World Development 151 (2022) 105757 Contents lists available at ScienceDirect World Development journal homepage: www.elsevier.com/locate/worlddev ‘To prevent this disease, we have to stay at home, but if we stay at home, we die of hunger’ – Livelihoods, vulnerability and coping with Covid-19 in rural Mozambique Judith E. Krauss a,⇑, Luis Artur b,y, Dan Brockington a,y, Eduardo Castro Jr b,y, Jone Fernando Jr b,y, Janet Fisher c,y, Andrew Kingman d,y, Hosia Mavoto Moises d,y, Ana Mlambo d,y, Milagre Nuvunga d,y, Rose Pritchard e,y, Natasha Ribeiro b,y, Casey M. Ryan c,y, Julio Tembe b,y, Clemence Zimudzi f,y a University of Sheffield, Western Bank, Sheffield S10 2TN, UK b Universidade Eduardo Mondlane, Mozambique c University of Edinburgh, UK d MICAIA Foundation, Mozambique/UK e University of Manchester, UK f University of Zimbabwe, Zimbabwe a r t i c l e i n f o a b s t r a c t Article history: Non-pharmaceutical interventions (NPIs) such as social distancing and travel restrictions have been Accepted 19 November 2021 introduced to prevent the spread of the novel coronavirus (hereinafter Covid). In many countries of Available online 25 November 2021 the Global South, NPIs are affecting rural livelihoods, but in-depth empirical data on these impacts are limited. Keywords: We traced the differentiated impacts of Covid NPIs throughout the start of the pandemic May to July Mozambique 2020. We conducted qualitative weekly phone interviews (n = 441) with 92 panelists from nine contrast- Covid-19 ing rural communities across Mozambique (3–7 study weeks), exploring how panelists’ livelihoods chan- Rural livelihoods Vulnerability ged and how the NPIs intersected with existing vulnerabilities, and created new exposures. Coping The NPIs significantly re-shaped many livelihoods and placed greatest burdens on those with precar- ious incomes, women, children and the elderly, exacerbating existing vulnerabilities. Transport and trad- ing restrictions and rising prices for consumables including food meant some respondents were concerned about dying not of Covid, but of hunger because of the disruptions caused by NPIs. No direct health impacts of the pandemic were reported in these communities during our interview period. Most market-orientated income diversification strategies largely failed to provide resilience to the NPI shocks. The exception was one specific case linked to a socially-minded value chain for baobab, where a strong duty of care helped avoid the collapse of incomes seen elsewhere. In contrast, agricultural and charcoal value chains either collapsed or saw producer prices and volumes reduced. The hyper-covariate, unprecedented nature of the shock caused significant restrictions on livelihoods through trading and transport limits and thus a region-wide decline in cash generation opportunities, which was seen as being unlike any prior shock. The scale of human-made interventions and their reper- cussions thus raises questions about the roles of institutional actors, diversification and socially-minded trading partners in addressing coping and vulnerability both conceptually and in policy-making. Ó 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction prevent the spread of the novel coronavirus (SARS-CoV-2, here- inafter Covid for short). Many have raised questions about the Non-pharmaceutical interventions (NPIs) such as social distanc- equity implications of Covid and associated restrictions in terms ing and travel restrictions have been introduced worldwide to of how they affect the Global South (Klassen & Murphy, 2020; Leach et al., 2021). Emerging empirical research on the impacts of Covid especially in rural settings (e.g. Córdoba et al., 2021; ⇑ Corresponding author. Gupta et al., 2021; Janssens et al., 2021; Mahmud & Riley, 2021; E-mail addresses: judith.e.krauss@gmail.com, j.krauss@sheffield.ac.uk (J.E. Krauss). y Puerta Silva et al., 2020a, b) is confirming prior fears of significant Author names are indicated in alphabetical order. https://doi.org/10.1016/j.worlddev.2021.105757 0305-750X/Ó 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 repercussions for already precarious livelihoods (Ravallion, 2020), 2021 claimed nearly 4 million lives globally (World Health particularly among vulnerable groups (FAO et al., 2020), and new Organization, 2021), prompting governments around the world forms of poverty have emerged especially in Sub-Saharan Africa to adopt NPIs including distancing, masks, and hygiene to prevent (Sumner et al., 2020). transmission. A global recession has ensued, increasing the number We contribute to this discussion with in-depth empirical data of people in poverty globally and particularly in Sub-Saharan Africa from Mozambique that demonstrate across diverse social groups, (Valensisi, 2020). There is a need to understand how the pan- occupations, genders and age groups how the repercussions of NPIs demic’s significant socio-economic repercussions manifest across are manifest within rural livelihoods, i.e. the means of gaining a diverse social groups, ages and genders in light of common concep- living (Chambers and Conway, 1991: 5) in rural contexts. This is tualizations of vulnerability and coping. important not only to illuminate lived experiences with NPIs in Turner et al.’s (2003) seminal framework defines vulnerability rural areas, but also because identifying vulnerability to current as ‘the degree to which a system, subsystem, or system component shocks and stressors produces a starting point for understanding is likely to experience harm due to exposure to a hazard, either a future vulnerabilities to environmental and social change in light perturbation or stress/stressor’ (Turner et al., 2003: 8074). The of the socio-economic and biophysical circumstances (Adger, strength of their framework is its capacity to account for the inter- 1999; Brooks & Adger, 2003; Eriksen & Silva, 2009; Ribot, 2014). dependent, nested relations between human and environmental To this end, we ask two interconnected research questions: influences at a world, regional or place-based level, the way such hazards interact, and the differentiated vulnerabilities of individu- 1) How is coping with Covid affecting livelihoods in rural als and groups within a certain place. In this framework, vulnera- Mozambique? bility is viewed as a function of exposures to stresses and 2) How do these livelihood repercussions intersect with exist- stressors in terms of frequency, magnitude and duration, the sen- ing vulnerabilities, e.g. in terms of differences between ages, sitivity to them depending on interrelated human and environ- genders, occupations, value chains and communities? mental conditions (e.g. different types of capital and endowments), with resilience then determined by coping and We argue that, despite no reported direct health impacts from responses to produce adjustments (Turner et al., 2003: 8076– Covid infection in our study communities during the interview per- 8077). In addition, we take from O’Brien et al.’s (2007) contextual iod, NPIs significantly re-shaped many lives and livelihoods. By vulnerability framework an awareness of how responses are investigating the differentiated repercussions for the lives and liveli- nested within political and institutional structures and changes, hoods of 92 panelists across five different social groups in nine study given their role in shaping crisis contexts. In our analysis, we take communities through weekly qualitative phone interviews from account of the multi-scalar, interrelated, social, political, economic May to July 2020 (n = 441), we show that Covid restrictions have and environmental changes requiring coping and adaptive exacerbated interconnected, existing vulnerabilities resulting from responses by individuals, communities and sectors (Bennett age, gender and precarious incomes, or created new ones. et al., 2016). This is because analyses of Covid-related impacts have To what extent does Covid challenge what we know about emphasized the importance of institutional factors such as state shocks, vulnerability and coping strategies? We will show that support (Dutta & Fischer, 2021), tax exemptions, credit and food institutional NPIs crucially shape vulnerabilities and coping by relief (Kansiime et al., 2021) or existing structures of government interacting with existing exposures and risks across the individual, neglect in shaping vulnerabilities especially in rural contexts household and community levels. As Dutta and Fischer (2021) put (Puerta Silva et al., 2020b). it, institutional responses are as important to understanding Covid In Sub-Saharan Africa (Quinn et al., 2011) and specifically as biological, demographic and economic insights. Fundamentally, Mozambique (Hanlon & Smart, 2008), a multitude of simultaneous, we see Covid not as an idiosyncratic, individual-level hazard, yet as structural, interconnected risks and hazards produce a need for a covariate, i.e. community-level (Dercon, 2000; Günther & populations to prioritize stress responses. This is reflected by Harttgen, 2006) shock. Arguably, it is a hyper-covariate shock, research into Covid effects in Sub-Saharan Africa highlighting a given its impacts on the community level as well as on all commu- particular propensity to see incomes reduced and crisis responses nity links to regional, national or global levels. The all- be required among precarious populations (Janssens et al., 2021; encompassing magnitude of this pandemic (Leach et al., 2021) Kansiime et al., 2021; Mahmud & Riley, 2021; Valensisi, 2020). This raises some questions about the suitability of common coping complex context requires a multi-layered analysis of how environ- and adaptation strategies including livelihood diversification, as mental challenges and economic processes interact in terms of cre- only socially-minded, adaptive value-chain partners protected pro- ating double exposures (Leichenko & O’Brien, 2008), with a ducers from NPI repercussions in our research. However, as our concomitant awareness of the role of human agency and interven- main emphasis is empirical, we invite future research to explore tions undertaken by governmental and non-governmental actors the conceptual implications of our empirically-based conclusions. (O’Brien et al., 2009) in shaping both vulnerability contexts and After defining key terms including vulnerability and coping in responses at the individual and household levels. To account for the next section, we introduce our research design and particularly the temporal dimension of coping, we used Quinn et al.’s (2011) the methodological metamorphoses required under Covid. Based distinction between coping strategies, i.e. shorter-term ways of on our research questions, we discuss in-depth the differentiated dealing with stresses especially linked to economizing on food, impacts of NPIs experienced by our panel, emphasizing variations and adaptation mechanisms, i.e. longer-term adjustments involv- between occupations, ages, social groups and genders, value chains ing alternative work or support from governments or projects. Cog- and communities. The final section discusses the implications of nizant of existing adaptation strategies to ongoing environmental our findings for debates about vulnerability, coping, diversification and economic challenges, such as the production and sale of char- and value chains. coal in some study districts (Eriksen & Silva, 2009; Smith et al., 2019; Vollmer et al., 2017), we thus analyze what short-term cop- ing strategies were used at the individual, community and sector 2. Defining vulnerability and coping in times of Covid levels in response to NPIs. In our analysis, we also recognize that one prominent strategy The highly infectious Severe Acute Respiratory Syndrome promoted by governmental and non-governmental actors to Coronavirus-2 (SARS-CoV-2 or Covid) from late 2019 to June reduce rural vulnerability has been diversification, i.e. pursuing 2 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 