The UBC study cited in the article conducted three different experiments related to homelessness in British Columbia. Firstly, they gave selected homeless individuals one-time cash transfers of $7,500 and measured various outcomes over one year. Secondly, they measured the general public’s perception of homeless people’s ability to manage money. Thirdly, they measured the general public’s perception of policies advocating for cash transfers, when presented with arguments that those policies can be effective.
A National Post article has suggested that a UBC study has a flawed scientific methodology that has undermined its results.
The National Post author argues that the study is biased since the researchers pre-screened their participants based on their likelihood of success. The researchers excluded individuals who had been homeless for more than 2 years, were substance abusers (to avoid the possibility of overdose), and those with serious mental illness. They also limited participation to those sleeping in shelters and not on the street. However, the point of the study was to compare how individuals within the same set of conditions would do with a cash transfer, relative to a control group who did not receive any money. They also explicitly cite that their findings are limited to those homeless populations who are “high functioning”, and that these results may not apply to the bottom 69 percent of the homeless population.
Another main claim of the author was that the UBC communications staff disregarded the results of the study and misrepresented the findings in their commentary article. However, the UBC article only discusses a component of the study which looked at public perceptions of how homeless people spend their money. They reported the public predicted homeless people would spend around 80 percent of their money on “temptation goods” such as tobacco and alcohol. The cash-transfer component of the UBC study showed that they did not spend more money on these goods than the control group did, contradicting public opinion on homeless people’s economic predispositions. Although they did not discuss the entire study, they didn’t misrepresent any findings.
The author also points to the fact that the study had “null results” of several different measurements including cognitive functioning. They indeed found null results for cognitive impact, which runs counter to previous studies done in developing countries which have had some positive results related to cash transfers. However, they note in the study that the cash amount of $7,500 was only 12 percent of an average yearly income in BC, while the other studies had more significant cash transfers representing more than double that relative income.
Finally, the author also suggests that the main significant finding of the study - that those given cash transfers found housing faster - could have been explained by the availability of social housing which increased during the time of the experiment. This claim is supported by the study itself, which points to this contextual factor as potentially influencing their results. They then point to the necessity of repeating similar studies to assess the significance of their results, which is a standard process in the social sciences. This is especially the case for experimental or novel studies.
The UBC study also discusses how the control group, who did not get into housing faster, eventually caught up with the cash group, since shelter clients typically exit the system within 1-3 months using existing supports. However, the benefit of cash transfers appears to be the net total savings relative to the cost of housing the homeless in traditional homeless shelters. The results indicated that homeless people given the cash transfers reduced their reliance on social supports and ended up saving the system $777 per person.