Gig worker organizing: toward an adapted Attraction-Selection-Attrition framework

Gordon B. Schmidt (Department of Management, The University of Louisiana Monroe, Monroe, Louisiana, USA)
Jestine Philip (Pompea College of Business, University of New Haven, West Haven, Connecticut, USA)
Stephanie A. Van Dellen (Department of Organizational Leadership, Purdue University Fort Wayne, Fort Wayne, Indiana, USA)
Sayeedul Islam (Farmingdale State College, Farmingdale, New York, USA)

Journal of Managerial Psychology

ISSN: 0268-3946

Article publication date: 21 November 2022

Issue publication date: 26 January 2023

1511

Abstract

Purpose

As conventional practices of working continue to be modified in the gig economy, more theoretical work examining the experiences of gig workers is needed. Relying on person-based fit and levels of analysis literature, this paper proposes an adaptation to the traditional Attraction-Selection-Attrition (ASA) framework to the gig economy.

Design/methodology/approach

Drawing on the ASA framework, this conceptual paper explores how gig workers join, leave and could be retained by gig employers.

Findings

The authors recognize an intermediary “organizing” phase within the ASA framework for gig workers. Using examples of appwork and crowdwork, the authors show that workers tend to self-organize through third-party websites to help gig work become economically sustainable, avoid being exploited and enhance gig workers' sense of community and identity.

Practical implications

The practical implications of this research lie in gig employers understanding how workers experience gig employment and in helping employers be successful in attracting, selecting and retaining quality workers and thereby lowering permanent attrition.

Originality/value

The authors propose a novel adaptation to the conventional ASA framework to include organizing as a phase in gig worker employment. This research defines gig attraction and attrition at the individual-level, selection at the individual- and task-levels based in person-job (PJ)-fit and the various aspects of gig organizing as encompassing fit with one's job, organization, and environmental (i.e., PJ-, PO-, PE-fit) at the individual-, task-, and network-levels.

Keywords

Citation

Schmidt, G.B., Philip, J., Van Dellen, S.A. and Islam, S. (2023), "Gig worker organizing: toward an adapted Attraction-Selection-Attrition framework", Journal of Managerial Psychology, Vol. 38 No. 1, pp. 47-59. https://doi.org/10.1108/JMP-09-2021-0531

Publisher

:

Emerald Publishing Limited

Copyright © 2022, Emerald Publishing Limited


Introduction and the need for organizing

Even with large numbers of the American workforce now participating in the digital gig economy, common worker-related issues continue to persist across gig platforms. Many gig workers accept jobs with no minimum wage guarantees and experience insecure and precarious working conditions (Tan et al., 2021). Another wage-related concern specific to the gig economy is search costs, i.e. workers not getting paid while searching for the next gig and hence being underemployed (Maarry et al., 2018; Watson et al., 2021). Although digital platforms offer more flexibility and autonomy than traditional full-time jobs with long-term contractual constraints and fixed working hours (Ashford et al., 2018), exposure to risky working conditions due to existing (or lacking) employer policies have compelled workers to adapt in order to survive in the gig economy. In light of these ongoing issues, our conceptual examination reveals how workers are adapting by engaging in online gig communities. Furthermore, we contextualize such adaptations made by gig workers within the Attraction-Selection-Attrition (ASA) theoretical framework.

The purpose for this research is twofold. With the primary goal to expand and contextualize the ASA model to the gig economy, we also reveal how and why gig workers are engaging in online communities. To this end, we present a vital intermediary stage for the gig economy that exists between the selection and attrition stages of ASA and name it “Organizing,” as it involves gig workers connecting and networking with each other (and potentially with allies) in the pursuit of improving the quality and viability of their gig experience. Moreover, with ASA's relevance to various forms of person-based fit (with one's job, organization and environment) (Kristof-Brown et al., 2005), this research frames specific types of fit when examining each stage in the model. Organizing is articulated as additional step that gig workers might choose to undergo to satisfy their lack of person-job fit or person-organization fit, their psychological contract expectation mismatch with the employer and to seek or share information and support from other workers.

Using examples of various gig platforms, we discuss how workers and allies have come together and created websites and applications that help deal with these significant threats to gig worker survival. Due to such a free-flowing relationship between the gig worker and third-party sites, the employer and its human resources function have chosen to stay out of these issues, possibly due to fears of having to convert gig worker employment status to full time and to avoid potential impacts that government mandated work benefits could have on these employers (Meijerink and Keegan, 2019). Examining these independent efforts by gig workers as an adaptation of the ASA framework offers a theoretical basis for understanding individual behaviors in the gig economy and also offers gig employers valuable insights into worker experiences. In doing so, we note that although this research does not compare and contrast traditional employment with gig work or offer a reinterpretation of ASA as such, we still draw from aspects of traditional employment that are relevant for gig attraction, selection and attrition.

