Robert L Axtell

University of Oxford | OX · Mathematical Institute

This report is a compendium of expert insights regarding opportunities for investing in science and technology to increase U.S. ability to engage in long-term competition in undergoverned spaces. This exploration marks an initial step toward developing a functional perspective on determining whether new approaches to strategy and engagement are war...

Using detailed administrative microdata for two countries, we build a modeling framework that yields new explanations for the origin of firm sizes, the firm contributions to unemployment, and the job-to-job mobility of workers between firms. Firms are organized as nodes in networks where connections represent low mobility barriers for workers. Thes...

Models of human dimensions of fisheries are important to understanding and predicting how fishing industries respond to changes in marine ecosystems and management institutions. Advances in computation have made it possible to construct agent‐based models (ABMs)—which explicitly describe the behaviour of individual people, firms or vessels in order...

Modeling is essential to characterize and explore complex societal and environmental issues in systematic and collaborative ways. Socio-environmental systems (SES) modeling integrates knowledge and perspectives into conceptual and computational tools that explicitly recognize how human decisions affect the environment. Depending on the modeling pur...

We develop an alternative theory to the aggregate matching function in which workers search for jobs through a network of firms: the labor flow network. The lack of an edge between two companies indicates the impossibility of labor flows between them due to high frictions. In equilibrium, firms’ hiring behavior correlates through the network, gener...

Sustainable management of complex human–environment systems, and the essential services they provide, remains a major challenge, felt from local to global scales. These systems are typically highly dynamic and hard to predict, particularly in the context of rapid environmental change, where novel sets of conditions drive coupled socio-economic-envi...

We develop an alternative theory to the aggregate matching function in which workers search for jobs through a network of firms: the labor flow network. The lack of an edge between two companies indicates the impossibility of labor flows between them due to high frictions. In equilibrium, firms' hiring behavior correlates through the network, gener...

Models of human dimensions of fisheries are important to understanding and predicting how fishing industries respond to changes in marine ecosystems and management institutions. Advances in computation have made it possible to construct agent-based models (ABMs)—which explicitly describe the behaviour of individual people, firms, or vessels in orde...

Succeeding under rampant software piracy is a critical task for software publishers today. Instead of treating software piracy as a destructive force, a few scholars have realized the positive influence of software piracy, that is, it may accelerate the diffusion of innovations. Furthermore, fewer scholars have tried to develop strategies to contro...

Certain elements of Hayek's work are prominent precursors to the modern field of complex adaptive systems, including his ideas on spontaneous order, his focus on market processes, his contrast between designing and gardening, and his own framing of complex systems. Conceptually, he was well ahead of his time, prescient in his formulation of novel w...

Complex systems theory and evolutionary theory hold important insight for economics, yet to date they have played a limited role in shaping modern economic theory. This chapter reviews different notions of equilibrium and explores four distinct areas relevant to the incorporation of evolutionary and complexity ideas into economics, finance, and pol...

Two widely heralded yet contested approaches to economics have emerged in recent years. One follows an older, rather neglected approach which emphasizes evolutionary theory in terms of individuals and institutions. The other emphasizes economies as complex adaptive systems. Important concepts from evolutionary theory include the distinction between...

We develop a theory of unemployment in which workers search for jobs through a network of firms, the labor flow network (LFN). The lack of an edge between two companies indicates the impossibility of labor flows between them due to high frictions. In equilibrium, firms' hiring behavior correlates through the network, modulating labor flows and gene...

The zero intelligence trader model is a well-known agent based representation of supply and demand relations for the study of the emergence of market clearing. We have parallelized it using a variety of languages and technologies, including pthreads in C, C11 threads, OpenMP in C, Clojure, Erlang, Go, Haskell, Java, Python, and Scala. For each, sig...

We construct a data-driven model of flows in graphs that captures the essential elements of the movement of workers between jobs in the companies (firms) of entire economic systems such as countries. The model is based on the observation that certain job transitions between firms are often repeated over time, showing persistent behavior, and sugges...

A model is described in which large numbers of simple agents organize into groups that empirically resemble U.S. firms. The agents work in team production environments, regularly adjust their work effort, and periodically seek better jobs or start new teams when it is in their self-interest. Nash equilibria of the team formation game exist but are...

