Can we develop a ‘stress test’ for national food systems?
If you follow me on Twitter (and if not, why not?!), then you will know that I have been promoting the idea of a ‘stress test’ for national food systems, most recently at the interesting IFPRI Conference on Building Resilience, held in Addis Ababa in May.
What is a stress test?
The idea of a stress test sits within the often overlapping conversations about vulnerability, risk and resilience, but I have borrowed it specifically from the debate about banking regulation. In this context, a stress test is a ‘technology’, usually managed by an independent body like a central bank or a regulatory authority, designed to examine whether banks are sufficiently capitalised and otherwise equipped to withstand specific shocks or combinations of shocks, like a rise in unemployment or a drop in equity prices. Examples can be found in the EU, the UK, the US, China, Japan and Brazil, among others. The IMFInternational Monetary Fund has a stream of work on the topic.
The UK Bank of England describes the purpose and process of its stress test as follows – with apologies for an extended quote:
‘The main purpose of the stress-testing framework is to provide a quantitative, forward-looking assessment of the capital adequacy of the UK banking system and individual institutions within it. To facilitate assessment of the system as a whole, stress tests will be carried out concurrently . . . The annual stress tests will deliver an integrated process for deliberations around bank capital adequacy, both at a system-wide and an individual-institution level. And they will provide a device through which the Bank can be held accountable to Parliament, and the wider public, on its financial stability objective, by allowing the FPC and the PRAParticipatory Rural Appraisal Board to articulate the resilience standard against which they hold the banking system. Regular stress tests should also bolster public confidence in the stability of the system, by demonstrating the range of severe, but plausible, stresses that authorities expect banks to be able to withstand.
Under the framework proposed . . . stress tests would explore a range of scenarios, including: (a) common scenarios applied across all banks taking part in the exercise — these would be designed by the FPC, in consultation with the PRAParticipatory Rural Appraisal Board; and (b) bespoke scenarios, designed by individual banks and approved by the PRAParticipatory Rural Appraisal Board.
The Bank expects that the stress-testing framework will use a suite of models to translate these scenarios into projections of bank profitability and capital ratios. This will involve modelling by Bank staff as well as by individual banks themselves. This approach seeks to guard against the risk that the stress test becomes excessively exposed to the unavoidable weaknesses of any single model. It also seeks to mitigate the risks associated with excessive reliance on banks’ own modelling: banks may face incentives to be overly optimistic about the impact of stress scenarios on their capital position to achieve a more favourable result in the stress test. But the suite of models does leave a greater role for judgement in combining model outputs to reach an overall view on capital adequacy.
. . In addition, the stress tests will produce a range of other information that is expected to inform . . . . Board judgements. This will include a qualitative assessment of banks’ own stress-testing and capital management processes and governance.
It is important that credible policy actions are taken in response to the results of the stress tests. To this end, the results of the stress-testing exercise are expected to be used to: (a) inform the FPC’s assessment of the resilience of the financial system and, in doing so, aid formulation of policy responses; and (b) support PRAParticipatory Rural Appraisal Board decisions and actions on individual banks, taking into account any system-wide actions.
Crucially, the results of the stress tests are not expected to be mechanically linked to policy responses. This is not intended to be a simple ‘pass-fail’ regime. Rather, it aims to deliver a more graduated policy framework, where the magnitude of remedial actions taken would be a function of policymakers’ judgement around the adequacy of banks’ capital plans.
. . . A key principle underpinning the proposed framework is that the outcome of, and analysis associated with, the annual stress tests should be made public. Transparency over methodologies adopted, stress-test results as well as policy and supervisory responses . . . will support the credibility of the framework, facilitate accountability of policy interventions and incentivise banks to engage fully with the exercise.’
That was a long quote, apologies again, but key words include: system-wide; independent; transparent; mixed method; judgement-based; authoritative; action-focused; and with teeth.
Probably, you don’t want to know more about bank stress tests, but in case you do, I will append a typical risk register, produced by the European Banking Authority, complete with an analysis of risk drivers and a set of traffic light risk indicators. Alternatively, read Timothy Geithner’s just-published account of the financial crisis, appropriately called ‘Stress Test’.
Why food policy stress tests?
Why do I think that stress tests would be useful in the world of food policy, and especially why do I think that they should be applied to national food systems? There are four reasons.
