Introduction

This article discusses the articulation of the use of the Integrated Food Security Phase Classification tool (IPC) with the registry and media relays. As a global international standard, the IPC is the main mechanism for declaring famine, when the principal users are donor agencies at country, regional and global levels. The IPC regularly publishes figures and categories expressing the size of populations threatened by food and nutrition insecurity by country, and quantifying the gravity of crises. Once officially published, these data are relayed through humanitarian communication channels and by national and international media.

On the one hand, the main argument of the article shows how the IPC is an expert, evolving and perfectible system. Indeed, while famine diagnoses have long been subject to political and media games, due to the absence of an objective framework, the IPC, established in the 2010s, has made it possible to rationalise these extreme crisis situations. Since its creation, the stakeholders of this mechanism have worked to continuously improve the integration of food security and nutrition indicators and the enrolment of parties – as well as the capacity to manage uncertainty. These advances give the system a form of maturity, taking into account the institutional and sociological realities of collective decisions. However, in the case of inaccessible fields, the ambiguity of the IPC data (linked to the lack of field data, foresight exercises, etc.) and the delay of their publication can still contribute to over-interpretation or, conversely, to the lack of treatment by humanitarian communication or the media, as we will show in this paper for the crisis in Yemen (2018) and Madagascar (2021). Hence, the methodological limitations of the IPC in those fields lead us to interpret the figures as ‘poor numbers’ (Jerven, 2013). We will question how the methodological weakness of these data can be mitigated or extended when they are incorporated into humanitarian or media communication schemes, in order to facilitate the launch of appropriate responses to crises.

The article intends to show how the expert and quantified registers of the IPC are a partial element of the logics of humanitarian or media communication, combined with emotional registers, that are not taken into account in the purely technical scheme of the IPC. In a sense, this statement proposes to extend the field of work on the politics of numbers, by questioning the use of numbers beyond their genesis, through their mediatisation. Based on Boltanski’s thesis on the ‘politics of pity’, which puts into tension the ‘topic of denunciation’ and the ‘topic of sentiment’ (Boltanski, 2007), it highlights that the triggering of responses requires the articulation of quantitative data with emotional registers through media and humanitarian communications. These observations lead us to question the gap between the expert and quantitative register of production of the IPC figures, and the media and communication register, as well as the consequences of this gap on the aid response. This reflection invites practitioners to articulate expert knowledge and its mediatisation in a more strategic way, thus strengthening their advocacy policies for appropriate responses to crises.

The author, doctor in socio-anthropology, is a former humanitarian practitioner (MSF, UNICEF). In several areas of intervention, he repeatedly faced a striking discrepancy between local perceptions of crises and their treatment by experts or the media, which can lead to inappropriate decisions regarding the targeting of aid. His work aims at improving humanitarian practices, and at better reasoning the logic of targeting and triggering aid. His research has focused on food aid targeting and early warning systems (EWS), particularly in Ethiopia (Enten, 2008, 2017). The writing of this article is based on a series of interviews conducted with humanitarian professionals (non-governmental agencies (NGOs) and UN agencies)¹ and a bibliographic study on a literature on food crises and EWS.

The first part of the article briefly recalls the stakes of the IPC, underlining that it should reach technical and political consensus both for the production of figures and an official declaration of famine. The second part exposes the elements of media rhetoric to account for a famine through the coupling between the rational logic of the figure and the emotional logic of the image. The third and fourth parts illustrate two cases, in Yemen and Madagascar, leading respectively to an overexposed crisis and a silent crisis.

Famines and ‘Poor Numbers’: Not Strong Enough to Trigger Emergency Aid?

For a long time, the use of the term ‘famine’ has not been limited by any terminology or precise technical criteria. Ambiguity has prevailed between the different theoretical models measuring the scales of severity and geographical gradations of famines. This ambiguity has led to misuse of the term when applied to situations of malnutrition or food crisis (Howe and Devereux, 2004). As a ‘political expression’ (de Waal, 1997), the spectre of famines can be invoked by politicians or aid actors to revive the inertia of public opinion and the international community, tipping into an emotional register partly fuelled by the imagery of past crises. Indeed, the historical famines were characterised by massive deaths² whose ranges were dizzyingly high in hundreds of thousands or millions of deaths,³ and whose last episodes dated from the 1970s and 1980s.

But, since the 2010s, famines have been defined according to technical and quantified criteria, through the analyses produced by the IPC tool, which aggregates information from a number of sources (e.g. about food consumption, nutrition, and mortality outcomes) into five Phases, from Phase 1 (no acute food insecurity) to Phase 5 (famine). Famine is not a rhetorical and emotive term anymore: it is determined by three major quantified parameters: mortality,⁴ wasting⁵ and food consumption.⁶ Then, when various degrees of ‘famine’ could be settled according thresholds⁷ (Howe and Devereux, 2004), the IPC distinguishes different stages of risk assumptions: (1) ‘Emergency Crises’, for extremely severe situations requiring urgent action to save lives and livelihoods, (2) ‘Likely Famines’, when reliable evidence is not yet proven, but some minimally adequate evidence is available indicating that a famine may or will occur, and (3) ‘Famine’, when reliable evidence is collected and validated (IPC, 2021).

This classification is the result of a long process of rationalisation of crisis management, particularly international food aid, whose initial logic was driven by political and economic issues, moving progressively from an initial surplus disposal regime (1950–75) to a development regime (1975–90) and then to a humanitarian regime (1990–2010) (Hopkins, 2009). Intended to rationalise and quantify food security diagnoses in order to prevent crises, famine EWS were created in the 1970s and 1980s by the United Nations (UN) and the United States Agency for International Development (USAID) to counterbalance the political and economic instrumentalisation of aid. But their imprecision, both in the production of data and in its processing and analysis, still left room for political games and the circumstantial interpretations of public actors, humanitarian organisations and the media (Buchanan-Smith and Davies, 1995). While these mechanisms give a scientific stamp to diagnoses and aid policy choices, they are regularly put into question when unexpected food crises trigger institutional crises due to inappropriate diagnoses, swept away by the emotional registers of collective mobilisations (Bonnecase, 2010; Enten, 2017). Today, much of contemporary analysis of famine has been consolidated under the IPC, intending to follow a ‘technical consensus’ process based on the convergence of evidences from various sources to classify the severity of the crisis, and to estimate numbers of people affected, the IPC includes projections – providing a current snapshot and a forecast three and six months into the future (IPC, 2019).

