Societies and economies are complex systems, but the theories used to inform economic policies predominantly neglect complexity. They assume for example representative agents such as a typical consumers, and they also assume that the future is risky rather than uncertain. This assumption allows for the application of the probability calculus and a whole series of other techniques based on it.
In risk situations, all potential outcomes of a policy can be known. This is not the case in situations of uncertainty, but human beings, policy makers included, cannot escape having to take their decisions and having to act facing an uncertain future. The argument is one of logic. Human beings cannot know now what will be discovered in the future. Future discoveries may however impact and shape the consequences of their current decisions and actions. Therefore, they are unable to come up with an exhaustive list of potential outcomes of a policy decision or action.
Properly taking into account the complexity of the economy and the uncertainty of the future implies a paradigm shift in economics. That paradigm does not need to be developed from scratch. It builds on modern complexity science, neo-Austrian economics (in particular Hayek and von Mises), as well as the work of Keynes and Knight and certain strands of cognitive psychology (for example, Kahneman 2011). There is no room here to elaborate on the theory and the claim that it entails a paradigm shift. Rather, I will discuss the implications for economic policy that follow from this paradigm.
This starts with the recognition that the future cannot be predicted in detail. We should be modest about what can be achieved with economic policy. This is the “modesty principle”. Economic policy cannot deliver specific targets for economic growth, income distribution, inflation, the increase of the average temperature in four decades from now, etc. Economic policy makers would be wise to stop pretending that they can deliver what they cannot. This insight implies that many current policies should be discontinued. To mention just one example: inflation targeting by central banks does not pass the modesty test.
This principle also implies refraining from detailed economic forecasts as a basis for policy making and execution. Policies should not be made on the assumption that we know the value of certain variables which we cannot know. An example here is the income multiplier in relation to changes in fiscal policy. The modesty principle also flashes red for risk-based regulation and supervision.
What economic policy can do is contribute to the formation and evolution of a fit economic order, and avoid doing harm to such an order, what I would call the “do no harm principle”, and be as little as possible a source of uncertainty for private economic agents.
Order is a central concept in the alternative paradigm, replacing the (dis)equilibrium concept in mainstream economics. An order is the set of possible general outcomes (patterns, like growth, inflation, cyclicality, etc.) emerging from purposefully acting and interacting individuals on the basis of a set of rules in a wide sense (laws, ethics, conventions…), together called a regime. Economics can analyse the connection between changes in regime and changes in economic order. Economic policy can influence the economic order through changing the regime.
However, this knowledge is not certain. There is always the potential for surprises (positive and negative; opportunities and threats) and unintended consequences. Policy can therefore not be designed first and then just be executed as designed. Policy making and execution have to evolve in a process of constant monitoring and adaptation. This would also allow for evolutionary change. An economic order that is not allowed to evolve may lose its fitness and may suddenly collapse or enter a crisis (as described by Scheffer for critical transitions in society). This mechanism may have played a role in the Great Moderation leading up to the financial crisis of 2007/2008 and in the crisis of fully funded pension systems. It is also a warning against basing sustainability policies on precise temperature targets decades in the unknowable future.
Fitness of an order has five dimensions. The first is an order in which agents are acting as described in the previous paragraph – policymaking involves a process of constant monitoring and adaptation. In addition to that, fitness is determined by alertness of agents (the ability to detect mistakes and opportunities); their resilience (the ability to survive and recover from mistakes and negative surprises); adaptive capacity (the ability to adjust); and creative capacity (the ability to imagine and shape the future). Policies may be directed at facilitating economic agents to improve these capacities, although constrained by the “modesty” and “do no harm” principles. Note that the concept of stability does not appear in the definition of fitness. This marks a difference with current policies which put much emphasis on stability.
In its own actions the government should be transparent and predictable. The best way to do that seems to be to follow simple rules. For example in fiscal policy, balance the budget, perhaps with clearly-defined, limited room for automatic stabilisers to work.
