U.S. flag

An official website of the United States government, Department of Justice.

Dot gov

Official websites use .gov
A .gov website belongs to an official government organization in the United States.


Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Domestic Violence Shelters: The Experience of the Survivor

Angela Moore, Acting Deputy Director for Research and Evaluation, National Institute of Justice; Marylouise Kelley, Director, Family Violence Prevention and Services Program, U.S. Department of Health and Human Services; Eleanor Lyon, P.h.D., Director of Institute for Violence Prevention and Reduction and Associate Professor in Residence, School of Social Work, University of Connecticut; Anne Menard, Director, National Resource Center on Domestic Violence

Panelists will present findings from a comprehensive study of domestic violence shelters in eight states. Data were collected from 3,410 residents in 215 domestic violence shelters — 81 percent of the shelters. The first of its kind, this descriptive study seeks to fill a gap in current knowledge about the needs and experiences of domestic violence survivors who turn to shelters for help and the type of help they receive. Implications for policy and programming will also be addressed.

Angela Moore: If you haven’t done so already, please grab handouts on the back table. Thank you.


Good afternoon everyone — thank you, I like people that talk back, except when they don’t say what I want them to say, but, on behalf of the National Institute of Justice, I would again like to welcome you to our annual conference, and I hope you all are having an enjoyable time and you’re getting a lot of useful information out of the conference. And I also hope that you found a way to get some caffeine with your lunch so that you can stay awake during our session. I know it’s gonna be very exciting and we’re very interested in it, but I do understand what happens right after lunch, so if you need to stand up, that’s fine; you need to get some water, please do so.

My name is Angela Moore, and I serve as the Acting Deputy Director for Research and Evaluation at NIJ, and I’m delighted to welcome you to this session. We are gonna be talking about domestic violence shelters, the experience of the survivor. So, I hope you all are here for that session, and even if you’re not, this is where you need to be. I am going to introduce our three panelists in the order that they going to be presenting, which is not necessarily the order in your book; we like to break all the rules.

So, but before I introduce our panelists, I just wanna say, this study that you’re gonna hear about today grew out of a long-standing partnership that NIJ has had with the Department of Health and Human Services, in particular, the Administration of Children and Families. It’s been a pleasure working, in particular, with Marylouise Kelley on this partnership, and I’m looking forward to the great things that we’re gonna be doing in the future, and perhaps she’ll talk about some of those things when she presents. So, first I’m gonna start with Marylouise Kelley, who is the director of the Family Violence Prevention and Services program within the Department of Health and Human Services. FVPSA — that’s the acronym, and every program has to have an acronym, and even if it doesn’t have one, we’ll make one up for it — supports a host of programs dealing with domestic violence, including the National Domestic Violence Hotline. In addition, they support a variety of shelter programs and services for women and children and men who are victimized.

I came to know Marylouise actually from the work that she used to do when she was at the Department of Justice Office on Violence Against Women, so, it’s, it’s been a great working relationship that we’ve shared. After Marylouise, you’ll hear from Dr. Eleanor Lyon. Dr. Lyon is director of the Institute for Violence Prevention and Reduction, and she’s also associate professor in residence at the School of Social Work at the University of Connecticut. Dr. Lyon specializes in evaluating programs for battered women and their children, children who have been abused sexually, as well as evaluating interventions in public schools. And actually one of my earlier experiences at NIJ was serving as a grant manager for one of the projects that Eleanor conducted.

Rounding out our panel will be Anne Menard, and Anne is the director of the National Resource Center on Domestic Violence, where she directs training and technical assistance and resource development to support domestic violence prevention and intervention efforts. A pioneer in the anti-violence against women field, she has worked on policy practice and research issues affecting domestic violence survivors since 1976. And on a personal level, I’ve had an opportunity to work with all three of these women throughout my career at the Department of Justice National Institute of Justice, and I can honestly say that I’m all the better for it. So with that, I’ll turn it over to Marylouise.

Marylouise Kelley: Thank you. If you’re able to hear me here, I think I’ll stay here at the table. Does this work? All right. Also, for those who just came in the room last few minutes, there are handouts in the back that you should pick up; there are also a couple of seats up front if you want to sit down with a table in front of you, so feel free. And Angela, thank you very much, that was very nice introduction, and it really has been a pleasure working with you for lots of years now, and I look forward to continuing to do that.

Well, in my role at the Family Violence Prevention Services Program at Health and Human Services, I was lucky enough to commit to this project after Bill Riley, the former director, decided to fund this project jointly with National Institute of Justice. And, we felt it was important because since the passage of the FVPSA Act in 1984, our program has been the primary federal funding source that’s dedicated to domestic violence shelter and supportive services throughout the country.

So we fund nearly all of the 1,200 shelter programs and about 300 nonresidential programs around the country. And I think we have gone — we’ve seen the work that is done there; we hear the stories of people’s lives that are changed in shelter programs, and like you, we are very convinced of the importance of this work. We were lucky enough to know from the one-day census of domestic violence services that’s conducted by the National Network to End Domestic Violence that there certainly is a need for these services. When they did a one-day count of services offered, they found that over 60,000 victims were served in a single day by domestic violence programs and that there were nearly 9,000 unmet requests for services — 21,000 hotline calls, 30,000 people trained. This is in a single day. So, clearly there’s a demand out here. This isn’t a service where people are looking for program participants at all.

But for our office and trying to be good stewards of federal funds, funding these programs for all these 20-some odd years, we also had to be able to answer the question, “Are these programs effective? Are these programs meeting the needs of victims of domestic violence?” We thought so. We told stories of the numbers served, but we really had a little difficulty answering those questions I feel until, until this study was conducted, and it’s been so helpful in that regard. But just as important as saying, “We have to ask the question, ‘Are these effective? Is this what survivors need?’” — it’s important to ask the question in the right way. If this were easy, it would have been done a long time ago. We would have had all kinds of effectiveness evaluations on domestic violence services. And when you’re in federal government, sometimes you focus on those widgets so that you get, you know, these program accountability measures that will, will tell us if we’re doing enough of this and is it cost effective?

You know, it was suggested that we should report on the cost per bed night, but if we reported on the cost per bed night in these underfunded programs, then you have to ask the question, “To improve on that measure, what should we do? Provide fewer services? Make shelters more sparse? Not serve those victims and children quite as well?” So we didn’t think those were the questions that needed to be asked.

Sometimes, people wanted to evaluate shelter programs by asking, “Did the, did the victims of domestic violence leave their abuser?” Well, without knowing, we all know that separation violence is a time of more violence for victims of abuse. Leaving isn’t always the answer. Maybe leaving at that moment isn’t the right answer, and maybe it’s not what she wants to do. So, it was very important to us that if we were going to look at shelter services, it was important to ask the right questions, and what I’m so pleased about with this study, is that it really is done from a survivor-driven services approach.

It asks domestic violence survivors for their assessment of whether they received the services they need. What was the change in their lives because of that? And it tried to get at the issue of, “Is this going to improve their safety and their well being?” The real important measures of whether or not these services are effective.

So we are very pleased because this approach was built on questions and instruments that were developed in consultation with the domestic violence field over a number of years, and Eleanor will, will explain some of that to you. It got at survivors’ assessment of the services, and it also gave us information on whether shelters were accessible, welcoming and effective for survivors from different backgrounds, different abilities, different sexual orientations, all important information for us to have.

So this shelter study answered many of these questions.

It gave us some reassurances, many reassurances, but as any good study does, it also pointed to the areas in which we could seek to improve our services and it raised questions for future research because we know that half of the domestic violence services that are offered in a given day are provided in shelter settings and another half are provided, are nonresidential services. Some of our next questions that we’re queuing up are, “What do we know about the effectiveness of nonresidential domestic violence services?” So, that is our, those were our goals and what we were looking for in this study, and I’m pleased to introduce Eleanor Lyon, who’ll share more with us about this study and what, what she found.

Eleanor Lyon: And since I am doing the, the heavy lifting with the PowerPoint, I’m gonna stand up here. Can you all hear me well? And everybody, those of you who are standing at the back should know that there are all these lovely empty seats toward the front. OK, good. You always feel a little strange when people just sorta float toward the back and don’t want to come near.

I wanted to start with this, even though that’s, this is not the title of today’s presentation, this is the title of the overall study and the final report, and so if you are looking for it, we’ll give you at the end the, a way to access it and a variety of other information online. But, I wanted to show you both so that you would see that title, but also to see symbolically the extent of collaboration that was involved in this study because not only were all these organizations and agencies involved, but in addition, clearly, survivors and advocates who are working in the field. So, this is the official title slide of today’s presentation.

So, wanted to give you a quick overview, even though Marylouise has just done a little bit of the overview, because clearly it was a highly collaborative effort; it did involve both the National Resource Center and the University of Connecticut working directly on this study, but also collaboration with domestic violence coalitions and advocates across the county. You’ve already heard about the funding, but we worked with coalitions and programs in eight states, which I’ll detail in a moment.

Just a quick little bit of background — we did provide training to the coalitions at a meeting in Atlanta in July; they were, they were all present. They, that was one of their opportunities to provide feedback, and we did make some modifications as a result of their feedback, but then that training presentation in PowerPoint was then given to each of the eight state coalitions so that they, in turn, could use it to provide training to the participating programs in their state. We did maintain ongoing connection through a variety of mechanisms. We provided biweekly reports back to each of the states so that they knew how many results were coming in and from which of their programs, so all of that coordination. We were also providing technical assistance throughout the study. There were two forms that were involved, one that was completed at around entry by survivors and another at exit, and I’ll explain that in a bit, more in a moment.

One of the things that we’re particularly proud of is that all materials that survivors saw were translated into 11 languages. Those were the languages that were identified by the coalitions as those non-English languages that were most commonly spoken by the survivors who came to their programs. We also had a contract with a national hotline so that people who were unable to complete a written form in their language other than English, we could involve their translation, their access to translation services in, I don’t know how many languages they have accessible to them, but it’s a massive, a massive number. So we did a great deal to try to enhance participation and access to this study.

Marylouise described some of the goals, but these, more precisely, were the goals for this study. It’s basically a descriptive study; it’s not controlled experimental design. And what we wanted to know was about the experiences of a diverse sample of domestic violence survivors in residential shelter programs. I need to underline “in residential shelter programs right now.” This is not a study that included people who were staying in safe homes or who were provided safety through motels or through other means in their communities, but these were the regular, residential shelter programs. We also wanted to learn more about what survivors want when they come to programs and the extent to which they get what they want. We wanted to learn about how their experiences and how their needs might vary across demographic and program characteristics, and we also hoped to be able to develop recommendations for programs about improving services.

This shows you something about the participation that we obtained. These were the eight states; you’ll see that we have two states from each of the major geographic regions of the country. We have two from the West, two from the North Central, two from the South Central, and two from the East. In all, there were 215 programs that participated in this study, which was a response rate of available programs in those eight states of 81 percent. There were some individual programs that were participating in other studies or had other pressures upon them at that time, and so they declined to participate. States were clearly selected to maximize the diversity of programs and survivors so that it would be as representative as possible. Clearly something that involves a pure random design under these circumstances would be incredibly difficult to accomplish, particularly if you have voluntary participation of survivors in the, in the a, in the shelter programs. So in this way we were able to vary by geographic region, by the type of economy. We had communities with very much concentrations on high-tech kinds of industries, as well as agricultural, as well as industrial. The rural urban combination I think was well represented and the, and we were able to obtain good representation of the racial, ethnic diversity in the programs in the states involved.

The coalition contacted the staff members in each program and requested their participation and then as I said, provided the monitoring. A description of our instruments, and I should say that as Marylouise indicated, these instruments were revised over time beginning in 1998 — working with researchers, state administrators, survivors and advocates, both local and state level over time — and were pilot tested in four states before this study was conducted. So we felt pretty confident in the utility and the comprehensiveness, at least for these questions of these particular instruments. 

Shelter one, as we refer to it, was offered within 48 hours of entry. It was clearly not administered while a new resident was in crisis, and it includes a wide variety of kinds of information in two short pages, because we were also not wanting to burden survivors with excessive involvement in answering questions. The staff had guidelines for how to talk to each of the survivors about their participation in the study. They were given an information sheet, again, translated into all those languages that explained that this was voluntary participation, and so all of that was included for both shelter one and shelter two. It also covers the needs, survivors’ needs, and the help that they wanted in addition to these other issues.

