded and remained high from  onwards.An analysis of mortality data later concluded that more than , excess deaths had occurred between  and  (Warsame One continuity throughout the s and s is who experiences food insecurity and famine.From colonial times, the maintenance of power has depended on the marginalisation and exploitation of particular groups in Somalia.These groups have included in particular the Bantu and Gabaweyn riverine groups along the Juba and Shebelle rivers and the Digil and Mirifle in Bay and Bakool regions.They live in some of Somalia's most fertile areas, but have been excluded from systems of power or governance, until recently, and have been subjected to various forms of land-grabbing (African Rights ; Samatar ).These marginalised clans in southern and central Somalia experienced the most severe famines or crisis in , in , , again in  and are most at risk now.Throughout this period they have been systematically excluded or marginalised in aid distributions.They consistently showed the highest acute malnutrition rates ( ).This section examines how the nature of vulnerability and exploitation of marginalised groups, or weaker clans, has changed over time with changes in the political marketplace.

Changes in land-grabbing
Famine and displacement are closely related to land-grabbing, the nature of which has changed over time.During Siad Barre's regime, it changed from well-connected clans benefiting from state farms along the rivers, to more direct land grabs when land tenure was liberalised (Samatar ).Production along the rivers (banana) was highly commodified and depended on cheap and exploitable labour.The  conflict and famine were associated with large-scale and violent looting and resource capture, including land.The riverine and inter-riverine populations were vulnerable and displaced because they were unarmed.Hawiye (dominant) clans from Mogadishu claimed land along the Shebelle rivers (sometimes as liberators), while the Marehan and Ogaden (also dominant) claimed land along the Juba river (De Waal ).The original landowners became a group of smallholders and labourers, and as a consequence more vulnerable to drought, floods or food crisis (African Rights ).These trends continued throughout the s, while business (including livestock, charcoal and banana export) boomed (Hagmann & Stepputat ).Weaker clans were subjected to violent land grabs, displacement and diversion of aid (Jaspars ).Since , much of the riverine areas and Bay and Bakool have been controlled by Al-Shabaab.While initially Al-Shabaab appeared to protect Somalia's marginalised groups, over time its punitive taxation and the purchase of land by commercial actors forced many farmers to move out.Violent conflict, destruction, displacement, as well as obstruction and exclusion from aid, caused crisis and famine.
From  onwards, the purchase of land by commercial actors and the concentration of aid in urban centres, have been key factors in displacement.Farmers had to sell to more poweful clans because their situation had been made impossiblewithout aid, services or a voice in local government (IDP representatives -,  ; Aid workers -,   Int.).Rather than outright violence, it was a combination of business interests and pressure from dominant clans (often with business, political and aid connections) that forced people to move during times of drought or floods.According to the International Crisis Group (), Somalis identified one cause of famine as insufficient local production due to land-grabbing by businessmen connected to powerful clans and the switch to cash crops.

The development of IDPs as a business opportunity
Diversion of aid has long been a way of gaining funds, boosting authority and rewarding family and friends in Somalia.Much of the aid to Ethiopian refugees in the s and s was diverted (Askin in De Waal ), and manipulation of aid continued during violent conflict in the s (Jaspars ).In , ICRC established soup kitchens because so much food was looted from dry distributions (Jaspars ), but even then, elders registered ghost kitchens (De Waal ).From the late s, taxation of IDPs by gatekeepers became a particular concern (Narbeth & McLean ).During the  famine, aid workers rediscovered gatekeepers because of large concentrations of displaced in Mogadishu where gatekeepers were prominent, and the taxation of aid rife.The number of displaced increased dramatically from  million in  to  million in  (UN OCHA ).By , as seen in the opening section, gatekeeping had developed into a form of entrepreneurship.Encouraging IDPs into camps has economic benefits as it attracts aid and increases the value of land.It appears that businessmen had learnt from the opportunities that displacement provided in .From  onwards, the number of internally displaced has continued to increase and stood at about  million in  (IDMC ), or about % of the population (Jelle et al. ).
