79401 AUTHOR ACCEPTED MANUSCRIPT PRELIMINARY INFORMATION Coping and Resilience during the Food, Fuel, and Financial Crises Accepted for publication in Journal of Development Studies To be published by Taylor and Francis THE FINAL PUBLISHED VERSION OF THIS ARTICLE WILL BE AVAILABLE ON THE PUBLISHER’S PLATFORM This Author Accepted Manuscript is copyrighted by the World Bank and published by Taylor and Francis. It is posted here by agreement between them. Changes resulting from the publishing process—such as editing, corrections, structural formatting, and other quality control mechanisms—may not be reflected in this version of the text. This Author Accepted Manuscript is under embargo for external use and is made available for internal World Bank use only. It is not for distribution outside the World Bank. © 2013 The World Bank COPING AND RESILIENCE DURING THE FOOD, FUEL, AND FINANCIAL CRISES Rasmus Heltberg*, Naomi Hossain**, Anna Reva*, and Carolyn Turk* *World Bank **Institute of Development Studies, Sussex Acknowledgements: We are grateful to Veronica M. Joffre for helpful comments and to the local researchers who carried out the country studies surveyed here. The research was funded by DfID and the trust Fund for Environmentally and Socially Sustainable Development housed at the World Bank and funded by the governments of Norway and Finland. 1. Introduction How and how well have people with low incomes been coping with global economic volatility since 2008? This question is addressed through a synthesis of qualitative research in 17 countries on how selected groups of people coped with the local effects of the global food, fuel, and financial crises from 2008-11. The data are a uniquely large body of qualitative evidence that afford insights into how people affected by a common global economic shock coped as it transmitted to their diverse social and economic contexts. The paper describes and categorises the impacts of coping responses used, analysing the distribution of the costs of coping, the effects on social and gender relations, and the sources of support, including public policy, non-state support and informal social support. The paper proceeds as follows. The next section sets out motivations and key concepts. Section 3 describes the approach, setting out the methodological advantages and limitations. Section 4 describes and categorises the impacts and coping responses they elicited. Section 5 analyses sources and types of support identified by the research. Section 6 draws concludes. 1 2. Conceptual approaches to analysing crisis coping Coping with global shocks This paper aims to update the evidence on coping responses in light of recent global economic shocks. We examine the local responses to stresses originating in the global economy, drawing on qualitative research on people with low incomes or precarious livelihoods; the commonality in research approach toward the shared experience of global economic shocks enables us to draw out common characteristics of people’s coping and vulnerability in this era of global vulnerability. A distinguishing mark of this period is the global reach of crises: integration in the environmental, financial and policy spheres explains the complexity and velocity with which these crisis effects spread (see Evans et al 2009; Goldin and Vogel 2010; Held et al 2009). While the crisis also unleashed political upheaval and major fiscal contractions, this paper focuses on coping with the first and second round economic impacts from the ‘Triple F crises – the food, fuel and financial shocks that overlapped in the period 2008-2011. This crisis period can be traced back to 2007, is ongoing at the time of writing (early 2012) and is widely estimated to have adverse impacts on developing countries and people living in poverty (Devarajan and Kasekende 2011; Green et al 2010; Naude 2010; Ravallion 2008; Ruel et al 2010; Tiwari and Zaman 2010; UNDP 2009; Wodon and Zaman 2009; World Bank 2008; World Bank 2010; World Bank/IMF 2012; World Bank 2011). The main object of interest is ‘coping’, or the short-term behaviours adopted in response to economic stress (Davies, 1993; Adams et al 1998). The analysis presented here was framed by the poverty dynamics and social mobility literatures (e.g., Narayan, Pritchett, and Kapoor, 2009; Krishna, 2004 and 2007; Krishna and Shariff, 2011; Dercon, 2004; Heltberg and Lund 2009. People experience: (a) a first wave of direct impacts (through wage and employment shocks, price rises, reduced remittances) and (b) a second wave of second order impacts, such as: the effects of aggregated coping responses (e.g. increased competition in the informal sector) and individual or household coping strategies (e.g., effects of women working longer hours on their health or the care of infants). Our analysis aims to show how second waves of crisis responses feed back into both individual and covariate impacts, the latter stemming in part from the accumulated effects of individual responses (see for example, Horn 2009; WIEGO 2010). Attribution A point on attribution: the focus here is on short-term responses to sudden stresses, and not on longer- term adaptation to structural change (see Miller et al 2010; Davies 1993). Yet many of these respondents already live in poverty. This means that some coping behaviours will resemble the responses to the idiosyncratic shocks of chronic poverty. A realistic view of the impacts of global crises on the lives of people living in poverty recognizes that responses to covariate shocks involve behaviours that are broadly similar to those of everyday poverty, if perhaps more accentuated and less effective (see below). How individuals behave when the crisis is global is very like how they manage the situation 2 when husbands fall ill or cows die: reducing consumption, stretching resources, and trying to beg, borrow, sell, steal or earn more. The challenge is then how to distinguish between coping with chronic poverty and responses to global crises? Two inter-related features of the research design help here – although they do not resolve the problem entirely. The first is timing: both the IDS and the World Bank data collection periods immediately followed major global shocks, during moments when hard data - food, commodity price, financial, export and other production indicators - showed that impacts were being felt at the national level. The timing of this research is one of its unique qualities, and it was only possible because of the rapid qualitative methodologies used. This means that the research offers insights into responses to shocks as they were unfolding. Most of the study sites also involved repeat visits, allowing changes in behaviour over time to be observed. This made it possible to distinguish, with a reasonable degree of