Shelter two was given when a resident showed signs of leaving or at the time when most residents usually left. If any of you — and I should ask that. How many of you are advocates or work in shelters? Any of you? OK, well those of you in particular will know that there is often not a planned exit moment, that sometimes survivors in shelters simply leave in the middle of the night, and that’s why we had this kind of guideline. If there was this sort of, if most people in their shelters stayed for a month, they would be asked to complete it again at about the one month point, and then if they knew when they were about to leave, they were asked to complete it again, and then the interim form was eliminated from the study. But, sometimes survivors give signs like they’re making arrangements for housing that they’re going to return to, or they’re organizing transportation or they’re organizing other kinds of things that are consistent with leaving, and so that would be an indication as well.

Shelter two includes the number of days they’d spent, same demographics, their experiences and their outcomes — very short-term, immediate, self-reported outcomes. And then the extent to which they had obtained the services that they had wanted. Those are the — that’s the information we obtained from survivors. We also obtained information from programs, each one, each program completed an information sheet that asks about their, the services that they provided, the kinds of advocacy they provided, something about their staff background, the accessibility to the shelter, and a variety of other questions. And then we also obtained census information about the county in which the shelter was located ’cause very often people are drawn to a shelter not necessarily from a specific town but from a larger area.

Want to emphasize some of the safety and confidentiality precautions that we took. Each program was assigned a code number by the coalition. The coalition maintained the record of the link between the program number and the program location. When we received the completed forms it had the, the number on it of the program, but we had no idea which program the survivor was coming from. We did know the state, but we did not know which program. Each, each resident created their own personal ID. We provided a suggestion as a way to do that but also allowed them to vary it if they wanted to use some other mechanism, if that made them feel like they were more easily identifiable, and that enabled us to link the shelter one with the shelter two, and without compromising the identity of — we didn’t even know which program they came from.

We also — the residents were provided with preaddressed and stamped envelopes, so they could send surveys to the research staff, and again, the staff received written guidelines for implementing this entire process, so it was very clear that they did not complete the form in the presence of a staff member. They did not hand their survey back to a staff member. They had another repository location. We’re getting to the results in a second, but this is all important stuff to understand as a background.

So, we, as I  said, we had 215 programs. The average number of staff was 16 and a half, but the range was dramatic, so we could have reported a median, but we reported the average here, but it ranged from 1.25 staff to 99, so we had tiny rural programs participating and very large urban programs as well. You can see that the average number of volunteers was 15. Average capacity as far as beds, you can see again, a dramatic range between four and 102. The average number of adults ranged from two to 2,300; that’s clearly one of those larger programs. And an average number of children, the range was one to 1,242. So we had some very, very busy programs as well as some much smaller programs.

We asked about the language capacity of the staff in the programs, and this is the list. Thirty-six different languages were spoken by staff in these 215 programs. Clearly the larger the program, the more likely the staff were to speak a language other than English. I can say 82 percent of those programs had at least one person who spoke a language other than English, and 72 percent spoke Spanish — of the total, had someone who spoke Spanish.

Looking at the program services offered by the, by the, the shelters, 98 percent reported that they could accommodate at least one kind of disability. And the, in the, the most common, clearly were physical disabilities, but over 80 percent could accommodate hearing disabilities and then slightly less than that, visual disabilities and had other mechanisms as well. Of nine different types of advocacy that we asked about, the percentage of the programs that offered those ranged from 95 percent to 73 percent. Housing was the most common kind advocacy followed by — and I don’t highlight that here but for this audience I should — 82 percent had civil court advocates; 81 percent had criminal court advocates as part of their program.

Housing — divorce issues were least common with 73 percent. So, you can see that a wide variety of advocacy was offered, and of the six different types of services that we asked about, the range was 97 percent who offered support groups to 50 percent who offered child care, before in between our crisis counseling, individual counseling, parenting classes and child counseling. So, those are the ones that fall in between the support groups and child care.

OK. We get to now to the survivors themselves. We did not have all of the survivors complete both forms as you can see, and you see that break down — 565 completed both shelter one and shelter two, and when we did a comparison of that group with the others, they were remarkably similar. The, there were no statistically significant differences between those two groups. There was a slight tendency, interestingly, for those with less than an eighth grade education to have completed both forms, so it was quite contrary to some of our notions that, “Well it’s only the college-educated folks that are gonna participate in the study” kind of thing, so we felt pleased about that. But if you put those numbers together, you end up with 2,446 who completed shelter one, and 1,429 who completed shelter two.

Gonna give you a quick overview of some of the, the background demographic characteristics. You see just slightly over half are white or are Caucasian and the remainder are people of color, 22 percent African-American and you can read those. Although you don’t have a handout that is of these slides in particular, much of this information is provided in the handout that you, that you received. There’s some other information that is not in there, however, which I’d be happy to provide to anyone.

This gives you a breakdown of their ages; they are fairly young, but not a really young group. Ten percent are over 50 as you can see, 70 percent are in that sort of middle range of 25 to 49. This was a piece of information that we found particularly interesting and somewhat poignant because you can see that there’s a difference between the number who report having children and the number who report having children with them. They are — 78 percent reported that they had children under the age of 18, but just 68 percent had children with them in the shelter. Spent some time talking about this and would love to hear some of your reactions during the comments, but it seemed to us that probably this is a reflection of some of them having older children who might be able to stay with friends or other family members and stay in the same school. In some cases, the shelter resident may not have had custody of their children, and so while they have children, they do not have custody of those children, and so they are not with them. And there may be a host of other reasons as well.

This provides information about the sexual orientation of the respondents in this study: 93 percent heterosexual or straight as you can see. The others were often sort of culturally specific designations that described a different kind of sexual orientation or in some cases they said that because of the abuse, they no longer felt like they were sexual at all and that they had lost the sexual parts of themselves, and so they didn’t claim an orientation in response to this question. So, we do have qualitative information that does explain some of these findings.

