Complexity Pier Paolo Saviotti •ECIS (Eindhoven Technological University), •INRA GAEL, GREDEG CNRS, Department of Economics, Hohenheim University Prepared for the meeting: Advances in Economic Dynamics and Development: Economics and Complexity – to be held in Curitiba on 11-13 November 2013 Complexity Development_Curitiba_11-13 November 2013 Vs evolutionary theories Evolutionary theories related to • many different research traditions/disciplines including complexity • probably undergoing increasing fragmentation (Robert, Yoguel) Here I will opt for a path which • (a) will draw from several but not all research traditions mentioned in RY • (b) is related to my (and co-authors’) work on long run economic development • (c) will necessarily neglect some of the possible contributing research traditions Complexity Development_Curitiba_11-13 November 2013 Complexity economics vs equilibrium economics (Arthur, 2013) Complexity Economics (a) different from Equilibrium Economics (b) • (b) cannot understand certain types of phenomena that (a) can understand • Are (a) and (b) complementary? Or can (a) provide a different interpretation of the same phenomena that (b) understood in particular way? Answer (B.A.):Complexity Economics (a) more general than Equilibrium Economics (b) • The economy is not necessarily in equilibrium • Economic agents (firms, consumers, investors) constantly change their actions and strategies in response to the outcome they mutually create • This further changes the outcome, which requires them to adjust afresh. Complexity Development_Curitiba_11-13 November 2013 Complexity Economics vs Equilibrium Economics (2) Equilibrium economics • • • • order determinacy deduction stasis Complexity economics • • • • contingency indeterminacy sense-making openness to change Complexity Development_Curitiba_11-13 November 2013 Patterns • …interacting elements → overall patterns & overall patterns cause the interacting elements to change or adapt (feedback loops, positive or negative) Interacting elements Complexity Development_Curitiba_11-13 November 2013 Patterns Main concepts • • • • • • • • • Dynamics Equilibrium Interactions Uncertainty Emergence Qualitative change Discontinuities Transformation Creativity Complexity Development_Curitiba_11-13 November 2013 General equilibrium Equilibrium vs innovation Innovation Systems destroying continuously (general) in movement equilibrium (Arthur) (Schumpeter) Restless capitalism (Metcalfe) Endogenous General innovations equilibrium as destroy an attractor general (Nelson, equilibrium 1995) (Kaldor, 1972) Complexity Development_Curitiba_11-13 November 2013 General equilibrium as an attractor General equilibrium An innovation The new equilibrium position • only as an attractor • present but not necessarily visited by the economic system • creates a new (potential) equilibrium position • at a distance from the present state of the system • can drive the system towards it • but at a finite speed Complexity Development_Curitiba_11-13 November 2013 Two processes • An equilibrium can be attained only if • Rs ˃ RI • General equilibrium good approximation only if as an innovation creates a new (potential) equilibrium position the system moves instantaneously towards it, when: • RS ˃˃ RI • Most of the times system not in equilibrium Innovation (attractor) Movement of the equilibrium position (Ri) movement of the system towards the equilibrium position (Rs) Further innovations (growth) Further movement of the equilibrium position (Ri) Further movement of the system towards the equilibrium position (Rs) Maturity, Saturation Slowing down of innovation (d(dInn))/dt ˂ 0) Complexity Development_Curitiba_11-13 November 2013 Is this state really stable? System, equilibrium, innovation S1 = state of system at t Movement of System occurring at RS Movement of Innovation occurring at RI When the innovatin emerges it creates an attractor which drives the system towards itself at a rate Rs . If Rs ˃˃ RI the system will eventually, at a finite rate, reach the attractor and rest there (equilibrium) Complexity Development_Curitiba_11-13 November 2013 Examples • Railways • • Cars + petroleum industry • ITC + Internet etc Complexity Development_Curitiba_11-13 November 2013 Life cycles Birth, novelty Growth Maturity Complexity Development_Curitiba_11-13 November 2013 Saturation Life cycles Emergence of a radical • System driven at finite rate towards the new attractor innovation introduced • Most of the times system out of equilibrium by a Schumpeterian entrepreneur → new • Imitation bandwagon, complementary industries, infrastructures, institutions = co-evolution sector (attractor) Birth, growth, maturity, saturation • At saturation attractor attained? Equilibrium? • But, what if saturation induces new innovation(s) which can give rise to new sectors? • One or more new attractors likely to be established and so on Complexity Development_Curitiba_11-13 November 2013 Order, change and chaos Economic systems are generally not completely random or chaotic They have order and structure We need to explain both the stability Type of order and of some structures structure changes for considerably in time (structural long periods of change, time and their transformations, changes, transitions) sometimes discontinuous and disruptive Complexity Development_Curitiba_11-13 November 2013 General equilibrium vs change General equilibrium • freezes the system and creates no incentives for any agents to change Change in these conditions • can only occur if there are exogenous shocks and even then: Only if • the rate of movement of the system towards the equilibrium position (Rb) is (much) greater than the rate of movement of the equilibrium position (Ra) Complexity Development_Curitiba_11-13 November 2013 Order & steady states Disruption of existing order • Emergence of innovation • some incumbent activities and institutions disappear Growth, maturity • appropriate institutions created, subsequent