How to change a system from the inside

Link post

The linked post is focused on the UK civil service yet I think much of it would be applicable to people in other contexts, organisations and countries (e.g. staff at large AI labs, financial institutions, other governments, etc). I find this an interesting topic and I hope this will inspire others to be optimistic about improving systems. Below I have copied the original text, removed the Civil Service jargon and made a few other tweaks.

***

HOW CAN YOU IMPROVE THINGS WHEN YOU ARE A JUST ONE SMALL PART OF A MUCH BIGGER ORGANISATION?

Working in policy teaches you to understand systems. But it also involves working in and being part of a system: a national government. And as you spend time in that system you may start to notice all the little cracks, all the ways it does not quite work or can lead to bad decisions being made.

I joined the UK Civil Service in 2012 and I started to notice the cracks. I noticed there was a lack of class diversity in the Treasury, with good staff from poor backgrounds being pushed out. I noticed how Overseas Development Aid funds received scrutiny and oversight when being spent by the Department for International Development (DFID) but not when being spent by other departments. Yet now, a decade later, these things are changing. The Cabinet Office has started tracking departments’ class diversity statistics, and the Independent Commission for Aid Impact increased its scrutiny of non-DFID aid spending (before the DFID merger with the Foreign and Commonwealth Office). None of this because of me: some people, somewhere, were seeing the problems and fixing them.

Sure, sometimes it is two steps forward and one back. But our system gets better, in both big and small ways. We have a better functioning government and Civil Service in the 2020s than we did in the 1920s, and we will be better still in the future. It is improving because people are seeing and fixing the cracks.

And you, in your organisation, can similarly play a part in fixing the cracks.

You are most likely allowed to try to fix things, more than you realise. As with any change, you may get pushback. But at some level many big bureaucracies do allow change and will champions those who genuinely try to improve things, to innovate, to make workplaces function well.

Furthermore, trying to make such changes will in many cases be good for your career. To make change you will likely be networking with very senior stakeholders, building skills and experiences and developing stories that will make you sounds great in future job interviews.

Story: After I graduated from pharmacy school I went in to my first job as a very junior intern pharmacist at St Vincent’s Hospital … I started trying to figure out how I could improve the workflow of all of the other people that I worked with … I figured out that pharmacists were spending 30% of their time on these kind of ad hoc drug requests which meant that that was time that they weren’t spending with their patients … So what I did was to try to link hospital admissions data to the pharmacy dispensing system so that we could predict how much we would use of various different antibiotics or other drugs at different places in the hospital at different times and make sure that those lifesaving drugs never went out of stock … My colleagues at the hospital were all astounded at what I did and they were like, “This isn’t your job, why are you doing this? Why are you putting in so much effort? Who is allowing you to do this?” And the thing that I said is, “You don’t need permission. You don’t need to be allowed to do something that’s not in your job description if you think that it’s gonna make your company or your organization more successful and more efficient, you can often just go and do it.” (Source).

About this guide

This is a guide on how to improve the system from the inside. This could range from changing the rules and protocols which your organisation uses to function and make decisions, to changing the culture and ethos of your organisation and the way things work in practice.

This guide was originally written by UK civil servants for UK civil servants with a focus on improving the UK Civil Service. This version has been adapted for this Forum, as it is possible that much of it would be applicable to other contexts, other organisations and other countries. We are confident that this represents a reasonable approach within the UK government, however those applying this guide in other contexts may want to approach these kind of actions with caution and an awareness that rocking the organisational boat may have risks to their career

It is for staff working in any area of a large organisation (not just for those few staff who happen to have organisational design in their job description). It is primarily written for staff in the junior to middle manager range (not for Directors etc.).

The guide is split into 6 key steps (each split into 2-5 substeps):

Depending on what you want to change, not every step is always necessary. Nor do the steps need to be carried out in the order given. Sometimes after understanding a problem (Step 1) you can just fix a problem yourself (Step 4), and sometimes you need collaborators (Step 5) to work with you. For some changes senior staff buy-in (Step 6) is not necessary yet other times senior staff buy-in to the change is the change, and once they have agreed then the problem will be fixed.

There are of course limits to what you can and should change working from inside the system. These are discussed at the end of the guide. See:

Limitations

This guide is about improving the functioning of organisational systems, not about trying to influence an organisations outputs (e.g. not on how to change policy by working in government). Staff in big organisations can definitely have a direct positive effect on their organisations outputs (e.g. policy) though their work, and how to do this is covered (at least from a UK policy angle) in the Impactful Government Careers Network separate guides on Where to work in government and How to have an impact by doing your job well in government.