certainty, more routine responses from the fallout from globally-generated downturn and food price inflation. The second feature of the research design that gives credence to claims of being able to identify causal impacts is that in both the IDS and the World Bank approach to the research, the analysis and data collection undertook to trace the effects of global shocks through to national impacts and right down to everyday life. The qualitative evidence presented here is not disembodied voices from focus groups claiming to have experienced impacts, but the situated accounts of people in occupations and social groups whom secondary sources confirm were being collectively impacted in the ways they report. Secondary data were collected and key informant interviews conducted to verify and refine understanding of the transmission channels (see Figure 1) through which global shocks were being transmitted to the local macro economies and through to people’s lives. Data were collected and analysis conducted to triangulate findings about how, for example, export manufacturing sectors such as garments in Cambodia and Bangladesh and commodities such as cashmere in Mongolia and rubber in Indonesia were being affected by changing global demand, stimulus packages, and changing industrial practices. However, the methods and type of data used here have at least two clear limitations in establishing impact and attributing it to any particular shock. The first is that of scale and representativeness. When we state that ‘children spoke of stealing food’, we do not, of course, mean that a representative sample of children in the 17 countries mentioned stealing food from each other, but that children in one or more focus group spoke of this. We do not provide figures because the data were neither collected nor processed to permit aggregation, but also because the value of a finding of this type is not in how many people it applies to, but in the insight it provides into the quality of life and relationships among children in this community as it experiences economic crises. But while this data provide insights into quality of life and processes of change, it can offer no evidence of how widespread such impacts might be, other than of how many countries they were found in. The second limitation is also, to some degree, a strength. This is that the data comprise reported experiences of life in a time of extraordinary global volatility: they are not the same kind of data as the objective impact data used in conventional economic analysis. But they are unusually rich accounts of 3 the experiences of crises that provide a timely window onto how people lived through these shocks and how they interpreted what they did to manage. So while the paper does not shirk responsibility for establishing causal impact of crises on poor people’s lives wherever the evidence is strong enough to do so, it is also clear about the status of the evidence it uses. We have not included all findings of changing behaviours, just those for which we feel adequate triangulation and supporting data are available to provide a credibly robust case of impact. At the same time, we are aware that the primary data are subjective, and so the value of the analysis is chiefly that of enabling interpretation of the experience. Interpreting resilience A key concept is ‘resilience’, or strategies for adjusting to short-term shocks that do not result in long- term harm. The term was recently imported into development studies (see Miller et al 2010), and became prominent during the recent crises (e.g., Green et al 2010). We use this concept critically, with close attention to whether its use conceals the real burden of coping and the regressive nature of crisis impacts. This is an important potential explanation of otherwise undetected adverse impacts, as we know from the ample literature on women as ‘shock absorbers’ of crisis (Chant 1994; Elson 2009; Espey et al 2010; González de la Rocha 1988; Holmes et al 2009; Jiggins 1986; Smith 1987). As women’s unpaid care work goes largely unmeasured in official statistics, coping is easily misclassified as ‘resilience’, when in fact it costs (mainly women’s) time, effort, health, and wellbeing (Floro 1995; Razavi 2007). A critical, gender-sensitive approach to resilience (Adams et al 1998) does not preclude the possibility of full recovery, but is attuned to how hardship is distributed within households, and how community institutions are affected by people’s coping responses (e.g., coping strategies like crime, sex work or abandoning families can permanently change social norms and relationships; Hossain 2009). The evidence here enables us to see how and by whom some of the largely hidden costs of surviving the crisis have been borne, thereby highlighting the seriousness of recent global shocks, and the inability of entire societies to cope without incurring losses of wellbeing, development and social cohesion (also Global Pulse 2010; Hossain and McGregor 2011). Instead of seeing coping as functional and (mainly) successful adaptive processes, our work tells us that at the local level, the coping responses on which people customarily lean during tough times can be fundamentally overwhelmed by protracted systemic shocks (Campbell 1990; Miller et al 2010). 3. Approach and methodologies We drew on thousands of pages of reports and focus group discussion (FGD) and interview transcripts from up to four rounds of qualitative research in research sites in Bangladesh, Cambodia, Central African Republic (CAR), Ghana, Indonesia, Jamaica, Kazakhstan, Kenya, Mongolia, Philippines, Senegal, Serbia, Thailand, Ukraine, Vietnam, Yemen, and Zambia. So far as we know, this constitutes the largest body of qualitative data on crisis impacts and coping ever collected in developing countries.i 4 Figure 1: Framework for assessing crisis transmission channels Countries were selected partly on the basis of expected vulnerability to crisis as well as on demand for qualitative evidence of crisis impacts.ii Most of the countries did not have relevant quantitative data to assess crisis impacts in a timely manner, sparking interest in rapid qualitative and participatory research. Research sites, respondent samples, and questions were chosen to explore hypotheses about which population groups might be affected in each country, assessing likely transmission channels from the global crisis through to vulnerable groups (see Figure 1 and also Lustig, 2000 and Ramesh, 2009). Sampling and site selection were not designed to be nationally representative but to cover an illustrative range of social and occupation groups presumed to be vulnerable to crisis. Country case studies were undertaken by local researchers in collaboration with IDS or the World Bank (Heltberg, Hossain, and Reva, 