The gig economy and gig work

The gig economy is a working environment that contains elements of contingent labor, independent contracting, short-term employment engagements and freelance work. It has been conceptualized as utilizing online digital platforms to connect employers with workers and workers with consumers (Harris, 2017). In general, a “gig” is considered a small, single task or micro task (Aguinis and Lawal, 2013), which a worker is hired to perform.

Gig work has gained popularity – both from the perspective of workers, as well as employers – for several reasons. For one, gig workers enjoy the freedom from being committed to a supervisor or organization and not having to show up at an office every day (Brenoff, 2010). Certain gig work allows workers to perform tasks at their own time and workers can voluntarily discontinue accepting gigs from an employer any time (Keith et al., 2019). Additionally, hiring workers for a gig rather than as contractual employees is beneficial for employers, as it allows them to adjust wages in real-time based on task demand. Since the evolution of the gig economy, employers have access to a workforce for which they do not have to provide compensation-related benefits like they are legally required to do for traditional employees in many countries (Meijerink and Keegan, 2019). Another reason for the popularity of gig work is the cheap, yet efficient services - like Uber and Airbnb - that this form of employment can offer for consumers, wherein quality services are priced lower than their conventional counterparts (i.e. taxi cabs and hotels).

Types of gig work

Scholars identify the digital application platform (e.g. Uber app) as the most common aspect of gig work that distinguishes it from other forms of contingent work (Duggan et al., 2020). These platforms offer a new way of organizing productive work by acting as the digital intermediary between several parties in the gig economy, such as the employer, workers, suppliers (like restaurants in the case of UberEats) and consumers (Gramano, 2019). With various types of gig work currently existing in the gig economy, prior literature has classified them as contingent work types and forms of contract work (Kuhn, 2016). The current discussion of gig work is limited to those that fall into two classifications offered by Duggan et al. (2020) namely, appwork (in which the platform is a service-providing intermediary where algorithms mediate work and customers pay for services offered locally, e.g. transport and food-delivery like Uber, Doordash, Grubhub and Deliveroo) and crowdwork (where digital platforms facilitate work dispersion to individuals in various geographical locations, like Amazon MTurk, Fiverr, Survey Junkie and Upwork).

Crowdwork involves completing simple tasks that require some skills and, therefore, may appeal to individuals looking for additional work that needs little effort. In appwork, the platform algorithm assigns a gig to an available worker who then provides a service (like food or grocery delivery) to the requesting customer. Crowdworkers are selected by the customer/requester (based on skillsets or work quality) and tasks are performed digitally (Williams et al., 2021). Even with such variations, there appear to be overlapping similarities between these two types like the gig employer setting the parameters for who is selected to perform the gig and minimum performance and service quality standards. Both appwork and crowdwork also involve concerns about finding worthwhile gigs and hence are relevant to the current research. It is noted that the third type of gig work offered by Duggan et al. (2020), capital platform work, falls outside the purview of the current research because such a form of gig work involves individuals using the online app to sell their goods or lease assets (like Etsy and Airbnb) and constitutes more of a business-to-business relationship between the app user and platform rather than the completion of tasks by a worker, which is more common in appwork and crowdwork.

From ASA to AS(O)A for gig work

The classic ASA model (Schneider et al., 1995) is based on the premise that organizations are known by the collective characteristics of the people in them, which get defined over time from the cyclic progression of the attraction, selection and attrition phases of employment. As a person-based model, ASA concerns the individual's contractual relationship with the employer and is, therefore, primarily analyzed at the individual-level. A more recent application of the ASA model has been in understanding the framework for communities (at the group-level) (Link and Jeske, 2017). Abedin et al. (2019) found that online communities with increased engagement and value perception had more long-term users. Butler et al. (2014) modeled the impact of ASA on online communities and found that community size impacted members' decision to stay or leave.

Following such trends of advancing ASA for new contexts, we adapt the framework for the gig economy. Knowing that in innovative contexts, the focal level of analysis is often dependent on other levels (Rothaermel and Hess, 2007), we retain ASA's focal individual level while integrating other tangential levels relevant within gig work. Hence, this research accounts for a gig worker's contractual relationship with more than one gig employer, task level variations in performing different types of gig work and the online social networks that influence gig workers' collective actions. Furthermore, we incorporate the concurrent role of person-based fit within gig ASA as it pertains to an appworker's or crowdworker's relationships with the employer (PO-fit), job (Person-job, PJ-fit) and social environment (person-environment, PE-fit).