multiple livelihood strategies concurrently, given an assumption that diverse livelihoods are less vulnerable than undiversified ones (Ellis, 2000). This assumption has been problematized due to mixed evidence on additional incomes generated (Torell et al., 2017) or distribution of incomes (Gautam & Andersen, 2016), being most effective for better-off households and reliant on a dynamic economy (Ellis, 2006) . Nevertheless, international organizations (OECD, 2007; World Bank, 2013) and scholars continue to support diversification as a progressive livelihood strategy (Martin & Lorenzen, 2016) and/or as a viable safety net, especially for poorer Sub-Saharan African smallholders (Alobo Loison, 2015). A key factor shaping vulnerabilities and coping strategies related to Covid is communities’ and individuals’ integration into markets (Eriksen & Silva, 2009, for one of our study districts) and value chains, i.e. the succession of transactions governing the cre- ation of products and services (cf. Barrientos, 2019; Horner & Nadvi, 2018; Neilson & Pritchard, 2009; Oldekop et al., 2020). Dif- fering levels of integration into networks, market knowledge or capital (Eriksen & Silva, 2009; Jones, Ryan & Fisher, 2016) crucially shape the degrees to which communities can access markets in times of crisis. Equally, the make-up of value chains plays an important role. Convention theory emphasizes that there are con- siderable differences between dominant value-chain stakeholders prioritizing market-based interests around price, or more civic- based regimes emphasizing ethical conduct (Renard, 2003; Krauss & Barrientos, 2021). While in the former case, the key crite- rion is price, affecting negotiations between stakeholders through- out the value chain, civic-based mindsets will prioritize ethical conduct and socio-environmental production circumstances, pro- ducing very different outcomes for value-chain interactions and local livelihoods especially in times of crisis. Finally, Quinn et al.’s (2011) and Pritchard et al.’s (2020) emphasis on environ- mental resources as pockets of agroecological resilience is vital, particularly when simultaneous exposures render some coping alternatives unavailable (Pritchard et al., 2020). 3. Methods Fig. 1. Map of six study districts across Mozambique. Source: Casey Ryan. 3.1. Study area and the Covid context access, and livelihoods, ranging from cultivation and livestock to The authors conducted a total of 441 qualitative phone inter- charcoal production or trading (cf. Table 1). A key difference was views over up to seven study weeks in the period between May the study communities’ exposure to cyclone Idai in March 2019 and July 2020 (cf. Table 1). The research project covered nine com- (Phiri, Simwanda & Nyirenda, 2020), as significant loss of lives munities in six districts across Mozambique (cf. Fig. 1), with the and livelihoods had been sustained in the two study communities study communities represented by abbreviations rather than full in Sussundenga (SMC/SMU), leaving them in a fragile situation names to safeguard confidentiality. even prior to Covid. By contrast, the Mabalane, Mapai and Mabote The study communities, located in differing regions, ecosystems communities lost crops and livestock in the 2015/16 drought, and climatic zones (cf. Fig. 1), vary also in terms of ease of water requiring humanitarian intervention. Table 1 Study districts, communities, panelists, number of study weeks, key livelihood activities and recent major hazards affecting the districts. Source: Authors. Districts Communities # of pane- Study Main livelihood activities Recent major lists weeks hazards Guro & Tambara1, Manica GN 10 7 Agriculture (maize), horticulture, livestock, baobab Province TL 10 7 Agriculture (millet), horticulture, livestock, baobab Mabalane, Gaza Province HC 10 5 Livestock, charcoal, rain-fed agriculture (maize) 2015/16 drought MV 10 6 Livestock, charcoal, flood plain agriculture (maize) 2015/16 drought Mabote, Inhambane Province2 MB 10 3 Livestock, rain-fed agriculture, some irrigation-fed agriculture 2015/16 drought (vegetables) Mapai, Gaza Province BR 10 4 Charcoal, livestock 2015/16 drought MF 10 5 Charcoal, livestock, rain-fed agriculture (maize) 2015/16 drought Sussundenga, Manica SMC 11 7 Agriculture (maize, banana), horticulture, bee-keeping Cyclone Idai 2019 Province SMU 11 7 Agriculture (maize), horticulture (vegetables), bee-keeping Cyclone Idai 2019 1 Guro and Tambara are separate districts, but combined here given their adjacent location. 2 Mabote district has only one panel of participants. It was added after the beginning of the project in other communities, resulting in a lower number of overall study weeks. 3 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 Table 2 Panelists disaggregated by study districts and categories (vertical), gender and number of panelists/interviews (horizontal). Source: Authors. Overall panelists4 # of panelists Interviews F# F interviews M# M interviews Total 92 441 45 205 47 236 Guro & Tambara 20 138 9 63 11 75 Mabalane 20 79 12 48 8 31 Mabote 10 14 5 8 5 6 Mapai 20 56 12 37 8 19 Sussundenga 22 154 7 49 15 105 Vulnerable 17 83 14 71 3 12 Microbusiness 19 91 9 28 10 63 Market-oriented smallholders 14 69 9 43 5 26 Traditional influence 16 74 2 13 14 61 Modern influence 23 119 10 47 13 72 4 Three panelists, who were added after the project start, did not fit into any categories; they and their interviews have not been allocated to any of the categories. The non-pharmaceutical interventions instituted by the cross-check panelists’ profiles1. We also selected communities Mozambican government in late March and early April 2020 to represent a diversity of rural experiences in terms of liveli- affected all study communities. After the World Health Organiza- hoods, climatic and geographical locations, demography and tion identified Covid as a pandemic, a Presidential Decree declared socio-economics. a state of emergency on 30 March. This was ratified by law the fol- The panelists agreed to take part in weekly phone calls with the lowing day, 31 March, with decree 12/2020 by the Council of Min- research team in local languages, compensated by receiving isters on how to operationalize the state of emergency following on phones and phone credit. The social groups to be included in our 2 April (GoM – Government of Mozambique, 2020a). Decree panel were selected to cover a range of livelihoods and experi- 12/2020, across all professional and private spaces, required social ences, and were identified based on prior work in study communi- distancing, frequent handwashing and face masks in public. At the ties (e.g. Baumert et al., 2016; Smith et al., 2019; Vollmer et al., same time, it closed borders, banned activities in public spaces 2017) as well as consultation with partners2. The categories chosen including the consumption of alcohol, suspended religious wor- were: ship, and limited private and public transport of people and goods, including restricting collective transport to 1/3 of occupancy. 1) Vulnerable people based on precarity of income, age, gender Trains were also suspended (Radio Moçambique, 2020; 2) Microbusiness owners, e.g. bakers, small-scale traders and SaudeMaisTV, 2020). With transport thus unavailable or signifi- vendors cantly more expensive, movement was more difficult or impossible 3) Market-oriented smallholders, including charcoal, livestock altogether, exacerbated by a significant fear among panelists of or agricultural producers selling to value chains/markets engaging with any strangers entering their space. Schools were 4) Traditionally influential individuals, including traditional closed during most of the state of emergency and only to be reo- local leaders, traditional healers pened if basic conditions of hygiene were met (Deutsche Welle, 5) Modern influential individuals, including teachers, church 2020) . Subsequently, the state of emergency was prolonged sev- pastors, health agents eral times (GoM – Government of Mozambique, 2020b) and then replaced by an indefinite state of public calamity nationwide from Given the disproportionate effects expected on vulnerable 7 September 2020 onwards (UNICEF Mozambique, 2020). Towards groups and women through the Covid pandemic (Ahmed, the end of our data collection period (May-July 2020), discussions Ahmed, Pissarides, & Stiglitz, 2020; Quisumbing, Kumar, about lifting restrictions had begun, particularly on reopening Meinzen-Dick, & Ringler, 2020; United Nations, 2020), care was churches with limited capacity from mid-August (Vatican News, taken to safeguard inclusion of different ages and genders within 2020). Panelists reported following these restrictions for fear of the ‘vulnerable’ category, and of women across all categories (cf. the pandemic given a lack of available medical treatments against Table 2).3 Covid. The questionnaire design built on prior work (Pritchard et al., 2020; Smith et al., 2019), while also accommodating the specific 3.2. Research design: methodological metamorphoses circumstance of phone data collection (Block & Erskine, 2012). The study used a detailed questionnaire in Week 1 with predomi- The research design selected study communities to reflect nantly closed questions to establish baseline information about diverse rural experiences, based on prior work (Baumert et al., individuals, their livelihoods, households and communities, with 2016; Jones, Ryan & Fisher, 2016; Pritchard et al., 2020; Smith et al., 2019),. Given significant travel and transport 1 Although the research team were initially concerned about potential political restrictions, we needed to work with study communities with influence on the selection of panelists and resulting bias, analytical rigor and diversity which partnerships and rapport had already been built in prior of rural experiences across the five social groups were the key determinants in collaborations. Significant existing rapport was vital firstly to choosing panelists. Despite interviewers and the research team being sensitized to the possibility of influence, there was no noticeable political or selection bias safeguard free, prior and informed consent, secondly owing to identified in terms of answers given. remote data collection by phone, and finally due to the co- 2 Panelists were only allocated to one category in light of what best fit their creative, dialogue-based design (Dearden & Kleine, 2020; situation, i.e. there were no overlaps e.g. between traditionally influential and modern Horvath & Carpenter, 2020), which relied on participants and influential individuals. 3 stakeholders to feed back on the salience of our research 3 to 7 study weeks in different locations depending on availability of interviewees in light of pandemic restrictions, cf. Table 1 for details. For instance, in some sites, enquiry in a novel pandemic. To select panelists and identify there were logistical difficulties with making phones available to respondents. relevant research questions, the research team used prior data- Equally, there were sites with poor mobile reception, requiring alternative sets and also coordinated directly with local governments to arrangements. 