Gender Differences in Earnings Rewards to Personality Traits in Wage-employment and Self-employment Labour Markets

Abstract

In an extensive review of wage determination papers, it is concluded that the standard demographic and human capital factors explain little of earning differentials. Consequently, there is a growing interest among economists to include non-cognitive skills measured by personality traits in recent empirical literature to explain variations in earnings. In a bid to contribute empirical evidence to this strand of literature, this study examines the associations between the Big-Five personality traits (i.e., agreeableness, conscientiousness, openness, extraversion and neuroticism) and earnings, using the World Bank’s Skills towards Employment and Productivity (STEP) data on Ghana. The study employed regression techniques to estimate a series of semi-logarithmic wage equations that include demographic and human capital factors and the Big-Five personality traits to determine how important these factors are in explaining wage and self-employment earnings. Furthermore, the estimations of the wage equations are done separately for males and females to highlight any gender differences in the way personality traits contribute to earnings. Findings are largely consistent with the literature but uniquely demonstrate that in a power-distant culture like Ghana, where, traditionally, girl-child education has been relegated to the background, agreeable females, and not males, are rewarded in the formal wage employment labour market. However, in the informal self-employment labour market, conscientious males, and not females, are positively rewarded with higher earnings. These unique findings contribute to our understanding of the gender differences in the relative importance of non-cognitive skills in the formal and informal labour markets.

JEL Codes: J31, J24

Keywords

Non-cognitive traits Big-Five factor model earning differential incentive-enhancing

Introduction

Traditionally, economists have explained differences in earnings, using human capital variables, such as education, job training, experience and tenure based on the canonical model. While this approach has been successful at explaining part of earning differentials, there is still a substantial part of earning differences or wage inequalities that is unaccounted for by variations in human capital factors (Bowles et al., 2001; Cubel et al., 2016; Nyhus & Pons, 2005, 2012). Consequently, there has been an ongoing research interest to discover other factors that can be used, together with human capital variables, to explain earning differences. In a seminal work by Becker (1964), it is argued that cognitive skills, which are mostly measured by verbal, mathematical, and critical or problem-solving ability, can account for a significant portion of the variations in earnings. With increasing availability of data on cognitive abilities (skills), researchers have examined the importance of these abilities (skills) to earning differentials (Cawley et al., 2001; Heckman et al., 2006; Murnane et al., 1995). These studies have generally offered support for Becker’s (1964) argument albeit the share of wage variance attributable to cognitive abilities is found to be modest (Cawley et al., 2001).

In the past few decades, economists have begun to focus research attention on the role of non-cognitive skills (i.e., personality traits) in determining earnings (Fletcher, 2013; Heckman & Kautz, 2012; Mueller & Plug, 2006). For example, Mueller and Plug (2006) have demonstrated that the magnitudes or effect sizes of personality traits on earnings are similar to those commonly found for cognitive skills. Different from cognitive skills (and abilities), which are usually associated with mental capability to reason and solve abstract or complex problems, the term non-cognitive or soft skills is identified mainly with personality traits. It encompasses a wide range of factors such as self-motivation, planning capabilities, industriousness, self-control or self-esteem. With its origin from the field of psychology, a widely accepted taxonomy of personality traits is the Big-Five factor model (Digman, 1990; John & Srivastava, 1999; Roberts et al., 2006). This five-factor model identifies the dimensions of personality traits to consist of extraversion, conscientiousness, agreeableness, stability and openness, with each trait having a polar opposite. Each of these traits combines several interrelated facets to measure the wide range of thoughts, actions and behaviours.

In the analysis of labour market outcomes, the Big-Five personality traits have received wider application than any other classifications of personality. Part of the reasons for this out-turn is that the Big-Five personality traits are considered to be largely uncorrelated with cognitive skills; hence, they constitute truly distinct factors, which make them less prone to endogeneity problem unlike other measures of personality traits (Cubel et al., 2016). In addition, they are reported to be robust measures of personality across different cultures and samples (Barrick & Mount, 1991) and relatively stable after biological maturation or early adulthood (Cobb-Clark & Schurer, 2012). Empirical studies linking the Big-Five personality traits to labour market outcomes like earnings or wages have produced mixed results. Some of the papers have reported significant positive associations between the traits of conscientiousness, agreeableness, and emotional stability and earnings, while others have not found any significant relationships (Bowles et al., 2001; Cubel et al., 2016; Fletcher, 2013; Heineck, 2011; Nyhus & Pons, 2005).

While the literature documenting the effects of personality traits on earnings is growing and evolving, studies conducted so far are not only inconclusive but also skewed towards developed countries (Bowles et al., 2001; Cubel et al., 2016; Nyhus & Pons, 2005), with studies in Africa still lacking. Consequently, it is imperative to ask whether non-cognitive skills, particularly, personality traits, contribute to earnings variations in an African labour market setting, focusing on Ghana, where there is rich data set from the World Bank’s Skills towards Employment and Productivity (STEP) survey. In fact, given that labour market earnings are the main source of income for the vast majority of people (Brunello & Schlotter, 2011; World Bank, 2012), it is important to understand, from the perspective of a developing country, which personality traits are rewarded or punished in the labour market. In an attempt to address this critical gap, this study argues that the effect of the Big-Five personality traits may not only differ by gender as demonstrated in previous studies, but, also, the gender differences in personality could contribute differently to differences in wage and self-employment earnings, which has not been addressed in any existing study. Thus, the purpose of this study is set against these gaps and seeks to examine the effect of the Big-Five personality traits on wage employment and self-employment earnings in Ghana. In line with this purpose, the following two research questions are addressed:

Are personality traits important in explaining wage and self-employment earning differentials in Ghana?

Are there gender differences in the associations between personality traits and earnings from wage and self-employment in the Ghanaian labour market?

In addressing the above-mentioned research questions, our study departs from similar studies in two main respects, which constitute the basis of the contributions of this study. First, unlike other studies that have used the STEP data (e.g., Nomura & Adhikari, 2017; Tognatta et al., 2016; Valerio et al., 2016), the effect of non-cognitive skills on earnings is yet to be examined in Ghana. The studies by Tognatta et al. (2016) and Valerio et al. (2016) focused on seven countries (namely Armenia, Bolivia, Colombia, Georgia, Kenya, Ukraine and Vietnam), while the focus of Nomura and Adhikari’s (2017) study was on Bangladesh. Findings from these studies suggest that there is significant heterogeneity across countries in terms of how non-cognitive skills are valued in the labour market, and hence the departure to focus on Ghana. Second, aside the focus on Ghana, this study departs from Nomura and Adhikari’s (2017) study, which focused only on the formal sector to broaden the scope to examine the associations between the non-cognitive skills and earnings in both formal (wage employment) and informal sectors (self-employment), given that over 80% of economic activity and employment occur in the informal sector in Ghana (Koto, 2015; Osei-Boateng & Ampratwum, 2011).

Given the aforementioned differences, the findings of this study make important contributions to the literature. First, although the findings are largely consistent with the literature, some associations between the traits and earnings may appear to be specific to Ghana. For example, the findings have unearthed that in a power-distant culture like Ghana, where, traditionally, girl-child education has been relegated to the background, agreeable females, and not males, are rewarded in the wage employment labour market where typically males are dominant. However, in the self-employment labour market, where females are typically dominant, conscientious males, and not females, are positively rewarded with higher earnings. These unique findings contribute to our understanding of the relative importance of non-cognitive skills in the formal and informal labour markets.

The rest of the article is organized into the following sections: literature review on personality traits and earnings, methodology, results and discussion, conclusion and limitations and suggestions for future research.

Literature Review

Wage Determination and Non-cognitive Skills

In the analysis of wage or earning differentials, human capital theory features prominently (Becker, 1964; Mincer 1958). This theory posits that individuals possess some cognitive productive skills that are innate or developed through education, training and work experience, which provide input into the production process. Such bundles of skills are sold to firms at equilibrium wage rate. Thus, any systematic difference or variation in earnings or wages can be attributed to differences in human capital attributes since compensations or earnings are determined by the skills set of an individual. Empirical studies focusing on human capital factors and cognitive skills have found evidence to support the human capital theory, but the share of wage variance attributable to these factors is modest (Cawley et al., 2001). For example, from the Dutch Brabant and the Netherlands’ Organization of Strategic Labour Market Research (OSA) surveys, Levin and Plug (1999) reported that the explained variance of wages was 0.26 and 0.35, respectively. Hartog et al. (1993) also found that the explained variances of wages were 0.22 and 0.39 for the public and private sectors, respectively. These studies, together with the literature survey conducted by Bowles et al. (2001), show the inadequacy of human capital factors in explaining wage differences. The authors noted that individuals who have similar characteristics still received quite different earnings.