The proposed Attraction-Selection-Organizing-Attrition - AS(O)A - framework and attributes of conventional attraction, selection and attrition apply to the two categorizations of gig work discussed previously - appwork and crowdwork. This research is a theoretical extension of Duggan et al.'s (2020) work, in that we further investigate the commonalities and differences in appwork and crowdwork through the modified ASA framework. Additionally, this research extends Gramano's (2019) discussion on new ways of organizing work using digital platforms. We introduce organizing as an intermediary that exists in gig ASA, which could potentially buffer gig worker attrition, hence the acronym AS(O)A. The parenthesis for the letter O indicates that organizing is not a mandatory step that every gig worker should necessarily go through, rather it is a provision that has, over time, become available for gig workers offering them the opportunity to connect with other workers should they choose to do so.

From an ASA perspective, we explain ahead how engaging in organizing would help workers remain in the gig economy, thereby reducing worker attrition. Table 1 lists the generalizable and differentiating features of appwork and crowdwork pertaining to AS(O)A. As seen in the table, even though the two forms of gig work vary slightly in regard to attraction and selection, features for organizing and consequently, attrition are similar, thus making organizing a generalizable and influential aspect for appworkers and crowdworkers.

Attraction

The “homogeneity hypothesis” in the ASA framework posits that because organizations strive to be homogenous, the people that find themselves being attracted, getting hired and remaining in an organization will have similar personalities, values and attitudes and, consequently, better fit (Schneider et al., 1995). Employment advertising has an impact on which types of applicants become attracted to different types of work. Recruitment and marketing efforts also likely result in attracting a specific group of gig workers (Meijerink and Keegan, 2019). Stevens and Szmerekovsky (2010) found evidence of signaling effect of language on potential applicants, in that language plays a role in what individual personality types apply for a job (e.g. conscientious language attracts conscientious applicants). The content, descriptions, and terms and conditions provided on a gig platform's website play a key role in attracting the types of individuals who see themselves as being a good fit for that form of gig work (Williams et al., 2021). Even words like “crowdsourcing” and “sharing economy” that are used to describe gig work are attractive because they signal low time investment, which workers see as time available to engage in seeking conventional jobs. These are clear signals by gig platforms about what type of individuals they wish to engage in their platforms.

Subsequently, gig work may be appealing to potential job applicants who are underemployed or unemployed as many workers indicate that they perform gig work to supplement income or to maintain a consistent stream of revenue (Ashford et al., 2018; Keith et al., 2019). There is a positive association between unemployment rates in the traditional labor market and the supply of online workers residing in the same county (Huang et al., 2020). Digital gig platforms allowing workers to find, and oftentimes, complete work online may also explain attraction to gig work. The autonomy and flexibility afforded in performing certain gigs allow part-time workers to engage in both gig and traditional employment simultaneously (Ashford et al., 2018). Hence, both appwork and crowdwork that fit with an individual's personal or professional schedule are enticing to those looking to make any or additional income. Moreover, just as increased homogeneity of a workforce is seen within organizations in ASA, we contend that appwork and crowdwork would each attract a different homogenous group of people who find appropriate fit in either type of gig work.

According to the ASA model, individuals get attracted to potential employers through person-organization (P-O) fit, i.e. the match between an individual's personality and an organization's values (Schneider et al., 1995). P-O fit plays a role in gig attraction, wherein an individual may become attracted to another employer that offers similar types of gig work to the ones they are currently performing over other types depending on similarities in tasks and hiring policies. For example, a worker looking to increase income from gig work might choose to drive more (perhaps for a second ridesharing company knowing that both companies have similar hiring procedures for drivers) rather than considering doing crowdwork-related gigs. Crowdwork employers like Fiverr and Upwork select individuals with good PJ-fit (i.e. good match between individual and job skills), particularly those who possess the skills and abilities to perform services like graphic design, copywriting and website development. There is a positive relationship between workers' employability and P-J fit for Fiverr workers (Carr et al., 2017).

Thus, each type of gig work is distinguished by unique characteristics that would appeal to different psychological characteristics of workers who seek PO-fit or PJ-fit. Attraction in gig AS(O)A constitutes individuals becoming attracted to appwork or crowdwork, with both gig types being distinguishable by their workers' characteristics.

Proposition 1.

Gig workers seeking appropriate fit will be more attracted to appwork or crowdwork that reflect their values, personal characteristics and interests.

Selection

Selection in the ASA model establishes how similar minded employees tend to select and hire other individuals that share their own personal attributes and consequently, the organization's values and interests (Schneider et al., 1995). In gig economy, however, workers are rarely hired based on personal characteristics. That said, the individual-level of analysis is still relevant in gig selection because personality type might play a role as individuals self-select into appwork (which may involve frequent social interactions) or into skill-based crowdwork. Personality-task misfit occurs when individuals with high need for cognition perform non-creative tasks (Wronska et al., 2019), suggesting that PJ-fit plays a role in whether a worker would self-select into appwork or crowdwork.

