beneficial “child labor”: the impact of adolescent work … · beneficial “child labor”:...

BENEFICIAL “CHILD LABOR”:

THE IMPACT OF ADOLESCENT

WORK ON FUTURE

PROFESSIONAL OUTCOMES$

Marjan Houshmand, Marc-David L. Seidel and

Dennis G. Ma

Adolescent Experiences and Adult Work Outcomes: Connections and Causes

Research in the Sociology of Work, Volume 25, 191�220

Copyright r 2014 by Emerald Group Publishing Limited

All rights of reproduction in any form reserved

ISSN: 0277-2833/doi:10.1108/S0277-283320140000025007

191

$This research was supported by funding from the Social Sciences and HumanitiesResearch Council of Canada, and the Sauder School of Business. It was made possi-ble through Statistics Canada providing access to the microlevel data through theResearch Data Centres program. The data for this study were accessed at the Inter-University Research Data Centre at the University of British Columbia, with thekind support of Lee Grenon and Cheryl Chunling Fu. We would like to thankHoward Aldrich, Nancy Langton, Claus Rerup, and Martin Schulz for helpful com-ments on earlier drafts of the paper. All mistakes remain the responsibility of theco-authors.

ABSTRACT

Purpose � Theories of income inequality frequently cite child and ado-lescent labor as a societal problem. In contrast to such theories, we pro-pose that path dependency coupled with enhancement of human andsocial capital enables some adolescents who work to find more attractivejobs later in life.

Methodology � Using the longitudinal Youth in Transition Survey(YITS) spanning over 10 years, we find support for a positive relation-ship between adolescents’ number of work hours and future desirable pro-fessional outcomes such as being employed, income, person-organizationfit, knowing where to look for a job, and career networking.

Findings � The positive relationship, in many instances, is curvilinearand highlights the downfall of working excessive hours. We also explorewhether adolescent work for a stranger or family member leads to differ-ent outcomes, and find that working in a family business leads toenhanced later life utilization of career networks as well as better person-organization fit.

Social implications � While we find that adolescent work intensity islinked to positive later life outcomes such as higher income, better fittingjobs, and better career networks, we also find maxima whereby addi-tional hours worked have a diminishing effect on the outcomes. This sug-gests the need for societal norms and/or laws to avoid excessiveadolescent work.

Value of chapter � The findings in this chapter shed light on the role ofearly life work experiences in future professional outcomes. We showthat certain types of adolescent employment can enhance future careerprospects, counter to much of the established literature on the detrimen-tal impact of youth labor.

Keywords: Adolescent work; career development; career networking;human capital; social capital; family business

192 MARJAN HOUSHMAND ET AL.

I live on broken wittles—and I sleep on the coals.

� Charles Dickens, David Copperfield, Ch. 5

Managers and policy makers face the ongoing question of if it is a goodpractice to hire adolescent workers. An underlying assumption in the eco-nomics literature is that there is a trade-off between child labor and humancapital (Baland & Robinson, 2000). The literature suggests most child laborhappens due to the demands of poverty and income inequality of the survi-val household (Basu & Van, 1998). While this assumption likely applies tomany families in poverty, the staggering high number of adolescents work-ing in more developed countries suggests this phenomenon is more than aby-product of income inequality and poverty. For instance in the UnitedStates, more than 50 percent of high school senior students and 6 millionteenagers are employed (Hirschman & Voloshin, 2007).

Similar to many other social phenomena, child labor has gone throughmany changes to reflect de facto societal structure and culture. What wasonce considered to be the “dark satanic mills” (Hindman, 2009) of childexploitation by parents and business owners (Horrell & Humphries, 1995;Zelizer, 1985), is now viewed as educational and developmental opportu-nities for adolescents in preparation for the real world (Brockhaus, 2004;Mortimer, 2003). The change in the meaning associated with adolescentsworking requires a deeper understanding of the historical, economic, andsocial embeddedness of child labor, as well as the true impact of it on laterwork outcomes.

In preindustrial societies, it was normal for children to help in theirfamily firms and child labor mostly implied working alongside with parentson the family farm or business (Kett, 1971). However, as the nature ofwork altered for everyone including parents, children started working fornonfamily members. Until the late 20th century, many families dependedon the financial contributions of these young employees (de Regt, 2004). Inthe past few decades, in North America and some European countries, thenotion of child labor during adolescence has transformed from exploitativeto developmental. The majority of adolescents in these countries experiencesome form of employment and contrary to the past, many have the optionof keeping their income with themselves (de Regt, 2004). While some ado-lescents in this context still work out of necessity or are forced by their par-ents, a significant portion seeks employment for other reasons such ascareer and personal development or a degree of financial independence. As

193Impact of Adolescent Work on Future Professional Outcomes

a result, in many more developed countries, a significant number of adoles-cents work by choice. The form and implications of child labor subse-quently become a function of the national economy, the child laborlegislation in place, and the socio and economic status of parents.

Adolescent experiences have considerable weight on identity develop-ment (Lerner & Steinberg, 2004). As such the notion of adolescent workfuels an ongoing debate in the family studies literature on the risks andbenefits of work during the adolescence stage. While some scholars offamily studies share a negative view of adolescent work and caution againstits harmful effects (Greenberger & Steinberg, 1986), such as stress and alower level of family commitment (Steinberg, Grennberger, Ruggiero,Garduque, & Vaux, 1982), others highlight the positive outcomes, such asa great sense of independence (Mortimer, 2003).

The debate has highlighted the complexity of adolescent work.Numerous contradicting theories as well as scientific evidence suggest thatthe simplistic question of whether adolescent work is good or bad does notadequately address this phenomenon (Mortimer, 2010). Rather, studyingadolescent work requires a deeper understanding of the nature and inten-sity of work in which adolescents engage and its impact on other aspects ofadolescent life both in the short term and in the long term. We know verylittle about the detailed influences of adolescent work on career develop-ment (Staff, Messersmith, & Schulenberg, 2009). The purpose of this studyis to answer questions surrounding this issue and shed light on some of thepotentially positive consequences of adolescent work in the long run. Wetheorize that the concept of adolescent work extends above income inequal-ity and under some circumstances could lead to desirable professional out-comes through enriching adolescents’ human and social capital. We use arich recent longitudinal survey, the Youth in Transition Survey (YITS),that spans over 10 years to test our theory. We conclude by discussing boththe theoretical and practical implications of the relationships between ado-lescent work and career development.

THEORETICAL PERSPECTIVES AND RELATED

RESEARCH

Income Inequality and Adolescent Labor

Child labor is frequently viewed as exploitative in a wide variety of litera-tures. The economics literature discusses child labor as an outcome of


income inequality and has adopted the assumption that parents would notsend their children to work if their own income were sufficiently high(Basu & Van, 1998). Numerous studies argue that child labor indicates ineffi-ciency (e.g., Baland & Robinson, 2000; Ranjan, 2001) and point to a positivecorrelation between child labor and poverty, particularly in developing coun-tries (Ray, 2000). Other literatures suggest social and psychological costsassociated with child labor are not only endured by working children andtheir families (Greenberger & Steinberg, 1986), rather, the costs are alsoimposed on other population groups in the society (Majumdar, 2001). Theysimilarly suggest that child labor is a by-product of lack of family resourcesand a potential challenge to receiving formal education (Walters & Briggs,1993). The exploitation view is often studied within the context of developingcountries such as Pakistan (Bhalotra, 2007), Kenya (Buchmann, 2000),Zambia (Jensen & Nielsen, 1997), Peru (Patrinos & Psacharopoulos, 1997),and Bolivia and Venezuela (Psacharopoulos, 1997) where many children arerequired to participate in the labor market.