First, national and international policy-makers have a legitimate interest in vulnerability to shocks and in the resilience of food systems. Unwise, indeed, would be the national food minister who did not have a risk register at the heart of operations. Even more unwise, by the way, but that is another story, would be the country that did not have a food minister at all, responsible for overseeing the production, marketing and distribution of food: this is not always a job for which Ministers of Agriculture are well-suited.
Second, as with banks, it is the system-wide risks that pose the greatest threat. Shocks can be domestic, for example drought-induced; but equally, and perhaps even more frighteningly, international, as for example during the food crisis of 2008. Simultaneous shocks affecting many actors. Contagion. Large players that are ‘too big to fail’. It is not for nothing that the international agencies, including the WTO, laid such emphasis during the 2008 crisis on the dangers of ‘beggar-thy-neighbour’ protectionist policies.
Third, the food system is unlike the financial system in that severe shocks cannot be dealt with by printing money, or in this case food. Lives are at risk when food runs out, and lags in the system mean that it takes time to catch up. Supply of grains has risen sharply since 2008, in response to rising prices, and stocks have been rebuilt. We have been lucky, but might not be so again.
Fourth, there has been investment since 2008 in better information, building on the existing work of FAO, for example monitoring food markets and prices - but without quite the authority and independence, and certainly without the pressure for change, of an official regulator. For example, the G20 has established the Agricultural Market Information System, but see how carefully its objectives are worded:
‘Established at the request of the Agriculture Ministers of the G20, the Agricultural Market Information System (AMIS) is an inter-Agency Platform to enhance food market transparency and encourage coordination of policy action in response to market uncertainty. The initial focus of AMIS is on four crops that are particularly important in international food markets, namely wheat, maize, rice and soybeans.
AMIS seeks to strengthen collaboration and dialogue among main producing, exporting and importing countries. Apart from G20 members plus Spain, participants in AMIS include seven major producing, consuming and exporting countries of commodities covered by AMIS. Together, these countries represent a large share of global production, consumption and trade volumes of the targeted crops, typically in the range of 80-90 percent. In addition, AMIS reaches out to other key stakeholders in international food markets such as commodity associations and institutional investors in commodity markets.’
What, then, would a stress test look like?
This is work in progress. However, the banking model suggests a relatively straightforward approach, modelling the impact of various shocks on a dependent variable - in the case of the banks, capital adequacy – and then working through a response. That amounts to a seven step procedure:
- Identify the main features to be protected and the level of protection required;
- List the risks;
- Estimate the potential impact of changes in the risk variables;
- Identify remedial measures to achieve the desired level of protection;
- Mandate the implementation of the remedial measures;
- Monitor the impact; and
- Repeat at regular intervals.
There is plenty of material on most elements of this approach, in the food arena and more widely. The IFPRI Resilience Conference had multiple sessions on different kinds of food and nutrition shocks, including drought, conflict and international price spikes. Nineteen background briefs were prepared. There was also a session on measuring resilience. This turns out to be something the World Bank has a special interest in. I am told that ‘the World Bank is committed to developing an indicator of resilience to natural disasters by December 2015. This indicator will serve as an important and actionable tool to measure countries' progress towards achieving resilience. To facilitate this process, the Climate Change Group (CCG) and the Global Facility for Disaster Reduction and Recovery (GFDRR) of the World Bank are organizing a small workshop of researchers and practitioners working on this topic to discuss how to build on existing efforts and develop a robust indicator moving forward.’ If anyone is interested, it is in July! Perhaps they will discuss a stress test . . .
More generally, there is a literature on macroeconomic shocks and resilience, including of course to underpin the design of stress tests for banks. As I observed last year in writing about the EU’s approach to resilience, ‘There has also been an increase in work on macro-economic vulnerability and resilience. As early as 2000, the Commonwealth Secretariat published a state-level Vulnerability Index, referring especially to small states. The global financial crisis of 2008 led to more work of this kind. For example, ODI work on the crisis specifically examined vulnerability to shocks and recommended a new ‘compact for crisis-resilient growth’. Subsequent work in 2011 examined the need for and experience of macro-economic shock facilities, including IDA facilities and the EU’s own V-FLEX.’ I was then pointed, by John Burton of KPMG, to a useful IMF paper by Abdul Abiad and others, which demonstrates greatly increased resilience in developing countries (and also demonstrates why those interested in resilience need to focus on both macro-economic and household issues). I will paste in the summary at the end. 'Policy space' turns out to be a key factor.
