Dani Rodrik, an economist from Harvard, wrote ‘Economics rules’, a book on economics (2015). In his book, he tries to explain what economics is as a science and what it is not.

Mathematics, and sometimes difficult math, has become very important in economics, giving it a sense of an exact science. But Rodrik is  clear on this: economics is a social science, so it is not possible to come up with fundamental laws of economics, rules that are universally valid. 

In economics, one comes with a model (another term for an economic theory) that tries to describe a particular aspect of how society functions. A model is always a simplification of reality, otherwise it would be too complex and not very helpful to gain insights in this particular aspect. Simplifications are necessary, but can also be dangerous. One needs to be careful what simplifications are accepted and what are not. And, importantly, also accepting simplifications is context specific: in one context a simplification can be accepted, whereas it is not in another context. Therefore, a good economist will frequently answer “it depends on the specific situation”, when asked a general economic question. 

Another important advantage of models is clarity (or transparency). With a model, basically a set of mathematical equations, everything is clearly spelled out. Misinterpretations and moving the goalposts when defending your model is not possible. Using models also helps to think clearly and logically about the problem at stake, not only with others but also for yourself.

In my opinion, the main message of the book, however, is that economics is not a science that builds on previous research to refine the former research. So, it is not like physics where, for example, Newton came up with some universal rules which were refined centuries later by Einstein (so the new refined rules also hold for objects traveling at very high speeds, which was not the case with Newton’s laws). In economics, Rodrik explains, one doesn’t build vertically (like in physics), but horizontally, adding new models to give answers on new questions or on the same questions in a different context. Seldomly, a model is being dismissed. An example that Rodrik gives here, and that I also remember from following Paul Krugman’s blog after the financial crisis in 2008-2009, is the use of an old Keynesian model (IS-LM) to describe what was happening after the financial crisis. This macroeconomic model was much simpler than the complex models used by central banks, but was pretty adequate in forecasting the absence of inflation even though there was a relatively large stimulus being done. The old Keynsian model was not intrinsically better than the new and more complex ones, but it was better in this context. Good economics depends highly on the choice of the richt model for the specific context.

Economics as a science that adds models to explore and explain ever more aspects of society (or in different situations) also has its importance when designing and regulating markets, something governments and other public agencies do all the time. Rodrik refers to Jean Tirole, a Nobel-winning economist, who said that market regulation and design is industry-specific. Also here, it depends. Context is everything.

So, politicians should know this, and maybe they do, but they often miss to apply this wisdom. Probably this is because “it depends” is a humble and nuanced message. And with such a message, it is diffecult to incite your tribe or convince a disinterested public.

Rodrik provides a good lesson to economists to be more humble, but also to appreciate the power of economic science. A good economist has learned to think clearly and thoroughly, with the help of mathematics, about complex problems that are important for society. It reminded me of this quote by Ben Bernanke in 2013: 

“Economics is a highly sophisticated field of thought that is superb at explaining to policymakers precisely why the choices they made in the past were wrong. About the future, not so much. However, careful economic analysis does have one important benefit, which is that it can help kill ideas that are completely logically inconsistent or wildly at variance with the data. This insight covers at least 90 percent of proposed economic policies.”