Domestic Outsourcing of Labor Services in the U.S.: 1996-2015
David Dorn, University of Zurich, CEPR, IZA, & CESifo
Johannes F. Schmieder, Boston University, NBER, IZA & CESifo
James R. Spletzer, U.S. Census Bureau (retired)
Lee C. Tucker, U.S. Census Bureau
Abstract
A growing amount of anecdotal and qualitative evidence suggests that outsourcing causes a deterioration of many aspects of job quality, yet quantitative evidence on the prevalence and effects of outsourcing in the United States is scarce. In this paper, we use the Census Bureau’s Longitudinal Employer Household Dynamics (LEHD) data to provide the first estimates of the effects of outsourcing on a variety of job quality measures in the U.S. that are credibly identified through the use of matched employer-employee information. In pursuit of this goal, we also develop an improved methodology for identifying outsourcing events in the data, generalizing the strategy of Goldschmidt and Schmieder (2017) to identify unexpected mass flows of workers across source-destination pairs using a random-flows model. And, we construct new metrics by which we can identify a broader set of worker destination industries that are likely to be associated with outsourcing. These innovations allow us to identify a substantially broader set of outsourcing events than prior research. We then use nearest-neighbor matching to a suitable counterfactual of workers who are not outsourced in order to estimate the causal impacts of outsourcing on outsourced workers. We find that on average, workers who are outsourced earn small positive earnings premia relative to their matched controls after outsourcing, and that these premia are persistent for up to ten years afterward. However, industry-level heterogeneity is important; workers in Food and Cleaning as well as Employment Services see substantially lower earnings after outsourcing. Although correspond to changes in the firm-level wage patterns of outsourced workers’ employing firms, the disproportionately negative effects of outsourcing on Food and Cleaning and Employment Services workers is not explained by the wage premia of their source firms alone. An analysis of linked responses to the American Community Survey also suggests that this industry-level heterogeneity in earnings impacts is unlikely to be driven by differences in hours or insurance benefits.
The Effect of Outsourcing on Remaining Workers, Rent Distribution, and Inequality
Daniel Mark Deibler, Federal Trade Commission∗
Abstract
Firms can decide whether to produce goods and services in-house or purchase them from the market. Increas- ingly, they are purchasing from the market—using outsourced labor. Low-wage workers’ wages decline when they are outsourced, which increases inequality, but little is known about how outsourcing affects remaining workers. If firms are rent sharing, outsourcing might increase remaining workers’ earnings because there are more rents or fewer workers to share them with. This paper measures the impact of occupational layoff (OL) outsourcing on the earnings and separations of workers who remain at those firms. Using employer-employee data based on German social security records in a matched difference-in-differences design, outsourcing increases remaining workers’ long-run earnings by 7.1% in a sample of 210 OL outsourcing events. The effect is larger for lower-paid workers in the firm, and increased earnings only occurs in firms with collective bargaining agreements (CBAs). Outsourcing decreases the year-over-year probability of remainers switching firms by approximately 8 percentage points. These results are consistent with a wage-setting model where CBAs give workers leverage when firms outsource. Analyzing the impact of outsourcing on overall wage inequality using Recentered Influence Functions, increasing the share of workers part of an outsourcing event by 1 percentage point increases the top of the earnings distribution by approximately 0.2%, and the variance of log wages by 0.5%. Remainers are relatively high-wage, and outsourcing increases their earnings. By not accounting for the effect on remainers, prior analyses underesti- mate the impact of outsourcing on earnings inequality.
∗ The views expressed in this article are those of the author and do not necessarily reflect those of the Federal Trade Commission or any individual Commissioner. Many thanks to Bentley MacLeod, Suresh Naidu, and Miguel Urquiola for their helpful advice. Many thanks to Tanya Avilova, Iain Bamford, Ihsaan Bassier, Michael Best, Sandra Black, Sydnee Caldwell, Cynthia Doniger, Andrew Garin, Len Goff, Lucas Husted, Simon Jager, Matt Mazewski, David Rosenkranz, Johannes Schmieder, Gregor Schubert, Anna Stansbury, Pablo Warnes, Ron Yang, and the Columbia Applied Micro Seminar.