In light of the significant variations in earning or wage differences that are unaccounted for by human capital factors and cognitive skills, there has been an ongoing growing research interest to discover other factors that can explain earning variations among individual of similar cognitive productive skills. Early research efforts in this area among economists and psychologists have shown that non-cognitive skills like self-esteem and self-directness can positively influence wages (Murnane et al., 2001; Osborne, 2000) and in a higher order of magnitude than human capital factors (Goldsmith et al., 1997). More broadly, the term ‘non-cognitive’ skills have been defined as personality traits or patterns of thought, feelings and actions or behaviours that are relatively stable and enduring characteristics of human beings (Borghans et al., 2008). In their works, Goldsmith et al. (1997) and Bowles et al. (2001) present strong arguments why non-cognitive skills or psychological capital should be included in studies on earnings. They even concluded that any analysis of the effect of human capital variables on wages that does not include or address psychological capital could be biased.

As argued by Bowles et al. (2001), employers may value some personality traits because they promote labour productivity by facilitating the production of positive work efforts. Hence, they referred to personality traits as ‘incentive-enhancing preferences’ that allow an employer to induce effort or productivity at lower cost. In consonance with this argument, the effect of personality on productivity or job performance has been investigated in occupational psychology and found to be significant (Barrick & Mount 1991; Tett et al., 1991), providing justification for the economist to believe that earnings could be related to personality traits. Another reason why personality traits may be believed to have incentive-enhancing quality and for that matter determine earnings as argued by Bowles et al. (2001) is that employers have lately conducted personality tests as part of their recruitment and selection processes.

As the inclusion of personality traits in earnings studies has gained wide acceptance, the emerging evidence is that different personality traits predict earnings differently and according to the gender of the worker (Mueller & Plug, 2006; Roberts et al., 2006). Therefore, in this study, we seek to examine the associations between personality traits and earnings for wage and self-employment and possible gender differences in the associations. We focus on personality traits for the reason that these are relatively enduring characteristics or dimensions of individual differences that have the tendencies to show consistent patterns of thought, feeling and actions (Costa & McCrae, 1990; Roberts, 2009). In psychology, there seems to be a controversy over which personality traits are the most important; however, the Big-Five factor model (Digman, 1990; John & Srivastava, 1999) by far appears to be the most comprehensive categorization and robust taxonomy of personality traits (Mueller & Plug, 2006). In economic psychology, consensus seems to be increasing in favour of the Big-Five factor model as the personality categorization that influences a wide variety of economic outcomes (Barrick & Mount, 1991; Goldberg, 1993; Heineck, 2011; Judge et al., 1999). According to Goldberg (1993), one advantage of the Big-Five factor model over other taxonomies of personality traits is that it provides a unifying, hierarchical taxonomy of the traits, which can be used in either broad or narrow studies of personality influences (Marshall et al., 1994; Smith & Williams 1992).

Table 1.

The Big-Five Personality Traits

Dimension	Facet (and correlated trait adjective)
Extroversion versus introversion	Activity (energetic), assertiveness (forceful), excitement-seeking (adventurous), gregariousness (sociable), positive emotions (enthusiastic) and warmth (outgoing)
Agreeableness versus antagonism	Altruism (warm), compliance (not stubborn), modesty (not showing off), tender-mindedness (sympathetic), trust (forgiving) and straightforwardness (not demanding)
Conscientiousness versus lack of direction	Achievement striving (thorough), competence (efficient), deliberation (not impulsive), dutifulness (not careless), order (organized) and self-discipline (not lazy)
Neuroticism versus emotional stability	Angry hostility (irritable), anxiety (tense), depression (not contented), impulsiveness (moody), self-consciousness (shy) and vulnerability (not self-confident)
Openness versus closedness to experience	Actions (wide interest), aesthetics (artistic), fantasy (imaginative), feelings (excitable), ideas (curious) and values (unconventional)

Source: Adopted from Mueller and Plug (2006, p. 5).

By way of description, the Big-Five factor model consists of extraversion, conscientiousness, openness to experience, emotional stability and agreeableness with corresponding polar opposites such as introversion, unconscientiousness, unintellectual/not open to experience, neuroticism and hostility, respectively. The characteristics defining these traits are presented in Table 1.

Hypothesized Relationships Between the Big-Five Personality Traits and Earnings

Researchers (e.g., Heineck, 2011; Lee & Ohtake, 2012) have made theoretical conjectures about the link between the five-factor personality traits and earnings by invoking the arguments by Mueller and Plug (2006), Heckman et al. (2006) and Borghans et al. (2008) that non-cognitive skills, like cognitive skills, may exert direct and indirect effect on productivity (Cubel et al., 2016), which may result in job performance differentials (Heineck, 2011). Robust associations between some of the five-factor personality traits and job performance have been documented by Barrick and Mount (1991) and Salgado and Rumbo (1997). As a direct channel to job performance, personality and its effect through behaviour may be considered as an individual’s set of productive/unproductive skills and rewarded or punished accordingly. However, as an indirect channel, personality may affect job performance through its effect on educational attainment (Borghans et al., 2008; Heckman et al., 2006) and occupational choice and attainment (Cobb-Clark & Tan, 2011).

While each of the personality traits has been conjectured to be related to job performance and for that matter earnings, Judge et al. (1999) note that there appears to be an agreement in organizational psychology literature that neuroticism, extraversion and conscientiousness are the most relevant personality traits to job performance. From the perspective of activation theory (Gardner & Cummings, 1988), neurotic or emotionally unstable individuals (characterized by being anxious, tense, irritable, impulsive, moody or depressed) may respond to external stimulus either too much or too little, leading them to probably perform their task or job poorly and/or be less suited for high-level, demanding or complex and stressful jobs (Heineck, 2011; Spector et al., 1995). As a consequence, neuroticism may contribute negatively to earnings.

In a business environment, social capital is important, and extraverts are more likely to have the social skills to establish social networks that are needed to succeed. As Judge et al. (1999) note ‘extraverts tend to be socially oriented (outgoing and gregarious), but also are surgent (dominant and ambitious) and active (adventuresome and assertive)’ (p.624). These qualities may more likely be rewarded than punished in the business environment as such workers tend to be likeable by colleagues and may even take on leadership roles (Heineck, 2011).

As labour-related skill, conscientiousness has been reported as the most consistent construct of the Big-Five Factor model that is related to performance across jobs (Judge et al., 1999). According to Judge et al. (1999, p. 624), ‘conscientiousness is related to an individual’s degree of self-control as well as need for achievement, order and persistence’. It is ‘manifested in three related facets—achievement oriented (hardworking and persistent), dependability (responsible and careful) and orderliness (planful and organized) (Judge et al., 1999, p. 624). Apart from these facets, studies demonstrating the importance of conscientiousness at work have linked the construct to effective job-seeking behaviour (Wanberg et al., 1996), retention (Li et al., 2014) and attendance at work (Judge et al., 1997).

Although the other two facets of the five-factor model (i.e., openness to experience and agreeableness) have not been found to be consistently related to job performance across occupations in distinctive ways, it certainly seems possible that they could be related to job performance and earnings in either ways for some occupations. Judge et al. (1999) note that openness can have both positive and negative effect on job performance and earning. On the one hand, they argue that ‘the flexibility, creativity, and intellectual orientation of open individuals may be instrumental to success in many occupations’ (p. 625). On the other hand, they argue that ‘open individuals may be prone to job hopping or may be unhappy in conventional occupations’ (p. 625), which will be detrimental to labour market success. According to Heineck (2011), openness is associated with autonomy and non-conformity. It indicates an individual’s tendency to make independent decisions and degree of initiative and control, which may promote or hinder labour market success.

Similar to openness, the effect of agreeableness on earning could be positive or negative. As an incentive-enhancing characteristic, agreeable people are more likely to respond positively to employer’s incentives and be associated with higher earnings (Bowles et al., 2001). Judge et al. (1999) also argue that ‘the cooperative nature of agreeable individuals may lead to more successful careers, particularly in occupations where teamwork or customer service is relevant’ (p. 625) and therefore contribute positively to earnings. On the contrary, it is also conceivable to argue that high levels of agreeableness may be detrimental to labour market success. Judge et al. (1999) reckon this point by arguing that ‘extremely agreeable individuals may sacrifice their success in pleasing others’ (p. 625). According to Nyhus and Pons (2005), ‘there is also a chance that agreeable people are less likely to claim higher wages’ (p. 368).

Drawing from the above-mentioned literature, it is hypothesized that non-cognitive skills, as measured by the Big-Five personality traits, will be related to earnings. specifically, it is expected that stability, extraversion and conscientiousness will be positively related to earnings, while openness and agreeableness could have either positive or negative association with earnings.