This participatory approach assumed by the IPC stakeholders is part of an institutional maturity of taking into account the value of figures as a result and support for social constructions, with a double objective: a search for stabilisation of a social consensus between the stakeholders around the published figures, as well as an effective demonstration through evidence. Figures are important both for their social value of generating a ‘technical consensus’ and for their intrinsic value (Ogien, 2010). However, recent experience suggests that information gathering and analytical processes may still be subject to political influences that prevent or limit good analysis. In the declaration of famine, a balance must be struck between the fear of a ‘false positive’ versus the fear of a ‘false negative’. Nevertheless, numerous parties have an interest in shaping, influencing and sometimes blocking or suppressing information about these outcomes for political reasons. The threats can come from nearly all quarters – governments, donors and humanitarian agencies and in some cases, local government or representatives of affected communities. As far as consensus-building is ‘one of the most valued yet challenging dimensions of the IPC analysis’, political compromise may affect the results, while field teams involved in the analysis process ‘are not sufficiently senior to manage such pressure’ (FAO, 2019: 4; Maxwell and Hailey, 2021).

Other factors affecting the use of IPC include timeliness in relation to key decisions, and dependent upon the prompt release of its findings. In some cases, tensions between the analyses and findings of expert groups clash with the agendas of governments, leading to the revision or even cancellation of portions of the analyses – as was the case in South Sudan (FAO, 2019; Maxwell and Hailey, 2021). In this case, both analysts and some donors admitted that they maintain two sets of ‘books’ or figures on populations in need: one that is ‘official’ and can be talked about publicly and the other private, but which contains one’s best estimates as to ‘real’ figures. Because, in many cases, the highly political word ‘famine’ is forbidden, as far as states and governments don’t want to admit that crises have deteriorated under their administrations. Famine not only connotes an extreme crisis of food insecurity, malnutrition and death; it also connotes a failure of governance and of humanitarian action (FAO, 2019; Maxwell and Hailey, 2021). In such circumstances of decision blockage, situations are translated into the IPC by intermediate phases of ‘likely famine’ – notably when field data are unavailable or of poor quality.⁸

According to this observation, despite the continuous strengthening of expert and institutional investments, IPC figures related to these inaccessible grounds, still belong to this category of ‘poor number’, as described by Morten Jerven as ‘a number that is a product of a process in which a range of arbitrary and controversial assumptions are made’ (Jerven, 2013: 121). However, in addition to this intrinsic weakness of numbers, we must also question the effectiveness of the IPC in transmitting, popularising and translating information, from a register of expertise and figures to the media and communication actors.

In its Technical Manual, IPC is depicted as providing clear protocols for communicating potential critical situations, informing early relief planning to prevent or limit the severity of forecast acute food insecurity and acute malnutrition (IPC, 2021). Communication is considered an integral part of the food security analysis process, by communicating the core aspects of the situation in a consistent, accessible and timely manner to inform strategic decision-making. The IPC Technical Manual develops in detail all the elements of the communication strategy: the content of the IPC reports, standard tools and supports, associated diagrams and maps, concrete examples. It reminds us of the importance of providing all key figures, numbers and percentages (current and projected situations), as seen in Figure 1 (IPC, 2021).

View Full Size

IPC Communication Template Module 7

Citation: Journal of Humanitarian Affairs 5, 1; 10.7227/JHA.101

• Download Figure
• Download figure as PowerPoint slide

The technical manual’s communication protocols are extremely precise, both in terms of the nature of the data, the dynamic analyses, the rigour of the sources and their degree of imprecision, as well as the formatting and clarity of the messages (map, tables, icons, etc.).

It requires working in close collaboration with the communication professionals of the agencies involved. We are therefore in a double register of expertise here – in the production of the figures, as well as in their translation, in terms of communication. All of this is based on an investment in form according to a rational expert professional logic.

To what extent is this expert and quantified register of IPCs relayed by the media and the communication of humanitarians? In what way is it consistent with the communication logics specific to humanitarians or the media to inform the general public about crises? Or even decision-makers? Are the famines numbers strong enough to trigger emergency aid while crossing the communication and media channels?

Communication Channels and Data: The Mechanics of Storytelling on Famines

A reading of the course of the former Sahelian crises – prior to the emergence of the IPC and having contributed to improved EWS – reveals that the reasons for triggering alerts and emergency food measures are based on a coupling of technical and numerical standards with the emotional register of media treatment of crises. Indeed, the institutional responses to food crises can be subject to brutal irruptions of media hype exacerbating the emotional registers of spectacular and dramatic representations of famine. Then, the media coverage of crises can lead to the questioning of the expertise and institutional responses in place. This was the case in Ethiopia in the 1970s, 1980s and 2000s, or in Niger in 2005, where the media, in contradiction with the EWS, provoked a sudden shift from ongoing actions to emergency actions, and even contributed to the overthrow of the regimes in place (Delcombel, 2008; Gill, 2010). These cases reveal two forms of representation of food crises, which are the spectacular treatment of the media and the technocratic treatment; at the same time they respond to each other, oppose each other, complement each other or feed each other. They are inseparable in understanding the strategic alliances that are played out between institutional actors in the politicisation of food crises. The two registers shape and influence each other both in content and in tempo (Janin, 2010).

