Economists have been vilified for their inability to anticipate the Great Financial Crisis. In retrospect, the absurdity of the years 2006-8 was flagrant. Economists did not see the obvious right in front of them, namely the huge shadow banking system which, uncontrolled, led to the Great Recession. Paul Krugman argues that the 2008 crisis was above all a ‘failure of observation’. Since then, economists’ track record has barely improved. Just think about the inability to explain the lower-than-expected inflationary pressure in the US (Mrs. Yellen infamous ‘mystery’, in her words) or the weak productivity growth in developed markets despite apparent technological advances. Frequent public arguments amongst leading economists do not help improve the image of the profession.
A peak behind the curtain of macroeconomics is warranted. The latest issue of the Oxford Review of Economic Policy entitled ‘Rebuilding macroeconomic theory’ provides an exceptional perspective on the challenges faced by economists as they seek to assess economic trends and make predictions. There appears to be two diametrically opposed approaches to macroeconomics today:
A – Seek to establish a model with a high degree of empirical coherence which is deemed to be ‘externally consistent’. This is achieved for example though vector autoregressions (VARs) which look at the historical interaction between a bunch of macroeconomic variables, noting the pitfalls of ‘Dangerous Correlations’ (February 4 note). In addition, the validity of empirical models depends upon the availability and quality of relevant data; or
B – Put emphasis on theoretical (as opposed to empirical) coherence by building a macroeconomic model which relies upon the aggregation of interrelated microeconomic theory-driven subsystems for each of the relevant economic actors (consumers, firms, the government and central banks). The macroeconomic model is thus built upon so-called ‘microfoundations’. The ‘Dynamic Stochastic General Equilibrium’ model or ‘DSGE’ falls in that category and is said to be ‘internally consistent’. The DSGE model is at its core a simplified replica of an entire economy as explained in this report published by the New York Fed. It relies upon the mid to large scale modelling of various agents and of the way they can be expected to interact as part of a system. Shocks (e.g. technology) can be simulated to observe system-wide implications as the model seeks to find a new equilibrium alongside business cycles. The ECB has developed a ‘New Area-Wide Model’ for the Eurozone which is based upon these principles and is used as one of the staff’s forecasting tools. The Chicago Fed or the New York Fed have their own
At the very core of the tension between the two possible starting points is the fact that micro-economic theory cannot be systematically reconciled with actual observations. Historical data cannot be agreed with theoretical models; Conversely, theoretical models do not explain historical data. Fundamentally, the actual behavior of economic actors such as consumers remains a puzzle to economists. As a result, combining approaches A and B have represented an unsurmountable challenge. Take the DSGEs. To improve their explanatory power under the circumstances outlined above, economists toy with their models by adding a multitude of variables and (often questionable) constraints. As their complexity grows, so does their opacity. DSGE models tend to evolve as ‘monsters’ (dixit an economist I spoke to) which cannot be controlled by mortals. Joseph Stiglitz concludes that the failing of DSGE models can be traced to the misguided attempt to reconcile macroeconomics with microeconomics.
The biggest mystery of all is not inflation or productivity, it is Man. Economists will no doubt seek to refine their models in years to come through research in microeconomics and behavioral science. Vast amounts of resources will continue to be spent to make little progress as the law of diminishing returns is disregarded. The mystery of Man’s behavior will not be cleared up. Despite its efforts, economics will remain an art more than a science.
A beautiful art.