Qualitative meta-analysis of economic models of climate change: informing policy development strategies

Global climate change refers to a “dynamic, multidimensional system of changes in environmental conditions that will likely influence human behaviour” (Evans, 2019, p. 2), which is an important current issue (Han and Ahn, 2020; Mikhaylov et al., 2020). Its impacts are pervasive on all scales – global, regional, national and local levels – and necessitate multi-level intervention [Intergovernmental Panel on Climate Change (IPCC), 2022; Ministry of Mahaweli Development and Environment, 2016]. Climate change creates cross-repercussions in many sectors, and agriculture is particularly vulnerable due to its reliance on stable weather, exposure to extreme events and limited adaptive capacities (Feulner, 2017; Ngcamu, 2023; Yazdanpanah et al., 2024). Climatic impacts on agriculture affect global trade, food security and accelerate migration (Bibi and Rahman, 2023; Carleton and Hsiang, 2016; Ziervogel et al., 2014), undermining environmental (Toukabri and Youssef, 2023), economic (Carleton and Hsiang, 2016) and social (Gawith et al., 2020) sustainability, adversely affecting livelihoods and emerging as a key cause of poverty (IPCC, 2022; Moellendorf, 2024). Although primary impacts include the destruction of harvests or delayed planting times, the “slow burn” of secondary impacts also cause further devastation for communities. Ratnasiri et al. (2019), for instance, state that about 75% of the total economic loss in agriculture is during the immediate aftermath of floods and droughts that, over time, impacts the livelihoods of heavily dependent communities. The overall impact of climate change on agricultural communities can be multifaceted affecting income, other personal economies and social considerations (Roson, 2014; Sánchez, 2018). In accordance with the United Nations (UN) Sustainable Development Goals (SDGs), Toukabri and Youssef (2023) investigated the problems associated with climate change and the socioeconomic impacts on communities’ livelihoods, addressing SDG1 (no poverty), SDG2 (zero hunger) and SDG11 (sustainable cities and communities) as priorities.

Economic models of climate change (EMCC) within agriculture extend beyond mere economic assessment to also evaluate how climate shapes livelihoods (Roson, 2014). These are considered in association with broader socioeconomic issues and impacts on a country’s food security goals (SDG2) (Teklewold et al., 2017). Pindyck (2017) notes that these models follow a structured methodology, linking agricultural outputs and outcomes with climatic data and initial assumptions while simplifying complex economic systems to forecast future conditions and support decision-making (Ouliaris, 2011). Additionally, Carleton and Hsiang (2016) emphasise that such models incorporate shifts in welfare, economic conditions, demographic responses (e.g. migration) and social interactions, which are challenging to predict due to their subjective nature. Ultimately, these outputs, including yields, profits and productivity, can be expressed as the costs and benefits of climate change that guide strategies on coping with its economic impacts (Ngcamu, 2023; Tol, 2018).

Studies on climate change can be distinguished from those predominantly based on climate science and those that transcend discipline boundaries to examine direct and indirect impacts. Amongst the latter, studies examining climatic change and economic models have tended to develop as a separate body of knowledge to climate science (Sánchez, 2018). Each economic model typically uses a set of economic and climatic conditions and output determinants to suit certain specific contexts (Hashida and Lewis, 2022; Roson, 2014). Despite numerous studies on EMCC, the multifaceted impacts of climate change, particularly those related to agriculture, seem disorganised (Teklewold et al., 2017). Thus, this literature review, and new knowledge emerging from the meta-analysis, aims to fill this gap by consolidating existing knowledge on EMCC to identify parameters. It also considers the effective use of models that prioritise and organise socioeconomic factors pertaining to agricultural communities’ livelihoods. Accordingly, four sub-questions (SRQs) have been formulated to address this aim as follows:

• SRQ1: What economic models are available to explore on the socioeconomic impacts of climate change?

• SRQ2: What EMCC characteristics are used to manage the impacts of climate change?

• SRQ3: Which EMCC outcomes address the socioeconomic impacts?

• SRQ4: What factors should be considered when using EMCC as a tool to identify the socioeconomic impacts of climate change?