This analysis of the combination of registers is supported by the classic thesis of sociologist Luc Boltanski, developed in his book La souffrance à distance, first published in 1993. In it, Boltanski analyses the ‘politics of pity’ underlying the communication strategies of humanitarian and media organisations, which are necessary institutions for linking local crisis situations with the general public. He describes the process of communication as the articulation of ingredients of an ‘emotional coordination’ (2007:109), put in tension through the ‘topic of indignation’, and the ‘topic of sentiment’ and ‘aesthetics’. The first one is founded on a rhetoric of the proof, whose argument is of technical order, even scientific, to unwind a speech at the same time ‘indignant and meticulous, emotional and factual’, but where the emotion remains controlled. The second draws on emotional registers, where ‘emotion makes truth’ without having to ‘unroll an apparatus of material (and) objective proofs’ to put forward the unfortunate person ‘open to gratitude, waiting for help that must come to him’ (2007:109). This topicality proves to be effective in triggering donations, thanks to which the donor participates in collective emotions and adheres to collective frameworks of solidarity (Dessinges, 2008).

The media treatment of the Niger food crisis in 2005 remains a textbook case. It was the words of NGOs and the media, based on the simplistic equations ‘drought + locusts = lack of cereals = famine = malnourished children’ (Delcombel, 2008: 84), that shifted decision-making procedures to an emergency regime. While humanitarians stuck to a technical nutritional argument highlighting nutritional ‘crises’ and ‘emergencies’, the national and international media systematically switched to the emotional and simplifying register of ‘famine’, supported by photos of malnourished children and reports in nutrition centres. Faced with the media pressure facilitated by this simplistic equation, the institutional mechanisms and partnerships in place do not hold up. The focus shifted from technical dialogue to political registers, where ‘the issue at stake was more the response to the humanitarian emergency described by the media’ rather than the one perceived by field actors, including their own agents (Delcombel, 2008: 86). In the end, it turns out that the combination of emotional media coverage in the form of advocacy and rational, numerical formatting reinforces the effectiveness of the appeal for aid, by pressing political levers, often at the expense of information and objectivity (Michiels and Egg, 2007). In an inseparable way, they both produce ‘appeal products’ (Janin, 2010: 18) whose combination accelerates the institutional recognition of a crisis and the triggering of emergency responses. Thus, their effectiveness is measured more in their triggering effects than in their informative and objective quality. The norm of effectiveness of institutional alert of the numerical data, combined with the discursive forms proper to the emotional registers prevails then on its norms of objectivity.

According to Boltanski, the politics of pity must both generalise a given situation and be eloquent in order to mobilise emotion and sentiment. It proceeds from a simultaneous combination of the two registers or norms, relative to numbers and images, through a discourse that must ‘bring out singularity in order to give substance to suffering’ while transporting ‘a plurality of situations of misfortune’ (2007: 36). As generality does not inspire pity, the requirement of affective and emotional investment is based on the singularity of the cases, proceeding to a staging of ‘sufferings made manifest and touching by the accumulation of details’ – which takes place by the staging of close-up photos of children, and the narration of their singular situation (family history, first name, etc.). In order to be able to ‘merge into a unified representation’ (2007: 37), it relies in particular on statistical instruments that meet the requirements of impartiality and generalisation. This quantified objectification thus makes it possible to tear oneself away from local situations, by constituting ‘a sort of procession, an imaginary demonstration of unfortunate people gathered together in what they have in common and what is most singular’ (2007: 36).

We are thus in presence of a communicational mechanics, where the figure must combine with emotional registers to be performative. This leads us to consider emotion as a serious factor in an early warning strategy. ‘The emotion is a sociological object in that it constitutes a tool of communication, of setting in relation with others which accompanies a process of mobilization by devices of sensitization’ through the more or less articulated speeches, the mise-en-scène, etc. (de Sommier, 2009: 202). While since the 2000s, the humanitarian and media drama had diminished in comparison with the stories of the 1990s, which were decried for their ‘often unheard-of display of suffering’ and their ‘indecent’ and ‘scandalous’ character (Boltanski, 2007: 24), sometimes described as ‘humanitarian pornography’ (Pérouse de Montclos, 2009: 762), we are witnessing a resurgence of sensationalist discourse mobilising dizzying figures, superlatives and images of emaciated children, summoning up the collective memory. Indeed, the abuse of victimhood and misery imagery had become commonplace in critical analyses of humanitarian and media communication, where the lines between marketing and testimony were blurred, exacerbating the victim-centred discourse in order to play on the register of emotion and indignation (Collovald, 2001; Hours, 1998; Pérouse de Montclos, 2009; Ryfman, 2008). The narratives were progressively neutralised in the 2000s, attenuating the images towards a euphemisation of suffering, by exposing resilient figures rather than victim figures (Dauvin, 2009). Today, while we are witnessing this return of old emotional tricks in narratives, unlike the 1980s and 1990s, these messages are disseminated on a large scale and almost instantaneously via the Internet and social networks (Dauvin, 2009), while combining a share of technical jargon drawn from international food security assessment schemes (Enten, 2017). Whatever the contexts and motives that justify their use, these types of montage via the image proceed from narrative artifices, which, all in all, remain effective since they are used in a quasi-systematic way by the codes of communication of humanitarians and the media.

If the Niger crisis of 2005 allowed the precision of EWS to evolve by integrating malnutrition criteria, up to their evolution to actual IPCs since the 2010s, how does the integration of IPC data into the media register take place? As a result of the objectification of IPC data production, are we witnessing a rebalancing between the registers of sentimental and denunciation topics? The comparison of current narratives on famines still allows us to draw some elements of a narrative mechanism, the mainspring of which is based, on the one hand, on an implicit analogy between elements whose association makes ‘famine’, and on the other hand, on the use of ambiguous expert references. Information on crises in Somalia (2011, 2017), Sahel (2012), South Sudan (2016 to 2019), Yemen (2016 to 2018) and Nigeria (2017), Kasai (2018) or in Zimbabwe (2019) and Southern Africa (2020) are published both by sites, through the publication of reports or press releases of the UN or NGOs, and then relayed by the media. All the narratives are based on three ingredients: a leap of proportion of the data of populations in the grip of crises, the use of images of children as a symbol of famine and a historical mise en abyme.

