10.3  Personal characteristics and mobility

The personal characteristics that economists focus on when explaining wage mobility are overwhelmingly those that may be characterised as dimensions of human capital. Indeed, the theoretical and empirical traction that can be obtained from the notion of human capital has overpowered other lines of thinking about the sources of advantage and disadvantage. Thus there is not much literature on the link between, say, family background and wage mobility. Earnings equations routinely include human capital measures such as formal education, years of employment experience (and tenure) and sex as explanators of differences in wages, if not of wage mobility directly. In some countries, notably the US, race is also included. The outcomes of such estimations systematically find that, in addition to the human capital variables, sex and race have significant effects on wages. Specifically, women earn less than men, and minorities earn less than whites, in Western countries (with the occasional exception of Asians). In the US, the group that fares worst in the labour market, and by inference in wage mobility, is young black men with little formal education. In France, the group with the worst wage and employment prospects is young unqualified women. Ryan (2001:44) concludes that

. . . while disadvantage runs along similar lines in all countries, the distance that it travels, particularly along the tracks of ethnicity and scholastic achievement, is greater in the United States and the United Kingdom.

There is strong evidence that recent rising wage inequality is predominantly to be found within groups of workers who have the same observable characteristics (of education, sex, experience etc). This has caused economists at least to acknowledge that personal attributes of motivation, ability, personality, character and appearance are probably important in affecting wages and employment. But they have yet to delve deeply into what is still largely a black box.

There is, however, an interesting literature on the role of physical appearance in affecting wages. Again, this is not directly linked to wage mobility, but the literature concludes that more beautiful and physically attractive people have higher earnings, other things equal. Harper (2000:771), for example, using longitudinal data from the UK Household Panel Survey, finds that

. . . physical appearance has a substantial effect on earnings and employment patterns for both men and women. Irrespective of gender, those who are assessed as unattractive or short, experience a significant earnings penalty. Tall men receive a pay premium while obese women experience a pay penalty. The bulk of the pay differential for appearance arises from employer discrimination, although we find evidence for productivity differences among occupations.

Pfann, Bosman, Biddle and Hamermesh (2000) conclude that Dutch firms that have more beautiful executives are thereby more profitable and pay their executives more. It might be reasonable to infer that attractive low wage workers are more likely to be upwardly mobile than unattractive ones.

Psychologists conclude that there are strong interconnections between what happens within the individual on a psychological level and what happens in the social environment within which they grow up and develop throughout their life-course (Weiten, 1995). Both environmental and social factors are expected to be predictors of subsequent employment status. These factors include demographic background variables (such as gender, geographical location, ethnicity, type of school attended, and socio-economic-status [SES]), and family background and peer variables (such as family dysfunction, family structure, parent's educational and occupational status, and peer relationships). Personal psychological factors that are expected to affect subsequent employment status include personality variables (such as self-esteem, locus of control, vocational identity, achievement motivation, attitudes to work, and optimism); mental health and behavioural variables (such as depression, delinquency, drug use and abuse); and intellectual/cognitive variables (such as cognitive ability, IQ scores, school performance, educational attainment and job-seeking skills).

These social and psychological factors are expected to influence the level of educational attainment obtained and the amount of job seeking activities, which in turn will determine the subsequent employment status. (Kokko, Pulkkinen and Puustinen, 2000; Lynd-Stevenson, 1999; Lynn, Hampson and Magee, 1984; Winefield, Tiggerman, Winefield and Goldney 1993). This suggests a mediating effect of educational attainment and job-seeking behaviour. The suggested links are highly plausible. The main contribution that careful empirical research can make is to quantify the size of the expected effects. As in most areas of policy, the key question is whether the factor in question has a large or a small effect on outcomes. The psychology literature is much better at identifying statistically significant causal relations than it is at identifying the magnitude of the effects in question.

We could find no psychological literature that tried to explain what caused some people to end up in low wage jobs. The nearest equivalent is the literature on what causes some people to be unemployed. We only report findings from longitudinal studies because causal conclusions about the relation between psychological variables and employment status can not be drawn from lesser studies. With a few exceptions (Leana and Feldman, 1995, Lynd-Stevenson, 1999), much of this research has investigated youth during their transition between school and employment, where the aim has been to predict, on the basis of information gathered at school age, who will subsequently end up unemployed and who will succeed in finding employment (Kokko et al 2000).

Demographic factors that have been found to predict subsequent employment status include type of school attended (Sanford, Offord, McLeod and Boyle 1994; Winefield et al 1993; Woodward and Fergusson, 2000); and ethnicity (Winefield et al 1993). Minority ethnicity youth from lower SES backgrounds who attended public schools are at higher risk for subsequent unemployment.

Family factors that have been found to predict subsequent unemployment include family dysfunction; growing up in a single parent family; lower status occupation and qualifications of parents; and unemployment in the family. Peer relationship problems stemming back into childhood are also predictors of youth unemployment (Caspi, Wright, Moffitt and Silva 1998, De-Goede, Spruijt, Mass and Duindam 2000; Winefield et al 1993, Woodward and Fergusson, 2000).