It is expected that this study will provide a platform to allow more expansive developments in EMCC and to further explore the socioeconomic conditions of communities whose agriculture-based livelihoods are affected by climate change.

Systematic literature reviews (SLRs) leading to meta-analyses have emerged as rigorous and structured research methods and have gained prominence across various scientific disciplines (Ngcamu, 2023). SLRs ensure a systematic selection of literature sources and the minimisation of biases from random searches. Qualitative meta-analyses further synthesize and integrate findings from multiple studies to identify patterns, generate theoretical insights and enhance the understanding of a given phenomenon while maintaining methodological integrity (Levitt, 2018; Timulak, 2009). Thus, this study uses an SLR followed by a qualitative meta-analysis. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework has been used to ensure an internationally recognized, evidence-based, transparent review process (Geekiyanage et al., 2020; Monash University, 2020; Ngcamu, 2023). Furthermore, the qualitative meta-analysis was structured using the Levitt (2018) approach, a widely adopted method in social science research (see Figure 1 for detail, which also illustrates the methodology used in this study).

As shown in Phase-01 (P1) in Figure 1, to obtain a versatile set of structured results, three databases of peer-reviewed literature have been used, which are highly ranked in the social science field. Further, PICO has been incorporated as the search strategy as it enables more sensitive results (Methley et al., 2014) and is commonly used to extract qualitative findings. “Title-abstract-keyword” search criteria have been used to obtain a comprehensive set of results; however, this study adopts a “title” only search for the control with (climat*) representing “climate”, a term prominently used in the abstracts of many articles that brought numerous misleading results (e.g. indoor climate, comfort levels, etc.). Additionally, proximity operators (W/0 or NEAR/0) and wildcards (*) were used to increase the consistency and accuracy of the results.

Following PRISMA 2020 guidelines, articles were screened, as shown in Phase-02 (P2) of Figure 1. To increase the rigour of the study, six articles were added manually, through a citation search, and 34 articles finally included as suitable, comprising 28 journals, four journal review papers and two book chapters. These results were limited to studies that consider variations of temperature, rainfall and wind as climate events as these are primary drivers of climatic hazards, such as droughts, floods, landslides and storms. Focusing on these variables ensures a comprehensive assessment of how fundamental climatic shifts influence agricultural systems, economic stability and community resilience (ability to build back better after a climatic disaster event (Geekiyanage et al., 2020), aligning with the study’s aim of evaluating the socioeconomic impacts of climate change on agricultural communities. The next phase (P3) describes the data analysis process of the qualitative meta-analysis. The first four steps of Levitt’s (2018) approach was achieved in P1 and P2, and represented as an equal sign (=) in Figure 1. The findings and discussion have been described in Section 3.

Figure 2 illustrates the annual distribution of the 34 articles accepted for this study.

Figure 2 indicates a notable peak in studies on EMCC in agriculture during 2017 and 2018. The World Meteorological Organization (WMO) (2019) reported severe droughts in 2017, which prevailed across many parts of the world, and potentially prompted increased research activity. In 2018, the accelerated pace of climatic events, aligned with emerging IPCC reports, further stimulated scholarly output. A modest increase in 2020 could reflect contributions to the 2021 IPCC report, while the global pandemic (COVID) likely resulted in fewer publications in 2021 due to the diminished need for investigations consequent to less air traffic.

The initial SLR analysis identified different forms (or types) of EMCC and was supported by the cited literature. The study conducted an inductive categorisation of EMCC in agriculture based on the primary purpose (e.g. land based, integrated assessment, etc.), as shown in Figure 3. While previous literature identifies various EMCC in agriculture, few studies classify them by purpose, to offer a comprehensive overview that supports socioeconomic decision-making based on parameters appropriate to the purpose.

The main characteristic of Figure 3 is the emergence of EMCC in determining its overall impacts on communities’ livelihoods. The detailed review and analysis of the models based on the five categories identify their common characteristics.