This alternative paradigm places highlights some methods and analytical techniques, including narrative techniques), network analysis), evolutionary logic), qualitative scenario thinking, non-linear dynamics (Scheffer), historical analysis (development of complex systems is path dependent) and (reverse) stress testing.
Economic policies developed along these lines help people to live their lives as they wish. They are good policies for good lives.
The OECD is organising a Workshop on Complexity and Policy, 29-30 September, OECD HQ, Paris, along with the European Commission and INET. Watch the webcast: 29/09 morning; 29/09 afternoon; 30/09 morning
The city is humanity’s greatest invention. An artificial ecosystem that enables millions of people to live in close proximity and to collaborate in the creation of new forms of value. While cities were invented many millennia ago, their economic importance has increased dramatically since the Industrial Revolution until they now account for the major fraction of the global economy. All human life is there and so the study of cities crosses boundaries among economics, finance, engineering, ecology, sociology, anthropology, and, well, almost all forms of knowledge. Yet, while we have great knowledge in each of these domains individually, we have little scientific knowledge of how they come together in the overall system of systems that is a city. In brief: How does a city work?
Such knowledge would be helpful in the coming decades. In the last sixty to seventy years, globalisation has spread the Industrial Revolution ever more widely, creating in cities new opportunities that attract hundreds of millions of internal and international migrants. This process is lifting many of these migrants out of deep poverty, while causing cities from London to Nairobi to struggle in differing ways with the unending influx.
Further, cities are responsible for large fractions of greenhouse gas emissions, for the consumption of natural resources such as water and air, and the resulting discharges of pollution into the environment. If the battle against climate change is to be won, it will be won in cities. Cities are also the principal centres for innovation and economic development, both of which are needed to continue lifting migrants out of poverty.
While the roots of urban planning can be traced back more than three thousand years in terms of the master plans of cities, it was the tremendous growth of cities in the late 19th century that transformed that field into considering the many services and affordances that are required for urban dwellers. But urban planning emerged mainly from the humanities and works primarily through extensive case studies, although it has adopted many digital tools. The notion of the city as an object of scientific study is more recent and still in its infancy, triggered in part by developments in complexity theory such as network theory, scaling laws, and systems science, and the growing availability of urban data.
Urban scaling laws have been explored at least since the early 20th century, when cities were found seen to be an example of Zipf’s law. In this case Zipf’s Law states that “for most countries, the number of cities with population greater than S is proportional to 1/S”. The understanding of scaling was greatly expanded in recent years by the works of West and Bettencourt and Batty. Their work showed that many properties of cities such as the number or lengths of roadways, the numbers of amenities such as restaurants, and so forth follow scaling laws over population ranges from ten thousand to tens of millions. Moreover these scaling laws have exponents in the ranges 0.85 to 1.15 that show large cities to be more productive, innovative, efficient in energy consumption, expensive, but also better paying than small cities. Likewise negative attributes such as crime, disease, and pollution also scale superlinearly, that is they don’t rise in strict proportion to the increase in city size. For example, GDP is proportional to the Size (S) of a city raised to a power that is slightly greater than 1, thus S1.15, while other attributes like energy consumption per capita scale sublinearly, at S0.85. Network laws also describe well the evolution over long time scales of roadways and railways in cities.
While scaling laws and network laws have great descriptive power, opinions vary on whether they apply across different countries or have predictive power. That is, the scaling of attributes is a snapshot of frequency versus size at a given time. If a city grows and “moves up the scale”, it may not achieve, in the short term, all of the positive benefits and negative impacts described. Nor do the laws provide explanations for the observed behaviours. Nonetheless, this is an important area for planners and developers seeing their cities growing or shrinking.
