Why is macroeconomics studied through "Shocks"?

"Why is modern macroeconomics all about identifying this shock and that shock?"

Coming from a theoretical econometrics background, I have recently delved into empirical macro literature. I am a bit confused about the meaning and significance of the notion of shock. Why we are trying to see the macro-world through the lens of shocks? What is the nature of these shocks that gives them this much explanatory power?

The only explanation I could come up with is that this obsession with shocks is a result of trying to fit macro problems into the framework of Wold decomposition theorem. Any covariance stationary process can be decomposed into a deterministic part and an $$MA(\infty)$$ part. The deterministic part fits well with the idea of steady-state of the endogenous variables, while the stochastic part gives rise to the idea of seeing the macro variables fluctuation as a result of MA shocks.

What is confusing to me is that it seems macroeconomists are trying to find "real" meaning for this merely-statistical notion of shock. The impulse response function of tax shock, government expenditure shock, etc., are how they see the world.

What am I missing?

• (+1) However, I'm not sure what's to be confused about the fact that "macroeconomists are trying to find "real" meaning for this merely-statistical notion of shock". The same can be said about much of economics and economic theory in particular, which is to find some economic (if that's what you mean by "real") meanings for some corresponding mathematical notions, such as equilibrium for fixed point, marginal product for derivative, and so on. Commented Aug 1, 2019 at 21:15
• @HerrK. But don't you think the proper course for economic (and scientific) investigation should be finding the right mathematical tool to model the problem at hand? What you described here "... which is to find some economic (if that's what you mean by "real") meanings for some corresponding mathematical notions", is the opposite, and that's the confusing part: we have a nice general statistical model and we try to squeeze our real (economic) problem into its framework. Commented Aug 2, 2019 at 0:11
• Economics isn’t just about predictive modelling, theres a lot to be learned about considering situations in which prices or supplies rapidly diminish. What happens to X country if oil prices plummet whilst their currency is highly correlated to oil prices? This can be answered through considering such a shock to a model. Commented Aug 2, 2019 at 2:28
• we try to squeeze our real (economic) problem into its framework. >> You are right. And unfortunately, this is exactly how a lot of modern economics works.
– user18
Commented Aug 2, 2019 at 2:45
• @Econ_guy: I agree with you about the proper course for scientific investigation. But I thought your question is why is shock, as opposed to other economic mechanisms, used to drive many theories in macro. Your comment however suggests a question that's slightly different: Why is one particular mathematical framework so often used to model macroeconomic phenomena than other mathematical frameworks. Perhaps I misinterpreted what you intended to ask. Commented Aug 2, 2019 at 2:45

Im assuming you're discussing some of the models associated with Real Business Cycle Theory.

These models have a structural underpinnings (meaning that there is some assumptions/empirics with regard to consumer behavior). Starting with the understanding that consumers are also suppliers of labor which face a intertemporal consumption-leisure trade off and Firms are profit maximizing and use capital,labor and technology to produce.Using this setup we can do some standard calculations for what equilibrium and Balanced Growth Paths would be.

Now this as it stands is not a theory of fluctuations. The way Fluctuations are accounted for in this model is VIA exogenous shocks that enter through Technology,A overall change in aggregate preferences and increases in interest rate ect. This enters in the error term of all our equations.

This is the justification for using Impulse response functions, Its to analyse what the implications are for surprise changes in our variables in terms of consumer and producer behavior.

TL;DR: The main point of this analytical process of shocking is to be cognizant for how economic shocks enter the economy and what the implications are for consumer and producer behavior. This is based on the structure of that is imposed by economic theory.

• Interesting discussion and this is probably pretty dumb but what I don't understand in econometrics ( I have some background in econometrics but formal background is stat ) is that, a lot of the time, there will be multiple predictors on the RHS of some equation and one error term, say $\epsilon_t$. But how does one know whether the error term or "shock" term if you will, represents a shock to the specific predictor if there are multiple predictors there ? Wouldn't you want a shock term for each predictor ? Commented Aug 2, 2019 at 5:05
• Noe that this concept of "analyzing shock terms" is not really done in statistical time series, atleast for the most part. Commented Aug 2, 2019 at 5:07
• @markleeds are you talking about VARs? Time series analyisis uses Impulse response function, ive seen statisticians like Terrance C Mills talk about it in his book.
– EconJohn
Commented Aug 2, 2019 at 15:03
• VARS are a good example but I don't see how the noise term-shock can be attributed to a specific variable. How to know which variable ? Thanks. I realize it's a detailed answer so a good reference is appreciated as an alternative. Commented Aug 2, 2019 at 21:31
• @markleeds In basic real Business Cycle models, the error term ans the consequent shock is always associated with the Total Factor Productivity, namely, the thing that we don't know what the hell it is. So it is reasonable to say that a "surprise" will come from the component about which we are the most ignorant. Commented Aug 4, 2019 at 16:34