Can I be a Know-It-All?
This story is the first in a series of 3 posts (note that they need not be read in a particular order). Each of them is a re-publication of a post that I wrote for a University Scholars’ module entitled “Making Connections”. We were told that the posts needed to have a thematic focus, and mine was “Economics: Thinking about the Basics”.
While I was initially sceptical about the aim of the module (to get us to reflect on our undergraduate experience) it did force me to think carefully about how my different learning experiences were connected, and I was mildly, but pleasantly, surprised at the results. I thought I would publish the posts here as well since, generally speaking, accessing them via Medium is much easier.
Perfect information is a concept in Economics, where all relevant information is presented to an individual to make a decision. In this post, I talk about how we enjoy a form of perfect information in school, and how the lack of it creates uncertainty in problem-solving at work. I also look at ways to mitigate this uncertainty, and their ultimate consequences and efficacy.
As an undergraduate at university, approximately two-thirds of my year is spent attending modules. While the remaining time is officially set aside for vacation, the truth is that I, like most others, spend my summer and winter breaks at internships accumulating work experience, in order to solidify my chances of securing employment upon graduation. Larger organizations usually have structured programmes, with each batch of interns being put through the same training courses or tasked with similar assignments. My internship experiences have been slightly different, given that thus far I have worked with smaller firms that don’t necessarily recruit interns on a regular basis.
Rather than undergoing a template programme that had been refined over several internship cycles, my responsibilities often involved research into areas that my employers themselves were not completely familiar with. This research was often undertaken as a response to a particular problem/question posed by my employers. For example, “How will the value of the aging market in Singapore change by 2030?” or “What will the Chinese football market look like in a decade?” I was baffled when I first heard these questions, uncertain as to how I should proceed. Looking back on these experiences, I realised that the cause of this uncertainty was two-fold.
First, I had little to no familiarity with such problems. While the questions were somewhat related to my field of study (Economics), these were not the types of questions I had been exposed to, and as such, most of the tools I had been equipped with in university did not seem applicable here. There was however, a second issue — the very nature of the problems I was facing was vastly different. My schoolwork involved problems where, more often than not, the following conditions were fulfilled:
- a concrete answer existed,
- I had an idea what the answer was, and
- there were established processes that could be used to arrive at the answer.
This was not true of the problems I tackled as an intern.
An example would be most useful here.
Looking at a weekly assignment for my Macroeconomics Module, EC3102, the question is asking me to explain what would happen to a hypothetical country’s economy following its conquest of another country. The solution sheet later proves what I know for a fact — that a specific answer exists. Moreover, there are established processes that I can use to arrive at the answer ie. The Solow Growth Model, a graphical representation of an economy in the long run (pictured below), taught in a lecture which I know is linked to this tutorial. To top it off, I have an inkling of what the answer should look like: I know the curves have to move, and I know I should eventually arrive at a steady state.
Referring to my project on valuing the aging market however, these conditions are no longer satisfied. One could certainly argue that a definite mathematical answer must exist, but any attempt to find it would only yield estimates, that may or may not be close to the true value. No professional I know would dare say they had derived the exact value. As one of my employers said, “Better to be vaguely right than exactly wrong”. Moreover, I didn’t necessarily know what the answer would look like. There was little done in the way of prior research (or at least, free research that I could access) and even then, all results from those previous analyses would have been estimates. If those estimates were considerably off, having answers comparable to them was of little comfort. Consider the following example:
While I found the estimate for 2014 to be believable, the 2030 one was unconvincing. The number of senior citizens in Singapore is projected to triple in the next 15 years, so the estimate itself should approximately grow by that much. And unfortunately, there were no processes listed/stated in a document/textbook that I could refer to.
All of the above meant that there was a lot more uncertainty in solving the problems I was presented with. One reason for this uncertainty, and the failure to fulfil the conditions listed above, was the lack of perfect information, an oft-mentioned and practised concept in Economics. A simple Google search yields the following definition:
In the problems presented to us in our tutorials, we benefitted from a form of perfection information as well. All the necessary assumptions and variables were made available in our questions. Our sole task was to recall the concepts we were taught (the equations, the cause-effect models) manipulate the data as needed, and derive the result. In my internships and, I suspect, any future work, this is not a state of affairs I can rely on.
Mitigating this uncertainty involves alleviating this lack of perfect (or any) information ie. One would need to gather the information oneself. This forced me to ask the following questions:
- What is my objective?
- What information is needed/relevant?
- What are the potential sources of information?
The first question is often easy to answer. For example, in valuing a particular market, I’m essentially asking “How much money do the people in this market have and are willing to spend?” The other 2 questions tend to be more problematic. There could be different approaches to answering the question, and each one requires different information that needs to be mined from different sources. For example, looking back at my work on valuing the aging market, my opening slide lists 4 different approaches:
A related problem that is equally pertinent: how does one know what approaches are available to begin with?
As far as I can tell, there are no textbooks to guide us in these matters, unlike the bag-splitting documents that we rely on in university. Our guidance must then come from elsewhere, and the most convenient, and perhaps reliable source, are other individuals who have been in our shoes. Compared to school, where learning involves consulting a textbook first before looking elsewhere for answers, here the process of learning is inverted. Individuals become our repositories of knowledge, collaboration become the first means of data gathering. From there, we branch outwards, going down the paths indicated to us by our guides.
This process itself generates its own uncertainty ie. In the midst of trying to increase the information available and reduce uncertainty, we inadvertently create more of the latter. In the face of differing opinions and advice, whose do we heed? We need to create our own criteria for assessing the reliability and credibility of the advice proffered. Moreover, when embarking on research of our own, we need to distinguish for ourselves which data sources to rely on, and within each source, which data is relevant.
The end result of this process is the realisation that perfect information, like most (if not all) ideals, is a fantasy, a theoretical convenience used to simplify problem solving in school, because the focus is on arriving at and understanding the implications of the answers. While non-academic work is also bent on arriving at an answer, the final outcome is an unknown, and the means to get there can be equally mysterious. Navigating this uncertainty is possible, through consultation and co-operation, but the uncertainty is never fully resolved. If anything, more of it finds its way into the mix — a mythical hydra, where the decapitation of one head results in the appearance of more.
Moving away from university thus entails the acceptance of the loss of this ideal, and the necessity of accepting the perennial existence of uncertainty, and the accompanying struggle to quell it. The process of problem solving thus never fully reaches fruition. Information is never perfect, we operate with what we have and can find, and upon the arrival of previously unknown information, we have to rethink our processes and our derivations, always seeking to improve, and never settling.
