The role of sharing data in ending homelessness

First Annual Canadian Homelessness Data Sharing Initiative

 By:  Nick Falvo

On May 4, 2016, approximately 40 people attended the First Annual Canadian Homelessness Data Sharing Initiative, sponsored in Calgary by the Calgary Homeless Foundation and the University of Calgary’s School of Public Policy.  Those attending included government officials, researchers and students.  Here are 10 things to know about the event.

1.  The Data Sharing Initiative was jointly organized by the Calgary Homeless Foundation (CHF) and the University of Calgary’s School of Public Policy. A major goal was to bring together persons with access to various forms of homelessness data for a one-day event for the first time.  Both CHF and SPP provided cash and in-kind support to make this inaugural event happen.
2.  When researchers want to access data on homeless persons in Canada, there isn’t a single point of access. Rather, data is both collected and accessed in a variety of ways.   When it comes to data collected about persons experiencing homelessness, there are three main types of data to understand.  First, there is administrative data (such as the data presented at the May 4 event by officials from Calgary, Ottawa and Toronto); this is collected on an ongoing basis by front-line professionals and often used by social workers.  Second, there is data gathered from Point-in-Time (PiT) counts (such as the data presented at the event from Montreal).  Third, there is survey data collected by researchers (such as the data presented at the event on Managed Alcohol Programs, and data presented at the event that was gathered from a study on homelessness in Northern Ontario).  The May 4 data-sharing event included discussions about all three types of data.
3.  Canada’s federal government has data on homeless persons from roughly half of the country’s homeless shelters. This includes data gathered via the Homeless Individuals and Families Information System (HIFIS) software; it also includes data gathered via data sharing agreements (which it has with the City of Toronto, the Province of Alberta, and BC Housing).  The federal government then uses this data for policy development, measurement and program evaluation.  The federal government uses some of the data for its Community Progress Indicators Reports, which it shares with its 61 Designated Communities. (Most of these reports aren’t publicly available, but the one for St. John’s can be found here.)    The federal government uses some of the data for public documents (e.g. the National Shelter Study). 
4.  All municipally-funded Toronto homeless shelters are required to use the Shelter Management Information System (SMIS) database system. This is a web-based software that tells administrators when people enter homeless shelters and when they leave.  It must be used by all 60 of Toronto’s city-funded homeless shelters.  Among other things, it helps   administrators tell where there are empty beds.  It also helps front-line staff to manage beds in their own programs.  And it helps staff in Toronto’s central access services to refer people in need to available beds.  In the future, city officials hope to use the SMIS system to assess how much ‘social work’ support a person will likely need once they’re referred to housing; they also hope to use the system to track how people do after they’re referred from emergency shelters to housing. (It’s also thanks to the SMIS system that municipal officials in Toronto know that, on a typical night, there are more than 4,000 persons staying in Toronto shelters, and that in a typical year, more than 17,000 persons sleep in a Toronto shelter for at least once night.)  More information on Toronto’s SMIS system can be found here.
5.  The City of Montreal has no centrally-coordinated database system for homeless shelters. However, the three main men’s shelters are using HIFIS, and other individual shelters in Montreal do keep data on their clients.  But those data are not coordinated or kept by one central body.  Likewise, Montreal-based programs that receive funding from the Homeless Partnering Strategy all keep data (which they have to provide to the federal government), though not necessarily using HIFIS.  Readers should be mindful that the homeless file in Quebec is first and foremost a provincial file; municipalities play only a minor formal role in program administration (e.g. the municipal government provides some funding to various social programs—such as day centres—that serve homeless persons).  All of Montreal’s homeless shelters are private, non-profit entities; in general, 60-70% of their budgets come from private sources (and those private funders have very few stipulations in terms of what kind of data must be kept).  Montreal did conduct its first PiT Count last year, and municipal officials have access to the dataset.  What’s more, in the summer of 2015, the City of Montreal commissioned a more detailed survey of homeless people; it was administered by four teams of paid professionals.  (Officials with both the City of Montreal and the Douglas Institute have access to data from both of these studies).
6.  Ottawa homeless shelters have an open system. Ottawa homeless shelters must use the HIFIS system as a condition of receiving funding from the City of Ottawa; and the centralized database is hosted by the City of Ottawa.  It is an open system in that staff at one shelter can see client records elsewhere in their respective sector—that is, staff in a men’s shelter can see information gathered on one of their clients by another men’s shelter.[1] Data, once collected, is shared on request by municipal staff with a variety of external actors, including University of Ottawa researchers and the Alliance to End Homelessness Ottawa.  (Note: municipal staff only release aggregate data—that is, no personally identifiable information.)  This open system has been in existence for almost 10 years.  Ottawa has never conducted a PiT Count.
7.  Throughout Canada, there are many examples of data repositories in the natural sciences and some in the social sciences, but very few in social policy. A trusted digital repository is another term for a certified digital archive (and an archive is a place you can deposit data for the long term, where it will be managed across space and time, and where it’s backed up and has a sustainable financial plan and technology transition plan).  One example in the natural sciences is the repository created for International Polar Year (IPY) data archives.  Another example with social science data is the Irish Qualitative Research Data (IQDA) archive.  With homelessness data, since there’s no central place to deposit them, such a data repository to consolidate data resources from all sectors and levels of government would be helpful.  In addition to the actual data, the repository would include information on who collected the data, methodological guides and a data dictionary with definitions; and these would all be kept together.  This could allow for a one stop shop of shared data—some data could be kept private, with restricted access for researchers only and other datasets could be open data.  For such an initiative to get started, someone would have to do the collecting, data, curating and cataloguing of the data.  They’d also have to contact sources of the data.  Librarians and archivists are experts in this area; they would need to be very involved in the early phases of the project.  Such an initiative would require resources (i.e. time, expertise and money) but in the long run could provide the sector with data for national, provincial and local planning.
8.  Canada’s major funding body for social sciences research has started mandating data management plans. Indeed, the Social Sciences Research Council of Canada (SSHRC) will soon start mandating data management plans.  They will also start mandating the preservation of data.  This will apply to research projects collecting large amounts of data; but it will not apply to municipal governments simply collecting administrative data, as these are not SSHRC funded research projects.  When research proposals go to SSHRC, SSHRC will expect to see money budgeted for data management; SSHRC will also ask researchers for data management plans.  Other national funding bodies may follow suit.
9.  The one-day event allowed for some invaluable networking. Canada is a big country, and there are staff in major cities who administer homelessness data who have never met their counterparts in other cities.  This event helped us to overcome some of those silos.  For example, one provincial official discussed HIFIS with a federal official during the event, and exploratory conversations have since begun about the possibility of that province introducing the HIFIS system into its homelessness sector. 
10.  The Data Sharing Initiative will become an annual event.