evolution occurs smoothly, with ‘good’ values • Giving rise to steady states, • characterized by constant set of institutions, organizations and interactions Saturation • Increasing ‘rigidity’ of the economic system • Inducement to innovate out of the system • Innovation, fluctuations arise and invade the system Complexity Development_Curitiba_11-13 November 2013 Structural change Structure • components and interactions • system as network Structural change • emergence of new components/nodes or disappearance of older ones • new interactions (links) • new forms or levels of aggregation Complexity Development_Curitiba_11-13 November 2013 "weak emergence" and "strong emergence". Weak emergence • A type of emergence in which the emergent property is reducible to its individual constituents • Example: phase transition Strong emergence • The emergent property is irreducible to its individual constituents • Strong emergence → Qualitative change → Creativity → (Radical) uncertainty (Knight) Complexity Development_Curitiba_11-13 November 2013 Levels of aggregation Micro (minimum) to Macro (maximum) Complex systems can have many meso levels of aggregation Social systems with the same micro and macro states • +, many possible intermediate (meso) ones • But, only some intermediate levels of aggregation are meaningful • 15 people are an aggregate but they are meaningful only if they constitute some, organizational form (a firm, a cooperative etc) • Macromolecules, cells, organs in a biological organism • Firms, sectors, schools, hospitals, cities, regions, countries governments etc • can behave differently if • they have different meso states Complexity Development_Curitiba_11-13 November 2013 Biological, social Different ontology or different micro levels? • Human beings are biological organisms but • Do their biological properties have an impact on their social or economic behaviour? Biologically rooted necessities but • Transition to imaginary worlds Complexity Development_Curitiba_11-13 November 2013 Biological, social (2) Biological analogies in evolutionary economics Generalised Darwinism • Variation, selection inheritance, replicators, interactors, fitness etc (Hogdson, Knudsen) both Biological organisms and social systems • are complex systems, to be explained based on complexity Complexity Development_Curitiba_11-13 November 2013 Differentiation • Necessary condition for the long run continuation of economic development • Growing variety • Complementary with growing efficiency • Distribution of differentiation amongst different countries • See Imbs, Wacziarg, 2003; Hidalgo et al, 2007, 2010; Frenken et al 2007; Saviotti, Pyka, 2004, 2008, 2013; Saviotti, Frenken, 2008) Complexity Development_Curitiba_11-13 November 2013 Differentiation Differentiation as • Necessary condition for the long run continuation of economic development • Growing variety, complementary with growing efficiency • Distribution of differentiation amongst different countries Variety and Complexity • Variety (i): N° distinguishable components • Complexity: N° distinguishable components + Interactions/linkages (constraint, degrees of freedom) Complexity • Max N° linkages → rigid system (uninteresting) • Min N° linkages (= 0) → trivial system, no interactions • Intermediate degrees of freedom : interesting complex systems with evolutionary potential Complexity Development_Curitiba_11-13 November 2013 Variety vs complexity When variety increases • more components but are there more interactions? • In a dynamical system the rate of creation of new components/nodes is not necessarily equal to the rate of creation of new links it does not necessarily rise or fall all the time • See network density, connectivity Complexity Development_Curitiba_11-13 November 2013 Variety vs complexity (2) Countries need to differentiate to grow • The probability that a country starts to export a new (to the country) product is proportional to • the proximity (similarity) of the new to the already exported products, or • related variety more important in the short run DCs vs LDCs • Developed countries export more differentiated and more complex products • Less developed countries export less differentiated and less complex products Complexity Development_Curitiba_11-13 November 2013 Network density 35 16 30 14 12 Centralization Index Centralization index 25 R2 = 0, 9198 20 15 10 R2 = 0, 8969 10 8 6 4 5 2 0 0 1988 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Years Years 16 14 Centralization Index 12 R2 = 0, 8969 10 8 6 4 2 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Years Evolution of network density for (a) 1st biotechnology generation (top left); 2nd biotechnology generation (top right); the two biotechnology generations combined (bottom) Complexity Development_Curitiba_11-13 November 2013 1999 .055 .05 .045 density .06 .065 Knowledge networks density in biotechnology 1980 1985 1990 1995 2000 Year Dynamics of Network Density for Biotechnology (Krafft, Quatraro, Saviotti) Complexity Development_Curitiba_11-13 November 2013 Creation of increasing variety • • • • Open and closed systems (Prigogine etc) Equilibrium vs non-equilibrium Dissipative systems Auto-catalysis and the emergence of new structures • Co-evolution Complexity Development_Curitiba_11-13 November 2013 Open vs closed systems Closed systems Open systems • No flow of energy, matter and information across system boundaries • Equilibrium • Maximum disorder • Flow of energy, matter and information across system boundaries • Steady states • Emergence of structure • changes of structure • But near equilibrium linrear region Complexity Development_Curitiba_11-13 November 2013 Distance from equilibrium Distance from equilibrium • Measured by the rate of flow through the boundaries of the system • Equilibrium corresponds to zero rates of flow through the boundaries of the system • Near equilibrium behaviour similar to equilibrium Emergence of new structures • beyond a given