On Models, Behaviour and Data: on the books by Shiller, Page & Hand.

Predictive models have become an integral part of modern society, helping us make decisions about everything from financial investments to medical treatments. However, it's important to remember that these models are only as good as the data and assumptions that go into them. As Robert Shiller, the Nobel laureate in economics, discusses in his book "Narrative Economics," people's behaviour is often driven by stories and narratives, which can be difficult to capture in a predictive model.

Similarly, in his book "The Model Thinking," Scott Page argues that models can help us understand complex systems, but they can also be limiting if we rely on them too heavily. In this post, we'll explore some of the insights from these two authors and consider how predictive models can be used effectively to understand and predict human behaviour.

One of the key ideas that #Shiller discusses in "Narrative Economics" is the role of stories and narratives in shaping people's beliefs and behaviours. He argues that economic events are often driven by changes in people's beliefs and expectations, which can be influenced by the stories and narratives they encounter. For example, the stock market bubble of the late 1990s was fuelled in part by stories of easy profits and the belief that the internet would revolutionize the economy.

This suggests that predictive models that rely solely on data and statistical analysis may be incomplete, since they may not fully capture the influence of these narrative factors. Shiller argues that we need to consider the role of stories and narratives in shaping people's behavior when building predictive models.

Similarly, Page discusses the importance of considering multiple models and perspectives when trying to understand complex systems, such as human behaviour. He argues that relying on a single model or perspective can be limiting, as it may not capture the full complexity of the system.

Instead, he suggests that we should think about models as tools that can help us understand different aspects of a system. By considering multiple models and perspectives, we can get a more complete understanding of the system and make more informed predictions.

So, how can we use these insights to improve our predictive models and better understand human behaviour? One approach is to incorporate a wider range of data and perspectives into our models. For example, in addition to traditional economic data, we could consider factors such as social media trends, sentiment analysis and other type of alternative data, which can provide insight into the stories and narratives that are shaping people's beliefs and behaviours.

Another approach is to be mindful of the limitations of our models and recognize that they are only one tool among many for understanding complex systems. We should be open to considering multiple models and perspectives and be willing to adapt and update our models as new information becomes available.

Overall, the insights from Shiller and Page suggest that predictive models can be powerful tools for understanding and predicting human behaviour, but we need to be mindful of their limitations and consider the role of stories and narratives in shaping people's beliefs and behaviours. By incorporating a wide range of data and perspectives and being open to multiple models and approaches, we can build more accurate and comprehensive predictive models that can help us make more informed decisions in a complex and rapidly changing world.

Both of these books highlight the importance of understanding the underlying factors that shape our understanding of the economy and the world around us. They both encourage us to question our assumptions and to look beyond the surface-level data in order to gain a deeper understanding of the forces at play.

David Hand's book "Dark Data" takes this idea a step further by exploring the concept of "dark data," or data that goes uncollected or unanalyzed due to various limitations or biases. Hand argues that this "dark data" can have significant implications for our understanding of the world and that by considering a wider range of data, we can gain a more accurate and complete picture of the world around us.

All three of these books highlight the importance of looking beyond the surface-level data in order to gain a deeper understanding of the world. They encourage us to question our assumptions and to consider a wide range of data and perspectives in order to gain a more accurate and complete picture of the world. By doing so, we can make more informed and effective decisions, whether we are individuals trying to make sense of the world around us or policy makers trying to shape the direction of the economy.

PS: Yes, at the bottom of the photo I had the letter that Peter Drucker sent me in person in 1998 (when letters were still sent)