The Effects of Outsourcing on Workers when
Employment Protection is High: Evidence from Italy
Diego Daruich, USC Marshall; Martino Kuntze, Bank of Italy; Pascuel Plotkin, University of British Columbia; Raffaele Saggio, University of British Columbia and NBER
Abstract
This paper exploits a novel identifier of outsourcing events recorded within Italian social security data to shed new light on the consequences of outsourcing for workers. Since 2005, Italian employers are required to motivate the reason behind a job separation. Outsourcing, and in particular the practice of transferring a particular branch to an external firm, is listed among these reasons. The availability of this information permits us to complement the “flow” approach popularized by Goldschmidt and Schmieder (2017). In particular, we can study the consequences of outsourcing even for workers that are not outsourced in business service firms in the Cleaning-Food-Logistics-Security sectors. Moreover, our approach does not need to impose that the outsourced worker remains employed in the outsourced job. Both of these margins appear important. More than two-thirds of workers being outsourced are transferred to a non-business-service firm and thus would not be easily captured in a ow approach. Event-study estimates suggest that outsourcing has large and persistent negative consequences for workers’ earnings – a majority of which is explained by the extensive margin of employment. Although the average worker loses from outsourcing, this varies depending on their industry. For example, restaurant workers are among the most negatively affected while technology workers may even gain from outsourcing, suggesting that the motive behind outsourcing may matter for the effects on workers. We argue that in Italy a motive for firms to rely on outsourcing practices is that the latter can constitute a “loophole” to fire workers without paying large firing costs
Comparing Survey and Administrative Measures of Self-Employment Income: New Implications from the Health and Retirement Study
Joelle Abramowitz, University of Michigan
Abstract
This paper uses the 2004-2016 Health and Retirement Study (HRS) linked to administrative earnings records to examine differences in reported self-employment earnings across the two sources as well as compared to the Current Population Survey Annual Social and Economic Supplement (CPS). Correctly identifying work-related income in general and self-employment income in particular is important to understand the full picture of the labor market as well as to estimate levels and changes in wellbeing, inequality, and poverty. However, a challenge for exploring these questions is that different data sources lead to different estimates and trends. In particular, survey measures of self-employment often diverge from one another, and administrative records provide yet another source of discrepancy.
While administrative data represent the gold standard in measuring certain outcomes, like program participation, they may provide a less reliable measure for measuring self-employment and self-employment income. This is due to the fact that much of self-employment income, especially at smaller amounts, is reported to the IRS only at the discretion of filers (as with survey data). Tax filers may also have incentives to under-report self-employment income to tax authorities to reduce their tax burden or to over-report self-employment income to take advantage of tax incentives. Furthermore, less formal work arrangements may be shorter term and less stable, and as a result, may be less likely to be reported as taxable income and identified in administrative records. In addition, to examine effects of these work arrangements on wellbeing, administrative records lack information on many demographic, health, and family characteristics and outcomes of interest.
These considerations point to the need to better understand the nature of survey reports of self-employment income, how they relate to survey reports of wage and salary income, and how they compare to administrative records. The existing literature has found reporting of any self-employment income underreported as compared to administrative earnings records in one survey, the Current Population Survey Annual Social and Economic Supplement (CPS) (Abraham et al., 2021).
This paper adds to the literature by using another survey linked to administrative earnings records, the 2004-2016 HRS, to examine comparisons using a different source of survey data focusing specifically on older workers. The paper finds that, for same-aged individuals, the HRS identifies more self-employment earnings than administrative records, which identify more self-employment earnings than the CPS. Results suggests that while 29.4 percent of inconsistent HRS self-employment reports represent misreported wage earnings, more than double, 60.8 percent, represent true self-employment not captured in the administrative records, likely driven by strategic reporting of earnings (for example, to avoid taxation). Further results suggest these are disproportionately reported by respondents who consider themselves to be retired, likely reflecting more informal work that is not reported to tax authorities. These findings highlight (1) administrative data may provide less reliable measures on self-employment and self-employment earnings; (2) the capacity for survey data to identify self-employment otherwise overlooked in analyses using only administrative records; and (3) heterogeneity in reporting self-employment earnings across survey instruments.
