There is one dog, watching with interest as another dog is frantically digging. After a few minutes, and seeing the digging dog disappear beneath the ground level in their hole, the first dog has to ask. What are you doing? The digging dog replies I’m looking for a bone I buried. The first dog then says ‘what is it’s not there if you’re in the wrong spot?’. Without hesitation, the second dog replies ‘ I’m this deep already, I think I’ll keep digging’.
Hey, it’s Andrew, and this is Safety on Tap.
Since you’re listening in, you must be a leader wanting to grow yourself and drastically improve health and safety along the way. Welcome to you, you’re in the right place. If this is your first time listening in, thanks for joining us and well done for trying something different to improve! And of course welcome back to all of you wonderful regular listeners.
Sunk cost is the bias in many animals, humans and pigeons included, to continue with a course of action or commitment on the basis of costs already incurred. These costs might be time, money, energy, or even social capital (being seen to be committed). The more we perceive that we have invested, the more likely we will continue down the same path even in the face of contrary information, changing circumstances or diminishing chances of success.

During Homer Simpson’s barbecue with all his friends and neighbours, Lisa steals the centrepiece of the barbeque, the rotisserie suckling pig, using the riding lawnmower to push the grill out of the yard and up a steep slope, while Homer and Bart chase after her. When Lisa gets to the top of the slope, the grill starts rolling downhill, building up speed. Homer and Bart now chase after the grill, while it rolls into a street and through a hedge
Homer: It’s just a little dirty. It’s still good, it’s still good!
[Homer and Bart keep running after the grill. The grill rolls into traffic (miraculously missing every car) and crashes into a bridge railing. The pig keeps going and ends up splashing into the river.]
Homer: It’s just a little slimy. It’s still good, it’s still good!
The pig floats downstream to a dam where it gets caught in the spillway, blocking it. Water builds up behind the pig until the pressure builds up, and explodes through the hole, flying through the air into the distance.
Homer: It’s just a little airborne. It’s still good, it’s still good!
Bart: [crestfallen] It’s gone.

Homer: [even more crestfallen] I know.
We’ve talked a little bit about cognitive biases before on the podcast (like episode 81), and we talk a lot about growth which leads to improved performance, that’s in every single episode.
Which raises the question, are sunk cost and persistence the same thing?
No, they are not, and the distinction is important for your growth.
There are two scenarios I would like to present you with, reflecting the difference between sunk cost and persistence. Imagine a 2×2 matrix. As a safety person, I thought you would like a matrix.
On the vertical, or y-axis, the higher end means continue or do more, and the lower end means to stop or do less. After all, sunk cost and persistence involve the decision about our forward momentum – stop, slow, or speed up?
On the horizontal, or x-axis, the lower or left end means poorer outcomes, and the right or higher-end means good or better outcomes.
This kind of matrix is often used to represent how errors occur when testing a hypothesis. I won’t go too deep on this, needless to say, the two scenarios I’m focussing on, are both errors – when we make the wrong decision.
So on our 2×2 matrix, first, we have the top left quadrant – do more, for poorer outcomes. This is continuing irrationally. Sunk cost lives here. In an experiment this would be called the false positive – we think we should keep going and find reasons and evidence to do so, despite not getting the result we were expecting.
That’s scenario One: continuing irrationally.
Second, we have the bottom right quadrant. Do less, for better outcomes. This is giving up too quickly. This is quitting. In an experiment the equivalent of this would be called the false negative – we think we should stop, and find reasons and evidence to do so, despite getting the result or having the potential to get the result we were expecting.
That’s scenario Two: giving up too quickly.
To to recap our two scenarios:
Continuing irrationally (false positive)
Giving up too quickly (false negative)
My theory is that I think that there is too much of both of these in our practice in health and safety. We continue with things irrationally when we shouldn’t, and we give up too quickly when we should keep going.
Here’s why I think this.
Continuing Irrationally
The dog at the beginning, the one in the ever-deeper hole, was motivated by sunk cost. ‘I’ve come this far, might as well keep going’. One of the reasons why this is considered irrational, is studies have proven that we mentally account for losses to be much greater than potential gains, when the losses and gains are directly comparable, such as with money. A $10 loss will feel like $30 or even $50 compared to the actual, equivalent, $10 gain. This effect is called loss aversion.
So when the reality is that our new online safety software is crappy and no one uses it, we feel pain even thinking about changing it. We worked so hard to get the budget, to get the IT people on board, to get the senior leaders interested in the potential it holds for better data and reporting. It hurts to think maybe we got it wrong, maybe we need to drastically change this situation.
When the reality is we ask for and receive a million pre-start checklists completed, but they all get filed in large boxes, mountainous records of tick and flick behaviour, never to be looked at again, it’s painful to think maybe it’s not working.
As humans, we are quite predictable. We don’t like loss, and so we are irrationally driven by sunk cost. Which is kind of ironic given that our ethical responsibility is to do things which actually effect health and safety, yet as humans, it perversely makes sense for us to continue with the effective activity and down dead-end paths fuelled by irrational sunk cost.
Which doesn’t change how wrong it is, and maybe even unethical. That’s continuing irrationally.
Giving up too quickly
Working in safety is hard. In episode 109 I opened up to you about what it feels like from my perspective, talking about ‘the person in the arena’ to adapt Teddy Roosevelt’s powerful metaphor. I used the phrase battle-hardened, revealing the kinds of metaphors working in safety – it is like a battle. This is a battle which can toughen us up, give us thick skin, or it can damage us or break us. It’s done both to me.
As an anxious perfectionist, I know all too well the feeling of cognitive dissonance when you live in the space between where we know a business has the potential to go, dare I say needs to go, and where they are now.
But for he vast majority of us, this work is out choice. No one is compelling us to do this against our will. And yet I think these battle metaphors, and even the person in the arena, which alludes to a gladiator metaphor, don’t help us. No one can blame a soldier who has left the armed forces, because it was tough, it was damaging.
We aren’t at war.

