This is one of a series of columns that were produced for Moneyweb Investor in which Stuart Theobald explores the intersection of philosophy of science and finance. This followed an earlier series for Business Day Investors Monthly on the same theme. This column was first published in March 2016.
Just what is economics and how does it work? Pretty much everyone who studies a commercial degree in South Africa has to do some economics along the way. Many arts students pick it up too. Often they take just enough to get a rough idea of some big ideas in economics – demand and supply, micro and macro. Those certainly provide useful insights into the way economies work, but how?
Just about everything presented in those elementary courses comes in the form of “models”. Here are a few that may jog your memory: the “model” of demand and supply, with downward sloping demand curves and upward sloping supply curves, showing that higher prices induce greater production while lower prices induce more purchasing until they find an equilibrium. The IS/LM model shows how interest rates and money supply affect aggregate output, a graphical representation of Keynes’ general theory. The older Ricardian model of international trade, illustrates how countries that trade benefit because of “comparative advantage” even if one country is more efficient in every respect than another. Then there is the Lewis model of development showing how urbanisation draws labour from the agricultural sector.
All of these models are the basis for lots of debate when it comes to making sense of the world. Some apply quite clearly, illuminating trends we can see in actual economies, while other situations appear to not match at all. But what is notable is that economics uses this method of arguing through models.
Economics is not the only science to reason in this way. Physics also has many models, ranging from the solar system to the “standard model” of particle physics that proposes that the world is ultimately made up of quarks, photons and other elementary particles.
Often models are criticised for being “unrealistic” but that seems to miss something important. Models are useful precisely because they are simplified representations of systems. The social world is messy, even more so than the physical world. Physics models are also simplified, for example, assuming points, vacuums and dimensionless planes, all things that don’t actually exist. Economics does something similar by assuming things like perfect information, zero transaction costs and rational decision makers, all things that don’t really exist either. London’s underground map is an unrealistic model too – it has little resemblance to the actual distance between stations, but provides a good tool to understand how to navigate through the system to get where you want to go.
(Incidentally, some neuroscientists have recently made quite provocative arguments that our cognition of the outside world is really just a model. They are argue that evolution is efficient – it wouldn’t give our brains the ability to comprehend “reality” if an abstract model that required less cognitive machinery would serve just as well in our efforts to survive and procreate. This might be why our brains find it so hard to comprehend some arguments in physics, like the simultaneous wave/particle nature of light and the 10 dimensional space time proposed by superstring theory. Maybe our mind’s model of the world is unrealistic. Luckily it is still useful.)
The Harvard economist Dani Rodrik, who was a key advisor to the South African government during the economic policy debates of the 1990s, has recently published a fascinating book exploring the nature of reasoning with models (called Economics Rules). He argues that economics progresses horizontally by widening its library of models rather than vertically by developing ever more accurate models. That library of models can sometimes flatly contradict each other, but in figuring out which one to apply we learn things about the economy we are studying.
Ever more accurate models of a particular economy would be ever more complex, perhaps as complex as the whole economy itself. But such a model would be quite useless to an economist wanting to understand other economies. The more detailed a model becomes to fit the particular, the less useful it is to understand the general. Simplified models may give us more insight to understand novel situations. And the more models we can draw on to study how they may apply to a situation, the more “truth” we may be able to discover.
When it comes to investing, we rely on economic models to help understand trends in the broader economy, but we have models to understand investment principles too. The “mean variance optimisation” model of the 1950s tells us why we should invest in a diverse portfolio. The Capital Asset Pricing Model of the 1960s tells us we should demand higher returns from riskier assets. But models are only a reasoning tool. We have to still do the hard work of understanding investment opportunities directly. Models can sometimes lead us astray, as the disastrous “value at risk” models did before the financial crisis, which allowed central bankers to think all was ok with the financial system. By understanding just how economics reasons with models, we may be better at avoiding the pitfalls.