2012; Hossain, 2009; Turk, Mason, and Petesch, 2010). The studies did not follow a single template, but similar approaches were adopted, and methodological lessons were shared across the projects. Occupation groups selected for research participation included farmers, farm owners, informal sector workers, and formal export sector workers. Respondents included some relatively well-off occupational groups (skilled factory workers in Southeast Asia), low-income groups (e.g., smallholders, vendors), and the destitute (e.g., beggars in Bangladesh). The research also undertook key informant interviews with social workers, staff of nongovernmental organizations (NGOs), business people, officials, chiefs and community leaders to gain an official or institutional perspective or triangulate findings, and collection of relevant secondary data of aggregate crisis impacts. The field activities typically involved participatory focus group discussions. The emphasis on participatory approaches aimed to enable people to set out their own concerns without being overly driven by researchers’ assumptions. Tools included prioritization exercises such as matrix ranking and 5 scoring, well-being ranking, time use assessments, food baskets and consumption analyses, occupation and social group ranking, and institutional mapping to identify sources of support. Individual semi- structured interviews were undertaken for household case studies, often over repeated research visits. Between 200 and 300 people were involved in each country, usually in several rounds. In all, the data surveyed here draws on the participation of more than 3,000 respondents, although the variety of approaches and tools mean that research findings cannot be robustly aggregated. Interview topics covered livelihoods and coping strategies, changes in paid and unpaid work, migration, borrowing, asset sales, social relations, community and private charitable support, and social protection and other government programmes. For the present paper, data were categorized and synthesized across the 17 countries. The evidence on crisis impacts focused on who and what was impacted. The evidence on coping was categorised according to the type of response cited as important to respondents, gender and age differences, the effects on social and gender relations, and the sources of support, including formal, informal, and community-based sources. Qualitative monitoring data of this kind can complement quantitative data and modelling approaches (Benson et al. 2008; World Bank/IMF 2012). Its advantages include speed, direct observation over estimation, and broader scope of research themes and topics, including subjective and relational well- being, community cohesion, informal and customary sources of social protection, and women’s unpaid care work. Unlike ex-post research, qualitative monitoring enables a focus on process as well as outcome. However, qualitative approaches are disadvantaged by small scale and purposive respondent selection: we cannot say by how much poverty, school enrolment, or community cohesion worsened. The findings are illustrative of the processes and experiences, which when situated within their contexts, enable a contribution to both the empirical evidence and the conceptual understanding of vulnerability, resilience and coping with economic crisis. 4. impacts and coping responses An overview of impacts and responses While economic growth rates barely dipped for many developing countries during the financial crisis, people on low incomes within those countries often experienced hardship during the period 2008-11. For some, these were severe and protracted, but others proved less vulnerable and more resilient. The impacts varied in timing, duration and severity, reflecting how people were positioned with respect to the transmission channels (see Figure 1). Workers in the formal economy were affected first and most directly through layoffs, reduced working hours and benefits. Yet, formal sector workers turned out to be relatively resilient: some received severance payments or lived off savings. Agricultural and informal sector workers experienced the commodity price shocks (e.g., for rubber, cashmere, recycled waste products) of early 2009 but were affected more indirectly through reduced demand, increased competition from laid-off factory workers, and reduced remittances. Many informal sector workers struggled even before the crises, in general, were affected the most and the longest, and struggled to recover, as later rounds of research found. 6 Working longer hours and more informal jobs or self-employment was commonly reported, but the rewards tended to be small, partly because of sharper competition and falling demandFood insecurity was by far the most severe and widely felt impact. Food insecurity could be attributed to the global food price spikes because findings over the rounds of research corresponded to global and national price movements: people reported stress when high global food prices transmitted into higher national prices, and when prices dropped (in late 2009 and early 2010), many experienced this as a moment of recovery. Food insecurity was most pronounced where the global crises were compounded by local shocks such as prolonged drought and political violence (Kenya) and an energy crisis (Central African Republic). Respondents described major stresses securing regular meals, and in the worst cases, death by starvation: “I lost my two children a year ago. Since their birth they never knew a happy life, they were always sick. The last time I sent them to the health center, the diagnostic was the lack of blood cells and malnourishment. The nurse told me to feed them with more milk and beans. Because I didn’t have enough money to buy these products, they died.” 51 year old woman, mother of six children, CAR (August 2010) Severe food insecurity experienced as weakness, dizziness and lack of energy were reported in Zambia and Bangladesh in 2009. In Southeast Asia and elsewhere, extreme hunger was not reported, but in common with households in South Asia and sub-Saharan Africa, people reduced the quantity and quality of food, and adults forewent meals so that children could eat. Less drastic reductions in food expenditure were reported in Eastern Europe and Central Asia (ECA). Coping with food insecurity led to sharp rises in stress as people struggled to feed families, maintain remittance levels, or repay loans: ”We have so little money we can almost not survive. How can we be happy? Now there are only headaches and tears.” FGD participant in Baboang Village, Cambodia (July 2009) In Eastern Europe, use of counselling and social worker assistance increased: "People started to ask more for [psychologist and social worker] consultations and the number of such