Hiring procedures in appwork and crowdwork provide insights into the selection processes for gigs. For appwork like Uber, selection involves basic procedures like completing an online form, passing driving and criminal history background checks and meeting minimum requirements like legal driving age and excludes determinants of organization fit like personality tests that are part of traditional selection processes. Gigs are not automatically assigned to every approved apprworker. Appworkers increase their chances of being selected when they are physically close to the locally available appwork and operate on a first-come first-served basis of responding to the app's gig offer.

Crowdwork, on the other hand, wherein the worker is selected by the requestor, often requires specific knowledge, skills and abilities (KSA's). Competing crowdwork platforms like Fiverr, Upwork and Freelancer.com are known to have similar selection criteria for freelancers, wherein gig workers with skills like writing, designing, video editing, voiceover work and knowledge of social media marketing, can set a price for their services and act as sellers (Green et al., 2018). The selection process is conducted by the businesses or bidders who hire workers with adequate KSA's.

Moreover, individual performance ratings in gig work influences the selection process in appwork because of how algorithms are designed to select workers when a task becomes available. Ridesharing drivers spend long hours driving, particularly during peak hours, to maintain high ratings (Duggan et al., 2020), which the algorithm then uses to identify good drivers and to route more gigs to their queue. The same logic also holds for crowdwork because workers with higher customer ratings are more likely to be selected by the task requestor. Hence, selection in AS(O)A can be assessed both at the individual and task levels.

Proposition 2.

Gig workers selected for more gigs, based on their proximal availability or KSA's, will have higher customer ratings which, in turn, increase the workers' chances of future selection.

Organizing

As ASA pertains to full-time employment and employee fit with a single organization, the framework does not discuss workers not getting appropriate or sufficient tasks. However, in adapting ASA for non-traditional gig work, inherent aspects like underemployment and low wages must be given consideration. Thus, organizing becomes highly relevant as gig workers - from the same or different appwork and crowdwork platforms - support each other by sharing information about available tasks and discussing frustrations about their working conditions.

Organizing to find worthwhile gigs

While gig employers may provide an avenue where gigs are posted (like MTurk or Upwork websites) or directly assign potential tasks (like ridesharing algorithms do), they offer little information on whether the gig is worthwhile to the worker in terms of compensation or about the worker's chances of successfully completing the task. When asking a driver to accept a ride request, the Uber app only provides information on the availability of a ride, not where a rider is going or if they might cancel while the driver is enroute to pick them up, or even whether the gig is a worthwhile use of the driver's time (Wells et al., 2020). We contend that organizing plays a role in helping the worker decide what types of ride requests are beneficial for them. Hence, organizing to find worthwhile gigs is appropriated at the task level with workers more likely to complete gigs that have good PJ-fit (i.e. gigs deemed worthy of their skills, time and effort).

Some third-party sites are organized exclusively for one gig company, offering advice on gigs tied to that particular app (Schmidt and Jettinghoff, 2016). In crowdwork like MTurk, workers share their experiences - through forums like Daily HITs - about the gigs they performed and found financially worthwhile (Maarry et al., 2018; Schmidt and Jettinghoff, 2016). For appwork sites like Uber, gig workers may be physically present in the same location; hence, there is more direct real-time competition for gigs. The sub-reddit “Uber Drivers” (https://www.reddit.com/r/uberdrivers/) includes both a FAQ page on general strategies to be successful, as well as individual users making threads about specific strategies and questions.

Organizing to avoid bad employers

Gig workers also organize to avoid bad employers. Gig platforms often provide greater control and resources to requestors than to workers, often leaving workers with inadequate compensation, unfair negative ratings that result in them getting less work, being deactivated by the gig platform (Wells et al., 2020), or even suffering monetary penalties (Feng, 2020). MTurkers organize to avoid bad employers through Turkopticon (https://turkopticon.ucsd.edu/) and Fiverr workers do so through sub-reddit (https://www.reddit.com/r/Fiverr/), where workers rate their experiences with certain requesters (Chan et al., 2019). Similarly, appworkers organize through message boards, with Uber drivers often posting about pitfalls of bad customers and warning signs, like the amount of control a driver had over refusing passengers who failed to follow protocols (MondayMan, 2020). In effect, gig workers organize online to avoid “bad” employers or requestors because they lack any form of relationship, trust, or compatibility with site requestors and are seeking appwork or crowdwork platforms with which they might share common goals, values and, subsequently, PO-fit.

Organizing as collective action

In the organizing phase, gig workers can also come together as a collective to demand better working conditions from their employer. Wells et al. (2020) analyzed how Uber drivers - both in local regions and globally - united to protest their working conditions. Freelance workers connected with allies like the “Freelancers Union” (Graham et al., 2017), which offers opportunities for freelancers to purchase health insurance and other benefits not provided directly through freelance work (Freelancers Union, 2020). Similarly, Working Washington (http://www.workingwa.org/) rallied for $15/h wage for fast food workers (Working Washington, n. d.) and joined forces with Instacart workers to advocate for better pay and working conditions (Covert, 2020). Such actions are consistent with research showing a positive association between a gig worker's social interactions on online communities and their increased interest in joining labor unions (Maffie, 2020). However, the formation of legally defined unions or labor associations for gig workers remains to be seen.

