INTRODUCTION: DYNAMIC STOCHASTIC GENERAL EQUILIBRIUM MODELLING AND THE STUDY OF ASIA‐PACIFIC ECONOMIES

Abstract

It is no exaggeration to state that dynamic stochastic general equilibrium (DSGE) modelling has become the dominant approach in quantitative macroeconomics. Yet it is often misunderstood as being rigid and narrow in its method and applications. Some of this stems from the roots of the DSGE approach in real business cycle (RBC) models epitomized by the path-breaking work of Kydland and Prescott (1982). Over the years, both the basic RBC modelling framework and the calibration approach to quantifying the model's predictions expanded to include, for instance, new Keynesian frameworks with nominal rigidities and more formal econometric methods, such as Bayesian estimation. DSGE modelling now embraces many strands and perspectives of modern macroeconomic reasoning. What is common is a shared modelling philosophy captured by the components of the DSGE acronym. DSGE models are explicitly dynamic. Economic agents in a DSGE world, households, firms, government institutions, formulate plans about the future. Crucially, they take into account the evolution of economic state variables, such as capital, money or wealth, so that decisions embody intertemporal trade-offs. DSGE models are (mainly, although not necessarily) stochastic. They recognize the fact that economic actors operate in an environment of uncertainty, that there are foreseen and unforeseen external disturbances, and that agents take this uncertainty into account when formulating expectations about the future and setting their plans. DSGE modelling requires the specification of the stochastic environment, typically by making assumptions about the nature of the exogenous stochastic processes. DSGE models also follow an equilibrium approach. In its most basic sense, this means that things have to add up, that budget and social resource constraints have to be obeyed, and that prices and quantities are jointly determined. This does not necessarily mean that in a DSGE model an economy has to be exhaustively modelled or that for particular questions some variables, such as interest rates, cannot be taken as exogenous. It only imposes the requirement that within the chosen framework, decisions are internally consistent with preferences, technologies and budget sets. This special issue of the Pacific Economic Review collects several papers that show the breadth and depth of the DSGE modelling practice. The focus of this issue is on the study of economies in the Asia-Pacific region, running the gamut from small city states, such as Hong Kong, to large highly developed countries, such as Japan. It is demonstrated that the DSGE approach can be applied to different economies (ancient China, Japan, Taiwan, Hong Kong and New Zealand) as well as different issues (including the rise and fall of nations, fiscal policy, the labour market and the exchange rate system). By necessity, this is not an exhaustive list. We hope that this special issue will stimulate further research on the Asia-Pacific economies with a distinct DSGE flavour. The paper by Kenneth Chan and Jean-Pierre Laffargue on ‘Foreign threats, technological progress and the rise and decline of imperial China’ serves as a powerful illustration of the reach of DSGE modelling, even into economic history. The paper starts from the famous Needham question; namely, why scientific progress in China suddenly seemed to stop while the counterpart of Europe took off, and quickly surpassed China. Many answers have been proposed, often veering into the realm of cultural and political differences. Chan and Laffargue cast a more prosaic eye on this question. They start from the observation that Imperial China had to contend with numerous external threats that could only be averted at considerable expense. At the same time, the Emperor had to satisfy the demands of a vast bureaucracy and an Imperial Court, which they label the Emperor's consumption. The Emperor's choice is thus a simple trade-off between consumption and investment. Consumption keeps the Court happy, but removes resources that could be invested in infrastructure and technology that are needed to fend off China's enemies. Chan and Laffargue overlay this problem with a simple dynamic political economy framework and use long-run historical data to calibrate the parameters of their model. Using numerical solution techniques, they find that the model admits two equilibria conditional on the level of foreign threat. If the latter is relatively low, then an agrarian economy emerges, whereas under high threat technological advances follow. This fascinating study thus provides an economic rationale for the saying of Heraclitus that ‘War is the father of all things’. Chen and Laffargue effectively propose a theory for the rise and fall of nations, which demonstrates how far DSGE modelling has come. Moving ahead several centuries, and firmly planted in a highly developed and technologically advanced economy, the paper by Yasuo Hirose and Takushi Kurozumi, ‘Do Investment-specific Technological Changes Matter for Business Fluctuations? Evidence from Japan’, sheds light on why growth in Japan has slowed over the past two decades. Previous studies, such as Hayashi and Prescott (2002), emphasize a slowdown of productivity growth as an explanation for the prolonged recession of the Japanese