Introduction

The effect of climate change on hydrology and water resources is possibly one of the most important environmental challenges worldwide. Climate change is anticipated to intensify the hydrological cycle and to change the temporal and spatial distribution pattern of water resources. It is predicted to increase the frequency and intensity of extreme hydrological events, such as heavy rainfall, storm surges, droughts, and floods (Huntington, 2006). Water-related hazards or hydro-hazards occur due to complex interactions between atmospheric and hydrological systems. These events cause economic crises, societal disturbances, and environmental impacts, which can pose a major threat to human security (Kundzewicz & Matczak, 2015).

According to statistics from the Centre of Research on the Epidemiology of Disasters (CRED), practically every year, there are more flood events than all other disaster events combined. The flood events are also very damaging. In 2020, out of the top ten highest mortality events, six were floods (all the others were heat waves). In terms of economic damage, three flood events were in the top ten most damaging ones in 2020, with five others being associated with tropical cyclones/hurricanes, one with an earthquake, and one with a wildfire (CRED & UNISDR, 2019).

The economic impacts of extreme hydrological events can be separated into direct damage and indirect losses. Direct damage includes the damages to fixed assets and capital (including inventories), losses of raw materials, crops, and extractable natural resources, and, most importantly, mortality, morbidity (injuries and sickness), and population displacement. All can be a direct consequence of extreme hydrological events (i.e., heavy rainfall, a storm surge, a flood, or a drought). Indirect losses are the reductions in economic activity, particularly the production of goods and services—which will be greatly reduced after the disaster and because of it (Cavallo & Noy, 2009). Indirect damages may be directly caused by the immediate impact (a first-order effect), or they may be caused because post-disaster reconstruction pulls resources away from the usual production activities (a higher-order effect) and consequently damages suppliers (Noy & duPont, 2018). These damages can also be divided between the short run and the long run (Cavallo & Noy, 2009). This article focuses on a specific hydro-meteorological hazard—drought— and the direct damages and indirect losses from droughts in agriculture.

Drought is recognized as being associated with some of the most high-profile humanitarian disasters of past years, threatening the lives and livelihoods of millions of people, particularly those living in semi-arid and arid regions. Notable recent events, for instance, were the drought in the Horn of Africa in 2011 and that in the Sahel in 2012 (Food and Agriculture Organization of the United Nations [FAO], 2015). Droughts depend on a set of factors: high temperatures, low humidity, the timing of rain during the crop growing cycle, and the intensity and duration of precipitation, as well as its onset and termination (Mishra & Singh, 2010). While drought has wide-ranging effects on many economic sectors, the agricultural sector bears much of its impact because it is very dependent on precipitation and evapotranspiration. Approximately 1.3 billion people—40% of people in the world—rely on agriculture as their main source of income (FAO, 2017). In developing countries, the agriculture sector absorbs up to 80% of all direct damages from droughts, with numerous impacts on water availability and agricultural production (FAO, 2017). Droughts are a major threat to food security and rural livelihoods, increasing local poverty, dislocating large numbers of people, and hindering progress toward the Sustainable Development Goals (SDGs). Even so, as Wilhite and Pulwarty (2017) pointed out, “global figures for the trends in economic losses associated with drought do not exist.”

Direct and Indirect Impacts of Droughts on Agriculture

The prolonged soil moisture deficit that is the main characteristic of droughts can reduce crop yields, crop hectarage, and pasture because of its negative effects on plant growth (Ray et al., 2018). It can lead to adverse effects on revenues from crop and livestock sales, to increasing production costs, and possibly to decreases in farm profits, in employment, and in life in rural communities more generally (Kuwayama et al., 2019). Crop failures and pasture losses are the main direct effects of drought on the farming sector. Water deficit causes poor pasture growth and may also decrease fodder supplies from crop residues. Lack of fodder leads to weight loss and increased mortality among livestock. As the period of drought-induced food deficit lengthens, farmers start selling livestock to reduce feed demand, thereby reducing their households’ wealth (Toulmin, 1987). Drought-induced production losses constitute a negative supply shock, and ultimately the interaction between that and the price elasticity of demand will determine their overall impacts.

