Tim Benton and Catherine Thompson of Leeds University explore the problems of food supply in a changing world and identify the need for multi-metric analysis to estimate risk and embed sustainability and resilience in food supply chains.
Growing demand for food in a changing world
In 2007/8, when there were historically low food stocks, a relatively small weather-related production shock led to rapid increases in the prices of the major agricultural commodities. This increase was compounded by some countries imposing export barriers in order to maintain their own food security. Overall, the FAO food price index increased by over 100%. Three years later, in 2010/11 a similar spike occurred, partly influenced by weather in Eastern Europe and Russia. In 2012, the worst drought to hit the American Midwest for half a century triggered comparable spikes in international maize and soybean prices [1].
These spikes led to significant impacts worldwide. In rich countries, food price inflation was marked and the poorest suffered, resulting in people trading down on food quality or quantity and in the process spending significantly more. In poorer countries, especially those with fragile governance, rapid food price inflation undermined civil order and, in part, was a spark for the Arab Spring and the consequences that have followed. UNEP (United Nations Environmental Programme) modelled a doubling of the FAO food price index and showed that globally a food price spike is overwhelmingly economically negative. In a subset of about a quarter of countries, a doubling of commodity prices leads to food price inflation of more than 10%. These countries include Morocco, Egypt, Bangladesh and Indonesia, which suffered ‘food riots’ in 2007-8.
A world where food is available at the prices we have come to expect cannot be taken for granted.
This sequence of price spikes, and their consequences, underlined two things. First, that a world where food is available at the prices we have come to expect cannot be taken for granted. Sir John Beddington’s powerful analogy of ‘the perfect storm’ – of rising demand for food, water and energy whilst climate change creates increasing constraints – became a call to action on how to manage demand growth in a changing world. Second, the rapid and significant price spikes created a new focus on volatility.
There is now irrefutable evidence that the climate is changing. Most people picture climate change as a gradual trend in the average weather, but perhaps more importantly, on a year-to-year basis, climate change can also impact variability in weather. As an example, if the trend is for annual rainfall to decrease in an area, one might expect flood risk to go down. However, if rainfall becomes more variable, this could lead to more droughts and more floods. There is already good evidence that extreme weather events, from intense storms to droughts and heatwaves, are increasing in frequency and severity at a considerable rate and the likelihood is that extreme weather will increasingly become unprecedented in severity and impact [2].
Extremes, by their very nature, tend to disrupt human systems and have the potential to interact with a whole range of other factors, thereby acting as a risk-multiplier. For example, droughts or floods can reduce food production and availability and lead to a food-price spike, which in turn can cause civil unrest in unstable economies, an increase in de-forestation (and thus carbon emissions) as rising prices promote land-clearing for agriculture, and intensification of production, impacting on water availability and quality. Bad weather can also influence pest and disease outbreaks, exacerbating production issues. As weather becomes more variable and unprecedented extremes occur, there is increasing potential for volatility in production and prices. This creates a need to focus on building systems that buffer against shocks and are resilient.
Image may be NSFW. Clik here to view.
Resilience theory
Resilient systems are stable: either they are robust to perturbations or quickly return to a pre-perturbed functional state. Whereas nonresilient systems, once perturbed, take some time, or even never, return to the pre-perturbed state.
Resilience theorists often explain resilience in terms of imagining balls. The system’s functioning (e.g. the agricultural yield in an area) is measured by the position of the ball on a surface, so when a ball is stationary it is stable. The external environment – for example weather – creates perturbations which knock the ball. When a ball is in a dip in the surface, it is more likely to remain in one place and to be stable when random perturbations knock it. If you drop the ball into a cone (Figure 1A), it immediately stops. If you drop the ball into a saucer (Figure 1B) it will roll around until it stops. Furthermore, a perturbation to the ball in the cone will not make the ball move much, but the same shock to the ball in the saucer will. The cone system is more resilient to a perturbation and, if the ball is moved, it very rapidly regains stability.
If the ball is perturbed enough, it bounces out of the dip and comes to rest somewhere else – either on a flat surface, where any perturbation will send it off again, or in another dip. A shock can bounce the ball from one stable state to another (Figure 2).
Image may be NSFW. Clik here to view.
Applying resilience theory to agricultural systems
The dips, or stable states, can be created by technology or natural processes. An example of the former is that plant protection products (pesticides) can make an agricultural system both high yielding and stable. With access to lots of pesticides, pests can be controlled and annual yields can be stable year-by-year. However, without pesticides – perhaps because society bans them or pests evolve resistance – yields would be much more variable and the overall system less resilient. This example illustrates that ‘stable states’ can change as the environment changes – what may once have been more stable may become less so (the dip evolves from cone to saucer over time).
