Generations of American kids have gotten an early sense of adulthood through a 150-year-old board game. Players begin on the opening square of The Game of Life with the same modest endowment. Whether you ultimately retire a millionaire or bankrupt depends on a few key decisions—and somewhat more on lucky or unlucky spins of the wheel.
As we march through similar milestones, real life gives us more power than a plastic spinner to alter our destinies. But of the countless decisions and forks in the road, which matter more? How will a particular move today affect our choices tomorrow? Real-life players also don’t start with an equal hand; economic research regularly reinforces how birth factors, such as parental income, race, place, and gender, connect to lifetime outcomes. If we want to balance the odds for all players, how could we update the rules or direct our public investments?
Taking in the Minneapolis Fed’s recent Education, Labor Market Outcomes, and Inequality conference stirred up my own Milton Bradley nostalgia. From various angles, the presented papers explore how lifetime outcomes hang on decisions we make—or decisions made for us—before we start our first day on the job. Here is a taste of what these economists are pursuing, Game of Life–style.
Fall in with the wrong crowd. Lose one turn.
We all know kids can be affected by the company they keep. University of Pennsylvania economist Francesco Agostinelli works to actually quantify “the influence of bad apples” and the subsequent responses of parents (“It Takes a Village: The Economics of Parenting with Neighborhood and Peer Effects,” with Matthias Doepke, Giuseppe Sorrenti, and Fabrizio Zilibotti). They make use of rich survey data in which U.S. high school students identify their close friends (who also took the survey) and state whether their parents dictate whom they hang out with. With student test scores, grades, and detailed family demographics, the economists have the raw materials to decipher the interactions of peers, parents, and outcomes.
Friendships with lower-performing peers appear to be a drag on students’ performance. But that’s just the start of the story, as “authoritarian” parents clamp down.
The data confirm that friendships with lower-performing peers appear to be a drag on students’ performance. But that’s just the start of the story. In the economists’ model, parents can respond by investing more time and energy in their kids, becoming stricter in managing friendships, or both. More “authoritarian” parenting reduces negative peer effects on performance, albeit with a trade-off: “Meddling in the choice of friends puts a strain on the parent-child relationship,” they write, “making the child less receptive to other parental interventions.”
Translation: Teens rebel. The net effect of the trade-off depends on how detrimental the peer environment is.
The insights from their model and data lead to a practical policy experiment: How would things play out if students are relocated—say, through bussing—from one neighborhood school (with one set of peers) to another? Per their model, parents of higher-performing children will turn more authoritarian, perceiving a higher risk from friends and controlling whom their kids mix with. “The parents are trying to undo the policy,” said Agostinelli. While bussing was supposed to encourage different kids to mix, “parents exacerbate segregation within the school through their behavior.”
Decide you’re not a “math person.” Go back two spaces.
During our education, we sort ourselves—and are sorted by others—onto paths that affect everything after. Work in progress by Minneapolis Fed Senior Research Economist Anusha Nath tracks five recent cohorts of Minnesota students through high school, college, and three years into the workforce—by which point distinct gender and racial earnings gaps have emerged (“Racial and Gender Differences in School-College-Career Paths,” with Katarína Borovičková and Alessia Leibert). The economists explore how students are sorted at critical “nodes” along the way. They measure which of these decisions matter more than others, and for whom.
Controlling for abilities, they find female and minority students sort onto paths with lower average earnings. Among the other initial findings:
- The smartest female and Black high school students are less likely than similarly skilled male or White students to take Advanced Placement courses in the STEM fields (science, technology, engineering, and mathematics).
- Black students who do follow an AP track in high school are more likely than White students on an AP track to choose a four-year college.
- At the college level, women are less likely to choose degrees, such as engineering, that lead to higher average earnings.
Nath and her co-authors calculate that sorting onto one path or another (as opposed to differences between students on the same path) accounts for 43 percent of the gender earnings gap and 25 percent of the Black-White gap. “Because of the difference in sorting patterns, policy changes that try to encourage minorities and women to make different decisions have the potential to be more effective in reducing gender gaps than race gaps,” said Nath. Earlier intervention also matters, but for these Minnesota students, sorting into college majors seems particularly important.
During our education, we sort ourselves—and are sorted by others—onto paths that affect everything after. Female and minority students appear to sort into paths with lower earnings.
High schoolers are aggressively sorted in Sweden, where they are explicitly placed into vocational, academic, or STEM paths from ninth grade and directed into colleges and college-major tracks through a centralized, national application process. University of Chicago economist Juanna Schrøter Joensen uses this unique environment (and test results from a required male military service exam) to understand how initial skills and early high school sorting amplify the later effects of college major and career (“Complementarities in High School and College Investments,” with John Eric Humphries and Gregory Veramendi).
The economists find such “dynamic complementarities” are strong, including in STEM fields. Early exposure and skill development in math and science supercharge the gains that students experience if they pursue related careers later.