people increased three-fold during last year. We even resumed the operation of the telephone line for psychological counselling, which had been closed long time ago…” From an FGD in Slavutych, Kiev region, Ukraine (August, 2010) A common finding was of resorting to self-medicating, traditional healers, and unlicensed medicines. In a CAR focus group, people were said to have died because they could not afford healthcare. In Kenya, people living with HIV/AIDS were taking antiretroviral drugs on empty stomachs. Respondents in Bangladesh postponed serious medical treatment. Increased indebtedness and asset sales were common. In Cambodia, Thailand, and the Philippines people had lost or feared losing their land or homes to pay off unserviceable debts. Consumer goods 7 (jewellery, electronic appliances) were sold in South and Southeast Asia. In Kazakhstan, respondents sold houses, cattle and cars to pay off debts incurred before the crisis. In Ghana, shea nut gatherers first sold off shea nuts (a form of savings), and then sold livestock in order of value, starting with poultry. In Kenya and CAR sale of assets was not commonly reported, because poor households had already exhausted their resources and had little left to sell. Distress asset sales disadvantage the poor, as the Mongolian case illustrates. Mongolian herders hit by declining world cashmere prices sold their herds at a low price. Wealthier herders waited for cashmere prices to recover by drawing on savings and alternative incomes, taking advantage of better knowledge of price movements and access to transportation to sell at higher prices in urban markets. Gendered effects Both men and women were affected by job losses, higher cost of living, harder work and worsening employment conditions. Yet in most countries, the impacts affected women doubly, because they were concentrated in low-paying and low-skilled occupations and bore the main burden of unpaid household or care work. In the formal sector, Southeast Asian women workers suffered disproportionately from industrial downsizing as they were overrepresented among the low-skilled, and companies prioritized retaining skilled employees who were mainly men. In Mongolia and Indonesia, workers noted that hiring decisions became more discriminatory than before, as labour markets tightened during the crisis. In Thailand, women said they had received less in severance payments than men. Indonesian export factories were said to prefer women workers after the crisis, as women were believed to be more amenable to longer hours without secure contracts and pressure for productivity. In Serbia and Ukraine, women with small children said they found it more difficult to find work, and were more likely to be laid off; Serbian single mothers said employers tried to exploit their vulnerability by asking them to work extra hours for low pay, or for sexual favours. Job losses and higher prices meant many women sought additional income. This typically meant work in the low-paid informal sector, involving long hours leaving small children in the care of older siblings. For some women, as in Mongolia, new involvement in paid work improved their self-esteem, role in decision making and social life; for others, as in Yemen, women’s new roles in paid work were a threat to social norms and to men’s employment. The widely-reported emotional pressures of facing hungry children often made women more entrepreneurial: in many sites, women were perceived to more often diversify into new activities than men and more prepared to accept low status, physically demanding and demeaning jobs, like cleaning, laundry and sex work. In Cambodia, garment workers that lost their jobs went to work in karaoke bars and massage salons, low status jobs with risk of sexual harassment. Sometimes care work constrained women’s coping options. In ECA, women complained about the lack of kindergartens, while in Southeast Asia women accepted poorly paid subcontracts that allowed them to combine home-working with their care responsibilities. Women were generally seen as better able to manage on tighter household budgets and to save and remit money to relatives during the difficult times, highlighting the role of women as shock absorbers. 8 The work of stretching meagre household budgets was mostly done by women, including shopping around for bargains, walking further to collect fuelwood or wild foods, and recycling clothing. Despite longer hours in paid work, women typically bore the burden of these more onerous household responsibilities, and women’s total working hours were said to be longer and to have increased more than men’s. In CAR an interviewee said “It is the sleep which drags her away from her daily housework”; in Kenya rural women reported 18 hour working days. The strains of coping with reduced incomes and rising unemployment contributed to increased domestic violence, by men against women and by women against children. In a Cambodian village, focus group participants linked an increase in domestic violence to reduced incomes as a result of the crisis: “Many families seem to have very frequent arguments and most of the cases relate to income … Some wives cannot stay in their house because they are afraid of their husbands”. FGD in Donn Vong Village, Cambodia (July 2009) In Mongolia, male unemployment and alcohol abuse was given as reasons for increased domestic violence: Sometimes we find women having black eyes. I observe that unemployed men usually spend their time drinking alcohol and playing game. When they are at home, drunken men often get into conflict with family members and start fighting. th A Social worker of 25 khoroo, Songino Khairkhan district, Ulaanbaatar, Mongolia (November 2010 – January, 2011) Female household headship increased in some places, as men migrated for work (without always sending remittances), were jailed (crime rates reportedly rose in most contexts) or in severe cases, abandoned their families. Men who lost their jobs were sometimes deeply affected by the emotional stress of the inability to provide for the family. Many men migrated, did hard physical labour without adequate nutrition and risked their personal safety by saving on accommodation (construction workers and cyclo drivers in urban Cambodia slept on the street) Impacts on children and youth Across contexts, parents sought to protect their children’s food consumption and education, with mixed success. Early on in the crisis, the evidence was that parents were reducing school costs (e.g., by transferring to cheaper schools), but prioritizing keeping children in education. This response differed from previous crises, and may reflect improved access to schooling since the 1990s (e.g., fee abolition, cash or food transfers) and changing values regarding the importance of