Business also benefits from aid that displaced populations bring, in terms of contracts for transport, distribution, money transfer, and from the cheap and flexible labour they provide.In some cases, the same business may be involved in providing aid, construction and agricultural production; the politics-aid-business link has become more dispersed and sophisticated, with more actors, a more diverse business sector, and sources of political finance.Until recently (see below) this arrangement has maintained a situation of precarity for the displaced although famine has been prevented.In addition, from a political perspective, IDP concentrations help maintain the government's hold on towns, as aid provision requires the presence of AMISOM and the Somali National Army.This in turn depends on a continued threat from Al-Shabaab (Harper ).Keeping people out of rural areas can also be seen as a form of counter-insurgency, as it reduces the population under Al-Shabaab's control.As mentioned earlier, Al-Shabaab also benefits in the collusive political marketplace because of its ability to tax business and aid organisations (indirectly) both within and outside of its own territory.The effects of aid, in particular in protracted humanitarian crises, thus go well beyond those intended by humanitarian organisations.In Somalia, there are a number of dimensions to the politicsaid-business link, of which the political marketplace is one.Large numbers of displaced people in  have become part of Somalia's rent-seeking economy and a means of maintaining funds, labour and power.
From , political changes in Somalia and in the Horn of Africa have threatened the informal political compact.These changes stemmed from a new President in Somalia in  (Farmajo) and in Ethiopia from , and a new political alliance among FGS, Ethiopia and Eritrea.The alliance in - contributed to a return to a more aggressive centralising stance  S U S A N N E J A S P A R S E T A L .from the FGS, increasingly asserting its power by force; for example as seen in the regional elections under Farmajo's Presidency (Majid et al. ).Some of the elites that were part of the Federal system in earlier Presidencies were marginalised.For example, Farmajo attempted to bring the logistics economy under state control.According to Elder (), he was rightly enraged about the logistics contracting that had undermined the formation of a legitimate centralised authority.Rent-producing structures such as port, airport, counter-terrorism funds remained fiercely fought over, and pushing for the auctioning of oil and gas concessions became more prominent as a source of political finance (Gundel ).In early , the President's term expired and it took until May  for a new President to be elected, with a period of instabilityand a risk of renewed civil warin between.In , several factors appear to once again raise the possibility of famine.In the  elections, lead businesses won parliamentary and senatorial seats highlighting the ongoing business-politics links.Security deteriorated during the election period, in part manufactured by Farmajo's opponents (Elder ).The delay in elections contributed to a decline in foreign aid (other reasons for a decline in aid to Somalia included changed geopolitical priorities with the Covid- pandemic, Brexit, and increase in demands elsewhere).A decrease in international diaspora remittances since the pandemic, a protracted drought and higher global grain prices as a result of the war in Ukraine, further constrained resources.Displacement remains high as does the exclusion of the weaker clans from aid and services (Majid et al. ).The five years under Farmajo highlighted the risk of the political marketplace to become more competitive with a resultant increased risk of famine.The re-election of Hassan Sheikh may yet lead to a return to a more collusive arrangement.However, it remains to be seen whether a new arrangement can help prevent famine, or change the role of displaced populations.Other risk factors such as access to Al-Shabaab controlled areas, and the international community's inability to find a way of working with them, the challenges of IDP returns, and the current prolonged drought, also need to be considered.This will need to be closely monitored, also taking into account changing regional and global political and economic dynamics.

C O N C L U S I O N S
The aim of this article was to examine whether and how governance analysed as a political marketplace can help explain Somalia's humanitarian crises and response.In analysing both the political marketplace and the nature of humanitarian crisis, we suggest that the violent and competitive political marketplace in  and in  contributed to humanitarian crisis and famine.Not only was this a time of conflict, displacement, and violent resource extraction, but food aid was a key business opportunity and source of political finance.Contracts to transport food aid created massive wealth for already elite businessmen and helped finance political careers, including for two Somali presidents.Politically marginalised groups were excluded or received less aid than aid organisations intended.However, a political marketplace analysis does not explain all aspects of famine.A key part of explaining the famine in  was the political ideology of Al-Shabaab which prohibited aid, and placed restrictions on movement, along with US anti-terrorist legislation.Its role in the political marketplace may have been limited to its policies contributing to displacement, which led to concentration of aid in towns, and increased the scope for diversion.A political marketplace analysis, however, focuses on elites, and a famine analysis requires more in-depth exploration of the vulnerability of particular groups and how they can become part of political and economic processes which create power for some.Elite-level politics alone does not explain who is vulnerable, exploited or marginalised and how.
