 All right. Thank you everyone. So last session of the day we're just going to start with a paper by Polina Orblinskaya at the rise and fall of violence in England and Wales. Polina is a research fellow at the Violence and Society Centre at City University working with the crime survey for England and Wales on a project violence, health and society that's a vision. Her research focuses on assessing the implications of methodological variations for accuracy and bias of estimating violence, exploring the differential risks of violence among groups in the population and investigating the differential impacts of violence on health. Thank you. Hello everyone. Thank you for the introduction. So, as Cathrin said, I'm working on the consortium called Violence, Health and Society of Vision. And I'm here with two of my colleagues who are also presenting in this session. So I just thought I'll give you a little bit of an overview, not an overview, just a snapshot of the project just so you see where our work comes into. So this project is funded by the UK Prevention Research Partnership and it's a five-year project. We are around halfway through it and we have multiple objectives in this consortium. Firstly, it's called Violence, Health and Society and our aims are to improve theories, so coordination of theories, theories of change to identify, classify, profile and compare different measurements and risks of violence to integrate. So we look to try to link data from multiple data sources. That's part of the consortium's aim as well. And then lastly to do some cost effectiveness and what works work. And sort of activities throughout the consortium are aligned with these objectives and with the aim to better measure violence, better understand violence to reduce it and to reduce the impact of violence on health and mental health. And the work consists of multiple strands. So we are here from the crime strand. So our work is predominantly on the crime survey for England and Wales and we come under the Crime Injustice Services group of colleagues working on it. And then we have health and health services strand. We have specialist services using data from services that provide support for, for example, victims of domestic abuse. We have a strand on inequalities and intersectionality that looks at socioeconomic, for example, impacts of the relationship between violence and health and mental health. And then the last one is the integration and that looks at integrating data from different sources. So yes, so we are the crime group and all three presentations will be using the crime survey for England and Wales. So this is a paper that we are currently working with, I'm currently working with Niels Blom who is in the audience and will be presenting shortly Vanessa Gash. And it's work in progress, so apologies for it not being complete, but I thought it would be a good chance to present to you so you're aware it's happening and also to collect any comments and useful questions. So just to acknowledge our funder. So there's been a considerable attention that's been paid to the crime drop, including violent crime from about mid 90s and it was particularly shocking because of the increase in crime preceding that period. The desegregation of trends by various individual characteristics has been emphasised as important in order to adequately describe this trend, but also it has been identified as a gap in the literature. More recently there has been a flattening of the trend in violent crime that has been reported in national statistics and most studies today explore the overall crime trends or trends for particular type of form of crime, for example violent crime or violence by domestic perpetrator. Either for the entire population or for a limited number of characteristics, population characteristics and mainly concentrating on sort of one of the trends, so either the rise or the drop in violent crime. But yeah, mostly the fall in crime. So from this we were thinking that the description of the trend in violence has been overlooking the differential experiences of different groups in the population over different periods of time. And also just to come back again to the violent crime, so the emphasis has been on violent crime predominantly and that is understandable because this is what crime survey is particularly good for it to look at a consistent definition of violence over time that will be more or less comparable with the police recorded data so that we can see who for example the police data is missing. But this means sort of focusing on the violent crime definition is underestimating the amount, sort of the number of people that are actually impacted by violence and I will talk about it in a minute. So some of the particulars of the violent crime measure before I sort of move on to the measure of violence that we are using in this work. So first of all the onus violent crime measure is based on the violence against the person definition and includes wounding, common assault and wounding with social motive including serious and so on. And so it excludes sexual offences which are sort of reported separately and they include such as rape and attempted rape for example and it has been argued that sexual offences should be included in the violence against the person measure and sort of the reporting on violence. As it is particularly impacting women so it's sort of biasing the violent crime measure. Another particular sort of about the violent crime measure is that it is based on offence codes that exist in priority order. So if a crime takes place and it has multiple offences in it, the principal offence will be reported and recorded and that is based on a specific order of seriousness or a specific reporting order. So sometimes and work of Polaris and Phoenix showed that the way the priority coding offences works is that it could be putting certain violent crimes and non-violent offences and that particularly impacts the estimates for women again. So another thing about violent crime measure is it excludes robberies and excludes threats which could be also argued that they have sort of a violence element in them so robberies were included in the national figures up until a certain point I think 2013. And robberies are characterized by either having a violent force element in them or threat and it's sort of more force than needed to just simply snatch something. So we'd argue it to sort of capture a broad experience of violence that it could also be included in the violence measure and threats of violence equally so. So additionally just to add that violent crime also excludes some non-physical forms of violence and it only includes physical forms of violence which is fine and you know it's for the needed reason but such violence and abuse of psychological economic violence because of control obviously not in the measure and we can't really do much about it because of how the data is collected in the face-to-face questionnaire. And it also does not include a more accurate measure of domestic violence and abuse including sexual violence that is in the self-completion part of the questionnaire. So this is all the details of what violent crime doesn't include and what we would like is to use