The young work literature has not adequately considered the path-dependent and training components of early work that can increase humanand social capital and start an adolescent down a path of a positive futurecareer trajectory. Counter to the line of research that explicitly presumes atrade-off between child labor and human capital (Doepke & Zilibotti, 2005;Hazan & Berdugo, 2002), we argue that with some employment opportu-nities adolescent work could potentially be a positive experience. Given theunstructured nature of transition between school and work in many devel-oped countries (Rosenbaum, 2001), we contend that adolescent work canfacilitate this transition and better prepare an adolescent to find moreattractive jobs as adults. We position our arguments in the organizationaltheory and family studies’ literatures and argue for the positive impact ofadolescent work on career development leading to later life outcomes suchas being employed, having higher incomes, increased person-organizationfit, knowing where to look for a job, and stronger career networking.

Path Dependency from Adolescent to Adult Work

Hall (1904) coined adolescence as a “new birth” in which the environmentplays a key role in shaping rapid psychological and social development. Inthe preindustrial era children commonly helped labor in family farms andshops. Attaining apprenticeship at age of 14 in the care of a relative orneighbour or helping with the farm activities until the age of 21 was a com-mon practice in the 19th century (Kett, 1971). After the industrial


revolution, alongside parents, children became factory workers working inharsh conditions (de Regt, 2004). Law makers changed legislation to legallyprohibit children from working under physical and emotional exhaustion.The reduction of labor force in the market led to an increase in adult wagesand consequently compensated for the loss of children’s financial contribu-tions to the family (de Regt, 2004). In the early 20th century, while laborreformers contested child farm work, they were empathetic toward childrenworking in their own farm families (Mortimer, 2003; Zelizer, 1985). Childlabor laws that protected children from excessive or dangerous work werelater expanded to include working in family farms and businesses(Mortimer, 2003).

Yet today a large segment of adolescents around the world engage inpaid labor. For example in the United States, around 80�90 percent of ado-lescents work sometime during their high school years (Marsh, 1991).Similarly in 2006, 25 percent of all Canadian workers were youth under theage of 20 (Statistics Canada, 2010). While these statistics show that a largenumber of adolescents work, the general improvement in the family finan-cial pool in the recent decades has weakened the need for adolescents to befinancial contributors to the family. Many of these young workers haveauthority over saving and spending their earnings and working in adoles-cence has become a developmental preparation for later life (de Regt, 2004).

While many scholars agree that work shapes adolescents in both theshort term and the long term (Mortimer, 2010), there is less agreement onwhat aspect of adolescents’ lives is influenced by employment, the degree ofimpact, and even the desirability of the impact. An important impact ofadolescent employment is on professional outcomes later in life, an under-studied topic in the literature (Staff et al., 2009). Early professional experi-ence situates adolescents on a career path dependency shaping their workexperiences later in life. Controlling for other factors, we contend thatemployment options in adulthood relate to the type of aspirations andexperiences to which adolescents are exposed (Packard & Nguyen, 2003).Path dependency describes the relationship between future events on initialconditions (Mahoney, 2000). We postulate that those adolescents who gainearly professional experience tend to follow a career path dependency thatkeeps them employed. Not only do adolescents who work develop a habitof working and having access to money, their work experience gives theman edge in finding and securing job opportunities.

H1: The number of work hours during adolescence is positively relatedto being employed later in life.


Adolescent Work and Human Capital Enrichment

The concept of adolescent employment is closely related to human capitalacquired by adolescents during their teenage years. Becker (1962) explainedhuman capital as a set of intangible skills, knowledge, and competencyaccumulated in people that has the capacity in attaining future economicvalue. A common example of human capital is education that offers knowl-edge and well-recognized qualifications as a precursor to numerous profes-sional jobs.

Similar to other types of capital, the return on investment in humancapital depends on situational factors as well as the nature and intensity ofinvestment in a particular type of human capital. It is not difficult toimagine that the magnitude of return on the number of years spent in edu-cational training varies considerately across different disciplines, universi-ties, countries, and economic and political cycles, and it is contingent uponindividual differences such as age, gender, socioeconomic backgrounds,and social ties. Even though there is uncertainty associated with the magni-tude of return, the literature suggests an overall positive relationshipbetween education and future economic value (Becker, 1964). To under-stand if adolescent employment categorized as a form of human capitalleads to future economic value, we need to first establish adolescentemployment as a form of human capital. The majority of adolescents findemployment in retail and service sectors and engage in a variety of tasks(Staff, Mortimer, & Uggen, 2004). They engage in activities that are differ-ent in nature from those in their school or home. They assume responsibil-ities and exchange their labor for monetary rewards. They take part insocial interactions with their employers, colleagues, and customers. Theyalso become more competent in managing their time while balancing thedemands of school and work (Mortimer, 2010). All these skills, knowledge,and competency gained through work embed intangible resources in ado-lescents that can help them in securing future jobs and earnings.

Comparable to job skills training, another type of human capital, ado-lescent employment prepares individuals for richer and more rewardingfuture work experiences. A few empirical studies have explored the rela-tionship between work intensity during adolescence and earned incomeyears later and provide support for the human capital argument. Whilesome show that there is no relationship between the intensity of workwith future income above and beyond working in general (Mortimer,2003), other studies demonstrate the link between intensity of work onpersonal and professional outcomes later in life (Mortimer, 2010; Staff


et al., 2009). One notable study using data from 1979 focuses on thepositive impact of work intensity during senior year on earned incomelater in life (Ruhm, 1997). The study showed that those who work longerhours during the last year of high school are more likely to earn higherincomes later in life. Given the main focus of the study on work duringthe senior year when adolescents are closer to becoming adults, and thepotential change of job opportunities and adolescents’ attitude towardsjobs since 1979 (Mortimer, 2010), the question of the impact of employ-ment in earlier years on economic attainment in the modern day is stillopen for updating.

Other research explores the relationship between work intensity andfuture income from the perspective of self-selection. Previous findings indi-cate that the adolescents’ socioeconomic status and academic performanceplay a role in adolescent employment such that those with higher socioeco-nomic status and higher average grades are more likely to be engaged inwork classified as “good jobs” characterized by fewer hours of work andhigher status (Hirschman & Voloshin, 2007). In the same vein, Staff andMortimer (2008) assert that there is a relationship between the degree ofadolescent involvement in work and their subsequent post-secondary edu-cational and economic pursuits. They categorize adolescent work into fourgroups based on the intensity of work hours per week and duration ofwork during a year. Their findings suggest that the most fruitful workengagement is steadily working 20 hours a week or less which is mostlyobserved in adolescents with higher class backgrounds. They argue thatthe relationship between adolescent employment and their behaviourbecomes more a function of self-selection dependent on adolescent back-ground and interest in school (Staff & Mortimer, 2008).