We still observe a disconnect between, on the one hand, the technical register and the precautions of nuanced formulations on the part of the UN based on the IPC, and on the other hand, the raw and sensationalist register of the media, and even of some NGO communication messages. The year 2020 began under these auspices, when the press release issued by the World Food Programme (WFP) ‘Southern Africa in the grip of a climate emergency: 45 million people go hungry in the region’ (WFP, 17 January 2020⁹) becomes by La Croix ‘Africa: 45 million people threatened by famine, according to the UN’ (La Croix, 16 January 2020¹⁰): the notion of ‘hunger’ stated by the WFP, still covering the whole spectrum of food insecurity, is interpreted by the media in its most extreme form: ‘famine’. A similar stall occurred in 2017. Pockets of famine were effectively declared by the UN in two counties of South Sudan, affecting nearly 100,000 people, when three other countries were in a food emergency or crisis: Nigeria, Somalia and Yemen, totalling 20 million people at risk (Food Security Information Network, 2019¹¹). The media seized the information, publishing articles with shocking titles, such as ‘The return of famine’ (Ouest France, 26 February 2017¹²), ‘Conflict and climate: why famines are back’ (Le Monde, 28 March 2017¹³), ‘20 million people are threatened by famine’ (Brut, 27 February 2017¹⁴). A consortium (UNICEF, Médecins du Monde, Handicap International, Caritas, Plan International) is launching an appeal campaign ‘Famine 12-12’ explaining the situation of children threatened with ‘starvation’, coupled with videos superimposing images of emaciated children, round numbers and heartbreaking or distressing music. The same simplifying principle is at work in UNICEF’s messages on the web, ‘Famine alert! 1.4 million children in danger of dying’,¹⁵ or in the Caritas donation mailings ‘More than 20 million people threatened by famine in Nigeria, South Sudan, Madagascar…’, for which you have to type ‘famine’ to send donations by phone. We are faced with a disconnect between the initial report – already amply dramatic – of the pockets of famine in Sudan and the accounts deployed by the press and the majority of NGOs, where the overbidding participates as much in a simplification of the media message, as in an erroneous interpretation, leading to staggering figures.

The second ingredient relies on the use of images to trigger emotion, filling the implausibility of the figures associated with the horrific burden of famines. The suffering inscribed in his exhausted body and his distraught look, the child assumes the role of emblematic victim of the crisis. All the more so as the tight or close-up framing suggests, when accompanied by the shocking titles and figures, that this zoom effect on an isolated martyr reflects the whole of the children or the population. This vector makes it possible to pass in a thrifty way, without laborious calculation or medical-technical precautions, from the case of a single child to a cohort, or even to a population of an entire region. Some reportages, such as ‘South Sudan on the verge of famine’ (AFP, 15 April 2016¹⁶) are coupling the poignant image of a child with the numbers ‘2.8 million people and tens of thousands at risk of hunger’. This is the case for the Yemen crisis in 2016 and 2018, where assertions of a risk of ‘losing an entire generation’ or the imminent death of ‘hundreds of thousands of children’ are illustrated with poignant close-up photos of a skeletal child, sometimes named – Salem Abdullah Musabih (The Guardian, 4 October 2016¹⁷) or anonymous (Al Jazeera, 12 December 2016¹⁸ and BBC News, 21 September, 2016¹⁹), or in 2018, when the New York Times illustrated the tragedy of war with a close-up photo of a skeletal child (26/10/2018). The New York Times editorial board justifies this unprecedented use of ‘disturbing’ and ‘brutal’ photography as a way to ‘bear witness, to give voice to those who are otherwise abandoned, victimised and forgotten’ to reveal the ‘horror’ experienced by Yemenis, showing not sanitised images, but images that reflect a degree of suffering that ‘words alone cannot describe’ (New York Times, 26 October 2018²⁰).

Finally, the last element is the reference to previous, recent or historical crises, which allows the imagination to be based on a known and already mediatised situation. These reminders are a way of bringing the situation into line with the ‘pre-conventions’ and the shared imagination of public discourse. Indeed, public discourse is structured by ‘pre-conventions’ and a shared imaginary. Informed by myths and tales, but also historical narratives, fictions and reports, these forms of expression shape the common sensibilities (Boltanski, 2007). These references to these historical facts add a touch of veracity, to validate the facts stated, despite the weakness of the evidence on which they are based. In this way, these memories of cohorts of starving and dead replace the images of numbers necessary to understand the predicted famine, which was missing from the argument at work. The 2017 East Africa crisis referred to the ‘2011 famine that killed 260,000 people in East Africa, more than half of whom were children under 5’ (La Croix, 28 April 2017²¹). The crises in Africa that have ‘left in memory some of the worst humanitarian dramas of recent decades are stated: Biafra (1967–70), Sahel (1969–74), Somalia (1991 then 2011), Ethiopia (1983–85), Sudan (1998)’. In 2020, references to Biafra with photos of the time are still resurfacing (La Croix, 20 January 2020²²). In this way, the current humanitarian narrative is self-citing and self-legitimising, by mobilising the founding and constitutive narratives of modern humanitarianism.

We can therefore observe that the IPC figure constitutes an indispensable, but tenuous, element in these media and communication arguments. Indispensable because it facilitates this objectivation and generalising distancing of the politics of pity – but for all that, taken in isolation from the emotional registers, in particular by allowing the use of the photo to embody the individual suffering at the limit of the bearable. But also, by making reference to the historical famines, constitutive of the imaginary ones associated with the contemporary humanitarian actions. We now propose to illustrate these narrative mechanisms around numbers from the IPC in two situations, one in Yemen and the other in Madagascar.