Empirical Literature on Personality and Earnings

Despite the theoretical developments on the link between non-cognitive skills and labour market outcomes, empirical studies on the effects of personality, especially the Big-Five factors, on earnings are relatively few and evolving. Unavailability of data has been ascribed as the main challenge. However, in the past two to three decades, some surveys have begun the collection of individual information on non-cognitive skills paving way for the few empirical studies to be conducted. For example, using the five-factor model, based on the Dutch National Bank Household Survey (DNBHS), Nyhus and Pons (2005) estimated a standard Mincerian wage equation to analyse which personality traits are rewarded in the labour market. The wage equations were estimated separately for men and women to account for gender differences in traditional wage settings and personality analysis. The study found significant positive association between emotional stability and earnings for men and women. However, agreeableness was negatively associated with women’s wages. As tenure increases, men are rewarded for autonomy, but conscientiousness is rewarded when tenure is low. Furthermore, the study found that reward to the personality traits is also dependent on the level of education.

Using the five factors provided by the Wisconsin Longitudinal Study, Mueller and Plug (2006) arrived at the conclusion that the effects of personality traits on earnings are comparable in magnitude to those commonly found for cognitive skills. The results indicated that all five factors have statistically significant positive or negative effects on earnings, suggesting that personality is an important determinant of earnings for both men and women. Among men, being antagonistic (the obverse of agreeable), emotionally stable (the obverse of neurotic) and open to experience were substantially associated with earnings advantages. However, among women, being conscientious and open to experience were associated with earnings advantages. The earnings advantages of openness were very similar across men and women, suggesting that this trait was equally important for both genders. Of the five traits, significant gender difference was observed in terms of agreeableness. Men were considerably more antagonistic (non-agreeable) than women and received higher earnings for this trait.

Heckman et al. (2006) and Fletcher (2013) have also linked job performance and wages to the Big-Five factors, using survey data. Results of their analyses show that conscientiousness, antagonism (inverse of agreeableness) and emotional stability (the inverse of neuroticism) are positively associated with earnings. Analysing data for Japan and the USA, Lee and Ohtake (2012) found that the effect of personality traits on labour market outcomes differs between genders but not between the two countries analysed. Males with high degree of conscientiousness appear to earn more, while emotional stability acts as an important factor in determining female wages. Earnings of females are also predicted by extraversion.

Aside the five-factor model, some studies have focused on wage determination, using different personality dimensions. One such study is Osborne (2000) who used the US National Longitudinal Survey of Young Women (NLSYW) and the UK National Child Development Survey (NCDS). Locus of control was the only personality variable considered from the NLSYW data, while two orthogonal personality variables (aggression and withdrawal) were extracted from the NCDS data. The findings indicate that differences in earnings are accounted for by personality, which is also related to sex and position in the occupational hierarchy. In a similar study, Coleman and DeLeire (2003) used a National Educational Longitudinal Study (NELS) data to examine the impact of locus of control on human capital investment decisions. The results indicate that locus of control influences education decisions and operates through teenagers’ expected earnings.

Using data from STEP Skills Measurement Survey, Tognatta et al. (2016) examined the contributions of cognitive and non-cognitive skills to gender wage gap in seven low- and middle-income countries. The study found that both cognitive and non-cognitive skills affect men’s and women’s earnings differently, and these effects varied across wage distribution and countries. Employing the STEP data on Bangladesh’s formal sector, Nomura and Adhikari’s (2017) study concluded that non-cognitive skills appear to be correlated with a worker’s likelihood of achieving success in the labour market. Their analysis revealed that although many of the findings were consistent with the literature, there appears to be a unique positive association between emotional stability and wages in Bangladesh, especially in the manufacturing sector. The study also found correlations between personality traits and wages to be more prominent among large firms than among small- or medium-sized firms.

To summarize, even though the literature on the effect of personality on earnings is evolving, the empirical evidence so far tilts towards developed countries, with emerging conclusion that the relationship between personality and wages differs by gender and not universal. A noticeable gap in the literature is the missing evidence from a developing country perspective, particularly Africa. As such, it is imperative to ask whether non-cognitive skills, particularly, personality traits, contribute to earnings in the African context by focusing on Ghana. In fact, given that labour market earnings are the main source of income for the vast majority of people (Brunello & Schlotter, 2011; World Bank, 2012), it is important to understand which personality traits earn higher or lower returns in the African labour market so that the results could be used to inform educational policies at the early childhood level.

Methodology

Based on the original Mincer’s (1974) model of human capital earnings function, and consistent with Nyhus and Pons (2005), the empirical model adopted in this study is based on the study by Bowles et al. (2001). The model is a semi-logarithmic wage equation specified as follows:

\ln W = X β + μ

(1)

where, lnW is the logarithm of weekly earnings, X is a row vector of productivity characteristics, β is a column vector of coefficients and μ is the error term. The vector of explanatory variables or observed worker characteristics, X, include the Big-Five personality traits and other control variables such as gender, type of education, tenure, potential experience, type of occupation,¹ dummy for underemployment and regional/geographical dummies.² The empirical model estimated in this study is thus specified as:

\begin{matrix} \ln W = β_{0} + β_{1} M a l e_{i} + β_{2} V o c a t i o n a l E d u c_{i} + β_{3} G e n e r a l E d u c_{i} + β_{4} V o c a t i o n a l a n d G e n e r a l_{i} \\ \begin{array}{l} + β_{5} T e n u r e_{i} + β_{6} T e n u r e_{i}^{2} + β_{7} P o t e n t i a l \exp_{i} + β_{8} P o t e n t i a l \exp_{i}^{2} + β_{9} H i g h l y S k i l l e d_{i} + \\ \begin{matrix} β_{10} L o w S k i l l e d_{i} + β_{11} C r a f t a n d T r a d e_{i} + β_{12} U n d e r e m p l o y m e n t_{i} + β_{13} E x t r a v e r s i o n_{i} + \\ β_{14} C o n s c i e n t i o u s n e s s_{i} + β_{15} O p e n n e s s_{i} + β_{16} S t a b i l i t y_{i} + β_{17} A g r e e a b l e n e s s_{i} + \\ \sum_{i = 18}^{26} β_{i} Re g i o n a l d u m m i e s_{i} + μ_{i} \end{matrix} \end{array} \end{matrix}

(2)

where lnW is the natural logarithm of weekly earnings with all covariates, both cognitive and non-cognitive, identified in literature to influence earnings. Tenure and potential experience variables are included as proxies to capture two aspects of human capital development related to on-the-job training. That is, firm-specific on-the-job training is captured by tenure, while general on-the-job training is captured by potential experience variable. The squares of these two variables are also included in the model to capture the quadratic effects of tenure and experience and approximate the well-known concave profile in earnings studies (Nyhus & Pons, 2005).

Data

The analysis in this study uses data from the World Bank’s surveys on STEP on Ghana, which is designed to better understand the interplay between skills, on one hand, and employability and productivity, on the other (Pierre et al., 2014). The STEP surveys are conducted in low- and middle-income countries which include Ghana. The data from these surveys have been widely used in empirical studies in other countries and acclaimed to be of high quality and comprehensive (Nomura & Adhikari, 2017; Tognatta et al., 2016; Valerio et al., 2016). Pierre et al. (2014) provide a comprehensive description on the design of the survey instruments and the constructs measured, as well as the technical standards and implementation protocols adopted to ensure data quality and comparability across countries. An important aspect of the STEP surveys is that they target the urban working-age population (15–64 years).

Table 2.