As urban data has become more pervasive, it is now possible to study cities as complex systems of interactions. We may view the city as a myriad of interactions among its inhabitants, its infrastructures and affordances, its natural environment, and its public, private, and civic organisations. Some of these interactions involve the exchange of goods or services for money, but many of them involve the exchange or transmission of information, enabling inhabitants and organisations to make choices. Public transportation is often studied in this way, revealing for example that small and medium sized cities evolve networks enabling commuting between small numbers of residential and business districts, while very large cities, such as London, have much richer networks that permit greater flexibility in where people live and work.
The operation of cities is also modelled using synthetic populations of software agents that represent the distribution of behaviours or preferences of much larger, real populations. Such agent-based models, with agents representing patterns of origin, destination, travel times, and modality preferences, are used to examine the overall impact of new services such as London’s Crossrail.
As the Internet of Things provides greater visibility into how inhabitants choose to exploit the opportunities offered by a given city, we may hope to discover abstract principles about how cities work. We may envision being able to construct agent-based models representing the complete spectrum of choices a city’s inhabitants make at timescales from minutes to years and spatial scales from meters to kilometers. Equally, given the increasing availability of real-time information, we might hope one day to understand the effective use of a city’s services in terms of a Nash Equilibrium, a game theory concept (often used to describe poker games), where no player can gain anything by changing their chosen strategy if other players don’t change theirs – all the players’ strategies are optimal. These are far in the future, but the EC’s Global Systems Science programme is the beginning of that journey.
The OECD is organising a Workshop on Complexity and Policy, 29-30 September, OECD HQ, Paris, along with the European Commission and INET. Watch the webcast: 29/09 morning; 29/09 afternoon; 30/09 morning
Fifteen years after 9/11, the world is now facing the threat of systemic terrorism. Apparently mindless, random attacks are in fact part of a strategy developed over a number of years, whose origins can be traced back to three major turning points, one ideological, one political, one military, that occurred at the end of the 1970s.
Traditionally, terrorism was the work of relatively small groups with clearly identifiable political or ideological goals, ranging from national liberation to animal rights. It was used as a bargaining counter to attain a clearly defined objective such as the freeing of prisoners or the withdrawal of the army from an occupied zone, or for vengeance. Of course, there were campaigns designed simply to destabilise the political climate, but these were the minority. The terrorism of Al-Qaeda represented a radical break from this, in that its aim was sustained opposition to the entire “Western” economic, cultural, and belief system, with no negotiable end to their campaigns, and whole populations seen as legitimate targets. Attacks, and the possibility of attacks, are supposed to change enemy policy by means other than the traditional method of battlefield superiority. One of their aims is to convince public opinion that the price for supporting a particular policy is too high, as well attracting support from potential sympathisers following retaliation for the initial attack.
Conflicts are fought worldwide in a complex arena across the whole spectrum of political, social, economic, and military networks, and involve a mix of national, international, transnational, and subnational actors, motivated not only by politics or ideology, but also profit. This grey area, combining aspects of traditional warfare with organised crime, is a major aspect of 21st century terrorism. But the major way in which terrorism has evolved beyond the Al-Qaida model is the strategy of Daesh to create a state by conquering and holding territory, using traditional military confrontation in some cases, and isolated attacks far from its main bases in others. The roots of this strategy can be found in the combination of the three events mentioned in the introduction.
In November 1977, Egypt’s President Sadat travelled to Jerusalem to prove his willingness to sign a peace deal with Israel. This marked the end of pan-Arabism as a viable ideology. Sadat also broke with the USSR and encouraged the rise of the Muslim brotherhood to counter the influence of the left, especially in the universities. Shortly afterwards, the Shah of Iran was overthrown by a popular uprising that the Islamists came to dominate, eventually creating an Islamic State. Then the Soviets invaded Afghanistan, but were defeated by a Western-backed coalition that the Taliban came to dominate.