Given this range of topics and wealth of expertise, I am hesitant to prescribe what a stress test might look like. However, there are six points to take into consideration.
(a) Setting the objective: a well-performing food system
First, the starting point is to ask what the objective or dependent variable might be. In the case of banks, the design of stress tests has focused on capital adequacy. A possible counterpart in the field of food security might be access to food or nutrition status. However, it would also be useful to ask about the characteristics of a well performing food system. That is a question Rachel Slater and I addressed in our work on the New Food Policy, back in 2003. Having reviewed different approaches, we identified 19 criteria against which a food system might be judged, including technical and allocative efficiency, but also a range of health-related and social indicators:
We concluded that
‘the very number of criteria, and their diverse character, immediately illustrate a challenge of aggregation in evaluating food systems, whether globally or locally. Some of the criteria are economic and financial, so that it might be possible to hope for a quantitative summary, using money as a numéraire. Others, however, are qualititative, and some are subjective. No single cost-benefit analysis is likely to be possible, even with heroic assumptions about valuation, weighting and time preference. As an alternative, the way forward may be to use multiple-criteria tables, as has been done before in evaluating food policy interventions. ‘
No doubt, the table could be simplified, but it captures an important point, that a food system cannot be judged simply by whether or not there is enough food available at national level to meet notional national demand. Distribution matters, and nutrition quality, and dignity, and a contribution to long term resilience. Of course, there will be trade-offs within this set. That is why a stress test has to be judgement-led.
(b) Specifying the level of protection required
Second, the design of a stress test requires not just identification of the dependent variable(s), but also the level of protection required. Think, for example, of designing a bridge. Should it be strong enough to resist a 50-year flood, a 100-year flood or a 500-year flood? In the field of food security, it might be obvious that a food system fails if anyone goes hungry or slips into a state of under-nutrition. That would certainly be consistent with implementing the right to food, and with strong statements on hunger, such as the ‘zero hunger’ challenge launched by Ban Ki Moon at the Rio Summit in 2012. However, national food systems in most countries, including developed countries, would fail that test in normal years, even without taking account of shocks. So, should there, for example, be a focus on deviation from a (hopefully improving) trend?
The literature on economic and social rights may be of some help here, with its focus on ‘progressive realisation’: ‘in the short-term, states have not a maximum, but a minimum obligation, which means providing the ‘maximum available’ resources. In practice, this means taking at least some steps, which are deliberate, concrete, targeted, and appropriate’. International duty-bearers have a role to play. Could a stress test be based on agreed progress towards food and nutrition targets, in the context of post-2015 sustainable development goals?
(c) Designing the risk register
Third, a key step in designing the stress test will be to produce a comprehensive risk register, estimating both the likelihood of risks and their impact. It is important to range widely in identifying risks, and not just to look within the confines of the food system. As the sociologist Rodolfo Stavenhagen once observed, ‘if there is no meat in the kitchen, the answer is not to be found in the kitchen’. That is why macro-economic and global geopolitical shocks need to be taken into account.
Personally, I like the approach to global risks of the World Economic Forum, in its annual risk survey. The focus is on global risk,defined as an occurrence that causes significant negative impact for several countries and industries over a time frame of up to 10 years. A key characteristic of global risks is their potential systemic nature – they have the potential to affect an entire system, as opposed to individual parts and components’. Systemic risks are characterized by: (a) modest tipping points combining indirectly to produce large failures; (b) risk-sharing or contagion, as one loss triggers a chain of others; and (c) “hysteresis”, or systems being unable to recover equilibrium after a shock.
The WEF Global Risk Report 2014 examines 31 risks in five categories, with the results summarised in the figure below. There are ten risks of greatest global concern, with food crises entering at number 8, below unemployment and inequality, but above bank failures. Note the importance of many food-related issues in the top right – hand quadrant, including water, climate change, loss of biodiversity and fiscal crises.
Would anyone like to have a go at a country-specific food security risk register along these lines? And with the model-based level of quantitative analysis recommended for bank stress tests?