education. . Yet dropout rates rose in several places, particularly in the Bangladeshi and African sites, and linked by respondents to food price rises. Concerns over rising teenage pregnancy rates were raised in Bangkok, Nairobi and Lusaka. Even when dropout did not result, hunger and casual work opportunities meant children’s attendance 9 became more erratic compared to pre-crisis with likely adverse long-term consequences for educational attainment and transition to higher levels of education. The situation was markedly better in the ECA and Mongolia sites, with their long traditions of universal school education. School dropouts and increased absenteeism were not reported, but parents reduced expenditures on school, including transport and school supplies. In Jamaica and Yemen, women admitted that they took their frustrations out on children: “Women beat their children more than men do. When women have nothing to eat, they beat their children and verbally abuse them. Women are more stressed.” FGD participant in Kingston, Jamaica (2009) “When I don’t have anything in the house and I am upset, I beat up the children.” FGD with a woman in Yemen (2010) Longer working hours and lack of childcare facilities meant children were often left alone or in the care of older siblings. In Serbia, single mothers complained that their children developed behavioural and school problems as they had no time to support them. In CAR, illness was said to have increased among small children being cared for by siblings. The effects on Kenyan children seemed particularly acute. Families in the Kenya sites could afford one or two meals per day through the 2009-2011 periods. Teachers reported declining student attendance and more children in waste collection, crime, and sex work. Children spoke of stealing food from each other. There were also concerns about abandoned children, as parents migrated to urban areas for work. School feeding programmes were highly valued and raised school attendance and performance. Youth were often seen as particularly vulnerable. In Mongolia and Kazakhstan, better-off students continued in education, while poorer students terminated or interrupted their studies, in a sign that coping responses could lead to increased future inequalities. Reports came from Ghana and Jamaica of students deferring or dropping out of college. In Vietnam, increased competition for jobs drove increased enrolment in vocational schools as workers took advantage of flexible class schedules to combine work with studies. Increased alcohol and drug use and engagement in criminal and anti-social activities by youth at this time were widely cited, and causally linked to the crises. In both Kenya and Zambia, people feared that the incidence of sexually transmitted diseases would rise, as sex work and drug use increased. Youth crime was a specific concern: “The [youth] who have completed school just stay at home, without jobs even when they have completed primary school. Most of these were not able to proceed to secondary school. A lot of these youth have ended up organizing themselves into groups. They end up on the centres and in town engaging in crime to survive. They have to steal, even violently; because they have no food and they must have at least a meal each day” Interview in Lango Baya, Kenya (March 2009). 10 Increased drug use and sex work were also observed in Thailand, where interviewees reported that children as young as ten were being employed to sell drugs. Community level impacts Increased economic hardship was widely associated with increased crime, drug and alcohol abuse and weaker local solidarity across the communities. The extent to which social norms were eroded varied, depending on the severity of the impact and the strength of local institutions to begin with. Yet in all contexts there were indications of strain: “In the black market when the working day ends, people all get drunk. They sell something and drink the money. Homeless people come to collect leftover food and steal things…” A candy and fruit trader from Narantuul market, Ulaanbaatar, Mongolia (August-September 2009) Crime was a source of concern in rural and urban sites in most countries. In the Philippines, women were concerned by the rise in crime, including kidnapping and rape. Many rural people worried about increased crop and livestock theft. In rural Kenya, a farmer said that theft of food crops from the fields discouraged cultivation. In Mongolia, animal theft was reported to be on the rise, and the poor were more vulnerable as they had to leave animals unattended. Competition for scarce work and resources sometimes divided people: for instance in Thailand, port workers were hostile towards illegal migrants from other countries who worked for lower pay; Serbian Roma reported increased discrimination by employers; and in Kenya, there were criticisms of mosques that assisted Muslims but not people of other faiths. 5. Sources of Support Credit In Kazakhstan, Serbia, Ukraine and Mongolia employed people could access bank credit, but its usefulness in crisis coping varied. Bank credit and microfinance institutions enabled herders and waged workers in Mongolia to cope. Bank credit also helped some people cover basic consumption needs in Ukraine. But unsustainable debt was a concern for the Roma in Serbia and in Kazakhstan, where interviewees sold assets, hid or moved to escape debts. Access to formal finance was limited elsewhere. In Africa, few respondents had access to banks or microcredit, while in Southeast Asia, informal sector workers and farmers relied on family or high interest moneylender loans. In the Philippines and Indonesia, community-based savings groups helped, but in Indonesia, interviewees reported that traditional savings groups met less frequently or stopped at the height of the crisis. In West Africa, women’s and professionals’ tontines (rotating savings and credit groups) were often the main source of financial assistance, although the credit amounts were inadequate for new businesses. In Senegal, many 11 poor people ended their membership in tontines due to inability to make contributions. Shops were a common source of customer credit, but this sometimes meant losses for business owners and even closures. Unserviceable debt to microfinance institutions and moneylenders was a source of distress in Cambodia, Thailand, the Philippines and Bangladesh. In Cambodia, rural people took moneylender loans to repay microfinance institutions; many finally sold motorbikes, boats, livestock, land or even their homes to pay debts, while others lived in fear of creditors: ‘We sell whatever we have to get the money to cope with the loan but it is still not enough … we are afraid of the credit agent like we are afraid of a tiger. ’ FGDs in rural Cambodia (July 