For what it's worth, I'm very much in Chris' "... really confused about what this model can reveal" camp.

Model building in situations like this is very challenging and constrained in all sorts of ways, so I'm reluctant to criticise too much without knowing any background to the authors work. But my list of questions about their modeling choices is rather long and I see very few caveats elucidated in the paper. I'd love to hear a justification for why this model can do what the authors claim it can ("identified and interpreted the unique contribution of international student number to rental cost") - I don't see it myself.

In the UK where I am now we are seeing many similar issues (HEIs trying to increase international student numbers for financial reasons; government feeling under pressure to reduce immigration and in various ways seeming to try to resist that increase) - I'd be very interested if anyone knows of any other attempts to address these sorts of issues/relationships quantitatively.

I recently read an ABC headline “International students not to blame for rising rents, Australian study finds”. The bi-line was “There is no link between international student numbers and the cost of rent, according to the findings of a new Australian study that examined rental data between 2017 and 2024.” No link at all?

This sounded so counter-intuitive to me that I took a look at the paper (published in Higher Education and attached). It was based on 76 monthly time points from 2017 to 2023. The response variable is “rental costs of Australian residents” obtained from from SQM research. The predictors were (a) international student numbers, (b) the rental vacancy rate, and (c) the CPI for Rent.

I am really confused about what this model can reveal. CPI for rent is an ABS index of rental costs. It strikes me that this is not much different to the response variable rental cost. The correlation (from their tables 5 and 6) is 0.86. They then add vacancy rate (also from SQM research) and find it has a strong negative coefficient. And lastly they add international students which is not significant.

I guess there are different causal diagrams you could draw for this, but it strikes me that greater demand from international students would decrease the vacancy rate which would increase rents. So you have controlled away one of the key pathways by which demand could affect price. And I cannot really get my head around what this means when you also include CPI for rent.

The authors also mention that international student suffer from “their non-competitiveness in the private rental market due to lack of recognised rental history and credit” which one would expect requires them to offer a higher than market rent.

I contacted the author(s) about this and he responded: “Really appreciate your critical feedback on our work, and thanks for providing alternatives to look at the issue. We’re still learning and your sharing will definitely inform our further analysis.”

So not any push back at all. Am I missing something?