But I wonder whether sometimes we hide behind the battle as a reason not to keep going. What this looks like, is dulling our strategic aspirations, softening how we hold people to account, pulling back on the hard emotional labour of building relationships and we fall short of asking for what we really want.

Here’s a test anyone can do if you are on LinkedIn. Look in your feed, and check out the profile of the first safety person you see. Jot down the amount of time they have been at their last three jobs. Then look at another person, randomly from your feed. And another. Then take the average tenure, see what you find….
What I think you will find, over a large sample, that the average length of tenure of safety professionals is surprisingly short. On a small sample which I admit is not statistically significant, my calculations came up with 2 years six months. We would need a bigger sample to make definitive statements about the average length of tenure, but my point is not to make a definitive statistical statement but to get you thinking about your own career, your own experience.
We all agree that change is a long term process. That relationships need time. That trust builds slowly. How can we expect to make meaningful progress if we give up too quickly, if we say ‘well, that didn’t work, it’s time to move on’, if we point fingers at all the external factors contributing to our challenges, and don’t look enough at our own role in success or failure.
I am a safety professional, so I know what it is like. This is not at all a criticism, but a question for you to answer honestly for yourself: do you give up too quickly? Do you stop short of what the world needs from you?

Tackling irrational effort and boosting persistence

So how can you tell the difference between when you are motivated by sunk costs and need to stop, or when you feel like quitting but need to persist? It’s really hard to – science has not revealed this to us yet.
But knowing what you know now, you can improve.
Here are my tips for reducing irrationally continuing with sunk-costs. I’m calling these sunk-cost challenges.