4 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 this strategy also employed to ease participants into unfamiliar 4.1. Key changes due to NPIs phone interviews5. Subsequent weeks posed fewer, but more open questions to ascertain changes from the previous week in three Four key changes resulted from Covid and were reported across key respects: information flows, coping strategies with Covid, and the panel: prevention measures, transport restrictions, access to livelihood impacts. In the final two weeks of phone interviews, ques- consumables, and repercussions for money and income. Firstly, tions were added in light of incoming data and feedback, focusing on Covid communication through village gatherings, posters, TV, radio vulnerabilities to Covid and NPI livelihood implications, as well as and project-provided phones and radio programs had successfully collective coping strategies.6 produced awareness of the importance of wearing masks, distanc- In terms of data collection and coding, the researchers con- ing and hand-washing, though information channels hampered by ducted qualitative phone interviews in local languages and subse- social distancing caused erroneous or lack of knowledge especially quently translated them into Portuguese. The summarized among the elderly as discussed below. The heavy reliance on hand- Portuguese versions were analyzed and coded across sites through washing produced additional expenses for soap or difficulties pur- Nvivo 12 following Mikkelsen’s approach (2005): an initial round chasing it (MV-06-02; TL-04-02), though many used ash as a more of open coding, i.e. analyzing initial key themes to inform question- available alternative. Additional expenses were also incurred for naire revisions. This was followed by axial coding, i.e. going face masks. through the data to assign both higher-order categories in light Transport, mentioned 321 times across our 441 interviews, and of the research themes and questionnaire headings, including the impact of Covid NPIs on it, was a key concern in our rural study ‘livelihoods’, ‘information’, ‘coping strategies’ or ‘vulnerabilities’, communities. The reduced availability of travel complicated access and sub-categories. The third ‘selective’ step meant selecting to goods, including clothing, food or consumables in all districts answers to illustrate key empirical results (Mikkelsen, 2005). Given because of local vendors partly being unable to travel to wholesale the novelty of Covid, there was a conscious effort to be driven by markets (HC-05-03, HC-08-05, SMC-01-01), or traders no longer participants’ answers in identifying, naming and clustering codes visiting rural communities (BR-05-04, HC-03-04, HC-06-05, MB- and analysis. 05-03, MV-02-06, TL-10-01). A range of panelists reported higher prices as a result: 3.3. Limitations ‘Now traders don’t come here anymore. And local vendors are raising prices because there is no competition.’ (HC-05-05) Although great care was taken to include a diversity of rural ‘Those who have little stalls here in the community used to tra- experiences across social groups, ages and genders, a key limitation vel to the city [to buy wholesale] and then resell here, but now of our study is the make-up, size and availability7 of our panel many do not have products to sell any more.’ (MF-05-04) which shapes what we can report. Equally, while our research team worked together very closely to undertake data-driven analysis Consequently, some individuals, despite the inherent risks, felt across all sites, the pandemic rendered impossible personal engage- they had to travel to larger settlements to restock or obtain key ment between the entire research team across preparation, data col- consumables (BR-05-04, HC-08-05, HC-09-05, MF-06-05). lection and analysis. This inevitably limits the insights we have gained. Thirdly, as explained above, our empirics consisted of quali- ‘I would like to just stay at home, but . . . I need to put myself at tative interviews rather than any quantitative data collection. All risk and go to the city.’ (MV-06-05) findings are based on a mixture of qualitative data, particularly interview quotes8, and analysis in terms of how frequently different Finally, the restrictions led to diminished cash and non-cash issues were raised by panelists, while contextualizing these frequen- income in many sectors:9 cies with qualitative quotes or in relation to the composition of the ‘The interruption of trains and restriction of car circulation has wider panel. We do not claim these to be representative of the led to a lack of money among community members here. We broader population, but present them to indicate prevalence within are all disadvantaged, no-one has benefited . . . Here in the com- our panel. munity, we have nowhere to buy soap, flour, all food products. I cannot go to [another settlement] because there too, they do 4. Results: Covid-19 effects on livelihoods and vulnerabilities in not want any strangers. We only survive on forest food products rural Mozambique now.’ (HC-01-05) After an initial discussion of key changes due to Covid, results No Covid infections in our communities were reported in the are structured in accordance with our research questions to reflect interview period and it appears there were no significant health variations between ages, social groups and genders, divergences impacts from Covid infections, though diets were compromised between occupations and value chains, and variations between and accessing treatment for chronic illnesses became more difficult communities. due to NPIs. 5 4.2. Variation between social groups, ages, and genders The original questionnaire is available from: https://blogs.ed.ac.uk/miombo- matters/wp-content/uploads/sites/961/2020/07/CwC_Datacollectionsheet_EN_PT. xlsx. Firstly, on variations between social groups, Fig. 2 highlights 6 This new set of questions was asked of 75 out of 92 panelists for at least one week with what frequency transport and movement were mentioned of study (36 female, 39 male panelists; 14 vulnerable, 16 microbusiness, 11 market- across our panel, with microbusiness owners and market- oriented smallholders, 13 traditional influence, 19 modern influence, with two panelists in Mabote unallocated to any categories). The questionnaire can be found 9 here: https://blogs.ed.ac.uk/miombo-matters/wp-content/uploads/sites/961/2020/ We acknowledge that there may be businesses who benefited from the crisis and 07/200710_CwC_Datacollectionsheet_NewQs_EN_PT_V6.xlsx. may not have wished to acknowledge that in interviews. However, based on both 7 Please cf. Table 2 for details on how many interviews could be conducted in interviews and on-the-ground observations from researchers and civil-society different communities, by gender and social groups. partners, our panelists’ unanimous description of either neutral or adverse livelihood 8 Interview quotes (xx-xx-xx) are indicated with interview codes, which reflect the impacts in response to open-ended interview questions was accurate for our sample. study community, the interviewee’s number, and the study week. Where more than We detail in ’Results’ the degree to which other value-chain stakeholders benefited in one interview code is stated, they are ordered alphabetically. light of power asymmetries. 5 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 ‘People who are most affected are people who are not aware, the elderly, the people who are not up-to-date on what is going on in the world.’ (HC-03-04) ‘[Most vulnerable to Covid are] elderly people who live alone . . . some have relatives, but others only depend on neighbors.’ (MF- 05-04) This perceived vulnerability for the elderly due to lack of knowl- edge was confirmed in our interviews by poor or erroneous knowl- edge on how to identify or prevent Covid. Although only 26 out of 90 panelists11 were over 55 years of age, half the interview state- ments demonstrating no or erroneous knowledge occurred in the Fig. 2. Number of references to transport or travel in interviews (light gray) across 55+ age range (43 out of 83 instances in interviews). While poor the five different social groups (left to right: microbusiness, market-oriented smallholders, modern influence, vulnerable, traditional influence); juxtaposed with or erroneous knowledge on Covid was therefore not highly preva- total number of interviews (gray) and panelists per group (dark gray). Colors lent, knowledge gaps nevertheless appeared to expose the elderly arbitrary. Source: Authors. differentially to risks in a novel pandemic. On gender, there was a clear difference in using environmental resources as a coping strategy. While NPIs changed almost all oriented smallholders dependent on selling e.g. charcoal or agricul- aspects of our panelists’ lives and livelihoods, interviewees across tural produce most vocal on the effects of NPIs on transport: all genders and social groups stated that their access to natural In terms of lived experiences differentiated by age, there were products was unchanged, ranging from timber for construction, significant concerns across the panel about the education that chil- fuelwood, forest food products, to baobab for collection and sale. dren were missing on account of NPIs resulting in school closures, Crucially, panelists in more vulnerable situations, be it due to and the degree to which this would affect their vulnerability long- pre-existing precarity, difficult water access or NPI-eliminated term. Consequently, the majority of parents10 wanted their children livelihoods, reported relying more heavily than before on natural to return to school when they reopened, for example ‘so that they do products. This included using forest food products as a means of not become illiterate like me’ (MV-01-05), or for girls to complete survival due to food insecurity, as well as recourse to collecting 12th grade before they got married (SMC-10-07, GN-07-07). How- and selling baobab, a livelihood only available to women in our ever, numerous parents were concerned about letting children go study communities. This increased environmental dependence back to school. was particularly reported by women across all social groups (cf. ‘When they reopen schools, my children will only go back if pre- Fig. 3): vention is in place. Distancing, water, disinfectant, mask obliga- A related, final key difference between genders was the degree tion need to be implemented.’ (HC-03-04) to which Covid made panelists vulnerable to food insecurity and hunger. One reason is that both ends of the age spectrum were seen by ‘. . . the government should give us food; if they give me food, I many panelists as being particularly at risk of contracting Covid. will never leave the house. The prevention measures are good, Children were considered in danger as lacking in prevention, but since we are hungry, we need to leave the house to get awareness of prevention strategies, and willingness to listen money to buy food.’ (MF-02-04) (MV-01-06, SMC-01-07): ‘To prevent this disease we have to stay at home, but if we stay at home, we die of hunger.’ (MF-04-05) ‘Some of the children and young people . . . don’t understand that they have to be careful and use masks.’ (MF-07-05) Across 441 interviews, Covid-related hunger was mentioned 115 times. However, across all social groups, women referenced ‘It is very hard to stop [children] from playing outside.’ (SMU- food insecurity and hunger due to NPIs more frequently than 05-06) men (83 compared with 32 mentions; total panel: 45 female and 47 male panelists), with vulnerable and smallholder panelists In some cases, against the backdrop of children being home equally citing food insecurity more often than traditional or mod- from school and economic hardship experienced, parents were ern influential individuals. Covid-related hunger and food insecu- considering early marriages for their daughters in two of our study rity also varied by community, as we explore further below. communities, in Sussundenga SMC and Guro-Tambara GN. In one case, a children’s rights non-governmental organization was called 4.3. Variation between occupations and value chains in to reverse an early marriage of an underage female adolescent. A second gender aspect of children being home from school was that, Through our qualitative interviews, considerable differences with children and adolescents now at home and in the charge of were reported in terms of how NPIs affected different occupations their mothers (GN-10-06, HC-03-05; HC-05-05, MB-05-03, MF- and value chains. Firstly, transport restrictions eliminated some 09-05, SMC-04-06), women’s workload significantly increased livelihoods entirely. These included purchasing wholesale to sell due to added unpaid reproductive labor. locally (MF-07-05, GN-08-01): ‘Many people from the community At the other end of the age spectrum, the perception of the used to go to [a bigger settlement] to buy products wholesale, but elderly being particularly at risk stemmed from the stay-at-home now that the train is not running, they have lost their businesses, orders exacerbating their level of isolation, and limited information especially women’ (MF-05-05). This equally applied to those rely- flow about Covid: ing on tourism revenue close to a protected area in Sussundenga (SMC-06-01; SMC-07-01), and to casual labor associated with e.g. 10 transporting bananas (SMC-02-04; SMC-06-06). BR-05-04; BR-07-03; GN-10-06; HC-08-05; MB-06-03; MF-05-04; MF-07-05; MF-09-05; MV-01-06; MV-02-06; MV-04-05; MV-06-05; MV-07-06; SMC-01-07; 11 SMC-04-06; SMC-07-06; SMC-10-06; SMU-05-06; TL-05-06; TL-07-06. For two panelists, ages were not known; they were excluded from this analysis. 