Without management, most ecological systems are fairly stable as ecological processes create a community of co-existing species that compete and complement each other. Agriculture, by creating monocultures, promotes systems which are inherently lacking ecological or ‘natural’ resilience; resilience is supplied by human interventions to control pests, diseases and nutrients. The history of agriculture is replete with examples of stable systems becoming unstable and the system transitions from one stable state to another. Some are climate-driven, such as transitions from wetter to drier climates, and have led to the collapse of civilisations like the Maya. Others are weather-related, such as the dustbowl in the mid-west. In this region, a once stable and productive agricultural system was initially perturbed by drought. Failure of crops and thus farm enterprises left land bare leading to significant soil erosion (over 20m ha of land lost up to 10cm of soil), which intensified the drought through surface-climate feedbacks. Loss of functionality from erosion led to very long term impacts on productivity and land value, which remains depressed 80 years on.
Our food system is, however, more than agriculture. Most processed food requires a complex set of supply chains, bringing ingredients from all around the world. Some of these are major commodities like maize, wheat, rice and soya; others are produced in smaller volumes but are nonetheless crucial for flavour or quality, such as chocolate, herbs, spices and salt. As in 2007-2012 (Figure 3), the restriction of supply by external circumstances can lead to significant month-on-month volatility in international markets. This lack of resilience may be created by, for example, a shortfall in production relative to demand, or a transport bottleneck in one commodity. This can spill-over to markets and reduce availability of other products, simply through loss of market confidence.
Image may be NSFW. Clik here to view.
Figure 3 - A measure of volatility in the FAO monthly food price index over the last 35 years. This is calculated as the square root of the difference between the monthly price index and the 12 month average [2], divided by the average. This is effectively a moving average of the standard deviation of prices over a year. The blue line is the deflated time series, the red the nominal
The lesson of the last decade of commodity prices is that our markets are not as resilient as we would like due to changing weather and its impacts on agriculture or supply chain logistics, correlated changes in markets (like oil and grains) or geo-political stability. As we expect perturbations to increase in magnitude and frequency in future, we should concentrate on building systemic resilience.
It is axiomatic, that reducing the variance in the system – by managing resilience – comes at a cost. Theory suggests that as volatility increases, the costs of coping with it also increase and eventually a tipping point is reached, where investing in reducing variance becomes more profitable than expecting supply chains to be stable. The question is then how to build resilience?
Routes to resilience
From the perspective of a single product, like an orange, a resilient supply chain would be one that manages to put oranges on a retailer’s shelves despite perturbations in production (or transport logistics). There are four main classes of solution to ensuring resilience.
1. Insurance production (making sure more is grown than needed, allowing for a loss contingency)
Insurance production is one of the greatest challenges for fresh produce importers as when demand is high and product is in short supply, growers prefer flexible pricing and not being fixed into a contract so as to maximise their profit (Figure 4). When demand is low, growers prefer fixed premium price contracts and importers suffer the financial implications of product wastage. Whilst insurance production may contribute to resilience in the short term, any increase in wastage as a result reduces overall sustainability (and arguablyresilience in the long term).
Increasing grower resilience requires a close working relationship, where an ongoing market commitment is assured.
2. Investing in making the supply chain resilient to perturbations
This may involve investing in the producer to ensure that production can proceed whatever the weather (e.g. sharing weather-related risks, investing in soil carbon to enable water retention in dry periods, investing in relationships along the supply chain). An important class of actions here is to assess supply chain risks to resilience through measurement and then reduce risks through actions. Increasing grower resilience requires a close working relationship, where an ongoing market commitment is assured (Figure 4). In essence the importer needs to offer the grower reassurance that the financial risk of specialisation is worth a premium ongoing price. The importer needs to be assured that the product will be readily available and of a high quality. The gamble is the effect of future climate change. Could the grower become fundamentally limited by its location to a degree un-mitigatable by technology or procedure? Is it also possible that the optimum growing conditions will be available more cheaply at another location? To use the above analogy, climate change may be making the dip shallower.
Image may be NSFW. Clik here to view.
