Assessing Internal Police Demand

Transcript of interview with Natascha de Leeuw

SKN

Hello everyone, this is Susanne Knabe-Nicol from Police Science Dr, and I'm doing an exclusive Magazine interview now with Natasha de Leeuw from the Netherlands Institute For Studying Crime and Law Enforcement NSCR, we're going to talk about how she managed to assess internal police demand and what we can do with that information. Welcome to the station, Natascha

NDL

Thank you very much for the invite, Susanne

SKN

Well, thank you for being here today. Why don't you go ahead and tell us about who you are and what you have achieved so far?

NDL

You already introduced me as Natasha De Leeuw, I work at the NSCR, the Netherlands Institute For Studying Crime and Law Enforcement, and I work there in a research group that revolves around evidence-based policing and I'm doing my PhD there and my PhD projects focus on criminal investigations. This particular project I'm talking about today has focused on assessing internal police demand, which has everything to do with the activities being done by the police after the response has been for an emergency call, so after the response to an emergency call, what happens next to actually investigate and close the incident.

SKN

I showed the image of the article that you published recently on this, it's actually an Open Access article isn't it, so everyone can read the whole article, no subscription walls?

NDL

Yes, exactly we're very much pro open science and also trying to get this out there to the practitioners in policing.

SKN

Okay, excellent, can you tell us why it is important for police to know what the internal police demand is? Why would we care?

NDL

I think this is a very important question because we work together with the Netherlands Police, actually, we were asked to join this project by the Netherlands Police, so it was very much instigated from the Netherlands Police themselves, because they felt like they didn't really have a very good grasp on what actually happens in those criminal investigations and how much time it takes. I mean to a certain extent, but if you really get down to the nitty-gritty details of it, what is actually happening and how much time and effort does it take? And it's important to know these types of things because - a little bit of context is the fact that we joined this project with The Bureau For Impact Analysis in the Netherlands Police, and they are tasked with trying to evaluate the impact of upcoming changes in policy and criminal law, so for example if something is going to be changed in criminal law, or policies of police change, maybe their policy to arrest or when to investigate, this team is tasked with trying to evaluate and estimate the impact that will have on the organisation. Because the police only has a limited amount of resources to distribute, so they want to know up front, they want to have an idea of ‘What will this take out of us’, to try and prepare or be ready for this change in policy and criminal law. And sometimes, assessing this impact is also a way of talking back to politics, because it allows the police to actually hand over an impact analysis that says like ‘We are not equipped to actually take on this extra task’ or that this change in policy will mean to us, so that is a bit of context. This was a team that actually contacted us to do this impact analysis for the police to have a good grasp on um what amount of capacity and what amount of demand is this taking out of us? They felt like these impact analyses and these evaluations could be more precise and they could be better, so they actually started this project a couple of years before trying to visit police teams to get an idea of what all the activities are that police actually does after the response to a call for service. Because some of it is registered, but a lot of it really isn't. So, all those little activities such as conferring with the prosecutor or conferring with your colleagues or doing other little tasks and investigative activities in-between, not all of them are registered in the systems. So, they really wanted to get a good idea of what this was and work on a way to better measure this, so this really helps them to assess the impact that changes will have on the police organisation, will allow them to calculate and evaluate the impact with this.

SKN

That's interesting I don't think we have a team or department in the UK that is just tasked with assessing potential changes on policing that result from legislation changes or changes in practice. I think that's probably a very good idea to have. Can you give us just a tiny bit of context, what is the population of the Netherlands and how many police officers and staff have you got there?

NDL

Right, well this is embarrassing because the population in the Netherlands I would really have think but it's around 17 to 18 million people in the Netherlands. And our police force, I believe this year we have over 60,000 FTE [full-time equivalent] in police, so some of that is part-time of course and some employees are full-time, but over 50,000 employees and over 60,000 FTE in the police organisation. It is the largest employer in the Netherlands, so they have a lot of people, a lot of hours to account for, a lot of budget to distribute But of course, then the contrary part is that even though they have a large budget, that it's just similar to police in other countries, it's actually never enough to investigate every single case, so they have to be very aware of the scarcity of their resources, which is what this team is actually about trying to be preventative in the distribution of resources and trying to look to the future and what's going to change, although they also answer different questions from police management. It's not all they do, the impact analysis, but they also do different types of research on that.