This gives you a sense of the education level of the survivors: 27 percent have not yet graduated from high school, you can say it that way. You can say, well, 56 percent have a high school education or less. Or you can say 44 percent have some school or training beyond high school. So, all different ways to look at that particular piece of information.

OK, now we’re getting to the responses from the survivors themselves. And this tells how they heard about the shelter, number one being a domestic violence advocate, but second being police. There were, these were the most common sources of information; there are many others, which I can provide for you. Thirty-four percent provided two or more responses to this question; they may have heard from multiple sources, 26 percent had heard from police and, or court put that together, so that raises it a bit more. But, mental health, health care flyers, church, child protective services and welfare services were other primary sources of information about the shelter.

This describes when the survivor had heard about the shelter. And there’s sort of an interesting potential set of stories just to, for this particular piece of information. You can see that 27 percent had first heard of the shelter more than a year previously. Well, it’s not just because they had stayed in the shelter before, although that was true for 54 percent of that particular group, so a little over half stayed in the shelter before. But for others it may have been a reflection of a fair amount of safety planning, looking at resources, sort of planning when will be a good or a safer time to make the move into a shelter. Clearly 25 percent who had first heard of the shelter within a day or two of getting there, that’s a reflection of a very immediate crisis and not a lot of advance planning or investigation of resources.

So, of the total that we talked about, nearly a quarter had stayed at the shelter before, and the majority of that group, it had been at least a year since they had stayed in the shelter.

This looks a little bit more specifically at people who heard about the shelter from either the police or from court staff. They were more likely to be Latina or Caucasian, the other women of color, the African-Americans, Asian and Pacific Islanders, Native Americans, and multiracial folks were much less likely to have heard about the shelter from people in police — from either the police or people in court. They were somewhat more likely to have heard a long time ago, and they were more likely to have stayed in a shelter in the past. Interestingly enough, just one of the men in our study had heard about the shelter from someone, from the police or from someone in court. It also suggests that perhaps the people that have known about the shelter for a longer time and who have been engaged in help-seeking for a longer period of time, have been more likely than the others to have also included criminal justice system contact.

We also learn that nine percent of the survivors had tried to stay in the shelter before, but they hadn’t been able to do so. Overwhelmingly that was because there was no room at the time that they tried, but there were also other reasons, and most of that had to do with information or, or eligibility issues. As many of you know there are eligibility criteria that some shelters have that include not admitting teenage sons or concerns about medication, and so some of those come from liability concerns, and there may be a host of other issues that are involved with reasons that, in the past, someone may have not been able to come to shelter.

A full quarter of the people reported that they had concerns about coming to the shelter. These are the types of concerns that they reported, this is stuff that we, that we coded from an open-ended question, and so most of it having to do with not being familiar with the shelter, but the next one I’m going to show you are some examples that illustrate those types of the people who say they have concerns; 95 percent provided some elaboration about their concerns, and these are some examples from what they said. You can see that they didn’t know what to expect, concern about feeling judged, fear of their own, about the potential for their own safety while they are there, concern about deportation or child custody and then rural issues — small town not being private. So, that really is a good illustration of the range of reasons that they had concerns about coming.

We also asked them what they would have done if the shelter didn’t exist. And again, these are our categories. You can see that we put them into basically five different categories, but again, those are the researchers collapsing the categories, and I think that one of the strengths of this study is there were a lot of qualitative opportunities for people to describe. And so I’m going to ask my two colleagues to provide some voices that provide illustration of quotes from the survivors that responded.

What would you do if this shelter didn’t exist?

Female Speaker 1: “Being at homeless shelter or begging friends to stay with them and putting them at risk of abuse by my ex-boyfriend.”

Female Speaker 2: “Continue to run and move from place to place and constantly watch my every move, wondering if he’s around.”

Female Speaker 1: “Perhaps I would continue putting up with the bad life I had with my abuser, in a few words, resign to live like this.”

Female Speaker 2: “I’d be sleeping in my car, lose my children and my job.”

Female Speaker 1: “Probably, I would have been killed ’cause I had no place else to go. If I would have stayed with my other half, I would have been killed most likely.”

Lyon: We asked them about their entry experience, and you can see that by and large, their entry experiences were really quite positive, that especially the responses of staff made them feel welcome, treated them with respect. A little bit less high marks for the “space felt comfortable,” a little bit less than that “it seemed like a place for people like me,” and least likely was an indication that the other residents made them feel welcome. Now of course it is not the job of shelter residents to welcome newcomers, and we know that, but we want, ; did want to have a sense of what that initial experience was like because it can provide an influence over subsequent experiences.

For the researchers in the room, these categories are a result of factor analysis, and so they are the categories of what the survivors wanted in entry. We had 38 different items that we asked them to respond to individually, and 97 percent said they needed some kind of support. You might think it would be 100, but it was 97 percent. These included such items as emotional support, learning about their options, making connections to people who could help them, help understanding domestic violence, reconnecting to their community, help leaving their relationship and obtaining counseling. You can see that those are the most prominent and the most frequently indicated.

Economic help was way up there and that was — this was done actually before the major economic downturn that this country’s been experiencing. We finished collecting data in March of 2008, so it was just before. Things were beginning to get a little nudgier, but it wasn’t at the current level of, of crisis. And you can still see that it was very prominent; 61 percent needed at least three of the five types. This included a job or job training, safe and affordable housing, opportunities for school for myself, help with transportation, help with budgeting. 
Help with regard to their children involved — was indicated by 87 percent of the mothers. That involved safety for my kids, school for my children, counseling for my children, help learning to respond to upset children and help with children’s health issues.

And, criminal justice system needs, we just had three that were specifically related to the criminal justice system. Just over a third needed one of those kinds of help. And that included help with a restraining or protective order, help dealing with my abusers arrest and help dealing with my own arrest.

There were also other clusters, and one that could be noted was help with health issues and benefits, and that had to do with their own health issues, their disabilities, help with TANIF and help with other government benefits and help with their own physical injuries.

This is the top 10, looking individually. So you can see the highest is 85 percent, so the top 10 goes down to 66 percent for safety planning, the average number that was selected out of the 38 was 14, so people came in with multiple and complex needs. A few others that would be worth noting: a job or job training, 57 percent; help with transportation, 52 percent; help with leaving my relationship, 49 percent — so just under half — help with my abuse-related injuries, 27 percent; help with my disability, 19 percent; and help with immigration-related issues, 7 percent.