Discrete choice economic model (DCEM), identified by Hashida and Lewis (2022), is a land management model that evaluates how climate change affects specific commercially grown tree species (e.g. Douglas-fir, Hemlock) across various land plots. The model elucidates the impacts on land values and, by extension, on the livelihood conditions of commercial landowners, thereby providing essential data which informs replanting strategies. By contrast, Ricardian model (RM) explores the relationship between the extent of land management and its potential for agricultural adaptation to climate change (Falco et al., 2018). Additionally, RM examines the relationship between land value or net revenue from land (based on the regressive nature of land values) and agro-climatic factors (Mendelsohn et al., 2004), with a key strength being its incorporation of private adaptation measures, such as adjustments to crop mix, planting and harvesting schedules and other agronomic practices (Deressa, 2007). Additionally, hedonic model (HM) derives its meaning from the term “hedonic” and denotes implicit (observed) prices where both the consumer and seller are kept at equilibrium (Rosen, 1974). It has been used by Mu et al. (2017) to determine how and whether variations influence net agricultural production, and subsequently affect land values. Some of the land-based models consider factors such as population density and methods of irrigation, which add perspective to land-based economic models (Mu et al., 2017). The incorporation of irrigation data provides further opportunities to build adaptation within the equation, and land-use share model (LUSM) falls in this category. The model assumes that land usage for agricultural purposes is decided on the expected market returns. Agro-economic model (AEM) is also considered within this category as it analyses the impacts of different climatic conditions on the performance of agricultural lands over time (Antle and Stockle, 2017). Agent-based rural land-use New Zealand model (ARLUNZ) is agent-based and developed for New Zealand; it provides insights into farmers’ adaptations to reduce vulnerability to climate change using farmer agents. Accordingly, farmers have the agency to determine the bio-physical features of land (biological productive potential, current land-use capacity), changes in economy and climate and the social and economic impacts on them from their adaptations. Multinomial endogenous switching regression model (MESRM) analyses how changes such as agricultural water management, fertilisers and modern seeds improve soil characteristics help to increase the performance of agricultural lands (Teklewold et al., 2017). Ethiopia’s economy-wide multi-market model (EEMM) is another country-based model like ARLUNZ which predicts climatic conditions in a particular area; its outputs suggest agricultural actions for farmers and economic decisions for the country (Zhang et al., 2020).

Whilst land-based models predominantly accommodate the local conditions of land value, integrated assessment models (IAMs) are widely used in a global context to determine the impacts of climate change on agricultural livelihoods. Revesz et al. (2014) used IAMs to determine the impacts of climatic conditions on a country/regional level based on future (climatic) scenarios. Rising et al. (2022) potentially proved that the world is at greater threat from future climate change than its already visible impacts. Khabbazan (2022) incorporates a risk assessment to identify the value of “climatic information” by integrating several factors to prepare for climate change. Khabbazan (2022) uses probabilistic integrated model of climate and economy (PRICE), which examines the magnitude of climate sensitivity, to analyse the cost of welfare and risks on agriculture and integrate climate and economic factors in an integrated assessment supporting decision-making. PRICE is therefore beneficial in decision-making processes by managing the welfare cost of people at a global level. In summary, IAMs provide a comprehensive framework to assess and mitigate the impacts of climate change on agricultural livelihoods and welfare costs on a global scale.

Crop models are “mathematical models which describe the growth and development of a crop interacting with soil” (Wallach et al., 2006, p. 3). They offer efficient support to research and development activities by aiming to reach maximal production potential through forecasting crop output while accounting for growth and development-related aspects like climate change (Divya et al., 2021). Global gridded crop model (GGCM) is a crop model that combines the yield (different types of crops), which is calculated by dividing the land into small grids with changing soil conditions (due to climate impacts). Bio-economic farm model (BEFM) is also used as a modelling tool to assess the impact of climate change on agriculture (Schuler et al., 2020). It uses household survey data, crop experiment data and long-term price and yield data to identify changes (and rationale thereof) in crop yields (Bobojonov and Aw-Hassan, 2014). Based on studies conducted by Islam et al. (2016) and Palazzo et al. (2017), international model for policy-analysis of agricultural commodities and trade (IMPACT) is a similar type of crop model, but determines the yields based on changes in climatic conditions and social trends to derive conclusions on crop productivity (technology use, location of field, population and income affect the yields). Probabilistic decision model (PDM) is a decision-making model, where decisions on crop adjustments are made based on the climatic influences of factors such as market changes and resource consumption (Choi et al., 2015). Global trade analysis project model (GTAP) is used to determine crop yield changes due to climate change and how this enables an assessment of impacts to welfare as a whole (Moore et al., 2017; Wang et al., 2021). Factors such as the nature of the land, crop productivity, intermediate use of crops and product prices can be varied to gain an understanding of adjusted yields/new crops due to climatic impacts. Overall, the above crop models attribute economic and welfare impacts (levels of poverty) to changes to crop production quantities and product prices.

