System dynamics model of inflation

Effect of money on output is an empirically observed phenomenon. Yet, the theoretical foundations and implications of this phenomenon is a debate in economics. In this study, System Dynamics Model of Inflation is developed in order to explain these foundations and implications based on a disequilibrium economic framework. System dynamics is a modeling and simulation methodology which is suitable for analyzing the dynamic behavior of complex systems. In system dynamics models, systems are defined as the structural components and their interrelations. These interrelations are defined on time dimension, thus the whole system may show non-linear and unintuitive behavior. Historically, system dynamics methodology is applied in various areas, from biological systems to social systems. Economics is also a suitable area to apply system dynamics principles for modeling, because of the complexity of the economic system structure, and the counter-intuitive behavioral observations in the system. Although there are many studies in the literature about the usage of this methodology for economic modeling, a complete system dynamics model could not be found which aims to explain the theoretical nature of the empirical relation between monetary and real economic variables. There are several empirical observations related to the scope of this study. First, a monetary shock appears to affect economic activity and unemployment rate. This effect sustains and accumulates for more than a year. Secondly, there is a negative relation between unemployment rate and inflation rates. Finally, inflation is persistent to monetary shocks and there is a phase lag for inflation in the business cycle. These behavior patterns are all related and show that money has different effects on the real economic indicators. Due to its ability to explain dynamic behavior of complex systems, system dynamics methodology is suitable to explore the structural reasons of these empirical regularities. However, the currently mainstream new-consensus model is not applicable in system dynamics models. Mainstream economic view is based on rational expectations and strict equilibrium assumptions, which conflicts the fundamentals of the system dynamics methodology. Theoretical knowledge for a suitable framework can be found in the disequilibrium theories of Keynesian and Post-Keynesian schools of thought. A remarkable example is Phillips Machine, a hydraulic analog computer built by Bill Phillips in 1949, in London School of Economics. It simulates a set of differential equations of a Keynesian disequilibrium economic model. The beginning of this study is dedicated to Phillips Machine, because it is an impressive example of using system dynamics methodology in economic modeling, and it is closely related with the purpose of this study. In this study, this machine is argued to be the first system dynamics model in the history of economics. The study follows with a discussion chapter about different economic views and incompatibilities on the concept of equilibrium. For the first time, this historical debate is entitled as 'Time Controversy'. This discussion chapter is followed by introductory chapter, dedicated to introduce the system dynamics methodology. After that, literature review of inflation is given from a historical perspective. Finally, description of the proposed model is given, model behavior is analyzed, and validation tests are conducted. System Dynamics Model of Inflation includes six sectors, or sub-models. Money sector explains the determination of demand-side by the circular flow of income. In the goods sector, supply-side is formulated and both sides are linked in real terms. In the price sector, pricing behavior of firms is explained. In the labor sector, employment dynamics are captured. In the wage sector, wage determination is explained by the feedback from labor sector. Finally, in the investment, investment decisions are described. The proposed model is tested for two types of monetary shock, which are named as money shock and inflation shock. The simulation results show that, the model is able to generate the related behavior patterns similar to empirical observations. Accordingly, unemployment rate is inversely related with wage inflation, price inflation, unexpected inflation and real wage level. The simulated scatter diagrams share some properties with the empirical observations. Moreover, the model is able to generate the persistent behavior of inflation, and the phase lag between inflation and growth rate. Depending on the parameter set, the model may generate limit cycles. These cycles share the behavioral properties of the conventional business cycles. The amplitude of this limit cycle is around 5% and the duration of cycles is approximately 4 years, for reasonable parameters. For building confidence on the proposed model, some validation tests are applied. In the structure confirmation test, each equation in the model is evaluated for its relevance to empirical and theoretical knowledge. In the parameter confirmation test, model parameters are questioned for being empirically reasonable and theoretically consistent. In the direct extreme-condition test, model equations are hypothetically forced to extreme conditions, in order to compare the analytical results with the real life anticipations for these extreme conditions. In the dimensional consistency test, the units of the variables are questioned to be meaningful and consistent with each other. In the integration error test, the sensitivity of model behavior to the simulation step is considered and the accuracy of the integration process is questioned for validation. In the extreme condition test, model behavior is analyzed for extreme condition scenarios, which are not likely to observe in real life but theoretically possible. In the parameter sensitivity test, the sensitivity of model behavior to different parameter sets are analyzed. Finally, in the behavioral validation test, the simulation results are discussed to be consistent with empirical observations. Model behavior shows that, labor market is the main source of economic instability. Likewise, sensitivity analysis show that, the behavior is significantly sensitive for some parameters, and the cyclic behavior is mainly the result of labor market conditions. Specifically, when wage rates adjusts easily, the overall economic system become more unstable. As a result, governmental controls and regulations for wage determination are argued to be a reasonable policy for economic stability. The proposed model in this study is the first system dynamics economic model which aims to explain the effect of money on output, and is the main contribution of this study. Other than that, the theoretical implications of the model behavior are also important for researchers and policy makers. Finally, the theoretical discussions about equilibrium and time controversy are scientifically valuable for the theoretical research in economics and system dynamics.