Contenu du sommaire : Agent-based models and economic policy

This paper argues that the path followed by modern macroeconomic theory excludes the analysis of major endogenous movements in macroeconomic variables. Rather than persist with models based on the idea that the economy behaves like a rational individual we should build models of the economy as a complex system of interacting agents. Profiting from the advances made in computer science we can now build agent based or computational models which allow us to model the out of equilibrium behaviour of such systems. They allow us to remove many of the restrictive assumptions of standard models and to incorporate the heterogeneity of economic agents and the evolution of the network that governs the interactions between the individuals and firms in the economy. Such models can help fill the theoretical void with which policymakers declare that they have been faced in the current crisis.

This paper emphasizes the importance of considering the mechanisms that coordinate economic transactions in a decentralized economy, namely the role played by a self-organizing network of entrepreneurial trading firms, for theories aimed at guiding macroeconomic policy. We review a research program that aims to understand how, and how well, trading activities are coordinated in various circumstances by employing agent-based computational (ACE) models of stylized economies where these activities take place in a self-organizing network of markets created and operated by profit-seeking business firms. We discuss how such a research program can yield important policy-relevant insights, beyond those that can be offered by conventional dynamic stochastic general equilibrium (DSGE) models, into several macroeconomic phenomena including the emergence of monetary equilibria in a decentralized economy, the microfoundations of the multiplier process, the costs of a higher trend rate of inflation, and the role of the banking system in economic crises.

The Great Recession seems to be a natural experiment for macroeconomics showing the inadequacy of the predominant theoretical framework—the New Neoclassical Synthesis—grounded on the DSGE model. In this paper, we present a critical discussion of the theoretical, empirical and political-economy pitfalls of the DSGE-based approach to policy analysis. We suggest that a more fruitful research avenue to pursue is to explore alternative theoretical paradigms, which can escape the strong theoretical requirements of neoclassical models (e.g., equilibrium, rationality, representative agent, etc.). We briefly introduce one of the most successful alternative research projects—known in the literature as agent-based computational economics (ACE)—and we present the way it has been applied to policy analysis issues. We then provide a survey of agent-based models addressing macroeconomic policy issues. Finally, we conclude by discussing the methodological status of ACE, as well as the (many) problems it raises.

The restrictive assumptions imposed by the traditional methods of aggregation prevented so far a sound analysis of complex system of feedback between microeconomic variables and macroeconomic outcomes. This issue seems to be crucial in macroeconomic modelling, in particular for the analysis of financial fragility, as conceived in the Keynesian and New Keynesian literature. In the present paper a statistical mechanics aggregation method is applied to a financial fragility model. The result is a consistent representation of the economic system that considers the heterogeneity of firms, their interactive behaviour and the feedback effects between micro, meso and macro level. In this approach, the impact of micro financial variables can be analytically assessed. The whole dynamics is described by a system of dynamic equations that well mimics the evolution of a numerically solved agent based model with the same features.

Maximum entropy predictions are made for the Kirman ant model as well as the Abrams-Strogatz model of language competition, also known as the voter model. In both cases the maximum entropy methodology provides good predictions of the limiting distribution of states, as was already the case for the Schelling model of segregation. As an additional contribution, the analysis of the models reveals the key role played by relative entropy and the model in controlling the time horizon of the prediction.

This paper presents a model of asymmetric (S,s) pricing. We investigate implications of such a behavior for the effectiveness of the monetary policy. We discuss two types of asymmetric responses to monetary interventions. One is the symmetry in the responses to positive and negative monetary shocks. The other is the variance in responses to monetary shocks during booms and recessions. The conclusion is that first type of asymmetry can be attributed to the asymmetry in adjustment bands, while the second kind of asymmetry is a result of firm heterogeneity, and asymmetry of (S,s) bands does not contribute to it.

Basel III is a recently-agreed regulatory standard for bank capital adequacy with focus on the macroprudential dimension of banking regulation, i.e., the system-wide implications of banks' lending and risk. An important Basel III provision is to reduce procyclicality of present banking regulation and promote countercyclical capital buffers for banks. The Eurace agent-based macroeconomic model and simulator has been recently showed to be able to reproduce a credit-fueled boom-bust dynamics where excessive bank leverages, while benefitting in the short term, have destabilizing effects in the medium-long term. In this paper we employ the Eurace model to test regulatory policies providing time varying capital requirements for banks, based on mechanisms that enforce banks to build up or release capital buffers, according to the overall conditions of the economy. As conditioning variables for these dynamic policies, both the unemployment rate and the aggregate credit growth have been considered. Results show that the dynamic regulation of capital requirements is generally more successful than fixed tight capital requirements in stabilizing the economy and improving the macroeconomic performance.

