Dynamic stochastic general equilibrium

Structure of DSGE models

Furthermore, as their name indicates, DSGE models are dynamic, studying how the economy evolves over time. They are also stochastic, taking into account the fact that the economy is affected by random shocks such as technological change, fluctuations in the price of oil, or changes in macroeconomic policy-making. This contrasts with the static models studied in Walrasian general equilibrium theory, applied general equilibrium models and some computable general equilibrium models.

For a coherent description of the macroeconomy, DSGE models must spell out the following economic 'ingredients'.

Preferences: the objectives of the agents in the economy must be specified. For example, households might be assumed to maximize a utility function over consumption and labor effort. Firms might be assumed to maximize profits.

Technology: the productive capacity of the agents in the economy must be specified. For example, firms might be assumed to have a production function, specifying the amount of goods produced, depending on the amount of labor, capital and other inputs they employ. Technological constraints on agents' decisions might also include costs of adjusting their capital stocks, their employment relations, or the prices of their products.

Institutional framework: the institutional constraints governing economic interactions must be specified. In many DSGE models, this might just mean that agents must obey some exogenously imposed budget constraints, and that prices are assumed to adjust until markets clear. It might also mean specifying the rules of monetary and fiscal policy, or even how policy rules and budget constraints change depending on a political process.

Traditional macroeconometric forecasting models used by central banks in the 1970s, and even today, estimated the dynamic correlations between prices and quantities in different sectors of the economy, and often included thousands of variables. Since DSGE models start from microeconomic principles of constrained decision-making, instead of just taking as given observed correlations, they are technically more difficult to solve and analyze. Therefore, they usually abstract from so many sectoral details, and include far fewer variables: just a few variables in theoretical DSGE papers, or on the order of a hundred variables in the experimental DSGE forecasting models now being constructed by central banks. What DSGE models give up in sectoral detail, they attempt to make up in logical consistency.

By specifying preferences (what the agents want), technology (what the agents can produce), and institutions (the way they interact), it is possible (in principle, though challenging in practice) to solve the DSGE model to predict what is actually produced, traded, and consumed, and how these variables evolve over time in response to various shocks. In principle, it is also possible to make predictions about the effects of changing the institutional framework.

In contrast, as Robert Lucas pointed out, such a prediction is unlikely to be valid in traditional macroeconometric forecasting models, since those models are based on observed past correlations between macroeconomic variables. These correlations can be expected to change when new policies are introduced, invalidating predictions based on past observations.

Given the difficulty of constructing accurate DSGE models, most central banks still rely on traditional macroeconometric models for short-term forecasting. However, the effects of alternative policies are increasingly studied using DSGE methods. Since DSGE models are constructed on the basis of assumptions about agents' preferences, it is possible to ask whether the policies considered are Pareto optimal, which is a state of allocation of resources in which it is impossible to make any one individual better off without making at least one individual worse off., or how well they satisfy some other social welfare criterion derived from preferences (Woodford, 2003, p. 12).

Schools of DSGE modeling

At present two competing schools of thought form the bulk of DSGE modeling.

Real business cycle (RBC) theory builds on the neoclassical growth model, under the assumption of flexible prices, to study how real shocks to the economy might cause business cycle fluctuations. The paper of Kydland and Prescott (1982) is often considered the starting point of RBC theory and of DSGE modeling in general. The RBC point of view is surveyed in Cooley (1995).

New-Keynesian DSGE models build on a structure similar to RBC models, but instead assume that prices are set by monopolistically competitive firms, and cannot be instantaneously and costlessly adjusted. The first to introduce this framework were Rotemberg and Woodford in 1997. Introductory and advanced textbook presentations are given by Galí (2008) and Woodford (2003). Monetary policy implications are surveyed by Clarida, Galí, and Gertler in 1999.