Note: stories are anonymous and from a variety of sources.

Step 1: Understand what needs to change

Take your time. A significant chunk of time should be invested in understanding the problem. Do not try to change things immediately after beginning a new job, or starting in a new department. If you have just joined the organisation, give it 6 months. You’ll need this time to understand why things are the way they are and the written and unwritten workplace rules, and to decide which problems should be fixed, and what solutions should be tried.

Understand the systems you are in

Be observant. Keep notes and records of things that are or are not working for you or other people. Talk to your colleagues, ask them about their gripes and what they would change. Think about where you can add value.

Learn relevant topics. Learn about systems thinking, organisational design, diversity and inclusion best practice, and so on. Read around these topics and try to find examples of best practices and well functioning systems that you can compare your workplace to.

Prioritise. Brainstorm ideas. You may end up with a very long list of things you wished would function better. Think about the things you want to change and prioritise between them. Make sure the thing you are working on is the highest value thing you can focus limited energy onto. Your top priorities might be the changes that you think would have the greatest impact on the civil service and society at large, but how tractable and achievable the changes are may also be relevant. (For more on this see the section on prioritisation in our career guide).

Story: Every 3 months I would take a few hours out, away from my desk, and list all the things I was working on at the minute. I included both the day to day work and the corporate objective projects I had. I would then try to ignore the day to day pressures and just prioritise them by which I thought would be having the biggest impact on the world. This was helpful for me to adjust my time towards what mattered most.

Understand why the system is not working and why it has not been fixed.

If it looks like a system is broken, maybe it is, but maybe it is actually that way for a good reason. (See: Chesterton’s fence). Remember not everything that looks broken needs to be fixed, and not everything that needs to be fixed is for you to work on. Before trying to change something try to understand:

Why the system is the way it is
The good reasons, the legitimate reasons the responsible staff have for not fixing it
The bad reasons, the misaligned incentives that may stop the responsible staff from fixing it
How it may have already been changed and why that might not have worked

This will help you decide if this is something to fix as well as help you understand (and later influence) the incentives of those who might have a say in fixing it.

Have evidence. If there is a problem you want to fix, make sure you have evidence that it is actually a problem, that it has some cost to the organisation, to public or to staff. Collect data and anecdotes. Also seek evidence that solutions are possible, such as examples from other departments or other countries or external think tank research. (When searching for solutions, remember that money is guarded and it is often easier to affect processes rather than funding.) And once you start changing things, track impact – collect stories of how your change helped save money or support staff and so on. Evidence will be particularly useful for persuading senior staff to engage with the problem.

Story: When I joined the Civil Service I was interested in encouraging people to give more to effective charities. I joined an existing group focused on workplace giving. I remember being embarrassed at a meeting by suggesting that we make charitable giving opt-out for all staff, an idea that would clearly never fly. After that I kept my suggestions to myself for a few months, volunteered to help others, and worked out what practically I could change. A while later, when I volunteered to find ways to make Payroll Giving easier and to put advice on the intranet on how to give effectively, everyone agreed and encouraged me to make those changes.

Step 2: Understand how the system is changed

Volunteer. If new to the Civil Service or a department, spend some time with the people there who are already changing things in ways seen as legitimate. There will always be such people doing volunteer corporate stuff slightly outside the standard day to day jobs. For example, you could volunteer to help with:

LGBT networks, Capability Action Network, or other diversity groups
Carer’s network or other support groups
Organisational changes, such as the panel of people smoothing the implementation of the new IT system, or similar.
You could also try to work in roles that involve setting up new teams or otherwise making planned organisational changes.

Identify precedents. Look at how change has happened in the past. Look for groups or people with unusual roles and work out how they came to be. Talk to your department’s Innovation Team if there is one. Learn generally about past reform of your organisation, how change has happened, and what has worked and not worked to date.

Story: I wanted to talk to other civil servants about how to have more impact in our careers. I had been going to the Civil Service weekly Jewish ‘lunch and learn’ events and thought that they created a good network. So following that precedent I decided to set up a similar lunch and learn event for civil servants interested in Effective Altruism.

Step 3: Create the space for action

Make sure you are doing your day job well. Your day to day work still matters. This might be the biggest thing that will get in your way – if you cannot do your day job to a reasonable standard then your manager or someone else is likely to block you from driving other changes. Furthermore, performing badly on the day to day work will reflect poorly on you and the others who are trying to change things with you.