Using the “Keynes+Schumpeter” (K+S) agent-based model developed in Dosi et al. (2010, 2012) we study how the interplay between firms' investment behavior and income distribution shapes the short—and long-run dynamics of the economy at the aggregate level. We study the dynamics of investment under two different scenarios. One in which investment is fully determined by past profits, and one in which investment is tied to expectations about future consumption demand. We show that, independently from the investment scenario analyzed, the emergence of steady growth with low unemployment requires a balance in the income distribution between profits and wages. If this is not the case, the economy gets locked either in stagnation equilibria, or into growth trajectories displaying high volatility and unemployment rates. Moreover, in the demand-led scenario we show the emergence of a non-linear relation between real wages and unemployment. Finally, we study whether increasing degrees of wage-flexibility are able to restore growth and unemployment and reduce the volatility in the economy. We show that this is indeed the case only when investment is profit-led. In contrast, in the scenario where investment is driven by demand expectations wage-flexibility has no effect on either growth and unemployment. In turn, this result casts doubts on the ability of wage-flexibility policies to stabilize the economy.

This paper describes the out-of-equilibrium approach to the analysis of economic processes. We argue that such an approach is adapted to study qualitative (or structural) changes, like technical progress or changes in preferences. Truly sequential analyses manage to capture the essential features of qualitative change. In particular, we show how this approach shifts the focus from the issue of optimality to the one of viability of the processes of change. The objective of the paper is, first, to highlight the analytical elements of an out-of-equilibrium approach, so as to serve as a guide for the construction of this type of models; second to show, how this analysis allows to see controversial phenomena, like for example the debate on wage rigidity or the productivity paradox in a new and different light ; third to identify the real causes of the on-going crisis.

We propose an agent-based computational model defining the following dimensions of structural change—organisation of production, technology of production, and product on the supply side, and income distribution and consumption patterns on the demand side—at the microeconomic level. We define ten different parameters to account for these five dimensions of structural change. Building on existing results we use a full factorial experimental design (DOE) to analyse the size and significance the effect of these parameters on output growth. We identify the aspects of structural change that have the strongest impact. We study the direct and indirect effects of the factors of structural change, and focus on the role of the interactions among the different factors and different aspects of structural change. We find that some aspects of structural change—income distribution, changes to production technology and the emergence of new sectors—play a major role on output growth, while others—consumption shares, preferences, and the quality of goods—play a rather minor role. Second, these major factors can radically modify the growth of an economy even when all other aspects experience no structural change. Third, different aspects of structural change strongly interact: the effect of a factor that influences a particular aspect of structural change varies radically for different degrees of structural change in other aspects. These results on the different aspects of structural change provide a number of insights on why regions starting from a similar level of output and with initial small differences grow so differently through time.

Long term economic development is characterized not only by increasing efficiency of economic activities but also by qualitative change within industries and increasing variety concerning the existence of different industries. Traditional economic growth models do not cope with the complex amalgam of these three trajectories of economic development nor could comprise the interactions among them. Furthermore, economic development is not a process which is spurred by supply-side effects but driven by the co-evolutionary interplay of supply and demand side forces. With our TEVECON model we analyze economic development driven by efficiency and quality improvements together with structural change and the co-evolution between innovation and demand. The first part of the paper introduces to the basic model and some general results. The second part of the paper deals with policy experiments which are undertaken by comparing different numerically analyzed scenarios.

Environmental innovations heavily depend on government policies and consumers' behaviour. This paper addresses the issue of how these two factors interact in shaping the transition to a green technology. We extend models of technological selection with heterogeneous agents and learning by including a weak hierarchy between green and polluting goods. For general distributions of agents' income and the explicit inclusion of a carbon tax, the model is not analytically tractable so we derive our results using numerical simulations. Given the level of income, carbon taxes are more effective when technological improvements brought about by wealthy pioneer consumers suffice in inducing the remaining population to buy the green good. In this case, a negative relationship between income inequality and tax effectiveness emerges. Taxes on polluting production have a regressive effect since they are mainly paid by poorer people who consume less of the green good. For these people, a negative wealth effect strongly contrasts the standard substitution effect of the tax. Finally, both lower inequality and taxes have the expected effect for intermediate levels of the learning parameter.

In this paper, we build a realistic large-scale agent-based model of the Italian electricity market and run simulations to investigate how a significant increase in wind capacity can affect electricity prices at the national level when the wind resource is geographically concentrated, as in the case of Italy. The simulator implements both cost-based and oligopoly models in which electricity companies learn to bid strategically. We compare a scenario based on the 2010 wind supply and a scenario based on the maximum potential wind capacity as estimated in technical reports. Results confirm the beneficial effect of low-cost renewable energy in reducing average market prices, but simulated power flows in the grid suggest that congestion in the electricity network induced by high wind penetration creates market power opportunities that can offset the price reduction effects.