Example

The European Central Bank (ECB) has developed a DSGE model, often called the Smets–Wouters model, which it uses to analyze the economy of the Eurozone as a whole (in other words, the model does not analyze individual European countries separately). The model is intended as an alternative to the Area-Wide Model (AWM), a more traditional empirical forecasting model which the ECB has been using for several years [when?].

The equations in the Smets-Wouters model describe the choices of three types of decision makers: households, who choose consumption and hours worked optimally, under a budget constraint; firms, who decide how much labor and capital to employ; and the central bank, which controls monetary policy. The parameters in the equations were estimated using Bayesian statistical techniques so that the model approximately describes the dynamics of GDP, consumption, investment, prices, wages, employment, and interest rates in the Eurozone economy. In order to accurately reproduce the sluggish behavior of some of these variables, the model incorporates several types of frictions that slow down adjustment to shocks, including sticky prices and wages, and adjustment costs in investment.

Controversy

Willem Buiter of the London School of Economics has argued that DSGE models rely excessively on an assumption of complete markets, and are unable to describe the highly nonlinear dynamics of economic fluctuations, making training in 'state-of-the-art' macroeconomic modeling "a privately and socially costly waste of time and resources".

N. Gregory Mankiw, regarded as one of the founders of New Keynesian DSGE modeling, has also argued that

'New classical and new Keynesian research has had little impact on practical macroeconomists who are charged with [...] policy. [...] From the standpoint of macroeconomic engineering, the work of the past several decades looks like an unfortunate wrong turn.'

Michael Woodford, replying to Mankiw, argues that DSGE models are commonly used by central banks today, and have strongly influenced policy makers like Ben Bernanke. However, he argues that what is learned from DSGE models is not so different from traditional Keynesian analysis:

'It is true that the modeling efforts of many policy institutions can reasonably be seen as an evolutionary development within the macroeconomic modeling program of the postwar Keynesians; thus if one expected, with the early New Classicals, that adoption of the new tools would require building anew from the ground up, one might conclude that the new tools have not been put to use. But in fact they have been put to use, only not with such radical consequences as had once been expected.'

Narayana Kocherlakota, President of the Federal Reserve Bank of Minneapolis, acknowledges that DSGE models were not very useful for analyzing the financial crisis of 2007-2010. Nonetheless, he argues that the applicability of these models is improving, and that there is growing consensus among macroeconomists that DSGE models need to incorporate both price stickiness and financial market frictions.

The United States Congress hosted hearings on macroeconomic modeling methods on July 20, 2010, to investigate why macroeconomists failed to foresee the financial crisis of 2007-2010. Robert Solow blasted DSGE models currently in use:

'I do not think that the currently popular DSGE models pass the smell test. They take it for granted that the whole economy can be thought about as if it were a single, consistent person or dynasty carrying out a rationally designed, long-term plan, occasionally disturbed by unexpected shocks, but adapting to them in a rational, consistent way... The protagonists of this idea make a claim to respectability by asserting that it is founded on what we know about microeconomic behavior, but I think that this claim is generally phony. The advocates no doubt believe what they say, but they seem to have stopped sniffing or to have lost their sense of smell altogether.'

V.V. Chari pointed out, however, that state-of-the-art DSGE models are more sophisticated than their critics suppose:

'The models have all kinds of heterogeneity in behavior and decisions... people's objectives differ, they differ by age, by information, by the history of their past experiences. '

Chari also argued that current DSGE models frequently incorporate frictional unemployment, financial market imperfections, and sticky prices and wages, and therefore imply that the macroeconomy behaves in a suboptimal way which monetary and fiscal policy may be able to improve.

Commenting on the Congressional session, The Economist asked whether agent-based models might better predict financial crises than DSGE models.

Noah Smith, economics professor and author of the blog 'noahpinion', observed that "DSGE fails the market test." That is, financial modelers who would benefit directly from superior market returns uniformly do not use DSGE models, thus strongly suggesting that DSGE models are not useful for macroeconomic prediction.