(d) Making sure the analysis is systemic
Fourth, the use of a global risk register like that proposed by the WEF reminds us that the key feature of banking stress tests is their focus on systemic failure: the risk that one bank failure will trigger collapse of the entire financial system. This is certainly a problem that food system planners should be aware of, especially in the short term. Some players are very large in terms of their impact on global markets, and might be thought of as ‘too big to fail’. The US? Australia? Russia? Crop failure in any one of these can set off a crisis in global markets, as happened in the food crisis of 1972-4, largely triggered by crop failure in the Soviet Union. Crop failures also played a big part in the food crisis of 2008.
In the financial sector, the response is to carry out stress tests for all significant players at the same time. Probably, the same should happen also with national food systems.
(e) Designing a response
Fifth, it is hard to specify remedial measures without having first carried out the risk analysis, but it is easy to imagine that the list of instruments would need to cover both financial and regulatory, and that governance issues would be central.
Just as an illustration, IFPRI’s Director General, Shenggen Fan, working with Joanna Brzeska, produced a Brief for the 2020 Resilience Conference, dealing with the response to food price spikes and volatility, which is one of the ways in which food systems are subject to shocks. They concluded that there were some areas of stagnation, some areas of progress, and various pathways to building future resilience. The areas of progress and stagnation are summarised in the figure below: high marks for increasing investment in agriculture, for example, and for decreased use of trade restrictions, low marks for the elimination of distortionary price restrictions and for climate change mitigation.
More generally, Fan and Brzeska identify six pathways to building resilience: (a) promote mutually beneficial trade; (b) reduce biofuel-food competition; (c) establish and expand rational and global food grain reserves; (d) increase adoption of climate-smart technologies and practices; (e) invest in productive and cross-sectoral social safety nets; and (f) develop market-based price stabilisation.
(f) Who is the regulator?
Finally, there is a discussion to be had about who would play the role of the Federal Reserve or the European Central Bank in food system stress tests.
At country level, this could be an independent body, equivalent in status to a central bank. Alternatively, it could be something like the Climate Change Committee in the UK. This is an independent, statutory body established under the Climate Change Act 2008. Its purpose is to advise the UK Government on emissions targets and report to Parliament on progress made in reducing greenhouse gas emissions and preparing for climate change. It is an authoritative voice on the topic.
At global level, the obvious vehicle is FAO, which has a Committee on Food Security. This has an emphasis on global frameworks and policy coordination, but is still work in progress – and, frankly, it would be a surprise if an FAOFood and Agriculture Organisation of the United Nations Committee managed the same level of authority as a bank regulator. I’ve appended a description of what the Committee aims to achieve. In Phase II, it says, it will ‘develop a Global Strategic Framework for food security and nutrition in order to improve coordination and guide synchronized action by a wide range of stakeholders.’ ‘Guide synchronised action’: I wonder what that means.
A test of the CFS would be whether its actions, including its Global Strategic Framework, meet the standards of a stress test. Remember: system-wide; independent; transparent; mixed method; judgement-based; authoritative; action-focused; and with teeth. Perhaps. In which case, the CFS could own the regulatory function and manage the stress test.
Or perhaps a special purpose institution needs to be created?
Conclusion: what should happen next?
I don’t pretend that this note is definitive, in the sense of exploring all the subtelties of global and national food system risk registers. I had a different purpose, which was to reflect on whether the specific technology of a bank regulator’s stress test had any application in the field of food security. My conclusion is that it does, but that the idea needs to be piloted at national level. If that suggests the idea has mileage, then the next step would be to think about governance, both nationally and internationally. Any volunteers?
Summary of Abiad et al: The Rising Resilience of Emerging Market and Developing Economies
Our main findings can be summarized as follows. First, EMDE resilience has increased markedly over the past two decades. These economies are spending more time in expansion, and downturns and recoveries have become shallower and shorter. The performance of the past decade was particularly impressive for many EMDE regions, with emerging Europe being a notable exception. In fact, the past decade was the first time that EMDEs spent more time in expansion, and had smaller downturns, than advanced economies.
Second, various external and domestic shocks are associated with the end of EMDE expansions. Among external shocks, sudden stops in capital flows, advanced economy recessions, spikes in global uncertainty, and terms-of-trade busts all increase the likelihood that an expansion will end. Among domestic shocks, credit booms double and banking crises triple the probability that an expansion will shift into a downturn by the following year.
