The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Backrooms Monetary policy Stability Robustness and fit Model Uncertainty Roundtable Discussion James Bullard President and CEO Federal Reserve Bank of St. Louis 27 May 2008 Model Uncertainty and Monetary Policy Design Bank of Korea Views expressed are those of the author and do not necessarily reflect the views of the Federal Reserve System. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability The nature of macroeconomics William Poole: James Bullard — Federal Reserve Bank of St. Louis Robustness and fit Backrooms Monetary policy Stability The nature of macroeconomics William Poole: At the start (1998): James Bullard — Federal Reserve Bank of St. Louis Robustness and fit Backrooms Monetary policy Stability The nature of macroeconomics William Poole: At the start (1998): “Macroeconomics is normal times punctuated by adjustment to shocks.” James Bullard — Federal Reserve Bank of St. Louis Robustness and fit Backrooms Monetary policy Stability The nature of macroeconomics William Poole: At the start (1998): “Macroeconomics is normal times punctuated by adjustment to shocks.” At the end (2008): James Bullard — Federal Reserve Bank of St. Louis Robustness and fit Backrooms Monetary policy Stability Robustness and fit The nature of macroeconomics William Poole: At the start (1998): “Macroeconomics is normal times punctuated by adjustment to shocks.” At the end (2008): “Macroeconomics is adjustment to shocks punctuated by normal times.” James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit The nature of macroeconomics William Poole: At the start (1998): “Macroeconomics is normal times punctuated by adjustment to shocks.” At the end (2008): “Macroeconomics is adjustment to shocks punctuated by normal times.” Lesson: Adjustment is everything. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit The nature of macroeconomics William Poole: At the start (1998): “Macroeconomics is normal times punctuated by adjustment to shocks.” At the end (2008): “Macroeconomics is adjustment to shocks punctuated by normal times.” Lesson: Adjustment is everything. Macroeconomy may be vulnerable to “big ticket losses” during adjustment. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. In studying simpler models, we encounter lots of complications. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. In studying simpler models, we encounter lots of complications. We infer that large models must have a lot of unanalyzed complications. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. In studying simpler models, we encounter lots of complications. We infer that large models must have a lot of unanalyzed complications. Result: We do not trust large models. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. In studying simpler models, we encounter lots of complications. We infer that large models must have a lot of unanalyzed complications. Result: We do not trust large models. Result: We are uncertain about the correct model of the macroeconomy. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. In studying simpler models, we encounter lots of complications. We infer that large models must have a lot of unanalyzed complications. Result: We do not trust large models. Result: We are uncertain about the correct model of the macroeconomy. How can we cope with these doubts? James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sources of model uncertainty We cannot write the “full” macroeconomic model down and study it. Much is missing in any analysis. We understand this. In studying simpler models, we encounter lots of complications. We infer that large models must have a lot of unanalyzed complications. Result: We do not trust large models. Result: We are uncertain about the correct model of the macroeconomy. How can we cope with these doubts? In particular: Since the models are about people, are our doubts also their doubts? James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). Caricature of the backroom versus the frontroom: James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). Caricature of the backroom versus the frontroom: Backroom: States and parameters are known by all, and policy is menacingly complicated. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). Caricature of the backroom versus the frontroom: Backroom: States and parameters are known by all, and policy is menacingly complicated. Frontroom: States and parameters are known by none, but policy itself is easy. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). Caricature of the backroom versus the frontroom: Backroom: States and parameters are known by all, and policy is menacingly complicated. Frontroom: States and parameters are known by none, but policy itself is easy. Backroom: Stationarity assumptions commonplace, but data are often untrusted. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). Caricature of the backroom versus the frontroom: Backroom: States and parameters are known by all, and policy is menacingly complicated. Frontroom: States and parameters are known by none, but policy itself is easy. Backroom: Stationarity assumptions commonplace, but data are often untrusted. Frontroom: Stationarity considered unlikely, but thirst for data is unquenchable. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Backrooms and frontrooms Sargent’s amusing description of academia (backroom) vs policymakers (frontroom). Caricature of the backroom versus the frontroom: Backroom: States and parameters are known by all, and policy is menacingly complicated. Frontroom: States and parameters are known by none, but policy itself is easy. Backroom: Stationarity assumptions commonplace, but data are often untrusted. Frontroom: Stationarity considered unlikely, but thirst for data is unquenchable. Implication: The focus in the policy world is on determining the state of the system by looking at lots of data. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. Important to stress that forecasts are made "under appropriate policy." James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. Important to stress that forecasts are made "under appropriate policy." FOMC has wrestled with this issue for a long time. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. Important to stress that forecasts are made "under appropriate policy." FOMC has wrestled with this issue for a long time. Staff forecasts sometimes made under "constant interest rate" projections, thought to be a "no policy change” benchmark. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. Important to stress that forecasts are made "under appropriate policy." FOMC has wrestled with this issue for a long time. Staff forecasts sometimes made under "constant interest rate" projections, thought to be a "no policy change” benchmark. Sargent correct to stress specification doubts, driven by large model distrust described above. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. Important to stress that forecasts are made "under appropriate policy." FOMC has wrestled with this issue for a long time. Staff forecasts sometimes made under "constant interest rate" projections, thought to be a "no policy change” benchmark. Sargent correct to stress specification doubts, driven by large model distrust described above. Also due to titles like "Making Macro Models Behave Reasonably"? James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Sargent meets the Romers A forecasting exercise comparing FOMC to staff forecasts. Important to stress that forecasts are made "under appropriate policy." FOMC has wrestled with this issue for a long time. Staff forecasts sometimes made under "constant interest rate" projections, thought to be a "no policy change” benchmark. Sargent correct to stress specification doubts, driven by large model distrust described above. Also due to titles like "Making Macro Models Behave Reasonably"? It would be the Romer’s pure forecast exercise if we take the Prescott view. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. These strike me as big ideas. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. These strike me as big ideas. Those that follow monetary policy in the U.S. understand when an old expected interest rate path is discarded and replaced with a new expected interest rate path. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. These strike me as big ideas. Those that follow monetary policy in the U.S. understand when an old expected interest rate path is discarded and replaced with a new expected interest rate path. Perhaps because of a key piece of data. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. These strike me as big ideas. Those that follow monetary policy in the U.S. understand when an old expected interest rate path is discarded and replaced with a new expected interest rate path. Perhaps because of a key piece of data. Or because of an official statement. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. These strike me as big ideas. Those that follow monetary policy in the U.S. understand when an old expected interest rate path is discarded and replaced with a new expected interest rate path. Perhaps because of a key piece of data. Or because of an official statement. The entire path changes. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Information flows Eric Leeper: information flow assumptions and realities affect interpretations of empirical work on the effects of fiscal policy changes. Related to Beaudry and Portier concerning the timing of information and changes in beliefs. These strike me as big ideas. Those that follow monetary policy in the U.S. understand when an old expected interest rate path is discarded and replaced with a new expected interest rate path. Perhaps because of a key piece of data. Or because of an official statement. The entire path changes. Related work by Hamilton. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit More stability analysis I would prefer to see more stability analysis than I did at this conference. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit More stability analysis I would prefer to see more stability analysis than I did at this conference. Stability is more important than you think as a policy goal. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit More stability analysis I would prefer to see more stability analysis than I did at this conference. Stability is more important than you think as a policy goal. Policymakers should strive first to avoid the big ticket losses associated with indeterminacy and expectational instability. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit More stability analysis I would prefer to see more stability analysis than I did at this conference. Stability is more important than you think as a policy goal. Policymakers should strive first to avoid the big ticket losses associated with indeterminacy and expectational instability. In the recent financial crisis ... James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit More stability analysis I would prefer to see more stability analysis than I did at this conference. Stability is more important than you think as a policy goal. Policymakers should strive first to avoid the big ticket losses associated with indeterminacy and expectational instability. In the recent financial crisis ... ... the threat might be best described as the possibility of a transition to a steady state with a low level of financial intermediation services. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. See Bullard and Suda (2008). James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. See Bullard and Suda (2008). Idea: Suppose the forecasting community is Bayesian instead of classical. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. See Bullard and Suda (2008). Idea: Suppose the forecasting community is Bayesian instead of classical. Standard analysis goes through, but with an additional term in the actual law of motion. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. See Bullard and Suda (2008). Idea: Suppose the forecasting community is Bayesian instead of classical. Standard analysis goes through, but with an additional term in the actual law of motion. E-stability conditions are unchanged. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. See Bullard and Suda (2008). Idea: Suppose the forecasting community is Bayesian instead of classical. Standard analysis goes through, but with an additional term in the actual law of motion. E-stability conditions are unchanged. Even Bayesian learning implies some type of expectational stability condition. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning “Recursive least squares” just one in a family of recursive algorithms. New Bayesian interpretations exist of the standard recursive learning exercise. See Bullard and Suda (2008). Idea: Suppose the forecasting community is Bayesian instead of classical. Standard analysis goes through, but with an additional term in the actual law of motion. E-stability conditions are unchanged. Even Bayesian learning implies some type of expectational stability condition. Careful readers of Woodford’s paper at this conference would see expectational stability in play there as well. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. Mehra-Prescott. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. Mehra-Prescott. Twisted beliefs of Cecchetti, Lam, and Mark are brought on by the Great Depression. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. Mehra-Prescott. Twisted beliefs of Cecchetti, Lam, and Mark are brought on by the Great Depression. Agents are Bayesian learners with respect to the probabilities of states. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. Mehra-Prescott. Twisted beliefs of Cecchetti, Lam, and Mark are brought on by the Great Depression. Agents are Bayesian learners with respect to the probabilities of states. Bad state is rarely visited, so that pessimism remains influential for decades. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. Mehra-Prescott. Twisted beliefs of Cecchetti, Lam, and Mark are brought on by the Great Depression. Agents are Bayesian learners with respect to the probabilities of states. Bad state is rarely visited, so that pessimism remains influential for decades. The equity premium is large and declines only very slowly to the “true” value of zero. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Learning and robustness A thought about robustness: Agents could “learn not to be so pessimistic.” But consider Cogley and Sargent (2007) concerning the equity premium. Mehra-Prescott. Twisted beliefs of Cecchetti, Lam, and Mark are brought on by the Great Depression. Agents are Bayesian learners with respect to the probabilities of states. Bad state is rarely visited, so that pessimism remains influential for decades. The equity premium is large and declines only very slowly to the “true” value of zero. The 1970s as a similar beliefs-twisting event? James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: Kocherlakota (2007, FRB-St. Louis Review). James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: Kocherlakota (2007, FRB-St. Louis Review). Argues by example that a better-fitting model can give a worse answer to a substantive policy question. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: Kocherlakota (2007, FRB-St. Louis Review). Argues by example that a better-fitting model can give a worse answer to a substantive policy question. What we want is the answer to the policy question. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: Kocherlakota (2007, FRB-St. Louis Review). Argues by example that a better-fitting model can give a worse answer to a substantive policy question. What we want is the answer to the policy question. I see the recent literature described by Sims as an attempt to reach an intelligent compromise in this area. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: Kocherlakota (2007, FRB-St. Louis Review). Argues by example that a better-fitting model can give a worse answer to a substantive policy question. What we want is the answer to the policy question. I see the recent literature described by Sims as an attempt to reach an intelligent compromise in this area. The fit to data gives us confidence that we are on the right track with our economic concepts. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability Robustness and fit Fitting the data Professor Sims spoke eloquently about model fit versus the interpretability of DSGE models. Good model fit is often taken as absolutely essential in the “frontroom.” But not necessarily in the backroom: Kocherlakota (2007, FRB-St. Louis Review). Argues by example that a better-fitting model can give a worse answer to a substantive policy question. What we want is the answer to the policy question. I see the recent literature described by Sims as an attempt to reach an intelligent compromise in this area. The fit to data gives us confidence that we are on the right track with our economic concepts. But we do not want to push so hard in getting a good fit that we lose our economic grounding altogether. James Bullard — Federal Reserve Bank of St. Louis Backrooms Monetary policy Stability An excellent conference Thanks to Governor Lee and the Bank of Korea staff. James Bullard — Federal Reserve Bank of St. Louis Robustness and fit Backrooms Monetary policy Stability Robustness and fit An excellent conference Thanks to Governor Lee and the Bank of Korea staff. This has been an excellent conference on critical topics at the research frontier of macroeconomics. James Bullard — Federal Reserve Bank of St. Louis
@FedFRASER