Back to Square One: Identification Issues in DSGE Models

By Fabio Canova, Luca Sala. Social Science Research Network, 2009

610 citations recorded by the source corpus.

We investigate identifiability issues in DSGE models and their consequences for parameter estimation and model evaluation when the objective function measures the distance between estimated and model impulse responses. We show that observational equivalence, partial and weak identification problems are widespread, that they lead to biased estimates, unreliable t-statistics and may induce investigators to select false models. We examine whether different objective functions affect identification and study how small samples interact with parameters and shock identification. We provide diagnostics and tests to detect identification failures and apply them to a state-of-the-art model

Subfields: Control and Systems Engineering, Mechanical Engineering, Signal Processing

View Paper: https://www.semanticscholar.org/paper/6628c390ff348fc1d1d9001ec21472514d899267

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Control and Systems Engineering

In Mechanical Engineering

In Signal Processing

Citations in the corpus

Yield curve in an estimated nonlinear macro model (2011, methodology): “These issues can arise either from data-related reasons or model-related reasons according to Canova and Sala (2009). Estimation results of the DSGE model using data simulated from the same model do not show evidence of such a bi-modality.”
Monetary Policy Rules and Macroeconomic Stability: Some New Evidence (2010, background): “model’s parameters, see Canova and Sala (2005) and Mavroeidis (2004).”
How Structural Are Structural Parameters? [with Comments and Discussion] (2007, neutral): “At the same time, Cogley and Sargent (2005) defend that there is still vari ation in the parameters of a vector autoregression (VAR), even after con trolling for heteroskedasticity. We are currently working on an exten sion of the model with both parameter drifting and changing volatilities. We buil”
Local identification in DSGE models (2010, methodology): “This is partly due to the widespread use of Bayesian methods, which, as Canova and Sala (2009) point out, may serve to conceal existing identification problems.”
Appendix: Monetary Policy Drivers of Bond and Equity Risks (2020, neutral): “…rates, an integrated treatment of bonds and stocks, especially with risk premia not driven by counterfactually extreme heteroskedasticity in macroeconomic fundamentals, has so far proved elusive.3 We use our model to quantitatively investigate two candidate explanations for the empirical…”
The Price Puzzle Revisited: Can the Cost Channel Explain a Rise in Inflation after a Monetary Policy Shock? (2009, background): “In a recent paper Canova and Sala (2005) provide some diagnostic tools to detect identification problems when the objective function measures the distance between empirical and model impulse responses.”
Macroeconomic Shocks and Their Propagation (2016, background): “Work by Canova and Sala (2009), Komunjer and Ng (2011), and others highlights some of the potential problems with identification in DSGE models.”
The State of Macro (2009, neutral): “Near nonidentification is frequent, with different sets of parameters yielding nearly the same value for the likelihood function—which is why pure maximum likelihood is rarely used [on nonidentification or weak identification, see Canova & Sala (2006) or Iskrev (2008)].”
Conditional Linear Combination Tests for Weakly Identified Models (2016, background): “As Canova and Sala (2009) highlight, however, if there is limited information about the structural parameters θ these approximations may be quite poor.”
Identification and frequency domain quasi-maximum likelihood estimation of linearized dynamic stochastic general equilibrium models (2012, neutral): “Substantial progress has been made recently, notably by Iskrev (2010) and Komunjer and Ng (2011), and by Canova and Sala (2009), Consolo, Favero, and Paccagnini (2009), and Fukac, Waggoner, and Zha (2007). Komunjer and Ng (2011) documented that an inherent difficulty in the identification analysis i”
Monetary policy and uncertainty in an empirical small open-economy model (2010, background): “1”
Bayesian Analysis of DSGE Models (2006, neutral): “Lack of identification is documented in papers by Beyer and Farmer (2004) and Canova and Sala (2005).”