Sam Stemper

Lots going on in the intervening years, but it is true that (in many places) foot traffic in Auckland is still far from its pre-COVID levels. K Road, Upper Queen Street particularly sluggish. Stronger recovery in Britomart/Commercial Bay area.

Main result: Colleges have large effects: college establishment increases a county’s notability rate by over 80%, and the number of notable people per birth year by over 3x.

Notability rates vary substantially across places. Here's a map of average rates from 1820 to 1980. Explore the data yourself: notability.samstemper.com/

Changes in 3G coverage are associated with decreases in test scores across all subjects. Our preferred estimates are 0.05 to 0.08 SDs: roughly equivalent to one quarter of a year of learning.

We first show that 3G indeed accelerated internet use among adolescents. 15 year-olds PISA respondents with 3G coverage are 7 percentage points more likely to use the internet daily, and 12 percentage points more likely to have a smartphone at home.

We link geospatial data on 3G coverage over time with 2.3 million student-level PISA records from 82 countries. PISA urbanicity data + pop. density data + 3G coverage shapefiles = within-country variation in 3G coverage over time. Here’s how this linkage looks in Czech Repub.

We study the rollout of 3G between 2000 and 2018. Compared to 2G, 3G allowed for faster data transfers, enabling the mobile internet browsing we know (and love?) today. 3G coverage and student use show similar over-time trends.

With the release of PISA 2022, the long-standing global decline in PISA scores has many of you asking: is it the phones? Ronak Jain and I have a new working paper that tries to shed some light on the question. Details below… 🧵