The What and How of Measuring Electronic Platform Work
Anne E. Polivka, Bureau of Labor Statistics
Abstract
Recently, there has been a great deal of interest by researchers, policy makers and the general public in electronic platform workers – a specific form of contract work. However, despite this interest, measuring electronic platform workers has presented unique definitional and operational challenges. This paper lays out 8 dimensions of work that could be considered when constructing a measure of platform workers - 1) the location of workers and customers (i.e. in-person versus completely online), 2) the types of customers for whom the work is done (i.e. peer-to peer, peer-to business, business-to-business), 3) the type of activities to include based on the amount of labor services embodied in these activities, 4) platform companies’ control over workers and the assignment of work, 5) the time period during which workers engaged in platform activities, 6) the intensity with which workers engaged in platform work during a specified time period, 7) thresholds, if any, of the proportion of total personal income platform work earnings constitute, and 8) workers’ commitment to working (i.e. registering to work without doing any work, number of hours of worked, frequency of work, when last worked) and the availability of platform work to them.
After discussing the dimensions of work that could be included in a measure of platform workers, how various countries’ national statistical offices, private sector entities, and academic researchers have incorporated these dimensions into their measurement of platform workers is reviewed.
The paper concludes with a discussion of lessons learned from various attempts to measure platform work. In this discussion there is a particular emphasis on BLS’s first attempt to measure platform workers in the 2017 Contingent Worker Supplement and the subsequent design work done for the May 2023 Contingent Worker Supplement to the Current Population Survey. The latter includes a discussion of the results of cognitive testing of the questions and modifications that were made to proposed questions based on this testing.
Form-1099K for Gig Workers: Quantifying the 1099-K Gap and Impacts on Tax Compliance
Andrew Garin, University of Illinois Urbana-Champaign; Emilie Jackson, Michigan State University; Dmitri Koustas, University of Chicago
Abstract
One challenge for measuring platform gig work in tax data after 2016 is the so called “1099-K gap,” where many platform gig workers earning less than $20,000 no longer received an information return. We combine federal tax return data with returns from states that introduced lower reporting thresholds to study the effects of this change. Incorporating state information returns in this setting allow us to make two key contributions in this paper: we quantify the size of the 1099-K gap and analyze how receipt of a 1099-K affects tax compliance. First, we present a new framework using returns from states that have lower reporting thresholds to provide a more complete estimate of total platform work. In order to assess the size of any resulting reporting gap, we merge state 1099-K information returns from Massachusetts and Vermont that are subject to a lower reporting threshold to the federal tax return data. We use the information available from these two states to impute the national trend in platform work beyond 2016 assuming they would have evolved in parallel. Our imputation methodology suggests raw counts underestimate platform work by approximately 770,000 workers by 2018. Second, we evaluate to what extent receipt of a 1099-K form affects tax compliance. This is policy relevant as The American Rescue Plan Act of 2021 lowered the 1099-K reporting threshold to $600 with no minimum number of transactions but has yet to be implemented. To examine whether receiving an information return affects tax filing behavior, we use a border design to compare tax outcomes of platform gig workers in two states with state laws that closed the 1099-K gap, with neighboring states where the federal threshold was binding. We explore the spillovers of changes in 1099-K reporting to individual reporting of self-employment earnings on Form 1040 Schedules C and SE. We find that platform gig workers who received an information return reported on average $420 more in self-employment profits.
Contracting out Labor Market Dynamism
Andrea Atencio-De-Leon, University of Illinois at Urbana-Champaign
Abstract
This paper investigates how domestic outsourcing affects plant-level labor responses
to revenue productivity shocks and biases the measurement of aggregate job reallocations. I develop a methodology to transform reported expenses on temporary and leased workers into plant-level outsourced employment using comprehensive administrative data on the U.S. manufacturing sector. I show that plant-level outsourced employment is twice as responsive as payroll employment to revenue productivity growth deviations and adjusts more quickly. The evidence indicates that domestic outsourcing is an important margin of adjustment that plants use to modify their workforce while they learn about the permanency of the shock. These micro implications have significant macroeconomic measurement consequences. I show that the measured pace at which jobs reallocate across workplaces is underestimated. On average, every year, we omit the equivalent to 15% of payroll reallocations. The
extent of mismeasurement varies with the business cycle, falling in downturns and increasing in upturns. My findings suggest that the increasing use of labor market intermediaries accounts for a substantial portion of the measured decline in labor market dynamism, and further reflects structural adjustments in the choice set of firms when facing shocks.