Variables Definitions and Measurement

Variable	Definition	Measurement
Dependent variable
Wage earnings and/or self-employment earnings	Weekly earnings from wage employment and/or weekly earnings from self-employment	Natural logarithm of the weekly earnings from wage employment and/or earnings from self-employment
Independent and control variables
Gender	Sex of the respondent	Value is 1 if male, 0 otherwise
Education	Type of education	1 if up to basic, 0 otherwise
		1 if vocation, 0 otherwise
		1 if general, 0 otherwise
		1 if vocational and general, 0 otherwise
Tenure	The number of years that the worker has spent on the current job	Number of years since worker started his/her current job
Potential experience	The number of years the worker has been in the labour market	Number of years since worker started his/her first paid job
Occupation type	Level of skill associated with job	1 if elementary occupation, 0 otherwise
		1 if highly skilled white-collar manager/professional/technician, 0 otherwise
		1 if low skilled, 0 otherwise
		1 if craft and related trades workers, plant and machine operator and assemblers, 0 otherwise
Underemployment	Measured by the number of hours the individual works in a week (less than 40 h per week)	1 if underemployed, 0 otherwise
Extraversion	Respondent’s self-evaluation regarding talkative; prefer to keep quiet or keep opinion; and outgoing, sociable and makes friends	Average score to the questions after reverse scoring (scale is from 1 ‘almost never’ to 4 ‘almost always’)
Conscientiousness	Respondent’s self-evaluation regarding careful when doing task; prefers relaxation more than hard work; and works well and quickly.	Average score to the questions after reverse scoring (scale is from 1 ‘almost never’ to 4 ‘almost always’)
Openness	Respondent’s self-evaluation regarding—comes up with ideas other people haven’t thought of before; interested in learning new things; enjoy beautiful things, like nature, art and music.	Average score to the questions after reverse scoring (scale is from 1 ‘almost never’ to 4 ‘almost always’)
Stability	Respondent’s self-evaluation regarding—relaxed during stressful situations; tends to worry; gets nervous easily	Average score to the questions after reverse scoring (scale is from 1 ‘almost never’ to 4 ‘almost always’)
Agreeableness	Respondent self-evaluation regarding—forgives other people easily; very polite to other people; prefer to keep quiet or keep opinion; and outgoing, sociable and generous to other people with your time or money	Average score to the questions after reverse scoring (scale is from 1 ‘almost never’ to 4 ‘almost always’)
Region	Regional dummies for all 10 administrative regions in Ghana at the time of survey	For all 10 regions, 1 if region is the dummy region, 0 otherwise

Source: International Labour Organization (2012).

Notes: Underemployment is measured by the number hours the individual works in a week. Using the International Labour Organization’s (ILO’s) definition, if an individual works for less than 40 hours a week, then he or she is considered to be underemployed. Elementary occupations consist of simple and routine tasks, which mainly require the use of hand-held tools or physical strength and/or endurance. They involve tasks such as selling goods in streets and public places, cleaning, kitchen assistants, digging, garden labourers, lifting and carrying materials or luggage, sorting or assembling goods by hand, operating non-motorized vehicles, picking fruits and vegetables, property watching, etc. Craft and related trades workers, plant and machine operator and assemblers involve the performance of tasks such as operating machinery and electronic equipment, driving vehicles, and maintenance of and repair of electrical and mechanical equipment. They involve sewing machinists, dressmakers, hairdressers, butchers, bus drivers, building electricians, motor vehicle mechanics, etc. Low skilled white-collar occupations involve the performance of tasks that require relatively advanced literacy and numeracy skills and good interpersonal communication skills. They include secretaries, accounts clerks, police and prison officers, services and sales workers, skilled agricultural, forestry and fishery workers, etc. Highly skilled white-collar manager/professional/technician occupations involve the performance of tasks (practical, technical and professional) that require extensive body of factual, technical and procedural knowledge in a specialized field or complex problem-solving, decision-making and creativity base on an extensive body of theoretical and factual knowledge in a specialized field. They involve shop managers, marketing managers, medical laboratory technicians, legal secretaries, computer support technicians, commercial sales representatives, broadcasting and recording technicians, civil engineers, secondary or tertiary school teachers, medical practitioners, lawyers, accountants, computing systems analysts, etc.

The Ghana sample design uses a four-stage sampling procedure in which, at the first stage, primary sample units (PSUs) uniquely defined by region and enumeration area are selected to ensure implicit stratification by region. The second-stage sampling partitions large PSUs into smaller areas after which one partition is randomly selected for full listing. In the third stage, the sample unit is household, and the sampling method is systematic and random. In the final stage, the sample unit was individuals aged 15–64 (inclusive), drawn from each selected household, using simple random method.³ At the end of the sampling process, the overall response rate was 83.2% of the targeted 3,000 individuals in the urban areas of Ghana.

The surveys collect data on a wide range of variables. However, for the purpose of this study, the selected variables are listed in Table 2. The main variables of interest are the non-cognitive skills measured by the Big-Five personality traits (extraversion, conscientiousness, openness, stability and agreeableness). These variables were measured using multiple items rated by the respondents on a 4-point Likert scale from 1 to 4, where 1 represents almost never, 2 represents some of the time, 3 represents most of the time and 4 represents almost always. For each trait, a sample question is given as follows: are you talkative? (extraversion); do you prefer relaxation more than hard work? (conscientiousness); do you come up with ideas other people have not thought of before? (openness); do you get nervous easily? (stability); and are you very polite to other people? (agreeableness). The responses to each set of questions relating to a particular personality trait were averaged into a composite index after reverse scoring of some of the items. The definition and measurement of all variables used in the analysis are provided in Table 2.

Results

Table 3 provides the descriptive statistics of the variables that are used in the regression analysis. These results show that our sample was almost equally split between those in wage employment (50.2%) and those in self-employment (49.8%) (see employment status in Table 3). However, there are more males (56.4%) in the sample than females (43.6%). In terms of earnings, wage-employed individuals on the average earn more (mean = 111.44; standard deviation = 118.47) than the self-employed (mean = 94.08; standard deviation = 112.49). Gender wise, the differences in earnings between wage- and self-employed persist; however, the gap is narrower for males (i.e., wage-employed males vs. self-employed males is 113.41 with standard deviation of 100.94 vs. 110.03 with standard deviation of 105.15) than it is for females (i.e., wage-employed females vs. self-employed females is 107.19 with standard deviation of 103.73 vs. 81.35 with std dev of 116.65). This comparison also shows that average weekly earnings are higher for males than females. While the full sample is almost equally split between wage-employment (50.2%) and self-employment (49.8%), there are more males (60.8%) than females (339.2%) in wage-employment and more females (63.6%) than males (36.4%) in self-employment.

In terms of education, more than half of the respondents went beyond the basic level. Close to 17% of the respondents had vocational education, while 33.4% had general education. On average, the respondents had been in the labour marker for almost 19 years, out of which close to 7 years had been spent on the current job. Majority of the respondents (39.1%) were engaged in low-skilled occupation, while high-skilled workers constituted 27.9%. Less than 7% of the respondents were found in elementary occupation. It is also worth noting that the distribution of males and females across the groups of occupations suggests that there are more males in highly skilled occupations than females. Conversely, the distribution shows that a larger proportion of females (60.7%) than males (22.3%) report working in low-skilled occupations. Further, the distribution of male–female with regard to underemployment indicate that a higher proportion of females are underemployed as compared to males.

Table 3.

Summary Statistics

Variable		Sample (std. dev.)	Male (std. dev.)	Female (std. dev.)
Gender
	Male	0.564 (0.496)
	Female	0.436 (0.496)
Type of education
	Up to basic	0.458 (0.499)	0.419 (0.494)	0.509 (0.500)
	Vocational	0.172 (0.378)	0.188 (0.391)	0.152 (0.359)
	General	0.334 (0.472)	0.354 (0.479)	0.308 (0.462)
	Vocational and general	0.036 (0.185)	0.039 (0.195)	0.030 (0.172)
Tenure		7.397 (7.829)	7.658 (7.845)	7.059 (7.805)
Potential experience		18.992 (10.898)	19.525 (11.464)	18.301 (10.090)
Occupation type
	Elementary occupation	0.066 (0.249)	0.101 (0.302)	0.021 (0.144)
	Highly skilled	0.279 (0.448)	0.332 (0.471)	0.210 (0.408)
	Low skilled	0.391 (0.488)	0.223 (0.417)	0.607 (0.489)
	Craft and trade workers	0.264 (0.441)	0.343 (0.475)	0.161 (0.368)
Underemployment
	Not underemployed	0.879 (0.327)	0.897 (0.304)	0.855 (0.352)
	Underemployed	0.121 (0.327)	0.103 (0.304)	0.145 (0.352)
Personality traits
	Extraversion	2.517 (0.602)	2.521 (0.593)	2.511 (0.614)
	Conscientiousness	3.260 (0.547)	3.333 (0.531)	3.165 (0.552)
	Openness	3.053 (0.586)	3.101 (0.591)	2.990 (0.575)
	Stability	2.737 (0.561)	2.819 (0.524)	2.630 (0.589)
	Agreeableness	3.021 (0.643)	3.034 (0.644)	3.002 (0.639)
Outcomes variables
	Total earnings (per week)	102.794 (107.535)	112.09 (102.53)	90.768 (112.677)
	Earnings (wage-employment)	111.439 (118.465)	113.409 (100.9)	107.186 (103.727)
	Earnings (self-employment)	94.078 (112.486)	110.029 (105.2)	81.352 (116.648)
Employment status
	Wage-employment	0.502 (0.500)	0.608 (0.489)	0.392 (0.489)
	Self-employment	0.498 (0.500)	0.364 (0.482)	0.636 (0.482)
Region
	Western	0.076 (0.266)	0.072 (0.259)	0.082 (0.274)
	Central	0.071 (0.257)	0.072 (0.259)	0.070 (0.256)
	Greater Accra	0.360 (0.480)	0.345 (0.476)	0.381 (0.486)
	Volta	0.068 (0.252)	0.056 (0.230)	0.084 (0.278)
	Eastern	0.081 (0.274)	0.074 (0.262)	0.091 (0.288)
	Ashanti	0.209 (0.407)	0.218 (0.413)	0.196 (0.398)
	Brong Ahafo	0.066 (0.249)	0.072 (0.259)	0.058 (0.235)
	Northern	0.047 (0.211)	0.067 (0.249)	0.021 (0.144)
	Upper East	0.015 (0.123)	0.016 (0.127)	0.014 (0.118)
	Upper West	0.005 (0.071)	0.007 (0.085)	0.002 (0.048)
Sample		982	554	428

Source: Authors’ computations from STEP Data.