These instances suggest that an aspect of organizing exists at the network level that encompasses one's social environment. The PE-fit, in this case, pertains to a “demand-supply” match between a worker's needs and the resources available in the online communities and assumes that such a fit would impact a worker's psychological, attitudinal and behavioral outcomes (Edwards and Shipp, 2007). One's network ties in the environment also influence individual and group behavior (Mizruchi and Marquis, 2006). When organizing for collective action, we now understand that appworkers' and crowdworkers' networks and PE-fit tend to influence what kind of action they might take (like protesting, forming alliances etc.).

Organizing for identity

While their intentions to engage in organizing largely concern wages and working conditions, gig workers may also view organizing as beneficial for their own psychological contract needs and identity. Gig work is known for its “precarious identity”, as workers often lack a professional relationship or the feeling of being connected to a particular organization (Petriglieri et al., 2019), which then motivates them to create their own work identity that is meaningful and sustainable (Ashford et al., 2018). A number of third-party sites offer workers help with emotional support and collegiality. Building on prior research suggesting that gig workers generally feel less identity with their employers (Meijerink and Keegan, 2019), we posit that engaging in organizing would create individual-level identity specifically for workers who feel some degree of shared values and beliefs with their gig employers (i.e. PO-fit). An instance of such PO-fit based identification is seen among MTurk and TaskRabbit workers who call themselves “MTurkers” and “Taskers,” respectively (Panteli et al., 2020).

Overall, workers organizing for various reasons are theoretically framed within ASA because they capture the individual's ongoing contractual and (positive or negative) psychological relationships with their gig employers (e.g. organizing to avoid bad contracts or to build identity). In extending ASA for the gig economy, we acknowledge that worker outcomes are further influenced by the task (gig) itself and the online social networks that facilitate said organizing.

Proposition 3.

Gig workers on a particular gig platform who organize will have better fit, higher hourly pay rates, fewer negative experiences and better identity than workers on that gig platform who do not engage in organizing.

Attrition

Attrition in the ASA model focuses on a single employer and how employer characteristics mismatch with the employee's expectations results in the individual leaving that employer permanently. In the gig economy, however, attrition may involve leaving more than one employer, as gig workers tend to work more than one type of gig concurrently (Booty, 2017). Moreover, as gig work inherently involves micro tasks that can be completed within short spans of time, the possibility of workers leaving any one employer briefly must also be considered.

Therefore, we expand the term “attrition” in gig work to include the following three instances - a) workers leaving gig work permanently (and deactivating all their gig platform accounts), b) workers not engaging in/deleting all gig platform apps temporarily (still keeping their gig platform accounts active) and returning to gig work after an extended time gap (like several months), or c) workers exiting one gig employer's app momentarily (like certain times of the day or certain days of the week) to be active on another gig employer's app. In this last instance, workers may remain inactive on the first app or seek fewer gigs from it. Whereas conventional attrition does not account for temporary or momentary departures from routine employment, such brief exits in the gig economy could be considered synonymous to a traditional employee taking an unpaid sabbatical leave (sometimes to pursue personal interests like reading and writing). We deem such a non-permanent exit as gig workers taking a short or extended “pause” from one gig platform to engage on another platform or in other forms of employment.

Due to the vast number and types of gigs available today and the ease of switching from one employer's gig to another using a cell phone app or computer, instances of temporary and momentary exits from gig work are seen in both appwork and crowdwork. Some ridesharing and food delivery app workers switching between Uber/Lyft apps and Grubhub/Doordash/UberEats apps numerous times a day to maximize their chances of receiving ride or delivery requests (Helling, 2020) is an example of momentary attrition from either app. Such a form of attrition is consistent with research positing that appworkers are likely to experience fewer hours of underemployment due to this switching behavior (Watson et al., 2021). Similarly, college students who return to their MTurk accounts to perform online gigs during non-academic months (e.g. summer) exhibited temporary attrition from gig work during their school year.

In the ASA framework, expectation mismatch relating to compensation is known to be a reason for leaving. While it is likely that poor compensation could result in an individual leaving gig work permanently, many workers remain in the gig economy due to lack of options in other realms of traditional work. Hence, income-related expectation mismatch would more likely result in temporary and momentary attrition from gig work first before workers decide to leave permanently.