economy. Hirose and Kurozumi develop a fully-specified monetary DSGE model with various real and nominal rigidities and several exogenous shocks as the driving forces. They estimate their DSGE model using Bayesian techniques, instead of simply calibrating it as so many earlier studies have done. They focus their discussion on the relative contribution of productivity and investment-adjustment cost shocks in determining the growth path in Japan. Hirose and Kurozumi find that the estimated investment adjustment cost shock plays an important role and, moreover, that it strongly correlates with an index of firms' financial position. They relate their finding to the asset price crash in the 1990s, when the ‘lost decade’ began. Thus, instead of depending on an assumed exogenous slowdown in productivity, Hirose and Kurozumi make a first step to use the financial market conditions to provide a ‘micro-foundation’ for the productivity slowdown. The contribution by Guay Lim and Paul McNelis, ‘Macroeconomic Volatility and Counterfactual Inflation-targeting in Hong Kong’, shifts the focus to a highly developed small open economy, Hong Kong. The authors study the implications of its fixed exchange regime, specifically a currency board, with the US dollar on domestic variables. They first build a small open-economy DSGE model, which they estimate using Bayesian methods. The key finding is that exogenous movements in export demand and world inflation are the drivers of Hong Kong's GDP. They then conduct a counterfactual experiment, which assumes a flexible exchange rate regime with a monetary policy of inflation targeting. The model is then simulated for this alternative regime, but the previous estimated parameters and stochastic structure. Lim and McNelis find that the welfare gain from switching to an inflation targeting regime and flexible exchange rates is marginal. The main effect on domestic variables is more stable inflation, but much higher interest-rate volatility. Lim and McNelis thus demonstrate the usefulness of a DSGE approach in conducting this type of counterfactual policy experiment, where there is a clear notion of a loss function and isolation from Lucas-critique type concerns. In ‘Aggregate Labour Market Dynamics in Hong Kong’, Thomas Lubik also studies the Hong Kong economy, but he focuses exclusively on the labour market. He builds a small-scale DSGE model in which the labour market is characterized by search and matching frictions. Using data on unemployment and vacancy postings, he estimates the model using Bayesian methods. The model describes labour market dynamics remarkably well, arguably better than the same model does for US data. The estimation finds that Hong Kong is characterized by a low degree of worker bargaining power, but a high replacement ratio. Notably, Lubik finds a low degree of churning, or labour turnover, in normal times. However, recessions are characterized by rapid firings, with rapid hiring in expansions. Overall, Lubik confirms the public perception of the Hong Kong labour market as highly flexible and adaptable. Lubik's contribution demonstrates how a DSGE approach can be used to analyze a subsector of the economy and how the estimation can inform the discussion of historical episodes. The next paper in this special issue, ‘An Estimated Small Open Economy Model with Frictional Unemployment’ also deals with the labour market in a DSGE model. Julien Albertini, Gunes Kamber and Michael Kirker develop a small open-economy DSGE model and estimate it using New Zealand data and Bayesian methods. One of the main contributions of this paper is that it extends the modelling of search and matching frictions in the labour market to an economy that trades and borrows internationally. The striking finding is, however, that the labour market appears almost divorced from disturbances in the rest of the economy, that it evolves virtually autonomously. This is reminiscent of earlier findings in the literature (e.g. Krause and Lubik, 2007), and it creates a challenge for the identification of the monetary transmission mechanism. The paper also deserves special mention as it presents an example of how a DSGE approach can be used to analyze the transmission of shocks, policy and otherwise, in a framework that is usable for monetary policy analysis. Finally, in ‘Ramsey Taxes Meet Price Rigidity’, Wing Leong Teo and C. C. Yang study fiscal policy in a DSGE model. The paper goes farthest of the contributions to this issue in modelling the government as a decision-making entity. The authors work with a DSGE model with sticky prices in which the government (i.e. the fiscal authority) provides minimal services. On top of that, it has to choose an optimal financing scheme for this expenditure by levying Ramsey taxes; that is, taxes which maximize the expected value of the household's utility. The key additional assumption is that there is heterogeneity in terms of price rigidity across goods: some prices change frequently, others not so much. Teo and Yang show that Ramsey taxation should be designed to remedy the relative price distortion caused by this heterogeneity, specifically by imposing low tax rates on more sticky prices. The authors calibrate their model to US and Taiwanese data and demonstrate that a shift to Ramsey taxation in the presence of heterogeneity in price stickiness would result in a substantial welfare gain.