Any opinions and conclusions expressed herein are those of the author and not those of the U.S. Census Bureau. The Census Bureau's Disclosure Review Board and Disclosure Avoidance Officers have reviewed this information product for unauthorized disclosure of confidential information and have approved the disclosure avoidance practices applied to this release. This research was performed at a Federal Statistical Research Data Center under FSRDC Project Number 1808. (CBDRB-FY22-P1808-R10049)
The Rise Of The Contract Workforce In U.S. Manufacturing And Its Implications For Worker Skills Measures1
Matthew Dey, Bureau of Labor Statistics ; Susan Houseman, Upjohn Institute for Employment Research
Abstract
Research and policy analysis on labor market issues often relies on employment measures that only capture employees of firms. Although prior studies have pointed to high and growing levels of contract labor in some segments of the economy, lack of data has hampered research into the size of this workforce and its broader implications for labor market analyses. This paper examines such implications in a study of U.S. manufacturers’ use of contract workers and its effects on measures of skills growth. Using a new research data set developed at the Bureau of Labor Statistics, we exploit granular information on establishments’ location and employment by occupation from 2000 to 2019 to impute staffing services workers in core manufacturing occupations to manufacturing industries and other sectors. We find that the share of contract workers in the manufacturing sector’s core occupations rose 45 percent from 6.9 percent in 2000 to 10.0 percent in 2019. This growth in contract use coincided with sharp declines in manufacturing employment, explaining 8 percent of the decline in manufacturing direct-hire employment in its core occupations. Notably, industries experiencing steeper overall employment declines outsourced a higher share of their core workforce on average. We estimate that the replacement of direct-hire workers with contract workers, who are concentrated in low-skilled occupations, explains 19 percent of the skills growth among workers in core occupations and 17 percent of the growth in skills among manufacturing workers in all occupations.
1 Any opinions and conclusions herein are those of the authors and do not necessarily reflect the views or policies of the U.S. Bureau of Labor Statistics.
Outsourcing, Occupationally Homogeneous Employers, and Wage Inequality in the United States
Elizabeth Weber Handwerker, U.S. Bureau of Labor Statistics
Abstract
This paper develops economy-wide measures of outsourcing in the United States, using the homogeneity of occupations by employer, as measured in the detailed microdata of the Occupational Employment and Wage Statistics Survey conducted by the Bureau of Labor Statistics. These measures distinguish between two types of outsourcing, which may have differing impacts on wage inequality. When businesses outsource work to avoid monitoring, hiring, or other costs for occupations in which they have less expertise, there will be less variety overall in the occupations they employ. However, when businesses outsource work to narrow the wage distribution of their employees, the variance of wages predicted from the particular set of occupations they employ will decrease. The impact of the changing distributions of occupations and of employer occupational homogeneity are compared with the effects of other changes in employer characteristics (industry, size, and location) on the distribution of wages.
The occupational homogeneity of employers is related to workers’ wages. Workers in more occupationally homogeneous establishments earn lower wages. This relationship holds even after controlling for workers’ own occupations and observable characteristics of their employers and is strongest for workers in occupations typically paid lower wages.
Employer occupational homogeneity of employers increased somewhat from 2004 to 2019, despite changes in the overall occupational composition of the US that would have led to no change in homogeneity (by one measure) or decreased homogeneity (by the other measure). In the early part of this period, homogeneity increased among workers in typically lower-wage occupations employed in smaller establishments. Over the period as a whole, the most consistent and unambiguous increases in homogeneity were for workers in typically high-wage occupations, regardless of their employers’ size.
Changes in the distribution of occupational homogeneity are related to the growth in private-sector wage inequality observed in the data from 20014 through 2016 (OEWS data show wage inequality declining after 2016). A substantial amount of growth in ln(wage) variance during this period can be attributed to the growing occupational homogeneity of establishments over this period. Both measures of employer homogeneity—one based on the distribution of occupations by wage levels, and the other a more functional measure of employer homogeneity that ignores wage differences among occupations—matter for growing wage inequality. Although wage inequality in these data fell after 2016, the wage convergence observed from 2016 to 2019 would have been even greater in the absence of these changes in occupational homogeneity. Changes in the occupational homogeneity of workplaces are a particularly important contributor to growing wage inequality among workers in different employers. Growing sorting and segregation by occupation of workers into different employers is an important part of wage inequality growth.
