Skip to main content

The best policymakers are systems thinkers - here's how to get started

Posted by: , Posted on: - Categories: Learning
Policy makers using systems mapping
Policy makers using systems mapping

I’m a systems engineer at heart but also passionate about design. Systems thinking has grown on me over many years of practice and research and I now do all that I can to promote it. 

I’m often asked to define a ‘systems approach’ and where to get started with systems thinking.

Forget the textbooks. Don’t worry too much about the theory. I’m not saying these things aren’t valuable. I’m simply saying that they can wait.

Start with some simple questions 

I recommend that you start by asking some questions (which, to reassure you, are founded on research, many hours of thinking and theory). 

In my opinion, the best and fastest way into systems thinking, into any discipline for that matter is to ask questions.

No academic qualification is needed to wonder. Anyone can ask questions and doing so can reveal more to you than the insights of one academic or twenty books.

What you learn from the answers is not only more likely to stick with you, it will provide you with a much fuller picture … and many more questions.

Here are some questions I co-designed when working on Engineering Better Care with the Royal Academy of Engineering, The Academy of Medical Sciences and The Royal College of Physicians. They are focussed on the key perspectives of people, systems, design and risk.

Questions to ask

Asking these questions in an appropriate, yet flexible and repetitive order constitutes an ‘approach’, but I often don’t tell people that I’m using a systems approach because it can feel alienating, something only a specialist can use. 

Slipping the questions naturally into a conversation reveals much about a system and typically identifies numerous opportunities for improvement. If I had to pick just 4 to start with, I would ask the following:

  • Who are the stakeholders?
  • What are the needs?
  • How can the needs be met?
  • What could go wrong?

Don’t get hung up on a common language 

In some situations, a common language it’s vital. When a surgeon has to remove an appendix, it matters quite a lot that there is an agreed understanding of the words ‘remove’ and ‘appendix’.

When it comes to systems thinking, my answer might surprise you: it really doesn’t matter. I’m being provocative but here is a true story to illustrate my point. 

Years ago, I was invited to join a team who wanted to explore how patient safety could be improved using systems thinking. The team comprised medics, a designer, and me, a systems engineer. 

So far, so good. 

The practice of agreeing definitions at the start of a project is something of a habit (particularly for academics) so that’s what we did. 

There were moments when the designer and I almost came to blows because we couldn’t agree what ‘design’ meant. 

Design to the designer meant asking the questions: how does it look? how does it delight? what does it do to the senses? 

Whereas design to me, an engineer, it meant asking the questions: how does it work? how can we ensure its structural integrity? How can we manage the risk in its development? How do we minimise the cost?

In short, ‘design’ was both a noun and a verb. Or put another way, it was an outcome to the designer and both behaviour and a process to me. 

Ultimately, it didn’t matter that design meant different things to each of us. What mattered was that we finally understood what the other meant when the word was used. 

That’s a huge part of systems thinking and there are lots of brilliant examples of where different views on design coalesce and result in the creation of beautiful, functional things.

All too often I see people getting hung up and wasting time trying to agree on a common language. It is far better to spend your time understanding the points of view of others in your team. It’s better when there are a plurality of views. That’s systems thinking. It’s good practice in life generally to try and see things from multiple perspectives. 

This is something I embed in my team. We don’t need to have a common language - we just need to be good at translating. 

Turn the idea of constraint on its head

Something I’ve observed, particularly when it comes to designing policy is that people see the barrier around the thing they can control but that isn’t the same as their boundary of influence. 

For example it’s common to hear ‘As X, We can’t do Y because of Z’. 

I suggest we turn that idea of constraint on its head. Instead of saying ‘I can’t do this because…’ ask yourself ‘What is the ideal policy or behaviour that we want?’ What does ‘good’ look like?

It’s a good place to start when a new team comes together. I encourage policymakers to use this technique because putting too much of a focus on constraints crushes you before you get anywhere close to delivering better outcomes.

So when it comes to designing policy - do notice when you feel barriers and boundaries are, but see if you can go beyond that and then, and only then, think about what is pragmatic. Approaching a solution from beyond the boundary brings a rather different perspective than trying to reach it from within. Dream a little before the reality kicks in.

The benefits of systems thinking in government 

I’ve worked with teams in central government for many years and I thought it might be useful to share my ‘outsider’ view.

You might wonder whether it’s possible to be a systems thinker in government. The institution of government is designed in a departmental model and this has understandably led to silos. I would argue that effective systems thinking is possible even in a silo.