While we do not contest the role of socioeconomic background and aca-demic performance in future economic performance, we argue that adoles-cent work intensity uniquely contributes to development of adolescents fora professional world. We argue that the intensity of work, particularly dur-ing early phases of adolescence, provides adolescents access to valuablelearning opportunities and enables them to enrich their human capital(Mortimer, 2003) and enhance their soft skills (Staff et al., 2009), which actas a means to future financial gains. The work experience contributes tohigher income by going above and beyond the role of gender, race, socioe-conomic status, and average high school grades. Compared to their peerswith little or no work experience, adolescents who work longer hours havemore impressive resumes, better references, and a deeper knowledge of howorganizations operate in general, all of which lead to securing higherincome jobs than their peers later in life.


H2a: The number of work hours during adolescence is positively relatedto higher income jobs later in life.

We further posit that adolescent employment is positively related tofinding jobs with higher degrees of person-organization fit. Research find-ings in the organizational studies literature highlight the importance ofperson-organization fit on a number of organizational outcomes such asjob satisfaction, organizational commitment, turnover, organizational citi-zenship behaviors, and performance (Hoffman & Woehr, 2006; Verquer,Beehr, & Wagner, 2003). Mortimer (2003) argues that anecdotal evidencedemonstrates that early work experience supplies adolescents with theknowledge of what they like to do and what they do not like. She goes onto quantitatively analyze the career relevance of work seven years afterhigh school, yet found no significant impact of work intensity above andbeyond working in general on career relevance.

We argue that instead of focusing on career relevance, we should focuson person-organization fit. The more adolescents are immersed in the jobexperience, the more likely they will develop a deeper level of understand-ing of their own job preferences and work conditions. As they grow older,adolescents gain a more realistic understanding of the world (Staff et al.,2009). Their career aspirations and valued job characteristics become rela-tively stable (Blanchard & Lichtenberg, 2003; Hartung, Porfeli, &Vondracek, 2005). Work during this young age can also play an importantrole in shaping career values and aspirations (Porfeli, 2007). The deeperknowledge coupled with richer human capital prepares adolescents to selectbetter fitting working environments than their counterparts. Not only dothese assets facilitate adolescents selecting better fitting organizations, butalso yield sharper business acumen to help seize appropriate opportunitiesin the job market matching their preferences.

H2b: The number of work hours during adolescence is positively relatedto better person-organization fit later in life.

Adolescent Work and Social Capital Enrichment

Social ties can also play an important role in securing future jobs. Socialcapital, intangible assets arising from social ties, has been associated withmany desirable economic outcomes (Adler & Kwon, 2002) � from land-ing a job (Petersen, Saporta, & Seidel, 2000) to receiving promotions(Seidel, Polzer, & Stewart, 2000). One may find a job through friends orfamily working in a particular organization and have some degree of


power or knowledge arising from their ties to help in getting the job.Research suggests ties, particularly weak ties, play an important role forindividuals who are seeking jobs (Brown & Konrad, 2001; Granovetter,1973).

We theorize that adolescent work broadens and enriches adolescents’social capital, and similar to adults, such ties help individuals to findjobs. Through work, adolescents form new ties with people outside oftheir circles of family, friends, and school. These ties connect them toprofessionals across various positions and industries. Not only do theseties connect adolescents to employment opportunities, but also they canbe a fruitful source of knowledge on where and how to look for jobs.The supervisors and colleagues at work generally have more experienceto bank on when it comes to job search than the teenage friends of ado-lescents. Therefore, the ties formed at work become a source of informa-tion for adolescents to learn about available job opportunities and whereto look for jobs.

H3: The number of work hours during adolescence is positively relatedto having better career networks later in life.

METHODS

Data and Sample

We used a complex recent longitudinal dataset called the Youth inTransition Survey (YITS) to test our theory and hypotheses. The YITS wasconducted jointly by Statistics Canada and Human Resources and SkillsDevelopment Canada and employed rigorous research design and collec-tion guidelines. It has followed a sample of 15-year olds, representing apopulation of 246,661 Canadian adolescents for our analysis. This cohorthas been surveyed every two years and for this analysis we have data until2009 (Cycle 6) when the 15-year olds became 25-year olds. Per StatisticsCanada’s guidelines, we used their given longitudinal weights to accountfor attrition across different cycles. Additionally, as suggested by StatisticsCanada, we employed Stata BRR (balanced repeated replication) boot-strapping procedures. The use of longitudinal weights and Stata BRR boot-strapping procedures allows us to generalize our findings to the broaderpopulation represented by our sample.


Measures

Dependent VariablesBeing employed: We created a dummy variable for each cycle to capturethe employment status of respondents. We coded it 1 for those who hadany work involvement during the year, otherwise it was coded as 0.

Income: This variable was based on the total income from all sourcesbefore taxes and deductions in the previous year. Following establishedpractice (Manning, 1998), we transformed this measure by taking the nat-ural log of a respondent’s income plus 1.

Person-organization fit: We operationalized person-organization fit asreservation wage expressed in dollars and cents per hour. Reservation wageis the lowest wage a respondent would accept to begin a new full-time jobimmediately (Burdett, 1978). The difference between reservation wage andcurrent income assigns a dollar amount to all the nontangible benefitsreceived from a job from loyalty to the firm to the physical space of thefirm (Mailath & Postlewaite, 1990). In all models predicting fit, we there-fore controlled for current income to rule out higher reservation wage asmerely a linear function of income. By controlling for individuals’ currentincome in the models, the reservation wage includes but is not limited toother desirable factors besides income that makes each job more or less fit-ting for individuals. Following the established norm of the field (Manning,1998), we further transformed this measure by taking the natural log of arespondent’s reservation wage and added 1.

Career networks: We used two different measures to operationalizecareer networks based on the extent to which respondents knew where tofind a job and their ability to learn from others about a job. A large portionof job search happens through informal channels and contacts and themajority of contacts people utilize when looking for a job are work con-tacts that are made professionally as opposed to personal contacts(Granovetter, 1974).

In the first operationalization we used a single item measure which askedunemployed respondents who were actively looking for a job whether theyknew where to look for a job. The question was originally worded as “notknowing where to look for a job” and we reverse coded it for the analysis.We created a dummy variable knowing where to look for a job in such away that 0 indicated the respondent who did not know where to look and 1indicated they knew where to look for a job. For the second measure, weused the question, “outside of educational programs and training courses,during the last year, have you through your own initiative, watched others


work, or received advice or assistance from others, to learn for a jobor career?” We operationalized this as a dummy variable career networkingin such a way that “No” to the above question was coded as 0 and “Yes”as 1.

Independent VariablesThe survey had separate questions about employment during the schoolyear and employment during the summer time. For consistency and also totease out the difference in the impact of work during the school year andsummer, we created two independent variables. The number of hoursworked during the school year refers to the number of hours worked in atypical week during the school year, whereas the number of hours workedduring summer captured weekly work hours during summer. Additionally,consistent with previous findings (Lee & Staff, 2007), we suspect that therelationship between the number of hours worked and the outcomes maybe curvilinear. As such we created square terms for number of hoursworked during fall and summer and included them in the analysis. Finally,outliers were excluded by cutting off the top 0.5 percent of hours workedfrom each of the two seasons from our data, as those reported hours whichappear to be response errors.