I did pull out a couple that related in particular to the legal system, not just criminal justice, but the legal system, and those were the indications of those needs. Note that all, except for “help with my own arrest” were significantly higher for people who had heard about the shelter from the police or from court staff. We have some information particularly about what the mothers wanted; these are the needs related to their kids, and all of these items — this is all of the items for mothers of children. None of the men who responded to the survey reported needs related to their children.

We’re getting closer to, the sort of the shelter two now, this is the, sort of the second, second form. The average length of stay was 33 days; range was one to 624, and yes, that really was 624 — we did say, “OK, that must be a typo.” We called the coalition, the coalition knew, was able to, by code, identify the program, called the program, “yes we did have someone who stayed for 624 days.” So that’s worth noting. The median length of stay was 22 days, for mothers was 27. If you look at that in a different way, 63 percent stayed up to a month, and 14 percent stayed over two months. One of the things I did not report, which I should have, is that when we asked the programs about their policies, 18 percent had, by policy, that they could not stay — that the maximum length of stay was 30 days. And 21 percent of the programs reported that the length of stay — the maximum length of stay was established by their funders. So just as a bit of context there.

This provides you with some information about who wanted a particular kind of help and who got it. The way to read this is, “Safety for myself.” And this is at shelter two: 98 percent said they wanted help with safety, 91 percent got all of the help they wanted, 8 percent got some of the help they wanted so that means that there is 1 percent who said, “I didn’t get any help with this, and I wanted it.” So, that’s the way to — that’s the way to read it.

One of those things for those of you who are particularly attentive to all this litany of numbers here is that gee, these numbers seem to be substantially higher than you reported before. And that was a very interesting and consistent finding that we found that at point two, a substantially higher percentage of people wanted each of these items than they did at the point of entry, when they were closer to crisis, when they were less likely to be aware of what available services there were. When they —and, at this point, they’ve had more time to reflect on what they need in order to be able to create a safer life for themselves. It’s consistent with the one other study that looked at this particular issue in Canada, where the needs at the, at the end point, at the exit point were dramatically higher than they were at the entry point. In general, the longer someone stayed, the more likely they were to get the help that they wanted.

This looks at the same sort of information, particularly for legal system issues, and you can see that they are not quite as high as the others, but they are still quite high. If you add them up across, 91 percent got at least some help with a protective or restraining order, 82 percent with their abuser’s arrest and their own arrest. And looking at mothers, child protection system issues, 90 percent got at least some help, and 83 percent got at least some help with custody and visitation issues.

These are the short-term, self-reported outcomes. This is the way they were stated, and you can see that they are really quite positive. They involve categorically thinking that they got information; they gained confidence; they gained a sense of agency that they can do things for themselves; and they feel more hopeful. One of the consistent findings. I might also point out that two of these items are now being used by the FVPSA office and are being required as reporting for all the funded shelter programs across the state, and that is, “I know more ways to plan for my safety,” and “I know more about community resources,” and those are selected in part because while this is clearly not a controlled experimental design, one of the major controlled experimental design studies found that people who had more information and more knowledge of resources had longer — better longer-term outcomes. They were safer, they had better well-being, they had better connection with their community, and so this is particularly important in that context.

As for the children, the mothers reported and the — the mothers reported about their children that they believe their children feel more supported, have more understanding about what’s been happening and are better able to express their feelings without violence. I did bypass some of the voices, so let’s back up and the women — what the survivors said about the overall, because of their shelter experience, how they’re feeing.

Female Speaker 2: One woman said, “Being here has helped me develop a sense of worth, knowing I don’t have to take abuse of any kind from anyone.”

Female Speaker 1: “I am more confident in my decision-making process. I can count — I can count [inaudible]to all the trained professionals here at the safe place; I changed myself, but had a great deal of coaches. I feel that someone had — I feel happy that someone had thought of abused people and founded a shelter to protect us and teach us that we have worth as human beings. Thank you for the support.”

Female Speaker 2: “Since I’ve been saying in the shelter, I went from slamming dope to clean and sober. Thank you.”

Female Speaker 1: “I feel like I’m a new person, and my standards are much higher in every aspect of my life now. I’m ready to live again.”

Lyon: And, from the mothers about their children:

Female Speaker 1: “My children know now that we don’t have to live in fear.”

Female Speaker 2: “Because of separation from my abuser, as well as my experience with a particular counselor, I can already see a change in my daughter’s concepts of her personal power, and women’s empowerment and her pride in me has grown greatly.”

Female Speaker 1: “My son went from being in learning disabled to high honor roll student. More goals were achieved by both of my kids here than ever in their life before this.”

Lyon: OK, one of the notable things about this study is that we asked about other kinds of experiences as well, and about the problems that they had encountered their conflicts and the source of them. You can see that conflict with other residents leads the list as the most commonly encountered problem. Although notably, nearly three-quarters of those were resolved. This is actually the top half of the 18 specific items; these are the most common. The 18 specific items; 10 of those were related to rules and eight related to other kinds of issues.

You can see that the next three really have to do with resource issues, whether it’s resources within the shelter or resources based on just capacity and being able to provide the service. So, there’s quite a range. Choices of food: sometimes it just wasn’t good; sometimes it wasn’t culturally appropriate; sometimes there was concern that there was not food that was particularly useful for people who might be pregnant. Some of the child discipline issues had to do with other residents disciplining their children, so a lot of it has to do with issues of group living that certainly staff are quite familiar with.

This gives you an indication of what they said about the support of respect shown by shelter staff, and this is the “strongly agree,” if you add in the “agree,” the range of responses is 98 to 92 percent, so this is the — so we’re reporting the “strongly agree.” And, I have asterisked the bottom three, because we did some particular comparisons. We looked at people who identified as heterosexual and those who had other identification as far as sexual orientation and the percentages were the same. Caucasians compared to people of color, the response was the same. And, people who were under 18 or 65 or older compared to the rest, the response was the same. So, think that is a powerful testimony.