You can influence the system just by understanding your area better. You can do this by understanding where you sit in the system, who is above you, who your counterpart is in a different department. Siloed or not, systems thinking at the very least, can help you determine who you should be talking to.

At the moment I’m working on just that. I’m part of a team tasked with understanding how parts of government are connected around a particular topic. We’re asking policy makers to map areas of influence to policy to determine where conversations should be happening. 

It really does seem to me that the best policymakers are already using systems thinking in government. It’s the future. It’s important to develop a class of people who understand systems thinking and are given the opportunity, encouragement and protection to use it. Hopefully by writing this blog post, I’ve made it a little easier to get started.


Sharing and comments

Share this page


  1. Comment by Joseph Callender posted on

    I've been thinking lately about how government could actually be the systems thinking body for all social systems / problems. If you imagine all of the social forces as silos (economic, health, logistics, etc..) instead of industries as the silos, who is minding the system implications of these social models? No one. And it shows.

    In a sense, move the silo considerations up to the wider, emergent, system implications level to which they are more closely related.

    I imagine a systems thinking government body would serve to weigh decision-making beyond specific special interest or industry biases.

    • Replies to Joseph Callender>

      Comment by Dr Tim Woolliscroft posted on

      I applied a systems thinking approach in my PhD, and I fundamentally disagree with much of what is advocated in this article. The problem is that there are many different forms of systems thinking, some of which are philisophically opposed to each other. What is being proposed here is systems engineering. Like many systems thinkers I find this approach problematic as it risks not actually getting to the real problem or problems and as such risks failing to reach a solution that will be useful. Contrary to what is advocated in the blog I argue that it is essential to agree on definitions, failing to do so risks a lot of confusion and misunderstanding. Such misunderstandings can result in any proposed solutions failing as they are built on sand.

    • Replies to Joseph Callender>

      Comment by Joseph Callender posted on

      Dr Wooliscroft,

      I do not get that impression from this. To me, it portrays a very real "1,000 ft view" (Sorry, American still stuck in imperial measurements) about systems thinking.

      While I get the need for some standardization of terms and methods, to me, a systems thinking mindset would never come to a problem with a predetermined set of methods, techniques, or assumptions about what will dissolve the issue...if that is even possible. In comparison, it's what much of the agile world got wrong, leaping to the methods and concepts without the prerequisite mindset. I would suggest that much of it attempted to "engineer" an agile culture.

      I walk away from this article with a strong sense of systems thinking being proposed / applied.

  2. Comment by Geoff Elliott posted on

    John, if i put my systems engineering hat on I agree with. If, however, I put my wider systems thinking hat on there are several points I would disagree with. Firstly I would argue there is no such thing as The System. Too paraphrase Checkland. I spy systems which I can reenginer. Or , I spy systems as mental constructs which I use as a learning process. Engineering often start with a conceptual model which doesn't necessarily exist as a tangible entity. A system is a personal construct dependent on where I choose to draw a boundary, include or exclude things. I agree with you it is useful to be a systems thinking when setting policy. The real challenge is cascading policy from S5 to S1 an operational level where work is carried out. This very often a major failing in government.

    • Replies to Geoff Elliott>

      Comment by John Clarkson posted on

      Geoff, I agree - there is no such thing as The system, we all have different perspectives on what it might be - the key is to have flexibility in the description of any boundary or any model, and to iterate constantly to update such views. For me, it is not really the model or boundary description in itself that matters, rather the journey that the team take in creating or discussing these issues and the understanding and consensus that this creates. Personal views when shared are useful as they help to articulate different perspectives. In the same way, models, real or conceptual, can add value to such a discussion before reality sets in.

      • Replies to John Clarkson>

        Comment by geoff elliott posted on

        Hi John, I agree its about the journey. Having "messed around" in ST for over 30 years across the world many people view, for example, ST just as Systems Dynamics (Senge et al) or say Systems Engineering. Perhaps more worrying is that we are now in the age of the instant "guru" where many people are just using surface language without bothering to explore deeper. Too me ST has always had 4 parts. 1, Different writers. 2, Underpinning concepts and ideas, there are about 40 or so? 3. Different approaches, there are about 15/20 ish. 4. non syntax sense making diagramming methods used to explore and communicate problematic situations. Over the past few years I have been told. Deming invented ST. Asked for ST method/methodology? And also been a variation on PDCA the and the ILO (international labour organisation Geneva) method study is ST. The words method/methodology are interesting. We have had over the past 40 years every consultancy, IT company, and accountancy practice offer their methodology/method as a solution focussed risk free process. In other words, their USP. Many people belief ST is a methodology/method which is sequential process needed to deliver a solution. I agree the journey is important

  3. Comment by George McConnell posted on

    Like Geoff I agree with much of what is said, but would echo his caution with regard to the intersection between systems engineering and systems thinking. Whilst there are, without doubt, many systems engineers who are excellent systems thinkers - in my experience the majority get 'stuck' on the engineering. Which is good - in SOME domains - where the goal is to create tangible systems.