6 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 The banning of alcohol also threatened the livelihoods of those previously brewing traditional beverages, meaning it dispropor- tionately affected women. ‘[Most affected by Covid restrictions, are] the women who pro- duce traditional drinks in their houses. They don’t have another source of livelihood.’ (GN-05-06) More generally, NPIs required shop or stall owners to provide water and soap or ashes at shop entrances (HC-03-01), and ensure that customers observed social distancing (SMC-03-01). Both cus- tomers and stall owners identified negative consequences of NPIs in terms of opening times, access to goods and additional stocks (HC-01-01; HC-05-04; HC-08-02), and elevated prices for key sta- ples including flour or sugar (GN-01-01; TL-04-01): ‘Things have changed - everything is more expensive. Food Fig. 3. Instances of reported higher environmental dependence in times of Covid by gender and social group (female left; male right; groups from left to right: stalls close at 17:00. To buy from there we have to wear masks, vulnerable, traditional influence, modern influence, market-oriented smallholder, wash hands before entering and enter one at a time.’ (BR-08-01) microbusiness). Colors arbitrary. Source: Authors. ‘[My access to food and goods] has been affected a lot; some- times when I go to the stalls, the stalls can be closed . . . And some people are raising prices a lot.’ (SMU-03-01) For those producing for or selling into domestic and interna- In addition, transport restrictions limited business for various tional value chains, degrees of market integration and prevalence other occupations. NPIs restricting transport operators to 1/3 of price-focused or civic-based priorities played a considerable role capacity entailed considerably higher fares (BR-05-04, HC-09-05) in shaping livelihood impacts as a result of NPIs. In some value given identical operating costs, yet fewer passengers. Conse- chains such as honey and baobab, socially-oriented investors prior- quently, the frequency of service use was diminished for transport itizing ethical conduct and socio-environmental benefits safe- business operator TL-07, despite an elevated risk of contracting guarded value-chain access and stable prices, despite transport Covid by virtue of being in the car with strangers (TL-07-01, TL- restrictions and distancing orders. The baobab buyer, ECO- 07-06). The reduced use of transport was also due to a fear of travel MICAIA, which is run by authors Kingman and Nuvunga, made (e.g. GN-06-07, HC-08-04, HC-09-05, MB-03-03, MF-01-05, MV- efforts to change its procurement and collection practices to adapt 01-05, SMC-04-06): to NPIs, observing social distancing and putting up Covid preven- ‘The main risk of contamination with this disease is during tra- tion and information posters in baobab purchasing centers (GN- vel to the city, because people sit so close together, and there 01-01; GN-07-01, TL-05-02). While baobab has in the past also are also a lot of people in the city.’ (HC-04-04) served as emergency food in crises (GN-04-06), its collection and sale, benefiting exclusively women, was seen as remaining con- More generally, the mobility required for livelihoods including stant despite Covid, and a key lifeline in the Guro/Tambara com- transport, trading or community leadership was seen as putting munities (GN and TL; GN-06-06; GN-09-06; TL-03-01): individuals at elevated risk of catching Covid (e.g. BR-01-03, GN- 02-06, HC-07-05, MF-07-05). A reliance on travel for business also ‘There is social distancing where we sell baobab. Selling baobab was seen as exposing households and communities to the effects of has not been affected by Covid.’ (TL-05-02) NPIs and changing buyer and seller behaviors (e.g. GN-02-06, GN- ‘I am just glad I can make money by selling baobab.’ (TL-04-05) 07-06, SMC-01-06, SMC-06-06, SMU-01-06, SMU-04-01, TL-06- In other value chains, NPIs drove down producer prices and 06): diminished opportunities to sell, including peanut, vegetable, ‘Traders cannot travel any more for their business. I have to stay banana and charcoal production. With a concomitant increase in at home and just sell small things.’ (SMC-10-06) retail prices for consumables including food (cf. sections 4.1, 4.2 ‘I used to sell bread outside of my community, but now I . . . only and below), NPIs effectively reduced purchasing power for parts sell from home.’ (TL-06-01) of our panel. The case of Tambara’s peanut farmers illustrates the dynamics Beyond transport restrictions, a key restriction imposed on of insecure market integration. Early in the study, there was a per- business owners concerned banning the sale of alcohol, including ception that peanut demand was low, prompting many farmers to traditional home-made drinks (HC-05-01). This restriction reflects sell early at lower prices (TL-03-02). When out-of-town buyers that a subset of panelists explicitly identified people who had con- came back later offering higher prices, many had already sold off sumed alcohol as more vulnerable to contracting Covid, given a their produce (TL-04-07). lack of inhibition (n = 5). This measure also had consequences for In the peanut and banana value chains, agricultural producers the livelihoods of traders: accustomed to selling their produce to non-local buyers, including ‘People are not circulating much anymore; they do not buy nipa in or from Zimbabwe or Malawi (TL-07-06, GN-03-06), saw prices [traditionally brewed drink] any more; there is no more busi- significantly reduced due to closed borders and a lack of transport ness going on.’ (SMC-01-01) (TL-01-01, TL-04-01, TL-09-01, SMC-02-01, SMC-05-01, GN-10- ‘I don’t sell like I used to any more. We used to be able to sell 01). Moreover, this led to a lack of casual labor opportunities alcohol, because that is where the money is. Now people don’t (SMC-06-06). have money. [It used to be that] when I buy product today, I ‘I used to sell bananas at good prices, but not anymore because can get the money to replenish the stock in three days, but that we are not allowed to leave the community to sell.’ (SMC-02- does not happen anymore.’ (BR-10-01) 01) 7 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 ‘I used to sell peanuts to buyers from [far-away locations], and water, because now, when she gets to the collection point, she they bought at good prices. Now I can only sell locally, and at has to wait her turn. Now to fetch water, it takes her more time, bad prices.’ (TL-04-01) and she has to wash clothes at home. . . . We used to use water just for drinking, cooking and washing, but now we have to A value chain affected in diverse ways by transport restrictions have a water bucket at the entrance to the house and we always and reduced purchasing power was charcoal. Charcoal is an impor- have to wash our hands. Now she collects five or ten water con- tant revenue in the Mabalane and Mapai study communities (BR, tainers, and it takes her a long time. She has to go fetch water HC, MF, MV) due to high-quality resources and demand from the twice now, in the morning and at the end of the day. Before this urban areas of Maputo and Xai-Xai; it is a longer-term adaptation disease, she only had to go once a day, and five containers were strategy which preceded Covid (Baumert et al., 2016; Eriksen & enough.’ Silva, 2009; Smith et al., 2019). This value chain was affected by NPIs firstly because of its heavy reliance on collaboration. This testimony highlights firstly, the time commitment related to water collection, and secondly, social distancing, which we will ‘Charcoal production requires working together to be effective, explore in turn. Fundamentally, our study communities state that but now we cannot help each other, and working alone is fetching water is the domain of women and children, mostly girls harder, especially for us women.’ (BR-07-03) (SMC-01-06, SMU-02-06, TL-09-06, GN-07-06, BR-07-03). In one Secondly, Covid’s travel and transport restrictions played a key community, men expressed that by tradition, men do not fetch role, as accessing urban charcoal markets through trains was no water (SMC-02-06, SMC-03-06) – ‘it would be a disgrace if a man longer possible. Consequently, sellers were reliant on itinerant went to fetch water’ (SMC-03-06). In other households, both male trucks buying from them. and female children help women with water collection (TL-09-06), though elsewhere, boys refuse to fetch water (GN-02-06, GN-07- ‘The livelihood which we have here to survive is producing . . . 06). charcoal. Now that the trains are no longer running, those that While there may be an assumption that, because water collec- depend on those livelihoods are bankrupt.’ (HC-10-04) tion is women’s labor and the Covid response relies heavily on ‘We now depend on vehicles which buy charcoal so we get any water, women’s workload and related time commitment increased money at all.’ (MF-07-05) significantly, there is a more differentiated picture in our panelists’ Moreover, charcoal sellers very explicitly linked transport responses. Among those believing there was no change (n = 16)13, restrictions to lower prices that passing vehicles were willing to one woman said: ‘Washing hands does not take a lot of water’ (MV- pay for charcoal: 01-06). Others said that the time commitment had increased (n = 15)14 – either because of higher volume needed, because of extra ‘Now that we have the disease, we are not selling well . . . With- time spent at the pump to guarantee social distancing, because of out the disease, we sell [a charcoal sack] at 500 Mts12, but now limits imposed on how much water could be collected at any one we sell it at 450 Mts by the roadside . . . Wholesale they used to time, or because of more time needed to do washing away from buy at 350 Mts, but now it’s 300 Mts . . . Revenues are lower . . . the collective water collection point. There is thus a mix of percep- now we can only buy flour . . . Now when we sell 10 sacks [of tions across diverse social groups, genders and locations. However, charcoal] - flour is 1500 Mts, oil is 280 Mts, so we have nothing it appears that in communities with more difficult water access left over. It feels like we are getting ripped off because of coron- (BR, GN, HC, MB, MF, MV), there is a stronger sense among panelists avirus.’ (BR-05-02) (n = 19 compared with n = 12 in other communities) of more time ‘Before this disease came, I used to sell alcoholic drinks; I being spent on water-related activities, pointing to wider, complex bought things in [bigger settlement] and sold them here. Since water scarcity dynamics. this disease came, trains are not running any more, and we can A definite change regarding water collection was noted due to only sell charcoal. . . . A sack of charcoal can stay by the side of social distancing requirements. Social distancing requirements the road for a month without being sold. Now . . . when cars do make it more difficult for women to carry the water home as it come, . . . they do not want to pay the price we tell them.’ (MF- makes mutual assistance with other women impossible (MV-04- 03-03) 04). In some cases, social distancing equally increased the amount This panelist, MF-03, thus adopted charcoal production, with of time spent collecting water due to waiting times (GN-03-06), which her neighbors had engaged as a longer-term adaptation with some women setting off around 3 a.m. for a five-hour round strategy, as a short-term coping strategy in response to Covid trip to fetch water (GN-04-06). Though this level of distancing is restrictions limiting other income sources. However, the lack of not observed everywhere (TL-02-06), it appears that where they integration into markets and the absence of socially-oriented part- are in force, social distancing requirements are respected (HC-03- ners in the charcoal value chain left charcoal sellers at the mercy of 04, BR-07-03): itinerant, non-integrated buyers, further diminishing in- ‘Before, women used to go in groups to get water, but now community value generation through charcoal. everyone goes in their hour, and . . . they practice social distanc- ing.’ (HC-05-04). 4.4. Variation between communities This social-distancing requirement chimes with some male A final, related dimension is variation between study communi- panelists viewing women as particularly at risk of catching Covid ties, which was particularly pronounced firstly around the use of because of water collection (e.g. GN-10-06, GN-01-07). For addi- water in the pandemic, and secondly the prevalence of Covid- tional protection, some women also have agreed to fetch water related food insecurity. Firstly, a range of NPIs affected water col- in masks (GN-03-04). Another common prevention strategy has lection and use, as panelist BR-07-03 explains. 13 ‘In my house it is my daughter-in-law who collects water; she is No change: MB-05-03, MB-06-03, BR-05-04, GN-01-06, GN-05-06, GN-09-06, MV-01-06, MV-07-06, SMC-02-06, SMC-04-06, SMC-05-06, SMC-06-06, SMC-08-06, the only one who does so. Now it takes her longer to fetch SMC-09-06, SMU-07-06, TL-08-06. 14 More time: BR-06-04, BR-07-04, GN-02-06, GN-03-06, GN-04-06, HC-04-06, HC- 08-06, HC-09-06, MB-03-03, MB-08-03, MF-02-04, SMC-10-06, SMU-05-06, TL-09-06, 12 At the time of data collection 1 USD was ca. 80Mts. TL-10-06. 8 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 been to separate locations of clothes washing from locations of 2021; Sapkota et al., 2016). As Covid-related hunger numbers, in fetching water (BR-01-03) or requiring women to do washing at our panel, were much higher among those already disadvantaged, home instead (MB-07-03), which was seen as further adding to our study thus confirmed estimates that Covid NPIs would entail women’s workload. significant impacts for vulnerable populations (FAO et al., 2020), Finally, the previously discussed observations around what with food insecurity reflecting and reinforcing socio-economic value chains and occupations were particularly affected by trans- inequities (Klassen & Murphy, 2020; Kansiime et al., 2021). Conse- port restrictions and reduced income opportunities equally trans- quently, these pre-existing vulnerabilities and inequalities merit late to a marked difference in terms of what communities most particular attention in all Covid-related analysis and policy recov- frequently report hunger and food insecurity as a result of Covid ery efforts. NPIs. Especially the border closures made coping strategies previ- By exacerbating existing vulnerabilities and creating new expo- ously employed in times of drought, such as buying food or seeking sures, the hyper-covariate nature of Covid and NPIs entailed con- employment abroad, unavailable (SMU-07-06, SMU-09-06, TL-06- siderable, yet varied consequences across social groups, 06). Given the above-explained dynamics in the charcoal value occupations, value chains, ages and genders, the nuance of which chain, which saw returns for small-scale producers grow lower is documented in our qualitative empirical data. At the community, and less predictable under Covid NPIs, communities in Mabalane household and individual levels, vulnerabilities and livelihoods and Mapai study districts mention food insecurity far more fre- were reshaped by Covid-reconfigured human and environmental quently (cf. Fig. 4): conditions, including the impact that travel, transport and trading restrictions exerted. Panelists commented on past coping strate- 5. Discussion: relevance to Global South research and policy gies to crises such as cholera or drought not being viable under Covid NPIs due to border closures and travel restrictions prohibit- 5.1. Vulnerabilities, old and new: intersecting precarities ing e.g. seeking employment or food elsewhere and abroad. In our charcoal communities, Covid NPIs combined with a lack of rain A key emerging theme relevant for research and policy concerns limiting cultivation opportunities to produce food insecurity and the ways that Covid NPIs exacerbated existing vulnerabilities. With reliance on forest food products especially among the vulnerable. respect to gender, age, poverty and other disadvantages, our study In addition, Covid created new economic vulnerabilities: firstly, thus confirms similar findings from other sites in the rural Global even larger-scale agricultural producers were affected as a result South (e.g. Kansiime et al., 2021; Mahmud & Riley, 2021; Puerta of transport restrictions. Livelihoods were eliminated or dimin- Silva et al., 2020a, b). Covid NPI-specific influences interacted with ished especially in the realms of vending, wholesale trading or existing exposures to climate change and variability, including the the traditionally female livelihood of making traditional beverages, prior shocks of cyclone Idai, the 2015/16 drought and limited rain- with the overall reduction in cash-income opportunities affecting fall restricting cultivation, to enhance sensitivity to NPIs. This con- especially those ordinarily engaging in casual labor. Unlike in some firms other studies which found that, more than exposure to the other Covid-affected contexts (e.g. Dutta & Fischer, 2021 for India; shock, pre-existing socio-economic and environmental conditions Gupta et al., 2021 globally; Kansiime et al., 2021, for Kenya and shaped vulnerabilities (Batterbury & Forsyth, 1999; Gupta et al., Uganda), there was no state support: the institutional level thus Fig. 4. Number of panelists mentioning hunger and food insecurity because of Covid NPIs, disaggregated by study community (left to right: MV, MF, BR, HC, MB, SMC, TL, GN, SMU; 10 panelists per community except in SMC and SMU, which had 11 each). Color arbitrary. Source: Authors. 9 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 only worked to restrict options, rather than providing livelihood gender via age to status and income, were accentuated and new support in a time of crisis. As environmental influences on vulner- exposures created through the hyper-covariate shock of Covid. By abilities and livelihoods in rural Mozambique and beyond are accentuating the contextual, institutional-political dimension of likely to expand due to accelerating environmental changes, the vulnerability (O’Brien et al., 2007), the findings raise the question unprecedented, hyper-covariate shock of Covid demonstrated a at what point and in what contexts the trade-off between illness concerning lack of safety nets especially for the more precarious due to the virus, and hunger and poor nutrition due to disruptions panelists and communities in our sample. to markets and institutions (Ravallion, 2020), needs to be The diverse vulnerabilities resulting from the exposure to NPIs rethought. Conceptually, the ways that Covid NPIs intersected with thus recall Bonilla and LeBrón’s (2019) idea of aftershocks. They preexisting vulnerabilities including gender, age, income raise highlight the importance of analyzing not only a shock, but the rip- broader questions about the degrees to which these axes need to ple effects and their myriad repercussions across social, economic be recognized more explicitly in crisis response policies and vul- and environmental domains, creating new urgencies and compli- nerability frameworks, as further explored below. cating recovery efforts. Our findings confirm that Covid is a health crisis with diverse impacts on food, social and economic systems 5.2. Coping and adaptation: rethinking diversification? (Swinnen and McDermott, 2020). Far beyond any impacts from the disease itself, which remained an unknown in our interview A further key theme concerned coping and adaptation strate- period and communities in terms of transmission or incidence, gies, and which approaches remained viable under Covid NPIs. the ripple effects of institutional restrictions, limits and bans The above-discussed reliance on natural products and the environ- accentuated vulnerabilities and created new exposures. ment highlights firstly the importance of environmental resources In gender terms, the above findings showed that women in to more vulnerable populations in crisis (Pritchard et al., 2020), Mozambique were affected by Covid NPIs in terms of food security, especially since access to natural products – unlike almost all other productive and reproductive labor. Women in our sample men- aspects of life – remained unchanged under Covid NPIs according tioned food insecurity as a concern more frequently than men, to panelists. The importance of natural products also highlights especially in the vulnerable and smallholder social groups. A that some, but not all, diversification strategies proved viable specifically female livelihood, producing and selling traditional under Covid. and alcoholic drinks, was affected by the government trading Across our panelists, diversified livelihoods came in different ban. Female reproductive labor equally expanded, confirming sim- manifestations, as respondents relied on some combination of agri- ilar findings from South Asia (Nichols et al., 2020; Agarwal, 2021). culture, bee-keeping, a stall or vending business, charcoal, tourist Specifically female roles including child care and water collection revenue, livestock or baobab collection. The degree to which these also were impacted considerably by Covid, as mothers’ workload strategies worked as Covid NPI coping strategies varied consider- was increased due to school closures, and water collection in ably. Tourist revenues were eliminated, stalls and vending busi- water-poor contexts grew more complicated and partly more nesses severely limited. Baobab collection, given a socially time-intensive due to NPIs. These findings highlight the more vul- inclined trading partner, continued providing income streams even nerable position of many women towards Covid and its implica- as many other value chains were affected by absent buyers. Food tions (Agarwal, 2021; Nichols et al., 2020; UN Women, 2020). crop cultivation was a lifeline for some, yet was limited in some Our study reiterates the importance of considering women of all areas by low rainfall in 2020. In Mabalane and Mapai, charcoal pro- ages, their higher levels of informal employment, their higher duction and sale is a major contributor to diversified livelihoods involvement in unpaid care and domestic work, and their higher (Baumert et al., 2016). Here, charcoal also proved to be a Covid levels of precarity, in all research and policy efforts to address coping strategy given other livelihoods such as brewing drinks or the socio-economic implications of Covid and Covid NPIs. having a stall being eliminated through NPIs. This has occurred In terms of age, both children and the elderly were affected par- even though charcoal production, given its reliance on collabora- ticularly by Covid NPIs in ways that increased their vulnerability. tion and its links to respiratory illness (Jagger and Shively, 2014) Children saw schools being closed and were thus significantly lim- as well as the cough it can