ECLAC climate impact assessment model (ECLAC-CIAM) by Roson (2014) highlights the negative economic consequences of climate change in the Caribbean. They analyse how economic parameters like labour productivity, job demand and physical assets are affected by climate change and support financial decision-making. Moreover, both dynamic computable general equilibrium model (DCGE) and computable general equilibrium model (CGEM) are based on the global-, national- and/or regional-level economic impacts of climate change (Breisinger et al., 2013) and develop future climatic scenarios and assumptions to determine the socioeconomic and climate impacts on labour markets (Matsumoto, 2019). Both models, therefore, predominantly link future climate scenarios with the economic loss and welfare of a particular region. Overall, economic simulation models help to assess the potential impacts of climate change on economic systems, which helps in planning the economy of agricultural communities.

The EMCC consists of factors that directly inform policymaking. For instance, rural socioeconomic model (RUSEM), which considers indicators such as public service availability, credit access and labour availability, supports rural climate change policies (Navarro and Tapiador, 2019). In comparison, global biosphere management economic model (GLOBIOM) feeds agricultural, forestry and bioenergy policies based on agricultural productivity, land use, production prices and rates of emission (Van Meijl et al., 2018). Moreover Falco et al. (2018) apply time-series or panel model (TSPM), which uses crop yields and weather information over the time by following the concept “time series” to investigate the relationship between climate change and migration. It thereby underpins policies that support rural development and sustainable agriculture, to address the issue of migration.

According to Adego and Woldie (2022), probit models (PM) determine the relationships – or correlation – between various climate adaptation strategies. The word “probit” derives from “probability” and shows the probability of adaptation practices impacting each other. It uses different climatic conditions and economic factors as parameters to focus on adaptation strategies that can inform policy design. PM have two types: multivariate and ordered probit models. In multivariate probit models, a variety of adaptation strategies can be identified and ranked. Hence, due to differences in probability, the outcomes provided in the models differ, even with the same set of data. Moreover, key factors emerge because of climatic impacts on communities, which are difficult to capture mathematically through models, although they are requirements to be considered within EMCC (Stern, 2016). For example, evidence of local-level adaptation by communities and migration activities can be included (Adego and Woldie, 2022; Yalew et al., 2018). The status of climatic conditions on agricultural lands is assessed using TAMIAT; thus, net production in farms, per capita income and poverty levels are analysed to determine household welfare amidst climate change (Mulwa et al., 2016). Collectively, these models offer critical insights for policy development by integrating both quantitative assessments and qualitative considerations of climate adaptation.

The EMCC identified in Table 1 were further analysed within socioeconomic parameters using the SLR data. These parameters and their groups are presented in Table 1.

We have ranked the parameters based on their frequency of occurrence (citations) as per the studies of Gasparri et al. (2023) and Geekiyanage et al. (2020), which were conducted in similar social science disciplines using the PRISMA (meta-analysis). Accordingly, this provides a quantitative analysis of parameters, which helps to identify the frequency of use within the EMCC.

The highest cited parameters in each of the four groups are as follows:

1. crop waste and losses/status of cultivation/productivity of output/productivity of land/changes in yield;

2. population growth and density;

3. wage (income); and

4. poverty.