a wide definition of violence as we can to just show what the prevalence of violence is in the population, what it looks like and how it changes for different groups of people. And so we are applying a wider definition of violence in this paper that goes beyond violent crime and this is to fill the gap in the evidence of what we know about, experience of different groups and to basically look at some of the protected characteristics of people so for example gender, age, disability status, marriage status but also some of the other characteristics that I'm not going to talk about here but we'll look at so for example migration status and employment status and so on. And here I'm going to present really broad findings and concentrating on just the trend mainly on how it just looks different for men and women. Everything is desegurated by men and women because they just have quite different results. So the data I'm using is obviously the crime survey, I'm not going to introduce you, we've been talking about it today. The measure of violence, I'll just go over it again, that we're using is based on the ONS, violent crime definition so offences including serious wounding and other wounding assaults and a wounding with sexual motive are included in our definition of violence and measure of violence as well as sexual offences, sexual violence such as rape, attempted rape and indecent assault, we're also including experience of robberies and threats of violence. So the analysis is descriptive mainly and we're looking at prevalence, no incidents or concentration and then we look at different groups over time. All the data is available from the start of the crime survey so we just use what we can and we also run analysis controlling for certain social demographic characteristics of the population. But they look quite similar I have to say the trends even after controls but I will note where they might not look similar. And there is another bit of analysis that's not going to be in the paper we are working on and that we'll look at. I did composition analysis so we'll see so other changes in trends that we see actually because there's been an increase in the risk for particular group or because perhaps it's the composition of the group has changed over time and that made an impact on their trend but I won't be talking about it today. I hope I can explain this well and I don't know if you can see the different colours but basically the blue colour here, the blue line for men and women is showing the trend for violent crime as per ONS definition. So we can see that both women and men experience an increase in violence up to the 90s and then a decline in violence a bit steeper for men and I mentioned that there's been a bit of a stall in the decline of violence in the last five to six years of data. We are using data to 2019-20 and it is particularly noticeable for women there is no change since 2014 in the prevalence of violence for women but there is a small one for men. But what is interesting is that during this time and it's particularly what you can see for women is that the prevalence of threats has increased significantly and it's actually been increasing since the 2000s. There's also increased in experience of sexual violence among women. We are not presenting it for men because the figures are quite low but we are including experience of sexual violence in the overall measure. So if we add all these experiences together we find that the prevalence has obviously increased because we are including experiences of threats, robberies and sexual violence and it's about more than doubled for both men and women. But what's interesting is that the trend, the latest sort of period, the trends in the latest period have changed quite, the picture has changed quite a bit for men and women and that is that there's been an upturn of prevalence in violence for women and for men it remained pretty flat since about 2014 and this is mainly because of the inclusion of threats that affect women more than men and we saw that the threats have gone up quite significantly. So I'm really going to just focus in the last minute on the sort of the last period's trends and what we see is that actually so just bearing in mind that the overall trend has gone down since the 90s we can see that for the youngest group and for I'm using quite a large group of those under the age of 50 just to really highlight the older population a little bit more, we see that there's been a decline in the prevalence of violence using our broader measure for the younger group, slight upturn here for women but there is quite, you can see there is an increase especially for women in prevalence of violence for the older age groups and actually for the oldest of the old there's been no change for quite some time and an increase right at the end. So no, obviously all age groups are following the same downward or flat trend as the overall. Let me just keep some. So there is an increase in violence again among men but especially women with disabilities in the sort of last period of relative stability in prevalence of violence and for women it's sort of started going up a little bit earlier and if we sort of control for age and other socio demographic characteristics we still see the same trend women's risks of violence increase significantly if they have disability over the past five, six years and then so the last bit of results is looking at risk of violence by marital status and I'm just going to talk about women because it has more interesting pattern for women than for men, for men it's pretty flat for all marital status groups but for women here we have just proportions so unadjusted trends and we can see that there's been an increase in prevalence of violence for married and coupled women as well as for separated and divorced women so it's those in partnerships or ex partnerships that the risk of violence increased for them and it remained broadly stable for the single never married women and it's the same sort of pattern of trends if you adjust for other socio demographic characteristics it's just that we see a much higher risk for the separated and divorced women as simply going up and again it's sort of the time period is from around 2014, 2015 so in summary we find that a greater proportion of people experience violence using our wider definition obviously compared to the honest violent crime definition and while the overall trend sort of of the increase and decline in violence in the previous years they are sort of similar using both measures they just might have a little different timing but it is particularly the last five, six years of data that have quite a different pattern of trends using the two different measures of violence and it is mainly because of the increase in threats that actually impacted all the groups such as women, those with disabilities, ethnic minorities which I didn't show and those in partnerships or ex partnerships as opposed to single never married individuals so the inclusion of different offences does make a difference to the trends and also to the risks experienced by different groups. Sorry I had to rush through it now quickly. I thought I didn't have enough material for 20 minutes. Thank you.