SKN

I've got a figure here from your article, can you maybe talk us through what we're seeing on there?

NDL

Yes, so for this study we collected individual estimates of the time spent on different investigative activities We went around the Netherlands to different teams and we asked them by means of a survey they would fill out, to fill out for each investigative activity on average how much time in minutes they spent on this. So, collecting this from a large sample in the Netherlands allowed us to get an overview of the average time spent on each activity, and this figure is actually a fictional representation of what you could do with that information. So, we asked them for time estimates regardless of the crime type or the case they were working on, but just on average - interviewing a witness - how long does this take, and this allows you, if you have estimates for all these investigative activities, it allows you to create these types of police paths which gives you a rough estimation of how much time actually goes into, in this case a typical high volume crime. So, for example shoplifting, you could check the systems, which type of investigative activities are very frequent in the police path or the criminal investigation of shoplifting for example, and you could lay it out like this and you would have time estimates to go with each investigative activity to give you a more precise idea of how much time and effort goes into working a call for shoplifting from start to finish.

SKN

And what do you think could be changed if police were able to map internal demand in that way and assess how much time is spent on which task, what could they do with that information?

NDL

From my conversations with the people we work with, the Bureau of Impact Analysis, it was very much about getting a better grasp on what is actually happening. Because, as we explained in the article there's been some research on external police demand, so calls for service, because that's very well registered, as I said. And internal demand, so everything that happens after this call for service, is actually not very well registered. And therefore, it has not really been looked into that much because it's really hard to measure this and to get a grasp of this. So, for police it's been very valuable to get a better grasp and the method we used in this article which, I'll explain in a minute, gives a way for replicating this every few years, or every year if the police organisation would want to keep it updated. Because of course police activities also change over the years, new technologies may become available, other activities may cease to exist or become less frequent because of technical advancements. So, this methodology actually allows them to get a more precise measure of the internal police demand and it allows for replication and for keeping that data set pretty much updated. And for the Netherlands Police, it will provide input into their dynamic simulation model that actually tries to predict the capacity needed for the police in upcoming years, and this will also be one of the forms of input for that dynamic simulation model, hopefully making it a little bit more precise in its predictions.

SKN

If some force somewhere in South America, Africa, Asia, anywhere hears this and they think ‘Oh, we should be doing that’ - how would they go about it, can you talk us through that process?

NDL

This is actually the best part about it, because I do believe that in most police forces, they would be very able to also employ this method. So, what um the Bureau of Impact Analysis did a few years before they met with us and before we started this project was they went around visiting police forces in the Netherlands, these were frontline teams but also district teams, different investigative teams in different parts of the organisation and different parts of the hierarchy. They pretty much asked - when you get a call for service or a call for something else, what do you start with? And I asked about talking me through all the most frequent activities that take place whenever they receive a call. So, when we met with them, they had a certain set of um of activities that they still wanted the time estimates for, but doing the work to create this set of - we call them building blocks - for a criminal investigation, so a set of these investigative activities and everything that surrounds it that's not necessarily investigative but maybe also part of the process, such as informing the victim who reported the crime. Collecting this set is pretty much a one-time job that takes probably much of the work, much of the time, but can be easily done by visiting different teams and asking them and collecting all these activities and then seeing what the most frequently named activities are that take place.

Once you have this dataset of activities, you go to the next part, which we did, and that is through a survey collecting individual estimates for each activity. So, we use a software called Mentimeter, which is actually interactive presentation software you can use but you can use any other type of survey as well, where we just ask each person to fill out how much time in minutes they on average spent on each activity. The bigger sample, the more idea you have of on average how long it takes to actually conduct an activity, and collecting the building blocks first, you would only have to update it every few years or every year if you want to see if there's any new activities that you can add to the dataset or if there are any old activities that are really not relevant anymore that you want to lose. And then it's also just a question of sending out that survey, we visited teams to do the survey with them for the first time, because this was part of also a different type of data collection for another study,  but you wouldn't even necessarily have to visit all these police teams, you could send out that survey for example every year and collect all those time estimates, and from year to year you would have an updated dataset and an updated average time spent on the most frequent investigative activities that you could use for very different types of analysis actually.

SKN

So, first of all you guys went out you spoke to the different people who did a specific job and mapped the process - what do you do first, what do you do then, what do you do then, and from that you created a survey, sent it out to those people so they could then input how much time they spent on each of those activities?

