As an academic discipline, Macroeconomics has been criticised for not predicting crises; for using simplistic, out-of-date models; for ignoring data that challenged stylised theories; and for failing to acknowledge that economic theory has little to offer without a clear, socio-political and historical context. The purpose of this course has been to counter such criticisms, not by reinventing the wheel but rather by showing that Macroeconomics, carefully and intelligently deployed, can offer helpful guidelines in addressing society’s key challenges for the 21st century.
In understanding the macro world around us we have deployed models. As we noted towards the beginning of the course, all models are wrong but some are useful. Models help citizens and their representatives to make better choices by organising thoughts, suggesting strategies, clarifying definitions and exposing assumptions. Accounting identities impose discipline as do long-run budget constraints – the issue of sustainability. Models require us to marshal facts and figures; more than that, to understand and interpret such data both critically and intelligently. In many cases, models can reveal unexpected consequences, forcing a rethink of strategies and policy options.
Initially we gained understanding of long-run growth processes through models developed by Solow and Romer. The classic Loanable Funds model provided additional insights. For business cycles, we have enlisted the help of multiplier-accelerator models, the Frisch-Slutsky paradigm and the AD-AS apparatus involving the IS curve, the Taylor Rule, the Phillips Curve as well as the Okun labour market relationship. Taking an international perspective we have extended our analysis to include exchange rates and capital flows. Specifically, we have deployed the Mundell-Fleming model and its extensions to embrace the realities of potential inflation and variable risk premia.
Models of expectations have extended our understanding of how the global economy works. Volatile beliefs, especially about an uncertain future, typically impact the present. Stability is no longer assured, a unique steady state might not exist; the world can be populated by multiple equilibria, bad as well as good. Efficient markets can become highly inefficient; recall Minsky’s ideas about the financial system. Crises can reflect bad luck, but often shocks are home-grown, endogenous responses to systemic frailties. Messy reality poses serious challenges for conventional macroeconomic analysis and these are explored in greater depth in two highly recommended books – Andrew Lo’s Adaptive Markets and William Janeway’s Doing Capitalism in the Innovation Economy.
A focus of our final session will be DSGE and structural models that are widely used in central banks and elsewhere for forecasting and simulating macroeconomic processes. They are by no means the only games in town and alternative approaches – such as Agent-Based Models and machine learning – are becoming increasingly popular. A key theme of the session is that you can never have enough models.
James Surowiecki’s Wisdom of Crowds and Scott Page’s The Difference speak to a particularly modern motif – the power of diversity. The diversity prediction theorem says that many models are better than single models; diversity can trump ability. As prosaic macro examples, Norway’s central bank has found that combining different forecasting models actually produces better inflation predictions. Weighting together the GDP and inflation forecasts of several models is often found to be a more robust procedure than relying on a single approach. The Bank of England is advocating interdisciplinary models that respect complexity, heterogeneity, networks, and heuristics.
At a more general level, the diversity prediction theorem implies that you should not be afraid to think and speak differently – you’re doing society a favour. More diversity is better than less; otherwise an undiversified “wisdom” leads to the badness of crowds; herds, manias and bubbles. A related issue is the benefit of humility: admitting ignorance is better than nursing an illusion of knowledge. Excess confidence can lead to bad decisions: the underestimation of true risk, the cultivation of complacency. In addition, do not overlook the value of simple rules of thumb that can work in the context of radical uncertainty. Better to be roughly right, than precisely wrong. Not a bad piece of advice, perhaps, for your finals!
3 Dec 2020