produce, could be seen as unsuitable ited in their education while seeing their chores at home increase, in Covid times. Equally, the growth in particularly female involve- particularly for girls in terms of fetching water in water-poor con- ment in charcoal production in our study mirrors Jones, Fisher and texts. In some communities, girls were also at increased risk of Ryan’s (2016) finding of significant female participation in charcoal early marriage. Both children and the elderly, according to our pan- production in central Mozambique. Charcoal as a Covid coping elists, were particularly in danger of contracting Covid on account strategy was severely impacted by transport NPIs, leading to a of a lack of knowledge and awareness, with the elderly, particularly squeeze in purchasing power and comparatively higher levels of those living alone, frequently cut off from reliable information food insecurity. Thus a popular diversification strategy pre-Covid channels on a novel disease, its symptoms and transmission. became the sole cash income source, despite a lack of market inte- The dynamic of Covid NPI-related hunger disproportionately gration and strong networks failing to safeguard value generation affecting vulnerable groups, including women, is relevant for both in the communities, which had already been low (Baumert et al., policy and research in two main ways. Firstly, NPIs led to reduced 2016). incomes and higher food insecurity among more vulnerable popu- Overall, our data and analysis thus emphasized a complex pic- lations in Kenya and Uganda (Kansiime et al., 2021), mirroring our ture in terms of which diversification approaches, and more gener- findings. Secondly, more vulnerable populations in terms of depri- ally, which coping or adaptation strategies, remained viable under vation in Brazil have been found to be at greater risk of contracting Covid restrictions, and what was required for each strategy to con- and dying from Covid due to poor housing and livelihoods that tinue being effective. Other studies on Covid impacts and diversifi- make it impossible to implement all preventive measures cation have produced diverse perspectives on economic (Tavares & Betti, 2021). Finally, it reinforces the need to analyze diversification potentially reducing vulnerability in rural contexts carefully the suitability of policy responses including stay-at- (Bassett et al., 2021), vs. Covid exceeding the scale of shocks for home policies for vulnerable people (Alon et al., 2020; Barnett- which households in Sub-Saharan Africa have previously used Howell & Mobarak, 2020) in the absence of state support or any diversification as a coping mechanism (Hilson et al., 2021). The other mechanisms that would soften the blow. Our findings show empirical data collected in the context of the hyper-covariate the nuance of how various dimensions of deprivation, ranging from shock of Covid thus produce a nuanced picture in terms of diversi- 10 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 fication reducing vulnerability (Ellis, 2000): as diversification can by communities across value chains would be vital to explore depend on a dynamic wider economy (Ellis, 2006), the broad eco- going forward. nomic ramifications of the hyper-covariate Covid shock both sig- In all value chains prioritizing market-based and price-focused nificantly affected economic productivity and dynamism overall, mindsets, Covid NPIs accentuated significant power differentials to and created new vulnerabilities including for wealthier panelists. the detriment of our panelists. Though agriculture-oriented value Consequently, those that would generally be seen as better-off – chains (vegetables, bananas, peanuts) fared somewhat better than e.g. a transport operator or large-scale banana cultivator – were charcoal, producers were just as dependent on buyer behavior as in unable to use diversification as a shield from the impacts of Covid, charcoal communities. Across all these value chains without challenging existing findings on diversification often benefiting the socially-oriented stakeholders with civic-based priorities, there better-off (Alobo Loison, 2015; Ellis, 2006). We thus encourage fur- were reports of transport restrictions leading to lower quantities ther empirical and conceptual work e.g. at system level to identify being purchased, lower prices being paid, while producers them- what coping, adaptation or diversification strategies, in Covid and selves were at the mercy of higher prices for food and other con- beyond, have remained viable in the face of what types of shock. sumables. As such, these findings confirm Reardon, Bellemare and Zilberman’s (2020) analysis that Covid impacts are likely to 5.3. Covid-proofing value chains? affect especially the commercialization of farm products, while driving up food prices. On power and power asymmetries in value Crucially, the empirical data about the highly diverging experi- chains and production networks (Coe, Dicken, & Hess, 2008; ences of value chains under Covid emphasizes questions of equity Henderson, Dicken, Hess, Coe, & Yeung, 2002; Krauss & (Leach et al., 2021) and ‘who gets what’ between value-chain Barrientos, 2021), this reiterates the question of how the corporate actors (Barrientos, 2019; Krauss & Krishnan, 2021; Oldekop et al., power of organized buyers, and the institutional power of state 2020). For charcoal, Baumert et al. (2016) previously showed that interventions, rendered the collective power of producers largely little value remains with local communities, who have limited con- irrelevant. The design of sustainable, inclusive markets as a precur- trol over commercialization. Our study confirms these to be crucial sor for well-being improvements to safeguard no-one being left issues given charcoal producers reporting lower prices being paid, behind (Smith et al., 2019) thus remains as crucial as it is elusive, and commercialization links to the capital ceasing to work reliably especially under and after Covid (Córdoba et al., 2021, Puerta Silva under Covid NPIs, creating unpredictability both for volume and et al., 2020a, b). A highly relevant question for national and devel- frequency of sales. Across charcoal-producing communities, ques- opment policies is to what extent formalization of producer asso- tions were raised about why trucks stopped coming, and why char- ciations to augment bargaining power and enhance value-chain coal prices reduced at a time when prices for consumables accountability through local involvement, as well as the promotion including flour increased. Given significant multidimensional vul- of solidarity-based economic models (Córdoba et al., 2021), would nerability, which charcoal production previously has been found improve producers’ standing, for charcoal and agriculture alike. not to alleviate sufficiently (Vollmer et al., 2017), the parallel crises of Covid and poor harvests caused considerable levels of food inse- 5.4. Reframing vulnerability and coping? curity and hunger among our panelists, confirming similar findings from Kenya and Uganda (Kansiime et al., 2021) and Colombia The empirical data presented above both confirms and chal- (Puerta Silva et al., 2020a, b). lenges existing literatures. We can confirm, in keeping with The dynamics seen in our empirical data reiterate the impor- (Turner et al. (2003)) observations, that individuals in the same tance of market integration (Eriksen & Silva, 2009) and, crucially, community, as well as communities in the same country, could integration into civic-based markets and value chains (Renard, not be assumed to have broadly similar vulnerabilities, as varia- 2003; Krauss & Barrientos, 2021). There was a clear discrepancy: tions occurred between occupations, genders, social groups, ages, baobab and honey value chains, which involve socially-oriented value chains and communities. This was even more the case amid stakeholders prioritizing civic-based mindsets including ethical the unprecedented scale of shock that was Covid (Hilson et al., conduct and reliable socio-economic benefits for producers, made 2021). The considerable reliance on environmental resources necessary adjustments to maintain fair operations and producer among our panel equally highlights their importance as lifelines livelihoods even under NPIs. This mirrors findings from the Andes, and safety nets to cope with crisis, supporting Pritchard et al.’s where organizations prioritizing fairness in trading and solidarity- (2020) findings. Finally, the complex interplay between national- based understandings of society were reinvigorated under Covid as level policy g and variegated local repercussions emphasized the a counter-movement to purely market-oriented practices (Córdoba importance of assessing diverse factors affecting contextual vul- et al., 2021). Conversely, value chains following exclusively the nerability, exposure as well as capacities for resilience (O’Brien, market-based regime, which prioritizes price, either produced Eriksen, Nygaard, & Schjolden, 2007; Turner et al., 2003). A surpris- adverse outcomes including lower prices or largely broke down, ing finding was the considerable role of integration into value given producers’ lack of integration into upstream buyers or retail- chains and markets in shaping livelihood vulnerability and resili- ers. Even communities selling vegetables or peanuts, whose links ence in our sample. While baobab and honey, operating under a to out-of-town buyers were stronger than for charcoal, partly civic-based regime prioritizing socio-environmental production reported selling lower quantities and at lower prices, yet the conditions and ethical conduct (Renard, 2003; Krauss & impact of transport NPIs was considerably less than in charcoal- Barrientos, 2021) provided a crucial reliable cash income source producing communities. For charcoal, the lack of integration into under Covid, especially charcoal, a solely price-oriented, market- urban charcoal retailers meant that both prices and selling oppor- based value chain, proved highly unstable for small-scale produc- tunities proved extremely unpredictable for producers, as they ers due to transport-restricting NPIs. Based on our novel data and were unable to safeguard fair prices and engagement from more findings, there is a need to explore further what factors in value- powerful value chain actors, who sold at elevated prices in urban chain governance and composition can boost livelihood resilience centers. Given the elimination of livelihood options such as tradi- in crisis. tional brewing under Covid, some panelists began engaging with However, the empirical data also challenged existing litera- charcoal to survive, yet many panelists reported significantly tures. New vulnerabilities arose for certain diversification or adap- reduced prices, frequencies of sales, and partly a need to rely on tation strategies which had previously been considered wise forest food products to survive. Opportunities for cross-learning precautions, including trading or vending and charcoal production, 11 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 and affected the better-off, who ordinarily would be expected to CRediT authorship contribution statement benefit from diversification (Ellis, 2006). In terms of challenging conceptual lenses, our empirical data showed how critically Judith E. Krauss: Conceptualization, Methodology, Data cura- human-institutional interventions, i.e. NPIs such as restricting tion, Formal analysis, Conceptualization, Writing – original draft, transport, shaped local vulnerabilities as well as coping strategies, Writing – review & editing, Project administration, Funding acqui- emphasizing the importance of analyses cognizant of such institu- sition. Luis Artur: Conceptualization, Methodology, Investigation, tional circumstances. Finally, Covid NPIs exacerbating vulnerability Writing – review & editing, Project administration, Supervision, resulting from