NDL

Yes so just to clarify a little bit, the team from the Bureau of Impact Analysis they did visit those investigative teams, those investigative units, they made sure to visit both the more generic investigative departments as well as the more expert departments, who for example only focus on sexual assaults or who are specialised in missing persons, those types of specialised departments. They visited both, and they just asked those people ‘What's a regular call that comes in? When you get this call, what do you do first, okay you go to the crime scene, when you get there, what do you do, okay what do you do next?’ And this is the type of like walk-through they did with the teams and they very quickly realised there were a lot of activities that were repeated for pretty much all the crime types, or at least multiple crime types, and there were also activities that were more specialised. And for our study, we asked for the time estimates of the most frequent activities, so not so much the very specialised ones, but it would be very easy to also send out the survey to the specialised departments to ask about the specialised activities.

SKN

And then how often would people have to fill in that survey, would it be for each case that they're working on for a week, or would it be a one-off or how would that work?

NDL

I think this is a nice part about it, it would be a one-off because this is how the Netherlands Police work, but the police departments we have, I mentioned the frontline teams which is both the patrol officers out on the streets as well as high volume crime investigation and then you have the district teams and then you have the regional teams, so these investigative units, they are in a type of hierarchy that is similar to the complexity, or is parallel to the complexity and the frequency of the crime types they handle. So, for frontline teams when you would ask them to fill out the time estimates for their for the investigative activities, these people would fill out interviewing a witness, not for all crime types, but they have this workload of very high volume and lower intensity crime such as shoplifting, simple assaults, DUIs [driving under influence], so per team filling out those time estimates you would also be able, in the Netherlands at least, to link it to a certain set of crime types that is very frequent in their workload. So, we would not ask them to think of one case, we would ask them ‘On average in your workload, how much time would it take you to interview a witness or to collect camera footage or to inform the victim of the progress of the case?’ Things like that.

SKN

And how have the police used that data, what have they done with your findings?

NDL

The police is now still in process of trying to integrate it with the dynamic simulation model, so that's one of the things they had in mind that they wanted to use it for as input data. And it also now has become a dataset that will be updated probably every few years, and it will be input for their specific impact analysis as well, and it will hopefully at least speed up the process, because beforehand, for each impact analysis they would have to visit the relevant teams to ask ‘What do you do, how do you think this will change, how much time would that take you extra or less if this change would be would be made?’ And now, they actually have numbers to already calculate with instead of just interviews and using expert opinions for those types of analysis. And of course, they did use numbers before, because they do have registrations in the systems, but hopefully this dataset will create a more easily accessible dataset on investigative activities and how much time they cost.

SKN

Have they made any efficiency changes, that for example they realised people were spending too much on this task, why is it taking so long, and maybe improving something, have they streamlined something or what was the ultimate goal, perhaps was it to free up some time, was it to identify where wasted time was or where activities were maybe unnecessary?

NDL

I think an important part of the research was the fact that we were not evaluating how well the police was doing in terms of time spent on things. I think this is also very important if you want to get honest answers from this type of survey, that that was not the aim of the research. But I think it was very informative, for example the team leads of the teams that participated, because they would get feedback on what are the activities that actually take the most time on average in your team and what are the least time-consuming activities? Which is important because they also have to answer questions on a yearly basis about capacity, about budgeting, about their vision for the next year, how they want to use their resources, so this was also I think a very valuable input for them. In the end of course these types of results can be used to see what is the most time spent on and is there anything that we might be more efficient in, but this study and the paper that we wrote was to present this methodology and present to the Dutch Police and to police forces outside of the Netherlands how you could use this method to measure internal police demand and keep it updated and replicated over the years.

SKN

Okay, and obviously you've explained your methodology here and also it is in this paper here again which is um fully available, Open Access, so anyone can just maybe Google your name or find out find the actual title here so thank you very much for explaining that. As my last question to you today Natascha, if you had a magic wand and you could wave it and make one change in policing, what would that change be?

NDL

I think I really loved my work with police so far because they're a very passionate group of people which is something I would absolutely not want to change about policing and how they do their work. I think it would be nice if I'm speaking on criminal investigations and internal police demand, I think it would be nice if this method would maybe not be relevant at all because the registration of internal police demand would become more automated and better registered. I think that would be that would be a very nice asset for both practitioners and researchers

SKN

Natascha de Leeuw, thank you very much for your time today.