The Independent Contractor Workforce:
New Evidence on Its Size and Composition and Ways to Improve Its Measurement in Household Surveys
Katharine G. Abraham, University of Maryland; Brad Hershbein, Susan N. Houseman, and Beth Truesdale, W.E. Upjohn Institute for Employment Research
Abstract
Good data on the size and composition of the independent contractor workforce are elusive, with household survey and administrative tax data often disagreeing on levels and trends. We carried out a series of focus groups to learn how self-employed independent contractors speak about their work. Based on these findings, we designed and fielded a large-scale telephone survey to elicit more accurate and complete information on independent contractors, including those who may be coded incorrectly as employees in conventional household survey data and those who are independent contractors in a secondary work activity. We find that, upon probing, roughly one in 10 workers who initially reports working for an employer (and thus is coded as an employee) is in fact an independent contractor. Incorporating these miscoded workers into estimates nearly doubles the size of the independent contractor workforce, to about 15 percent of all workers. Young workers, less-educated workers, workers of color, multiple-job holders, and those with low hours are more likely to be miscoded. Taking these workers into account substantively changes the demographic profile of the independent contractor workforce. Consistent with case study accounts, we also find evidence of multilayer contracting not captured in prior surveys.
Characteristics of Gig Workers in the U.S.: Evidence from the Entrepreneurship in the Population Survey
Rachel Marie Brooks Atkins, St. John’s University; Quentin Brummet and Katie Johnson, NORC at the University of Chicago
Abstract
This paper presents new results from the Entrepreneurship in the Population (EPOP) Survey, a new nationally representative survey of entrepreneurship and gig work activities across the U.S. In addition to a series of questions on the pathways to entrepreneurship, the survey collects information from all respondents on current work activities and whether respondents are part of the “gig economy.” Importantly, in addition to asking about general involvement in gig work, respondents are also asked more specifics about the name of the gig work platform and whether the platform is an online app. Using responses, our analysis provides a variety of measures of gig work in the U.S. For example, using the broadest measure of gig work that asks respondents if they engage in work that uses a platform to coordinate payment, 18.91% of respondents report that they engage in gig work. Using a more restrictive definition that limits gig workers to only respondents using an online platform app, we estimate 7.12% of individuals are engaged in gig work. Regardless of definition, respondents report that flexibility and supplementing pay are important reasons for engaging in gig work. We conclude with discussing future directions for the research and other potential uses of EPOP data.
The Evolution of Gig Work, 2020-2021
Andrew Garin, University of Illinois Urbana-Champaign; Emilie Jackson, Michigan State University;
Dmitri Koustas, University of Chicago
Abstract
In this paper, we explore the evolution of 1099-reported gig work during the COVID-19 pandemic. Overall, we find 1099 gig work fell between 2019-2021. However, this masks dramatic growth in gig work mediated by online platforms (“platform gig work”). COVID was a time of both record entry and exit from platform gig work, with a substantial net increase in the number of platform gig workers of around 3 million (150% growth relative to 2019), with nearly all of the growth coming from delivery platforms. This growth is primarily part-time and for small dollar amounts; we also find that the demographic composition of platform workers shifted significantly during the pandemic, becoming younger and more female. An increase in younger workers is consistent with the nature of COVID risk, and the increase in the female share may be due to changing tasks away from transportation towards shopping and delivery. The number with earnings from creator/influencer platforms also increased, but is still small comparatively. We next investigate 1099-contract work outside of platform gig work and find substantial declines. This is true across all NAICS 2 industries outside of platform gig work, although self employment outside of 1099-contract work increased modestly. As a result, the overall 1099-workforce declined over this period, and platform gig work now represents a much larger share of the alternative workforce. We next explore mechanisms: We probe the role of expanded unemployment insurance benefits for exits, finding that these benefits resulted in many individuals who were self-employed in 2019 not reporting any self-employment income in 2020-2021. However, for the case of platform gig work, these disincentive effects are swamped by new entry. Entry into platform gig work from wage work was largest among workers in industries most affected by COVID, suggestive that this work helped provide extra income to workers facing the largest shocks. In addition to unemployment insurance availability, we hypothesize other gig work may have decreased due to negative demand shocks for self-employed services (such as personal services), and changes in workplace practices that increased flexibility in traditional wage jobs.