And here’s the overall rating: very, very helpful or helpful, 94 percent. And we’ll give you some sense of just a couple of quotes from survivors on this.

Female Speaker 2: “I want every woman to be safe. I am here ’cause I’m 21 years old, and my boyfriend tried to stab me to death. He beat me so bad one time, all I could feel on my face one time was blood, but because of this shelter, I am alive today. So be safe, and God bless.”

Female Speaker 1: “I was able to get a good night’s sleep without thinking about who and when — when will come through the door and how. I wish there was a half-way house or shelter to go after my stay here.”

Female Speaker 2: “When you think of asking for help, you’re the only one in that situation, they’re gonna laugh at you, but when you arrive, everything is different. They respect you as a person, as a woman, as a mother, and above all, they let you know that you have a lot of options, that you don’t have to live in a situation you don’t want to. Housing, food and respect, there is everything here. They’re all understanding and very respectful and above all, they listen and don’t judge you. God bless you. You have saved a lot of lives and mine too. Thanks.”

Lyon: So in general, we concluded in fact that shelters do address a compelling need not met elsewhere. That survivors don’t turn to shelter as a first resort. Sometimes people think that it’s — that there are other motivations, but this study speaks to the concerns and the, the fears that, that many people had before they came to the shelter. It also demonstrates that programs respond to diverse people and a wide range of kinds of needs. We also learned that the primary needs that were expressed by survivors in shelter are very basic ones: safety, information, help with the children, help with the emotional stress and trauma that they’ve experienced, and help connecting to or creating economic resources.

Now, this was not necessarily true across the board, but shelters actually meet most needs for most residents. Definitely exceptions, but this as a general pattern was born out: that shelter staff do help to resolve most of the problems that arise. Sometimes the problems that arise are beyond their control because of resources and other times, maybe they need some more training relating to conflict resolution or other kinds of issues, but they resolve most of the problems that arise. And that survivors rate their experience quite highly and say that they have changed as a result of their time there.

Clearly, resource challenges do remain an issue. Mothers’ concerns about their children remain primary for those who are mothers. There’s evidence that continued staff training in conflict resolution is very important and would be useful on an ongoing basis. Given the concerns about eligibility and the prior experience of not being able to get into shelter, there are some shelter rules and eligibility criteria that in some circumstances weren’t for their consideration and that language and cultural competence remained crucial. So, I’d like to turn this over to Anne Menard, for some additional —

[Side talk.]

Lyon: Oh the next slide is really context information.

Anne Menard: OK, well, I’ll move to that and then — OK, thank you for coming to this workshop. We are obviously very excited about the study and, and what it tells us. I, before I start, and I’m just gonna talk for about five minutes, and then we should have about 20 minutes or so for questions so, we really did want to allow time for us to engage you in some conversation. We expect you to have some reactions to this, surprises or things that you hadn’t thought about before; that was certainly true for us.

I really want to add my thanks to Angela and her colleagues at the National Institute of Justice and to Marylouise and her colleagues at the FVPSA office over at HHS. Working with them was great, and we appreciated them providing, finding the resources for this study. And, also to the coalitions and programs who participated. This was, you know they took this on, on top of everything else they do all the time, and for those of you, I want to honor the advocates in the room and those of you that have provided support to domestic violence programs. There’s a lot on their plate every day, and for them to add this to it was really important to us. They participated because they believed with us that the findings were, were gonna be important, and I think they were, they were really pleased at the end of the day to have participated; so again, we’re very grateful to them.

And, of course to the survivors, who in addition to dealing with the crisis that they were, they were in, at the point that they came into the shelter took time to give us this kind of feedback. And as, and as Eleanor pointed out, in addition to doing the check boxes, really wrote a lot of, a lot of words for us, and that’s very special for those of us. This movement was really founded initially 30 years ago in response to women who spoke out about what was happening to them at home, and so it’s real important to many of us that this study captured not only numbers, but real voices, including voices of criticism and critique in addition to, “This saved my life,” “here’s what you can do better.”

In terms of implications, there are really two that I really want to focus on. One was the implications for policymakers and then implications for the advocacy community itself. In terms of policymakers — and, and by that I’m talking about Congress and other funders at a, at a federal level and a state level — they really, they’re asking reasonable questions about how the funds that they are providing through the Family Violence Prevention and Services Act, through the Violence Against Women Act, at a federal, state and local level, how, how are these funds that are available for a shelters, you know, how they play out. “Does it matter? Are shelters still vital and necessary?” I mean, it’s now 25 years later. “Are they still, do they still play a critical role in our response to domestic violence? Why? For whom? And what can we do better?” So those were all very reasonable questions that Congress and other folks are asking about this, and up to this point, we really haven’t been able to tell the FVPSA story, the story of the Family Violence Prevention and Services Act, which is the primary federal funding stream that, that supports shelters and, and the full range of domestic violence programs. We haven’t been able to tell that story well.

We have, you know, certainly the voices of advocates who say, “Of course, this is what I see every day. This is what I see every day.” And other individual advocate stories, but this is really the first comprehensive study of this type that has, again, captured numbers and voices that, that again, help us understand or create a picture of something that most people don’t get to see every day unless you, unless you work at a shelter, so that’s been, that’s been very important.

I think one of the things particularly for policymakers, one of the parts of the story that we’ve been struggling to tell is that while legal advocacy is incredibly important in the Violence Against Women Act of 1994, and 2000 and 2005 provided really vital resources for domestic violence programs to make sure that when survivors are going in to court for protection orders or, or in criminal proceedings that they have advocates with them. We know that it’s also important that advocates have help around housing, around getting jobs, around job training, around education, those other, what we call “economic advocacy issues,” and certainly the, the, the current recession we’re in has really underscored that and highlighted it in bright yellow how important that is, that, that a protection order doesn’t necessarily, you know, help you find a safe place to stay. It doesn’t help you replace the income that you were dependent on that the, that the abuser provided.