Pan-Arabism was promoted by Nasser, and the intellectual origins of today’s Islamist terrorism can be traced back to the writings of one of his opponents, Sayyid Qutb, an Egyptian intellectual jailed by the regime, although his books were not banned. For Qutb, the world is living in a state of ignorance and idolatry, Jahiliyyah, a term normally reserved for pre-Islam Arabia. This includes even those who claim to be Muslims, but who are in fact apostates and thus legitimate targets: “this is not Islam and they are not Muslims”. The evil is due to the fact that men have denied God one of his attributes, Hakemeyya, divine sovereignty. Muslim scholars are scandalised by the claim that man can deprive God of anything, but Qutb’s position is echoed by Daesh’s sinister black flag, where the “Mohammed-Messenger-Allah” you would expect is replaced by “Allah-Messenger-Mohammed”, if read in the usual top-to-bottom order.
Daesh are also influenced by Qutb’s idea that divine sovereignty will be restored by a self-proclaimed elite, and that the declaration of faith is not enough to define someone as Muslim, and must be completed by jihad. The practical manual for bringing about this new, truly Islamic state, was written by Daesh in the mid-2000s. The Management of Savagery: The Most Critical Stage through Which the Umma Will Pass sets out the thinking behind the terrorist campaigns we’re seeing just now. The idea is to create such chaos, by whatever means necessary, that the jihadi are seen as the only group capable of restoring and maintaining order, similar to the initial support for the Taliban regime from Afghanis exhausted by the corruption and incompetence of the warlords.
Terrorism is one part of this strategy and Daesh have learned at least one lesson from the totalitarian regimes in Europe before and after the Second World War, namely that terror succeeds best when it is accepted on its own terms by its enemies. It’s not possible to physically terrorise everybody, but if everybody thinks they could be the next random victim, that is just as efficient.
Another major strand of Daesh’s approach is finance. The recommendations of The Management of Savagery for winning people over emphasize: “Uniting the hearts of the world’s people by means of money”. The financial power of Daesh is another significant difference with previous terrorist organisations, with some estimates putting its annual turnover at around USD 2 billion. It obtains its income through extortion, theft, and the black market – the same means described in a 2010 Rand Corporation report into Al-Qaida’s finances for the US Office of the Secretary of Defense.
Black market sales of oil probably remain Daesh’s main source of income, but as this dries up due to the success of the military forces opposing them, they will turn to other means. (Al Shabab in Somalia for instance controlled the sugar trade). Whatever it is, corruption will still be the “enabling technology” that enables the terrorists to operate. Two of the 9/11 hijackers allegedly obtained fraudulent driver’s licences from a branch of Virginia’s Division of Motor Vehicles which they used as identity cards to board the aircraft. The same branch had also sold licenses to illegal immigrants in exchange for bribes. “Nigerian troops were denied weapons to fight Boko Haram and thousands of lives were lost because of rampant fraud in the procurement process”, Nigerian President Muhammadu Buhari declared when a corrupt multi-billion dollar deal for weapons and equipment was revealed in the press in November 2015. The deal has not materialised, leaving troops without proper equipment to fight terrorist groups.
The OECD published work on the economic consequences of terrorism as long ago as 2002, and since then has examined regional, sectoral and broader aspects of the issues, for example terrorism and conflict over resources in West Africa, the implications for the transport industry, and how to help fragile states. In an analysis published earlier this year, Terrorism, corruption and the criminal exploitation of natural resources, the OECD argues that since terrorism is a multidimensional challenge, tackling it efficiently requires integrating social, economic, and political factors into the security analysis and response. Speaking personally, I would add that since the aims of Daesh include destroying democracy and dividing society along religious grounds, we should not do this for them in the name of the “war against terror”.
A session at the 2016 OECD Forum entitled “Teaching & Learning with Robots” brought Nao, a humanoid robot, to meet with a class of young students from the Sections Internationales de Sèvres (SIS) school. Catherine Potter-Jadas, head of the primary school, noted the children’s reactions to the robot.