Nevertheless, drought-induced losses are not entirely borne by farmers; instead, some of the losses are transferred to consumers through increased prices. Thus, in principle, it is even possible that some farmers benefit from the drought impacts, becauses output price increases by a higher percentage than the supply falls (i.e., if the price elasticity of supply is high; Ding et al., 2011). These price hikes are not uncommon. For example, drought was considered one of the key triggers of the global food price crisis in mid-2008, a crisis that dissipated when the global financial crisis started in September of that year (Piesse & Thirtle, 2009).

Indirect impacts of drought can be more damaging than direct impacts, given that the impacts are then spread to other industries and production sectors and along supply chains (Gil et al., 2013). The indirect impacts can flow both upstream and downstream through supply chains. For instance, drought-induced production losses may reduce farm supplies to downstream industries, such as food processors, packaging plants, and ethanol plants. These industries then must reduce their own production due to the lack of inputs, or they may need to pay higher input prices. Upstream, farmers might reduce their demand for farm inputs like fertilizers and pesticides from their own suppliers.

Impact of Climate Change on Droughts

Global climate change modeling indicates that rising global temperatures will intensify the Earth’s hydrological cycle, including its uncertainty, worldwide monsoon precipitation, and the severity of wet and dry events (Intergovernmental Panel on Climate Change [IPCC], 2021). The lack of available moisture leads to increased drying and heating of the land surface (Mukherjee et al., 2018), increasing the risk of drought or prolonging periods of drought in drier land areas.

According to the IPCC, global droughts intensified and lasted longer over wider areas during the 20th century, mainly in association with higher temperatures and decreased precipitation (IPCC, 2019). Regional climate change projections suggest that drought is likely to be further aggravated in some regions under future climate change, particularly in the Mediterranean, Central and Southern Europe, central North America, Central America and Mexico, northeast Brazil, and Southern Africa (IPCC, 2019).

Predictions of future changes in precipitation are less certain than predictions of future warming. However, a large decline in precipitation over low and mid-latitudes is projected by most climate models. Thus, arid and semi-arid regions in the Mediterranean, Southern Africa, and the south-central United States are expected to experience 10% to 30% decreasing water availability by 2050 (IPCC, 2018). Global circulation models (GCMs) also project worldwide decreases in soil moisture over the same regions, Central and Western Europe, and Australia (van der Linden et al., 2019; Wang, 2005). This suggests that these regions will also experience an increase in agricultural droughts. Indeed, given the projected climate change, more frequent and severe droughts are predicted in the 21st century around the world (Haile et al., 2020; Schwalm et al., 2017; Trenberth et al., 2014). If vulnerability to droughts will not decrease, climate change is likely to exacerbate the impacts of droughts on agriculture and more broadly on many economies around the world (Dai, 2011; Trenberth et al., 2014).

Drought Definitions

Regional differences in hydro-meteorological elements, socioeconomic factors, and the local features of water demand have been obstacles to having a precise drought definition (Mishra & Singh, 2010). Drought refers to a temporary dry period, a period of below-normal rainfall lasting for months or even years. Drought occurs in almost all regions, even in very wet regions like New Zealand, regions with distinct wet and dry seasons like California, or very dry regions, like the Sahel.

Droughts differ from other natural hazards in several ways. Even a declaration whether a region is suffering from a drought can be controversial and subjective (Hollins & Dodson, 2016). Consequently, the exact onset and termination of a drought episode also often cannot be identified because the drought’s effects often accumulate over a long time period and may linger for years after the end of the drought event itself. Drought is therefore often referred to as a “creeping disaster” (Wilhite, 2003). Also, drought rarely occurs only in a limited region or over a very short time. Thus, its impacts are typically spread over large geographical areas and can last a while. Furthermore, a drought rarely leads to damage to human-made structures, in contrast to floods, hurricanes, and earthquakes. Accordingly, risk assessment and provision of relief are much more challenging for drought than for other natural hazards (Mishra & Singh, 2010; Wilhelmi & Wilhite, 2002). Drought events can be caused or can be intensified by human activities, including over-farming, land-use changes, unsustainable irrigation practices, over-exploitation of water sources, deforestation, and erosion (Mishra & Singh, 2010).