Yemen 2018: An Overexposed Crisis

The Yemeni conflict pits government forces in the south against Houthi supporters in the north, who have been active since 2004 due to the government’s perceived marginalisation of northern tribes since the country’s unification in 1990. In 2014, the Houthis, allied with forces loyal to former President Ali Abdullah Saleh, seized the capital Sana’a, forcing the resignation of President Abdrabbuh Mansour Hadi, who had been in power since 2012 following the Yemeni revolution and the overthrow of Saleh. Superimposed on this is a struggle for influence between Iran and Saudi Arabia allied with the United Arab Emirates: the conflict became international in 2015 with the intervention of a coalition of Sunni Arab countries led by Saudi Arabia, which pushed the Houthis back to the north, bombing the northern regions and imposing a blockade on the main port of Hodeida. The armed conflict has never made the headlines and has long been neglected by Western diplomats and public opinion. This ‘Hidden War’ stems in part from the complexity of the conflict and its pretence of local and regional alliance games, and the absence of foreign journalists on the ground. All that emerges is a ‘catastrophic discourse on humanitarian needs’, further blurring the image of the conflict and making more measured analyses inaudible (Bonnefoy, 2016, 2018; Mermier, 2017).

At the end of 2016, alerts were launched by the UN and relayed by the press in a bid to outdo each other, combining figures, catastrophic allegations and victimised images. UNICEF relays disproportionate ranges of figures – both in their amounts and their approximations – between 370,000 and even 400,000 to 500,000 children suffering from severe acute malnutrition and between 1.5 and 2.2 million suffering from hunger. These assertions of the risk of ‘losing an entire generation’ or the imminent deaths of ‘hundreds of thousands of children’ are illustrated by poignant close-up photos of skeletal children. In 2017, the dramatisation goes up another notch. At the end of 2017, the UN and many humanitarian organisations denounced the blockade, both as a violation of international humanitarian law, and as a factor of humanitarian crisis, due to the severe restriction of food, fuel and medicine to civilians. The figures are then deployed in order to argue for the urgent lifting of the blockade: ‘17 million people are in urgent need of humanitarian aid’ (ONU Info, 7 December 2017²³), when almost a ‘million (struck by) cholera’ (ONU Info, 22 December 2017²⁴) and ‘seven million are on the verge of starvation’ (ONU Info, 9 November 2017²⁵). Through Jan Egeland, head of the Norwegian Refugee Council, the WFP announces that seven million people are on the brink of a famine of ‘biblical’ proportions. He also calls the crisis the ‘worst famine in modern times’.²⁶

However, the imprecision of the data and methods of the IPC did not allow such clear-cut statements.²⁷ Indeed, IPC diagnoses have been informally criticised by aid actors (CRASH, 2019), and their limitations and biases have also been highlighted (Maxwell and Hailey, 2021). On the one hand, beyond the methodological limitations of the IPC, there are ‘reports of pressure from belligerents on some members’, especially national members, of the diagnostic bodies to influence the figures, most often by inflating them. ‘It is very difficult to monitor [these] indicators, since parts of the country are inaccessible for security reasons and access depends, in particular on the authorities’ goodwill’ (CRASH, 2019). Also, MSF field teams have reported cases of child malnutrition, the severity of which does not corroborate a famine or pre-famine situation. There is also a discrepancy between ‘allegedly very high levels of food insecurity and [relatively] low mortality or malnutrition indicators’ (CRASH, 2019).

According to MSF interviewees, the case of the ‘famine’ in Yemen represents a ‘case of a misdiagnosis and a bad response’. If the first declarations date from 2016, the year 2018 will become a kind of ‘total climax’, which will be the scene of an ‘off-the-wall bidding’ in cataclysmic terms: the UN and the media, diagnosing a ‘risk of humanitarian disaster’. It is a ‘crisis of all superlatives’ unfolding a ‘crushing narrative of famine’ combining a ‘misuse of words and data’, and a ‘hype over image’. Several components would be at work in this staging: a methodological imprecision of the IPC, a will of the UN to force a peace agreement, a media hype playing of the overbidding, and the fundraising of the NGOs that continue to treat ‘famines’.

These announcements are then over-dramatised by the press, exposing a photo of an emaciated child against a striking title to mention the release of international aid ‘The famine in Yemen’ (Le Monde, 25 April 2017²⁸), or highlighting an apocalyptic countdown with ‘a child dying every 10 minutes’ and a terrifying photo of a hospitalised child, Jamila Ali Abdu (Christian Post, 5 May 2017²⁹). In 2018, it is the words of Save The Children, based on data of equally impressive proportions – and above all both unrealistic and unverifiable – announcing 85,000 children who have died of hunger since the beginning of the war, and estimating that nearly 14 million people are at risk of famine. The New York Times headlines the tragedy of war to denounce the role of Saudi Arabia, where famine is used as a weapon of war, accompanied by close-up photos of skeletal children (26 October 2018³⁰). Relaying statements from Save The Children (21 November 2018³¹), a fake news site (disclose.tv) is even broadcasting the story, further blurring the boundaries between truth and exaggeration. The media have rushed into the breach opened by the UN and certain NGOs, echoing their data and alarmist speeches, without seeking to question them or their methodologies, but by overloading their dramatic character. These media communication logics – just like the logics of communication for fund-raising purposes – persist in borrowing from the ‘unchanged codes of the “skeletal child”’ and are ‘disconnected from other forms of evolution of the narrative on famines’. Still in 2019, UNICEF maintains that 370,000 children are suffering from severe acute malnutrition, but underscoring its rhetoric with images focused on the benefits of aid (UNICEF, November 2019³²).

