Thread by @angrist_noam

📣Excited to share our paper is now *published* in @Nature "Measuring human capital using global learning data" w/ Djankov @PennyG_Yale @hpatrinos

nature.com/articles/s4158…

We present learning data in 164 countries from 2000-2017 + explore key Qs in the human capital literature👇

The Harmonized Learning Outcomes (HLO) data is a public good, publicly available, and will be regularly updated on the @WorldBank data catalog:
datacatalog.worldbank.org/dataset/harmon…

We show using the HLO that even as enrollments have gone up and achieved high levels globally, learning levels are low and progress is slow. This gap in human capital formation has been referred to in the literature and global education community as ‘the learning crisis'

➡️Development accounting: w/ measure of ed quality, human capital accounts for 1/5-1/2 of cross-country income differences
➡️A middle-ground in a wide-ranging literature from 💯 to potentially none.
➡️Avg results mask heterogeneity -> important to include full global distribution

➡️We also show our measure of human capital is more closely associated with economic growth than measures in the Penn World Tables 9.0 & the @UnitedNations Human Development Index
➡️Likely because these measures of human capital largely omit learning.

👉By constructing learning data across 164 countries over time, we fill a key gap: broad coverage over nearly two decades and a measure of human capital with strong links to economic development.

The HLO data is a key ingredient in the World Bank Human Capital Index, combining learning and schooling through the Learning-Adjusted Years of Schooling education component of the index: openknowledge.worldbank.org/bitstream/hand…

worldbank.org/en/publication…

Read more about the latest version of the Human Capital Index here by Gatti and Kraay #InvestInPeople blogs.worldbank.org/developmenttal…

The HLO closely relates to a series of additional efforts to produce global learning data such as @UNESCOstat's effort (@montoya_sil) to construct a globally comparable proficiency scale and the Rosetta Stone project (@JustinSandefur @dev_a_patel cgdev.org/sites/default/…)

This paper builds on seminal work highlighting the gap between schooling and learning:

➡️@EricHanushek @Woessmann hanushek.stanford.edu/sites/default/…
➡️Pritchett's "rebirth of education" cgdev.org/sites/default/…
➡️@GEMReport
➡️WDR 2018 (+ the many included studies) worldbank.org/en/publication…

➡️The HLO complements the related Learning Poverty @WorldBank indicator focused on reading in primary school worldbank.org/en/news/press-…
➡️The slow/limited learning we find builds on multiple 'soundings of the alarm' highlighting the urgent need for proven solutions to boost learning

While learning is hard to budge, other work shows it *can* improve.

A review of 150 studies identifies approaches that deliver 3 years of high-quality schooling with substantial learning.

@DaveEvansPhD @deon_filmer @rglenner Rogers @shwetlena cgdev.org/publication/ho…

The Millions Learning report @BrookingsInst by @JennyPerlman @RebeccaWinthrop highlights a few principles and north star efforts to enable learning at scale. brookings.edu/wp-content/upl…

And a new panel, GEEAP, continually (!) reviews the evidence on cost-effective education approaches

@kwameakyeampong @AndrabiTahir Banerjee Banerji @dynarski @rglenner McGregor @karthik_econ @benlpiper @JaimeSaavedra22 @SYLVIASCHMELKES @HiroYoshikawaNY
worldbank.org/en/topic/teach…

Wonderful to see the "Measuring human capital using global learning data" paper out in @nature after a decade since starting work on the HLO and exciting to see so much complementary research and related efforts to promote learning for all.