Media Briefings

ECONOMIC MODELS AS ANALOGIES

  • Published Date: August 2014

Much of modern economic theory is concerned with generating insightful analogies about the way the world works rather than accurate rules. That is the central message of a new study by Itzhak Gilboa, Andrew Postlewaite, Larry Samuelson and David Schmeidler, published in the August 2014 issue of the Economic Journal.

Their research helps to explain why economists are rarely bothered by the common criticism that they analyse models whose assumptions they know to be false. Part of the reason, the authors suggest, is that economists’ ‘theories’ are not necessarily meant as general statements but as sources of analogies, which, by bearing some similarity to concrete real-life situations, can change the way we think about them.

The researchers conclude that since the application of economic models requires originality and creativity, it might be advisable for governments more often to seek the advice of basic economic scientists in addition to that of ‘economic engineers’.

Economics is an interesting phenomenon in today’s academic world, the study notes: it is considered to be a respectable social science, yet it is often criticised for scientific failure. On the one hand, it appears to have a unifying and acceptable paradigm and to use sophisticated mathematical tools more than other social sciences. On the other, it is the butt of many jokes and the target of many complaints.

Some of these complaints are misplaced. Economists should not be expected to predict financial crises any more than physicists are expected to predict earthquakes. Large complex systems that cannot be isolated in laboratory experiments pose serious challenges to science’s ability to predict – and we would do well to lower our expectations for what can be achieved.

By contrast, when we restrict attention to smaller isolated systems where analysis and experimentation are possible, economics can boast many successes. Importantly, it can provide guidelines regarding its own applicability, delineating the scope of applications for given theories.

But there is another reason for which economics is often criticised: it seems to make assumptions that are unrealistic and to proceed to analyse their implications, demonstrating considerable mathematical skill while blithely ignoring the falsity of these assumptions.

Psychologists have criticised the assumptions of homo economicus at least since the Nobel Prize-winning work of Herbert Simon, Daniel Kahneman and Amos Tversky, which has given rise in recent decades to the field of behavioural economics. Many a layperson believes that economic analysis should be redone from scratch, ignoring all conclusions that are based on false premises and starting afresh with assumptions that are, at long last, correct.

It is an interesting sociological fact that despite these criticisms, most academic economists feel that they understand reality and can even provide predictions and advice better, thanks to these models. Why is this the case? Why do economists in general – and economic theorists in particular – find incorrect models useful? This question has bothered economists over the twentieth century and it has received considerable attention from philosophers and historians of science in recent decades.

This study attempts to offer a new way to look at economic models that may help explain this puzzle in the sociology of science. According to this view, economic models are not always offered as general rules, making universal statements of the type ‘whenever price goes up, demanded quantity goes down’.

Rather, economic models are sometimes suggested as particular examples that can change the way we think about concrete cases by way of analogy. Thus, models are not necessarily about rules but about particular cases.

For example, Nobel laureate George Akerlof’s celebrated ‘lemons’ model (based on the market for used cars) may be viewed as a story or a ‘theoretical case’, which, by bearing some similarity to concrete real-life situations, can change the way we think about these situations without offering any general statements.

The view that economists sometimes engage in case-based or analogical thinking can explain some stylised facts about the practice of economic science. For example, economists seem to be relatively unperturbed by refutations of their ‘theories’. It is argued that part of the reason is that these ‘theories’ were not necessarily meant as general statements but as sources of analogies. Cases might refute general theories but they cannot refute each other.

Similarly, the scientific nature of economics is sometimes criticised. Using the analogical view, one may suggest that economic models do not necessarily make direct falsifiable predictions. Such predictions would necessitate a method for judging similarity between real-life cases and theoretical ones.

By contrast, economic models are often pre-scientific: they pave the road for analogical predictions but they do not yet commit to such. This might, in turn, explain why the application of economic models requires originality and creativity – and why governments should perhaps seek the advice of basic economic scientists rather than that of ‘economic engineers’.

ENDS


Notes for editors: ‘Economic Models as Analogies’ by Itzhak Gilboa, Andrew Postlewaite, Larry Samuelson and David Schmeidler is published in the August 2014 issue of the Economic Journal.

Itzhak Gilboa and David Schmeidler are at Tel-Aviv University. Andrew Postlewaite is at the University of Pennsylvania. Larry Samuelson is at Yale University.

For further information: contact Romesh Vaitilingam on +44-7768-661095 (email: romesh@vaitilingam.com; Twitter: @econromesh); or Andrew Postlewaite via email: apostlew@econ.upenn.edu