Replication Data to "Are average years of education losing predictive power for economic growth? An alternative measure through Structural Equations Modeling”

The model estimated in this document uses a set of variables that are available for a wide range of countries with different levels of development, resulting in a sample of 91 countries for the period 1970-2010. The file titled “Database PLS-PM” contains the data with which is possible to estimate the human capital index (ich) calculated in the paper.
The variables used and their notation is as follows:

FR= Fertility Rates

VAAS = value-added contributed by the agricultural sector to GDP

GNI = Gross National Incomes per capita

LE = Life Expectancy

MR = Mortality rate for children under five years

AYE = Average Years of Education

SPR = Student-Professor Ratio

EC = Energy Consumption per capita

PP = patent applications by residents per capita

Given the database is not complete for all countries or for all years, this missing data was complete through interpolation method. All variables were transformed by mean of logarithms, except GNI. In the case of EC and PP, block of returns on human capital, the manifest variables are transformed such that they may be retrieved in levels at a later stage.

2. Data to estimate the economic growth regressions

Cross-section: The file titled “Database – Cross-Section” contains the data with which it is possible to estimate the results shown in tables 1-5 of the manuscript. The variables used and their notation is the following:

grow = GDP per capita, rate of change

log(gdp75) = lag of GDP in 1975, logarithm

demo = a binary variable measuring the level of democracy in the countries

contes = indicators by principal component analysis to approximate the degree of contestation

inclu = indicators by principal component analysis to approximate the degree of inclusiveness

lnihc = human capital index estimated through PLS-PM, logarithm

lnaye = average years of education developed by Barro and Lee (2013), logarithm

lninves = investment in physical capital, measured as the average share of investment real to GDP, logarithm

lngov = average government consumption as a percentage of GDP, logarithm

lninfla = inflation measured by consumer prices, logarithm

lnpop = population growth rate, logarithm

lnich70, lnich75, lnape70, lnape75 lninves70 lninves75 lnpop70 lnpop75 = lags of lnich, lnaye, lninves and lnpop

dafri = dummy for African countries

Panel data: The file titled “Database – Panel data” contains the data with which it is possible to estimate the results shown in tables 6-9 of the manuscript. All variables are averages for the underlying period. The variables used and their notation is the following:

grow = GDP per capita, rate of change

lngdp75 = initial GDP in 1975, logarithm

demo = a binary variable measuring the level of democracy in the countries

contes = indicators by principal component analysis to approximate the degree of contestation

inclu = indicators by principal component analysis to approximate the degree of inclusiveness

lnihc = human capital index estimated through PLS-PM, logarithm

lnaye = average years of education developed by Barro and Lee (2013), logarithm

lninves = investment in physical capital, measured as the average share of investment real to GDP, logarithm

lngov = average government consumption as a percentage of GDP, logarithm

lninfla = inflation measured by consumer prices, logarithm

lnpop = population growth rate, logarithm

dafri = dummy for African countries