Refugee Refusals Based on Selective Reading and Cherry Picking the Evidence

As refugee lawyers we’ve seen this before: a Refugee Protection Division decision that twists the evidence into knots to justify rejection. One of the most frustrating errors is selective reading, where the member plucks out isolated bits from country reports while ignoring the broader picture that clearly corroborates risk. It’s sometimes more than a  mistake many times it’s intellectual dishonesty.

Selective reading (or cherry picking) happens when an RPD member relies on minor qualifications (or silence) in some parts of the evidentiary record (like general human rights reports) to override detailed, specific evidence from other parts of the record (like sources like Response to Information Requests in the National Documentation Package). For example, they might note that a U.S. Department of State report doesn’t explicitly mention mistreatment of similarly situated individuals outside a certain region, and use that absence to dismiss corroborative proof contained elsewhere.

But evidence should be weighed for what it does say, not what it omits. It’s rare indeed that absence of evidence should be preferred over extant evidence.

This error turns risk assessment upside down, prioritizing speculation over facts. And … it may be that personal bias has crept in, and the decision maker is colouring the analysis in ways that have no place in fair adjudication.

The Federal Court has called this out repeatedly. Take Magonza v. Canada (Citizenship and Immigration), 2019 FC 1478. There, Justice Grammond tore apart a pre-removal risk assessment for doing exactly that. The officer quoted a tiny excerpt from a DOS report about government efforts on gender-based violence, but ignored the report’s overall condemnation of ineffective state protection.

As the court put it at paragraph 87: “These statements convey a finding that state protection for victims of gender-based violence in Tanzania is ineffective. In contrast, the excerpt quoted by the PRRA officer simply states a fact, without any assessment of the effectiveness of the government action plan or the ‘gender desks.’ Thus, it was unreasonable for the officer to rely on an isolated quote from the report to sustain a finding that is the exact opposite of the report’s conclusions.”

Further on, at paragraph 92, the judge drew a line between mixed evidence requiring judgment calls and cases where the proof overwhelmingly points one way, but the decision clings to “some minor qualification or positive aspect” to flip it.

That’s a failure to engage with the record. The court stressed that officers/decision makers must discuss contrary evidence and explain rejections.

Why does this matter so much? In refugee claims, we’re dealing with risk. In a recent RAD that we filed, the RPD justified the refusal of the claim based on a flawed internal flight alternative finding which was based on cherry-picked evidence . We’ve seen similar reasoning and successfully challenged this in the past but Justice should be done at the first instance (and now especially imperative given the restrictions on refugee eligibility recently announced, there may not be a full appeal on the merits for a first time risk assessment).

It’s important for the decision maker to holistically address the NDP, including RIRs; they are not to sidestep countervailing items by pointing to “silence” in general reports. Fortunately, there’s an appeal for most refugee decisions to the Refugee Appeal Division and a timely appeal results in a stay of removal. It’s disappointing that some members seem more interested in what’s not there than in the clear warnings that are. This approach doesn’t just result in unreasonable decisions it also erodes trust in the system.