income, status, gender, age in diverse, but pro- Funding acquisition. Dan Brockington: Conceptualization, nounced, ways, equally raises questions about how systematically Methodology, Writing – review & editing, Project administration, these lenses are, or should be, taken into account by analyses Funding acquisition. Natasha Ribeiro: Conceptualization, Method- investigating particularly how crisis-proof coping and adaptation ology, Investigation, Conceptualization, Writing – review & editing, strategies including diversified livelihoods are. Project administration, Supervision, Funding acquisition. Julio Tembe: Methodology, Investigation, Data curation, Formal analy- sis. Eduardo Castro: Methodology, Investigation, Data curation, 6. Conclusion Formal analysis, Writing – review & editing. Jone Fernando: Con- ceptualization, Methodology, Investigation, Data curation, Formal In this article, we addressed the questions of how Covid NPIs analysis, Project administration, Funding acquisition. Janet Fisher: affected livelihoods in rural Mozambique, and how these impacts Methodology, Conceptualization, Writing – review & editing, Pro- intersect with existing vulnerabilities and crises across different ject administration. Casey M. Ryan: Conceptualization, Methodol- occupations, social groups and genders. Based on in-depth qualita- ogy, Conceptualization, Writing – review & editing, Project tive interviews with 92 panelists from 9 Mozambican communities administration, Funding acquisition. Andrew Kingman: Conceptu- at the start of the pandemic (n = 441, May–July 2020), we show alization, Methodology, Investigation, Writing – review & editing, that NPIs significantly reshaped lives and livelihoods. Unlike in Project administration, Supervision, Funding acquisition. Hosia some other contexts (e.g. Dutta & Fischer, 2021; Gupta et al., Mavoto Moises: Methodology, Investigation, Data curation. Ana 2021; Kansiime et al., 2021), there was a paucity of external sup- Mlambo: Methodology, Investigation, Data curation. Milagre port from the state or other safety nets for our panelists; state Nuvunga: Conceptualization, Methodology, Investigation, Data interventions such as restricting transport, trading and distancing, curation, Formal analysis, Conceptualization, Writing – review & only produced adverse impacts for our panelists. Stall owners had editing, Project administration, Supervision, Funding acquisition. opening times and permissible sales curtailed, eliminating the Rose Pritchard: Conceptualization, Methodology, Conceptualiza- specifically female livelihood of brewing traditional drinks. Travel tion, Writing – review & editing, Project administration, Funding restrictions significantly affected diverse occupations such as acquisition. Clemence Zimudzi: Project administration, Writing – transport operators, wholesale-to-retail vending, stall operators, review & editing. and charcoal producers, whose train links to wholesalers and urban customers ceased altogether. For many panelists especially in charcoal communities, this led to elevated levels of food insecu- rity and hunger (Puerta Silva et al., 2020a, b; Kansiime et al., 2021), Declaration of Competing Interest prompting panelists to rely more strongly on environmental resources including forest food products (Pritchard et al., 2020). The authors declare that they have no known competing finan- Akin to previous infectious disease outbreaks creating dispro- cial interests or personal relationships that could have appeared portionate burdens on certain groups, such as Ebola on women to influence the work reported in this paper. (UNDP, 2020), our empirical data shows vulnerabilities being accentuated or created by the hyper-covariate shock of Covid NPIs in ways that may require rethinking aspects of vulnerability and Acknowledgements coping. School closures and social distancing at water collection points meant greater workloads for women, children lost out on We would like to thank all involved in our research project very education and some underage girls were at risk of early marriage, much for sharing their time and expertise with us during a very while isolation of the elderly was accentuated. Charcoal produc- difficult time – this paper would not have been possible without tion, a rare opportunity for cash income in many parts of rural you. SSA (Zulu & Richardson, 2013), was not a highly effective coping We would also like to thank the Land Team from the University strategy under Covid NPIs given a breakdown of transport links, of Edinburgh, especially Geoff Wells and Ellie Wood, for construc- reliable pricing and sales opportunities, which prompted consider- tive engagement and comments on the manuscript. We thank also able food insecurity for some panelists. The better-off, such as World Development, especially Dr Christelle Dumas, as well as two transport operators or larger-scale agricultural sellers, who would anonymous reviewers for helpful thoughts that strengthened our ordinarily be expected to benefit from diversification (Ellis, 2007), article. All errors that remain are our own. were unable to do so given the hyper-covariate nature of Covid The ‘Livelihood impacts of Coping with Covid in rural Africa’ NPIs leading to border closures, distancing and transport restric- (CwC) project was funded by The University of Edinburgh, with tions. The only value chains which largely continued to function support from the Scottish Funding Council and the Global Chal- were those involving socially-oriented investors with civic-based lenges Research Fund (GCRF). Additionally, this work was carried priorities, including maintaining fair livelihoods for baobab collec- out with financial support from the UK Government’s Foreign, tors and honey producers, contrasting sharply with particularly the Commonwealth & Development Office and the International charcoal value chain. Abiding power asymmetries in value chains Development Research Centre, Ottawa, Canada. The views (Krauss & Krishnan, 2021), and the ways they were accentuated expressed herein are those of the creators and do not necessarily under Covid, merits further research. In sum, Covid NPIs in rural represent those of the UK Government’s Foreign, Commonwealth Mozambique thus exacerbated existing vulnerabilities in varie- & Development Office, IDRC or its Board of Governors, or any other gated ways, while creating new exposures. organization. 12 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 References GoM - Government of Mozambique (2020b). Decretos Presidenciais (Presidential Decrees). (Accessed 19/10/20)https://www.portaldogoverno.gov.mz/por/ Declaracao-do-Estado-de-Emergencia/Decreto-Presidencial. Adger, W. N. (1999). Social vulnerability to climate change and extremes in coastal Günther, I; Harttgen, K. (2006). Estimating vulnerability to covariate and Vietnam. World Development, 27(2), 249–269. idiosyncratic shocks, IAI Discussion Papers, No. 154, Georg-August-Universität Agarwal, B. (2021). Livelihoods in Covid times: Gendered perils and new pathways Göttingen. Göttingen: Ibero-America Institute for Economic Research (IAI). in India. World Development, 139, 105312. https://doi.org/10.1016/ Gupta, J., Bavinck, M., Ros-Tonen, M., Asubonteng, K., Bosch, H., van Ewijk, E., et al. j.worlddev.2020.105312. (2021). COVID-19, poverty and inclusive development. World Development, 145, Ahmed, F., Ahmed, N., Pissarides, C., & Stiglitz, J. (2020). Why inequality could 105527. https://doi.org/10.1016/j.worlddev.2021.105527. spread Covid-19. The Lancet Public Health, 5(5), e240. https://doi.org/10.1016/ Hanlon, J. Smart, T. (2008). Do Bicycles Equal Development in Mozambique? S2468-2667(20)30085-2. Woodbridge, Suffolk, UK; Rochester, NY, USA: Boydell & Brewer. doi:10.7722/ Alobo Loison, S. (2015). Rural livelihood diversification in Sub-Saharan Africa: A j.ctt1bh2m6z. literature review. The Journal of Development Studies, 51(9), 1125–1138. https:// Henderson, J., Dicken, P., Hess, M., Coe, N., & Yeung, H.-C. (2002). Global production doi.org/10.1080/00220388.2015.1046445. networks and the analysis of economic development. Review of International Alon, T., Kim, M., Lagakos, D., VanVuren, M. (2020). How should policy responses tot Political Economy, 9(3), 436–464. he Covid-19 pandemic differ in the developing world? NBER Working Paper Hilson, G., Van Bockstael, S., Sauerwein, T., Hilson, A., & McQuilken, J. (2021). 27273. (Accessed 19/10/20) http://www.nber.org/papers/w27273. Artisanal and small-scale mining, and COVID-19 in sub-Saharan Africa: A Barnett-Howell, Z. Mobarak, A.M. (2020). The Benefits and Costs of Social Distancing preliminary analysis. World Development, 139, 105315. https://doi.org/10.1016/ in Rich and Poor Countries, Working Paper arXiv:2004.04867. j.worlddev.2020.105315. Barrientos, S. (2019). Gender and work in global value chains: Capturing the gains? Horner, R., & Nadvi, K. (2018). Global value chains and the rise of the Global South: Cambridge: Cambridge University Press. Unpacking twenty-first century polycentric trade. Global Networks, 18, 207–237. Bassett, H. R., Lau, J., Giordano, C., Suri, S. K., Advani, S., & Sharan, S. (2021). https://doi.org/10.1111/glob.12180. Preliminary lessons from COVID-19 disruptions of small-scale fishery supply Horvath, C., Carpenter, J. (eds; 2020). Co-Creation in Theory and Practice: Exploring chains. World Development, 143, 105473. https://doi.org/10.1016/ Creativity in the Global North and South. University of Bristol: Policy Press. j.worlddev.2021.105473. Janssens, W., Pradhan, M., de Groot, R., Sidze, E., Donfouet, H. P. P., & Abajobir, A. Batterbury, S., & Forsyth, T. (1999). Fighting back: Human adaptations in marginal (2021). The short-term economic effects of COVID-19 on low-income environments. Environment, 41(6), 6–9. households in rural Kenya: An analysis using weekly financial household Baumert, S., Luz, A. C., Fisher, J., Vollmer, F., Ryan, C. M., Patenaude, G., et al. (2016). data. World Development, 138, 105280. https://doi.org/10.1016/ Charcoal supply chains from Mabalane to Maputo: Who benefits? Energy for j.worlddev.2020.105280. Sustainable Development, 33, 129–138. Jagger, P., & Shively, G. (2014). Land use change, fuel use and respiratory health in Bennett, N. J., Blythe, J., Tyler, S., & Ban, N. C. (2016). Communities and change in the Uganda. Energy policy, 67, 713–726. https://doi.org/10.1016/j. Anthropocene: Understanding social-ecological vulnerability and planning enpol.2013.11.068. adaptations to multiple interacting exposures. Regional Environmental Change, Jones, D., Ryan, C. M., & Fisher, J. (2016). Charcoal as a diversification strategy: The 16(4), 907–926. https://doi.org/10.1007/s10113-015-0839-5. flexible role of charcoal production in the livelihoods of smallholders in central Block, E.S., Erskine, L. (2012) Interviewing by telephone: Specific considerations, Mozambique. Energy for Sustainable Development, 32, 14–21. opportunities, and challenges. International Journal of Qualitative Methods. Kansiime, M. K., Tambo, J. A., Mugambi, I., Bundi, M., Kara, A., & Owuor, C. (2021). September 2012:428-445. https://doi.org/10.1177/160940691201100409. COVID-19 implications on household income and food security in Kenya and Bonilla, Y., LeBrón, M. (eds., 2019) Aftershocks of Disaster: Puerto Rico Before and Uganda: Findings from a rapid assessment. World Development, 137, 105199. After the Storm. Chicago: Haymarket Books. https://doi.org/10.1016/j.worlddev.2020.105199. Brooks, N., & Adger, W. N. (2003). Country level risk measures of climate-related Klassen, S., & Murphy, S. (2020). Equity as both a means and an end. Lessons for natural disasters and implications for adaptation to climate change. Working Paper resilient food systems from Covid-19. World Development, 136 105104. 