What is the idea and methodology of the DSGE model? How to apply

DSGE, whose full name is dynamic stochastic general equilibrium, is a modeling methodology that currently occupies an important (or even dominant) place in macroeconomics research, and is mainly used for discussing economic growth, the economic cycle, and discussing the effects of policy instruments (fiscal and monetary policy).

Traditional macro forecasting models are subject to the Lucas critique, which argues that the use of aggregated historical data to give policy recommendations and economic forecasts leads to dubious conclusions because the macro system itself is composed of a large number of dynamic micro subjects, which in this case can refer to both residential consumers- -They can change their consumption and saving behaviors according to price changes and income changes, etc., and they can also refer to enterprises - they can adjust their production according to changes in prices, costs, and market demand. If the forecasting model is built with no regard for the response of these micro subjects to policy and the actual market, but only mechanically makes quantitative forecasts in the statistical sense, the results will be very unreliable. The prerequisite for the establishment of statistical methods is historical reproduction, that is, the past and the present have the same mechanism in order to respond to the same economic effects, but both enterprises and residents even in the face of a similar economic situation will react differently, of course, for different reasons, they have the role of learning, trial and error, feedback and correction. So policies that worked in the past may not work today, or may even have the opposite effect.

As a direct result of Lucas's critique, the mainstream of macroeconomics has since abandoned the traditional practice of linking variables into a huge set of equations, solving for the parameters, and then making forecasts based on a "pseudo" law that describes how the macro variables vary with each other.

In fact, traditional macroeconomic forecasting models -- used by many central banks in the 1970s to make policy assessments -- try to establish the interrelationships between prices and quantities, typically in the form of a huge set of equations with thousands of variables, which are then run by computers to produce results. They are then run by computers to produce results. This so-called "correlation", even if it is correct, is a stopgap measure because you have no way of knowing what the relationship is between the independent and dependent variables on either side of the equation - is it causal, and if so, who is the cause and who is the effect, and if not, are there other variables determining the relationship? Are there other variables determining both? Questions such as these are not answered by traditional methods.

Macroeconomics must seek a micro basis, starting from the demand side and the supply side of the micro-decision-making, portraying their optimal decision-making in the face of their respective constraints, and then summing up the decision-making behavior of the various micro-bodies to get the so-called aggregate supply and aggregate demand, and then use the conditions of the market clearing to find out the prices, interest rates, and other key variables affecting both sides of the decision to enter the unemployment, the market wage, endogenous variables such as residents' income. The quantitative relationship established on this basis, although not necessarily reflect the real relationship between the variables, but at least can be used for trial and error.

What do I mean by that last sentence? What's the point of working on something when you don't even know if it's real?

In fact, what is the ideal macro model? It should be an economist to investigate each household and each enterprise's various decision-making behavior, long-term tracking, to establish a household and each enterprise's decision-making function, and then sum up to obtain the aggregate supply and demand, and then the government to change the interest rate, tax rate, the issuance of money and other variables, to see how the response of these microscopic subjects, and then select the best for the best.

Obviously, the above strategy is not operational.

What's more, if you think that interest rates and tax rates are real concepts, then do the macro concepts of aggregate supply, aggregate consumption, etc. have clear counterparts in the real world?

So the economist's strategy: start from some accepted assumptions about the micro-constituents, build a model and derive their decisions and sum them up, and then in turn "make up" the actual data by substituting it in. If some important observable quantities, such as unemployment, consumption as a share of GDP, investment as a share of GDP, real interest rates, etc., are well explained by the existing parameters, then the model is considered to have at least a good picture of the existing economy. Then this model is like a laboratory, and we can come to see what happens if we assume tax rates go up, or fiscal subsidies go up.

The microfoundations are based on setting preferences on actors, so DSGE can be used to analyze the welfare effects of policy changes.

As the name suggests, DSGE must firstly be dynamic, that is, any decision is intertemporal, not static. Second, it is stochastic, because in reality micro agents are living in a world of uncertainty, such as technological advances, price fluctuations etc. Finally, it has to be a GENERAL EQUILIBRIUM MODEL, that is to say, DSGE is inherited from the most important result of mathematical economics developed in the post-war period - the theory of general equilibrium.

DSGE is essentially a methodology, a major methodological innovation in macroeconomics since the post-war period.DSGE can be used by people from different schools of thought to construct models that respond to the ideas espoused by their own school (that is, make the assumptions you think are correct about the preferences, technology, and constraints of the micro-agents), and then follow the procedure to make the deductions that are just that.

The Rational Expectations school of thought, of which Lucas is a member, came up with the famous RBC model, or real business cycle, which is usually translated as real cycle theory. Of course, there is no equivalent of the so-called false cycle theory. The "real" here actually means that all the variables in this model are defined in real variables rather than nominal variables, or to put it more bluntly, there is no money in this model, and all the prices and wages are measured in real terms, without all the effects that money can bring.

Not assuming the existence of money is of course far from reality, but first of all this makes the model simple, controllable and solvable. Secondly, what exactly is the use of money, in itself, is a very tangled problem in economics, each school of thought has its own ideas, simply can not reach a unity. But everyone recognizes the fact that: currency if necessary, then the whole economy is certainly not complete market. here do not explain what is complete market, simply put, is that there must be some friction in the world, must rely on the circulation of money to solve the. So RBC automatically assumes that we are in a complete market.

The most important thing is that even though RBC took out the currency, the final conclusions can still be largely kept in line with the real data. This at least shows that taking out the currency did not cause the model to be distorted.

Of course, without money, you have no way of discussing the role of monetary policy - of course, the flip side of that is that since models without money can already explain the world, then perhaps monetary policy probably probably doesn't really do much!

This last statement angered Keynes' disciples - the Keynesians. Once the Keynesians learned the DSGE routine, they revived the ideas of the Keynesian bigwigs, a school of thought known as the New Keynesians. They add some non-competitive factors and other frictions to the model, such as monopolistic competition between firms and rigidity in wage adjustments.