It is conventional in labor economics to treat all workers who are seeking new jobs as belonging to a labor pool, and all firms that have job vacancies as an employer pool, and then match workers to jobs. Here we develop a new approach to study labor and firm dynamics. By combining the emerging science of networks with newly available employment mi...

Systemic risk must include the housing market, though economists have not generally focused on it. We begin construction of an agent-based model of the housing market with individual data from Washington, DC. Twenty years of success with agent-based models of mortgage prepayments give us hope that such a model could be useful. Preliminary analysis...

Agentization is the process of rendering neoclassical models into computational ones. This methodological tool can be used to analyze and test neoclassical theories under a more flexible computational framework. This paper presents agentization and its methodological framework. We propose that, by classifying the assumptions of a neoclassical model...

Supplementary materials for "The cause of universality in growth fluctuations".As supplementary materials we provide the following:In Section 1 we discuss a modification to our model incorporating a constraint over the number of elements. In Section 2 we derive and discuss in detail the logarithmic growth rate distribution arising from our model. I...

The Business Assessment Model (BAM) evaluates trajectories of People, Performance and Planet (3P) for a single firm and for the whole market system as a function of business decision of actors, exogenous events in the broader socioeconomic environment or both. BAM computes the cumulative difference between predictions of the perturbed and original...

Phenomena as diverse as breeding bird populations, the size of U.S. firms, money invested in mutual funds, the GDP of individual countries and the scientific output of universities all show unusual but remarkably similar growth fluctuations. The fluctuations display characteristic features, including double exponential scaling in the body of the di...

Phenomena as diverse as breeding bird populations, the size of U.S. firms, money invested in mutual funds, and the scientific output of universities all show unusual but remarkably similar growth fluctuations. The fluctuations display characteristic features, including heavy tails and anomalous power law scaling of the standard deviation as a funct...

Current trends in model construction in the eld of agent- based computational economics base behavior of agents on either game theoretic procedures (e.g. belief learning, c- titious play, Bayesian learning) or are inspired by articial intelligence (e.g. reinforcement learning). Evidence from experiments with human subjects puts the rst approach in...

New results on a class of distributed two-sided matching algorithms are described. In particular, these algorithms can dominate the matches produced by the Gale-Shapley algorithm in terms of both average social welfare and fairness. However, stability is not guaranteed and therefore a natural trade- off arises between stability and social welfare....

In this paper, we use a computational modeling approach to examine the long-standing social issue of tax compliance. Specifically, we design an agent-based model-the Networked Agent-Based Compliance Model (NACSM)-where taxpayers not only exist within localized social networks, but also possess heterogeneous characteristics such as perceptions about...

Kohn (The Cato Journal, 24(3):303–339, 2004) has argued that the neoclassical conception of economics—what he terms the “value paradigm”—has experienced diminishing marginal returns for some time. He suggests a new perspective is emerging—one that gives more import to economic processes and less to end states, one that bases behavior less on axioms...

A mathematical formalism is developed for the existence of unique invariants associated with wide classes of observed discounting behavior. These invariants are 'exponential discount rate spectra,' derived from the theory of completely monotone functions. Exponential discounting, the empirically important case of hyperbolic discounting, and so-call...

While macroeconomics has never denied the essentially microeconomic character of most economic activity, it is only in the past 30_40 years that the quest for microeconomic foundations for macro has assumed center stage. There appear to be myriad motivations for this quest, from logical necessity to ultimate completeness to elements of the ‘Lucas c...

Recently discovered facts concerning the size distribution of U.S. firms are recapitulated - in short, these sizes are closely approximated by the Zipf distribution, a Pareto (power law) distribution with exponent of unity. Interesting consequences of this result are then developed, having primarily to do with formulae for the distribution's moment...

Without Abstract

In this report we have summarized our research into the nature of firm growth in the U.S. over the 5 year period 1998-2003 using comprehensive data on U.S. businesses extracted from the Census database. We have analyzed these data with a particular eye to the departures from standard assumptions and results in economics generally, and within the fi...

The tendency for human beings, economic firms and social institutions to locate together has been the subject of intellectual inquiry for centuries. People and firms are not spread ubiquitously across the globe but have a powerful tendency to organize in well-defined geographic units, cities. First Auerbach (1913) and then Zipf (1941; 1949) observe...