Add the change you are working on to your objectives. Make it part of your day job. Check what precedents there are for this, maybe there is a formal 10% time for you to use how you see best, maybe it is common to add things like office recycling champion or fire warden to peoples to do lists or objectives, maybe you just think your manager is likely to support you trying to improve things. Then if you think there is a good precedent and if there is something you want to change, broach the subject with your manager and make the case that it should be added to your your objectives or your to do lists or your personal development plan, or similar.

Story: I wanted there to be better career coaching for junior civil servants. I started looking into coaching techniques, and began running career coaching sessions for people I knew, and then for others. I talked to my manager about this and explained it would be good for my career. He not only agreed to add it to my corporate objective but also said that the team training budget could be used to send me on a course on coaching skills.

Step 4: Fix things that you can fix

Start to fix things. At some point you need to start changing the things you can change. Start small if you need to: lead by example and change things in your team or area of work. You can always reach out to other staff or senior staff later to spread the change.

Ask for forgiveness not permission. Before taking an action, carefully think through the consequences and risks. But if in doubt about whether you can or should do something, take action! Sometimes if you try to ask for permission you will be stuck in limbo for ages as you wait for a response from people, or they may be risk averse and tell you ‘No, let’s stick with the status quo’. Yet no one wants to be the person to tell someone that they should not have done a good thing, after they’ve done it! You have more autonomy than you realise.

Story: A civil servant wanted to make her department work better. She spent a while talking to other staff to work out what bothered them most and what could be fixed. One thing that came up is that new staff got lost or confused when trying to find other staff’s desks in the broad open plan office. So she thought of a numbering system for the desks, got hold of a labelling device and went around and labelled the desks with numbers. She did not ask permission first, she just went out and corrected a problem she saw.

Step 5: Work with others – build a group of collaborators

Network. It is easier to change things or to spread your change more widely across your organisation if you have collaborators. Go to networking and social events; get to know your colleagues and make friends at work. For example find other people in the effective altruism community in your organisation or look for events which invite “innovators” or “change makers”. Consider the teams and senior staff it may be good to reach or have support from across the entire organisation (not just your area) and build a broad base of connections with the junior people in those teams.

Network with senior staff too. For example, get to know your Director or find a senior staff member to be your mentor.

As you get more buy-in and build credibility, consider also writing blogs for your organisations intranet (or even for industry magazines or groups like Apolitical) about what you are trying to do. You can also speak at events such as team meetings or conferences, and suggest that people interested in collaborating with you should get in touch.

Build a group of collaborators. Find people you can work with who want to change and improve the same things as you, and build them into a mini-team. Invite them to take part and let their views shape what you are working on. Give people clear roles, responsibilities and actions to take, whilst recognising the time constraints that their day jobs place on them.

Story: I wanted to run training courses for civil servants on how to make career decisions. So I did. I wrote the course, tested it, and ran it. (I did not ask permission, there was no push back, and senior civil servants seemed happy that this was happening). Then I started getting the people who turned up to re-run it in their departments, and so it spread.

Step 6: Involve senior staff

Slowly build credibility. Showcase your expertise: for example, by changing something small that everyone agrees needs to be changed before moving onto something bigger. Toot your own horn: for example, talk about what you have done at a team meeting or write on internal blogs. Use your senior allies to build credibility or amplify your message to other senior staff.

Work out who to talk to. Work out what levers can be pulled to drive change and which senior staff might be interested in the change you are making. This could be people in your management chain, or people with particular responsibilities (e.g. senior executives with relevant remits) or people in other teams or even other departments who could fix the problem (analyst teams, HR leads, etc). Remember that not everyone will have time for you. Look for senior staff with a reputation for being innovative or different (if you are not sure, talk to their junior staff). In some cases you might also consider getting support from trusted credible senior stakeholders outside the organisation, such as ex-senior staff or a well respected trusted external body such as a think tank or industry body.

Consider proposing a process not a solution. People often don’t like being confronted with a solution – especially if it comes from someone more junior than them. They need time to get used to the idea that there is a problem, and to get used to potential solutions. Creating a process avoids asking anyone to commit to anything at the beginning and it only needs a certain amount of recognition that there might be a problem. A ‘process’ might have some or all of: a name, a group of people working on it, possibly some senior oversight, an agreed set of questions to look into, an identified recipient of the findings, a series of meetings over a period of time, possibly some consultation with internal/external experts, and a report at the end with some recommendations. The process itself has legitimacy, it gives people confidence that proposals for change are coming from some deliberation, not off the top of people’s heads and when the recommendations finally come out, the fact that there has been a process makes it harder to ignore.