Control VariablesWe controlled for some basic individual attributes such as age, gender, andrace (Hirschman & Voloshin, 2007; Pabilonia, 2001). We also included anumber of family characteristics that play a role in the self-selection processof adolescent work (Mortimer, 2010), such as number of siblings, parents’socioeconomic status, and parents’ average age as well as whether respon-dent lived in urban or rural areas. Finally, we controlled for adolescents’GPA to account for and tease out their high school performance fromwork experience. Working for a stranger is different from working for arelative (Eckrich & Loughead, 1996; Gomez-Mejia, Cruz, Berrone, & DeCastro, 2011), and this difference may confound the influence of work onadolescent career development. To control for this effect, we included adummy variable coded as 1 when adolescents had done any work in theirfamily business during summer or school seasons.

Our analysis included all the control variables that were measured dur-ing adolescence and purposefully excluded contemporary job attributessuch as organizational size and tenure measured during adulthood. Weargue that those adolescents who work gain a competitive advantage laterin life and can select themselves into job opportunities with desirable


professional outcomes such as income and person-organization fit, andthat controlling for such contemporary adult variables would mask exactlywhat we are seeking to test. We, therefore, limited our control variables tothose factors that literature suggests shape adolescent experience, and allowthe selection effects to be captured in the dependent variables.

RESULTS

We used weighted regressions and logit to test our hypotheses as appropri-ate. In accordance with the Statistics Canada’s privacy restrictions, we arenot permitted to provide details about the sample size and correlationamong variables but rather report on the target population weightedappropriately. For our analysis, the target population covered 246,661Canadians. We examined the variance inflation factor (VIF) values of allvariables in all models � with the exception of squared work hour terms,and found no significant multicollinearity problems (VIF< 1.26).

We measured our independent variables when adolescents were 15 yearsold and conducted analyses on appropriate dependent variables for all theavailable subsequent years from when adolescents were 17 years old to 25years old. For each age cycle, for presentation simplicity, we present a fullmodel that contains all the control variables and independent variables.However, in the actual analysis we ran four separate models for each agecycle and dependent variables to examine the effect of important covariatesindependently. For the cases with significant variations among the fourmodels, we describe the results in the text.

Table 1 presents findings on the impact of adolescent employment onbeing employed later in life. Since being employed is a dummy variable, weconducted a logit regression to test H1. Table 1 captures the results forbeing employed measured when adolescents were 17�25 years old.

As seen in Table 1, at ages 17 and 19, we find strong support for a posi-tive relationship between hours worked (at the age of 15) and higher likeli-hood of being employed two and four years later. The significant negativesquared coefficients suggest that the effect is not linear and the likelihoodof being employed later decreases if adolescents exhaust themselves byworking too many hours during either school or summer seasons. Wefound incremental hourly benefits peaked around 25�31 hours per weekduring the school season, depending on the age cycle. The intersection(zero net effect) occurs at exactly double the maxima. This suggests that,


Table 1. Weighted Logit Predicting Being Employed for Adolescents from Age 17 to 25.

Being Employed at Different Ages

Age 17 Age 19 Age 21 Age 23 Age 25

Female 0.131+ −0.00406 0.147 0.0244 −0.128White 1.022*** 0.798*** 0.765*** 1.053*** 0.369

Urban −0.220* 0.0601 0.147 0.339* 0.142

Number of siblings 0.0171 0.00480 −0.0244 0.0891 0.0450

GPA 0.120** 0.146** 0.0749 0.144* 0.0483

Parent’s SES 0.118* 0.279*** −0.0810 0.114 0.120

Parent’s avgerage age −0.0158* −0.0187* 0.0239* 0.0179 0.00411

Family business any season 0.0763* −0.0391 −0.0692 0.104 + 0.00264

School hours worked 0.0641*** 0.0510*** 0.0252* 0.00225 0.0386*

School hours worked squared −0.00110*** −0.000967*** −0.000494* −0.0000371 −0.000613*Summer hours worked 0.0372*** 0.0325*** 0.0145 0.00151 −0.00674Summer hours worked squared −0.000410*** −0.000385*** −0.000127 −0.0000364 0.0000567

Constant 1.107** 1.846*** 0.588 0.428 2.163*

Prob>F 5.08e-75 1.16e-27 0.000000170 0.0000188 0.479

F 39.03 14.23 4.703 3.685 0.967

N 236,943.8 236,660.4 237,483.2 237,114.3 239,443.6

+p< .10; *p< .05; **p< .01; ***p< .001.

Note: All independent variables are measured when adolescents were 15 years old.

204

MARJA

NHOUSHMAND

ETAL.

though curvilinear in shape, working more hours during the school year(within reason) for an adolescent is beneficial in terms of their futureemployment prospects, though the benefit levels off as thisfigure approaches 25 hours. Taking into account that few 15-year-oldCanadians work these many hours, the results simply relay that there arefuture employment benefits to working more hours during the school year,characterized with diminishing marginal returns on each additional hourworked. At age 17�19 summer employment effects follow a similar patternwith an extended range. The maxima is between 42 and 45 hours, andshould/can be interpreted similarly, with the distinction that full-time workduring the summer months appears to have the maximum impact.

Table 1 also shows that there is mixed support for H1 in later years.While the number of working hours during the school season at age 15 issignificantly associated with higher likelihood of being employed at age 21,there is little evidence for number of working hours during summer andbeing employed in later years (though there appears to be a significanteffect in the age 25 model, the model F-statistic shows that the model itselfis not significant). There is also little evidence that those who work in theirfamily business are more likely to be employed later in life.

Table 2 depicts the results of the weighted regression for the logged over-all income variable when adolescents were 17�25 years old. For all ages,hours worked during school when adolescents were 15 years old led tohigher incomes in subsequent years from 17 to 25 years old. Furthermore,the number of hours worked during summer at age 15 was positively asso-ciated with higher income at ages 17, 19, and 23. The higher work intensityduring both seasons is positively associated with higher incomes later inlife, providing support for H2a.

We also observed significant negative coefficients for the squared terms ofhours worked during school and summer seasons. This implies the relation-ship between number of hours worked and future income is curvilinear.There exists a maxima in which working longer hours has a detrimentaleffect on future earnings. The maxima for the number of hours work duringschool season is 33 hours for both 17 and 19 ages, whereas the number goeshigher to 45 and 40 hours during summer, respectively. The maxima are 31and 33 hours, respectively, at ages 21 and 23 implying that additional hoursworked during the school season at the age of 15 yields higher income atages 21 and 23 until it reaches 31 hours (33 hours for age 23) and after thatevery hour decreases the marginal earned income. We observe a similar cur-vilinear pattern between the number of hours worked during summer andearned income at the age of 23 with a maxima at 46 hours.


Table 2. Weighted Regression Predicting Logged Overall Income for Adolescents from Age 17 to 25.