The study established that these four are key socioeconomic parameters with the highest number of citations by each of the groups in Table 1. The four key parameters across the four groups address aspects of welfare within a community, which was a key issue articulated in many studies (Breisinger et al., 2013; Gurgel et al., 2021; Hashida and Lewis, 2022; Khabbazan, 2022). More specific studies concerned the welfare of agricultural communities facing climate change risks and resulting in the reduction of people’s income (Breisingeret al., 2013) and changes in consumption patterns (Nikas et al., 2018). Some of the crop models in particular enable the measurement of welfare changes identified by Gurgel et al. (2021) and Wang et al. (2021). Khabbazan (2022) integrated crop waste and losses, the nature of the land, the nature of irrigation (agriculture group parameters), population growth (demographic group) and wage income (economic group) to collectively addresses the welfare of communities. This was observed in the PRICE model, which developed a methodology to determine welfare costs and incorporated a cost risk analysis scenario for that purpose.

The economic group has nine parameters, amongst which wage, GDP, lifestyle affected and product prices have the highest citations collectively. These parameters provide an understanding of the direct economic impacts of climate change, quantifying both shifts in income and consumption patterns and contributing the analysis of climate impacts (Bobojonov and Aw-Hassan, 2014). As the name implies (economic group), these parameters are expected to have a higher number of citations in studies of economic models. With fewer citations within the group, “alternative modes of employment” and “reshaping policies (market data)” relate to flexibilities in reshaping policies such as changing plant/harvest patterns, alternative crops, etc. based on market data. Despite fewer citations, these parameters are starting to make an impact within current economic models (of climate change) by analysing the impacts of alternative livelihoods when facing climate change and determining the contribution of current policies to the economy. For instance, when considering parameters across groups, the DCGE model accounts for output productivity (agriculture group), population growth (demographic group) and wage and poverty (economic group) when identifying predominant economic impacts on communities.

Social group has the second highest number of parameters (as per Table 1). However, other than poverty, the majority of parameters have low levels of citation. Parameters such as the level of risk and damage, vulnerability to hazards, willingness to adapt and migration collectively seem to identify adaptation measures against worsening climate change. Thus, these parameters can be integrated within the EMCC framework to identify adaptation barriers and establish a foundation for the development of targeted strategies. Gawith et al. (2020) identified adaptive constraints due to changing climatic conditions using parameters such as changes in labour intensity (social group) and GDP (economic group). The BEFM, for instance, considers a cross-group of parameters, with many belonging to the economic group. Furthermore, “irrigation nature” (agricultural group) applies the principles of adaptation where climate change impacts have hastened the need for more ways to irrigate areas and new irrigation schemes for areas that are presently rainfed (Bobojonov and Aw-Hassan, 2014; Schuler et al., 2020). In accordance with the PM, Adego and Woldie (2022) consider migration (social group) income and expenses and changes in income generating opportunities (all of economic group) as part of the adaptation measures against climate change. Other studies show that the output quality would be enhanced based on adaptation measures, which include delays to harvesting time, the improved diversification of seeds and better irrigation methods (Adego and Woldie, 2022; Gawith et al., 2020; Islam et al., 2016).

Agriculture group parameters stem from productivity (crop waste and losses/status of cultivation/productivity of output/productivity of land/changes in yield) together with soil conditions/nature of the land, which are more static parameters and tend to have the highest number of citations. These parameters facilitate the quantification of damage to agricultural crops and their growing environments, informing the development of mechanisms to enhance productivity in the context of climate change. Similar to the social group, the levels of individual citation seem low compared to the parameters, such as irrigation, changes in technology, varying crop types/alternative crops, which are part of the agriculturegroup. However collectively, they support undertaking adaptation against worsening climate change.

The final group is demographic, which includes four parameters, among which population growth/density have higher levels of citation and appear to be an influencing factor within the socioeconomic impacts of climate change (Khabbazan, 2022; Palazzo et al., 2017). These parameters play a crucial role in addressing climate change impacts, as their effects vary based on community characteristics. For instance, Teklewold et al. (2017) found that demographic factors significantly influence adaptive capacities, with younger populations exhibiting greater physical resilience to climate-related challenges than older populations. These variations underscore the importance of tailoring adaptation strategies to the specific needs of different community groups. Moreover, parameters within the demographic group were used in models such as AEM, RM, MESRM and IAM (Table 1). This study argues that the decision to expand the use of these demographic parameters in models is based on the degree to which they support specific policy contributions [e.g. the PM (Ogada et al., 2020)].

