Social and Economic Change as a Complex Dynamical System  (SEC) Session 2

Abstract: We develop a behavioral model for housing prices with heterogeneous expectations, with fundamental buying prices linked to housing rental levels to fundamental buying prices. Using quarterly data we estimate the model parameters for eight different countries, US, UK, NL, JP, CH, ES, SE and BE. We find that the data support heterogeneity in expectations, with temporary endogenous switching between fundamental mean-reverting and trend-following beliefs based on their relative performance. For all countries we identify temporary, long lasting house price bubbles, amplified by trend extrapolation, and crashes reinforced by mean-reverting expectations. The qualitative predictions of such non-linear models are very different from standard linear benchmarks, with important policy implications. The fundamental price becomes unstable, e.g. when the interest rate is set too low or mortgage tax deductions too high, giving rise to multiple non-fundamental equilibria and/or global instability. We also discuss estimation of similar non-linear switching models to other time series, such as stock? prices, commodity prices, exchange rates and inflation. Close

Abstract: The Netherlands has generally been a laggard in tackling the energy transition, to the extent that a judge Dutch State is acting unlawfully by not contributing its proportional share to preventing a global warming. A 2015?report (in Dutch)?co-authored with the?Netherlands Scientific Council for Government Policy (WRR) argues that a contributing cause to this state of affairs is from policy makers looking at the energy system in too much isolation. The report is a practical case study how taking a complex systems lens to the problem, may open the path to different and innovative approaches to Dutch energy transition policy. Close

Abstract: Network-based metrics are commonly applied in a wide range of real-world problems, such as ranking, recommendation, and information spreading. Classical methods inspired by physical processes (like diffusion) often neglect the temporal order of interactions and, as a consequence, turn out to be highly ineffective when applied to systems evolving in time which is particularly worrying as most real systems fall in this category. To devise improved and well founded metrics, it is critical to understand the nature of the shortcomings of classical methods in evolving systems. In this presentation, we focus on the problems of ranking and recommendation in growing complex networks. For both problems, we show that by understanding the temporal patterns of the studied systems and unveiling the consequent lack of performance of classical metrics, we are able to design time-dependent methods that significantly outperform their static counterparts in singling out the most valuable items in the system. In the case of ranking, we use growing network models to show that PageRank centrality metrics systematically exhibit a temporal bias towards old or recent nodes depending on the temporal scales of node aging. We introduce a rescaled score which is built on the PageRank score and at the same time it is not biased by node age. We use real and model data and show that the rescaled score allows us to identify high-quality nodes earlier than with other metrics. In the case of recommendation, we use data from Netflix, Yelp and Digg to show that classical methods tend to favor old nodes and, as a consequence, systematically fail to identify recent items that will be collected by users in the future. We design a new time-aware method built on network growth patterns and show that it markedly outperforms its static counterpart. The findings presented here support the idea that time-aware modifications of existing metrics can lead to improved results in finding the most valuable information in diverse real systems. Close

Abstract: In animal collective behavior, interactions between individuals and between an agent and the environment are usually seen to be ?through information?propagation. Cooperations can emerge with or without information sharing.? The information propagation in ants, fish and birds seems to be the?crucial?mechanism that regulates the global behavior of the crowd. This talk will introduce the research related to?information?functioning in collective?groups?and several recent works that demonstrate interesting collective patterns with information sharing playing pivotal roles. Close

Abstract: In this work we analyze the short- and long-run effects of fiscal austerity policies, employing an agent-based?model populated by heterogeneous, boundedly-rational firms and banks. The model, in line with the family of?"Keynes+Schumpeter" formalism, is able to account for a wide array of macro and micro empirical regularities.?In particular, it endogenously generates self-sustained growth patterns together with persistent economic?fluctuations punctuated by deep downturns. On the policy side, we find that austerity policies considerably?harm the economy, by increasing output volatility, unemployment, and the incidence of crises. In addition, they?depress innovation and the diffusion of new technologies, thus reducing long-run productivity and GDP growth.?Finally, we show that "discipline-guided" fiscal rules are self-defeating, as they do not stabilize public finances,?but, on the contrary, they disrupt them. Close

Abstract: We use citation data of scientific articles produced by individual nations in different scientific domains to build a bipartite country - scientific domains network to determine the structure and efficiency of national research systems. We characterize the scientific fitness of each nation ?that is, the competitiveness of its research system?and the complexity of each scientific domain by means of a non-linear iterative algorithm able to assess quantitatively the advantage of scientific diversification. We find that technological leading nations, beyond having the largest production of scientific papers and the largest number of citations, do not specialize in a few scientific domains. Rather, they diversify as much as possible their research system. On the other side, less developed nations are competitive only in scientific domains where also many other nations are present. Diversification thus represents the key element that correlates with scientific and technological competitiveness. A remarkable implication of this structure of the scientific competition is that the scientific domains playing the role of ?markers? of national scientific competitiveness are those not necessarily of high technological requirements, but rather addressing the most ??sophisticated?? needs of the society. We complement this analysis with a correlation study between the scientific impact of a nation with a normalized measure of RD funds and the level of internationalization. Close

Abstract: Systemic risk in financial markets arises largely through the interconnectedness of agents through financial contracts. We show that the systemic risk level of every agent in the system can be quantified by simple network measures. With actual central bank data for Austria and Mexico we are able to compute the expected systemic losses of the economy, a number that allows us to estimate the true cost of a financial crises. We can further show with real data that it is possible to compute the systemic risk contribution of every single financial transaction to the financial system. Based on these findings, we suggest a smart financial transaction tax that taxes the systemic risk contribution of individual transactions. This tax provides an incentive for market participants to trade financial assets in a way that effectively restructures financial networks so that contagion events become impossible.?It is possible to show the existence of a systemically risk-free equilibrium under this smart tax. More intuitively, with the help of an agent based model we can demonstrate that the Systemic Risk Tax practically eliminates the network-driven systemic risk in a system. Close

Abstract: We introduce a model of innovation in which products are composed of distinct components and new components are adopted one at a time. We show that the number of products we can make now gives a distorted view of the number we can make in the future: simple products get over-represented, and complex products under- represented. By applying this at the component level, we derive a strategy for making far-sighted innovation choices that increase the rate of innovation. We apply our strategy to real data from three different sectors?language, gastronomy and mobile technology.? Close