Educators will take comfort from views such as: “For the moment, the robot can’t replace teachers because, in a country like France, children are too immature, and because they need a real human to control them. A robot wouldn’t have the authority.” In fact, most of the children saw the robot as useful assistant, rather than a substitute: “I think it’s great that a robot helps children in schools. They’ll find it interesting and become more open” said one, echoed by a classmate who thought that “Nao could be really helpful to education.” And of course some were more attracted to the entertainment value: “When the presenter said ‘Nao can carry anything lighter than a wooden spoon,’ he flexed his muscles! I liked the way the robot laughed and showed his muscles, it made me think of an odd little creature.”
Robots first made their appearance in industry, starting in the automobile sector in the 1960s. For decades, industrial robots were bulky and expensive. They were operated from stationary posts inside the workshop, and they carried out a small number of repetitive tasks, sometimes dangerous ones like soldering and cutting metal. With improvements in technologies, a second generation of robots was born. Less bulky and expensive, more autonomous, adaptable and cooperative, these robots are programmable and can be used by workers without any specific qualifications. They can also play new roles in services, health (surgical operations), education, training, commercial information, services to the elderly…
The children’s comments and questions raised a number of issues about the evolving role of technology in society and the economy, and how to equip people to take advantage of the profound impacts digital technologies will have on all of us. The majority of concerns related to robots are based on loss of jobs in developed economies. That said, there could be a “relocation” of low-skilled jobs to countries that have robots.
It is quite clear that challenges to the development of robots remain, in particular in the areas of perception, specific object recognition in a visually cluttered environment, object manipulation, and cognition. But smarter and more autonomous robots will soon be a reality thanks to improvements in a number of areas, including computational performance, electromechanical design tools and computer numerical control machines, storage of electrical energy and energy efficiency of power electronics, availability and quality of local digital (wireless) communication, scale and effectiveness of the Internet, and data storage capacities and their computational power.
In the commercial and industrial sector, beyond the improvements in reliability of manufacturing processes, robots have already shortened delays in the manufacturing of finished products, which allows for greater reactivity to detailed variations in demand. The market for personal domestic robots is growing from year to year (20% a year), while the prices should drop in the near future.
But even if people with no particular training will be able to use the next-generation of robots, those who have not mastered ICTs will find themselves more and more limited in their access to many basic services, to rewarding jobs, and to opportunities to improve their skills through training. Without ignoring the problems that arise from the disappearance of certain jobs and the serious repercussions on people and society, we must acknowledge that these innovations are full of opportunities for productivity development; they could create the new jobs of tomorrow.
It stands to reason that workers who acquire the skills necessary to adapt to changes in their line of work will be less vulnerable to replacement by digital technology. The innovations sparked by digital technologies could also present the potential for development and management of social improvements, in areas like public administration, health, education, and research. The creation of huge amounts of data and the capacity to extract knowledge and information from this data (known as big data) will launch a new wave of innovation, the creation of new services, the emergence of new products and markets…
Employee skills management will become vital in order for companies to adapt to rapid technological change, with support from complementary public investment in, for example, education and training. Primary and secondary schools will be responsible for preparing young people for an interconnected world where they will live with people of different origins and cultures, an undeniably “globalised” world.
The children from SIS didn’t sound at all worried by the prospects of more and better robots, quite the contrary: “I found it quite amazing and fabulous that technology is able to do such things. The people who built Nao must be very proud of their invention.”
Rough waters for container shipping. Why Hanjin, the world’s seventh largest container line, went under
Sad news. After months – even years – of pain and suffering, the South Korean container shipping company Hanjin finally sank and passed away. Not just any casualty, but the largest shipping bankruptcy in history: Hanjin was the world’s seventh biggest container line with a fleet of 90 ships. Was this an accident, an isolated case of bad luck, or is something more structural going on?