So, I think what I’m most excited about, or one of the many things that I’m most excited about related to the study, is that women identified a range of issues, including the importance of legal — continued funding for legal advocacy, but the need to expand our ability to provide a full range of economic advocacy issues as well. And, and advocacy around immigration issues, and again, some of the other issues that we’re trying to raise the profile of because we, we  knew from our experience every day that they were important, but now we have again, both qualitative and quantitative data to really support that, so that’s, that’s very important. This gives us a good support for the, for the story we’re trying to tell and helps, I think, policymakers feel more confident that the investments that they’re making are, are right, and they need to continue and perhaps need to be expanded in some directions.

And then, of course the other group that, that I think this study has particular implications for are advocates who reasonably ask the question, “What difference does what we do every day make to the survivors that we work with?” And they hear that sometimes, but I think perhaps not in an organized way as the shelter site provides. They also want to know, “Does it matter? Why? To whom? And, what can we do better? And is there a match between what it is that we’re providing in our particular community through our  program and what it is that survivors coming through the door expect, need, want?”

And again, I think there’s affirmation that for, for, for many survivors that’s true, but there are survivors whose need — or coming, walking through that shelter door with a set of expectations and needs that we may not be as well prepared to respond to at this point. And, I’m just gonna pull out a few things for example, I think there was some really important information about the kinds of concerns that victims and survivors had before even reaching out to a shelter. That they’re concerned that they’re gonna be judged, that they’re gonna be found; they’re concerned, you know, that, that it’s not for someone like them. I think that provides a lot of important information to domestic violence programs about things that they can do to communicate better to both victims in the community that haven’t reached out yet, but also those who are in helping roles with them, their family so that they, so that they can be reassuring about what and close that gap between what people — you know, those concerns that people have and what their real experience will be.

I think there’s other areas. Eleanor mentioned conflict resolution — are we doing as good a job as we need to do to make sure that domestic violence program staff know how to resolve conflicts? Clearly conflicts arise. We know that anyone from a large family knows that conflicts arise in any kind of communal living situation; that’s even worse when you’ve got a whole bunch of people from different places, but, so conflict’s gonna arise. Do we anticipate that and make sure that, that staff have — feel confident in their ability to resolve different types of conflicts?

Do we need to look at our eligibility criteria and what — the, the reasons that we are screening women out of shelters? Of course we do. This gives us some, I think some very concrete guidance, and Eleanor’s doing some additional looking at the data, looking specifically at accessibility issues, and then we’re gonna immediately kind of get those to the, to the FVPSA grantees.

We very much wanted this study not to be driven by research, and some of you may have heard that phrase before but you know, where research is gathered, and then it, you know the people from whom it is gathered never hear about it again. We did — we’ve conducted a series of webinars for the advocacy community. We are posting —we’ve, I think — is that the next slide? — Right here? We’ve posted the shelter study findings, which includes the executive summary, this research in brief, the full study, the instruments; we’re putting up a copy of the webinar that Eleanor and many of us were on, a frequently asked questions, a bunch of other stuff, as well as for example, here’s the finding “conflict resolution is important training for staff to have,” here’s some resources that will, that you can use to make sure your staff get good conflict resolution training.

So, we’re really trying to connect the “here’s the data,” “here’s what survivors said,” “here’s resources that will help you.” In the next couple of weeks we’re gonna have a whole — there’s some pieces up there now — in the next couple of weeks, we’re gonna be adding more. We’re gonna continue to make, you know, do whatever we can to have this be a living document that the advocacy community and policymakers can use to shape the services they provide, as well as the policies and supports that we provide to these programs.

So, we’re excited about the, the shelter study; we’re excited about the ways that we can use it to improve services, and to make — enhance the responsiveness of domestic violence programs to the, to the issues that victims and survivors are identifying now. So I’m gonna turn over to Angela.

Moore: First, I just — I’m a break the rule, and I would like right now, especially for those of you who are gonna run out in a minute, to thank all the panelists for their presentation.


Moore: Thank you. OK. So, we do have some time for questions, and as a moderator, I get to take the liberty to ask the first question, and maybe this might be a hard question, but it’s one that, that we’re already getting and this is for, for Eleanor in particular. So critics, researchers, etc., will say, “Well, this wasn’t a randomized control study so how can we say that shelters are effective or that they really work?” What’s your response?

Lyon: Well, that was, that is, that is a tough question. Certainly if we had huge amounts of, of resources and an ability to draw a random sample for, from across the nation, we would have been interested in being able to do that. We have people in crisis, we have a wide variety of other reasons why people may or may not be able to or willing to participate. We do know that 81 percent of the programs in those eight states did participate. We did try to calculate a response rate compared to the numbers but, shelter programs don’t consistently keep those that are date bound in a way that we were able to do that.

It resonates very much with the advocates and the knowledgeable people who have heard the results so that this squares with their experience. We did have a wide range and a diverse range of people and, and regions who, who participated. I would not go to the bank and say, “Well, if we found 51 percent said this, then that means it’s exactly 51 percent across the nation of all survivors who felt this way or had this experience.” So, I see this as a start, it is — but given all the, all the constraints, I think we, we actually did a quite a good job of a, of a reasonably representative response.

Moore: Thank you. Any questions? OK. In the white right here.

Questioner: First of all, I want to say for shelter advocates, justice advocates, shelter director — thank you for all the work that’s involved in doing this. I know that it’s very time — and researcher — time consuming and, a, very important, and I think that the results are very indicative. Some things we can look at across the country around policy and shelter criteria. My question is somewhat related to the protective order or restraining order issue, there’s a lot of content in there and a lot of numbers — strong numbers around assistance with civil advocacy, around orders of protection, restraining orders and things like that. And my current role, I manage a, the first and only automated protective order notification system, which is being implemented across the country and a couple different states, so my question was, was there any indication that victims and survivors were entering shelter for fear of retaliation immediately following the issuance of an order because they were unaware or concerned, they were unaware of the, you know, the, the mindset of that abuser, and if in our designing of information, accessible information for victims of notification services, what can we do better to make sure that we’re addressing some of those needs?

Lyon: I think you highlight a very important need, it — that is one of many areas that we did not explore in depth. I think this is a kind of overview and gives some, some direction for much further research that would be useful to conduct. I mean there were lots of questions that I was dying to ask that we had to say, “no, we have to set a limit; we have people who are in shelter and in crisis, and so we need to, to keep it simple.” And so to answer the questions that we identified, that would have been something that would have been very useful, I think you probably know anecdotally that that does happen, and I couldn’t tell you the extent of which from the study.