In this paper we show how a simple modification of the Kesten's process (Kesten, 1973), taking into account common and idiosyncratic elements, allows to reconcile the aggregate-sector puzzles on firms' size distribution reported in the literature.

Laws and regulations have both benefits and costs, comparisons of which provide useful information. When such benefits and costs occur over time it is normal practice to discount them, with costs in the distant future considered less important than those occurring in the present, and distant benefits less valuable than present benefits. Conventiona...

The computational complexity of two classes of market mechanisms is compared. First the Walrasian interpretation in which prices are centrally computed by an auctioneer. Recent results on the "computational complexity" are reviewed. The "non"-"polynomial complexity" of these algorithms makes Walrasian general equilibrium an "implausible" conception...

In this research, we propose group formation that enhances the performance of multiagent systems in the Social Dilemmas. As the mechanisms for group formation, mutual choice mechanism for 2-IPD is extended to that for N-IPD. Furthermore, group split based on the metanorm of mutual choice is introduced. As a result of simulations with an evolutionar...

In this research, we propose group dynamics that promotes cooperative behavior in the so-called Social Dilemmas and enhances the performance of systems. If cooperative behavior among self-interest individuals is established, effective distribution of resources and useful allocation of tasks based on coalition formation can be realized. In order to...

In human societies diverse people act purposively with powerful but limited cognitive processes, interacting directly with one another through techno-logically-facilitated and physically-mediated social networks. Agent-based com-putational modeling takes these features of humanity—behavioral heterogeneity, bounded rationality, network interactions—...

Long House Valley in the Black Mesa area of northeastern Arizona (U.S.) was inhabited by the Kayenta Anasazi from about 1800 before Christ to about anno Domini 1300. These people were prehistoric ancestors of the modern Pueblo cultures of the Colorado Plateau. Paleoenvironmental research based on alluvial geomorphology, palynology, and dendroclimat...

Results on the formation of multi-agent teams are reviewed and extended. Conditions are specified under which it is individually rational for agents to spontaneously form coalitions in order to engage in collective action. In a cooperative setting the formation of such groups is to be expected. Here we show that in non-cooperative environments-pres...

Analyses of firm sizes have historically used data that included limited samples of small firms, data typically described by lognormal distributions. Using data on the entire population of tax-paying firms in the United States, I show here that the Zipf distribution characterizes firm sizes: the probability a firm is larger than size s is inversely...

A model of city formation and evolution is elaborated, based on a multi-agent model of endogenous firm formation. Agents have heterogeneous abilities, are boundedly rational, and interact directly with one another out of equilibrium in team production environments. Each agent works in a firm and each firm has a location. Agents periodically search...

Economists have traditionally studied aggregate behavior as the outcome of individual decisions made interactively, while sociologists have focused on the role of social influences on individual behavior. Over the past decade, however, the barriers between the disciplines have broken down, resulting in the new area of social economics. Social econo...

The many motivations for employing agent-based computation in the social sciences are reviewed. It is argued that there exist three distinct uses of agent modeling techniques. One such use — the simplest — is conceptually quite close to traditional simulation in operations research. This use arises when equations can be formulated that completely d...

The panel on Empirical versus Formal Methods was highly thought-provoking. The panel began with 10-minute presentations by the panel members. The first speaker was Doug Smith from Kestrel Institute. The main thrust of Smith’s presentation was that formal methods enable run-time matching of agent services and requirements. In particular, if agent se...

The essential idea is to show how norms can emerge spontaneously at the social level from the decentralized interactions of many individuals that cumulate over time into a set of social expectations.

The effects of distinct agent interaction and activation structures are compared and contrasted in several multi-agent models of social phenomena. Random graphs and lattices represent two limiting kinds of agent interaction networks studied, with so-called ‘small-world’ networks being an intermediate form between these two extremes. A model of reti...

We use panel data from the National Longitudinal Survey of Youth (NLSY) to examine how body weight changes with age for a cohort moving through early adulthood, to investigate how the age-obesity gradient differs with socioeconomic status (SES) and to study channels for these SES disparities. Our results show first that weight increases with age an...

A model in which heterogeneous agents form firms is described and empirically tested. Each agent has preferences for both income and leisure and provides a variable input ('effort') to production. There are increasing returns to cooperation, and agents self-organize into productive teams. Within each group the output is divided into equal shares. E...