Drought Classification

Droughts are traditionally classified into four categories (American Meteorological Society [AMS], 2004; Wilhite & Glantz, 1985): meteorological drought, agricultural drought, socioeconomic drought, and ecological drought.

A meteorological drought is defined by the magnitude and duration of a precipitation shortfall event. It is meteorological because decreases in precipitation depend on local atmospheric conditions. Most meteorological drought definitions relate to variation of actual precipitation in relation to the normal climatic conditions (average precipitation on monthly, seasonal, or annual time scales) for a specific region (AMS, 2004). Hydrological drought is associated with a period of deficit in surface water (e.g., rivers, lakes, reservoirs) and/or groundwater. The frequency and severity of hydrological drought are frequently defined at a catchment or river basin scale (Wilhite, 2003).

Agricultural droughts are defined by their agricultural impacts. The term is most used in reference to non-irrigated agricultural areas. Agricultural drought focuses on precipitation shortages, differences between actual and potential evapotranspiration, and soil moisture deficits—all factors that affect crop growth.

A socioeconomic drought is distinctly different from the other types because it specifically refers to the supply and/or demand of some tradeable good. The good can be cash crops, but it can also be other traded goods (e.g., hydro-electricity). Of course, socioeconomic drought has meteorological and hydrological elements. The term suggests that the magnitude of a socioeconomic drought depends not only on the meteorological phenomenon, but also on socioeconomic vulnerability to the precipitation shortfalls and water shortages (Wilhite et al., 2014).

In the early 21st century, a new concept, ecological drought, was introduced. This concept combines the ecological, climatic, hydrological, socioeconomic, and cultural dimensions of droughts into a more holistic framework. Ecological drought is an “episodic deficit in water availability that drives ecosystems beyond thresholds of vulnerability, impacts ecosystem services, and triggers feedbacks in natural and/or human systems” (Crausbay et al., 2017).

Drought Indices

In recent decades, numerous indices for monitoring and quantifying drought have been developed. Lloyd-Hughes (2014) counted over 100 drought indices that have been proposed for different types of droughts. A drought index is a crucial component in providing a quantitative assessment of drought impacts and drought characteristics, including intensity, duration, and spatial extent.

Drought typically lasts for more than a few months, and severe drought can persist for several years, or even decades, in so-called megadrought (Dai, 2011). For example, according to the United States Drought Monitor, almost 75% of the western United States is currently experiencing a megadrought that has lasted for more than two decades (National Oceanic and Atmospheric Administration [NOAA], 2020).

A large body of literature describes the range of drought indices designed to measure and detect drought (Svoboda & Fuchs, 2016). Given the complexity of droughts, various sources of drought-related elements, such as precipitation, vegetation growth conditions, soil moisture, and land surface temperature, can be integrated to indicate the spatial extent, duration, and intensity of droughts. It is apparent that the aggregation of all drought-related factors depends on the availability of climate and weather data. The World Meteorological Organization (WMO) defined a drought index as “an index which is related to some of the cumulative effects of a prolonged and abnormal moisture deficiency” (WMO, 1992). Drought indices are categorized by type and function, and they have been classified into the following groups: meteorological indices, soil moisture indices, hydrology indices, remote sensing indices, and composite or modeled indices.

Data and Methodological Challenges

Given that droughts threaten agricultural production, the ecological environment, and socioeconomic development, drought monitoring is a necessary step toward enhancing drought resilience and mitigating risks. However, there are some challenges in drought monitoring, assessing drought impacts, and predicting their occurrence.

Data-related challenges include both data availability and event definition (Brunner et al., 2021). The data availability of drought impacts is very limited; accordingly, a scarcity of comprehensive information is an obstacle to any impact modeling. The transition from hazard to disaster is generally more complex for droughts than other damaging natural hazards like floods or earthquakes (Wens et al., 2019). Moreover, due to the slow-onset characteristic of a drought event, the opportunity to initiate data collection is often lost before there is a realization that the event has started (National Research Council, 1999).

