Andrew Shepherd,3 Vidya Diwakar,4 Mary Lubungu,5 Richard Bwalya,6 Antony Chapoto,7 Lucia da Corta,8 Marta Eichsteller,9 Marja Hinfelaar,10 Arthur Monze Moonga,11 Brian Mulenga,12 Kate Pruce13 and Joseph Simbaya14
Abstract The past ten years have witnessed a greater intensity of climate-related crises in Zambia than ever before. Alongside this, the number of people living in poverty in Zambia continues to grow. This article investigates key drivers of resilience to poverty and to climate change in Zambia, and as part of this, what the key institutional responses are and policy opportunities to strengthen climate resilience alongside resilience to poverty. It relies on mixed-methods data from 2010 to the present in Zambia: the Rural Agricultural Livelihoods Survey panel (2010–19), the Living Conditions Monitoring Survey (2022), and qualitative data collection across nine provinces in Zambia (2019–24). Study findings indicate that climate resilience is in great part built on a foundation of escaping poverty and developing capacities to remain out of poverty. However, degrees of climate resilience can be enabled at lower levels of wellbeing through policies that support climate-smart agriculture and economic diversification.
Keywords poverty dynamics, climate resilience, climate-smart agriculture, Zambia, mixed methods.
Climate change has become a significant challenge for the population of Zambia, with droughts especially in the south and west (recently, in 2015/16, 2019/20, and 2023/4), and flooding particularly in the north (World Bank 2019; ACAPS 2025). These crises often have devastating impacts on people living in and near poverty as they may lack the capacities or support to remain resilient to poverty and climate change. Becoming resilient to the effects of climate change involves adaptation, the reduction of extreme poverty, continued investment in strengthened disaster risk management, climate proofing institutions, and macroeconomic policy to support upward mobility (Bahadur et al. 2015; UNDP 2025; GIZ 2021). Resilience to climate change is closely related to resilience to poverty when viewed from the perspective of poor and vulnerable people; shocks (including climate shocks) set people back when they have been making progress escaping poverty (Diwakar et al., this IDS Bulletin). These shocks may be sudden and acute, as in a flood or a period of power cuts, or slower onset and cumulative, as in a drought, where the effects can deepen a long time after the onset; or even longer-term changes in temperature, rainfall patterns, river flows, and associated fish and wild populations of other animals, soil quality, and lake and groundwater levels.
Longer-term resilience to climate-related stressors, such as slowly increasing average temperatures and a more variable, less predictable climate, is also likely to depend on the diversification of livelihoods that people can achieve, as argued in this article and that by Joseph (this IDS Bulletin). The capacity to diversify livelihoods is a function of savings and transfers allowing investment; strategies to exploit market niches; supportive spousal relationships and family networks; education; a family culture of risk-taking entrepreneurship; and the evolving state of local markets for labour, financial services, land, and other business assets (Diwakar and Shepherd 2022). However, climate change can also undermine the capacity to diversify successfully. For example, the wells that women throughout Zambia rely on for small-scale dry season gardening, as a form of agricultural diversification alongside farming, can also dry up in a drought year, thus removing a hard-won opportunity for diversification.
People’s climate and poverty resilience is affected by wider institutional considerations. This may include how robust the agricultural research and extension system is for technological innovation, so that crops, animals, and fish can endure hotter weather, longer dry spells, and higher/longer flood episodes; the quality of the disaster risk management system in terms of preparedness for sometimes multiple simultaneous disasters or sequences of different kinds of disasters, managing human welfare during disasters, and helping people to recover afterwards; the flexibility and robustness of financial services; and the degree to which education is preparing children and young people for diversification into non-traditional occupations (Diwakar et al., this IDS Bulletin). The resilience of individuals, households, and communities to poverty and climate change is partly determined by the wider resilience of the systems in which they are embedded: weak individual and household resilience may to some extent be compensated by strong system resilience and vice versa. Yet little is known about system resilience in Zambia, beyond the implications of technology and cash transfers (Berretta et al. 2023).
This article investigates key drivers of resilience to poverty and climate-related disasters in Zambia, and as part of this considers what the key institutional responses are and policy opportunities to strengthen climate resilience alongside resilience to poverty. It relies on mixed-methods data from 2010 to the present in Zambia: the Rural Agricultural Livelihoods Survey panel (2010–19), the Living Conditions Monitoring Survey (2022), and qualitative data collection by the Chronic Poverty Advisory Network (CPAN) and its partners across nine provinces in Zambia (2019–24). These data sets are iteratively analysed to identify key drivers of poverty reduction alongside climate resilience.