Questioner: I actually had a second question from my company: This is really regarding transitional housing. Of those programs that you had, were you able to do any, any analysis regarding those that had further housing needs [inaudible].

Lyon: We did not ask the shelter program whether they were part of a larger program that had a transitional housing component. We do know that — as you heard some of the, from some of the voices — and those were just a few — that many were talking about the time limits and how the time that they were there was not enough to, to make the changes and to, to put all the pieces together that they needed in order to, to make the kinds of changes that they want to complete, completely make the changes that they wanted to in their lives and that, that would be useful. I think that’s, that’s one of the clear implications of this — is that transitional housing services are critical and in dire need, but we weren’t able to do that kind of analysis.

Moore: Yes, the woman in the green?

Questioner: Yeah, I’m, I’m a police officer in Connecticut, and I was just kind of disturbed by the fact that you said that the majority of the women that heard about the shelter from police were either white or Latino. Is that a reflection on the officers not communicating this to minority women or are the minority women not involving the police in their domestic violence issues?

Lyon: I can’t answer that question based on the study; I only reporting that there were those differences among the, among the survivors and their — their account of how they had heard about the shelter. I think that it’s, it’s notable that nearly a quarter of them heard from the police and that’s, that’s only, it was only 5 percent below the rate at which they had heard from advocates. So they had heard about shelter from a lot of different sources, so I wouldn’t be quite as alarmed about, about that necessarily, but it may indicate different responses and interactions with police or the courts.

Questioner: [Inaudible] on the part of the police officers, or is it just the fact that the minorities are not involving the police and that’s why [inaudible].

Lyon: I don’t think we have too much evidence that people of color are not involving with the police necessarily. Certainly in the Connecticut court system, it’s a majority of people going through domestic violence cases and domestic violence special sessions.

Moore: Yes, right here?

Questioner: [Inaudible] University of Pennsylvania. Did you find many variations in quality between sites? I know you [inaudible] identified where you could see that some were lower and others were higher [inaudible] reporting the averages. It would make more sense then — and if so were they — were there some urban/rural differences?

Lyon: We were able to do some analyses by capacity of the shelter so, so the size and that becomes a kind proxy because when we were looking at the county-level data, it became harder to say where exactly this shelter was and whether it was an urban one. So I think a good proxy would be the number of staff and the number of beds and the size of the shelter. One of the ironies that, that I found was that it was the mid-size and smaller shelters that people reported more satisfaction with than the larger, which may mean that in some cases there’s more of an institutional kind of feel to the, to the larger shelter even though they offer more services in theory, so that we did find. But we didn’t do it, you know, by specific program. There’s still a good deal of very particular analysis that we hope to be able to do, but we have a lot of other kinds of conversions to do ’cause there are — there were some culturally specific shelters that were involved in the study just based on descriptions of the staff, and we want to be able to pull that out. So, you suggest a real interesting line of analysis.

Moore: Let’s go to the back here on this side.

Questioner: I just want to get back to the police officer’s question because I think [inaudible]. We’re presenting a study tomorrow that’s funded by the National Institute of Justice, [inaudible] focus group [inaudible] women, yet none of them knew about [inaudible] even though we were hosting them, holding them at the shelter [inaudible]. But I think what we learned from our following interviews [inaudible] but we think [inaudible] but I think for the women we serve, it’s more blurred than that so a police officers, victim advocate might call [inaudible] …

Menard: I’m from a family of 10, so you know, I mean conflict, it just  so, so makes sense that, that, that, of course, I mean who gets to use the bathroom first is a, is a huge issue in large families. So when you’re in, in, in a domestic violence shelters are immeasurably more complex in their dynamics.

Lyon: And we do know from studies staff and programs and coalitions that conflict resolution training is one category of training that, that folks have been trying to provide.

Moore: OK, in the back here.

Questioner: [Inaudible.]

Lyon: It’s easy to have missed that in the blur of numbers there. Ten percent were age 50 or older. It was less than 1 percent who were over the age of 65. And we did find that they reported satisfaction levels and respect, specifically based on their age at equivalent levels to the other survivors who responded.

Moore: OK.

Lyon: But there were, there were some other kinds of differences, which are fully elaborated in the full reports, so I’d urge you to take a look.

Moore: OK, I’m sorry; go ahead.

Questioner: As having worked with children before, I would be interested – getting back to the conflict resolution question again — to find out, and I don’t know that you asked this particular question, but to find out whether those individuals who thought their conflict was resolved satisfactorily was because the person they were having conflict with left the residence.


Lyon: A terrific question, and no, we can’t answer that one.

Menard: But that’s not really the solution is it?

Questioner: I actually have three questions, but I’m going to try to cram it into one. Great that you do this by the way — so needed. But it is sort of the client satisfaction kind of approach that we sort of don’t know, I always feel like. Do you have real world outcomes? [Inaudible] that tell you how well you can do [Inaudible] safety for the time you’re there. We have longer goals that require following them after they leave shelter continuing safety planning, for example, that someday that they’ll have, you know [Inaudible] because the reality is it depends on the user and the [inaudible], and so we’re trying to accomplish something that we have no control over, and it looks like we’re not doing anything. [Inaudible.] On the other hand, housing — there is no housing. So, you know, you can’t really make people’s lives better in terms of [Inaudible]. So, I did a little mini version of this study in New York City, which changes things a lot [Inaudible].

Lyon: So I think in that context some of the very concrete needs that were identified, the fact that they, that they reported job or job training helped with housing, safe and affordable housing, etc., were things that the majority received; I mean, I forget the specifics and the degree of it, is notable so that we did ask those kinds of questions. And I think that having – I have more confidence in these results because of their link to experimental, longitudinal studies that followed women over a period of a couple of years and found that those particular outcomes were associated with longer term well-being and safety. So that adds some strength and credibility, I believe, to the utility of these findings along those dimensions, but you raise some really good points.

Questioner: [Inaudible.]

Moore: Well. Thank you so much for that question, but we are out of time.


Moore: I hope you all enjoyed the session and thank our panelists again.

Date Created: August 14, 2019