One implication: There could be large returns to efforts long before college to divert promising students down a STEM path when they might have chosen (or been steered) otherwise. By the same token, efforts at the college level to encourage students to pursue STEM majors and careers could come too late to capture these complementarities. Investing early is not only a better way to push students into STEM but also to set them up for success in those fields.
Apply to a top 10 college, just in case … and get in! Leap ahead four spaces.
St. Louis Fed Economic Policy Advisor Oksana Leukhina and co-authors are working to solve what she calls the “undermatch puzzle.”
“A lot of high-ability students go to two-year schools and low-quality four-year schools,” Leukhina said. “We want to understand why.” The economists find that high-ability U.S. students get a 10 percent earnings bump from attending a second-tier college over a third-tier school, and a further 20 percent bump from going to one of the best colleges. Yet their data show only 9 percent of top-ability, low-income students go to top-tier schools. (“Sorting of Students into Colleges: Inefficiencies and Policy Implications,” with Lutz Hendricks and Tatyana Koreshkova).
High-ability students get a big earnings bump from going to a higher-ranked college. Yet only 9 percent of top-ability, low-income students go to top-tier schools.
While affordability plays some role, the economists’ model finds a larger role for information and motivation: In other words, getting low-income students to apply in the first place. The economists find that compared with providing money to low-income students and families, providing better information about school quality has a bigger impact and a moderate boost in their admissions rankings.
Lower-ability, higher-income students lose out in this scenario, as they are bumped to lower-ranked schools. But abilities and colleges are now better matched, taking advantage of the complementarities between the two. “With stronger, more meritocratic student sorting, we would wind up producing more human capital on aggregate,” Leukhina said.
Major in “underwater basket weaving.” Skip every other turn for the rest of the game.
The research on Minnesota students from Nath, Borovičková, and Leibert (discussed above) finds that while high school sorting matters, the most important “decision node” appears to be the choice of college major. The economists look at earnings three years after graduation.
The University of Michigan’s Kevin Stange looks further out, tracing students in the 10 most popular major categories roughly 16 years into the workforce. (“The Returns to College Major Choice: Average and Distributional Effects, Career Trajectories, and Earnings Variability,” with Rodney Andrews, Scott Imberman, and Michael Lovenheim). Using linked education and earnings data from Texas, the economists get a multidimensional view, uncovering how income grows as workers gain experience and how earnings are distributed for workers who had the same major.
Recently out of school, engineering majors make more than $22,000 more annually than liberal arts majors. Ten years later, the gap has grown to more than $32,000.
The nature of the patterns might not surprise, but the changes over time are striking to see. Liberal arts majors have the lowest earnings at all points and serve as the baseline against which other majors are compared. Recently out of school, engineering majors make more than $22,000 more annually than liberal arts majors (in 2016 dollars; importantly, these findings control for standardized test scores, demographics, and other factors). Ten years later, the gap has grown to more than $32,000. Business and economics majors make a little less than engineers but see similar wage growth. The biggest leaps come for biology and health majors—the effect of doctors who enter the workforce late and earn less as medical residents, but make rapid gains when they enter full practice.
Communications and social science majors, by contrast, do not make much more than liberal arts majors to begin with, and this small premium actually shrinks over time. A decade into the job market, the average communications major makes about $4,000 more annually than liberal arts; the average social science major just $2,600 more.
The within-major distributions are similarly eye-opening; majors vary significantly in how concentrated the earnings are. Earnings for business and economics majors are particularly concentrated among the highest earners. Earnings for engineers, by contrast, rise more evenly along the earnings range. The slightly higher average earnings for social science majors can be attributed largely to the very highest earners in this group; most social science majors actually earn less than liberal arts majors.
The economists also find majors with high returns have more stable earnings over time. In general, said Stange, the results show that choosing the right college major matters more to your earnings than the decision to go to college at all. The research is illuminating at a time “when there is a lot of concern about the value of college, writ large,” Stange said, and “many states are scaling back their investments in certain programs based on what investments are going to pay off in the labor market.”
The insights could inform policies to help with affordability. Consider two common policy ideas: capping tuition and student loan forgiveness. Stange says loan forgiveness looks more appealing, insofar as it helps insure against the wide variation in lifetime outcomes within and across majors.
College is harder than you thought! Drop out and work instead. Collect $200 (but skip two turns).
Many parents could attest that while teenagers think they know everything, they don’t. Arizona State University economist Esteban Aucejo explores the idea that students approach the critical college decision with imperfect information about their own skills, abilities, and prospects.
A high percentage of college students will drop out. Some of this is surely a matter of financial hardship. But others learn college isn’t for them—or at least that it is not worth it for them to stay.