Logged Income at Different Ages


Female −0.143*** −0.155*** −0.248*** −0.274*** −0.220***White 0.474*** 0.447*** 0.372*** 0.358*** 0.165*

Urban 0.0149 0.0582 0.0220 0.00127 −0.00673Number of siblings −0.0186 0.00309 0.00141 −0.0279 + −0.0306+

GPA 0.0714*** 0.112*** 0.0192 0.00673 0.0882**

Parent’s SES 0.0691* 0.108** −0.00888 −0.0199 0.0397

Parent’s average age 0.00409 −0.00373 −0.00186 −0.00854* −0.00734Family business any season 0.01000 −0.00842 −0.0107 0.0361** 0.0219

School hours worked 0.0392*** 0.0287*** 0.0189*** 0.0134*** 0.0141**

School hours worked squared −0.000598*** −0.000438*** −0.000304** −0.000201* −0.000259Summer hours worked 0.0210*** 0.0126*** 0.00401 0.0107*** 0.00514

Summer hours worked squared −0.000233*** −0.000157*** −0.0000480 −0.000116*** −0.0000397Constant 6.441*** 7.810*** 8.914*** 9.786*** 10.01***

Prob>F 8.64e-57 1.53e-29 8.93e-11 5.50e-43 1.14e-12

F 28.90 15.13 6.277 21.74 7.170

N 236,943.8 236,660.4 237,483.2 237,114.3 239,443.6

+p< .10; *p< .05; **p< .01; *** p< .001.


206

MARJA

NHOUSHMAND

ETAL.

These numbers should be interpreted with care as the majority of ourpopulation do not come close to reaching the maxima: the relationshipbetween early work and later income is a positive one, but with decreasingmarginal returns to each additional hour worked. Finally, as presented inTable 2, the findings also suggest that the observed positive effect of workon future attained income is not significantly influenced by the family rela-tionship between adolescents and their employers. Rather it is more a func-tion of number of hours worked and the gained work experience in generalas opposed to working for a family member.

Table 3 pertains to the impact of adolescent work intensity on person-organization fit, measured as reservation wage, from age 17 to 25 years old.In addition to the base control variables, we also controlled for earnedincome from the same year to assess the intangible benefits above andbeyond income. When adolescents were 17 years old, we find marginally sig-nificant support for H2b only for number of hours worked during summer.When adolescents were 19 years old, the number of hours worked duringsummer, four years prior at age 15, has a significant positive effect onperson-organization fit. This effect is marginally significant for the numberof hours worked during school season. The curvilinear effects are significantwhen adolescents are 17 and 19 years, respectively, only for number of hoursworked during summer with maxima at 31 and 46 hours, respectively.

The findings at ages 21 and 23 provide strong support that higher num-ber of work hours during the age 15 school year are positively associatedwith higher fit six and eight years later. We see a similar pattern for numberof hours worked during summer and higher person-organization fit whenadolescents were 21 years old. At age 25, the story is more complex. In anunreported model without the family business variable, the number ofhours worked during the school year has a direct effect but it is not curvi-linear.1 The effect disappears entirely when controlling for family business.There is no significant association between number of hours worked duringsummer and person-organization fit at age 25. The curvilinear relationshipwas not significant for any of the age cycles, suggesting that there is nodampening effect of hours worked at person-organization fit. Overall,we find consistent support for H3 that those who are employed duringadolescence tend to find better fitting jobs later in life, reflected by higherreservation wages that go above and beyond their earned income.The data suggests that summer work intensity has more impact on person-organization fit at earlier ages, and school year work intensity has its effectson person-organization fit on a longer term. Family business work experi-ence goes above and beyond the positive relationship between the number


Table 3. Weighted Regression Predicting Logged Person-Organization Fit for Adolescents fromAge 17 to 25.

Logged Person-Organization Fit at Different Ages


Female −0.0512*** −0.0805*** −0.0918*** −0.0878*** −0.0816***White −0.101*** −0.0703*** −0.0822*** −0.116*** −0.0714***Urban 0.0711*** 0.0362*** 0.0209* 0.0176 0.00228

Number of siblings −0.00476+ −0.00304 −0.00931** −0.00379 −0.00595GPA 0.00947** 0.00859** 0.0310*** 0.0598*** 0.0726***

Parent’s SES 0.0238*** 0.0178*** 0.0275*** 0.0695*** 0.0741***

Parent’s average age 0.000141 −0.000881 −0.00123 −0.000795 −0.000653Income 0.0165*** 0.0317*** 0.0685*** 0.102*** 0.101***

Family business any season 0.00624* 0.00484 + 0.00817* 0.00752 + 0.00882 +

School hours worked 0.000174 0.00129 + 0.00294** 0.00314** 0.00214

School hours worked squared −0.000000352 0.00000301 −0.0000325 −0.0000404 −0.0000253Summer hours worked 0.000756 + 0.00153** 0.00139* 0.00118 0.00143

Summer hours worked squared −0.0000121* −0.0000168* −0.0000151 −0.0000131 −0.0000202Constant 1.994*** 2.063*** 1.922*** 1.759*** 1.834***

Prob>F 3.83e-72 7.63e-120 9.73e-130 1.80e-119 2.46e-89

F 34.89 63.07 69.73 62.83 44.35

N 225,594.3 226,628.4 226,506.5 229,902.6 232,958.4

+p< .10; *p< .05; **p< .01; ***p< .001.


208

MARJA

NHOUSHMAND

ETAL.

of work hours during school and summer seasons and the person-organization fit variable. This is an interesting finding and suggests thoseadolescents who work in their family business have a different and uniqueexperience compared to their peers who work for others. These adolescentsare better equipped in recognizing and securing better fitting jobs in thefuture as suggested by their higher reservation wage.

Finally, Tables 4 and 5 present findings for H3 on the positive relation-ship between adolescent employment and better career networks. ForTable 4, the analysis was conducted over a subpopulation of our sampleand restricted only to those who were actively looking for a job in a givencycle (e.g., a target population of 39,693 at age 17). The findings showmixed support for the relationship. The evidence, however, is stronger inearly years suggesting the social capital created in early employment yearsfacilitates access to career networks in terms of knowing where to look fora job but does not carry a lasting effect. The number of hours worked dur-ing summer is marginally significant and positively related to knowingwhere to look for a job when adolescents were 17 and 19 years old andsimilar as before, this relationship is curvilinear with a maxima at 36 and38 hours, respectively. On the other hand, the number of hours workedduring the school season when adolescents were 15 years old is positivelyrelated to higher knowing where to look for a job when adolescents were21 years old but not for later age cycles.

A plausible explanation for our findings is that for those who are activelyseeking employment, prior work experience and social ties are a source ofinformation that has a more unique value in early ages than later phases inlife. This is consistent with our findings in Tables 1 and 2 that many adultsfind employment later in life regardless of their prior work experiences inadolescence. Given that a higher percentage of relatively older adults gainwork experience, have had educational training, and have had friends whoalso work than those who just enter adulthood, it is not difficult to assumethat they have a better understanding of where to look for job opportunitiesthan 19-year olds who are still fresh in the work force.

Table 5 shows the analysis for the second measure of career networks.Our results show partial support for the effect of number of hours workedduring adolescence on having better career networks later in life.Unfortunately data on the dependent variable was not available for the age19 cycle, so we only report the other four available ages in the data. Asseen in Table 5, the effect of the number of hours worked during schoolseason on career networking is marginally significant at age 21 with a max-ima at 26 hours. This relationship becomes significant at age 23 and again


Table 4. Weighted Logit Predicting Knowing Where to Look for a Job from Age 17 to 25.

Knowing Where to Look at Different Ages


Female −0.158* 0.125 −0.302 + −0.0064 −0.199White 0.392* 0.224 0.571* 0.251 0.0768

Urban 0.0289 −0.199 −0.216 −0.155 −0.375Number of siblings −0.108* 0.00485 0.0345 0.00346 −0.106GPA 0.0514 0.0941 0.0631 −0.141 −0.112Parent’s SES 0.161 + −0.256** −0.191 −0.122 −0.0420Parent’s average age −0.00184 0.0191 −0.00660 −0.0278 −0.0285Family business any season −0.0377 0.0209 −0.115 + 0.0668 0.0780

Hours worked during school 0.0129 −0.0191 0.0450** 0.0267 0.0305

Hours worked during school squared −0.000301 0.000966* −0.000639 + −0.000371 −0.000811Hours worked during summer 0.0205 + 0.0182 + −0.00356 −0.00328 0.0159

Hours worked during summer squared −0.000281 + −0.000242 + 0.0000811 −0.00000375 −0.0000883Constant 0.724 −0.549 0.640 2.593** 3.112*

Prob>F 0.00235 0.0000762 0.0155 0.0752 0.0536

F 2.568 3.371 2.088 1.641 1.742

N 38,301.2 38,386.2 37,835.4 33,985.2 35,647.3

+p< .10; *p< .05; **p< .01; ***p< .001.


210

MARJA

NHOUSHMAND

ETAL.

Table 5. Weighted Logit Predicting Career Networking for Adolescents from Age 17 to 25.

Career Networking at Different Ages

Age 17 Age 21 Age 23 Age 25

Female 0.0662 −0.00314 0.0368 0.0468

White −0.114 −0.0824 0.00635 −0.0978Urban −0.130* −0.0428 −0.0276 0.0869

Number of siblings 0.0334 + −0.0171 0.00206 −0.0136GPA 0.000921 0.0367 0.130*** 0.153***

Parent’s SES −0.00474 0.200*** 0.301*** 0.247***

Parent’s average age −0.00439 −0.0036 0.00138 0.0152*

Family business any season 0.0893*** 0.0759*** 0.019 0.0705*

School hours worked 0.00306 0.0104 + 0.0134* 0.0138 +

School hours worked squared −0.0000291 −0.000197 + −0.000308* −0.000264Summer hours worked 0.0105** 0.0024 −0.00263 0.00268

Summer hours worked squared −0.000120* −0.0000432 0.0000513 −0.0000143Constant 0.977*** 0.475 −0.18 −1.097**Prob>F 1.52E-12 7.81E-07 9.72E-14 5.66E-14

F 7.112 4.377 7.673 7.783

N 230,523.1 229,132.7 232,001.3 235,416.5

+p< .10; *p< .05; **p< .01; ***p< .001.


211

Impact

ofAdolescen

tWork

onFuture

Professio

nalOutco

mes

marginally significant at age 25. This relationship is only marginally curvi-linear at the age of 23 with the maxima at 22 hours.

Number of hours worked during summer, on the other hand, is signifi-cant for only age 17 with a maxima at 44 hours, but not in later years.Taking the outcomes of both measures together, we conclude that adoles-cent work experience during the school season plays a longer term role thanof employment during the summer in providing better career networks.Summer work seems to have shorter run impact on career networks.

Table 5 also portrays an intriguing pattern about the relationship betweenadolescent family business experience and having better career networks laterin life. While we, overall, found no significant relationship between familybusiness experience and knowing where to look for a job, we did find astrong significant positive association between working in a family businessand career networking. These results illuminate another noticeable differencein terms of career-related outcomes between adolescents who work for theirrelatives compared to their peers who work for strangers.

We ran a number of robustness checks. Given our study draws fromdata when adolescents were 15�25 years old, our main concern was therole of education in human and social capital development during thoseyears. We, therefore, conducted a series of additional analyses to assesswhether controlling for education and the transition between work andeducation from one time point to another would alter our findings. In thefirst wave of analysis, we repeated all the models with two additional con-trol variables � individual’s educational and professional engagements,measured at the same year as the dependent variables. In the second wave,we created more control variables by adding work and education dummiesfrom all previous cycles and included them in our models. Finally, we cre-ated one variable for each education and work engagement which countedall instances of education/work leading up to and including the dependentvariable cycle. We added these two variables to our original models as wellas we added the interaction between the two variables. In all of theserobustness checks we found the results were substantively unchanged, andtherefore report the simpler models due to space constraint.

DISCUSSION

How does adolescent employment shape career development outcomes? Inthis chapter, we theorized that adolescent employment could go beyond


issues surrounding income inequality as suggested in the economics litera-ture (Basu & Van, 1998) and has the potential to positively imprint long-term professional outcomes. To understand how early employment shapesfuture professional outcomes, we drew from literature in sociology, familystudies, and organizational theory to highlight the role of path dependency,human capital, and social capital to explain the observed positive effects.Given that adolescence is a critical stage in one’s life (Lerner & Steinberg,2004), early employment prepares adolescents for professional successmainly through enhancing human and social capital accessible to adoles-cents and acts as a competitive advantage in the labor market.

This study contributes to the family studies’ literature by providingempirical evidence for some of the positive influences of adolescent workfrom a long-term perspective. This study sheds light on an understudiedarea in the literature of adolescent career development and its findingscould be a valuable contribution to the current dynamic debate that dis-cusses the harms and benefits of adolescent work (Staff et al., 2009).Through our analysis and findings, we were able to carefully assert causal-ity and generalize the findings to a broader population by using a compre-hensive and recent longitudinal data set spanning over 10 years.

We simultaneously accounted for numerous causal factors previouslysuggested independently in the literature, by including a broad variety ofattributes such as gender (Stevens, Puchtell, Ryu, & Mortimer, 1992),academic performance, and socioeconomic background (Staff &Mortimer, 2008). This study contributes to our understanding by includ-ing the long-term impact of work independently of the demographic fac-tors. Our findings suggest that adolescent work intensity is linked to laterlife outcomes such as higher income, better fitting jobs, and better careernetworks. We were also able to draw deeper inferences about the rela-tionship between adolescent work and career development by examiningthe impact based on the number of hours worked rather than a relativelysimplistic dummy variable for any work experience. This allowed us tocalculate maxima whereby additional hours worked have a diminishingeffect on the outcomes.

Our findings suggest that the positive likelihood of being employed as aresult of adolescent employment diminishes over time. This is, however,not surprising given that as adolescents enter adulthood they increasinglyengage in labor market regardless of their prior work experience.Prior work experience sets adolescents apart in terms of what type of jobthey can secure and what kind of career networks they can establish anddraw upon.


We found strong support for the positive relationship between higherincome and number of hours worked during summer and school year suchthat adolescent work is positively associated with higher earned income infuture years. In our models the regression coefficients for number of hoursworked during school season are higher than the corresponding coefficientsfor number of hours worked during summer. In other words, an hourworked during the school year has a higher return than an hour workedduring summer. A plausible explanation is that those adolescents whostudy and work at the same time are more likely apt in time managementskills and can secure higher income jobs than their counterparts.

We also found strong support for the positive impact of adolescent workon person-organization fit later in life. Adolescents learn about what theylike and what they do not through early exposure to work (Mortimer,2003) and their understanding of the labor market enables them to bematched to better fitting work environments. Similar with income, weobserve the more central role of work during school season as opposed tosummer that helps adolescents in working in a better fitting organization.As such, working in summer had a positive outcome in earlier yearswhereas the impact of working in the school season was observed atlater ages.

Finally, work during adolescence provides access to better career net-works, although support was mixed. This study highlights and empiricallyanalyzes the important role work plays in developing the social capital ofadolescents � an area that has not been empirically explored within thefamily studies’ literature previously. The data suggests that work helps ado-lescents with establishing better career networks from which they can learnabout jobs and job opportunities. The effect is more visible and greater forwork during the school season. Evidently, not only does the intensity ofwork matter in adolescent work development, but also the seasonal impactof work with respect to other life obligations such as schooling is an influ-ential factor. Therefore, in investigating the impact of adolescent work, thetemporal seasonal dimensions of work also play a role in shaping theseyoung employees.

An important practical and theoretical implication of this study is thatwhile the findings suggest working longer hours is linked to more desirableoutcomes through enhanced human and social capital, there is a maximawhere the impact is reversed (Lee & Staff, 2007; Mortimer, 2010).For example, the data indicates that for income on an average there is anoptimum range of work hours during school and summer seasons, whichwe found to be around 33 and 43 hours per week, respectively. This implies


that for an average adolescent an increase in work hours up to the 33rdhour has a marginal benefit but after the 33rd hour the increase in workhours is negative.

Our findings are surprising given previous studies have chosen thereference group for work intensity to be 20 hours and have categorizedadolescent work into groups of working less than 20 hours and workingmore than 20 hours. While these studies usually point out to the benefitsof working fewer than 20 hours (Mortimer, 2010), our study highlightsthe need for future research to focus on including more concrete measureof work such as number of work hours rather than categories of workintensity. Practitioners and policy makers need to be careful that thisrange is the average observed pattern for the general population andfurther customization is required for each adolescent depending on theirunique personal attributes. As the data shows, it also varies during schoolyear and summer time. Therefore, while we encourage adolescents toengage in work, we caution them against working too many hours andexhausting themselves or losing the opportunity for other types of posi-tive development.

Our findings should be viewed cautiously and not be used in promotingchild labor without providing the necessary learning opportunities and safeenvironment. Our sample was taken from a developed country where thelaws and regulations offer some protection to children from being abusedand may not be applicable in societies with more lenient child labor laws.We acknowledge that even in this context some children may not have anoption and work due to necessity. In such cases, the number of work hoursis usually dictated to them by their financial situation. Therefore, as thefindings in this study suggest, the effect of work on future professional out-comes becomes detrimental when these adolescents work longer hours thanthe maxima points and the influence of human and social capital gainedfrom work experience during adolescence may be severely limited inextreme cases.

In this vein, we conducted supplemental analysis to find out whowould be more likely to work and also work longer hours. Given ourdata structure, the earliest information available to us on adolescents wasfrom the same time as when our independent variables were measured.Therefore, we were able to only conduct cross-sectional models. Weobserved that female white adolescents, who had access to their familybusiness, had more siblings, were born to younger parents, and hadhigher GPAs and higher socioeconomic backgrounds, were more likely towork during the school year than their peers. On the other hand,


adolescents with similar attributes but who had lower GPAs and lowersocioeconomic backgrounds were more likely to work longer hours. Thisis consistent with the previous literature that suggests social class isrelated to adolescent work intensity (Staff & Mortimer, 2008). Given thecross-sectional nature of the data, these findings suggest future studiesshould investigate who selects into working and who selects into workinglonger hours during adolescence, and explore whether similar mechanismsand attributes could also play a role in selecting the type of work indivi-duals engage in later in life.

Our study also offers unique insight about the role of the relationshipbetween adolescents and their employers and how family ties at work influ-ence adolescent career development � something that has not been pre-viously investigated in the literature. While we did not theorize about theinfluence of working in family business compared to working in nonfamilybusiness, we empirically explored whether the experience of working for afamily member had a unique influence that goes above and beyond theimpact of general work experience on professional outcomes. We observedthose adolescents who worked in their family business consistently foundbetter fitting jobs later in life and had better career networks. These find-ings suggest the work experience and career network available to adoles-cents in their family business is different from what their peers experiencein working for strangers. These results open interesting avenues of futureresearch both in family studies and in family business studies to furtherprobe into the difference across not only the type of work adolescentsengage in but also the nature of relationship between the adolescent andthe employers.

This study also contributes to the organizational studies’ literature byshedding light on the role of early life experiences on future professionaloutcomes. It is through early employment that adolescents develop a dee-per sense about the work and business world, acquire sharper business acu-men, and hone more concrete technical skills, all of which enhance theirfuture career prospects. The attained human and social capital becometheir competitive advantage over their peers and prepare them to betterselect jobs with higher income and better organizational fit. We hope thefindings of this study encourage organizational scholars to more deeplyconsider incorporating individuals’ early life experiences in later life workoutcomes.

Finally, our findings open many avenues of future research in exploringthe unique impact of work on future outcomes. What is the role of selec-tion in shaping and differentiating adolescent work experience? Specifically,


what factors influence adolescents’ decision to work, type of work,intensity, and seasonality of work? If these factors propagate beyondadolescence to adulthood, how can individuals escape them? Lastly, thereremains much to be done with respect to delineating the mechanisms whichimpact adolescent development in both the short term and the long term. Isit the time management skills, immersing in work, assuming responsibil-ities, conducting complex tasks, gaining deeper understanding of the careerworld, or developing social ties that is responsible for the unique effect ofwork on professional outcomes? Hopefully, these research questions will beprofitably addressed in future studies.

NOTE

1. Please note for the presentation simplicity our tables do not include thesemodels, but they are available from the first author upon request.

REFERENCES

Adler, P. S., & Kwon, S.-W. (2002). Social capital: Prospects for a new concept. Academy of

Management Review, 27(1), 17�40.

Baland, J. M., & Robinson, J. A. (2000). Is child labor inefficient? Journal of Political

Economy, 108(4), 663�679.

Basu, K., & Van, P. H. (1998). The economics of child labor. The American Economic Review,

88(3), 412�427.

Becker, G. S. (1962). Investment in human capital: A theoretical analysis. Journal of Political

Economy, 70(5), 9�49.

Becker, G. S. (1964). Human capital. New York: National Bureau of Economic Research.

Bhalotra, S. (2007). Is child work necessary? Oxford Bulletin of Economics & Statistics, 69(1),

29�55.

Blanchard, C. A., & Lichtenberg, J. W. (2003). Compromise in career decision making: A test

of Gottfredson’s theory. Journal of Vocational Behavior, 62(2), 250�271.

Brockhaus, R. H. (2004). Family business succession: Suggestions for future research. Family

Business Review, 17(2), 165�177.

Brown, D. W., & Konrad, A. M. (2001). Granovetter was right: The importance of weak ties

to a contemporary job search. Group & Organization Management, 26(4), 434�462.

Buchmann, C. (2000). Family structure, parental perceptions, and child labor in Kenya: What

factors determine who is enrolled in school? Social Forces, 78(4), 1349�1378.

Burdett, K. (1978). A theory of employee job search and quit rates. The American Economic

Review, 68(1), 212�220.

de Regt, A. (2004). Children in the 20th-century family economy: From co-providers to consu-

mers. The History of the Family, 9(4), 371�384.


Doepke, M., & Zilibotti, F. (2005). The macroeconomics of child labor regulation. American

Economic Review, 95(5), 1492�1524.

Eckrich, C. J., & Loughead, T. A. (1996). Effects of family business membership and psycholo-

gical separation on the career development of late adolescents. Family Business Review,

9(4), 369�386.

Gomez-Mejia, L. R., Cruz, C., Berrone, P., & De Castro, J. (2011). The bind that ties:

Socioemotional wealth preservation in family firms. The Academy of Management

Annals, 5(1), 653�707.

Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78(6),

1360�1380.

Granovetter, M. (1974). Getting a job: A study of contacts and careers. Cambridge, MA:

Harvard University Press.

Greenberger, E., & Steinberg, L. (1986). When teenagers work. New York: Basic Books.

Hall, S. G. (1904). Adolescence: Its psychology and its relations to physiology, anthropology,

sociology, sex, crime, religion, and education. New York: D. Appleton.

Hartung, P. J., Porfeli, E. J., & Vondracek, F. W. (2005). Child vocational development: A

review and reconsideration. Journal of Vocational Behavior, 66(3), 385�419.

Hazan, M., & Berdugo, B. (2002). Child labour, fertility, and economic growth. The Economic

Journal, 112(482), 810�828.

Hindman, H. D. (2009). The world of child labor: An historical and regional survey. New York:

M. E. Sharpe.

Hirschman, C., & Voloshin, I. (2007). The structure of teenage employment: Social back-

ground and the jobs held by high school seniors. Research in Social Stratification and

Mobility, 25(3), 189�203.

Hoffman, B. J., & Woehr, D. J. (2006). A quantitative review of the relationship between

person�organization fit and behavioral outcomes. Journal of Vocational Behavior,

68(3), 389�399.

Horrell, S., & Humphries, J. (1995). “The exploitation of little children”: Child labor and the

family economy in the industrial revolution. Explorations in Economic History, 32(4),

485�516.

Jensen, P., & Nielsen, H. S. (1997). Child labour or school attendance? Evidence from

Zambia. Journal of Population Economics, 10(4), 407�424.

Kett, J. F. (1971). Adolescence and youth in nineteenth-century america. The Journal of

Interdisciplinary History, 2(2), 283�298.

Lee, J. C., & Staff, J. (2007). When work matters: The varying impact of work intensity on

high school dropout. Sociology of Education, 80(2), 158�178.

Lerner, R. M., & Steinberg, L. (2004). The scientific study of adolescent development,

pp. 429�450. Hoboken, New Jersey: Wiley.

Mahoney, J. (2000). Path dependence in historical sociology. Theory and Society, 29(4),

507�548.

Mailath, G. J., & Postlewaite, A. (1990). Workers versus firms: Bargaining over a firm’s value.

The Review of Economic Studies, 57(3), 369�380.

Majumdar, M. (2001). Child labour as a human security problem: Evidence from India.

Oxford Development Studies, 29(3), 279�304.

Manning, W. G. (1998). The logged dependent variable, heteroscedasticity, and the retransfor-

mation problem. Journal of Health Economics, 17(3), 283�295.

Marsh, H. W. (1991). Employment during high school: Character building or a subversion of

academic goals? Sociology of Education, 64(3), 172�189.


Mortimer, J. T. (2003). Working and growing up in America. London: Harvard University

Press.

Mortimer, J. T. (2010). The benefits and risks of adolescent employment. Prevention

Researcher, 17(2), 8�11.

Pabilonia, S. W. (2001). Evidence on youth employment, earnings, and parental transfers in

the national longitudinal survey of youth 1997. Journal of Human Resources, 36(4),

795�822.

Packard, B. W.-L., & Nguyen, D. (2003). Science career-related possible selves of adolescent

girls: A longitudinal study. Journal of Career Development, 29(4), 251�263.

Patrinos, H. A., & Psacharopoulos, G. (1997). Family size, schooling and child labor in Peru:

An empirical analysis. Journal of Population Economics, 10(4), 387�405.

Petersen, T., Saporta, I., & Seidel, M.-D. L. (2000). Offering a job: Meritocracy and social net-

works. American Journal of Sociology, 106(3), 763�816.

Porfeli, E. J. (2007). Work values system development during adolescence. Journal of

Vocational Behavior, 70(1), 42�60.

Psacharopoulos, G. (1997). Child labor versus educational attainment some evidence from

Latin America. Journal of Population Economics, 10(4), 377�386.

Ranjan, P. (2001). Credit constraints and the phenomenon of child labor. Journal of

Development Economics, 64(1), 81�102.

Ray, R. (2000). Child labor, child schooling, and their interaction with adult labor:

Empirical evidence for Peru and Pakistan. The World Bank Economic Review, 14(2),

347�367.

Rosenbaum, J. E. (2001). Beyond college for all: Career paths for the forgotten half. New York:

Russell Sage Foundation.

Ruhm, C. J. (1997). Is high school employment consumption or investment? Journal of Labor

Economics, 15(4), 735�776.

Seidel, M.-D. L., Polzer, J. T., & Stewart, K. J. (2000). Friends in high places: The effects of

social networks on discrimination in salary negotiations. Administrative Science

Quarterly, 45(1), 1�24.

Staff, J., Messersmith, E. E., & Schulenberg, J. E. (2009). Adolescents and the world of work.

In R. Lerner & L. Steinberg (Eds.), Handbook of adolescent psychology (3rd ed.),

New York: Wiley.

Staff, J., & Mortimer, J. T. (2008). Social class background and the school-to-work transition.

New Directions for Child & Adolescent Development, 2008(119), 55�69.

Staff, J., Mortimer, J. T., & Uggen, C. (2004). Work and leisure in adolescence. In

R. M. Lerner & L. Steinberg (Eds.), Handbook of adolescent psychology. New Jersey:

Wiley.

Statistics Canada. (2010). Observed and projected percentage of youth, working age popula-

tion, and seniors, Canada, 1971 to 2056 (table). Healthy people, healthy places

(Vol. 2009, No. 1). Statistics Canada Catalogue No. 82-229-X. Retrieved from http://

www.statcan.gc.ca/pub/82-229-x/2009001/demo/dep-eng.htm. Accessed on August 20,

2012.

Steinberg, L. D., Grennberger, E., Ruggiero, M., Garduque, L., & Vaux, A. (1982). Effects of

working on adolescent development. Developmental Psychology, 18(3), 385�395.

Stevens, C. J., Puchtell, L. A., Ryu, S., & Mortimer, J. T. (1992). Adolescent work

and boys’ and girls’ orientations to the future. The Sociological Quarterly, 33(2),

153�169.


http://www.statcan.gc.ca/pub/82-229-x/2009001/demo/dep-eng.htm

http://www.statcan.gc.ca/pub/82-229-x/2009001/demo/dep-eng.htm

Verquer, M. L., Beehr, T. A., & Wagner, S. H. (2003). A meta-analysis of relations between

person�organization fit and work attitudes. Journal of Vocational Behavior, 63(3),

473�489.

Walters, P. B., & Briggs, C. M. (1993). The family economy, child labor, and schooling:

Evidence from the early twentieth-century south. American Sociological Review, 58(2),

163�181.

Zelizer, V. A. (1985). Pricing the priceless child: The changing social value of children.

New York: Basic Books.


beneficial “child labor”: the impact of adolescent work … · beneficial “child labor”:...

Documents