Dear All,
Hello. I major in macroeconomics.
In setting up macro models, it is usually difficult for me to see which is the control variable (consumption, ...) and which is the state variable (physical capital accumulation, ...), especially when the model is large.
Are there any "tips" for distinguishing these? Basically, I think of control variables as the variable(s) in the utility function.
Thanks in Advance,
Mizuki
It is always difficult to start anything when you don't have healthy knowledge about it. As i am not fully able to understand your question but i have attached the related reads for you. Hope you find these helpful.
https://minneapolisfed.org/publications/the-region/modern-macroeconomic-models-as-tools-for-economic-policy
http://macromodelbase.com/
http://www.reed.edu/economics/parker/s11/314/book/Ch01.pdf
Thanks!
Many thanks for pointing out these. Let me clarify.
Suppose that you have to solve the representative-agent model with money and government spending in the utility function. Then, it is obvious that control variables are consumption and money - government spending is not the control variable even if it is in the utility function, because it is out of the reach of the private sector. This is the easy case.
Next, suppose that you have to solve the two-country, continuous-time New Keynesian DSGE model with unemployment, Calvo sticky prices, sticky wages, borrowing constraint, capital adjustment costs, labor disutility, etc...typically, this class of models has more than 5 control variables (I know the model with 20 control variables!). And in this case, for instance, the evolution of prices is the control variable (!) and the price itself is the state variable (see http://www.princeton.edu/~moll/ECO521_2016/Lecture2_ECO521.pdf).
How can we see/distinguish, among (say) 30 variables, which variable is the control and which is the state? Only experiences help? Or are there any "magical" way to sort them out!?
Thanks,
Mizuki
Dear Mizuki,
The control variables must be chosen from the underpinning theory. If you are unable to find them from the theory, better to refer to the literature on said field and check which variables are used as control variables. As far as the main variables are concerned, your research objectives will guide you to select your main variables, however, you can choose among various proxies for measuring the variables.
Best wishes for your research.
Atif
Dear Atif,
Thanks a lot for your comment! Yours is what I really have wanted.
Yes, when things are not clear, it's better to read some papers that use the similar approach and to follow them.
Warm Regards,
Mizuki
Core variables should be chosen in line with the objectives. However, control variables can come from the theory as well as from existing body of knowledge in the form of empirical literature.
I've found the nice rule of thumb. Let me share:
"You wake up in the morning, look at your state variables, make decisions about your control variables, then go back to sleep until the next day."
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression