There’s quite a bit of research in this blog, but it’s not cited. If you want references or want to search for the sources of the data, drop me a line and I will do my best to find them for you. As usual, there’s quite a bit of culture stuff in here, but it wouldn’t be one of my blogs without that
I started working on Leadership specifically on the College of Police Leadership Review. Prior to that I studied it at Warwick for two years, and I’m back into studying it now. It’s a complex subject, and usually unfairly reduced to soundbites and catchphrases that people throw around with abandon. Classics like:
“Many hands make light work.”
Go hand in hand with:
“Too many cooks spoil the broth.”
It doesn’t take rocket science to realise that these clash quite significantly. This is true of much of the leadership literature. You can find books about keeping ‘grip and control,’ and place them directly across from books by leadership consultants/gurus like Simon Sinek. Sinek will talk about empowerment, and finding meaning, whereas the grip and control gurus will talk about managing process and keeping tight audit.
This clash makes the whole area of study contradictory, and there are few studies that actually begin to discuss how complicated the area of leadership actually is. When you dig underneath the myriad of discussions about leadership ‘styles,’ you realise that they are collections of lists of behaviours. These behaviours have been found to hold value when researched in particular contexts, but we don’t actually know if it was the context that made them successful, or the behaviours themselves…
Let me give you an example.
An organisation calls the researchers in because the staff survey says that working in that company at that time is pretty awful. They may like the work, but they don’t like the leaders, and they certainly don’t like the environment and culture. Unsurprisingly, a particular set of behaviours is absent, and the workplace needs them to begin to function in a balanced way. So… a really heavy command and control culture, is likely to need empowerment and far more communication up and down. The company needs them, because there has been a culture in place for some time, and that culture has self propagated behaviours in order to function in a particular way.
This happened with policing during the performance era. Performance became the by-word for success, and if you delivered performance you were held to be a ‘good’ leader. Those leaders then entered into positions where they got to select other future leaders, and as performance had always been important to them, it became important to other aspiring leaders too. When this begins to happen, you see a distillation of behaviour. It becomes palpably strong, almost themic.
The trouble is, the police is a pretty diverse organisation in terms of function. You really shouldn’t have command and control permeating through every department in the police, it’s damaging. You need different leadership styles in neighbourhood policing, CID, community cohesion etc.. As Command and Control remained the dominant leadership style, so it did grow into an unstoppable force, and it still remains immensely important in policing to this day.
Command and Control is a controversial subject. Some leaders will swear by it. They call it by all sorts of names and usually attach ‘Operationally Credible’ to its use. The demand stats show that the kind of work that needs it is actually really, really rare, yet it’s sticking around with a vengeance up and down the country. There are good reasons for this, as if the Command had been better at high profile incidents where there has been loss of life, the outcome may have been very different. If you screw up when you need it, the outcomes can be pretty atrocious.
So, what have we got so far? We have the need for Command and Control in fairly rare incidents, but it is of very high importance. Do we need it everywhere? Clearly not. Yet, as with many things in policing, the broad brush is applied, and usually all candidates will have to prove their ability in Command and Control during promotion processes etc.. This results in a very specific set of behaviours being applied as a filter through to areas of leadership.
And, I need to add some comment in here about measuring Command and Control (or investigation, or road traffic incidents etc. – whatever you choose to measure as a filter). If you are being measured on whether you reach the same decisions, as those that designed the exercises, then you are only being measured for whether you think like them. This is designed in group-think. Decisions have to be able to be made that are different to the ‘accepted solutions,’ whilst you measure how candidates got there, instead of where they ended up (this is a whole other blog!). This is Diversity of Thought, and you can allow for it in measuring candidates.
Want to be a leader in the police? Best be able to use Command and Control – and that’s OK, because we need it (in places).
Now, some people are not good at Command and Control. They have a different skill set, and may instead prefer leadership styles such as Collective Leadership.
What happens if someone who prefers Collective Leadership gets put through the Command and Control Assessment? Well, they won’t perform nearly as well as those who are naturally really good at Command and Control. In fact, they may be filtered out here, and all because their strengths in leadership lie elsewhere. It’s kind of like a control mechanism for a particular kind of leadership. It also significantly disadvantages leaders from departments who need a different leadership style – such as people with skills in forming partnerships and collaborations.
It establishes a behavioural hierarchy, and it attaches status to particular behaviours. If you don’t see assessments during promotion that look at problem solving, you won’t attach value to the behaviours that go along with good problem solving. It’s strange that you will see lots and lots of examples on application forms that detail dealing with public disorder or match command, but you rarely see them discussing forming strong links with local charities…
Back to measurement.
Hopefully the example discussed above sets a scene. If you have a strong kind of leadership in place, that leadership will often set the bar for selecting other leadership, and unless they are enlightened enough to look for challenge in their future leaders, they will select people that they are comfortable with. There’s loads of research on this, but here is a very quick run down on how good unstructured interviews in isolation are for selecting future leaders (varies depending on the study, but by and large, they are rubbish). Add to this the other Police favourite of the application form, and you have a combination that looks like this:
- Exercises that create behavioural bias in selection (such as command and control scenarios as a filtering method).
- Use of application forms (pretty awful in terms of reliability and validity).
- Unstructured interviews (again, fairly awful in terms of reliability and validity due to the fact that they are so subject to bias).
The resulting picture basically shows (and if you want a run down on how good particular exercises are for selecting good leaders, have a look here) that the traditional methods for leadership selection within constabularies up and down the country, are riven with bias, unreliable, and often invalid. These are technical assessment/selection terms for being poor quality.
A quick example of the hierarchy of exercises:
So, I hear you asking, if the current methods are pretty awful, what can you replace them with? Well, actually there’s all sorts of things that we could do better, but as an ex senior officer pointed out on Twitter last week, Assessment Centres were stopped in many constabularies because the senior leaders didn’t like the outcome of them. This stands to reason, because assessment centers often look at the whole of the candidate, and single exercise filters are forgone for several exercises that measure the a much wider picture. This means that some Command and Control heavy senior leaders, will be presented with successful candidates that do not have anywhere near as much of a Command and Control outlook as they do – thus putting them out of favour, or making it very difficult for them to succeed.
In short, solid assessment centers with a wide array of exercises, are more likely (certainly not certain) to measure wider communication skills, the ability to process information, and the ability to display empathy/values. They are more likely to select good leaders.
We can go even further into the building blocks of leadership, and look instead at things such as intelligence, emotional intelligence (how we interact with other’s and our own emotions) and cognitive load (how much info you can take in and process). This is relatively simple stuff, but it removes the behavioural bias that is built into many of our internal selection processes now. Serving cops often don’t like discussing measuring intelligence, they also dislike written tests, and they hate psychometric tests (I have direct experience of the use of all of these methods). All of these are far more valid and reliable than current methods though. How do we know? Huge amounts of research tells us, and it tells us convincingly too.
So what gets in the way of better selection and measurement? Well, the truth lies in something that is called ‘face validity.’ This is fancy wordage for how the culture sees the exercises. If the candidates do not feel that the test measures ‘being a good cop,’ then the outcome of the results will often not be trusted or given any value. This makes improving selection fraught with leadership challenges. If you introduce psychometrics as a leg of your assessment center, there will likely be a backlash by candidates who will tell you that they mean nothing and are a total waste of time. The research tells us different, but the people’s perceptions of doing the tests are clearly immensely important.
So, how do we navigate such a complicated landscape? The answer? With great difficulty and lots of iterative work – this means that things will fail, you learn from them, try it again with tweaks, and then learn from the second application, and so on. Eventually you end up with a product that the culture has begun to accept, that does a far more improved job of unbiased selection.
If you made it through to here, then you are doing well, because that was quite dry. What are the take-homes from the above?
- The current way we select people internally is hugely prone to bias and unreliability.
- There are many better ways to select out there that have been researched and tested repeatedly.
- These methods may be really counter-culture, so implementing them may be very difficult.
So, finally, what about the ‘Holy Grail?’ Well, nothing is perfect, you could search for the ‘right’ kind of selection exercises for a millennia. Will we ever find them? I don’t think so. I don’t think that searching for the perfect set of selection exercises is actually a worthwhile exercise. Why? Because it is all contextual. It depends on your environment, on your people, and on your culture. What is true however, is that better selection is not only important in order to find future leaders, it’s also immensely important to find different leaders. We will never find the hallowed ‘diversity of thought’ without doing a full reset of how we select our future leaders. Behavioural filters are used almost everywhere, and quite simply, they institutionalise behavioural bias. They’ve got to go, or become part of a suite of exercises that sees far more of the candidate. This means going from a past-the-post system (where each exercise is an opportunity to filter candidates) to a holistic system where a selection of exercises are used to see a wider picture of the candidate as a whole.
Dilbert (as usual) captures the bias that needs to disappear very nicely:
Nothing like a good challenge
Above all, there’s a duty to the people that work in the police, and to the public, to use the best methods that we can with the resources we have. That way, the likelihood of selecting the best leaders for the future will improve, despite the heavy cultural challenges that we may face. And just to finish… the above methods that have been discussed, don’t bring in the Values that put those candidates in the chairs in the first place, and if you don’t measure for that, it just may be possible that you don’t get the leaders that you would like holding the leadership positions… there’s work to do here, let’s get on with it