We contend that organizing to find worthwhile gigs (that offer PJ-fit) would reduce such a mismatch between workers' general expectations of gig income and the actual income they earn through a particular gig employer. In this case, organizing might have a buffering effect on their permanent attrition because other workers in the organizing community might offer their own perceptions of “fair” pay, thereby helping the gig worker to lower or adjust their psychological contract expectations. Moreover, if a gig worker engages regularly with others on online communities, they might come to know of gigs that are in high demand during certain time periods in specific locations (e.g. demand for ridesharing appwork at airports during holiday seasons). In this example, organizing helps reduce permanent attrition from ridesharing appwork for workers who perceive job fit because such workers might decide to drive seasonally rather than quitting driving entirely, which also allows them to maintain some engagement with that gig work. Such a claim about reduced attrition is supported by research as good PJ-fit is negatively related to turnover intentions when mediated by work engagement (Kristof-Brown et al., 2005).

Furthermore, engaging in organizing is helpful in managing expectations during momentary attrition - when workers switch between apps or complete online tasks on multiple crowdwork websites simultaneously - as workers can keep up with the latest online posts or ratings of other gig workers' experiences with particular requesters or customers. The various aspects of organizing buffer workers from leaving the gig economy entirely. Attrition in gig AS(O)A is more nuanced than conventional ASA as it entails workers attempting to manage employer expectation mismatch by engaging in varied forms of attrition.

Proposition 4.

In the AS(O)A framework for appwork and crowdwork, workers with person-job fit that engage in organizing will experience permanent attrition from gig work to a lesser extent.

Discussion

This conceptual research re-examined the classic ASA model for the gig economy and reveals a need for organizing as an additional phase for gig workers. Even with parsing ASA, examining it at levels beyond the individual and revealing the organizing phase for gig workers, our research maintains that such a theoretically anchored model of traditional employment still applies to contemporary gig work. Lacking sufficient fit and support with their gig employer(s), connecting with each other through third party sites and online communities is an important means by which workers sustain themselves in gig work. Hence, the organizing phase plays a crucial role in helping gig workers improve person-based fit, find gigs that maximize their pay outs and minimize their negative experiences with gig work.

We reiterate that the current research is necessary to advance both theoretical and practical knowledge of how workers experience gig work. Firstly, for scholarly literature, we contextualized the ASA framework for the gig economy by revealing the existence of an organizing phase occurring between selection and attrition. Framed within the concepts of person-based fit and levels, this research offered theoretical arguments for attraction, selection and three forms of attrition for crowdwork and appwork. Moreover, we presented four reasons (situated at various levels of analysis) for why gig workers would engage in organizing. It could be further argued that this research was needed to extend ASA's framing beyond its focal level of analysis (i.e. the individual) in order to contemporize ASA for modern-day work. Secondly, for management practitioners, we unpacked, named and explained organizing as an important phase occurring widely within the gig economy that employers have yet to take note of.

Theoretical implications

The theoretical contributions lie in the complex levels of analysis and fit that were considered in building propositional statements for gig AS(O)A. For appwork and crowdwork, this research defined attraction and attrition at the individual-level, selection at the individual- and task-levels based in PJ-fit and the various reasons for organizing as encompassing fit with one's job, organization and environment (i.e. PJ-, PO- and PE-fit) at the individual-, task- and network-levels (also shown in Table 1). Presenting the organizing phase in gig employment enabled us to incorporate arguments at the network-level (a level sparsely investigated in ASA). Scholars agree that incorporating multiple levels of analysis in a research topic brings to light the inherent gaps and issues, which further aids theory building and testing (Kozlowski and Klein, 2000). Furthermore, multiple levels in gig research allow future research questions to be answered about the individual gig worker's experiences, situations in which they operate and the work systems at play (e.g. socio-technical aspects of the gig economy) (Cameron, 2022).

Lastly, in claiming such theoretical contributions, we acknowledge the existing boundary condition of AS(O)A with regard to capital platform work. As this form of gig work offers asset-based services, where individuals share their self-assets like accommodation and act as small businesses rather than as workers (Duggan et al., 2020), we contend that these “workers” in capital platform work may be engaging in efficient sharing and consolidating (rather than purely organizing) with platform providers to improve their own strategic and psychological outcomes.

Practical implications and recommendations for gig employers

Our proposed adaption of the ASA framework to examine gig work informs gig employers of the relevance and importance of the organizing phase. Gig employers can derive implications for both micro- and macro-organizational levels of the gig economy through this research. At the macro level, it is revealed that P-O fit may be absent in crowdwork as skills and abilities determine employability in this type of gig work. Hence, employers would benefit from knowing that for crowdworkers who do not engage in organizing, such a lack of fit between personal characteristics and organizational values could result in permanent attrition. In seeking personal-organizational value alignment for their contingent workers, interested employers could advertise themselves as offering gigs suited to certain personality types, thereby delivering a more consistent experience for workers and customers alike.