In the full sample, the personality traits with the two highest scores were conscientiousness and openness, and this was also the case for the male subsample, but for the female subsample, they were conscientiousness and agreeableness, suggesting that more females tend to be agreeable.

To examine the correlations among the variables, Table 4 presents the correlation matrix, which basically shows low correlation coefficients among the variables; hence, multicollinearity should not be a serious issue. Formal tests for multicollinearity, which is not reported due to space limitation, indicated that there was no multicollinearity problem as the mean of the variance inflation factor (VIF) was less than 10 and the tolerance level was more than 0.1.

Tables 5 –7 report the Ordinary Least Squares (OLS) estimates of the relationship between measures of the Big-Five personality traits and the log of weekly earnings for both wage and self-employment combined (Table 5), and separately for each category of employment—wage employment (Table 6) and self-employment (Table 7). The association between weekly earnings and other variables such as education, tenure and potential experience are also reported in Tables 5 –7. Each table contains three OLS estimates for the full sample and separately for male and female subsamples. The estimated coefficients are heteroskedasticity consistent. In all the three tables, the coefficient of gender is statistically significant and positive, meaning that, holding all other factors constant, males earn more than females. This gender-based earning differentials may be attributed to gender wage discrimination as reported in some studies (e.g., Blau & Kahn, 2003; Pham & Reilly, 2007; Terrell, 1992) or ascribed to the high (over 60%) low-skilled occupations reported by females in this study.

Regarding the returns to cognitive skills measured by the human capital variables, we observe that the broad types of education (vocational, general, and vocational and general) contribute positively to earnings (see Table 5). Considering the coefficients of these variables, it can be noted that relative to respondents with up-to-basic education, respondents with both vocational and general education are the highest rewarded followed by those with only vocational education and, finally, general education. However, when viewed by employment type (wage or self-employment), general education is more rewarding in wage-employment than vocational education (see Table 6). By contrast, the self-employment labour market rewards vocational education but not general although general education combined with vocational education rewards better (see Table 7). In all the results, potential experience appears to consistently have a positive effect on earnings, but tenure does not, except for females in the wage- and self-employment data combined (Table 5). Nevertheless, potential experience square has a negative association with earnings, confirming the well-known concave profile in earnings studies (Nyhus & Pons, 2005).

Turning our attention to the non-cognitive variables, the results from the full sample in Table 5 show that extraversion (which is associated with social skills, sociable, assertiveness, outgoing, likeable, etc.), conscientiousness (which is associated with competence, orderliness, dutifulness, self-discipline achievement-oriented, etc.), openness (which is characterized by intellect, ideas, curiosity, imaginative, etc.) and stability (which is characterized by the opposites of being anxious, tense, irritable, impulsive, moody or depressed, etc.) are the personality traits that have significant relationships with earnings. All four variables contribute positively to weekly earnings, and hence, they can be considered to have earnings advantages. Of the four significant personality variables, openness seems to be the most rewarding followed by conscientiousness, stability and extraversion.

In the male and female subsamples in Table 5, where the data are disaggregated according to gender, the results reveal which gender benefits from what personality traits. Generally, we notice that the results are largely consistent with what we observe in the full sample in terms of signs and significance of variables in the model, although there are slight variations. First, conscientiousness is only significant in the male sample and the most reward trait but not significant in the female sample. Second, in the female sample, the most rewarded trait is openness.

Table 4.

Kendall Correlation Matrix

	1	2	3	4	5	6	7	8	9	10	11	12	13
1	1
2	−0.174*	1
3	0.219*	−0.077	1
4	0.087*	−0.053	−0.0679	1
5	−0.019	−0.025	−0.273*	0.386*	1
6	−0.113*	0.037	−0.294*	0.035	0.120*	1
7	−0.039	−0.067	0.091	−0.012	−0.08	−0.039	1
8	−0.078	0.064	−0.048	−0.022	−0.007	−0.007	−0.0382	1
9	0.0651	0.001	0.133*	−0.018	−0.076	−0.083	−0.007	−0.0417	1
10	0.098*	−0.135*	0.167*	0.019	−0.041	−0.090*	0.0381	−0.0541	0.055	1
11	0.107*	−0.087	0.164*	−0.018	−0.112*	−0.098*	−0.0242	−0.0438	0.0917*	0.2611*	1
12	0.042	−0.132*	0.046	0.058	0.067	−0.062	0.0245	−0.027	−0.0424	0.0808	0.029	1
13	0.068	−0.027	0.143*	0.039	−0.024	−0.130*	−0.0068	0.0007	0.1122*	0.2351*	0.2780*	0.0345	1

Source: Authors’ computations from STEP Data.

Note: 1 = Earnings; 2 = gender; 3 = type of education; 4 = tenure; 5 = potential experience; 6 = occupation type; 7 = region; 8 = underemployment; 9 = extraversion; 10 = conscientiousness; 11 = openness; 12 = stability; and 13 = agreeableness.

*Significant at the 5% level.

Table 5.

Regression Output for Both Wage- and Self-employment

Variable	Full Sample	Male	Female
Gender (1 if male, 0 otherwise)	0.3232***
Gender (1 if male, 0 otherwise)	(0.0745)
Type of education (base = up to basic)
Vocational	0.3915***	0.2522**	0.5413***
	(0.1001)	(0.1227)	(0.1688)
General	0.3530***	0.1840	0.5390***
	(0.0888)	(0.1120)	(0.1483)
Vocational and general	0.7401***	0.5489***	0.9426***
	(0.1492)	(0.1807)	(0.2744)
Tenure	0.0154	0.0001	0.0346*
	(0.0125)	(0.0174)	(0.0195)
Tenure squared	−0.0002	0.0001	−0.0008
	(0.0004)	(0.0005)	(0.0007)
Potential experience	0.0535***	0.0734***	0.0427**
	(0.0125)	(0.0164)	(0.0202)
Potential experience squared	−0.0010***	−0.0014***	−0.0007
	(0.0003)	(0.0003)	(0.0004)
Occupation type (base = elementary occupation)
Highly skilled	0.8654***	0.8335***	0.9113***
	(0.1494)	(0.1669)	(0.3518)
Low skilled	0.4815***	0.5290***	0.3845
	(0.1385)	(0.1569)	(0.3344)
Craft and trade workers	0.6003***	0.6194***	0.4576
	(0.1366)	(0.1464)	(0.3530)
Underemployment (1 if underemployed, 0 otherwise)	−0.1508	−0.1829	−0.1578
	(0.1054)	(0.1406)	(0.1633)
Personality traits
Extraversion	0.1776***	0.1499**	0.2498***
	(0.0514)	(0.0710)	(0.0779)
Conscientiousness	0.2119***	0.3383***	0.0504
	(0.0657)	(0.0804)	(0.1092)
Openness	0.2482***	0.2577***	0.2564**
	(0.0639)	(0.0786)	(0.1031)
Stability	0.1833***	0.1705***	0.2291***
	(0.0521)	(0.0643)	(0.0841)
Agreeableness	0.0454	0.0202	0.1059
	(0.0569)	(0.0725)	(0.0926)
Region (base = Greater Accra Region)
Western	0.2869*	−0.0325	0.7042***
	(0.1498)	(0.1776)	(0.2400)
Central	0.0430	0.0398	0.0937
	(0.1260)	(0.1661)	(0.2043)
Volta	−0.0942	0.0687	−0.2565
	(0.1616)	(0.1775)	(0.2737)
Eastern	−0.2431**	−0.0776	−0.3836**
	(0.1153)	(0.1757)	(0.1488)
Ashanti	0.0664	0.0346	0.1318
	(0.0872)	(0.1161)	(0.1395)
Brong Ahafo	−0.0107	0.2171	−0.2711
	(0.1476)	(0.1804)	(0.2665)
Northern	−0.4618***	−0.4872**	−0.1773
	(0.1736)	(0.2055)	(0.3587)
Upper East	−0.3324	−0.0005	−0.6341
	(0.2509)	(0.2932)	(0.4349)
Upper West	0.2415	0.2546	0.5671***
	(0.1729)	(0.1873)	(0.1647)
Sample size	982	544	428

Source: Authors’ computations from STEP Data.

