Wittgenstein Center for Demography and Human Capital

Project: Demography of Adult Skills

Introduction

This repository refers to the codes, data and procedures for constructing the dataset on the Skills in Literacy Adjusted Mean Years of Schooling for 185 countries in the period 1970-2015.

More details on the methodology can be found at:

Reiter, C., Özdemir, C., Yildiz, D., Goujon, A., Guimaraes, R., & Lutz, W. (2020). The Demography of Skills-Adjusted Human Capital [Working paper]. International Institute for Applied System Analysis. http://pure.iiasa.ac.at/id/eprint/16477/

Project Coordinators

Wolfgang Lutz

Anne Goujon

Researcher Team

Claudia Reiter

Caner Özdemir

Dilek Yildiz

Raquel Guimaraes

Usage instructions

Folder structure

This repository contains two folders:

1. SLAMYS_empirical
2. adj_fact

Estimation of the complete dataset starts with the SLAMYS_empirical, and then follows to the calculation of the adjustment factor.

Estimation of SLAMYS for 44 countries with empirical data (PIAAC/STEP)

The user can replicate estimation of empirical SLAMYS using "SLAMYS_empirical.R" in "SLAMYS_empirical" folder. Estimation includes four steps:

1. Estimation of standard of comparison
2. Calculation of base year SLAMYS
3. Estimation of standard ageing pattern
4. Reconstruction of SLAMYS 1970-2015

Original data can be found in Input folder. Large files are stored with Git LFS.

The output file is "slamys_1970-2015.csv" providing SLAMYS data in quinquennial time intervals for all 44 countries for the population aged 20-64.

Estimation of SLAMYS for the rest of the world (regression adjustment)

Attention: Data preparation files are renamed with the initial word "prep_"

The user can replicate SLAMYS calculations with the codes listed in "Estimation" section We also provide additional code used for data preparation and visualisation

The R Code for SLAMYS estimation can be found on the script "./code/slamys_estimation_1970_2015.R"

The data required are available "./data" folder, and regional averages are calculated in the R-script "./code/regional_averages_1970_2015.R"

We provide also additional files, as follows

• R Code to compute the QEI regression model: "./code/qei_estimation.R"
• Literacy is estimated in six steps starting with "prep_literacy_1_merge_unesco.R"
• Literacy data preperation also includes a Stata code at step 5
• Stata code for teacher-pupil ratio: "prep_tp_ratio.do"
• Stata code for education expenditure: "prep_edu_exp.do"

Folder description

./code: code for data preparation and for regression models ./data: raw and manipulated data ./figures ./results