26. Tyndall Centre for Climate Change Research. Norwich: University of East Krauss, J. E., & Barrientos, S. (2021). Fairtrade and beyond: Shifting dynamics in Anglia. cocoa sustainability production networks. Geoforum, 120(2), 186–197. https:// Chambers, R., & Conway, G. R. (1991). Sustainable rural livelihoods: Practical concepts doi.org/10.1016/j.geoforum.2021.02.002. for the 21st century. IDS Discussion Paper 296. Brighton: IDS. Krauss, J. E., & Krishnan, A. (2021). Global decisions and local realities: Sustainability Coe, N. M., Dicken, P., & Hess, M. (2008). Global production networks: Realising the standards, priorities and upgrading dynamics in agricultural global production potentials. Journal of Economic Geography, 8, 271–295. https://doi.org/ networks. Global Networks. https://doi.org/10.1111/glob.12325. 10.1093/jeg/lbn002. Leach, M., MacGregor, H., Scoones, I., & Wilkinson, A. (2021). Post-pandemic Córdoba, D., Peredo, A. M., & Chaves, P. (2021). Shaping alternatives to transformations: How and why COVID-19 requires us to rethink development. development: Solidarity and reciprocity in the Andes during Covid-19. World World Development, 138, 105233. https://doi.org/10.1016/ Development, 139, 105323. https://doi.org/10.1016/j.worlddev.2020.105323. j.worlddev.2020.105233. Dearden, A., & Kleine, D. (2020). Interdisciplinarity, self-governance and dialogue: Leichenko, R. M., & O’Brien, K. L. (2008). Global environmental change and The participatory process underpinning the minimum ethical standards for globalization: Double exposures. Oxford: Oxford University Press. ICTD/ICT4D research. Information Technology for Development. https://doi.org/ Mahmud, M., & Riley, E. (2021). Household response to an extreme shock: Evidence 10.1080/02681102.2020.1840321. on the immediate impact of the Covid-19 lockdown on economic outcomes and Dercon, S. (2000). Income risk, coping strategies and safety nets. Paper 136, The well-being in rural Uganda. World Development, 140, 105318. https://doi.org/ Centre for the Study of African Economies Working Paper Series. Berkeley 10.1016/j.worlddev.2020.105318. Electronic Press: Centre for the Study of African Economies. Martin, S. M., & Lorenzen, K. (2016). Livelihood diversification in rural laos. World Dutta, A., & Fischer, H. W. (2021). The local governance of COVID-19: Disease Development, 83, 231–243. https://doi.org/10.1016/j.worlddev.2016.01.018. prevention and social security in rural India. World Development, 138, 105234. Mikkelsen, B. (2005). Methods for development work and research: A new guide for https://doi.org/10.1016/j.worlddev.2020.105234. practitioners (2nd ed.). London: Sage. Ellis, F. (2000). The determinants of rural livelihood diversification in developing Neilson, J., & Pritchard, B. (2009). Value Chain Struggles: Institutions and Governance countries. Journal of Agricultural Economics, 51, 289–302. https://doi.org/ in the Plantation Districts of South India. Chichester: Wiley-Blackwell. 10.1111/j.1477-9552.2000.tb01229.x. Nichols, C., Jalali, F., Ali, S., Gupta, D., Shrestha, S., & Fischer, H. (2020). The gendered Ellis, F. (2006). Agrarian change and rising vulnerability in rural sub-Saharan Africa. impacts of COVID-19 amid Agrarian distress: Opportunities for comprehensive New Political Economy, 11(3), 387–397. https://doi.org/10.1080/ policy response in Agrarian South Asia. Politics & Gender, 16(4), 1142–1149. 13563460600841025. https://doi.org/10.1017/S1743923X20000483. Eriksen, S., & Silva, J. A. (2009). The vulnerability context of a savanna area in O’Brien, K., Eriksen, S., Nygaard, L. P., & Schjolden, A. (2007). Why different Mozambique: Household drought coping strategies and responses to economic interpretations of vulnerability matter in climate change discourses. Climate change. Environmental Science and Policy, 12(1), 33–52. Policy, 7(1), 73–88. Deutsche Welle (2020). Covid-19: escolas em Moçambique vão permanecer fechadas O’Brien, K. L., Quinlan, T., & Ziervogel, G. (2009). Assessing vulnerability in the (Covid-19: Mozambican schools will remain closed). 17 July 2020. (Accessed 19/ context of multiple stressors: The Southern Africa Vulnerability Initiative 10/20). https://www.dw.com/pt-002/covid-19-escolas-em-moçambique-vão- (SAVI). Environmental Science and Policy, 12, 23–32. permanecer-fechadas/a-54209108. OECD (2007), Promoting Diversified Livelihoods, In: OECD (ed.) Promoting Pro-Poor FAO, IFAD, UNICEF, WFP & WHO. (2020). In Brief to The State of Food Security and Growth: Policy Guidance for Donors, OECD Publishing, Paris. DOI: https://doi. Nutrition in the World 2020. Transforming food systems for affordable healthy org/10.1787/9789264024786-17-en. diets. Rome, FAO. https://doi.org/10.4060/ca9699en. Oldekop, J. A., Horner, R., Hulme, D., Adhikari, R., Agarwal, B., Alford, M., et al. (2020). Tavares, F. F., & Betti, G. (2021). The pandemic of poverty, vulnerability, and COVID-19: COVID-19 and the case for global development. World Development, 134, Evidence from a fuzzy multidimensional analysis of deprivations in Brazil. World 105044. https://doi.org/10.1016/j.worlddev.2020.105044. Development, 139, 105307. https://doi.org/10.1016/j.worlddev.2020.105307. Phiri, D., Simwanda, M., & Nyirenda, V. (2020). Mapping the impacts of cyclone Idai Gautam, Y., & Andersen, P. (2016). Rural livelihood diversification and household in Mozambique using Sentinel-2 and OBIA approach. South African Geographical well-being: Insights from Humla, Nepal. Journal of Rural Studies, 44, 239–249. Journal. https://doi.org/10.1080/03736245.2020.1740104. GoM – Government of Mozambique (2020a). Decreto No. 12-2020 (Decree No. Pritchard, R., Grundy, I. M., van der Horst, D., Dzobo, N., & Ryan, C. M. (2020). 12-2020). 2 April. (Accessed 19/10/20) http://www.open.ac.uk/technology/ Environmental resources as ‘last resort’ coping strategies following harvest mozambique/sites/www.open.ac.uk.technology.mozambique/files/files/Decreto_ failures in Zimbabwe. World Development, 127, 104741. https://doi.org/10.1016/ 12_2020_de_2_de_Abril_BR_64_I_SERIE_2020.pdf. j.worlddev.2019.104741. 13 J.E. Krauss, L. Artur, D. Brockington et al. World Development 151 (2022) 105757 Puerta Silva, C., Torres Muriel, E., Amaya Epiayú, R. C., Dorado González, A., Epieyú, Swinnen, J., & McDermott, J. (2020). Assessing impacts and policy responses for food F., Frías Epinayú, E., et al. (2020a). If the coronavirus doesn’t kill us, hunger will. and nutrition security. In J. Swinnen & J. McDermott (Eds.), Covid-19 and global Regional absenteeism and the Wayuu permanent humanitarian crises. Regions food security (pp. 8–11). Washington: IFPRI Books. & Cohesion, 10(3), 140–155. https://doi.org/10.3167/reco.2020.100312. Torell, E., McNally, C., Crawford, B., & Majubwa, G. (2017). Coastal livelihood Puerta Silva, C.P., Torres Muriel, E., Amaya Epiayú, R.C., Dorado Rosales, A., Epieyú, diversification as a pathway out of poverty and vulnerability: Experiences from F., Frías Epinayú, E., Ramírez Boscán, M., Romero Epiayú, J. (2020b). If the Tanzania. Coastal Management, 45(3), 199–218. https://doi.org/10.1080/ coronavirus doesn’t kill us, hunger will: The political ecology of Wayuu permanent 08920753.2017.1303718. humanitarian crises. Collaborative paper between Wayuu Activists and Turner, B. L., Kasperson, R. E., Matson, P. A., McCarthy, J. J., Corell, R. W., Christensen, Universidad de Antioqui researchers. Keynote at POLLEN20 (political ecology L., et al. (2003). A framework for vulnerability analysis in sustainability science. conference), September 2020. PNAS, 100(14), 8074–8079. Quinn, C. H., Ziervogel, G., Taylor, A., Takama, T., & Thomalla, F. (2011). Coping with UNICEF Mozambique (2020). UNICEF Mozambique Covid-19 Situation Report No. 9: multiple stresses in rural South Africa. Ecology and Society, 16(3), 2. https://doi. 27 August – 25 September 2020. (Accessed 19/10/20) https://reliefweb.int/sites/ org/10.5751/ES-04216-160302. reliefweb.int/files/resources/UNICEF%20Mozambique%20COVID-19%20Situation Quisumbing, A., Kumar, N., Meinzen-Dick, R., & Ringler, C. (2020). Why gender %20Report%20No.%209%20-%2027%20August-25%20September%202020.pdf. matters in COVID-19 responses — Now and in the future. In J. Swinnen & J. UNDP – United Nations Development Programme (2020). Unpacking the potential McDermott (Eds.), COVID-19 and global food security (pp. 88–90). Washington: socioeconomic impact of the coronavirus pandemic in Mozambique: A United IFPRI Books. Nations Situation Analysis and Policy Recommendations. 30 March 2020. Radio Moçambique (2020). COVID-19: Interrupção de comboios no Corredor do (Accessed 19/10/20) https://www.undp.org/content/dam/rba/docs/COVID-19- Limpopo dita paralisação (COVID-19: Interruption of trains in the Limpopo CO-Response/Socio-Economic-Impact-COVID-19-Mozambique-UN- Corridor to paralyse). (Accessed 19/10/20) https://es-la.facebook.com/ Mozambique-March-2020.pdf. radiomoc/posts/3295773777134186/. United Nations (2020). Shared responsibility, global solidarity: responding to the socio- Ravallion, M. (2020) On the virus and poor people in the world. 2 April 2020. economic impacts of Covid-19. March 2020. (Accessed 19/10/20) https://www.un. (Accessed 19/10/20) https://economicsandpoverty.com/2020/04/02/on-the- org/sites/un2.un.org/files/sg_report_socio-economic_impact_of_covid19.pdf. virus-and-poor-people-in-the-world/. UN Women (2020). Policy Brief: The impact of Covid-19 on Women. 9 April 2020. Reardon, T., Bellemare, M. F., & Zilberman, D. (2020). How COVID-19 may disrupt (Accessed 19/10/20) https://www.unwomen.org/-/media/headquarters/ food supply chains in developing countries. In J. Swinnen & J. McDermott (Eds.), attachments/sections/library/publications/2020/policy-brief-the-impact-of- COVID-19 and global food security (pp. 78–80). Washington: IFPRI Books. covid-19-on-women-en.pdf?la=en&vs=1406. Renard, M.-C. (2003). Fair trade: Quality, market and conventions. Journal of Rural Valensisi, G. (2020). COVID-19 and global poverty: Are LDCs being left behind? Studies, 19(1), 87–96. European Journal of Development Research, 32(5), 1535–1557. https://doi.org/ Ribot, J. (2014). Cause and response: Vulnerability and climate in the Anthropocene. 10.1057/s41287-020-00314-8. The Journal of Peasant Studies, 41(5), 667–705. Vatican News (2020). Mocambique: Estado de emêrgencia por mais 30 dias; cultos Sapkota, P., Keenan, R. J., Paschen, J.-A., & Ojha, H. R. (2016). Social production of retomam [Mozambique: state of emergency for another 30 days; services vulnerability to climate change in the rural middle hills of Nepal. Journal of Rural resume]. (Accessed 19/10/20) https://www.vaticannews.va/pt/igreja/news/ Studies, 48, 53–64. 2020-08/mocambique-estado-de-emergencia-por-mais-30-dias-cultos-retoma. SaudeMaisTV (2020). Covid-19 suspensa circulacao de comboios de longo curso em html. Mocambique (Covid-19 stops long-distance trains from running in Mozambique). Vollmer, F., Zorrilla-Miras, P., Baumert, S., Luz, A. C., Woollen, E., Grundy, I., et al. 22 March 2020. (Accessed 19/10/20) https://www.saudemais.tv/noticia/7371- (2017). Charcoal income as a means to a valuable end: Scope and limitations of covid-19-suspensa-circulacao-de-comboios-de-longo-curso-em-mocambique. income from rural charcoal production to alleviate acute multidimensional Smith, H. E., Ryan, C. M., Vollmer, F., Woollen, E., Keane, A., Fisher, J. A., et al. (2019). poverty in Mabalane district, southern Mozambique. World Development Impacts of land-use intensification on human well-being: Evidence from rural Perspectives, 7-8, 43–60. Mozambique. Global Environmental Change, 59, 101976. https://doi.org/ World Bank (2013). World Development Report 2014: Risk and Opportunity— 10.1016/j.gloenvcha.2019.101976. Managing Risk for Development. Washington, DC: World Bank. doi: 10.1596/ Sumner, A., Hoy, C., Ortiz-Juarez, E. (2020). Estimates of the Impact of COVID-19 on 978-0-8213-9903–3. Global Poverty. UNU-WIDER Working Paper 2020/43, https://www.wider.unu. World Health Organization (2021). Coronavirus disease (COVID-19) pandemic. (Accessed edu/sites/default/files/Publications/Working-paper/PDF/wp2020-43.pdf. 30/06/21) https://www.who.int/emergencies/diseases/novel-coronavirus-2019. 14