The initial PRISMA identified key EMCC, which were grouped into five categories (Figure 3). The initial aim of the study was to prioritise and organise socioeconomic factors pertaining to the livelihoods of communities in the agriculture sector. Therefore, as shown in Table 1, the socioeconomic parameters were identified. The discussion above was instrumental in identifying key socioeconomic outcomes that, along with their supporting factors (in brackets), are given below:

• “Welfare of the community” (identifying income reductions, changes to consumption patterns, measurement, cost calculations);

• “Making economically orientated policy decisions based on market data” (policies being flexible, and informed scientifically, economically and socially); and

• “Undertaking adaptation in the context of improving livelihoods” (considering adaptive constraints due to climate variability, including scientific and practical measures such as varying plant harvest times and diversifying crops, adapting land, migration, changes to income and output quality).

This will help to close the theory–practice gap by making targeted recommendations aimed at improving the socioeconomic experiences of agricultural communities. As our research questions and the SLR initially focused on the agriculture sector, it is interesting to note that Outcome 1 (welfare of the community), is a highly cited outcome from the EMCC. This demonstrates the heightened sensitivity of communities’ welfare to the growing challenges of climate change. Hence, the ability of existing EMCC to identify specific income reductions in communities’ livelihoods, the resultant consumption patterns, the ability to measure specific welfare changes and attribute any intangible welfare impacts in terms of cost are found to be of value in this research. Outcomes 2 and 3 address ways to improve the livelihoods of communities with specific policy-level interventions and methods of adaptation. The three key outcomes combined offer several benefits to policymakers, economists and authorities focusing on the socioeconomic impacts of climate change on agricultural communities. These outcomes are also aligned with 1, 2 and 11 of the UN’s SDGs, which concern enhancement to community resilience through developing the ability to build back better after a climatic disaster event and be self-sustaining over time.

Despite the extended benefits of the PRISMA, starting with the identification of key EMCC (Table 1), and the important socioeconomic parameters and outcomes applicable to the livelihoods of agricultural communities (Table 1), our conclusions are made within certain boundaries that are contextually dependent. Furthermore, some obstacles which hinder the full application of some of the socioeconomic parameters due to the variability of climate change. Additionally, the cultural and traditional mindsets of agricultural communities sometimes do not support the behavioural change needed when undertaking actions based on the outcomes of these models. Accordingly, these are discussed as factors that limit or vary the application of the EMCC outcomes:

• Problems with taking average rather than actual climatic data

Mulwa et al. (2016) stated that using average rainfall data (instead of actual) caused a negative income using the TAMIAT model, due to wide variations in rainfall in the data context (place or region). A similar view was expressed by Gawith et al. (2020) who highlighted that the average rainfall was used in models after experiencing difficulty with inputting real-time data; this subsequently led to inaccurate results. Similarly, Revesz et al. (2014) asserted that finding a way to incorporate the variability of weather is more important than using averages as inputs to IAM models. Accordingly, to develop accurate adaptation strategies, it is recommended that climatic information (rainfall, temperature, wind, etc.) is collected as point-based data rather than averages over a period (Falco et al., 2018).

• Standard methodologies used in adjusting values based on ground conditions.

When using IAM, Rising et al. (2022) emphasized the use of context-specific data as inputs to models (standard damage functions). Context-specific data more accurately accommodates heterogeneity within the regions, sectors or populations. Once the impact of different data are identified, IAM can then be calibrated to suit the specific damage functions (which are standard to a specific context) to obtain output values from the models. Moreover, IAM also capture non-economic costs as standard adjustments (e.g. loss of biodiversity, impact of migration). Khabbazan (2022) identified the importance of determining the value of some of the input data into EMCC (e.g. cost of welfare) and introduced a new cost analysis scenario for future use as a standard methodology.

