In 2017, three data volunteers, Eugene, Sid and Josh, have collaborated with Refugees Welcome Australia to help them match asylum seekers and hosts in the area of Sydney. How did we join our forces to leverage the power of the data for social good?
Refugees Welcome Australia (RWA) is the Australian branch of a global community that seeks to provide safe and welcoming accommodation to refugees. Their mission not only focuses on the primary needs of asylum seekers. It also aims to change mentalities on both sides and create an environment where refugees are not marginalised. This dynamic and resourceful team was trying to collect more data and leverage its insights on different levels: supporting the assumptions that certain areas of Sydney had potentially more rooms to provide than others; compiling and understanding the outcomes of their actions and creating an algorithm that could provide the best possible match for hosts and refugees. These different objectives would allow them to support their funding applications but also creating a more efficient strategy when signing up the hosts.
At our very first meetup, they presented us their three main challenges:
Data collection — gathering valuable data for the best match possible between refugees and hosts
Measuring their impact — what proxies can they use to assess the quality of their services to the refugees and the community
Domain insights — what insights can they leverage to target the most relevant areas for their campaigns
During the following months, Eugene, Sid and Josh communicated constantly with the RWA team to find the areas where they could help the most and to suggest ways to use these insights for their campaigns. The two streams that we ended up focusing on were: mapping empty dwellings in the Sydney area with open data and supporting their refugees-host matching process through data collection. We focus in the past months on the first part, finding and mapping empty dwellings in the Sydney area.
Mapping potential dwellings
RWA was interested in identifying dwellings in Sydney that were likely to be under-occupied and could potentially host refugees. Mapping such dwellings could help RWA decide where to focus in recruiting hosts, plus their total number could also serve to demonstrate the feasibility of their approach in the famously housing-poor Sydney area.
Luckily we were able to find the data we needed on the Australian Bureau of Statistics TableBuilder website which allows free access to a subset of data from the 2016 Census. It enabled us to find, by postcode, an approximate number of dwellings with the number of rooms greater than the number of usual occupants. We could also filter out households that were likely unsuitable for RWA — such as those with children (as RWA doesn’t yet have the resources to comply with child safety legislation). The output could then be visualised using Tableau’s freely available mapping tool.
Despite some limitations (such as the inability to distinguish adult vs young children) our ability to gather adequate information once again shows how open data can be useful to community organisations.
What are the results of the first stream?
RWA used the analysis provided to support their projects to local authorities. Having these proof in addition of the qualitative research they had already made helped the project have more credibility.