Redefining “Core Competencies”: Labor Market Intermediation in Outsourced Warehouses
Beth Gutelius and Nik Theodore, University of Illinois Chicago
Abstract
For nearly a half century, the questions of why and how firms navigate the “make-buy” decision have animated fields as varied as industry studies, labor relations, business management, and economic geography. Research on vertical disintegration has examined the makeup of inter-firm relationships, out of which arose an initial, non-mutually exclusive typology of outsourcing: capacity subcontracting, specialization subcontracting, and supplier subcontracting (Watanabe, 1971; Scott, 1983; Bernhardt et al., 2016). In addition, two central factors motivating the choice to contract out were identified: product market instability and access to technologies and economies of scale (Rubery and Wilkinson, 1981; Holmes, 1986). These intellectual advances were critical in making sense of the major changes underway in the organization of global production.
The idea of “core competencies” has more recently emerged as a significant factor explaining corporate decision-making processes, where any activity deemed outside of the central specializations of the firm is a possible candidate for outsourcing. Coupled with the intensifying focus on short-term profit-taking, corporate leaders have grown increasingly focused on shedding less profitable activities and shifting supply-chain risk. This, in turn, has led to high levels of lead firm influence over subcontracting markets and the cost-based competition that permeates them.
Despite long standing scholarly interest in the vertical disintegration of firms, the logistics activities included therein have remained largely outside the realm of analysis (Coe, 2021; Selviardis and Spring, 2007). Building on a growing body of literature focused on the outsourcing dynamics of supply-chain logistics (Aoyama et al., 2006; Rodrigue, 2012; Dittman and Vitasek, 2016), this paper examines the role of third-party logistics companies (3PLs) in providing competitive advantages to lead firms in the warehousing sector. More precisely, it argues that efforts to contain operational costs increasingly are focused on labor, which constitutes a substantial portion of the warehouse balance sheet, and that the ability to access and deploy low-cost labor is among the “core competencies” touted by many 3PLs in the warehousing sector.
Drawing on interviews with 118 warehouse operators and industry analysts, the paper seeks to contribute to the literature on outsourcing in two central ways. First, it demonstrates the role of lead firms in shaping the competitive markets in which 3PLs vie for contracts. Second, it analyzes the ways in which operators of subcontracted warehouses are reworking the social division of labor in response to competitive conditions within these markets. In doing so, we explore the ways in which outsourcing has contributed to the ongoing expansion of the social division of labor, and how this expansion has gone hand in hand with new patterns of labor market segmentation.
The Role Of Labor Market Intermediaries In The Job Matching Processes For Travel Nurses
Hye Jin Rho, Michigan State University; Christine Riordan and Ki-Jung Kim, University of Illinois Urbana-Champaign
Abstract
Research shows that labor market intermediaries (LMI)—such as temporary help agencies, online labor platforms, and contract firms—play a crucial role in influencing labor market outcomes for workers, including wage-setting processes (e.g., Dube and Kaplan 2010; Bidwell et al. 2013). However, their roles are not always distinct, and they are no longer confined to intermediating tripartite employment relationships. LMIs in healthcare provide an apt illustration of such multifaceted employment relationships. For instance, hospitals may outsource their staffing functions to contract firms (i.e., managed service providers), who in turn subcontract with smaller staffing agencies to fulfill the hospital’s staffing needs. Staffing agencies, in turn, may rely on multiple online platforms to attract qualified workers. Alternatively, service providers and online labor platforms may offer their own staffing services, in effect creating hybrid organizations that have multiple job matching functions. All these arrangements suggest a growing “distance” between nurses and the hospitals at which they work, raising issues about transparency in the job search process. Yet, we lack a sufficient understanding of how workers navigate such an opaque labor market with actors in more complex relationships that may influence outcomes such as wages.
