Are DSGE models used for the day-to-day government policy estimation and optimization? E.g. can DSGE model be used for estimation of the return of additional investments in education and how those education gains propagates to the whole economy (possibly DSGE in multi-sectoral setting this propagation will be sector-dependent). Can similar estimations be made for investments in healhtcare, for industrial policy, for investments in R&D, for estimation for welfare programmes? Can DSGE models be tuned monthly estimations, for making arguments about budget decisions?

Are there government papers (no matter which government, just sample could be acceptable) which uses DSGE models for such estimations?

If DSGE models are not used for government decisions about issues that are outside monetary and fiscal scopes, then - what is the reason why DSGE (as the best macroeconomic models available today) is not used for government policy optimization? Are parameter estimation costs so large?

  • 1
    $\begingroup$ I think it's always important to keep in mind (as Dani Rodrik would say) "not to confuse a model as the model." Different types of models are useful in different settings, and as advanced as DSGE models might be for considering the impacts of particular shocks on the economy, or testing optimal monetary policy, they aren't particularly well suited for the micro cases you suggested. For example, one of the most difficult modelling aspects of human capital problems is understanding the determinants of returns to education. Here, we need to consider a variety of different types of models, ... $\endgroup$ – AndrewC Jul 1 '18 at 3:12
  • 1
    $\begingroup$ micro theory models can propose mechanisms important to consider, while econometrically estimated models (such as regressions, or Mincer wage equations, etc.) provide insight into the actual observed forces at work. A DSGE model could only provide broad calibration results, largely dependent on the specific form of the model we consider. Furthermore, what we care about is different in the micro vs monetary cases. The education literature isn't particularly interested in things like impulse responses (a fundamental output of DSGE models, of course) since that's generally pretty obvious- on ... $\endgroup$ – AndrewC Jul 1 '18 at 3:17
  • 1
    $\begingroup$ impact, increasing the amount of skills workers have will generally lead to an increase in productivity. DSGE models give us IRFs that we try to align with observed data, while human capital models are more designed to have more predictive ability. $\endgroup$ – AndrewC Jul 1 '18 at 3:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.