Study findings suggest that the expanding social development policy agenda in the country creates possibilities for upward pathways out of poverty, especially for the very challenging transition between extreme and moderate poverty, for better risk management, and for climate resilience, in a context where risks are likely to continue to grow and resilience will be hard won. As many commentators have mentioned (World Bank 2024; Martin 2024), the narrow, volatile and capital-intensive base of Zambia’s economic growth, concentrated in copper mining with weak links to the broader economy, needs stretching into a more inclusive pattern. According to our results, this can be nurtured through strong support for agricultural development and other natural resource-based conservation and development, diversification into the rural and urban non-farm economy, and the institutional and physical infrastructure to support such initiatives.
The rest of this article proceeds as follows. Section 2 outlines the data sources and methods adopted. Section 3 discusses some emerging insights from this data analysis. Section 4 presents the key findings organised around two questions – what role do government policies and interventions play in supporting upward mobility? And what are the main drivers of climate resilience? Section 5 concludes.
2.1 Data
Quantitative and qualitative data were analysed in this study. Regarding the quantitative data, the nationally representative Living Conditions Monitoring Survey (LCMS) 2022 (conducted by the Zambia Statistics Agency (ZamStats) to monitor the effects of changes in policies on the living conditions of the population) was central. The LCMS covers 8,520 households and provides insights on self-reported household shocks (including drought), household livelihoods, socio-demographic characteristics, and household poverty status according to the national poverty line. Table A2 in the Appendix provides summary statistics from the survey.
This was complemented by the Rural Agricultural Livelihoods Survey (RALS) panel data (2012, 2015, and 2019), conducted by the Indaba Agricultural Policy Research Institute (IAPRI) in collaboration with Zambia’s Ministry of Agriculture and ZamStats. RALS covers a balanced sample of 6,625 households across Zambian provinces focusing on those cultivating less than 20 ha of land to obtain a comprehensive picture of Zambia’s smallholder farming sector. A stratified two-stage sample design was used where agricultural households were stratified into categories based on area of crops, cattle types, and income, followed by systematic sampling to select 20 rural households across three strata in each Sample Enumeration Area. RALS provides information on agricultural practices, household characteristics, income, and economic outcomes for smallholder farmers. Table A3 in the Appendix provides summary statistics from the survey.
The RALS was merged at the household’s geocoordinates with drought and flood incidence data from Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS) satellite data, creating a radius for households within which climate disasters were present. This was done to provide, with high spatial resolution, data on actual climate-related disasters to better understand the extent to which households were able to remain resilient to poverty in the presence of climate-related disasters.
Qualitative data in turn comprised life history interviews (LHIs), focus group discussions (FGDs), and key informant interviews (KIIs). A total of 137 LHIs in 2023 and another 74 LHIs in 2025 were conducted in rural and urban areas across nine of the ten Zambian provinces (Figure 1), to identify in-depth insights of critical transitions in wellbeing and the pathways and sequences through which individuals navigated poverty and climate resilience. Districts within provinces were selected based on an assessment of poverty rates in Zambia from 2010 when district level estimates were present, alongside provincial trends over time, and rural poverty trajectories using the RALS to identify sites capturing a range of poverty dynamics (see Bond et al. 2021 for details).
LHIs were complemented by 18 gender-disaggregated FGDs (two per site), to capture gender and generational perspectives on wellbeing alongside meso-level drivers of poverty reduction and climate resilience. Alongside this, in each study site, FGDs were conducted with long-term residents to identify changes in local livelihoods, public services, policies, and environmental factors over time. Finally, KIIs were held with local and national-level policymakers and other experts to better understand institutional responses and the broader environmental and policy challenges and opportunities in Zambia.
Poverty in the quantitative data is derived as noted above by comparing income or expenditures to a poverty line. Poverty trajectories were constructed to identify chronically poor households (persistently under the poverty line or in wellbeing levels 1–3; see elaboration below and Table A1 in the Appendix), transient poor households (falling into poverty or not experiencing a sustained escape over the last five years qualitatively or in the last two survey waves), and households resilient to poverty (escaping and remaining out of poverty over the last five years qualitatively or in the last two survey waves, or being never poor). Resilience to poverty conceptually is defined in this analysis as the capacity to resist downward mobility into poverty as a result of a climate-related shock (either in the year preceding a survey or in the qualitative research during the period after a climate shock).