Sell your problem and process/solution. Understand what the senior staff you are talking to want (if you are not sure, talk to their junior staff). Understand their capability, opportunity and motivation (COM) for acting. Remember to appeal both to senior staff’s incentives for improved systems, but also to their personal (or even misaligned) incentives, such as looking good, selling their work area, or having control over a bigger pot of resources.

Be humble. Remember you do not have all the answers. You may have identified an important problem which, if fixed, could make the organisation run better. However, you still may not know all the reasons for it. It may be helpful to have some suggestions to demonstrate that solutions are possible, but trying to find the solution to a complex problem is almost certainly a joint effort that needs to involve others and their expertise. This likely includes seeking input from the senior staff you are engaging with.

Story: We were running these training courses for other civil servants and wanted to make it more formal and maybe find capacity for someone to do it full time. So, we arranged to be on a panel at Civil Service Live about innovators in government. From there we met a Director who liked what we were doing. We asked and she said maybe her team could support us. She put us in touch with her Deputy Director. We pitched our ideas to her and she offered to help us.

Limitations

It is worth flagging some of the limitations of driving change from the inside.

There is a limit to how much you can change things from the inside. For example in governments big changes to systems need external or political pressure. From the inside you can probably play a role in bigger changes if there happens to be public or think tank pressure or political reasons for the change to be made.

Change is slow. On rare occasions, it’s possible to change things very quickly. But bigger changes can often be very slow. Say you want to be allowed to put 50% of your work hours into running a training project you set up for other civil servants. You might find you need to get sign-off from: your manager, your director, Perm Sec, legal, HR, comms, and so on. This could take months even if it goes smoothly; and in the meantime a change of Director could cause you to have to restart, or even scrap, the whole process.

You might fail. Finally, remember that making changes requires a lot of luck and being at the right place at the right time. You will not always succeed to change the things you want to change, and that is also OK. You will learn and maybe be a bit luckier and more skilled next time.

Story: I had this idea that we could have better Ministers if the annual Civil Service People Survey included questions about Ministers’ decision making, and the answers were aggregated and provided to No.10 so that the PM could make better calls about which Ministers were doing a good job. It might just be possible for a motivated grade 7 with the right connections to drive that kind of change, but it was a bit beyond me.

A note on risks – We are confident that this represents an acceptable approach for driving change for staff within the UK Civil Service. However we have not investigated other contexts/countries. I would encourage cautious action, as I think there is much to be learnt – yet those applying this guide in other contexts may want to approach this carefully, it is possible that rocking the organisational boat may have risks. It may be reasonable to: spend longer on steps 1 and 2, to try to change smaller things first, to not associate your efforts with others who have not signed-up to be a part (e.g. avoid presenting your efforts under the “effective altruism” brand), etc.

Final words

Go create the change you want to see. Trying to improve things can be good for you, your career, your organisation, and the wider world.

I would love for people to try this and see how it works. Some concrete next steps could be:

Meet up. Find 2 to 3 other staff at your organisation. Agree to meet every (or every other) week.
Discuss and learn. Take a month or two on Steps 1-2. Maybe have a complaints session focusing solely on the things that are not going well, then a session on possible solutions, then on comparing and prioritising a few things to work on, etc. And if you find a thing you think should change:
Go create change. Apply Steps 3-6 and see what you can achieve. Maybe start small and move slowly, but do start seeing what you can achieve.
Give feedback. Let me know what worked and help improve this guide.

You likely have more permission and more autonomy to do this than you realise. It will not always be easy but it can be done.

Good luck!!

A final story: As a junior civil servant, an HEO, I noticed our policy team barely spoke to our analysts. We were introducing massive reforms without any performance indicators. I volunteered to come up with the indicators. I created a virtual team of G7 policy people and analysts and got them working together on these measures. The team created a new impact report that went up to the Investment Committee as a pilot, and then got shared around the department as good practice. Other divisions followed suit. Policy people in my division felt more comfortable with the analysts and the culture changed – they started being involved in policy from the start, coming to meetings. Eventually they joined our division, and became a fundamental part of our team and work. I can’t take credit for all that, but seeing a problem and pro-actively trying to change the culture did have knock on effects.