Recent results on the computational complexity of Brouwer and Kakutani fixed points is reviewed. It is argued that the non-polynomial complexity of fixed-point algorithms makes Walrasian general equilibrium an unrealistic model of real markets. A radically more decentralized and distributed picture of markets involves repeated bilateral trade betwe...

This paper describes an innovative approach to stochastic modeling of variable annuity policyholder behavior, combining behavioral economics and agent-based computational modeling. This permits realistic rendering of behavior at a highly disaggregated level, the level of individual salespeople and annuity policyholders. The model has been implement...

An agent-based computational model of Long House Valley, in northern Arizona near Monument Valley, is described and demontrated. The model, that runs from about AD 400 to 1400, consists of artificial adaptive agents (households) who inhabit a digitized version of the Long House Valley landscape. A detailed paleoenvironmental record exists for Long...

What is anartificial society? What can such models offer the social sciences in particular? We address these general questions, drawing brief illustrations from the specific artificial society we call “Sugarscape.”

A statistical method to evaluate extreme events in earthquake precursory signals of electrical nature is proposed. The occurrence probability of seismo-electrical anomalies is computed by means of a parametric time series approach. In order to avoid the influence of external noise, a preliminary filtering of the electrical signals is performed. In...

Long-range R&D and capital investment projects are normally evaluated by means of a procedure (cost-benefit analysis) that involves a choice of time preference functions. Cost-benefit analysts and many economists typically assume that time preference is a behavioral fact of life and that the time preference function is essentially equivalent to a c...

How do social structures and group behaviors arise from the interaction of individuals? Growing Artificial Societies approaches this question with cutting-edge computer simulation techniques. Fundamental collective behaviors such as group formation, cultural transmission, combat, and trade are seen to "emerge" from the interaction of individual age...

How do social structures and group behaviors arise from the interaction of individuals? Growing Artificial Societies approaches this question with cutting-edge computer simulation techniques. Fundamental collective behaviors such as group formation, cultural transmission, combat, and trade are seen to "emerge" from the interaction of individual age...

This paper develops the concepts and methods of a process we will call "alignment of computational models" or "docking" for short. Alignment is needed to determine whether two models can produce the same results, which in turn is the basis for critical experiments and for tests of whether one model can subsume another. We illustrate our concepts an...

This paper develops the concepts and methods of a process we will call "alignment of computational models" of "docking" for short. Alignment is needed to determine whether two models can produce the same results, which in turn is the basis for critical experiments and for tests of whether one model can subsume another. We illustrate our concepts an...

Economists generally assume that systems of transferable property rights are preferable to non-market systems. This paper suggests that the design of a market-based policy that dominates a command-and-control regime is more subtle than is commonly believed, even in theory. The subtlety arises because identical approaches to monitoring and enforceme...

The ways in which exponentially increasing IT capabilities are reshaping the social sciences are briefly reviewed. Many of these changes are primarily increases in scale and scope and do not represent new methodology. However, one specific IT-facilitated development—multi-agent systems—holds out the promise of fundamentally altering the ways in whi...

This paper describes an innovative approach to stochastic modeling of variable annuity policyholder behavior, combining behavioral economics and agent- based computational modeling. This permits realistic rendering of behavior at a highly disaggregated level—the level of individual salespeople and annuity policyholders. The model has been implement...

Models in the classical theory of games and in neoclassical economics normally assume rational- ity in the sense that agents have compete and transitive preferences. The paper labels this fundamental rationality and distinguishs two other sorts of rationality pertinent to the study of strategic interaction: individual economic rationality (IER) and...

Zipf's law for cities is one of the most striking regularities in the social sciences. It states that, for a given region or country, a plot of log(city size) versus log(city size rank), yields a linear line having slope exactly -1. This is equivalent to saying that city sizes are Pareto-distributed with an exponent of 1. This statistical regularit...

A model of retirement decision-making is described in which are relatively small number of agents are rational. Such agents behave as if they compute the optimal age at which to retire. A small proportion of agents retire at random. A majority of the population engages in imitative behavior. An imitative agent retires once a certain fraction of the...

Thesis (Ph. D.)--Carnegie-Mellon University, 1992. Photocopy.