Implications at the micro-level (for workers that do not organize) include being underemployed, experiencing high psychological contract expectation mismatch and poor psychological outcomes. In regard to underemployment, when employers do not invest in rigorous selection procedures, more workers become part of the gig economy resulting in workers receiving fewer gigs during low demand, which in turn, either leads to their need to organize or permanently attrit. Even those who receive frequent gigs (like online survey takers) would fail to consistently find economically sustainable work if supply exceeds demand. Moreover, high performers - not being able to find challenging tasks - would eventually leave gig work entirely. Hence, our research emphasizes that underemployment (and its associated costs on workers) is an inherent part of gig selection and attrition, which employers could help mitigate by better understanding their workers' organizing efforts.

Organizing has also allowed well performing and interested workers to adjust their psychological contract expectations of their employer(s) and find gigs creatively to sustain themselves, thereby only resulting in their momentary and temporary attrition from the gig economy. Because the responsibility of retention is currently on the worker, workers are actively organizing through online communities and third parties to create collegial bonds with each other. It is crucial that gig employers engage with workers to learn about their grievances and psychological outcomes and to offer support and create better working conditions. One way in which gig employers might engage in worker organizing efforts is by offering organizing platforms in-house (as currently workers organize on third-party sites). Such platforms could be housed within company websites or apps and designed keeping in mind the personality characteristics, interests and psychological needs of their current and future workers. By creating such an internal space, gig workers can be expected to have a direct line of communication with their employer, which would eventually reduce their psychological contract expectations mismatch.

In sum, the practical implications of this research lie in gig employers understanding how and why workers organize as part of their gig employment. To this end, we offered suggestions to gig employers so that they may be successful in attracting, selecting and retaining quality workers and lowering permanent attrition. Doing so would also improve the employer's brand reputation in the employment market if they prioritized aspects like worker fit, psychological contract expectations and overall worker experiences.

Future directions and conclusion

Empirical research on gig worker selection is needed to know how they are benefiting from organizing and enhancing their chances of being selected for future gigs. Content analyses of online community sites may help reveal whether workers organize differently in appwork and crowdwork based on their needs. It is also important to compare empirically our claim that gig workers engaging in organizing (versus those that are not) are more likely to stay in the gig economy. Furthermore, the varying buffering effects that organizing is having on the three forms of gig attrition can be studied by examining whether workers perceive some degree of psychological contract fulfillment when they display momentary and temporary attrition. Lastly, empirical research could determine existing variances in individual-, community/group-, situational- and organizational-levels in gig work.

In conclusion, the current theoretically grounded research reveals important implications for employers as their workers navigate the gig economy. We proposed an adaptation to the conventional ASA framework to recognize organizing as a crucial intermediary phase in gig worker employment. For workers who engage in organizing activities, this stage plays a key role in improving fit with their job, employer and environment as well as in enhancing their overall experience and identity with gig work.

Generalizable and differentiating features of appwork and crowdwork for AS(O)A

AppworkCrowdwork
AttractionAttracts underemployed seeking PO-fitAttracts underemployed seeking
PJ-fit
SelectionPlatform selects based on proximal availability and ratings (PJ-fit)Requestor selects based on KSA's and ratings (PJ-fit)
(Organizing)To find worthwhile gigs and PJ-fit;
To avoid bad employers/requestors and find PO-fit;
For collective action from PE-fit;
For identity and PO-fit
AttritionThree forms from lack of PJ-fit: Momentary, Temporary, Permanent
Organizing mitigates permanent attrition

References

Abedin, B., Erfani, S., Milne, D., Beattie, A. and Fenerty, K. (2019), “Unpacking support types in online health communities: an application of attraction-selection-attrition theory”, in PACIS 2019 Conference Proceedings.

Aguinis, H. and Lawal, S.O. (2013), “eLancing: a review and research agenda for bridging the science–practice gap”, Human Resource Management Review, Vol. 23, pp. 6-17, doi: 10.1016/j.hrmr.2012.06.003.

Ashford, S.J., Caza, B.B. and Reid, E.M. (2018), “From surviving to thriving in the gig economy: a research agenda for individuals in the new world of work”, Research in Organizational Behavior, Vol. 38, pp. 23-41. doi: 10.1177/0022185619865480.

Booty, L. (2017), “A third of UK gig workers juggling multiple jobs”, available at: https://realbusiness.co.uk/third-uk-gig-workers-juggling-multiple-jobs (accessed 18 March 2022).

Brenoff, Anna (Jacobson) (2010), “Unemployment's down but something's up: welcome to the gig economy – DailyFinance”, available at: http://www.dailyfinance.com/2010/07/26/unemployments-down-but-somethings-upwelcome-to-the-gigecono/ (accessed 21 July 2021).