2.2 Methods
The quantitative analysis relied on a set of regressions to identify correlates of poverty status and its trajectories. Using the LCMS, logistic regressions were employed to account for the binary outcome equal to 1 if a household is poor and 0 otherwise, where:
Pr(poori =1) = Φ (β0 + β1 Hi + β2 Xi + εi ) (1)
where poor is equal to 1 if household i is poor and zero otherwise, β0 is a constant, H is a vector of household-specific controls, X is a vector of regional variables, and εi captures the error term. Key livelihood variables were interacted with self-reported disasters to understand the extent which livelihoods were able to offset drought or floods.
A similar process was carried out using the RALS data, where multinomial logistic regressions capturing the different poverty trajectory outcomes were employed. The richness of agricultural data in the RALS allowed us to additionally examine key climate-smart agriculture (CSA) variables and their role in moderating the relationship between drought and negative poverty trajectories. CSA variables included irrigation (household irrigated at least one field); conservation agriculture practices (minimum tillage: e.g. potholing, ridging); ripping, mulching for residue retention; adoption of drought-tolerant crops (e.g. cassava, sorghum, millet) or practising agroforestry; erosion/flash‑flooding control; and cultivation of improved varieties (especially for maize). Marginal effects were calculated and discussed for all results, and complemented by descriptive analysis of key variables, alongside various sensitivity analyses.
The qualitative data analysis relied on thematic coding and process tracing to identify pathways through which people experienced different poverty trajectories, including sustaining escapes from poverty, and in that process built climate resilience. As noted earlier, CPAN research suggests that climate resilience is both conceptually and empirically similar to making a sustained escape from poverty (Diwakar et al., this IDS Bulletin). In terms of climate resilience, this often refers to the ability of a socio‑ecological system to anticipate, reduce, or recover from the effects of a hazardous climate event. In terms of poverty resilience, in the qualitative research, households were categorised into six levels of wellbeing by community members (see Table A1). Wellbeing level 5 (WB5) is ‘resilient’, which translates into having the resources and strategies to avoid falling back into vulnerability (WB4) and poverty and food insecurity (WB1–3). WB1–4 groups are either very poor, poor, or vulnerable to poverty, and so not resilient: if they are resilient to falling back into poverty, which may be caused by climate-related shocks, then they are resilient to at least one very important consequence of climate change.
Moreover, while wealthier households might be better positioned to recover after a disaster, a wealthy household located in a floodplain is still very exposed to climate shocks such as flooding; nevertheless, the point is that these households can recover. Moreover, a poor household may be exposed and vulnerable to drought, but drought itself is not the only factor driving poverty; if that household has small landholdings, labour force, or infertile soil, the experience of drought would compound existing drivers and not be the only reason a household is poor. Indeed, there are other sources of downward mobility in Zambia, including ill health, limited landholdings, or dispossession of assets, economic shocks, and adverse gender norms (Ndubani et al. 2021). Achieving resilience to these shocks and stressors, including climate change, is accordingly a big challenge in a situation where the majority of Zambians are in WB1–4, and resilience to both poverty and climate change is associated with WB5. The extent to which poor and vulnerable people can become more resilient is a critical policy question.
The data was analysed iteratively, where for example the importance of livestock as a correlate in the LCMS supported its deeper probing in the qualitative data, which in turn led to a more disaggregated analysis of livestock by type in the quantitative data. The various data sets enabled triangulation for a richer, more rigorous understanding of drivers of poverty reduction alongside climate resilience.
An optimistic finding from both the quantitative and qualitative research is that there is an increasing number of policy and programme interventions that support poverty reduction in Zambia and that may be capable of building greater resilience, including to climate change. The expansion of social cash transfer (SCT) coverage is a key example, supported by increased social sector spending (Figure 2); we interviewed SCT recipients across study sites. However, the Living Conditions Monitoring Survey (LCMS) 2022, which was carried out after the SCT expanded to 974,000 households in 2021, recorded that coverage among the poor and poorest remained a major challenge (Figure 3). Nevertheless, the regular cash transfer at the time of the survey, and, in 2024, the emergency expansion, has resulted in significant stabilisation of many livelihoods, especially for the poorest recipients:
Social cash transfers provided some relief, helping her buy school necessities for her grandchildren, but debts often consumed these funds.
(LHI5, updated, Tilolele, Chipata, 2024)
Social protection has also been employed in response to climate‑related disasters, such as through emergency cash transfers and Cash for Work (Ministry of Local Government and Rural Development 2024), though with less effectiveness. Qualitative data points to implementation being challenged by delays in payments and inadequate government budgets, such that participation had to be limited, and participants could not rely on getting work when they needed it – a cardinal principle of successful Cash for Work programmes.
Agricultural policy is also making a slow transition from allocating fertiliser subsidies and the almost exclusive focus on maize to a more comprehensive and climate-smart approach. Small-scale farmers now have greater choice over the subsidised inputs they can buy through an e-voucher scheme, which was being rolled out nationwide in the 2025/6 season (World Bank 2026). Moreover, the Department of Livestock has a livestock distribution programme, as do a number of non-governmental organisations, but these are all small-scale programmes as yet.
Though there are positive developments, this has yet to translate into substantive poverty reduction. Both the overall incidence of poverty and the number of people in poverty increased between 2015 and 2022, especially in urban areas (ZamStats 2022). Even so, the provincial picture was varied (Figure 4).
Many households perceived that they were worse off in 2022 compared with 12 months previously, which is not surprising given the context of the Covid-19 pandemic. This was marked among extremely poor households, where over a third (34 per cent) felt that they were worse off, compared to a still high 28 per cent amongst non-poor households and 28 per cent amongst moderately poor households (analysis of LCMS 2022 data).
The year 2021/2 was also a peak of food and energy price inflation, which was one reason for being worse off, especially but not only in cities (Figure 5). While reasons for the downward mobility of non-poor households were most often economic (such as falling incomes and rising prices), agriculture- and climate-related reasons accounted for the downward mobility of extremely poor rural households. Nearly half of all households reported that a lack of agricultural inputs was the principal reason for their self‑perceived poverty, despite the extensive fertiliser subsidy scheme at the time (Figure 5).
The rural panel and satellite weather data collected during the 2010s (2012, 2015, and 2019, with satellite weather data from the season before in each case) indicates that drought and flood risks were widespread (Figure 6) – four-fifths of the population experienced them, with floods as significant as droughts, if not more so, especially in terms of their incidence and unmitigated effects. Qualitative data supports this. For example, in Mpulungu (Lake Tanganyika), flooding in January–March 2024 was among the worst experienced in decades, submerging homes and farmland. Some community members believed that the lake was ‘reclaiming its land’ as they observed that water had moved further inland by an estimated 20–50m.
At the same time, climate risks are only one of several kinds of risk that the population of Zambia has faced in recent years. LHIs across provinces evidenced economic shocks and inflation driving the long-term erosion of purchasing power, leaving chronically poor households struggling to afford basic needs and other households pushed into poverty due to sudden price hikes on essential goods. The business and market vulnerability characterising broader economic instability has exposed businesses to regular failure pushing individuals deeper into poverty or trapping them in a cycle of low-income employment or small-scale enterprises with minimal growth.
Despite its significant and widely welcomed expansion in coverage, still-inadequate social assistance and barriers to other government support and market access means that many poorer households remain reliant on their own efforts to diversify and manage risk, as well as often unstable informal networks for survival. The limitations of coverage, the inadequacy of payment levels, and the payment delays in social assistance have made it difficult for many households to escape poverty or demonstrate climate resilience. From the qualitative data, moreover, it appeared that the emergency cash transfers were also not as well managed as the SCTs, with delays in payment, as well as omissions of some vulnerable people:
The ECT [emergency cash transfer] was supposed to help people for six months, but the payments have been inconsistent. Some households have gone months without receiving any support.
(KII, Community Welfare Assistance Committee, Nsingo, Chipata)
We have seen that families use the ECT to buy food, but it only lasts a short time because prices keep rising.
(KII, Department of Social Work, Shangombo)
This section presents study findings, organised around two themes. First, it summarises key enabling factors supporting the upward mobility of households across the distribution, with a focus on the supportive role of government policies and programmes. This is considered as a complement to supporting household resilience to climate-related shocks and stressors. It then considers key direct drivers of climate resilience based on mixed-methods research results, namely climate-smart agriculture and economic diversification.