We explore these relationships in the context of the healthcare industry, using fieldwork and supplementing with unique proprietary data from an online labor platform, which we refer to as “TNRecruit”. TNRecruit provides a matching service between travel nurses, or temporary nurses who travel for their short-term work assignments, and the various entities that post job openings on the platform. These include staffing agencies and managed service providers, as well as TNRecruit’s own staffing services arm. Their business model is built on increasing information transparency for travel nurses, as their platform allows job seekers to compare compensation packages from multiple staffing companies for the same work assignment. TNRcecruit’s dataset contains about 2.5 million travel job postings from 135 staffing agencies, for more than 90,000 unique travel nurses seeking work through the platform between 2018-2021. It also includes information on the terms of staffing agencies’ job postings, (e.g., wages), as well as how nurses respond to the job postings (e.g., number interactions with recruiters, in the form of messages, inquiries, and applications).
Our fieldwork suggests that the lack of information transparency in the industry contributes to unequal bargaining power among actors in the network, which in turn, may have an impact on the wages and how nurses seeking travel assignments decide to apply for these jobs. We are in the early stages of using our proprietary data to assess this relationship, first by examining whether transparency in wages on the platform changes how staffing agencies post jobs, and second, by exploring whether variation in the terms of a given job posted by different types of LMIs influence job application rates of travel nurses.
Paul Osterman, MIT Sloan School of Management
Abstract
The importance of contracting can best be understood by contrast with what is perceived as the "standard" approach to organizing work. This "standard" model is heavily rooted in the idea of Internal Labor Markets in which people build their careers within an organization. Contract work is widely seen as an effort by firms to get out from under the constraints of ILMs and outcomes for contract workers are often seen as diminished by their lack of access to ILMs.
There are two difficulties with the foregoing. First, we do not have, and in fact have never had, a measure of how common are ILMs. Second, it is possible—and the literature supports this—that contractors may build a "standard" career within a staffing agency and hence be under the umbrella of an ILM.
In the proposed presentation I will use the 2022 nationally representative survey of employed adults which I have drawn upon in prior papers on contractors. This survey builds upon a careful definition of contracting (based on the work of Abraham, Hershbein, and Houseman ) and it also includes variables that permit an estimate of the incidence of ILMs. I will consider the consequences of ILMs building on a theory of complementary bundles of human resource practices. I will test that theory on multiple economic and attitudinal outcomes.
The measure of ILMs that I will utilize is whether the employer gives strong preference to insiders in filling openings. The outcome measures will fall into three categories: Human Resource Policies (employer provided training, internally focused compensation practices); Economic (earnings, pension provision, health insurance provision, job security); and attitudinal (job satisfaction, organizational citizenship measures).
respect to contractors I will report on the incidence of their access to ILMs and describe how this differs from the non-contractor workforce (I do not consider freelancers because they lack an employer who might provide an ILM). I will also consider what sub-categories of contractors are more or less likely to be in ILMs. I will also analyze whether contractors share in the (possible) outcome benefits of ILMs to the same degree as do standard employees.
In all of the foregoing I will control for standard human capital and organizational features (such as union status, employer size, and industry) and will also consider how the results vary by gender, race, and ethnicity. I will also consider whether the utilization of contractors increases the chances that standard employees have access to ILMs (based on a variable which asks standard employees whether their employer uses contractors in jobs similar to theirs).
Anat Bracha, The Hebrew University of Jerusalem; Mary Burke, The Federal Reserve Bank of Boston
Abstract
Based on seven consecutive annual surveys spanning 2015 through 2021, we investigate whether participation in informal or gig jobs in the United States varies systematically with the business cycle. Considering either the overall participation rate or average hours among participants, our measures of gig work appear to be acyclic. Furthermore, consistent with recent evidence from tax data, the participation rate in gig work was relatively stable over the time period we consider, although ridesharing increased. Nevertheless, our results suggest that there are two broad types of gig workers—one who works relatively few hours in gig activities and is relatively insensitive to the cycle (the typical case), and a more active type (in the top 25 percent of the gig hours distribution) whose hours increase during periods of higher unemployment. Interestingly, the highly active and responsive gig workers report having less money in savings (than other gig workers or non-gig workers) and are more likely to be classified as part-time employees. We also find evidence that the rate of recent entry into gig work increases during periods of higher unemployment.
