Economics · Inflation – who cares? Monetary Policy in Times of Low Attention · Oliver Pfauti, University of Bern. The student's answers here are a worked example; the tutor's questions and the feedback are exactly what the interviewer produced.
Tutor · question 1 of 6Welcome, and thanks for watching that clip. Could you explain in your own words what the researcher actually claims about an 'inflation attention trap'?
StudentRight, so basically he's saying that if people just don't really notice or pay attention to inflation, then the economy can kind of get stuck. Like at the zero lower bound, where interest rates can't go lower, and the central bank tries forward guidance to sort of guide expectations, but it doesn't work because nobody's actually listening or paying attention. So it's this trap where low attention keeps you stuck in a bad period. I think it's mainly about attention being the missing piece, like people being inattentive sort of undermines the policy.
Tutor · question 2 of 6You said forward guidance 'doesn't work because nobody's listening' — what evidence does the researcher actually give to show attention is low, rather than just assuming it? How would he measure something like that?
StudentYeah, so I think he uses surveys and stuff, like asking people about their inflation expectations, and maybe how much they read the news or search for inflation online. There's something about Google searches or attention measures showing people just aren't paying attention. I'm not totally sure on the exact data, but I think he looks at whether people revise their forecasts when news comes out — if they don't react much, that shows low attention. So it's measured rather than just assumed, though I guess it's hard to pin down attention directly.
Tutor · question 3 of 6You mentioned people not revising forecasts when news comes out. Suppose instead they revised a lot — would that still count as low attention, or would your measure break down?
StudentHmm, yeah, that's a good point — if they revised a lot, then I suppose that would look like high attention, so my measure wouldn't really work. But actually I think it's more subtle, because you could revise a lot just based on like one big news story and then ignore everything else. So maybe attention isn't just about revising, it's about how consistently you update. I'm not sure though — maybe he uses a combination of measures rather than just one, because a single one could be misleading.
Tutor · question 4 of 6You mention combining measures because one could mislead — can you think of another field, like how we judge someone's engagement in a seminar, where a single indicator fails and you'd need several? What would a good composite look like?
StudentYeah, like in a seminar — you might see someone nodding a lot, but that doesn't mean they're actually engaged, they could just be tired or copying someone. So you'd want a few things together, like whether they ask questions, take notes, and maybe how their essays go. A good composite would sort of combine behaviours over time rather than one snapshot. I guess it's the same idea with attention — one measure could be noisy, so you'd want several. Though I'm not sure exactly how you'd weight them.
Tutor · question 5 of 6You said you'd combine measures but weren't sure how to weight them — so if a student asks one brilliant question but never takes notes, is that high engagement? How would you actually decide the weights rather than just picking them?
StudentYeah, that's tricky. I guess you'd decide by looking at what actually predicts the outcome you care about — like if note-taking correlates with exam success, weight it more. So it's kind of data-driven rather than just picking. For the brilliant-question student, you might weight the question highly if it shows real understanding, but if they never take notes maybe that's fine because they engage differently. I'm not sure though — maybe you'd test which combination best matches some benchmark, though that sounds quite complicated to actually do.
Tutor · question 6 of 6You mentioned testing combinations against a benchmark — if you could design the next study on this inflation attention trap, what would you want to find out, and how would you test whether attention actually keeps the economy stuck?
StudentYeah, so I'd want to find out if low attention actually causes the trap, not just happens alongside it. Maybe you could sort of vary how much attention people pay, like giving some groups more info about inflation, and see if that changes how they respond to forward guidance. Or use regional differences where attention is higher versus lower and compare outcomes. I guess the tricky bit is proving causation rather than just correlation — I'm not totally sure how you'd isolate attention cleanly, but something like a natural experiment might help.
You demonstrated a solid grasp of the core idea and engaged thoughtfully with measurement issues, but you need to support your claims with more precise evidence and commit to concrete proposals rather than hedging.