4.1 Poverty escapes and synergistic drivers of climate resilience
Despite the strong downward pressures on livelihoods of the past decade, some poor and vulnerable people in Zambia have managed a degree of upward mobility, as summarised in Table 1. As noted earlier, what is noticeable here is the important but changing role of government policies and interventions as people progress through wellbeing levels. Moving from extreme to moderate poverty is very challenging and entails multiple layered or sequenced interventions, combining regular transfers with capital boosts. Moving from moderate poverty to being resilient to poverty is likely to be somewhat more self-propelled; there are virtuous cycles too: greater regularity of income enables stable participation in savings groups, which can provide small investment funds.
As suggested in Table 1, the WB3 to WB4 transitions identified in the qualitative research across study sites were not just important drivers of crossing the poverty threshold but were also important implicitly in developing climate resilience capacities. For example, crop diversification and improving soil health could help households spread risk and increase agricultural productivity, such that they are better able to withstand the effects of climate‑related disasters. Similarly, village savings and loan associations can offer households a critical means of liquidity to draw on to help smooth consumption during these shocks and stressors.
4.2 Climate-smart agriculture and economic diversification
The quantitative and qualitative research results discussed below point to a powerful narrative on rural climate resilience around agricultural development, especially CSA and livelihood diversification as twin potential sources of climate resilience that also support poverty escapes. In other words, one source typically requires the other in order to strengthen its potential, not just to sustain poverty escapes but concurrently to promote climate resilience at lower levels of wellbeing, even for households in extreme poverty.
4.2.1 Resilience to poverty and climate change through climate‑smart agriculture
Crop agriculture acts as a cushion for the poorest households even in drought years, although it does not provide an escape from poverty during drought periods. This may be related to the continued (policy) preference for maize in much of Zambia, which is highly vulnerable to drought. Qualitative research indicates that during the 2024 drought other more drought-resistant crops also suffered significantly reduced yields or crop failure; even micro‑scale vegetable farming, which relies on bucket irrigation, was constrained:
Gardening involves crops like tomatoes, onions, and leafy vegetables. It provides income when managed well, but the biggest challenge is water. When wells dry up, the gardens suffer, and profits fall… Farming is the main livelihood for women. We grow maize, sunflower, groundnuts, and vegetables. During the rainy season, we plant for home consumption and sell the surplus. After the harvest, many women continue with gardening.
(Women’s FGD, Chipata, 2024)
Nevertheless, a degree of climate resilience to the 2024 drought was achieved by some, through early planting, use of SCT for investment in irrigation, and a clever marketing strategy (Box 1). This is resilience at a lower level than that of a sustained escape from poverty, indicating that degrees of climate resilience can be achieved if specific resilience capacities can be developed.
The RALS panel surveys (2012, 2015, and 2019) help refine this picture: descriptively, across all poverty trajectories, the adoption of at least one CSA practice was consistently high, exceeding 80 per cent for most groups by 2019 (Table A4 in the Appendix). Households exposed to drought risks generally exhibited higher adoption rates. Residue retention was among the most frequently adopted CSA practices in the sample overall, with significant growth across all groups between 2012 and 2019. Households exposed to drought risks demonstrated particularly high adoption rates – nine in ten households that had escaped poverty temporarily by 2019 were adopters (Table A4).
Several of the individual CSA variables were associated with lower chronic poverty and higher poverty resilience. Regression results indicate that improved maize varieties were associated with a lower probability of chronic poverty, but a higher probability of both transient poverty and resilience to poverty for households facing both drought and flood risks (Figure 7). This could reflect the role of improved maize varieties in raising welfare and recovery capacity, though increasing exposure to loss during times of climate-related disasters.
Complementary practices also had poverty resilience effects. Ridging in flood-prone areas was further associated with a higher probability of household resilience to poverty and lower probability of poverty transience (Figure 7, top). Descriptively, ridging was a widely adopted practice, especially among chronically poor and impoverished households. Adoption rates for flood-exposed households peaked in 2015, reflecting its accessibility and relevance to water management, but had declined by 2019 (Table A4).
Irrigation instead had mixed results, associated with a lower probability of poverty transience but higher probability of chronic poverty (Figure 7, bottom). The former could reflect the role of irrigation in smoothing consumption during climaterelated disasters such as drought, yet its high upfront investment costs could be difficult to manage amongst chronically poor households.