Several emerging studies address big data and its applications in drought monitoring, evaluation, and prediction using heterogeneous sources of data, such as weather-sensor data, satellite (remote-sensing) data, weather forecasts, and water-usage reports. Balti et al. (2020) introduced and reviewed the challenges of big data analysis in drought monitoring. The challenges are similar to those with other big data uses: data collection (e.g., data structure, data deluge, heterogeneity in data, and data fusion), data processing (the choice of approach and the tools to use), and data management challenges (data storage, infrastructure costs, and memory loading). Balti et al. concluded that statistical and artificial intelligence (AI) approaches need to be used so that big data analytics can be beneficial in drought monitoring (Balti et al., 2020).

Assessing the impacts of droughts is difficult because identifying the hazard itself is hard, and a drought event usually does not cause visible damage to infrastructure, but there may be diffuse, delayed, and intangible effects (Kallis, 2008; Naumann et al., 2021). Several reasons make it difficult to trace these effects or attribute loss records to the drought event. First, most of the impact of droughts is through its secondary (indirect or second-order) consequences. For example, a study in Ethiopia found that changes in food prices can be more notable than primary impacts, such as losses in agricultural production (Holden & Shiferaw, 2004). In India, in another study, rural households were found to stop sending their children to school in drought events (Chatterjee et al., 2005), which leads to long-run adverse socioeconomic consequences. Thus, each of these secondary effects requires a specific (and maybe separate) framework of economic analysis.

Second, compound effects make it difficult to isolate drought’s effects in the aggregated macroeconomic variables. For example, it may be difficult to isolate the effect of drought on food prices from the effects of global supply and demand changes, agricultural or energy policy changes, or geopolitical pressures (Kallis, 2008; Markandya & Halsnaes, 2000).

Third, any accounting for impact and loss must rely on defining a baseline or counterfactual scenario (i.e., a scenario of what would have happened in the absence of the drought). However, constructing such a baseline itself depends on assumptions about any other changes in the determinants of exposure or vulnerability (Guha-Sapir et al., 2004).

Fourth, indirectly, some sectors or groups might benefit from droughts, so identifying the distributional consequences of the impacts is important. For example, farmers in nondrought areas compete with those affected by a drought and can therefore benefit from drought-induced increases in prices. New Zealand, for instance, is the market maker in the global milk powder market. Thus, drought events in New Zealand can increase dairy farms’ revenue and profits when drought increases global prices (Pourzand et al., 2020). Subsequently, other dairy exporters elsewhere indirectly benefit even more from the occurrence of a drought in New Zealand. Similar dynamics may be possible for other regions that grow a significant concentration of specific crops (such as California and almonds).

Fifth, aggregates conceal distributional effects. Droughts hit different groups or sectors in different ways. While drought effects that can be measured in monetary terms can also be compared, other intangible impacts, such as deaths or irreversible environmental and cultural damage, are more difficult to quantify or compare. Therefore, these impacts are often omitted entirely from drought assessment (Kallis, 2008; Markandya & Halsnaes, 2000). These methodological challenges in part prevent worldwide efforts to account for drought impacts and vulnerability (for further discussion of these points, see Gerber & Mirzabaev, 2017).

Economic Assessment Methods

Several methodological approaches are typically applied to assess different types of drought impacts—direct, indirect, and intangible—across scales (from intrahousehold or crop-specific to economywide or even global effects). Some of the methods are used for estimating only one impact type (e.g., only intangible costs), while other methods can be employed to evaluate more than one type of drought impact.

The most frequently applied approach for assessing drought impacts is based on econometric modeling of farm-level crop or livestock losses (Bastos, 2016; Birthal et al., 2015; Mare et al., 2018; Quiroga & Iglesias, 2009) as well as regional- or basin-level drought impacts (Gil et al., 2013; Kamber et al., 2013; Kirby et al., 2014). A review by Logar and van den Bergh (2013) concluded that market valuation methods (i.e., production function, market prices, avoided costs, replacement or repair costs) are the most suitable techniques for assessing direct tangible impacts of droughts. Their advantages include being easy to use, covering any economic sector, and providing reasonably precise estimations (Meyer et al., 2013). Still, they do not account for behavioral changes and input substitutions.