Notes: Standard errors in parentheses; * p < 0.10, ** p < 0.05 and *** p < 0.01.

Table 6.

Regression Output for Wage-employment

Variable	Full Sample	Male	Female
Gender (1 if male, 0 otherwise)	0.2301**
Gender (1 if male, 0 otherwise)	(0.0993)
Type of education (base = up to basic)
Vocational	0.4046***	0.2202	0.9171***
	(0.1357)	(0.1635)	(0.2404)
General	0.5116***	0.2949**	1.1058***
	(0.1179)	(0.1422)	(0.2005)
Vocational and general	0.8066***	0.5830***	1.3943***
	(0.1770)	(0.2038)	(0.3410)
Tenure	0.0139	−0.0074	0.0190
	(0.0166)	(0.0207)	(0.0280)
Tenure squared	0.0001	0.0006	0.0002
	(0.0005)	(0.0007)	(0.0009)
Potential experience	0.0599***	0.0683***	0.0715***
	(0.0156)	(0.0200)	(0.0262)
Potential experience squared	−0.0013***	−0.0014***	−0.0017***
	(0.0003)	(0.0004)	(0.0006)
Occupation type (base = elementary occupation)
Highly skilled	0.4594***	0.6316***	−0.2046
	(0.1522)	(0.1750)	(0.3522)
Low skilled	0.1437	0.3008*	−0.4612
	(0.1417)	(0.1696)	(0.3109)
Craft and trade workers	0.2629*	0.3243**	−0.1217
	(0.1450)	(0.1515)	(0.5075)
Underemployment (1 if underemployed, 0 otherwise)	−0.0932	−0.0505	−0.1427
Underemployment (1 if underemployed, 0 otherwise)	(0.1293)	(0.1644)	(0.2366)
Personality traits
Extraversion	0.2104***	0.2224***	0.1677
	(0.0627)	(0.0772)	(0.1182)
Conscientiousness	0.2511***	0.2619***	0.2289
	(0.0756)	(0.0949)	(0.1408)
Openness	0.2931***	0.3560***	0.1957
	(0.0723)	(0.0865)	(0.1280)
Stability	0.1928***	0.2575***	0.1477
	(0.0672)	(0.0796)	(0.1194)
Agreeableness	0.0435	−0.0403	0.2749**
	(0.0718)	(0.0901)	(0.1260)
Region (base = Greater Accra Region)
Western	−0.0562	−0.0325	−0.0074
	(0.1536)	(0.1930)	(0.2252)
Central	−0.1229	−0.1164	−0.1410
	(0.1415)	(0.1865)	(0.2213)
Volta	−0.1937	−0.0675	−0.4391
	(0.2059)	(0.2337)	(0.3740)
Eastern	−0.0618	−0.0595	−0.1671
	(0.1477)	(0.2133)	(0.2218)
Ashanti	0.0783	0.1107	0.0536
	(0.1171)	(0.1469)	(0.2084)
Brong Ahafo	−0.0193	0.0253	−0.1484
	(0.1847)	(0.2221)	(0.4092)
Northern	−0.5901**	−0.7581***	0.3435
	(0.2338)	(0.2598)	(0.2603)
Upper East	−0.2778	−0.2273	−0.1901
	(0.2741)	(0.2894)	(0.7436)
Upper West	0.0815	0.0940	0.1689
	(0.2058)	(0.2685)	(0.1776)
Sample size	493	377	156

Source: Authors’ computations from STEP Data.

Notes: Standard errors in parentheses; * p < 0.10, ** p < 0.05 and *** p < 0.01.

To determine if the personality traits are rewarded or punished differently in the wage employment market or the self-employment market, we turn our attention to the results in Tables 6 and 7. The results in Table 6 relate to the wage employment market. These results are largely similar to the ones discussed in Table 5. For the full sample, the significant personality trait variables are still extraversion, conscientiousness, openness and stability with the most rewarding trait still being openness. However, notable differences lie in the male and female subsamples: while the traits of extraversion, conscientiousness, openness and stability maintained their positive association with earnings in the male sample, only agreeableness had a significant positive effect on female earnings in the wage-employment. This finding may be interpreted to mean that in the Ghanaian cultural setting, females may be expected to be agreeable as a sign of submissiveness and hence rewarded, while the same attribute in males may be considered to be a weakness or sign of not having independent mind and hence not rewarded.

Contrasting the results in Table 7 to those in Table 6, the only consistent personality trait variable that is related to earnings in the full sample is stability. This variable has significant positive associations with earnings in the female sample as well. Similarly, extraversion and openness are desirable attributes for females in self-employment and are therefore rewarded. However, the only trait that is rewarded in self-employment for males is conscientiousness, which is associated with not being lazy, competence, orderliness, dutifulness, self-discipline, achievement-oriented, etc.

Table 7.

Regression Output for Self-employment

Variable	Full Sample	Male	Female
Gender (1 if male, 0 otherwise)	0.5604***
Gender (1 if male, 0 otherwise)	(0.1180)
Type of education (base = up to basic)
Vocational	0.5135***	0.4165**	0.5953**
	(0.1520)	(0.2003)	(0.2388)
General	0.1302	−0.0242	0.3503
	(0.1473)	(0.1959)	(0.2300)
Vocational and general	0.8322***	0.6941**	0.8349*
	(0.3120)	(0.3170)	(0.4877)
Tenure	0.0104	−0.0200	0.0242
	(0.0170)	(0.0291)	(0.0268)
Tenure squared	−0.0002	0.0003	−0.0005
	(0.0005)	(0.0008)	(0.0010)
Potential experience	0.0660***	0.1023***	0.0643*
	(0.0225)	(0.0334)	(0.0332)
Potential experience squared	−0.0010**	−0.0016***	−0.0009
	(0.0004)	(0.0006)	(0.0006)
Occupation type (base = elementary occupation)
Highly skilled	1.2078***	1.0297***	0.1107
	(0.3299)	(0.3382)	(0.1469)
Low skilled	1.0875***	1.0080***	0.1923
	(0.2610)	(0.2730)	(0.5247)
Craft and trade workers	1.1302***	1.1037***	0.2245
	(0.2576)	(0.2692)	(0.5353)
Underemployment (1 if underemployed, 0 otherwise)	−0.1764	−0.4040*	−0.1259
	(0.1520)	(0.2375)	(0.2053)
Personality traits
Extraversion	0.1133	0.0167	0.2876***
	(0.0812)	(0.1356)	(0.1056)
Conscientiousness	0.1240	0.4112***	−0.0358
	(0.1010)	(0.1412)	(0.1411)
Openness	0.1521	0.0797	0.2931**
	(0.0976)	(0.1437)	(0.1343)
Stability	0.1447*	0.0460	0.2682**
	(0.0783)	(0.1204)	(0.1096)
Agreeableness	0.0411	0.1023	0.0749
	(0.0897)	(0.1383)	(0.1233)
Region (base = Greater Accra Region)
Western	0.6059**	−0.0613	0.9445***
	(0.2469)	(0.3629)	(0.3007)
Central	0.1766	0.2238	0.2794
	(0.2168)	(0.3314)	(0.3392)
Volta	−0.0684	0.1325	−0.2364
	(0.2414)	(0.2489)	(0.3689)
Eastern	−0.4684***	−0.1097	−0.6200***
	(0.1649)	(0.3587)	(0.1921)
Ashanti	0.0289	−0.1489	0.2157
	(0.1329)	(0.1988)	(0.1903)
Brong Ahafo	0.0581	0.6093*	−0.3458
	(0.2202)	(0.3339)	(0.3196)
Northern	−0.2614	−0.0099	−0.8327
	(0.2600)	(0.3424)	(0.5297)
Upper East	−0.3274	0.6976	−0.7682
	(0.4381)	(0.5394)	(0.4924)
Upper West	0.3915**	0.2025	0.1020
	(0.1706)	(0.2622)	(0.1384)
Sample size	489	217	272

Source: Authors’ computations from STEP Data.

Notes: Standard errors in parentheses; * p < 0.10, ** p < 0.05 and *** p < 0.01.