    It can be much less helpful where the "system" that is under consideration is more peopleware than hardware and software. That must surely cover the majority of government. As Geoff also rightly says it must always be in the forefront of your mind that the system boundary that you have chosen is pretty much arbitrary - and probably wrong!

    I would agree with the point about a common language - but rather than just being "translators" (which is necessary) there needs to be a striving for a common understanding - a way of representing "the system" that everyone involved can relate to and - most importantly - does not contain any aspect that is "understood" by different people in different ways. Therein lies the root of many problems - "oh, I thought you meant....".

    There is one other thing to be borne in mind - in my experience not everyone is equipped to understand systems concepts - many find it hard to go beyond immediate cause and effect; or to imagine the result of feedback loops; or fail to understand that in the sort of systems that we are discussing here there are often "competing" systems as well as "co-operating" systems. Those people need to be able to contribute their knowledge to the conceptual model even when others provide the linkage from their part to "the whole".

    In the end the 'model' will illustrate how the silos interact - most importantly how they interact indirectly. It should also illustrate where they 'should' interact but do not. I completely agree that the understanding of systems concepts is essential for government - and looking from the outside I can only see a considerable lack of evidence of this happening just now.

    • Replies to George McConnell>

      Comment by John Clarkson posted on

      George, agreed - although systems engineering as a disciple has broadened its outlook considerable over he the years, it has its limits and I would propose a much wider perspective that borrows much from social science and design thinking. I also think that it is the discussion of the boundary that is more useful than the description in itself, although I would always argue that it is wise to take a broader view where possible, accepting that some limitation of scope is also wise - iteration and flexibility is key.

      Common understanding is particularly important - the ability to translate only truly comes if we understand not only the words, but also their meaning in a particular context. I would advocate the identification of a few specific features a system must (or must not) have that everyone can agree on and relate to their own worlds.

      I also agree that systems concepts can be difficult, even for engineers, and would suggest that systems leaders, or systems facilitators, can add particular value in this respect, guiding a team to take a more effective systems approach and translate these strange concepts into practical and insightful activities and discussions that will lead to better outcomes. This includes capturing and collating local knowledge as a means to discovering and highlighting 'hidden' connections across the wider system landscape - if only to suggest where there might be unforeseen challenges to be discussed or investigated further.

  4. Comment by Sally Bean posted on

    I agree with Geoff, and I think it would be a huge mistake to ignore the textbooks. Learning for Action by Checkland and Poulter is a concise, accessible book which clearly explains the fundamental difference between systems engineering and systems thinking and provides some examples to illustrate this.

    • Replies to Sally Bean>

      Comment by Pauline Roberts posted on

      I agree with Geoff, George and Sally. The power from systems thinking comes from the mindset adopted when undertaking such work. This is informed by the theory and systems thinkers undergo a continual theory informed practice which informs theory....... If you arent engaging in the theory then you move forward with the same mental models and patterns of thinking and miss out on the powerful change that can be achieved.

    • Replies to Sally Bean>

      Comment by John Clarkson posted on

      Sally, I could not disagree with your comment - these books are excellent in articulating an important difference between systems engineering and systems thinking. For me personally, systems engineering has been influential in bringing discipline to complex technical projects, but does not go far enough in getting to grips with wicked problems. I also find systems thinking too restrictive a term and prefer to talk of a systems approach, which borrows ideas from a much wider range of disciplines with a light tough that does not exclude the use of the underlying concepts, but hopefully adds new perspectives.

  5. Comment by john mortimer posted on

    Delving into the public sector, with systems thinking in your back pocket, is bound to be interesting, different, and a challenge. But having a go at it is a good place to start, together with learning from those who have gone before. There are many that have come in as systems thinkers, to do something.

    I have found that learning from those systems thinking initiatives that have gone on before will give a leap in learning. Link up with and create a network ion learning. As well as beating a new approach.
    My main comments would be around:
    1. What problem are you trying to solve? Why are you wishing to look through a systems thinking lens? And the answer might just be to explore to see what we can find.
    2. Use a systems thinking approach if you wish to affect real changes and that approach should be combined with whatever disciplines are needed; e.g. design, change, behavioural change, etc.
    3. Find a problem, and apply yourself. A good guide to succeeding in central gov is Digital Transformation at Scale by Loosemore. If you don't pay too much attention to the Digital part, the rest of it is exemplary, and proven.