A different approach relies on structural modeling. This approach can be based on partial equilibrium models (of a specific market), computable general equilibrium (CGE) modeling that attempts to comprehensively model more markets and their interactions (García-León et al., 2021; Horridge et al., 2005), and input–output (I-O) static or dynamic models (Howitt et al., 2015; Jenkins, 2013; Medellín-Azuara et al., 2016). Such modeling, unlike many of the econometric approaches, has the advantage of accounting for all (or most) sectors of the economy. It therefore has the potential to capture both direct and indirect effects (Freire-González et al., 2017). Since CGE models typically assume perfect adjustment to market-clearing equilibria, they result in overly resilient responses. Thus, the CGE models’ results may be considered a lower-bound estimate of the economic impacts (Rose, 2004).

Integrated assessment analysis that specifically focuses on agriculture is also possible. It can include biophysical-agronomic models, integrating crop models with the economic assessment (Fischer et al., 2005; Holden & Shiferaw, 2004; Rosenberg, 1993), and hydrological-economic models that link the economic assessment to a hydrological model (Basheer et al., 2021; Kahil et al., 2016; Ward et al., 2006).

A Ricardian hedonic price approach connects variations in farmland values across space with variations in climate, assuming each farmer has adapted to the local climate (Massetti & Mendelsohn, 2011; Mendelsohn et al., 1994). Whereas biophysical-agroeconomic models assume no adaptation, the Ricardian estimates reflect farm adaptation. However, when variables contributing to land value that is correlated with climate are omitted, they confound the estimation of the relationship between climate and land values (Deschênes & Greenstone, 2007).

Empirical Studies of Drought Impacts

Some recent studies have focused on climate-related risks, extreme weather, and agriculture (e.g., Moore & Lobell, 2014; Schlenker & Roberts, 2009). The focus of most studies has been the impacts of changes in temperature and precipitation on agricultural production. For example, Schlenker and Roberts (2009) estimated the relationship between weather and yields for corn, soya beans, and cotton in the United States. They found that there is a nonlinear relationship between yields and temperature in both the cross-section of counties and the aggregate annual time series. Ali et al. (2017) investigated the impacts of maximum temperature, minimum temperature, rainfall, relative humidity, and sunshine on major crops in Pakistan (wheat, rice, maize, and sugarcane) using time-series data for the period 1989–2015.

Kumar et al. (2011) examined the effect of monsoon drought on the production of, demand for, and prices of seven major agricultural commodities—rice, sorghum, pearl millet, maize, pigeon pea, groundnut, and cotton—in India. They showed that drought during the monsoon period has an adverse effect on agricultural production. Yet, loss of production also led to an increase in the prices of these agricultural commodities, thus ameliorating the impact on farmers’ balance sheets. The increased prices, however, have an adverse impact on consumers of the foodstuffs; so, clearly, any improvement in farmers’ profitability has to be compared to the decreased well-being associated with increasing prices for everyone else. Shakoor et al. (2011) and Barrios et al. (2008) showed a significant negative impact of rising temperatures on agricultural production and a positive impact of rainfall in lower income countries.

In contrast, Moore and Lobell (2014) found that European agricultural profits could moderately increase under climate change if farmers implement adaptation measures, but they could decrease in many regions if there is no adaptation. Broadly, studies aiming to assess the impacts of natural hazards, particularly droughts, can be summarized within two main concepts—economic loss assessment and financial loss assessment (Penning-Rowsell et al., 2013).

An economic loss assessment is generally implemented on a macro (aggregate) level, tracking macroeconomic variables, such as changes in GDP, production volume and value, the trade balance, and employment, for the entire country or for a region. Most nationwide ex post measurements of drought loss have been conducted in high-income countries: the United States and Canada (Howitt et al., 2014, 2015; Kulshreshtha et al., 2003; Medellín-Azuara et al., 2016) using I-O tables, Australia (Horridge et al., 2005) using CGE modeling, New Zealand (Kamber et al., 2013) using a time-series econometrics framework, and Italy (García-León et al., 2021) using a CGE model combining agronomic and economic components.

Quantifying the economic loss of droughts in low-income countries is more difficult, so there are fewer studies that have attempted it. For example, in the Sahel region, many countries have experienced a series of severe droughts potentially causing substantial socioeconomic costs (Mishra & Singh, 2010), but few studies assessed the economic loss. Taylor et al. (2015) estimated on average $237 million (USD) loss annually to droughts in Uganda during the previous decade. Sadoff et al. (2015) applied a fixed-effects panel regression to compute the cumulative impact of drought over time (1980–2012). They found that droughts reduced GDP substantially in a model calibrated for Malawi, and they demonstrated how the impact of droughts can compound over a long time.