Our rich survey data not only allow us to reveal heterogeneity across gig workers, they also help us to knit together seemingly inconsistent micro- and macro-level results from previous studies. Specifically, our evidence that the top gig workers engage in such work more intensively during periods of higher unemployment agrees with previous studies using non-survey data that find that some individuals use gig work as a means to smooth consumption in response to employment or earnings shocks; at the same time, since most gig workers in our survey would qualify as multiple job holders, the acyclicality of the aggregate participation rate in gig work agrees with a previous finding that the multiple job-holding (MJH) rate in the United States does not respond to unemployment rate. Our survey suggests that these two potentially contradictory results—that some individuals take on gig work to smooth consumption following employment or earnings shocks and that the MJH rate is acyclic—can coexist because the response of gig work to the business cycle does not push significantly more people into multiple jobholding, as it occurs largely along the hours margin and pertains only to select workers.
Work Flexibility in the Gig Economy
Liya Palagashvili, George Mason University; Paola A. Suarez, Seton Hall University
Abstract
New technology and digital platforms have ushered in a rise of gig, freelance, contract and other types of independent work. Independent work generally exhibits workers taking up commissioned tasks without guarantee of further employment. Independent workers also generally command greater flexibility over their work arrangements compared to their employee counterparts. While there is some research suggesting the value of flexibility for independent workers, there is ample variability in the margins of flexibility which independent workers command—from setting their own schedules and rates, to choosing where and how much to work. We seek to improve our understanding of work flexibility in the gig economy by identifying and measuring distinct margins of flexibility. Using data from a gig economy platform, we empirically examine how workers utilize various margins of flexibility. In particular, we focus on three margins of work flexibility: how much to work, when to work, and where to work. Existing literature shows that flexibility in the labor market is an important job amenity for women. We thus further examine whether men and women exercise margins of flexibility differently in the context of a gig economy platform. In doing so, we seek to gain further insight into the relative importance of distinct margins of flexibility for women in the context of the gig economy.
Understanding Non-Traditional Work Arrangements in the United States
Joelle Abramowitz and Andrew Joung, University of Michigan
Abstract
Over recent decades, mediated by new technologies, entirely new structures for work have emerged. Employers have facilitated a rise of more flexible work through the implementation of innovations such as on-call scheduling, outsourcing, and platform-mediated gig work. However, there exists a dearth of data on these arrangements, precluding understanding of their prevalence, trends, and effects on wellbeing. While administrative records and electronic transaction records provide valuable information on alternative work arrangements, they lack information on many demographic, health, and family characteristics of interest, and do not capture employment taking place outside of their own purviews. While traditional survey data include such information, they generally lack the specificity to identify particular work arrangements of interest.
This paper examines non-traditional work arrangements in the United States over 1997-2019 using the Panel Study of Income Dynamics (PSID). The PSID is a longitudinal survey that is updated every two years; by 2017, it included over 10,000 families and 24,000 individuals. This work uses machine learning to leverage internal data collected in the PSID on respondent narratives on industry and type of work as well as respondents’ employer names. The approach classifies work arrangements into several categories including informal self-employment, formal self-employment, business ownership, and wage and salaried employment.
Preliminary findings show disparate trends in the share of workers engaging in different types of self-employment work arrangements that would otherwise be masked. We find that, between 2003 and 2019, total self-employment rose but that this trend varied by type of self-employment. The share of workers in formal self-employment fell, but an increasing share of workers found work as business owners or through informal self-employment, though with different patterns. For business owners, a marked increase following the Great Recession has subsequently nearly returned to pre-recession levels, while informal self-employment has increased steadily since 2011. Further results suggest that, compared to those in other work arrangements, the informal self-employed generally tend to be less educated, are less likely to be male and non-Hispanic White, have less labor income, and have worse measures of wellbeing. Our findings also suggest that a slightly more male, and substantially more racially and ethnically diverse population has entered platform gig work.
Exploring these questions using these novel data provides unique insights into the changing nature of work relevant to policy considerations across the health, insurance, and poverty dimensions, among others. This study’s findings provide greater insight into the nature of alternative work arrangements and permit future work that will more thoroughly considering the causes and implications of differences in work arrangements. This work lays the groundwork for future research examining individuals' work trajectories leading to these roles, movement between different work arrangements, and how these are associated with different levels of economic, physical, and psychological wellbeing over the life course.