Like with any bereavement, there are the immediate arrangements to make. Terminal operators and maritime service providers were not paid for their services and need their money, so they have seized Hanjin ships in ports to have some sort of guarantee. Hanjin’s clients are eager to know that their goods will be delivered and not be stuck on ships. Competitors are circling around the deceased to pick up some of the ships that Hanjin leaves behind.
At the same time, people are starting to wonder how all this could have happened. Forensic analysts talk about the sluggish demand for container transport, hit by declining trade from China, the overcapacity in container shipping and the resulting low ocean freight rates that have made it very difficult to make profits in container shipping. All this sounds very logical, but also pretty abstract, and – more fundamentally – it obscures an uncomfortable truth: this was not an accident, but market forces at play – and it will happen again.
The story starts – in a way – in a corporate boardroom in Copenhagen in 2010. Then, the world’s largest container shipping company, Maersk Line, decided to order a set of new container ships that were larger than the world had ever seen, able to carry 18,000 standard containers. Putting more containers on a more fuel-efficient ship would save costs and thus give it a better position in a very competitive market.
For a weekly container service between Asia and Europe – the route on which the largest ships are deployed – ten to eleven ships are needed; a lot of capital that smaller companies would not be able to collect. As the order for the new mega-ships was placed while the global economic crisis was still unfolding, banks were unwilling to lend much to a risky business like shipping, especially the smaller ones with high risk profiles. Timing was excellent, with ship prices low due to overcapacity in shipbuilding yards. The new mega-ships were smartly marketed as “Triple E” ships, providing economies of scale, energy efficiency and environmental performance. They also provided a once in a lifetime opportunity “for the market consolidation that big players hoped for“.
Yet things worked out differently: other firms reacted by ordering similar mega-ships and by organising themselves in alliances. They agreed to share slots on each other’s vessels, which means they can offer networks and connections that they would not be able to offer if they would go it alone. Alliances had existed before, but the Triple E-strategy involuntarily resulted in stronger alliances in which more carriers were involved. These consortia were also used to share newly acquired mega-ships, so individual carriers would only need to buy a few of these, instead of having to shoulder a whole set of ten ships. Consequently, many carriers were able to rapidly catch up and also order mega-ships, many more than expected. The alliances became such powerful mechanisms that even the largest companies found themselves forced to find alliance partners.
This gave a different twist to the play, but with a similar outcome. The combined mega-ship orders in a period of sluggish demand created a sensational amount of overcapacity: way more ships than were needed. This overcapacity resulted in lower freight rates, lower revenues and several years of losses, which we have not started to see the end of yet. Whoever has the longest breath and biggest pockets will survive; the others won’t and will suffer death by overcapacity, like Hanjin.
There will very likely be more Hanjins. Hardly any container shipping line is making profit nowadays and the perspectives are bleak. Sputtering trade growth and gigantic ship overcapacity will continue to depress ocean freight rates. Banks, creditors and governments might well get impatient with some of the liners and cut life lines again.
Economic theory champions the notion of “creative destruction”, in which inefficient firms are replaced by more efficient ones. So, even if it is hardly any comfort for employees that lose their jobs in the process, one could consider it a natural thing that weaker shipping firms disappear.
There is just one problem. If this process continues, it will soon lead to a very small group of powerful carriers dominating an already concentrated market, enabling them to put a lot of pressure on clients and ports. We are starting to see what the results of this are: less choice, less service and fewer connections for shippers, the clients of shipping lines. The ports that accepted the offer they could not refuse and invested in becoming mega ship-ready may find out that they placed their fate in the hands of a few big players who frequently change loyalties at fast as the wind.
Hanjin is gone; the problem is still very much there.
The impact of mega-ships Olaf Merk on OECD Insights
The Hanjin case is a practical illustration of the complexity of sectors such as international shipping. The OECD is organising a Workshop on Complexity and Policy, 29-30 September, OECD HQ, Paris, along with the European Commission and INET. Watch the webcast: 29/09 morning; 29/09 afternoon; 30/09 morning