Why bad leaders think they are good leaders

Ask people “Do you have above average looks?” Research shows that roughly 90% will say yes. Ask “Do you have above average driving skills?” Again, 90% will say yes. “Sense of humour?” You guessed it, 90%.

It’s human to think we’re above average, but in leadership it’s a worrying phenomenon.

We’ve all come across the underperforming executive who thinks they are an outstanding leader. They don’t simply think they are good, they think they are very good.

We also strike the opposite. Excellent leaders who think they are nothing special. We assume they are just being modest, but are they? And why are these leaders sometimes the last to spot how hopeless other leaders are?

If you’ve experienced the above, it wasn’t a fluke. In fact, it has been so well demonstrated in studies that it is now known as the Dunning-Kruger effect. This is a cognitive bias where people of low ability assess their ability as greater than it is, because of their meta-cognitive inability to recognise their own ineptitude.

One way this phenomenon was tested was by having university students estimate their rank in class. They were also assessed (objectively) on their logical reasoning, grammar and sense of humour. Those in the bottom quartile on these factors hugely over-estimated their rank in class. On average, they sat at the 12th percentile. On average, they ranked themselves at the 62nd percentile. The same was true at the other end of the scale.

Across a large number of quite different studies, the finding is clear: people at all performance levels are equally poor at estimating their relative performance. Most concerning though, researchers found that “poor performers do not learn from feedback suggesting a need to improve”.

This problem is especially difficult to deal with in the realm of leadership. If someone doesn’t believe their rank in maths class, you can show them their rank. If someone doesn’t believe they are bad at chess, you can get them to play a few games. It may take a while, but the evidence eventually gets through.

However, leadership suffers from fuzzy causality. This has three characteristics:

• Cause and effect may be separated by a great deal of time
• Causes have multiple effects and effects have multiple causes
• You may never find out what the effect was

To put that in a leadership context, you may undertake several actions (e.g. holding meetings, undertaking one-to-one discussions, embarking on a new marketing campaign etc) to produce a desired result (e.g. improved business outcomes). However:

• It may take months or years for the effect of these actions to come to light (you might have left the job or the company by this time)
• The ultimate outcome will be influenced by much more than what you did

One of the difficult implications of fuzzy causality is that good leadership behaviours may go hand-in-hand with bad outcomes and bad leadership behaviours may go hand-in-hand with good outcomes (or no evidence of the bad impact for a very long time).

In this situation, the Dunning-Kruger effect becomes even greater than normal. Not only is the person unaware of their incompetence, but there may be no evidence to the contrary. Or worse, the evidence may suggest they are competent.

I’ve seen this first-hand. I’ve seen leaders promoted because of short-term actions that produced short-term effects that got them promoted. In one case, some got promoted five times in ten years, always moving on before the proverbial hit the fan. Finally, they were made the CEO and this time, they didn’t move on soon enough. The company was almost bankrupted.

So what can you do about it?

Firstly, how do you ensure that that fuzzy causality doesn’t lead your company to make bad decisions about your future leaders?

The key here is to focus on ‘lead’ as well as ‘lag’ indicators and to ensure that your approach is time-sensitive. All things being equal, the right behaviours will lead to the right results. As a rough rule of thumb, good actions today will see better employee results in around three months, better customer results in six months and better financial results in a year. Therefore:

• Ensure that behaviour is recognised as something important.
• Ensure that you are looking at "right behaviours" as a key component of performance measurement. This can be assessed by upward feedback, 360-degree assessment or employee engagement scores.
• Assess business outcome information at the right time. Don’t expect results too soon or you might incentivise the behaviours that are worst for the long-term health of your business.
• Be careful of the effect of fuzzy causality. Ask yourself: Are there reasons other than the leader’s effectiveness that have skewed results more positively or negatively?
• Combine results from all stages: behaviour, employee, customer, financial. If all are positive, the person has almost certainly led well. If the first three are positive but not financial, perhaps there are other factors or simply a little more time is needed. If only financial is positive, ask yourself if this might be a misleading result.

Secondly, how do you address leaders who have the ‘illusory superiority’ bias?

This one is much harder. Evidence is a good place to start, but not always sufficient. Training and development can help too, but it can’t be delivered in a way that causes people to simply confirm their existing views of themselves. A seminar on leadership frameworks and tools just won’t cut it.

One of the findings of my company’s Potentiology research showed that the best approach is to put leaders in situations where the level of discomfort is ratcheted right up, where the reality of their current capability stares them in the face and compels them to confront themselves and increase their self-awareness. It’s an approach we use because leaders can see, starkly, the impact of their actions and their effectiveness relative to those who are truly world-class. As this occurs, and as they start getting better results from trialling different approaches in a challenging but safe setting, they become more self-aware and much more willing to change.

Of course, there is still a "bottom 2%" who are unmovable. One shouldn’t jump to diagnoses about individuals, but as a group, such people are often described as having character disorders. They simply have such low self-awareness that they have no foundation for increasing it. In those cases, you may need a different and more definitive solution.

(Anthony Mitchell is the co-founder and Chief Potential Officer of Australian strategic leadership firm, Bendelta)

Read more posts on Business Insider Australia »