What Is Causation?
Many writers on LinkedIn and other places have been presenting very complicated and comprehensive strategy frameworks for a while now, including me. They present boxes and arrows pointing to other boxes, with some looping around on themselves. Are the connections of various boxes supposed to be logically related (that is – “this seems to affect that from everyday observation”), or are the relationships supposed to be correlated supported by data or causal supported by data? Most writers do not distinguish among these three options. I think as strategists we should be mostly about causation. Causation allows us to make predictions and as will see this is very beneficial.
Let me make a caveat before we begin. There is a class of problems called “messy problems”. These are problems characterized by an absence of a correct solution because multiple stakeholders have conflicting goals. We will table this discussion for now as in my view most strategic management problems do have a solution. This is in turn caused by key stakeholders desiring firms to remain healthy for the long term, thus coalescing on a key goal.
This short article (for me given my recent lengthy tomes) is a pre-amble to a comprehensive article I will publish in a few days (today is October 19, 2020). The upcoming comprehensive article will be on applying my full causal model of what increases a for-profit firm’s valuation as a firm moves to its Full Potential to the Rolex watch company. Given that few folks will read this to-be-published longer article (due to how busy people are) I thought I would carve out the discussion of what causation is and why I think it is so important.
Figure 1 depicts my attempt to make understanding causation easier:
Figure 1: How Causation Works: Example of Improved Employee Satisfaction Causing Improved Customer Satisfaction
I will discuss Figure 1 more below. But by way of introduction, causation is evidenced when a change in one variable will produce a change to some degree in another variable every time. Causation is different than correlation, as many know. Correlation is simply when two or more variables move up or down together or where one moves up and another down. To use a facetious example, it has been statistically shown that the annual rise in the Nile River is correlated with higher sugar cane yields in south Louisiana. This is non-sensical and is simply a result of coincidence. But in organizations we see two variables move together all the time with no apparent rational reason for one “causing” the other. An example would be when an office air conditioner intermittently breaks in August in south Louisiana over a two-week period, more people file complaints about their immediate supervisor. The broken air conditioner cannot be proven to have caused the increase in employee complaints about their immediate supervisor. There could be a causal relationship but other factors could have come into play. For example, a training program for all supervisors had them all exhibiting a new form of assertive control that the employees detested. All we can say is the intermittent breaking of the air conditioner is correlated with increasing complaints about direct supervisors.
Causation is what we are after in strategy work in my view. In my experience over the last forty years working with strategic planning departments as part of an overall competitive strategy project I have found very few who can prove what causes what. But I think this knowledge via proof is indispensable for making wise investment choices. Causation knowledge throughout a firm coalesces to what we think will have a direct up or down impact on firm valuation as the firm moves to its Full Potential.
Let’s delve into Figure 1 and discuss the example of improvements in Employee Satisfaction causing improvements in Customer Satisfaction. We know the logic here. Disgruntled employees in call centers (itself correlated with lower employee satisfaction) always lead to reported declines in customer satisfaction of some degree. To prove causation, we need to show that if Employee Satisfaction improves by X% then Customer Satisfaction will improve by Y% every time over a reasonable range of percentage changes. We can only have 100% of our employees becoming very satisfied as a goal. A similar grand goal for complete customer satisfaction is at most 80% from my experience. Some customers will never be very satisfied no matter what we do. Therefore causation should be expected to happen within “normal” limits. So wherever the current baseline measures of employee satisfaction and customer satisfaction are, we can expect improved changes that are reasonable but offer a stretch aspect. And in my experience Employee Satisfaction needs to have a higher percentage change to cause a usually lower Customer Satisfaction percentage change.
This kind of analysis should proceed for all the variables that could possibly have a causal relationship to firm valuation as it moves to its Full Potential. Figure 1 refers to these as Causer Variables (example Employee Satisfaction) and Causee Variable (example Customer Satisfaction).
I hope Figure 1 is fairly straightforward to assimilate. The vertical axis is Quantity of a variable and the horizontal axis is Time. You can see a Causer variable has a Current State Quantity and a Changed State Quantity. Same for any Causee variable. You can further see that there is a percentage change (the delta symbol) for each state calculated. I term the time during such change is happening as Flow. Notice then that both the Changed Quantity of Employee Satisfaction at the end of its Flow cycle and the activities during the Flow cycle can cause the Changed Quantity of Customer Satisfaction.
We should be able to prove a hypothesized causal relationship with data and statistical techniques to declare that relationship is causal and not of the correlational kind. If not, the management team can decide to claim a causal relationship via observations over many cycles that bear the relationship out. This is not ideal though. Some entity in the firm should have accountability to do this work for all variables thought to have a causal relationship.
You might be thinking this is “overkill” and why not just conjecture with observations that are confirmed by many over time and perform ‘back-of-the-envelope” type analyses. You could do this as I stated above but in today’s world, you will likely become confused about what causes what and what the investment in new initiatives for improvement in firm valuation should be. Resource allocation for new initiatives is where the rubber meets the road for increases in firm valuation that continues an organization on its way to its Full Potential. This is crucial for management teams and boards of directors to have data-driven knowledge of about causation.
For just a little more added work over any same time period that the conjectural and back-of-the-envelope approach takes, the management team and board of directors can have more assurance they are placing the best investment bets. Nothing is 100% certain, but a robust approach to causation analyses greatly improves certainty and peace of mind.
This article is part of a series on what causes a firm’s value to increase.
Dr. William Bigler is the founder and CEO of Bill Bigler Associates. He is a former Associate Professor of Strategy and the former MBA Program Director at Louisiana State University at Shreveport. He was the President of the Board of the Association for Strategic Planning in 2012 and served on the Board of Advisors for Nitro Security Inc. from 2003-2005. He is the author of the 2004 book “The New Science of Strategy Execution: How Established Firms Become Fast, Sleek Wealth Creators”. He has worked in the strategy departments of PricewaterhouseCoopers, the Hay Group, Ernst & Young and the Thomas Group among several others. He can be reached at bill@billbigler.com or www.billbigler.com.