The political marketplace framework also helps explain why  was not a severe famine.By , Somalia had a broadly accepted federal government based on an elite bargain and which was part of an informal political compact with business and aid actors.With more sources of aid, investment and finance, a more collusive oligopoly was necessary to keep external resources coming in.The manipulation and diversion of humanitarian aid was less important as a source of political finance, and its decentralisation meant it boosted the power of local authorities but also that it had a higher chance of reaching those in need.To maintain the informal political compact the threat from Al-Shabaab was necessary and violent conflict and famine would risk upsetting it (as this would reduce development aid and commercial investment).Al-Shabaab enters the political markeplace by maintaining external support for the federal system, and the need for aid, and thus sources of political finance for government and business actors.Al-Shabaab itself benefited from the elite compact as it was able to tax production, transport, trade and import/export in ports, as well as tax businesses for protection (including in government areas).Lower-level conflict or political instability was the norm during the - period.Post , a more aggressive centralising stance from the FGS, followed by increased instability and delays in elections, as well as a reduction in aid and investment, may have contributed to an increased risk of famine in the early s.
This status quo has requiredor at least functions bymaintaining displaced populations in a permanent state of precarity.This needs close monitoring and further exploration.Large concentrations of displaced populations benefit government and business in a number of ways.Camps yield income through gatekeeping and by increasing the value of land.Businesses benefit through contracts to supply or distribute aid or because IDPs are a source of cheap casual labour, whether for construction or for commercial agriculturewhich were two of the most profitable businesses in .Somali businesses need to be able to invest quickly in whatever is most profitable at a particular time, and for this they need a cheap and flexible labour force.Even though severe famine was avoided, therefore, the collusive political marketplace in Somalia maintains the ongoing marginalisation of Somalia's weakest clans, who form the majority of IDPs.The increase in IDPs between  and  (and  S U S A N N E J A S P A R S E T A L .onwards) coincided with an increase in commercial opportunities.The status quo that prevented severe famine in  maintains historically marginalised groups as IDPs in a permanent state of precarity; with their displacement facilitating land acquisition, providing a source of aid, the cheap and flexible labour force that Somalia's businesspeople benefit from, maintaining the government's hold on towns and benefiting Al-Shabaab through its ability to tax property, trade, movement.As Bakonyi et al. (: ) argued, urban camps in Somalia are globally governed formations of precarity: 'they may be at the edge of the world but they nonetheless remain part of it and are shaped by its socio-political logics and the multiple agencies that contribute to their emergence'.Furthermore, Somalia is not alone in having a large and growing precariatand this study provides one way of looking at their emergence, maintainance and functions in today's world.
This article has shown how a political marketplace analysis can contribute to an understanding of famine and precarity but also that further research is needed to sufficiently capture what is happening in Somalia. Understanding Somalia's protracted humanitarian crisis needs further on-the-ground in-depth analysis of the changes in the political marketplace, the role of Al-Shabaab within it, and the effect on marginalised populations.This includes further exploration on the political economy of displacement and its links with the rural and humanitarian economy.

N O T E S
.In addition, in our discussions and dissemination with UN organisations (including WFP), donors and NGOs, in Nairobi and Mogadishu in February , no one disputed the  diversions and collusion reported by the UN Monitoring Group.The diversions and collusion in  were accepted by all (Majid & Jaspars ).
. How and why aid practices fail to reveal the political dynamics of food insecurity and aid in Somalia is discussed extensively in Jaspars et al. ().
McCullough & Saed ; Jaspars et al. ).Gatekeeping increased in  as famine-affected people had to move to Mogadishu to receive aid (Hedlund et al. ).
(Hedlund et al. ; Maunder et al. ).Just like in our interviewees in , people had sold land, lost livestock and migrated to urban centres in search of aid or work (Hedlund et al. ; Maxwell & Majid ).This led to concentrations of IDPs and of aid provision in Mogadishu and other urban centres, which made its use as a source of power and political finance, and IDP exploitation, easier.Once again, most famine victims were from Bay and Bakool or minority clans along the Juba and Shebelle rivers.
Jaspars & Maxwell ; Majid & McDowell ; Jaspars et al.  S U S A N N E J A S P A R S E T A L .