Current Projects

The Workforce Science Project is currently sponsoring the following projects:

Paths to Economic Opportunity

For the past two decades, the United States has been plagued by growing economic inequality and declining social mobility, adding to the challenges faced by marginalized groups, including people of color and women. 

Paths to Economic Opportunity (PEO) is developing advanced data science tools to construct educational and career paths that provide useful new perspectives on important policy questions of economic opportunity and diversity.

PEO uses advanced machine learning to dramatically expand our understanding of why some entry level jobs lead nowhere while others are the first rung on about the ladder of success. We’re expanding this analysis to understand better the career paths of underrepresented groups.

Tools like job ladders illuminate many high-impact questions about economic opportunity. PEO’s results can help employers design human capital practices that promote diversity and economic mobility. For more information on our first results on college selectivity, see CLBE Workforce Science Report: When Does Selective College Matter? Making Your School and Major Fit. More information on Paths to Economics Opportunity.

How College Majors and Selectivity Affect Economic Opportunity 

Every year, millions of high school students decide whether to attend college, which college to attend, and what field to major in. A key consideration for many students is the role that college plays in their long-run earnings prospects.

Where should a high school graduate go to college? What should they major in? Admission to a selective college is viewed by some as a prerequisite for long-term economic success. Others dismiss selectivity as irrelevant compared with choice of major. The truth is, of course, much more complex. In this research, we find that the majors that provide a high return at less selective schools are very different from the high-return majors at more selective schools.

Our results can help employers design cost-effective tuition programs for economic mobility and reassess their recruiting practices to ensure that they are making the best use of the pipeline from less selective schools 

More information on this research, funded by Microsoft, can be found in the Executive Summary and the full report: When Does Selective College Matter? Making Your School and Your Major Fit. The full research analysis can be found at  https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3946507

Firm-Level Evidence of the Effect of Ban-the-Box Policies 

Employers regard a criminal record as a negative signal of future job performance. In response, numerous jurisdictions have enacted Ban-the-Box (BTB) laws that prohibit employers from making inquiries about an applicant’s criminal record until a conditional offer of employment has been made and that sometimes restrict the grounds on which employers can withdraw a conditional offer.

Previous research on the effectiveness of BTB policies has been hampered by the lack of internal firm data. In this research we use data obtained from an employer who had voluntarily implemented a BTB-like policy. Our results confirm the negative effect of even a minor criminal record on the likelihood that an individual receives both a conditional and a final offer of employment. We also find that the adoption of BTB rules have no or even a small negative effect on the rate at which individuals with criminal records receive conditional or final employment offers. Our examination of job performance finds that the overall promotion rate of individuals with records is the same as that of other employees. Individuals with criminal records have higher rates of involuntary separation than other employees, though individuals with more serious offenses do have lower voluntary separation rates. BTB policies did not seem to appreciably affect promotion, tenure or separation.

The Causes and Consequences of Distinctively Elite Names 

Many first names are associated with a particular socioeconomic class. A small industry provides baby-naming advice based on the premise that this relationship is causal, so that Finn and Isla will fare better in life than their siblings Brandi and Jaxon. Relatively little research illuminates whether this parental faith in the power of names is warranted. In this paper we examine the possible consequences of socioeconomically distinctive names. Using California birth certificate data, we create a measure of the socioeconomic class signaled by a name, and then match the birth certificates of women at the time of their own birth to the birth certificates of their children, using the data from the child’s birth certificate to construct outcome measures for the mother. Even after controlling for race and socioeconomic factors, we find a strong relationship between economic outcomes and the racial and socioeconomic signals of names, and especially for the interaction between these two. However, this relationship does not appear to be causal. When we control for the effect of a child’s family of origin, the significance of the name variables disappears.

Must De-biased Hiring be Anonymous? An Experimental Test of Job Simulations

A powerful body of empirical research raises concerns that current hiring practices facilitate discrimination. As a result of this research, some firms now review resumes on an anonymized (“blind”) basis. However, later steps in the hiring process cannot be anonymized. Job simulations are often proposed as a less discriminatory approach to non-anonymous hiring, but the empirical evidence supporting this is sparse. In this paper, we extend the experimental hiring literature to examine whether job simulations can produce non-discriminatory outcomes. Our study has three novel characteristics. First, our design reduces insofar as possible the element of prediction, and thus the possibility of statistical discrimination. Second, our design gives our subjects a strong incentive to make a non-discriminatory decision. Third, unlike most studies we control for the possibility of discrimination based on socioeconomic status as well as race.

Our results suggest that job simulations provide a potentially promising avenue for developing less biased approaches to hiring. The overall results of the study suggest no evidence of racial discrimination in the subject’s evaluation of work quality, although we find some evidence at low significance levels that discrimination may have occurred among participants who voted for Donald Trump in 2016. These results indicate that non-anonymized job simulations hold some promise for reducing discrimination in personnel selection.

What Privacy Rights Do Employees Value?

As with any use of Big Data, talent analytics raises the challenge of protecting individual privacy rights. With diminishing practical constraints on data collection and analysis, privacy advocates have increasingly turned to legislative and regulatory restrictions. This push for governmental intervention is based on guesses and assumptions rather than an empirical understanding of what data employers use, how employers use that data, or even of what monitoring most concerns employees. The current project attempts to shed light on these issues through a series of surveys of employee privacy attitudes. Using a list of current and proposed monitoring practices gleaned from interviews with talent analytics professionals, these surveys seek to determine what types of monitoring and data usage raise the greatest degree of privacy concerns.

The Complex Doctrine of At-Will Employment

The doctrine of at-will employment has been surrounded by much controversy. Some believe that the elimination of at-will employment is essential to providing employees with secure employment while others argue that eliminating the at-will doctrine will reduce overall employment.

A significant economic literature examines the effect of the at-will doctrine on a variety of outcomes such as employment levels. Variations in state law and changes in the law over time are used to untangle the causal effect of laws themselves from other factors that might affect the measures of interest.  This literature relies on coding the at-will doctrine in simple three criteria Yes/No format that ignores the complex differences among state laws. This study first develops a more complex and more nuanced coding that can be used for statistical analysis. Using this new coding, it exploits variations in state laws to investigate the extent to which the employment laws effect economic activity.

A second study examines the history of the at-will doctrine, tracing its origins back to the early 20^th century and finding multiple causes.