Practices such as mulching, digging planting basins, and ripping showed low adoption rates compared with residue retention (i.e. crop residues left in the field) and ridging (Table A4). These practices were more prevalent among drought-exposed households, particularly among those who only escaped poverty temporarily and those who became poor, but adoption remained below 10 per cent for most groups. Of these practices, mulching was associated with a higher probability of chronic poverty amidst drought, which could reflect its role as a low-cost practice accessible to the chronic poor though often without complementary, more effective technologies that could drive higher-return investments.
So, while CSA was definitely beneficial in the 2010s, it did not represent a complete answer to building resilience to poverty and/or climate change. This could partly be because of the limited adoption of CSA: conservation agriculture was typically only adopted on a small part of a smallholder farm due to its high labour or capital demands, or it was adopted and then ‘disadopted’ later (Kayula et al. 2022). Constraints include labour shortage, and limited irrigation and mechanisation. The government has now at least allocated additional budgetary resources to irrigation and has a mechanisation strategy which acknowledges the issue (Republic of Zambia 2024). Thorough and rapid implementation of these will be important.
4.2.2 Agricultural and off-farm diversification
Economic diversification is important in two of the three wellbeing transitions indicated in Table 1; and investment in irrigation, livestock, and mechanisation underpin the transition out of poverty and to climate resilience. However, agriculture alone (i.e. without CSA practices) is less often the solution, especially in contexts of climate-related disasters. Indeed, LCMS regression results indicate that employment in agriculture of the household head amidst drought is associated with a statistically significant higher probability of poverty (Figure 8), indicating the added challenges that emerge to agricultural livelihoods in the context of drought.
At the same time, households exposed to climate risks that have larger landholdings, livestock holdings, and asset values are less likely to be chronically poor and more likely to experience transient poverty and a degree of resilience to poverty (Figure 8). Effect sizes where significant are typically larger in the presence of drought compared to floods, suggesting that these investments work better in drought than flood situations.
Again, the somewhat unexpected relationship of some of these variables with transient poverty – where, for example, a higher asset value, more livestock, and larger landholdings are associated with a higher probability of transient poverty amidst either drought and/or flood risks – could reflect the role of these different forms of diversification as a necessary but not sufficient condition for preventing downward mobility during crises. Even so, the strong positive relationship of these variables with resilience to poverty (i.e. sustained poverty escapes or being never poor) and negative relationships with chronic poverty points to the importance of these investments in land, livestock, and asset development in supporting upward mobility, even if they may be less effective in guarding against downward mobility during floods and drought.
According to the LCMS data, the largest effect size in the extreme poverty specification is observed for the livestock variable, where ownership of livestock amongst households reporting drought was associated with a 10.6 percentage point lower probability of poverty in 2022 (Figure 9). This was an effect size that was, moreover, much larger and statistically significant compared to households not experiencing drought. Disaggregating by type of livestock, the presence of cattle in particular was associated with a statistically significant lower probability (by 10.1 percentage points) of poverty for households reporting drought. Instead, the presence of goats, pigs, and sheep was not a significant correlate though the expected directionality remained. That owning small stock does not appear to help smooth consumption in drought years is a finding which deserves further research.
Mixed-methods research nuances this picture. The RALS panel data showed how owning livestock had challenges amidst drought and flood episodes, as they were at risk of disease as well as shortages of pasture and fodder. The qualitative data also pointed to risks to livestock during droughts. For example, when drought struck, LHI4 had to stop farming. By February 2024, many of his seedlings failed to grow due to water shortage. By March, the river that supplied his farm had dried up, forcing him to halt all gardening activities. To cope with the crisis, he sold three oxcart-loads of cabbages for 2,100 kwacha (approx. US$90) before completely shutting down operations. With his farm no longer producing income, he relied on rental earnings and livestock sales to sustain his household. However, the lack of grazing land and water shortages weakened his cattle, reducing both their market value and milk production. Despite these setbacks, he remained determined to recover. By October 2024, he applied for a Constituency Development Fund loan to install a mechanised irrigation system, hoping this would improve his future harvests (LHI4, Shangombo).