Roughly 6 in 10 students who enter two- or four-year colleges in the U.S. fail to complete a degree within six years. Some of this is surely a matter of financial hardship. But others learn college isn’t for them—or at least that it is not worth it for them to stay. “I go to college and perhaps I receive some information,” said Aucejo. “I learn a little bit about my abilities, and I decide to drop out and go into the labor market.” Some students work while going to school, giving them further information about the value of college versus the alternative.
Aucejo and his co-authors construct an economic model that incorporates these realities, plus another important information gap: High school graduates who incorrectly think college is not for them, and don’t go. (“College Attrition and the Dynamics of Information Revelation,” with Peter Arcidiacono, Arnaud Maurel, and Tyler Ransom).
People in the economic model learn and adapt each year to new information. Go to college? Two-year or four-year? Work part time? Drop out? Science or nonscience major? Go to grad school? These information “frictions” are inefficient. When the economists remove them from the simulation—giving everyone full information about their abilities and prospects—dropping out plummets and the college graduation rate increases by nearly 40 percent. Students are better matched with college options according to their ability.
“Providing full information closes the gap in college completion between high-income and low-income individuals particularly because many low-income individuals thought college was not a good match for them,” said Aucejo. The outcome has complex implications for inequality. More low-income students go to college and experience increased earnings. However, better alignment between ability and higher education also widens the wage gap substantially between college graduates and noncollege workers.
Assert yourself: Ask your boss for a raise! Collect $5,000 and spin again.
After school, we enter the labor market with our credentials, experience, skills, and knowledge. We also tote along our personalities. Our success undoubtedly depends on how we act once we’re out in the world. But such “noncognitive” or emotional skills can be a squishy subject for economists to study, often left among the unobserved factors in statistical models of earnings and achievement.
Men who are more conscientious experience higher wages, faster wage growth, and quicker job-finding. Conscientious women, on the other hand, suffer worse outcomes in the workplace across the board.
As it happens, European labor researchers had the notion to administer personality tests and track the progress of thousands of Germans who became unemployed in 2007 and 2008. The workers assessed themselves for the “big five” personality traits established in the psychology literature: conscientiousness, agreeableness, openness to experience, extraversion, and neuroticism. New York University economist Christopher Flinn uses this novel data to unpack the role of personality for job-seeking and pay, with a focus on the gap between women and men (“Labor Market Returns to Personality: A Job Search Approach to Understanding Gender Gaps,” with Petra Todd and Weilong Zhang).
In the raw data, women skew higher than men in traits like agreeableness and conscientiousness, but score themselves as more neurotic. The economists find that for both genders, lower neuroticism is associated with higher wages and more stable employment. With other traits, however, the effects of personality for men and women look different in scale—or even push in opposite directions. Men who are more conscientious experience higher wages, faster wage growth, and quicker job-finding. Conscientious women, on the other hand, suffer worse outcomes in the workplace across the board.
In a simulation where women are provided the same bargaining position as men, the gender wage gap disappears, leading the authors to conclude that “gender differences in the labor market valuation of personality traits rather than in trait levels are the primary determinants of gender wage gap.” Rephrased: The problem is not just that men and women are different. It’s that the differences in personalities affect their bargaining positions—the way they are perceived and treated from across the table.
Get some help with child care. Collect your paycheck and advance three spaces.
Game of Life players will recall the blue and pink pegs you collect as your plastic car traverses the board. In real life, of course, these expensive bundles of joy have a way of changing and complicating our working lives.
The cost and availability of child care affects whether parents work, how much, and the paths of their careers. As the Minneapolis Fed and others have established, investments in care and education in early childhood bring lifetime benefits for the child and society. As lawmakers contemplated new federal child care subsidies in 2022 (part of the ultimately abandoned Build Back Better Act), University of Wisconsin economist Matthew Wiswall and a large group of collaborators developed a framework to assess the impact of varying subsidies on costs and the labor force (see “An Equilibrium Model of the Impact of Increased Public Investment in Early Childhood Education” for a prior version of this work).
Researchers find a broad child care subsidy would shift millions of mothers from nonwork or part-time work into full-time work.
Build Back Better contemplated new child care subsidies that fully covered care for low-income households. The subsidies declined as family income increased and phased out for households making 250 percent of median income. The researchers find this broad expansion would shift millions of mothers from nonwork or part-time work into full-time work; a narrower policy would have more modest effects. Although the sticker price of care rises with a new subsidy, most households experience a decline in net costs for child care. Child care teacher wages rise. Many families switch to formal, center-based child care, with big declines in informal, unlicensed care provided by relatives and friends.
While the group reported results in their earlier paper, Wiswall cautioned against quoting numbers while their work is evolving (and we don’t do so here). Although the Build Back Better child care plan died amid congressional negotiations, Wiswall said the episode led him and his co-authors to create new tools for evaluating specific proposals in time to inform political decisions. “This is the kind of work we can use to do better, faster policy analysis than what I think we’re doing now,” he said.