A financial loss assessment is conducted at the micro-scale (farm, household) or the meso-level (a local community). For assessing financial loss in agriculture, a range of economic indicators are typically used. The most frequently used indicators for estimation of financial loss due to droughts are the crop or livestock sales, with revenue and gross margin adjusted with variable costs (Edwards et al., 2019; Holden & Shiferaw, 2004; Lopez-Nicolas et al., 2017; Pourzand et al., 2020; Todorović et al., 2021). However, it is worth mentioning that one of the challenges of financial assessment of drought is that farms’ gross income can differ dramatically from year to year for reasons apart from drought events. Fluctuations in the income of farmers might be due to other types of hazards or extreme weather, changes in input prices, and changes in commodity prices (Edwards et al., 2019).

Regarding other financial indicators, farm business profit, profit at full equity, rate of return, and off-farm income were all calculated to assess Australian farms’ financial performance (Hooper et al., 2008). Moreover, Lawes and Kingwell (2012) used operating profit per hectare, return on capital, business equity, and the debt-to-income ratio to analyze the financial performance of farms affected by drought, while Kingwell and Xayavong (2017) also used retained profit per hectare, and Pourzand et al. (2020) used interest coverage as another business indicator. All these indicators except business equity vary a lot, as they are affected by seasonal and market conditions.

Some findings from the literature are worth noting here. First, it remains difficult to adequately characterize droughts, and there is no consensus on their definition, identification, and measurement. Second, the impacts of droughts on financial loss indicators may depend on the farm’s production type and vary across rain-fed and irrigated farmland. Third, farmers use various coping strategies, and these can ameliorate much of the adverse impact on farms’ balance sheets. Equally, drought-induced higher prices can allow farmers to be more resilient to the adverse production shock.

Kingwell and Xayavong (2017) explored how the incidence of drought (the number of drought years) affects Australian farms’ financial performance. They found a significant positive impact on the operating profit per hectare and retained profit per hectare in drought-affected agricultural regions. Lopez-Nicolas et al. (2017) assessed the monetary value of production in irrigated agriculture in the Jucar River Basin in eastern Spain. The effect of crop price volatility was also isolated from the losses due to water scarcity in their assessment of drought impacts. They demonstrated that an increase of crop prices can partially counteract losses from reductions in crop production due to water scarcity. In another study, Pourzand et al. (2020) investigated the effect of changes in drought frequency (an additional day of drought) on farms’ financial outcomes in New Zealand. They showed that drought events positively impact dairy farms’ revenue and profit per hectare, an affect they attribute to drought-induced increases in the global price of milk (New Zealand being the dominant seller in this market). Todorović et al. (2021) examined the impact of extreme weather events (floods and droughts) on farm balance sheets for the two most common farm types in Serbia (mixed crop–livestock and specialized crop farms). They showed that performances of both farm types were more sensitive to drought than to floods. They also demonstrated that the mixed crop–livestock farming systems, with diversified sources of income, were less vulnerable to extreme weather events than were farms focusing on a single specialized crop.

For lower income countries, in a study on household-level impact of drought in Ethiopia, Holden and Shiferaw (2004) examined increased risk of drought on household production, welfare, and food security. Their results showed that the indirect impacts of drought on household welfare through the impact on crop and livestock prices were larger than the direct production effects of drought, because even farming households were typically net buyers of crops for food during a drought, so they had to buy crops at a higher price. Holden and Shiferaw also showed that severe droughts caused livestock prices to decline, because drought-affected households were forced to sell animals to buy food. This led to a loss in livestock value.