Finally, as Figure 9 also conveys, ownership of non-farm enterprises is associated with climate resilience in drought situations. However, as the qualitative results indicate, such opportunities to diversify into non-farm businesses were mostly only available to sustained escapers and non-poor households, which had the capital, knowledge, and networks to sustain trading, services, and property rental enterprises. It is challenging for micro-enterprises to comply with business regulations and satisfy application processes sufficiently to get access to government loans; for example, through the Constituency Development Fund (CDF). Poorer households were restricted to crop diversification and other agriculture-related diversification, which was less effective in providing climate resilience. Many of the sources of non-farm income that poorer people rely on, such as petty trade, also dried up during drought periods. There is thus a major challenge for poor people to diversify away from farming. It is also a major current policy challenge in Zambia, where the Ministry of Small and Medium Enterprise Development has not yet supported many outcomes visible in the qualitative research carried out since the Ministry was formed.
The twin pillars of household climate resilience are diversification into non-farm enterprises and investment in climate-smart agriculture. These are also pathways to escape from poverty, and stay out of poverty, so these two objectives are intimately intertwined in the modern, crisis-prone context. Once people move from extreme poverty to moderate poverty, they can develop specific resilience capacities, which can help sustain them over a significant crisis without experiencing downward mobility or further impoverishment. Such capacities include diversifying livelihoods and making CSA adaptations (including seeds, micro-irrigation, mechanisation, livestock, and soil and water conservation). Such adaptations are challenging for people living in extreme poverty, however, and they will need significant external support.
In these efforts, special attention will need to be given to the large numbers of chronically poor and impoverished farm households, and women-headed households, that depend
In these efforts, special attention will need to be given to the large numbers of chronically poor and impoverished farm households, and women-headed households, that depend on natural resources. These households are more vulnerable to health- and climate-related shocks than moderately poor and non-poor households, as they are more exposed to price and income shocks. These households should be a focus for the development of specific poverty reduction and climate resilience capacities centred on small-scale irrigation, appropriate technology mechanisation, and soil and water conservation. Their small surpluses need aggregation: this is where ‘compassionate cooperatives’ can help, alongside competitive markets more generally.
Disasters will continue to happen, possibly at increasing frequency and intensity: droughts and flooding are almost predictable in terms of their frequency and severity. The Zambian government needs to be in a position to respond strongly when disasters strike, and also to prepare well for them and help people recover. Investment in the disaster risk management system is necessary, underpinned by sound macroeconomic management. Institutions supporting people’s resilience capacities (ministries and agencies for agriculture, livestock and fisheries, the green economy, non-farm enterprises, education, health, and community development) also need to examine their own resilience to climate change and disaster risks, and identify where these need to be strengthened. The Ministry of Finance and Planning should pay close attention to the resilience of these institutions, exploring the opportunities presented by the availability of climate finance to support adaptation, climate resilience, and poverty reduction.
1 This issue of the IDS Bulletin was supported by the UK Foreign, Commonwealth & Development Office. The opinions expressed are the authors’ own and do not reflect the views of the funder.
2 In memory of Dr Joseph Simbaya (deceased 7 October 2025), who contributed centrally to the writing of this article but sadly passed on before its publication.
3 Andrew Shepherd, Director, Chronic Poverty Advisory Network (CPAN), UK and Honorary Associate, Institute of Development Studies, UK.
4 Vidya Diwakar, Deputy Director, Chronic Poverty Advisory Network (CPAN), UK and Senior Research Fellow, Institute of Development Studies, UK.
5 Mary Lubungu, Research Fellow, Indaba Agricultural Policy Research Institute (IAPRI), Zambia.
6 Richard Bwalya, Research Fellow, University of Zambia.
7 Antony Chapoto, Executive Director, Africa Network of Agricultural Policy Research Institutes (ANAPRI), Zambia.
8 Lucia da Corta, Research Associate, Chronic Poverty Advisory Network (CPAN), UK.
9 Marta Eichsteller, Assistant Professor, University College Dublin.
10 Marja Hinfelaar, Director of Research and Programmes, Southern African Institute for Policy and Research (SAIPAR), Zambia.
11 Arthur Monze Moonga, Director, Institute of Research and Development (IRED), Zambia and Research Associate, University of Zambia.
12 Brian Mulenga, Executive Director, Indaba Agricultural Policy Research Institute (IAPRI), Zambia.
13 Kate Pruce, Research Fellow, Institute of Development Studies, UK.
14 The late Dr Joseph Simbaya, Research Fellow and former Director of the Institute of Economic and Social Research (INESOR), Zambia.