     

  The author wishes to thank the following individuals for invaluable assistance with this blog post:  Britany Ardelli, Steven Bulgin, Francesco Falvo, Louise Gallagher, Darcy Halber, Ron Kneebone, Eric Latimer, Tracey Lauriault, James McGregor, Kevin McNichol, Laural Raine, Aaron Segaert, Madison Smith and Shelley Vanbuskirk.  Any errors lie with the author.

[1] Ottawa’s homeless shelter system has four sectors:  men’s, women’s, family and youth.  Family shelter staff are allowed to see client data in all of the sectors because family shelters are the ‘overflow sector’ (meaning they sometimes take in clients from the other sectors).

Below are links to the minutes of the full day’s event, as well as the slide presentations.  Next year’s event will also be held in Calgary. 

MINUTES  May 4, 2016 Colloquium

The DI Data

Homeward Trust Edmonton Dataset

The At Home / Chez Soi Demonstration Project  Tim Aubry, PhD., C.Psych

The Health and Housing in Transition (HHiT) Study Tim Aubry, PhD., C. Psych. 

2016 Coordinated PiT Count – Overview and Data, HPS Dr. Patrick Hunter, Policy Analyst, HPS, ESDC

Dataset Description  HPS 2016 Coordinated Point-in-Time Count 

Data Description, The Mustard Seed Calgary   John Rook, PhD, Samantha Sexsmith

Patterns and Determinants of Housing Utilization and “Graduation” in Calgary  Ali Jadidzadeh, PhD and Nick Falvo, PhD

Homelessness in Northeastern Ontario Carol Kauppi, PhD   |   Activities

Montreal’s First PiT Homeless Count and Subsequent Survey  Eric Latimer, PhD

I Count MTL 2015 (Report)

Homelessness Data Discussion  Tracey Lauriault

Using Data for Evidence-Based Service Planning   Laural Raine, City of Toronto

Homelessness as a public issue  Aaron Segaert

Managed Alcohol Programs (MAPS): Implementation and Effectiveness  Bernie Pauly RN, PhD, Tim Stockwell PhD, Clifton Chow, MA, Kate Vallance MA, Ashley Wettlaufer, MA