Most vulnerability studies in drought-prone areas focus on micro-level mitigation actions involving households, communities, and individual businesses. Drought vulnerability differs considerably between low-income countries, where the economy depends on rain-fed agriculture and pastoralism and drought can cause hunger and livestock loss, and high-income countries with very diversified economies, where the concern is significant economic or asset losses (Kallis, 2008). For instance, Holden and Shiferaw (2004) found that better access to credit, financial services, and alternative savings mechanisms could help agricultural households to mitigate drought risk in Ethiopia, but these resources were not often available. Secure land tenure, efficient markets, and credit are the most critical elements of adapting to drought risk and pursuing mitigation actions (e.g., among farming households in Morocco; Kusunose & Lybbert, 2014). Similarly, Alam (2015) found that access to electricity and access to agricultural extension services (agronomic knowledge) are crucial for farmers’ adaptation in Bangladesh.

Another commonly used mitigation strategy is diversification. Diversification may include both agricultural production variation (mixing cropping and livestock farming), and income diversification with off-farm employment. In many drought-prone communities, diversifying livelihoods is considered the most important drought-resilient management practice (Antwi-Agyei et al., 2014; Bryan et al., 2013; Eriksen et al., 2005), especially diversification with off-farm income (Block & Webb, 2001). In addition to livestock’s value in diversification of income sources, livestock can also serve as a saving mechanism through periods of drought in West Africa (Kazianga & Udry, 2006). Lange and Reimers (2021), for example, found that some West African farm households finance their food consumption via divesting of livestock assets in response to drought conditions and declining incomes.

Beyond diversification, farm structural changes—such as land-use change, switching to more drought-resistant crops, and modification of cropping patterns—were also found to be useful options for building resilience against droughts in Europe (Huntjens et al., 2010), in China (Lei et al., 2014), in France (Willaume et al., 2014), in Australia (Kingwell & Xayavong, 2017), and in the United Kingdom (Holman et al., 2021).

Equally, improved irrigation techniques and the development of alternative water sources are frequently cited as mechanisms for coping with drought (Antwi-Agyei & Nyantakyi-Frimpong, 2021; Ncube & Shikwambana, 2016). For example, Dono and Mazzapicchio (2010) found in Sardinia that the impact of future droughts could be minimized by accessing groundwater resources. Similarly, groundwater wells are sometimes useful as emergency sources of water during droughts in Ethiopia (MacDonald et al., 2019), South Africa (Mussá et al., 2015), Spain (Lopez-Nicolas et al., 2017), and India (Shah, 2009). Although irrigation and tapping into groundwater resources buffer against short-run rainfall deficits, in the long run, reservoirs and aquifers may dry up, especially as a consequence of multiyear droughts and general overuse. Unfortunately, the groundwater deficit could, eventually, contribute to higher drought loss and more environmental damage (Eakin et al., 2016; Liverman, 1990). Therefore, when agriculture is intensified in irrigated areas, losses may end up being greater than in rain-fed farming areas (Eakin et al., 2016).

At the macro policy level, drought mitigation may involve various institutional and policy measures. Examples include the establishment of interregional water markets (Booker et al., 2005; Harou et al., 2010; Wheeler et al., 2014), an early warning system (Pulwarty & Sivakumar, 2014), investment in water and irrigation infrastructure and in drought preparedness plans (Zilberman et al., 2011), water conservation measures (Taylor et al., 2015), and crop insurance (Barnett et al., 2008).

Future Research

Ultimately, as detailed above, there has been a lot of research on the economic risk from droughts (and from other hydrological hazards not surveyed here). While there are many methodological and practical difficulties in pursing this empirical research, it has also offered many possible mitigation and adaptation suggestions for improving the ability to manage this risk. Policies like crop diversification and improved use of available water resources through better irrigation techniques have all been explored in various contexts, for various crops, and in various regions. Still, the economic risk from droughts is only increasing, as both the hazard itself and the population exposed to it are increasing. Especially worrying are places like the Sahel, where the hazard is intensifying, population is rapidly increasing (as high growth rates are decreasing slowly), and vulnerability is being exacerbated by conflict.

Creative solutions to these problems are being proposed and tried all the time—for example, in the African Risk Capacity (ARC) parametric insurance (risk pool) program. The ARC risk pool had 15 sub-Saharan African participant countries in the 2021/2022 risk pool (up from 13 the year before); it potentially offers a way to overcome some of the challenges described above. No single program or policy will, of course, be able to deal with the complex challenge of an increasing frequency and intensity of droughts in many of the least advantaged parts of the world. The need for further research is thus becoming even more urgent.