Third, good policies are associated with increased resilience. Specifically, greater policy space (as measured by low inflation and favorable fiscal and external positions) and improved policy frameworks (as measured by countercyclical policy, inflation targeting, and flexible exchange rate regimes) are associated with longer expansions and faster recoveries. There is no clear relation between resilience and structural characteristics , such as trade
patterns, financial openness and the composition of capital flows, and income distribution. Few of these characteristics are robustly associated with the duration of expansions and the speed of recoveries.
Fourth, the improvements in policymaking and the buildup of policy space in many EMDEs account for the bulk of the increased resilience during the past decade. Some shocks, such as spikes in global uncertainty, have become more frequent in the past decade, but many other shocks have become less frequent, such as banking crises and credit booms. Overall, fewer shocks account for about two-fifths of the improved performance in EMDEs. Greater policy space and better policy frameworks account for the remaining three-fifths.
The FAO’s Committee on Food Security
The terms of reference of the Committee are laid down in Rule XXXIII of the General Rules of the Organization (GRO). Due regard should also be made to the CFS Reform Document (CFS: 2009/2 Rev.2) incorporated in Part Q of Volume II of the Basic Texts of FAO. The terms of reference of the Committee on World Food Security are as follows:
a) Coordination at global level. Provide a platform for discussion and coordination to strengthen collaborative action among Governments, regional organizations, international organizations and agencies, NGOs, CSOs, food producers’ organizations, private sector organizations, philanthropic organizations and other relevant stakeholders, in a manner that is in alignment with each country’s specific context and needs.
b) Policy convergence. Promote greater policy convergence and coordination, including through the development of international strategies and voluntary guidelines on food security and nutrition on the basis of best practices, lessons learned from local experience, inputs received from national and regional levels, and expert advice and opinions from different stakeholders.
c) Support and advice to countries and regions. At country and/or region request, facilitate support and/or advice in the development, implementation, monitoring and evaluation of their nationally and regionally owned plans of action for the elimination of hunger, the achievement of food security and the practical application of the “Voluntary Guidelines for the Right to Food” that shall be based on the principles of participation, transparency and accountability.
In Phase II the CFS will gradually take on additional roles such as:
a) Coordination at national and regional levels. Serve as a platform to promote greater coordination and alignment of actions in the field, encourage more efficient use of resources and identify resource gaps. As the reform progresses, the CFS will build, as appropriate, on the coordination work of the United Nations High-Level Task Force (HLTF). One guiding principle to support this role will be to build on and strengthen existing structures and linkages with key partners at all levels. Key partners include national mechanisms and networks for food security and nutrition, the UN country teams and other coordination mechanisms such as the International Alliance Against Hunger (IAAH) and its National Alliances, food security thematic groups, regional intergovernmental bodies and a large number of civil society networks and private sector associations operating at the regional and national levels. In each case, the functional contributions they could make, as well as how the CFS could strengthen linkages and enhance synergy with such partners would have to be established.
b) Promote accountability and share best practices at all levels. One of the main functions of the CFS has been to “monitor actively the implementation of the 1996 World Food Summit Plan of Action” (WFS-PoA). Although countries are taking measures to address food insecurity, the specific programmes as they are presented do not necessarily help to report quantitatively on progress towards realizing the WFS-PoA objectives. The CFS should help countries and regions, as appropriate, address the questions of whether objectives are being achieved and how food insecurity and malnutrition can be reduced more quickly and effectively. This will entail developing an innovative mechanism, including the definition of common indicators, to monitor progress towards these agreed-upon objectives and actions taking into account lessons learned from previous CFS and other monitoring attempts. Comments by all CFS stakeholders will have to be taken into account and new mechanisms will build on existing structures.
c) Develop a Global Strategic Framework for food security and nutrition in order to improve coordination and guide synchronized action by a wide range of stakeholders. The Global Strategic Framework shall be flexible so that it can be adjusted as priorities change. It will build upon existing frameworks such as the UN’s Comprehensive Framework for Action (CFA), the Comprehensive Africa Agriculture Development Programme (CAADP) and the Voluntary Guidelines to Support the Progressive Realization of the Right to Adequate Food in the Context of National Food Security.