threshold distance the behaviour can change radically and • give rise to qualitative change, emergence of new structures • Equilibrium structures (crystals, structure and order ) vs Dissipative structures (waste, dissipation) Carnot vs Darwin • ..life is incompatible with Boltzmann’s order principle but not with the kind of behaviour that can occur in far-from equilibrium conditions (Prigogine, Stengers) Complexity Development_Curitiba_11-13 November 2013 Fluctuations In an equilibrium system • they tend to die out In a dissipative system • they can increase in amplitude and invade the whole system leading to a transformation Innovation • as an endogenous fluctuation leading to the transformation/emergence of new subsystems Complexity Development_Curitiba_11-13 November 2013 Open systems, structure and Differentiation • Bifurcations, transitions lead to the emergence of new structures subsystems • They occur only out of equilibrium • The number of possible states of the system can increase • Conditions leading to bifurcations, transitions? Complexity Development_Curitiba_11-13 November 2013 Co-evolution Definition Effect Coevolution of • Interactions and feedback between the components of the system • Can accelerate (positive feedback) or slow down (negative feedback) the changes of the system • Can facilitates the emergence of new subsystems • Vs cumulative causation (Robert, Yoguel) • Technologies and institutions • Technologies and infrastructures • Complementary industries, activities Complexity Development_Curitiba_11-13 November 2013 Auto catalysis and Co-evolution • Auto catalysis: occurs in a process in which one of the outputs is also one of the inputs A B AC D • Co-evolution acts in a similar way by exploiting the positive feedback between different components of the system • Can accelerate transitions towards new states/Phases/compone nts Complexity Development_Curitiba_11-13 November 2013 Examples of co-evolution Cars Rules, police tc Roads Petroche micals Service stations Fuel, oil refining Complexity Development_Curitiba_11-13 November 2013 Transition LQ→ HQ • (LQ HQ) transition can occur through a virtuous circle, but not under all circumstances. See for example effect of kqual • Wages as costs vs purchasing power Falling rate of population growth Increasing Quality & Diff Increasing education Higher product price Disposable income, demand Increasing Competencies Higher wages (Complementary) Trajectories Trajectory 1: The efficiency of productive processes increases during the course of economic development. • Efficiency = the ratio of the inputs used to the output produced, when the type of output remains constant. Trajectory 2: The output variety of the economic system increases in the course of time. • Here such variety is measured by the number of distinguishable sectors • A sector is defined as the set of firms producing a common although highly differentiated output • Arrow of time Trajectory 3: The output quality and internal differentiation of existing sectors increases in the course of time after their creation. • This means that if during the period of observation the type of output changes what we will observe is a combination of growing productive efficiency and of quality change Qualitative change, discontinuities (Strong) emergence • requires qualitative change and discontinuities Variety can only increase • if sectors are distinguishable = qualitatively different Emergence of new sectors • (Strong) emergence Complexity Development_Curitiba_11-13 November 2013 Differentiation by the emergence of new sectors Dynamics and competition • Sector dynamics: entry (Entrepreneurs + bandwagon of imitators) and exit (failure + M&A) of firms • Competition: from temporary monopoly to bandwagon of imitators (rising intensity of competition) to the circular flow • Competition: both inter- and intra-sector Emergence of new sectors development •Entrepreneurs •Temporary monopoly entry •bandwagon of imitators •Increasing ICi •failure •M&A Saturation Inducement new sector •Creation of new niche if technology available Equilibrium and development Closed system • moves towards equilibrium and rests there • with fluctuations around the equilibrium position Open system • reaches a given configuration which may be stable for a period of time but • fluctuations can increase in intensity, spread over the system and induce a transition Complexity Development_Curitiba_11-13 November 2013 Stability Closed system Open system • Equilibriums is stable unless • an exogenous shock opens up the system • Are steady states in an open system stable? Is the Schumpeterian circular flow stable? • Apparent stability : Constant number of distinguishable components of the system, steady rates of growth, no undesirable properties (unemployment, indebtedness etc) • But their size can vary, undesirable properties can grow → Fluctuations can induce transitions Complexity Development_Curitiba_11-13 November 2013 Stability (2) Modern (capitalist) economic systems The long run evolution of economic systems There can be a potentially large) number of short run situations in which qualitative change is almost absent • are open systems and • most of the time they are out of equilibrium • and the changes in composition that accompanied it • can only be interpreted by considering them open systems out of equilibrium • and in which the assumptions of equilibrium economics • can be an adequate approximation Complexity Development_Curitiba_11-13 November 2013 Stability and long run change Capitalist economic systems Equilibrium Open, dissipative, out of equilibrium systems • have high creativity by innovation but • they are intrinsically unstable and prone to crises • Restless capitalism (Metcalfe) • is not the appropriate conceptual framework • to understand their long run dynamics • are a more appropriate intellectual framework, • of which equilibrium economics would be subset Complexity Development_Curitiba_11-13 November 2013