Get others involved – increase your rationality and shroud the past.
If you are not a gambler, or at least not a regular gambler, you will know how to do both these things. You will look at someone sitting at a gaming machine in a casino, and think – these machines are designed for the house to win, almost always, and are programmed so over the life of that machine the house must win. When you are gambling, you might set yourself limits, putting so much gambling money in one pocket, and any winnings in the other. You are using rational planning and acting to help you succeed, and not get sucked into the sunk cost. You tut tut when you hear the other gambler saying things like ‘I’ve been on this for a few hours now, it’s warmed up and primed for a win’. You rationally know that is unlikely, and you are also uninvested in the hours and however many dollars the other person has put into that machine.
Getting other people involved to provide a more rational assessment, one which is not anchored to the sunk costs of whatever you are looking at, is really useful. And don’t assume because someone is external, they meet the criteria I am suggesting. Your external auditors are bad outsiders for this purpose because they are affected by sunk costs bias if they’ve been around for a while, if you pay them, they will inevitably be affected by not giving news so bad that it will affect their contract, especially if that audit result has commercial implications like for tender pre-qualification.
Regularly ask yourself, is [insert activity or project or initiative] really achieving what we want it to achieve? The level of sunk investment is correlated with the strength of the bias – the longer you’ve been doing something, the more people involved, the more money spent, strengthens the pull to keep going irrationally. So you have to push back hard on that, and ask tough questions. Edward de Bono called this black hat thinking, you or someone in your team can adopt a deliberately sceptical perspective. Large projects do this with approaches like red-teaming, or having a murder-board – google them.
Manage switching costs. Switching costs are the very real direct and indirect costs we experience – in all those currencies we value, like time, money, emotional energy, social capital – when we switch from one thing to another. In fact, the feeling of loss often is much larger than the feeling of gain, when all things are equal. This loss aversion effect is behind the saying ‘losses loom larger than gains’. It feels worse to lose, or potentially lose something, like switching costs. But these can be evaluated rationally, in relation to potential the benefits of the change. This is the tried and true method of cost-benefit analysis. My advice for you, however, harks back to point number one – get others involved.
We are not good at rationally evaluating switching costs either, because of the sunk cost bias, so don’t do it yourself. There are also very real risks associated with switching, mostly emotional. What will people think if I back out of my earlier decision? What will they say about the wasted resources? What will they do in the future when I decide or suggest a particular course of action? Will it damage my brand, my trust, my reputation? These are important things to consider. Your focus on gains needs to magnify these benefits, to compete with the feeling of loss from switching. So managing risks in switching are important – articulate the costs & benefits. Clearly make the rational case for change, show your homework in evaluating this, amplify the benefits significantly, don’t hark on the costs, and get ready and anticipate other people’s reactions to the change.
Now for the tips on persistence, perseverance, continuing when you feel like giving up. I’ll call these persistence hacks.
Treat your work as an experiment. Instead of saying to yourself, your team, and your internal customers, do this, and you will get this result, why not say ‘for these reasons, we think this action will deliver the desired result. Can we test that hypothesis?’. A subtle reframing of work as an experiment engages everyone in a different kind of thinking, where wrong and right aren’t nearly as important as ‘what are we learning?’. If you are learning, you know more, and your next actions become better tuned. It’s far easier to persist when you are seeking to answer an interesting question, than crossing your fingers and hope you are right all the time.
Assume there will be obstacles and challenges, and plan for them. Two kinds of planning, one in your mind, and the other more practically. Any competitive sport involving two sides competing at the same time will assume the plan won’t work, that they can’t know what will happen so they prepare to be agile, to pivot when necessary. In practical terms, learn from project planners – identify risks and issues upfront, and work out how you will address these. We understand risk management, right?
Taking these two approaches means that when your experiment provides you with learning which suggests you need to pivot, you are doing so based on something sound. Notice how I talked about experiments as how ‘we can test this hypothesis, what are we learning’ – getting other people involved will help you make better decisions about what to do with your new learning, and anticipating challenges and obstacles will make that far less surprising when they pop up, so instead of being signals to quit, they are simply another one of the innumerous problems we solve every day, and get on with the job.

A little more perseverance, and a little less continuing irrationally….what would that look like for your growth and performance?

Thanks so much for listening. Until next time, what’s the one thing you’ll do to take positive, effective or rewarding action, to grow yourself, and drastically improve health and safety along the way? Seeya!


Here’s your FREE reflection worksheet from this episode.

And here’s your FREE download of the full transcript of this episode.
Feel free to share this with your team/colleagues!