The impact of sustained attention on labor market outcomes: The case of Ghana

• DOI: http://doi.org/10.1111/rode.12553
• Date: 01 February 2019
• Author: Dhanushka Thamarapani,Chih Ming Tan
• Published date: 01 February 2019

REGULAR ARTICLE

The impact of sustained attention on labor market

outcomes: The case of Ghana

Chih Ming Tan

1

|

Dhanushka Thamarapani

2

1

Department of Economics, College of

Business and Public Administration,

University of North Dakota, Grand

Forks, ND, USA

2

Department of Economics, California

State University, Chico, CA, USA

Correspondence

Dhanushka Thamarapani, Department of

Economics, California State University,

Chico, 400 W 1st Street, Chico, CA

95928, USA.

Email: dthamarapani@csuchico.edu

Funding information

Greg and Cindy Page Faculty Distribution

Fund

Abstract

In this paper, we go beyond traditional measures of cog-

nitive abilities (IQ) in explaining labor market and social

outcomes in developing countries. We exploit a rich data-

set from Ghana that provides information on demograph-

ics, labor market outcomes, and a direct measure of

cognitive ability along with other test scores to construct

a measure of sustained attention. Our work is therefore

related to the broader literature in psychology on the

importance of executive function on individual behavior

and outcomes. We find that, at least for the case of

Ghana, after controlling for IQ and other covariates,

higher levels of sustained attention are associated with

higher educational attainment and a higher likelihood of

being employed in a white‐collar job.

KEYWORDS

cognitive and noncognitive abilities, education, executive function,

Ghana, occupational choice

1

|

INTRODUCTION

Following the seminal work of, notably, Heckman and co‐authors,

1

there has been a deep interest in

the impact of cognitive abilities on economic outcomes where economists have generally employed

standard measures of intelligence in psychology. These measures are principally based on psychomet-

ric tests; for example, the intelligence quotient (IQ) and general intelligence (g‐factor) which were

used to predict various socioeconomic outcomes. Cawley, Conneely, Heckman, and Vytlacil (1996)

note that there are other likely factors that potentially determine wages and occupational choice. Thus,

the emerging literature on noncognitive abilities has taken a prominent role. A more recent study by

Gill and Prowse (2016) finds that individuals endowed with noncognitive abilities (agreeability and

DOI: 10.1111/rode.12553

Rev Dev Econ. 2019;23:155–171. wileyonlinelibrary.com/journal/rode © 2018 John Wiley & Sons Ltd

|

155

emotional stability) are better performers overall and fast learners. A key measure of noncognitive

abilities has been the Big Five factors: openness to experience, conscientiousness, extraversion, agree-

ableness, and neuroticism (or emotional stability); see Borghans, Duckworth, Heckman, & ter Weel

(2008). The Big Five factors have therefore been employed to analyze the effects of noncognitiv e

abilities on a variety of outcomes.

2

One major drawback with the above approach is that they can sometimes suffer from limited

theoretical underpinning and measurement. For example, there is a lack of consensus on concepts

such as conscientiousness –how to define and measure it given the context specificity and subjec-

tivity of its nature, thereby limiting its predictive validity (Paunonen & Ashton, 2001). Further-

more, proxy variables for these noncognitive abilities typically rely on self‐reported data. Such

data could be misleading as respondents may bias their responses by selectively enhancing their

positive traits while downplaying negative ones (Hirsh & Peterson, 2008).

Our work is an attempt to overcome these limitations. We focus on a developing country

3

and

construct a measure of noncognitive abilities –specifically, sustained attention –that is not based

on self‐reported data. For this endeavor we use two datasets –the Ghana Living Standards Survey

wave 2 (GLSS 2) and the Ghana Education Impact Evaluation Survey (GEIES). Our variable of

interest is constructed based on respondents’performance on the Simple and Advanced versions of

Mathematics and English tests administered in GLSS 2 and GEIES.

4

We exploit the fact that the

difficulty levels between the Simple and Advanced tests were dramatically different to construct

our measure for sustained attention.

The Simple tests are elementary, consisting of eight questions, and correspond roughly to profi-

ciency levels that would be attained by fifth graders.

5

The Advanced tests consisted of more ques-

tions (29 for English and 36 for Mathematics) and correspond more closely to proficiency levels

attained by ninth graders. Based on the length and difficulty level of the Advanced tests, if a

respondent has done well on (i.e., passed) the Advanced test, there is no reason why he or she

should have made any mistakes at all on the Simple test, given the rudimentary nature of the

questions. We attribute any mistakes made on the Simple test in this case to a lack of sustained

attention on the part of the respondent.

6

More generally, our measure for sustained attention (so constructed) is associated with the broader

psychology literature on executive function.

7

Executive function is distinct from IQ and is defined by

the National Center for Learning Disabilities as “a set of mental processes that helps connect past

experience with present action.”Importantly, individuals have a limit on these functions and they are

not context‐specific unlike the Big Five factors (Morgan & Lilienfeld, 2000). Diminished executive

function “makes activities like planning, organizing, strategizing, remembering details, and managing

time and space difficult”for the individual. We contend that errors on the Simple test by respon dents

who have passed the Advanced test reveal potential deficiencies in executive function. We interpret

executive function as a measure of sustained attention because the psychological process most likely

to underlie both emotion regulation and executive function is attention (Rothbart, Posner, & Kieras,

2006; Bell & Deater‐Deckard, 2007).

It is important to note that both GLSS 2 and GEIES administered a standard IQ (Raven's Progres-

sive Matrices) test along with the Mathematics and English tests.

8

By supplementing the GEIES with

the GLSS 2, we have information on a rich set of individual, family, and community controls for each

respondent along with information on outcome variables such as educational attainment and occupa-

tional choice. We employ matching and propensity score methods to uncover the associati on between

higher sustained attention and labor market outcomes. We find that respondents with higher sustained

attention completed more years of schooling and are also more likely to be employed in a white‐collar

job as opposed to a blue‐collar job.

156

|

TAN AND THAMARAPANI