The interviewees explain that several factors would have led to such an overreaction. On the one hand, the impossibility of conducting solid assessments to produce quality data to announce an imminent famine, due to administrative blockages and security risks in Yemen, resulting in ‘ladle’ IPC estimates that facilitate such extrapolations. On the other hand, a willingness to advocate on behalf of the UN would have generated this ‘staging of a rupture’, by magnifying the features of a ‘simplistic and binary narrative’ into an equation ‘embargo = malnutrition = famine’, ignoring the war economy, and its speculations, black markets, especially fuel. Then, the positions of ‘primary anti-Saudi’ journalists accusing Saudi Arabia of causing famine. Finally, the limited access of journalists to the country controlled by the Houthi authorities has allowed the UN to maintain its monopoly on narrative, with the media becoming a sounding board for facts and figures that are difficult to verify.

Finally, this declaration and media coverage of the famine made it possible to launch ‘the world’s largest humanitarian operation to respond to the worst humanitarian crisis’, whose ‘impact is immediate’, because ‘in half of the districts facing famine, conditions have improved to the point where families are no longer at risk of starvation’ (ONU Info, 21 August 2019³³). But without being able to control distribution, the first reports of food-aid diversion and explicit corruption by the Houthi administration are becoming evident, with the latter multiplying the denial of access to populations for humanitarians, blocking aid convoys, interfering with beneficiary lists, etc. In May 2019, the WFP threatened to suspend aid to Sana’a because of ‘the obstructive and uncooperative role of some of the Houthi leaders in the areas under their control’.³⁴

How then to denounce the narrative imposture, when the situation is really serious? How to create a ‘counter-narrative’? This is what MSF tried to do, through press conferences, when their knowledge of the field based on hospital admission figures contradicted these famine warnings, to declare that ‘there is no imminent famine in Yemen’. Their position was not relayed by the media beyond the humanitarian community. But above all, ‘once the one-upmanship is established, one can no longer qualify’: a discourse structured by a quantified argumentation becomes inaudible in the face of the powerful and well-orchestrated ‘topic of sentiment’ fuelled by emotional imagery and raw data, and spreads like ‘a headless chicken that continues to run’, where the public only retains ‘the overwhelming reality’.³⁵

This case study shows how, in a context where data collection is hampered, IPC analyses lead to reliance on partial data, and to the production of scenarii based on conjectures rather than on factual cross-checks. This pragmatic approach to producing IPC data is not in itself open to criticism. On the other hand, when the figures are publicised, it leads to the ‘topic of denunciation’ being crushed by the ‘topic of sentiment’ through the use of spectacular imagery and the imaginary ‘biblical famine’. All of this contributes to evacuating any nuance or prudence from the media content, and from what should imply an objective approach claimed by the IPC to rationalise decision-making in crisis situations.

Madagascar 2021: From Silent Crisis to Climate Crisis

The Grand Sud in Madagascar, the rural populations of which depend on subsistence agriculture and zebu breeding, has remained ‘on the sidelines of development’ (Morlat and Castellanet, 2012). Suffering from a lack of accessibility, this poorest geographic area of the country is also affected by meteorological variations characterised by strong rainfall and spatial variability, resulting in chronic droughts and floods (Bidou and Droy, 2007). Associated with lean periods, kéré means ‘hunger’ in the Tandroy language, but in order to distance itself from the banalisation of its chronicity, it is translated by humanitarians into ‘famine’ to characterise the worsening of food and nutritional crises. Indeed, since the 1980s, drought has occurred four times, including a serious episode in 1989–92. From 2000 to 2004, already affected by irregular rainfall, family economies were weakened by post-election road blockades. More recently, climatic shocks were repeated in 2007, as well as in 2015, following the El Niño phenomenon. The historical recurrence of crises in the Grand Sud and its coupling with the repetition of kéré–which seem to be ‘part of the banal order of things’ (Janin, 2004: 921) – paradoxically maintain its invisibility in the eyes of humanitarian institutions and the media. The fragmented information context, as well as the geographical and temporal variability of the categories and spaces at risk, make it difficult to characterise the risk of food insecurity, to quantify it based on apparent indicators and to model it on a large scale according to the categories of crises and emergencies (Janin, 2004). This poses a double challenge for the IPC, both in its evaluative function, due to the region’s isolation and size, and in its alert function, due to the invisibility of the kéré due to its banalisation and low media coverage.

The IPC undertook, from 2016, the coordination of pre-existing systems to issue bulletins reporting on situations, regularly classified as Phase 3 (crisis) and 4 (emergency).³⁶ This was still the case in September 2020. Without predicting a disaster, these alerts led the Malagasy government, together with UN agencies, to launch a first response with food distributions and a nutritional and medical care centre. The WFP launched an appeal for donations to reinforce food distributions, and some NGOs on the ground launched individual appeals. In December 2020, the IPC bulletin for the Grand Sud and Grand Sud-Est identified nearly 1 million people in a situation of acute food and nutrition insecurity, still in Phases 3 and 4. The projection for April 2021 called for an increase in the number of people affected to 1.35 million.