ACAPS (2025) Zambia: Update on the Impact of Drought, Thematic Report, 16 January (accessed 21 April 2026)
Bahadur, A. et al. (2015) The 3As: Tracking Resilience Across BRACED, BRACED Working Paper, London: Overseas Development Institute (ODI)
Berretta, M. et al. (2023) Strengthening Resilience Against Shocks, Stressors, and Recurring Crisis in Low- and Middle- Income Countries: An Evidence Gap Map, Evidence Gap Map Report 23, New Delhi: International Initiative for Impact Evaluation (3ie)
Bond, V. et al. (2021) A Qualitative Understanding of Poverty Dynamics in Zambia, Chronic Poverty Advisory Network (CPAN) Working Paper, London: Overseas Development Institute (ODI) (accessed 23 April 2026)
da Corta, L. and Bwalya, R. (2025) Rural Pathways Out of Poverty, Chronic Poverty Advisory Network (CPAN) Evidence Paper, Brighton: Institute of Development Studies, DOI: 10.19088/CPAN.2025.002 (accessed 21 April 2026)
Diwakar, V. and Shepherd, A. (2022) ‘Sustaining Escapes from Poverty’, World Development 151: 105611, DOI: 10.1016/j.worlddev.2021.105611
GIZ (2021) Macroeconomic Models for Climate Resilience: An Economic Tool for Adaptation and Development Planning, Eschborn: Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ)
Hinfelaar, M. (2025) ‘Political Economy Analysis of Pro-Poor Policies in Zambia’, Chronic Poverty Advisory Network, 8 October (accessed 3 June 2026)
IAPRI (2019) Rural Agricultural Livelihood Survey, Indaba Agricultural Policy Research Institute
IAPRI (2015) Rural Agricultural Livelihood Survey, Indaba Agricultural Policy Research Institute
IAPRI (2012) Rural Agricultural Livelihood Survey, Indaba Agricultural Policy Research Institute
Kayula, F.; Maambo, O.; Mwasi, D. and Bwalya, I. (2022) Baseline Report: CSA Practices and Adoption in Zambia: A Baseline Evaluation for Climate Smart Agriculture Practices and Adoption in Zambia, Climate Smart Agriculture Alliance Zambia, Ministry of Agriculture and the Food and Agriculture Organization of the United Nations, 28 February (accessed 21 April 2026)
Martin, M.V. (2024) Economic Growth in Zambia: Key Issues and Trajectory, International Monetary Fund (accessed 21 April 2026)
Ministry of Local Government and Rural Development (2024) Cash For Work (CFW) Programme Operational Guidelines, Government of Zambia (accessed 21 April 2026)
Ndubani, P.; Chiti, B.; Billima, T.; Hineflaar, M. and Bond, V. (2021) Literature Review: Poverty Dynamics and Drivers in Zambia, 1964–2019, Chronic Poverty Advisory Network, 13 December (accessed 21 April 2026)
Republic of Zambia (2024) National Agricultural Mechanisation Strategy, 2024–2028, Lusaka: Ministry of Agriculture
UNDP (2025) Global Multidimensional Poverty Index 2025. Overlapping Hardships: Poverty and Climate Hazards, New York NY: United Nations Development Programme
World Bank (2026) A Smarter Way to Support Farmers in Zambia, 23 February (accessed 21 April 2026)
World Bank (2024) Zambia Country Economic Memorandum: Unlocking Productivity and Economic Transformation for Better Jobs, Washington DC: World Bank (accessed 21 April 2026)
World Bank (2019) Zambia – Transforming Landscapes for Resilience and Development Project: Project Appraisal Document, Report PAD3074, Washington DC: World Bank (accessed 21 April 2026)
ZamStats (2022) 2022 Living Conditions Monitoring Survey Report, Lusaka: Zambia Statistics Agency (accessed 21 April 2026)
ZamStats (2016) 2015 Living Conditions Monitoring Survey Report, Lusaka: Zambia Statistics Agency (accessed 3 June 2026)
ZIPAR and UN Zambia (2022) Analysis of the 2023 National Budget, Lusaka: Zambia Institute for Policy Analysis and Research (ZIPAR) and United Nations (UN) Zambia (accessed 21 April 2026)
© 2026 The Authors. IDS Bulletin © Institute of Development Studies | DOI: 10.19088/1968-2026.170
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International licence (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are credited and any modifications or adaptations are indicated.
The IDS Bulletin is published by Institute of Development Studies, Library Road, Brighton, BN1 9RE, UK. This article is part of IDS Bulletin Vol. 57 No. 2 July 2026 ‘Climate Resilience and Poverty Reduction’.