Figure 4 - The push and pull factors dictating the fresh produce supply chain
3. Increasing future flexibility or bethedging
This pushes against the drive for a simple supply chain with a single or small number of suppliers within a single location (to avoid ‘putting all your eggs in one basket’) (Figure 4). Importers must always keep one eye on the future climatic growing regions as well as regions less likely to be impacted by risk events, such as social unrest, earthquakes or volcanic explosions. For example, sparkling wine can now be produced in the UK, limes in Spain and blood oranges, which used only to come from the slopes of Mount Etna, can now be grown in many other places.
4. Reducing the systemic risks in the market
Some of the food price spikes were driven by public policy decisions (e.g. countries instituting export bans). Increasing recognition that volatility creates significant problems means that policy makers (e.g. across the G20) are focusing on how to ensure market functioning in times of perturbation. However, public policy is only one aspect of overall systemic risk. A strong focus on price-cutting and efficiency can create a ‘race to the bottom’ where resilience (and sustainability) are seen as secondary issues and too expensive for the current business environment [3].
Sustainability and resilience
A ‘sustainable’ food system is one in which production is sustained over time, which implies that there is a relationship between sustainability and resilience. If the system is not sustainable, it is reducing resilience in the long term and thereby increasing risk. To address resilience, quantitative, multi-metric elements of ethical and environmental sustainability have to be combined with a predictive element of risk.
Until recently, the focus of sustainability was via single factors, such as water use and carbon footprint analysis, with the challenge being to provide sufficient business justification (assessment of financial risk or commitment) for water use reduction or costly carbon footprint analysis procedure. One approach was to focus on target metrics which were crucial for food production in a particular region. For example, much of Spanish agriculture comes from regions exposed to severe water stress, where agriculture is a major user of water (irrigation is used for 60% of total production and in 80% of total farmer exports) [4].
However, a simplistic approach targeting a single metric and driving the efficiency of use can be counterproductive. Driving down water use in Southern Spain, via higher irrigation control, lower labour costs and concentrating on higher quality fruit, created a Jevon’s paradox: increasing efficiency stimulated demand. The improved irrigation efficiency allowed expansion of the growing area for higher revenue crops making it subject to more market volatility and also higher water consumption[3]. Similarly, whilst the technological advance of ferti-irrigation reduced water pollution levels, it has also markedly increased energy usage[3]. In Heluva, the shift to high revenue, water demanding, strawberry production has been in direct conflict with the health of the nearby Doñana National Park (reducing water availability by 50% with impacts on thousands of migratory birds).
To avoid the curse of unintended consequences, a ‘systems approach’ is needed via multi-metric assessment of sustainability and risk. Exploring the push and pull of multiple elements of risk and their knock on effects on business resilience, not just in terms of cost, but also environmental resilience, flexibility of supplier and the ability to meet the exacting social and ethical demands of the customer base, is where the current knowledge frontier lies (Figure 5).
Image may be NSFW. Clik here to view.
Figure 5 - The relationship between sustainability, risk and resilience in the fresh produce supply chain
Tackling the multi-metric assessment of risk and its relation to business need
The availability of data and handheld computers and sensors (also known as smart phones) offers a revolution in ‘big data’ and its analysis. Already, advances in social auditing have improved business culture and allowed an overspill of best practice into sustainability. Most growers and importers document water, energy and waste usage[5]. To demonstrate how these data can be utilised, we have developed a tool that combines environmental and social audit data, specific grower questionnaire data, varietal and growing specifics alongside importer level business metrics, such as profitability, reliability and product quality for each source grower. Our work has shown that we can place this wealth of data in the context of risk, using grower geo-references that select from a wide range of spatial risk maps, as well as future climate forecasts (namely precipitation levels and temperature change). Having accurate grower location data also allows the incorporation of current weather data and the frequency of business impacting weather events into the big data picture.
risk, resilience and sustainability are all corelated and should be aligned for assessing and managing food systems using a multi-metric analysis.
Conclusions
As the world changes rapidly, particularly in terms of increasing demand from finite resources and increasingly variable and extreme weather patterns, the ‘food system’ is likely to face unprecedented perturbations. If we do not make efforts to build resilience and focus only on profit and efficiency based on ‘normal conditions’, the system becomes more fragile. Our contention is that risk, resilience and sustainability are all co-related and should be aligned for assessing and managing food systems using a multi-metric analysis. Sustainability and resilience are essential properties of the sustainability and resilience of businesses themselves. Multi-metric analysis is increasingly possible in complex supply chains, estimating risk from metrics of production coupled with georeferenced data on environmental and social risks. The development of new business analytical tools can expedite business focused actions with a lasting positive impact for grower sustainability, food security and resilience alike.