In May 2021, while the new IPC report card announced that only 1.14 million people were acutely insecure, it reported a sudden deterioration, with pockets of 14,000 people in a famine situation (Phase 5) that are expected to grow to 28,000 people by the end of the year. This sudden jump in severity, which was revealed late in May 2021 and which ‘leaves one speechless’ because ‘degradation is generally progressive’ according to one of the NGO members interviewed, calls into question the capacity of the system to anticipate. It was not until May that the Malagasy government officially recognised the risk of famine in a joint declaration with the UN urging the mobilisation of emergency aid. Events accelerated in the following days, when on 15 June, a Development Plan for the Grand Sud was launched by the Malagasy President Andry Rajoelina and on 23 June, the WFP Executive Director, David Beasley, communicated on ‘the spectre of famine that threatens Madagascar’.³⁷

In terms of media coverage, at the end of 2020, posts on Malagasy social networks were relayed by the government,³⁸ but they remained associated with a worsening of the chronic kéré, without turning into a recognition of famine. The tone of the international press was mixed, sometimes evoking ‘drought and food crisis’ (RFI, 1 October 2020; Réunion la 1ère, 12 October 2020³⁹) or more occasionally, announcing ‘famine’ (La Croix, 25 November 2020⁴⁰). NGOs maintain a cautious tone, mentioning ‘hunger’ (Action contre la Faim, 15 January 2021⁴¹) or ‘chronic food and nutrition crisis’ (MSF, 6 April 2021⁴²) following the moderate tone of the UN media (ONU Info, 12 January 2021⁴³). The press took the plunge in January, referring more systematically to ‘famine’ (Le Monde, 19 January 2021; Le Figaro 19 May 2021⁴⁴). The official declaration of famine was published in May (WFP, 11 May 2021⁴⁵) in line with the IPC figures. Some television media show the distress and exhaustion of the population rather than the ‘famine’, explaining the multiple factors of the crisis and giving voice to the incomprehension, even anger, of local authorities in the face of the inertia of governmental and humanitarian institutions (France 24, 15 May 2021⁴⁶). In tune with the tempo of official announcements, even if they misuse the term ‘famine’ to describe a food crisis, the media do not question the results of the IPC so much as the delays in warning and aid, without triggering a change in international mobilisation.

A second wave of media coverage took place at the end of June when the country was closed to international journalists due to health restrictions linked to COVID-19. On 21 June, journalist Gaëlle Borgia, correspondent in Madagascar for AFP, TV 5 Monde and France 24, posted a video on Facebook showing women and children eating scraps of zebu leather from the Ambovombe market.⁴⁷ These elements show the extreme of people’s destitution, the ultimate testimony stripped of any technical and quantitative register from the IPC. The sequence quickly made a buzz and provoked a reaction from the Malagasy authorities, who orchestrated a campaign of denigration through the press to accuse it of anti-government propaganda.⁴⁸ Gaëlle Borgia’s post created a polemic showing the authorities’ reluctance to recognise the crisis and the media’s difficulty in covering the events. This controversy led the media to take a renewed interest in kéré as, on 23 June, the Executive Director of the WFP, David Beasley, sounded the alarm by calling on ‘the world not to look away’ from families suffering from hunger. His rhetoric uses a ‘topic of sentiment’ that echo the images of Gaëlle Borgia. According to him, ‘the very harsh reality prevailing in the south of Madagascar would bring tears to the eyes of the most hardened humanitarians’. But above all, he points to the direct responsibility of climate change in the crisis. This allegation was not justified by a demonstration of evidence or any data, not even from the IPC. Coupled with the power of the images broadcast at the same time on social networks, it provokes a renewal of the media coverage of the crisis, this time under the angle of climate change. On 29 June, France 2 reported on ‘Madagascar, a famine caused by global warming’, while on 6 July, France 24 reported on ‘a food crisis caused by exceptional climatic phenomena’ and Libération followed up on 9 July with an article ‘Madagascar hit hard by famine’.⁴⁹ At the end of the year, before the upcoming lean season in November 2021, the UN media⁵⁰ and international media⁵¹ exposed a ‘climatic famine’ with its cohorts of ‘climatic displaced persons’ and their meals made of boiled leather. This discourse on ‘climate famine’, placing climate change as the cause of all evils, would be at the centre of a statement by the Malagasy president at COP26 in November 2021.⁵² This reading tends to relegate to the background, or even forget, the political causes of a crisis that also originates in the lack of infrastructure and the isolation of the region. It also highlights the limitations of the IPC – those indicators measure only the impacts of these shortcomings in development policies. Finally, this univocal reading will be qualified by the researcher Paubert T. Mahatante⁵³ and will be challenged by a study concluding that the prolonged rainfall deficit is not so much caused by anthropogenic climate disruption as by natural variations (World Weather Attribution, 2021).

This review highlights the delay in assessing the crisis by the IPC, whose diagnoses in December 2020 only projected scenarios that fall short of a ‘catastrophe’, which were only made official six months later. In the meantime, humanitarian communication remained moderate, surfing on technical terms faithful to the IPC, when only a few media outlets evoke the risk of ‘famine’. Once the crisis was made official, the media accelerated, amplified by social networks with messages that were devoid of any expert register, but underpinned by a questioning of political decision-makers. The crisis took several forms over time: a ‘silent crisis’ until the end of 2020 when, together with the publication of the IPC, it became a ‘controlled crisis’ through the media actions of the state, and then a ‘climate famine’ when the latter, after the publication of the IPC, recognised the famine. The latter removed any direct responsibility of national authorities or international aid actors behind the fatal invocation of the force of climate change. Climate joined the panel of collective imaginations, capable of swaying opinions, just like the references to historical famines – without having to resort to quantified arguments.

If this gap between the temporality of the emergency and that of institutional responses could not be bridged by the IPC, partly because of inaccuracies and delays in diagnoses, it is also notable that NGO and media communication did not make the kéré crisis audible on an international scale. Indeed, the Grand Sud chronic crises remain a subject ‘subcontracted by the media’ giving them little coverage. Several factors contribute to this. First, the people interviewed emphasised the wear and tear of institutions in the face of the chronicity of crises in the Grand Sud, which makes them insensitive to acute crises. Some describe the kéré as a ‘silent crisis’ because ‘for years, there has been a kéré alert during the lean season, creating donor fatigue’. The December 2020 IPC results were interpreted more as ‘routine’ and ‘background noise’ rather than an announcement of ‘real kéré’. Second, semantic confusions contribute to making the crises inaudible. The reductive use of the term kéré by institutions was cited, systematically associating it with ‘famine’ when, for the peasant of the Androy, the term expresses ‘hunger’. This misunderstanding turns into a caricature because ‘even during surveys, the peasant uses kéré to say that he has just not eaten for a day’, leading the investigators to interpret it as a symptom of famine – but made little credible by the institutions because of its chronicity. Third, the incompatibility of reducing the complexity of a crisis and its treatment by the IPC to simple language is also cited. For the Malagasy media, and even for government interlocutors, there is confusion between food insecurity and malnutrition, which are translated into a single expression in Malagasy.