Personality factors that have been found to predict subsequent employment status include lower achievement motivation and aspirations (Capsi et al 1998; Winefield et al 1993); poor conscientiousness (De Fruyt and Mervielde, 1999); and how important having a job is to the individual (“work ethic”) (Feather, 1986; Lynd-Stevenson, 1999). Other personality precursors to unemployment generally involve certain ways of thinking which characterise personalities. These predictors include hopelessness about job prospects; lower self-efficacy or sense of competence; lower level of optimism; higher level of self-blame; poorer coping skills, external locus of control; poor control of emotions; passivity and; lower levels of extraversion; and poor identity development.[3]

Mental health factors that have been found to predict subsequent employment status include diagnosis of a major psychiatric disorder before the age of 16; psychoticism; greater perceptions of stress; neuroticism, anxiety and nervousness problems; more depressive affect; lower life satisfaction; antisocial, aggressive, and deviant behaviour; drug (ab)use; and attentional deficits[4] .

Intellectual/cognitive factors that have been found to predict subsequent employment status include the level of intelligence; level of academic potential; reading skills; high school grades; and cognitive development [5]. Individuals with lower IQs, poor reading skills, lower school performance and academic potential, and slower cognitive development thus tend to be at greater risk for subsequent unemployment. The quality of previous work experience has also been found to be an important predictor of future work status (Schneider, 2000).

From this research, the picture painted of the youth who is likely to end up unemployed is not a happy one. These adolescents are likely to have suffered lives filled with adversities such as family problems and a lack of resources, and they are likely to have mental health problems, a low opinion of themselves, and poor intellectual ability.

Fewer studies have investigated the paths of interrelationships between psycho-social factors and educational attainment on employment status. Kokko and colleagues (2000) found that passive and anxious behaviour measured at age 8 lead to poor educational achievement, which then lead to long term unemployment in adulthood. Capsi et al (1998) found that a number of social and personal factors affected employment status indirectly through the duration of education, but they also had direct effects on employment status. Woodward and Fergusson (2000)—in a study especially relevant to New Zealand found that childhood peer relationship problems lead to school related difficulties such as early school leaving, which then in turn increased the risk of youth unemployment. Thus it appears that psycho-social factors may influence future employment status through their effect on educational attainment, but that these psycho-social factors can have a direct influence on future employment as well. Some research findings have also supported the life-course perspective, in that social factors influence personal psychological factors, which in turn influence education and employment status. Bynner (1998) found that SES influenced the quality of identity development, which in turn predicted future employment status. Lynd-Stevenson (1999) found that background factors influenced hopelessness about job seeking and attitudes towards working, which in turn predicted future employment status. Lynn and colleagues (1984) also found that home background influenced a number of psychological variables such as psychoticism, work ethic, and intelligence, which in turn all influenced educational attainment, which then predicted employment status. However, many of the variables in this study had direct as well as indirect effects on employment status.

Most of the conclusions from this literature are to be expected. Employment prospects are better if you come from a high socio-economic status, well-adjusted two-parent family and are confident, motivated, intelligent and have good relationships with your peers when you are young. Perhaps only the last of these would not readily have been guessed at. These factors work on employment prospects both directly and indirectly via achievement in the education system.

The psychological literature reported above focuses on the personality and social characteristics that predict unemployment. In the absence of direct research on their links with low wages, and with wage mobility, it seems reasonable to suppose that the characteristics that predict unemployment will play a role in causing other poor labour market outcomes, including low wages.

Economists have also sought to understand the influence of personal attributes and family background on labour market outcomes. The need for comprehensive longitudinal data to enable causal relations to be identified has limited the number of studies that have been done. An important recent piece of research, Burgess, Garduiner and Propper (2001) draw on the US National Longitudinal Survey of youth (specifically, people who were aged between 14 and 19 in 1979). They are able to trace their subset for 17 years, to 1996. Their objective is to identify the link between family, school and neighbourhood characteristics of young people and their subsequent earnings capacity and risk of being poor. This is not the same as wage mobility, but, as with the psychological literature, it has sufficient family resemblance to be worth reporting (in the absence of more direct evidence). The biggest influence on future earnings came from the family, with area having little separate effect. The family variables that were significantly positively associated with higher earnings were mother and father’s levels of education. For women, fewer siblings lead to higher earnings. For neither sex did coming from a sole parent family have a significant impact on future earnings. We note though, that the total explanatory power of family variables was low—about 12% of variance for men and about 9% for women. These findings support the general conclusion of empirical research in economics, that low levels of parental education (and in some cases, poverty in childhood) have a negative impact on adult earnings (for a review of this evidence, see Haveman and Wolfe, 1995). Family background appears to do its work both directly and through its impact on educational outcomes. The low explanatory power of the empirical estimations suggest that many other factors are at work (or economists have not yet been able to capture the impact of family in a fully effective way).