2009) Social safety nets With the exception of the ECA and Mongolia sites, vulnerable people benefited little from formal social protection. The key problem was inadequate coverage, but other concerns included poor design and targeting, inadequate amounts and poor quality (of in-kind assistance). Free or subsidized education and school feeding programmes were popular and important sources of support, as were free or subsidized health services. In Serbia, Ukraine, Kazakhstan and Mongolia, many respondents benefited from universal old-age pensions, subsidies for utilities, child allowances, unemployment benefits, and employment services. Government transfers were too small to cover all needs and sometimes did not reach the most vulnerable (e.g. the Roma in Serbia and informal gold miners in Mongolia), but were generally perceived to have helped. In other countries, there were some programmes, but they generally did not address the key vulnerabilities (e.g. indebtedness in Cambodia and the Philippines, food insecurity in Senegal, CAR, and Zambia). In Kenya, those that accessed a food-for-work program found it a mixed blessing, as the food earned barely compensated for the hard work, particularly when claimants were already malnourished after years of drought. Other programmes were not targeted to the neediest: utility subsidy and loan refinancing programmes in Thailand effectively excluded slum dwellers that lacked access to basic utilities and poor borrowers from informal moneylenders. In several countries, political considerations and corruption were alleged to influence targeting. In Mongolia, official fodder and hay distribution never reached some areas governed by the opposition. In Cambodia, rural food aid distribution was linked to political loyalties rather than need. In Ukraine and Kazakhstan, bribery and nepotism were perceived to influence access to employment and vocational training opportunities. Benefits were often too small to matter. The government of Mongolia distributed 10,000 tugrugs a month (around US$ 8) to every citizen since 2010 - amounts that were too small to have any real impact. In rural Bangladesh, recipients of the old age and widows’ allowances asked: “Can anyone feed 12 themselves on only Taka 200 (US$ 2.89) per month?” In Zambia, the government halved the amount of fertilizer and seed distributed to double coverage: interviewees complained that the inputs received were too negligible to improve production; even so, many eligible households never received any fertilizer. Assistance provided in-kind was sometimes of poor quality. In the Philippines, fertilizer from government was said to have a negative impact on crops. In Kenya, the Ministry of Agriculture provided farmers with ‘improved’ male goats that did not survive in the local climate. In Bangladesh, women complained that subsidized rice smelled bad and spoiled if not eaten immediately after cooking. Official crisis stimulus schemes generally excluded informal sector enterprises. In Thailand and Vietnam, small and informal businesses were ineligible for the subsidized loans that bigger firms received. Workers laid-off from formal enterprises received severance payments, while informal sector workers did not. This group lacked access and knowledge of finance and markets, and were exposed to more crime and racketeering as a result of the crisis. NGOs and religious organizations NGOs and religious organizations played a positive but limited role in mitigating the effects of crisis. There was no evidence that activities changed or scaled up in response. NGOs projects were often well- regarded but with few beneficiaries and long-term goals, rather than the immediate support people needed at the height of the crisis (e.g., cash or food, work, affordable loans or credit refinancing). Religious organizations were appreciated in many sites for their moral and material support, but levels of support were often nominal, and fell as contributions declined and need grew. In Nairobi, school feeding programmes and other church assistance were open to all, yet assistance from mosques was reportedly limited to Muslims. In Thailand, a temple provided food to needy families, and motorcycle taxi drivers took turns to ask for food from the nearby temple. In Bangladesh, a Sufi shrine fed 500 people daily at the height of the food crisis. In Cambodia, the crisis saw a rise in claimants from and a decline in contributions to the pagodas: assistance became restricted to the extreme poor and the sick. The moral support provided by churches was valued by respondents in Zambia, but material assistance had declined compared to previous years. In CAR, religious organizations and mutual solidarity networks were a key source of support, more so than elsewhere. People appreciated the mutual moral support, the psychological relief from prayer, and the material support during events such as birth, marriage, death or illnesses. Reliance on religious organizations and adherence to faith reportedly increased during the crisis. Informal safety nets In each country, informal safety nets were a major –often the only – source of support for the poor. Informal support was usually between family, friends, neighbours, and within solidarity networks 13 formed along lines of common occupations or locality. Membership of a kin, occupational or community group made people stronger and helped them cope. But as the crisis unfolded, these solidarity mechanisms were weakening as funds dried up and needs grew, although close relatives generally remained important. Informal safety nets—altruistic gifts, remittances, free meals and so on to the needy—became less reliable as the better-off community members who would normally be donating faced hardships themselves. This led to local altruism becoming far more sharply targeted and sometimes rationed along ethnic or religious lines. “‘We help only elderly people who have nobody (no children or no relatives) to look after them – giving them food or anything they need to survive. If elderly people have children or relatives we do not help them, because we are also poor and have to take care of ourselves as well. But we will not allow any people to die from hunger in our village.” Interviews in rural Cambodia (July 2009) People did not report a decline in non-financial support - moral support, help with finding jobs, sharing tools. In Kenya, farmers in irrigation associations shared water pumps; Senegalese traders shared market spaces; Hanoi day labourers shared the available work so that nobody went entirely without. In Cambodia, construction and garment workers pooled money or rice and cooked and ate in large groups. These eating groups grew as the crisis unfolded. Homeless cyclo drivers and unskilled construction workers lived together to protect