Discussion

The findings of this study indicate that personality as measured by the Big-Five factors (extraversion, conscientiousness, openness, stability and agreeableness) contribute positively to earnings of males and/or females in the wage-employment or self-employment markets. This is evidenced by the fact that all the five traits are found to be associated with either male or female earnings, or both. Thus, in line with the theoretical arguments advanced by Goldsmith et al. (1997) and Bowles et al. (2001) and empirical evidence offered by other researchers (Heineck, 2011; Lee & Ohtake, 2012; Nyhus & Pons, 2005), the findings support our hypothesized relationships and suggest that non-cognitive skills or psychological capital can explain wage variations and therefore should be included in studies on earnings. While each of the five personality traits are conjectured in the literature review section of this study to be related to earnings, only three of the traits (i.e., extraversion, openness and stability) were found to be consistent for both genders (male and female) when data from the two markets—wage and self-employment—were pulled together. However, when these two labour markets are disaggregated, there was no common trait to males and females that was rewarded or punished in the labour markets. For the females, it was agreeable that was rewarded in the wage-employment market, while, for the males, it was conscientiousness. This shows significant gender differences in the reward to personality traits in both the formal and informal sectors of the Ghanaian labour market.

The findings also underscore the relative importance of openness, as the association between this trait—which is related to intellect, ideas, curiosity, imaginative, etc.—and earnings is not only positive and significant but also the highest in most cases. The implication of this is that individuals with high levels of openness, regardless of their gender and holding other factors constant, enjoy higher wages in the labour market as a whole. As Judge et al. (1999) note, openness can have either positive or negative effect on job performance and earnings. For it to have a positive effect, it means that ‘the flexibility, creativity, and intellectual orientation of open individuals may be instrumental to success in many occupations’ (Judge et al., 1999, p. 625). Heineck (2011) has also observed that because openness is associated with autonomy and non-conformity, it indicates that individuals who are open have the tendency to make independent decisions and take initiatives, which may end up translating into earning advantages.

Different from openness, agreeableness turns out to be positively associated with earnings only in the female sample for wage-employment. As indicated earlier, the rationalization for this finding could be that the Ghanaian society, which has traditionally relegated girl-child education to the back-burner and characterized by power-distant culture, recognizes agreeable females as submissive and therefore rewarded in the wage-employment market which is dominated by males. This finding is somewhat consistent with the literature, which posits that agreeable people are more likely to get along and respond positively to employer’s incentives and be rewarded with higher earnings (Bowles et al., 2001). Judge et al. (1999) also note that ‘the cooperative nature of agreeable individuals may lead to more successful careers, particularly in occupations where teamwork or customer service is relevant’ (p. 625) and therefore contribute positively to earnings. Supporting the arguments of Bowles et al. (2001) and Judge et al. (1999), this finding, thus, contradicts the argument of Nyhus and Pons (2005), who stated that ‘there is also a chance that agreeable people are less likely to claim higher wages’ (p. 368).

The findings on conscientiousness indicate a positive association with earnings when male and female data are combined without recourse to the market in which they are working. However, when the data is disaggregated, it is only males who are rewarded for conscientiousness. As a labour-related skill, conscientiousness has been reported as the most consistent construct of the Big-Five factor model that is related to performance across jobs (Judge et al., 1999); however the findings reported in this study do not seem to suggest that they are rewarded for females. These findings agree with Lee and Ohtake (2012) who reported that in both Japan and the USA, males with high degree of conscientiousness appear to earn more, while emotional stability and extraversion act as an important factor in determining female wages.

Implications

The results of this study have important implications for theory and practice. Theoretically, the findings provide insights and continuing support for the arguments on why non-cognitive skills or psychological capital should be included in earnings by economists. As argued by Goldsmith et al. (1997) and Bowles et al. (2001), personality traits may be rewarded or punished in the labour market because they serve as a canal through which employers can induce efforts or productivity at a lower cost. The results highlight that non-cognitive skills are associated with earning differentials. Therefore, further theoretical development is required to understand the mechanisms by which personality traits are related to earnings.

Practically, the findings hold at least three implications. The first of these implications is that the findings can be used to inform educational policies in the formative years, where the curricula can be designed to shape children’s personality along the lines of the traits that earn higher returns in the labour market. Thankfully, studies in psychology indicate that personality can be honed and shaped before biological maturation or early adulthood, after which they become relatively stable (Cobb-Clark & Schurer, 2012). The second equally important implication is that it is possible that those traits that earn higher returns in the labour market may be related to productivity as established in organizational psychology studies (e.g., Barrick & Mount 1991; Tett et al., 1991) and economics literature (Cubel et al., 2016). Hence, recruiters in Ghana may pay attention to the relevant traits when recruiting employees. Particularly, since openness has the highest positive coefficient, it may well be the case that openness should be one of the valuable non-cognitive skills that recruiters in the Ghanaian labour market must look out for. Third, for self-employment or entrepreneurial businesses, the findings imply that conscientious males are more likely to succeed, while, for females, it pays to have the traits of extraversion, openness and stability.

Conclusion

This article focused on examining the association between personality traits and earnings in wage- and self-employment labour markets with particular interest in identifying which traits are rewarding to males and females. After controlling for relevant human capital variables such as education, occupation, underemployment, tenure, potential experience and the quadratic effects of tenure and potential experience in addition to regional or geographical location, the results indicate that in both the wage- and self-employment markets, stability is the common trait that has earning advantages or is rewarded regardless of the gender of the employee. However, when the data are disaggregated by gender and employment status, agreeableness is the only rewarded trait for females in the formal wage-employment labour market, while conscientiousness is the only rewarded trait for males in the informal self-employment labour market. Extraversion and openness are significant in the full samples for both wage- and self-employment labour markets combined as well as for wage-employment alone. In all cases where openness was significant, it had the highest coefficient, implying that the characteristics associated with this trait like intellect and ideas are the most valuable in the Ghanaian labour market. Based on these findings, the study concludes that consistent with existing literature, non-cognitive characteristics of individuals like personality can help explain differences in earnings after accounting for variation in human capital variables. Therefore, it is recommended that future studies on earnings should consider the influence of non-cognitive skills, in general, and, particularly, personality, which is seen to be relatively more enduring characteristics of adults. For educational policy formulation or soft skill training, the findings can be used since they highlight which personality traits are rewarding and for what gender.

Although the findings of this study are comparable to those obtained in previous studies, the main contribution of this study is that it is the first to conduct gender analysis of the associations between personality traits and earnings from the perspective of wage- and self-employment in an African country context. Through this, the findings have unearthed that in a power-distant culture like Ghana, where, traditionally, girl-child education has been relegated to the background, agreeable females, and not males, are rewarded in the formal wage-employment labour market where men are typically dominant. However, in the self-employment labour market, where females typically dominate, conscientious males, and not females, are positively rewarded with higher earnings.

Limitations of Study and Suggestions for Future Research

While the findings of this study add fresh evidence from an African perspective to the myriad of studies from the Western world, there are notable limitations that create opportunity for further studies. First, although the study is based on sound theoretical expositions, the findings make limited contributions to the mechanism(s) by which non-cognitive skills or personality traits affect earnings. In other words, the study could not explore the channels or mechanisms through which personality traits affect earnings. In view of this, future studies can design laboratory experiment to test whether relationship between personality traits and earnings operates through productivity or other channels. Second, the study’s focus on a single country may limit the transferability of the findings, especially when studies in other countries suggest that the results are inconclusive. Thus, to contribute to the generalizability of these findings, future studies can replicate this study in other African countries, where data exits. Third, due to the non-existence of longitudinal data that would have allowed us to conduct dynamic analysis, this study used cross-sectional data. Therefore, future studies can examine the dynamic aspects when longitudinal data become available. Finally, our focus was on examining the effect of the Big-Five personality traits on earnings. Though this taxonomy of personality is the most popular and widely applied, it is by no means the only relevant way of classifying personality. Therefore, future studies in African context can explore the associations between other classifications of personality and earnings to enrich the literature.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

References

Barrick

M. R.

, & Mount

M. K.

(1991). The Big Five personality dimensions and job performance: A meta-analysis. Personnel Psychology, 44(1), 1–26.

Becker

(1964). Human capital: A theoretical and empirical analysis, with special reference to education. National Bureau of Economic Research.

Blau

F. D.

, & Kahn

L. M.

(2003). Understanding international differences in the gender pay gap. Journal of Labor Economics, 21(1), 106–144.

Borghans

, Duckworth

A. L.

, Heckman

J. J.

, & Ter Weel

(2008). The economics and psychology of personality traits. Journal of Human Resources, 43(4), 972–1059.

Bowles

, Gintis

, & Osborne

(2001). The determinants of earnings: A behavioral approach. Journal of Economic Literature, 39(4), 1137–1176.