On the other hand, for the media or NGO communicators, the competition between traditional and new media to ‘seduce’ young audiences requires a ‘balancing act of popularization’ between ‘scientific approach and simplification, without falling into a binary discourse’.⁵⁴ The differences between Phases 4 and 5 are not very audible and ‘the journalist will tend to use the word famine according to a definition from the dictionary’⁵⁵ knowing that ‘it will have more echo’ by making shortcuts. Factors related to communication techniques also contribute to the simplification of messages. Indeed, in a similar logic of capturing an audience of followers of influencers, humanitarian communication tends to reproduce the rules of instantaneous staging, spontaneity and proximity with the ‘beneficiaries’, giving the floor directly to ‘the guy in the tent’ without going through the filter of expert discourse and quantitative analysis, or even freeing itself from the circuits of internal validation and the fussy prudence required by the latter. Finally, the diffusion on social networks also formats the text and the images to ‘stick’ to the algorithms allowing visibility on Facebook. If these communicational logics of famine 2.0 take us away from the expert translations of the IPC, they can also bring us closer to the realities of the kéré, as the episode of Gaëlle Borgia’s post shows us, upsetting the Malagasy media and humanitarian establishment.

However, the flow of competing information to be treated by the media gives little room to marginal and ‘lead’ subjects, especially in the editorial offices, where ‘tropism makes us focus on other regions’. In this regard, the pandemic in 2020, in addition to having monopolised a ‘news obsessed with the management of COVID’, has prevented journalists from going to the field, due to the suspension of international flights and health rules, limiting media coverage of the crisis cut off from ‘live’ testimony. Indeed, regardless of the critical analysis of the IPC, but impacting the weak echo of its results on the media scene, the closure of the country’s air borders in March 2020, after the announcement of the first cases of COVID-19, prevented the arrival of international journalists, as well as the deployment of humanitarian operations. The closure of borders, weariness and competition for information and media did not encourage media coverage of the crisis, which remained weak and isolated compared to the crisis in Yemen in 2018.

This case study highlights the choreography between the production of IPC figures and their mediatisation, via traditional media and social networks, showing once again how the mobilisation of IPC figures is not enough to generate awareness of the seriousness of the crisis. Initially held back by the IPC’s methodological precautions, the crisis remains a ‘silent crisis’, which the media, despite the use of the term ‘famine’, does not manage to shift into the media register of the emergency. It was not until the publication of the official figures of the IPC that the traditional media hype was combined with a combination of the two topics, using climatic analogies, and with the late buzz of social networks, unrestrainedly deploying the raw emotional registers and the denunciation of governmental responsibilities.

Conclusion

The IPC figures (estimates of affected populations, mortality, malnutrition rates, etc.) and the development of crisis categories (food insecurity, humanitarian emergency and famine) result from a challenge to find a balance between ‘technical consensus’ and ‘political consensus’. Technical consensus to stabilise the decisions by considering the approximations or conflicts of methodologies, the absence of always reliable data, the contradictions of their interpretations and their extrapolations; political consensus integrating the political stakes between the institutions of the IPC on the one hand, and between the latter and the associated governments on the other. With regard to these intrinsic stakes of the IPC, an intermediate category of ‘likely famine’ makes it possible to ensure the early warning function of this system.

These two examples, Yemen and Madagascar, show how the publication of IPC figures requires their mediatisation – and their insertion in public information channels (humanitarian and media communication) to generate urgent responses. In the first case, the association of raw numbers with images and the orchestration of their media dissemination were remarkably effective, enhanced tenfold by references to biblical imagery. In the second case, the social networks took over, stripped of references to numbers but based on the shocking image, transmitting only an emotion doubled by anger. Anger that could be finally channelled by official speeches, by invoking the generic cause of the climate – thus neutralising the political dimension of the crisis and rallying to an operative collective imaginary.

Paradoxically, in the case of Yemen, the IPC figures were not consistent enough to find a technical consensus – but were nevertheless associated with communication campaigns, building rhetoric and a powerful persuasive force, which could no longer be countered by numerical counterarguments. In Madagascar, the figures produced by technical consensus only revealed the severity of the crisis at a very late stage – and their dissemination and impact were overwhelmed by the raw register of social networks, devoid of quantitative register. From one crisis to another, and from one institutional and political context to another, the IPC figure, despite the consistency as a ‘topic of denunciation’, will not have the same weight according to its anchoring, its diffusion and its media translation through the ‘topic of sentiment’ – making it possible to say that ‘there are some IPCs that we sit on’ and others that we do not.

The qualification of IPC data as ‘poor numbers’ stems from an intrinsic constraint assumed by the actors of this device when it evaluates situations in inaccessible areas. But when the use of these data is aimed at communication and mediatisation intended to trigger responses to crises, this limitation of ‘poor numbers’ is maintained, insofar as this system is constructed and justified only by the logic of experts, without taking into consideration the logic of ‘affective’ or ‘emotional’ rationalities of media communication. If the IPC has made it possible since the 2010s to reach maturity in the understanding and treatment of crises by humanitarians, the media system has not changed, using the same emotional strings. This deserves to be taken into consideration by opening the black boxes of its mechanisms in order to integrate, in a more strategic way, these emotional logics in the IPC’s communication strategies, or risk remaining captive of a ‘citadel of experts’ (de Waal, 1997), overvaluing the objectifying power of numbers in triggering responses to crises and famines.