against gang attacks. In Cambodia, Mongolia and Kazakhstan, people relied on friends and relatives to find jobs in urban areas. Support from immediate family members was often extensive, including remittances, sharing food and other in-kind assistance (e.g. lending livestock), interest free loans and paying tuition fees. Family ties were also used widely in job-hunting. There was no evidence of permanent damage to the social fabric underpinning mutual solidarity arrangements, but capacity to support the growing number of people in need was stretched. Remittances played an important role: “Migration is the most important solution for employment and incomes. If all migrants returned home, we would come back to eating porridge only”- said a local official in Vietnam. Yet, as the crisis unfolded and hit urban migrants, the amount and regularity of remittance transfers declined in some places, causing great hardship. The return of some urban migrants increased competition for jobs and reduced already low wage rates, while returning international migrants struggled to repay large debts incurred to go abroad. 6. Conclusions A key finding from this synthesis of qualitative evidence is that crises impacts and coping responses imposed severe hardship, and the main safety nets for most people were relatives, friends and mutual solidarity groups. These depleted as the crisis went on. The conclusion is therefore that formal social protection was inadequate for a global shock on the scale, complexity and magnitude experienced by people living in poverty in these places across 17 countries. Three years of compound crises depleted 14 the resources of vulnerable people, and many may not be able to take advantage of new economic opportunities without targeted assistance. Furthermore, food and fuel prices remained high or increased further with economic recovery, while incomes stagnated or rose insignificantly. The analysis of the case studies demonstrates the need for crisis response policies to go beyond safety nets, and include more effective and generous social protection systems, free health and education, support packages for informal enterprises and smallholder farmers, including access to credit and debt refinancing schemes, as well as measures to strengthen community cohesion. A methodological conclusion is that qualitative analysis of crisis impacts can complement quantitative approaches. The methods enabled a closer look at the impacts on the informal sector that employs the majority of population in developing countries yet is often ignored, and drew attention to those usually excluded from household surveys (e.g., beggars, informal gold miners, illegal migrants). It also enabled analysis of crisis impacts by age and gender, highlighting in particular the role of women’s unpaid work in household resilience, and helped assess a wider range of impacts beyond the material, pointing to psychological stress and its effects on family and community relationships, demonstrating the importance of communal solidarity in cushioning against shocks. Over time, the qualitative monitoring approach made it possible to assess the damage of prolonged economic stress to traditional support mechanisms. The study has shown that the analysis of macroeconomic indicators alone conceals important aspects of economic recovery. In particular, the resumption of GDP growth from late 2009 in many developing countries gave optimism that the impacts of the crises would be short-lived and that the poor would not be strongly affected. Our findings do not support this optimism. Despite the narrow employment base of export production in developing countries, formal sector workers fuel rural economies through remittances and provide a customer base for the informal sector: even temporary labour shocks in the formal economy have lasting impacts on poverty. The first wave of coping responses adopted by poor people (incurring debt, forgoing health care, diversification) led to the second wave (selling assets or increased competition in the informal sector). In many surveyed countries, poor people were living through this second round of negative impacts while national economies were showing strong signs of economic recovery. The evidence here indicates that the recovery of people and households depends in particular on replenishing the human resources depleted in the effort to absorb the shocks; this means focusing on rebuilding the household or care economy, and in future crises, paying closer attention to the impacts on this vital yet unpaid and uncounted sector. Although the coping mechanisms used to cope with the global food, fuel, and financial crises are similar to those used for more localized shocks, there is a sharp difference in their effectiveness. The traditional informal safety nets of the poor became increasingly depleted and ineffective because of the large and protracted crisis. The informal sector was hit by the crisis and exposed to suddenly increased competition and reduced remittances, and religious and mutual solidarity dried up as needs grew and fewer could afford to contribute. People often tried to identify new livelihood options and move to other sectors, localities, and occupations. These attempts were often unsuccessful, in part because they were simultaneously pursued by many other competitors, yet there were also examples of adaptation of business strategies and creative diversification that helped people cope through the bad times. 15 7. REFERENCES Adams, A. M. Cekan, J., and Sauerborn, R. (1998) ‘Towards a Conceptual Framework of Household Coping: Reflections from Rural West Africa’. Africa: Journal of the International African Institute, Vol. 68, No. 2 (1998), pp. 263-283 Campbell, D. J. (1990) ‘Strategies for coping with severe food deficits in rural Africa: A review of the literature’. Food and Foodways: Explorations in the History and Culture of 4:2, 143-162 Chant, S. (1994) ‘Women, Work and Household Survival Strategies in Mexico, 1982-1992: Past Trends, Current Tendencies and Future Research’. Bulletin of Latin American Research Vol. 13 (2) 203-33. Davies, S. (1993) ‘Are coping strategies a copout?’ IDS Bulletin Vol. 24 (4) 60-72 Dercon, S. (ed.), 2004, Insurance against Poverty, Oxford University Press in collaboration with WIDER. Devarajan,, S. and Kasekende, L. A. (2011) ‘Africa and the Global Economic Crisis: Impacts, Policy Responses and Political Economy’ African Development Review Vol. 421-38 no. 4l: 421-38 Elson, D. (2010) ‘Gender and the global economic crisis in developing countries: a framework for analysis’. Gender and Development Vol. 18 (2) Espey, J., Harper, C., and Jones, N. (2010) ‘Crisis, care and childhood: the impact of economic crisis on care work in poor households in the developing world’. Gender & Development 18 (2) 291-307 Evans, A., Jones, B. and Steven, D. (2010). ‘Confronting the Long Crisis of Globalization’, Brookings/CIC report Floro, M. S. (1995) ‘Economic restructuring, gender and the allocation of time’. World Development 23 (1) 1913-29 Global Pulse (2010), Voices of the Vulnerable: recovery from the ground up. United Nations. Goldin, I. and Vogel, T. (2010) ‘Global Governance and Systemic Risk in the 21st Century: Lessons from the Financial Crisis’. Global Policy Vol. 1 (1) 4-15 González de la Rocha, M. (1988) ‘Economic Crisis, Domestic Reorganisation and Women's Work in Guadalajara, Mexico’ Bulletin of Latin American Research Vol. 7 (2) 207-223 Green, D., King, R., and Miller-Dawkins, M. (2010). ‘The Global Economic Crisis and Developing Countries: Impact and Response’. Oxfam Research Report. Oxford: Oxfam International Held, D., Kaldor, M. and Quah, D. (2010) ‘The Hydra-Headed Crisis’. LSE Global Governance pamphlet. London: London School of Economics. 16 Heltberg, R., N. Hossain and A. Reva, eds(2012) Living through Crises: How the Food, Fuel, and Financial Shocks Affect the Poor, Washington DC: World Bank New Frontiers of Social Policy. Heltberg, R. and N. Lund (2009) “Shocks, Coping, and Outcomes for Pakistan’s Poor: Health Risks Predominate”, Journal of Development Studies (2009) 45:6,889-910. Holmes, R., Jones, N. and Marsden, H. (2009) ‘Gender vulnerabilities, food price shocks and social protection responses’. ODI Background Note. London: Overseas Development Institute Horn, Z. E. (2009). No Cushion to Fall Back On: The Global Economic Crisis And Informal Workers. Women in Informal Employment: Globalizing and Organizing (WIEGO)/ Inclusive Cities Project report. Hossain, N.et al (2009), Accounts of Crisis: Poor People’s Experiences of the Food, Fuel and Financial Crises in Five Countries. Report on a pilot study in Bangladesh, Indonesia, Jamaica, Kenya and Zambia, January-March 2009’. Mimeo. Brighton: Institute of Development Studies, UK. Hossain, N. (2009) ‘Crime and Social Cohesion in the Time of Crisis: Early Evidence of Wider Impacts of Food, Fuel and Financial Shocks’. IDS Bulletin vol. 40 (5) 59-67 Hossain, N., R. Fillaili, G. Lubaale, M. Mulumbi, M. Rashid, and M. Tadros. (May 2010). ‘The Social Impacts of Crisis. Findings from community–level research in five developing countries’. Mimeo. Brighton: Institute of Development Studies, UK. Hossain, N. and McGregor, J. A. (2011) ‘A ‘lost generation’? Impacts of complex compound crises on children and young people’. Development Policy Review vol. 29 (5) 565-84 Jiggins, J., (1986) ‘Women and Seasonality: Coping with Crisis and Calamity’. IDS Bulletin 17 (3) 9-18 Anirudh Krishna, For Reducing Poverty Faster: Target Reasons Before People, World Development, Volume 35, Issue 11, November 2007, Pages 1947-1960. Anirudh Krishna, Escaping Poverty and Becoming Poor: Who Gains, Who Loses, and Why?, World Development, Volume 32, Issue 1, January 2004, Pages 121-136. Anirudh Krishna, Abusaleh Shariff, The Irrelevance of National Strategies? Rural Poverty Dynamics in States and Regions of India, 1993-2005, World Development, Volume 39, Issue 4, April 2011, Pages 533- 549. Lustig, N. C. (2000) ‘Crises and the Poor: Socially Responsible Macroeconomics’, Economia 1 (1): 1-19. Miller, F., Osbahr, H., Boyd, E., Thomalla, F., Bharwani, S., Ziervogel, G., Walker, B., Birkmann, J., van der Leeuw, S., Rockström, J., Hinkel, J., Downing, T., Folke, C., and Nelson, D. (2010) ‘R esilience and Vulnerability: Complementary or Conflicting Concepts?’ Ecology and Society 15 (3): 11 Narayan, Deepa, 2000, The Voices of the Poor, OUP. 17 Narayan, Deepa, Lant Pritchett, and Soumya Kapoor, 2009, Moving Out of Poverty Volume 2 Success from the Bottom Up, Palgrave McMillan and the World Bank. Naudé, Wim (2010) "The Global Economic Crisis and Developing Countries: Effects, Responses and Options for Sustainable Recovery," Poverty & Public Policy: Vol. 2: Iss. 2, Article 8. Ramesh, M. (2009) ‘Economic Crisis and its Social Impacts: Lessons from the 1997 Asian Economic Crisis’, Global Social Policy 9: 79-99 Razavi, S. (2007) ‘The Political and Social Economy of Care in a Development Context: Conceptual Issues, Research Questions and Policy Options’. United Nations Research Institute for Social Development. Gender and Development Programme Paper Number 3 Ruel, M. T., Garrett, J. L., Hawkes, C., and Cohen, M. J. (2010) ‘The Food, Fuel, and Financial Crises Affect the Urban and Rural Poor Disproportionately: A Review of the Evidence’ The Journal of Nutrition. Supplement: The Impact of Climate Change, the Economic Crisis, and the Increase in Food Prices on Malnutrition Smith, J. (1987) ‘Transforming Households: Working-Class Women And Economic Crisis’ Social Problems no. 34 (5) 416-36 Tiwari, S., and Zaman, H. (2010) ‘The impact of economic shocks on global undernourishment’. World Bank Policy Research Working Paper Series No 5215. Washington DC: World Bank Turk, Carrie; Andrew Mason; and Patti Petesch (2010), Impacts of the Economic Crisis on Vulnerable Population Groups: Midcourse Findings from a Qualitative Monitoring Project in 12 Developing Countries. World Bank, processed. UNDP (2009) Human Development in Crisis: Insights from the Literature, Emerging Accounts from the Field and the Correlates of Growth Accelerations and Decelerations. Social Policy Working Paper. New York: UNDP. WIEGO (2011) Coping with Crises: Lingering Recession, Rising Inflation, and the Informal Workforce, WIEGO/Inclusive Cities Project report Wodon, Q. and Zaman, H. (2009) ‘Higher Food Prices in Sub-Saharan Africa: Poverty Impact and Policy Responses’. World Bank Research Observer 25:157–176; Ravallion, Martin (2008) ‘Lessons from World Bank Research on Financial Crises’, Policy Research Working Paper No. 4779. Washington, DC: World Bank. World Bank (2008) Rising Food and Fuel Prices: Addressing the Risks to Future Generations. Washington, DC: Human Development Network, World Bank. World Bank (2010) The Millennium Development Goals and the Road to 2015: Building on Progress and Responding to Crisis. Washington DC: World Bank. 18 World Bank/IMF (2012). ‘Global Monitoring Report 2012: Food Prices, Nutrition, and the Millennium Development Goals’. Washington DC: World Bank/International Monetary Fund. World Bank Thailand, 2010, Thailand Economic Monitor November 2010, Bangkok. World Bank (2011) ‘Food Price Watch report April 2011’. Washington DC: World Bank. i The Voices of the Poor (Narayan, 2000) and Moving out of Poverty (Narayan, Pritchett, and Kapoor, 2009) studies were larger but not focused on crisis and coping. Some of the underlying country case material has been published in Heltberg, Hossain, and Reva (eds, 2012). ii For a more detailed discussion of site selection, see chapter 1 in Heltberg et al eds (2012). 19