Bayesian inference on returns to schooling in Palestine

Other Title(s)

الاستدلال البيزي لتقدير العائد على التعليم في فلسطين

Dissertant

Ayyash, Muhsin

Thesis advisor

Sadiq, Tariq

University

Birzeit University

Faculty

Faculty of Graduate Studies

Department

Applied Statistics

University Country

Palestine (West Bank)

Degree

Master

Degree Date

2014

English Abstract

In recent decades, researchers focus on earnings and what determines it.

Education is an important investment tool, which can improve workers’ productivity and influence economic growth.

As an important determinant of labor productivity, it also therefore has an important inf_uence on earnings.

The human capital theory emphasizes the impact of human capital investment on labor market outcomes.

We can identify the earnings attainment derived f-om education as the return on an investment.

Therefore, the rate of return to education is known as the average increase in wages associated with one year increase in education.

The basic theoretical model for the human capital is considered by Mincer (1974), Backer (1964) and Schultz (1961).

There is various statistical inference methods have been applied to return to education.

Inference methods like (OLS, GLS, GLM and MLE) produce an inefficient estimation of the parameters.

Meanwhile, Bayesian inference methods can provide solutions to many problems in classical inference methods and may fit out them.

Furthermore, model comparison using Bayes factor produce more efficient results than classical methods (R-squared, AIC,^etc).

The aim of this thesis is to highlight the investigations about the estimates of returns to education in Palestine.

It’s mainly uses Bayesian inference to identify the relationship between education (schooling) and wage controlling for other variables that will affect the relationship like gender, experience or age, educational levels, labor sector and place of residence and compares between OLS and Bayesian estimation.

We find that Bayesian estimation method is more efficient than OLS estimation method since the standard error of the mean for Bayesian method is lower than those found in OLS method.

Finally, we correct for self-selection bias and we fnd that Bayesian estimation with Logit Heckman correction differ and greater than from those we’ve got form Bayesian estimation with Probit Heckman correction and Bayesian estimation without Heckman correction.

Main Subjects

Mathematics

No. of Pages

86

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-837121