  6. Comment by Michael C Jackson posted on

    I would add that we do need theory, along with practical experience of course, to understand that different systems approaches see the world differently and recommend intervening in it according to their particular perspectives. I think this account of systems engineering is quite sophisticated, but it is still SE, based on particular limited assumptions - effectively that agreement on a goal can be reached between stakeholders, the system modelled, and the best means of reaching the goal determined. It will struggle with ‘wicked problems’ where these assumptions do not hold. Other systems approaches are based on very different assumptions about the nature of the world and how to intervene in it e.g. VSM, SSM, CSH. Critical systems thinking argues that, since it is impossible to understand the ‘whole system’, it is useful to have this variety of viewpoints and ways of intervening in complex problem situations. We need them all. And, when possible, should seek to use them together in the same intervention to address the widest range of issues. To do this, we have to understand their different strengths and weaknesses. Theory, along with learning from practice, can help achieve this.

    • Replies to Michael C Jackson>

      Comment by geoff elliott posted on

      Hi Mike, …..Other systems approaches are based on very different assumptions about the nature of the world and how to intervene in it e.g. VSM, SSM, CSH..... I totally agree. If I can extend the discussion. One area where different systems approaches used either singularly or in combination is managing change. Unfortunately manging change in the UK is perceived rightly or wrongly as some form of cultural change which to me has always been an outcome of making an appropriate intervention into an unbounded problematic situation. That is, culture change is an outcome not a starting point. Too my knowledge the HR and Management Professional Bodies don't recognise or understand the role different ST approaches and the underpinning concepts and ideas play in managing change. Perhaps this an areas which needs to be explored - thoughts?

  7. Comment by Kevin Baylis posted on

    I remember getting invited to a government sponsored systems thinkers workshop on why change programs for the NHS fail. I think the final consensus was that they hadn't all failed, however they never had a chance to bed-in before a different change buried them. Not sure anything improved with too many political leaders in too much of a rush to employ evidence based approaches?

  8. Comment by Mark Parker posted on

    First of all, let me say this is a very good post. Systems thinking has physical engineering or technology as the tip of the iceberg. The silo’s (key constraints) are generally routed in organisational culture, reinforced within departments. Executive Directors quite rightly make the big decisions. Then on big projects systems thinking will kick in with stakeholders being ‘collected’ for the project. The weakness in this approach is that you are always starting from scratch. In a number of the organisations I have worked, Direct Reports to Directors (Heads of Dept) rarely meet each other formally, cross function, to discuss operational value chains, except when problems arise, which need resolving. Yet this cadre run the teams that run the operations on a daily basis. They feed their facts and figures for their departments to their Line Director, and they understand at an operational level how everything works. My point is, if systems thinking permeates the culture top to bottom, then changes and tweaks drive continual value. The foundation for larger change, when the need arises is already there. It’s lean and it drives continual improvement.

  9. Comment by Joseph Callender posted on

    There's a concept called Suspended Activation. It has 3 steps.

    1. SUSPEND what you think you know / understand

    2. REDIRECT to new possibilities, ideas, info, etc...

    3. LET GO of determining a solution right away

    It doesn't suggest never experimenting with a solution. It suggests never be too married to any single "answer."

    To envision it, I imagine it as handling any knowledge, ideas, or potential solutions lightly, never holding any single one so tightly that that is all that gets deployed. If you are familiar with the Globetrotters, imagine ideas and answers as the basketballs they handle. They maintain a very light touch so that the ball is easily transitioned to new stages.

    Given the inherent complexity of systems, approaching a system with any single set of ideas is ill-advised.

  10. Comment by Al posted on

    It is extraordinary that nothing is said on this site about John Seddon’s work. To my knowledge the only example of systems thinking that is delivering profound results. I think he would say all of what’s here is interesting rather than useful.

  11. Comment by Harald Kreher posted on

    Mike and Geoff have made some comments I would largely share.

    Maybe THE aspect to start, though, is around these questions:
    - what is "best", who "policymakers"
    - who defines the above
    - by what criteria

    As a pragmatic systemist, I, of course, would tongue-in-cheek agree that ST has some of the very best to offer. Well aware that the definition of ST itself is far from unanimous. So, what, I call it richness, requisite variety ... fun.


Leave a comment

We only ask for your email address so we know you're a real person

By submitting a comment you understand it may be published on this public website. Please read our privacy notice to see how the GOV.UK blogging platform handles your information.