Brunello

, & Schlotter

(2011). Non-cognitive skills and personality traits: Labour market relevance and their development in education & training systems (IZA Discussion paper 5743). The Institute for the Study of Labor (IZA).

Cawley

, Heckman

, & Vytlacil

(2001). Three observations on wages and measured cognitive ability. Labour Economics, 8(4), 419–442.

Cobb-Clark

D. A.

, & Schurer

(2012). The stability of Big-Five personality traits. Economics Letters, 115(1), 11–15.

Cobb-Clark

D. A.

, & Tan

(2011). Noncognitive skills, occupational attainment, and relative wages. Labour Economics, 18(1), 1–13.

10.

Coleman

, & DeLeire

(2003). An economic model of locus of control and the human capital investment decision. Journal of Human Resources, 38(3), 701–721.

11.

Costa

P. T.

Jr. , & McCrae

R. R.

(1990). Personality: Another ‘hidden factor’ is stress research. Psychological Inquiry, 1(1), 22–24.

12.

Cubel

, Nuevo-Chiquero

, Sanchez-Pages

, & Vidal-Fernandez

(2016). Do personality traits affect productivity? Evidence from the laboratory. The Economic Journal, 126(592), 654–681.

13.

Digman

J. M.

(1990). Personality structure: Emergence of the five-factor model. Annual Review of Psychology, 41(1), 417–440.

14.

Fletcher

J. M.

(2013). The effects of personality traits on adult labor market outcomes: Evidence from siblings. Journal of Economic Behavior and Organization, 89, 122–135.

15.

Gardner

D. G.

, & Cummings

L. L.

(1988). Activation theory and job. Research in Organizational Behavior, 10, 81–122.

16.

Goldberg

L. R.

(1993). The structure of phenotypic personality traits. American Psychologist, 48(1), 26.

17.

Goldsmith

A. H.

, Veum

J. R.

, & Darity

Jr. (1997). The impact of psychological and human capital on wages. Economic Inquiry, 35(4), 815–829.

18.

Hartog

, Oosterbeek

, & Teulings

(1993). Age, wages and education in the Netherlands. In: Johnson

, & Zimmermann

K. F.

(Eds.), Labour markets in an ageing European (182–211). Cambridge University Press.

19.

Heckman

J. J.

, & Kautz

(2012). Hard evidence on soft skills. Labour Economics, 19(4), 451–464.

20.

Heckman

J. J.

, Stixrud

, & Urzua

(2006). The effects of cognitive and noncognitive abilities on labor market outcomes and social behavior. Journal of Labor Economics, 24(3), 411–482.

21.

Heineck

(2011). Does it pay to be nice? Personality and earnings in the United Kingdom. Industrial and Labor Relations Review, 64(5), 1020–1038.

22.

International Labour Organization (2012). International Standard Classification of Occupations 2008 (ISCO-08): Structure, group definitions and correspondence tables. International Labour Organization.

23.

John

O. P.

, & Srivastava

(1999). The Big Five trait taxonomy: History, measurement, and theoretical perspectives. In Pervin

L. A.

, & John

O. P.

(Eds.), Handbook of personality: Theory and research (Vol. 2, pp. 102–138). The Guilford Press.

24.

Judge

T. A.

, Higgins

C. A.

, Thoresen

C. J.

, & Barrick

M. R.

(1999). The Big Five personality traits, general mental ability, and career success across the life span. Personnel Psychology, 52(3), 621–652.

25.

Judge

T. A.

, Martocchio

J. J.

, & Thoresen

C. J.

(1997). Five-factor model of personality and employee absence. Journal of Applied Psychology, 82(5), 745.

26.

Koto

P. S.

(2015). An empirical analysis of the informal sector in Ghana. The Journal of Developing Areas, 49(2), 93–108.

27.

Lee

, & Ohtake

(2012). The effect of personality traits and behavioral characteristics on schooling, earnings and career promotion. Journal of Behavioral Economics and Finance, 5, 231–238.

28.

Levin

, & Plug

E. J. S.

(1999). Instrumenting education and the returns to schooling in the Netherlands. Labour Economics, 6(4), 521–534.

29.

, Barrick

M. R.

, Zimmerman

R. D.

, & Chiaburu

D. S.

(2014). Retaining the productive employee: The role of personality. The Academy of Management Annals, 8(1), 347–395.

30.

Marshall

G. N.

, Wortman

C. B.

, Vickers

R. R.

, Kusulas

J. W.

, & Hervig

L. K.

(1994). The five-factor model of personality as a framework for personality-health research. Journal of Personality and Social Psychology, 67(2), 278.

31.

Mincer

(1958). Investment in human capital and personal income distribution. Journal of Political Economy, 66(4), 281–302.

32.

Mincer

(1974). Schooling, experience, and earnings. Columbia University Press.

33.

Mueller

, & Plug

(2006). Estimating the effect of personality on male and female earnings. Industrial and Labor Relations Review, 60(1), 3–22.

34.

Murnane

R. J.

, Willett

J. B.

, & Levy

(1995). The growing importance of cognitive skills in wage determination. The Review of Economics and Statistics, 77(2), 251–266.

35.

Murnane

R. J.

, Willett

J. B.

, Braatz

M. J.

, & Duhaldeborde

(2001). Do different dimensions of male high school students’ skills predict labor market success a decade later? Evidence from the NLSY. Economics of Education Review, 20(4), 311–320.

36.

Nomura

, & Adhikari

(2017). The influence of non-cognitive skills on wages within and between firms: Evidence from Bangladesh’s formal sector (World Bank Policy Research Working Paper, 8053). World Bank.

37.

Nyhus

E. K.

, & Pons

(2005). The effects of personality on earnings. Journal of Economic Psychology, 26(3), 363–384.

38.

Nyhus

E. K.

, & Pons

(2012). Personality and the gender wage gap. Applied Economics, 44(1), 105–118.

39.

Osborne

M. A.

(2000). The power of personality: Labor market rewards and the transmission of earnings (Doctoral dissertation, ProQuest, Paper AAI9988828). University of Amherst.

40.

Osei-Boateng

, & Ampratwum

(2011). The informal sector in Ghana. Friedrich-Ebert-Stiftung, Ghana Office.

41.

Pham

T. H.

, & Reilly

(2007). The gender pay gap in Vietnam, 1993–2002: A quantile regression approach. Journal of Asian Economics, 18(5), 775–808.

42.

Pierre

, Sanchez Puerta

M. L.

, Valerio

, & Rajadel

(2014). STEP skills measurement surveys: Innovative tools for assessing skills (World Bank Social Protection & Labour Discussion Paper, No. 1421). World Bank.

43.

Roberts

B. W.

(2009). Back to the future: Personality and assessment and personality development. Journal of Research in Personality, 43(2), 137–145.

44.

Roberts

B. W.

, Walton

K. E.

, & Viechtbauer

(2006). Patterns of mean-level change in personality traits across the life course: A meta-analysis of longitudinal studies. Psychological Bulletin, 132(1), 1–25.

45.

Salgado

J. F.

, & Rumbo

(1997). Personality and job performance in financial services managers. International Journal of Selection and Assessment, 5(2), 91–100.

46.

Smith

T. W.

, & Williams

P. G.

(1992). Personality and health: Advantages and limitations of the five-factor model. Journal of Personality, 60(2), 395–425.

47.

Spector

P. E.

, Jex

S. M.

, & Chen

P. Y.

(1995). Relations of incumbent affect-related personality traits with incumbent and objective measures of characteristics of jobs. Journal of Organizational Behavior, 16(1), 59–65.

48.

Terrell

(1992). Female-male earnings differentials and occupational structure. International Labour Review, 131(4, 5), 387–404.

49.

Tett

R. P.

, Jackson

D. N.

, & Rothstein

(1991). Personality measures as predictors of job performance: A meta-analytic review. Personnel Psychology, 44(4), 703–742.

50.

Tognatta

, Valerio

, & Sanchez Puerta

M. L.

(2016). Do cognitive and noncognitive skills explain the gender wage gap in middle-income countries? An analysis using step data (World Bank Policy Research Working Paper, 7878). World Bank.

51.

Valerio

, Sanchez Puerta

M. L.

, Tognatta

, & Monroy-Taborda

(2016). Are there skills payoffs in low-and middle-income countries? Empirical evidence using step data (World Bank Policy Research Working Paper, 7879). World Bank.

52.

Wanberg

C. R.

, Watt

J. D.

, & Rumsey

D. J.

(1996). Individuals without jobs: An empirical study of job-seeking behavior and reemployment. Journal of Applied Psychology, 81(1), 76.

53.

World Bank (2012). World Development Report 2013: Jobs. World Bank. https://doi.org/10.1596/978-0-8213-9575-2