• Use of real-time information

Gurgel et al. (2021) advocated that when using crop models to assess climatic impacts on agriculture in a specific region, the incorporation of a few selected crops could skew the results. Accordingly, multiple attributes (rather than a small number) would improve the accuracy of the results. For instance, in their PDMChoi et al. (2015) showed that with increasing information (more attributes considered), it was easier to determine the reasons for wage dispersions between innovative and traditional farms.

• Perceptions and behavioural patterns of communities to climate change

It is important to note that despite the scientific outcomes of EMCC, communities’ preferences and perspectives might eventually determine the actions undertaken. The probit model (Adego and Woldie, 2022), for instance, identifies that adaptation strategies are optimal when there is decreased rainfall and increased temperature. However, in practice, farmers are reluctant to make behavioural changes at the onset of fewer rainfall episodes, but react quickly and are more sensitive to increases in temperature when they modify their planting and harvest times (Adego and Woldie, 2022). Overall, it can be argued that resistance to action to manage climate change occurs because of communities’ current knowledge and perceptions, and their level of understanding, rather than scientific determinants.

• Comparison with different models/theories to increase the validity of the models

To improve the value of results, the use of multiple models is recommended when studying/applying EMCC. For instance, Hashida and Lewis (2022) examined changes in agricultural land values consequent to the effects of climate change and identified that the DCEM provides more heterogeneous plot-level results and larger negative impacts than the RM, which provides country-level results. Accordingly, the authors used findings from both approaches to develop a method that estimates the welfare impact of climate change on landowners using post-harvest land values. Hence, it is important to identify the best-fitting approach to a particular circumstance by studying different models and theories.

Overall, this section provides criteria for consideration when referring to a particular economic model of climate change to ensure the accuracy, practicality and reliability of the model and its results.

This paper prioritized and organized socioeconomic factors pertaining to the livelihoods of communities in the agriculture sector. It identified: EMCC available in the agricultural context; the parameters of EMCC used to manage the impacts of climate change; outputs derived from EMCC; and factors for consideration when selecting (using) EMCC. Accordingly, 23 economic models were systematically selected and reviewed to identify their inputs (climatic conditions and parameters used) and outputs (socioeconomic impacts that they address).

According to the findings of the SLR, EMCC can be categorized as land-based, global integrated, crop or economic simulation models, or as models with specific policy contributions. Furthermore, this study identified a set of 26 parameters to determine the socioeconomic impacts of climate change on agriculture under particular agricultural, demographic, economic and social groups. The main socioeconomic impacts within the economic models (outcomes) were identified under the categories of “welfare of the community”, “making economically orientated policy decisions based on market data” and “adaptation in the context of improving livelihoods”.

The findings of this study can be used by policymakers on an international level to identify the areas that need consideration in policy formulation. Researchers should identify the parameters of impacts that need further investigation, and model developers should identify the aspects (problems) to be addressed in future economic models. This offers insights into the factors for consideration when developing reliable, accurate and practical models to determine the socioeconomic impacts of climate change on communities; it will also help to refine existing and developing new models. Moreover, this serves as a risk management strategy, which is a mechanism to identify the parameters to focus on when assessing climate impacts, suggesting the actions required to overcome the limitations of existing EMCC. This provides comprehensive guidance on climate change management and decision-making and reduces the risk of not focusing on key parameters (see Table 1) in climatic events. Lastly, this study helps to bridge the gap between theory and practice by identifying improvements to existing EMCC. Furthermore, these findings help to raise awareness of communities and critical areas (parameters) for attention when addressing climate impacts, and the actions required to enable resilience (e.g. in-depth investigations on migration and improving community willingness to adapt). The study addresses the UN’s SDGs, specifically 1, 2 and 11, which are priority areas.

There are a few limitations to this study. The SLR was based on keyword searches in three databases, and the findings are limited to the keyword combinations. Hence, there is a possibility that other useful articles were not included. Moreover, although we have followed structured selection criteria, there may still be biases in the screening process. Lastly, the findings of the studies reported in languages other than English have not been considered.