Blog post

The DSGE Model Quarrel (Again)

Dynamic Stochastic General Equilibrium models have come under fire since the financial crisis. A recent paper by Christiano, Eichenbaum and Trabandt –

Publishing date
11 December 2017
Silvia Merler

A recent paper by Christiano, Eichenbaum and Trabandt (C.E.T.) on Dynamic Stochastic General Equilibrium Models (DSGEs) has generated quite a reaction in the blogosphere. In the paper, C.E.T. argue that pre-crisis DSGE models had shortcomings that were highlighted by the financial crisis and its aftermath. But over the past 10 years, progress has been made incorporating financial frictions and heterogeneity into DSGE models and C.E.T. foresee that DSGE models will remain central to how macroeconomists think about aggregate phenomena and policy, because there is simply no credible complete alternative to policy analysis in a world of competing economic forces.

Much of the criticism of the paper refers to the first version published online – which is, however, no longer available (the latest version is dated November 27). Noah Smith has extracts of the earlier version, in particular a sentence in which C.E.T. referred to people who don’t like DSGE as “dilettantes”, because they only point to the existence of competing forces at work – and informally judge their relative importance via implicit thought experiments – but can never give serious policy advice. Smith argues that C.E.T.’s defense of DSGE as the only way to make quantitative predictions about the effects of policy changes is wrong, because there are at least two other approaches in common use – sVARs and SEMs. A structural model is also not always needed to make quantitative predictions about policy, as this can often be done in reduced form. When policy changes can be treated like natural experiments, their effects – including general equilibrium effects – can be measured directly instead of inferred from a structural model. But C.E.T. ignore the existence of natural experiments, despite the rapidly rising popularity of the natural experiment approach in economics.

Bradford Delong points out that new Keynesian models were constructed to show that old Keynesian and old Monetarist policy conclusions were relatively robust, and not blown out of the water by rational expectations. They were built to show that the irrelevance of real variables to systematic policy results were extremely fragile. Lucas and company then followed Prescott into the land of Real Business Cycles (RBCs), taking a residual error and claiming it was their fundamental driving exogenous variable. The DSGE framework was then constructed so that new Keynesians could talk to RBCites. None of this has, so far, materially advanced the project of understanding the macroeconomic policy-relevant emergent properties of really existing industrial and post-industrial economies.

Jo Mitchell at Critical Finance thinks that what C.E.T. are attempting to do is argue that anyone doing macro without DSGE is not doing it “properly”. But on what basis is DSGE macro “done properly”? There are two places to look for empirical validation – the micro data and the macro data. Thirty years of DSGE research have produced exactly one empirically plausible result – the expectations-augmented Phillips Curve. It was already well known. There is an ironic twist here: the breakdown of the Phillips Curve in the 1970s gave the Freshwater economists their breakthrough. The breakdown of the Phillips Curve now – in the other direction – leaves DSGE with precisely zero verifiable achievements. C.E.T.’s paper is welcome in one respect: it confirms what macroeconomists at the top of the discipline think about those lower down the academic pecking order, particularly those who take a critical view.

Lars Syll thinks that ‘rigorous’ and ‘precise’ DSGE models cannot be considered anything other than unsubstantiated conjectures as long as they aren’t supported by evidence from outside the theory or model, and no decisive empirical evidence has been presented. Advocates of DSGE modelling want to have deductively automated answers to fundamental causal questions. But to apply ‘thin’ methods we have to have ‘thick’ background knowledge of what’s going on in the real world, and not in idealised models. Conclusions can only be as certain as their premises. The modelling convention used when constructing DSGE models makes it impossible to fully incorporate things that we know are of paramount importance for understanding modern economies. Given all these fundamental problems for the use of these models and their underlying methodology, it is beyond understanding how the DSGE approach has come to be the standard approach in ‘modern’ macroeconomics. DSGE models are based on assumptions profoundly at odds with what we know about real-world economies. That also makes them little more than overconfident story-telling devoid of real scientific value.

Chris Surro at Pretense of Knowledge argues that the problem with DSGE models is not that they are unable to explain specific economic phenomenon, but that they can explain almost any economic phenomenon you can possibly imagine and we have essentially no way to decide which models are better or worse than others except by comparing them to data that they were explicitly designed to match. All that the DSGE model itself adds is a set of assumptions which everybody knows are false, that generate those intuitive results. C.E.T. do nothing to address this criticism. Surro argues that macroeconomics should be exactly the opposite: start by getting the assumptions right. Since we will never be able to capture all of the intricacies of a true economy, the model economy should look very different from a real economy. However, if the assumptions that generate that economy are realistic, it might still provide answers that are relevant for the real world. A model that gets the facts right but the assumptions wrong probably does not.

Brian Romanchuk at Bond Economics thinks that the the recent attempt at a defence by C.E.T. was such a spectacular intellectual failure that it is not worth taking seriously.One could argue that we need to use a modelling strategy similar to the one used by DSGE modellers (i) to account for a shifting policy environment, and (ii) to take into account macro relationships between all variables. Although those are reasonable points, it does not mean that DSGE macro actually fulfils those objectives. One could easily raise doubts about other methodologies, but the paper by C.E.T. went completely off the rails by arguing that no other economic modelling methodology even exists.

About the authors

  • Silvia Merler

    Silvia Merler, an Italian citizen, is the Head of ESG and Policy Research at Algebris Investments.

    She joined Bruegel as Affiliate fellow at Bruegel in August 2013. Her main research interests include international macro and financial economics, central banking and EU institutions and policy making.

    Before joining Bruegel, she worked as Economic Analyst in DG Economic and Financial Affairs of the European Commission (ECFIN). There she focused on macro-financial stability as well as financial assistance and stability mechanisms, in particular on the European Stability Mechanism (ESM), providing supportive analysis for the policy negotiations.


Related content

Opinion piece

How Europe can sustain Russia sanctions

Russia's war in Ukraine has underscored the need for Europe finally to invest more in its own defence and security. Such an outrageous act of aggressi

Silvia Merler, Francesco Nicoli, Simone Tagliapietra and Ana Palacio
Blog post

The microeconomics of Christmas

Review of major contributions to the literature on the controversial topic of the deadweight loss of Christmas.

Silvia Merler