Butler, B.S., Bateman, P.J., Gray, P.H. and Diamant, E.I. (2014), “An attraction–selection–attrition theory of online community size and resilience”, MIS Quarterly, Vol. 38, pp. 699-729.

Cameron, L.D. (2022), “Making out” while driving: relational and efficiency games in the gig economy”, Organization Science, Vol. 33 No. 1, pp. 231-252.

Carr, C.T., Hall, R.D., Mason, A.J. and Varney, E.J. (2017), “Cueing employability in the gig economy: effects of task-relevant and task-irrelevant information on Fiverr”, Management Communication Quarterly, Vol. 31, pp. 409-428.

Chan, E., Baez, A. and Irani, L. (2019), “The promise and limits of tailorability for turkopticon”, Designing Interactive Systems Conference 2019 Companion Publication, pp. 141-145, June.

Covert, B. (2020), “Like Uber but for gig worker organizing”, available at: https://prospect.org/labor/like-uber-but-for-gig-worker-organizing/ (accessed 22 September 2021).

Duggan, J., Sherman, U., Carbery, R. and McDonnell, A. (2020), “Algorithmic management and app‐work in the gig economy: a research agenda for employment relations and HRM”, Human Resource Management Journal, Vol. 30, pp. 114-132, doi: 10.1111/1748-8583.12258.

Edwards, J.R. and Shipp, A.J. (2007), “The relationship between person-environment fit and outcomes: an integrative theoretical framework”, in Ostroff, C. and Judge, T.A. (Eds), Perspectives on Organizational Fit, Jossey-Bass, San Francisco, pp. 209-258.

Feng, E. (2020), “For China's overburdened delivery workers, the customer — and app — is always right”, available at: https://www.npr.org/2020/12/01/938876826/for-chinas-overburdened-delivery-workers-the-customer-and-app-is-always-right (accessed 22 September 2021).

Graham, M., Hjorth, I. and Lehdonvirta, V. (2017), “Digital labour and development: impacts of global digital labour platforms and the gig economy on worker livelihoods”, Transfer: European Review of Labour and Research, Vol. 23, pp. 135-162, doi: 10.1177/1024258916687250.

Gramano, E. (2019), “Digitalisation and work: challenges from the platform economy”, Contemporary Social Science, Vol. 15, pp. 476-488, doi: 10.1080/21582041.2019.1572919.

Green, D., Walker, C., Alabulththim, A., Smith, D. and Phillips, M. (2018), “Fueling the gig economy: a case study evaluation of Upwork. com”, Management and Economics Research Journal, Vol. 4, pp. 104-112.

Harris, B. (2017), “Uber, Lyft, and regulating the sharing economy”, Seattle University Law Review, Vol. 41, pp. 269-285.

Helling, B. (2020), “How to drive for uber and lyft (at the same time)”, available at: https://www.ridester.com/drive-for-uber-and-lyft/#:∼:text=You'll%20need%20to%20have,both%20at%20the%20same%20time.&text=This%20is%20a%20subtle%20way,still%20receive%20Lyft%20ride%20requests (accessed 22 September 2021).

Huang, N., Burtch, G., Hong, Y. and Pavlou, P.A. (2020), “Unemployment and worker participation in the gig economy: evidence from an online labor market”, Information Systems Research, Vol. 31, pp. 431-448.

Keith, M.G., Harms, P. and Tay, L. (2019), “Mechanical Turk and the gig economy: exploring differences between gig workers”, Journal of Managerial Psychology, Vol. 34, pp. 286-306, doi: 10.1108/JMP-06-2018-0228.

Kozlowski, S.W. and Klein, K.J. (2000), “A multilevel approach to theory and research in organizations: contextual, temporal, and emergent processes”, in Klein and Kozlowski (Eds), Multilevel Theory, Research and Methods in Organizations: Foundations, Extensions, and New Directions, Jossey-Bass, San Francisco, CA, pp. 3-90.

Kristof‐Brown, A.L., Zimmerman, R.D. and Johnson, E.C. (2005), “Consequences of individuals' fit at work: a Meta‐analysis of person–job, person–organization, person–group, and person–supervisor fit”, Personnel Psychology, Vol. 58, pp. 281-342.

Kuhn, K.M. (2016), “The rise of the “gig economy” and implications for understanding work and workers”, Industrial and Organisational Psychology, Vol. 9, pp. 157-162, doi: 10.1017/iop.2015.129.

Link, G.J. and Jeske, D. (2017), “Understanding organization and open source community relations through the attraction-selection-attrition model”, Proceedings of the 13th International Symposium on Open Collaboration, pp. 1-8.

Maarry, K.E., Milland, K. and Balke, W.T. (2018), “A fair share of the work? The evolving ecosystem of crowd workers”, WebSci’18, May 27-30, 2018, Amsterdam, Netherlands, doi: 10.1145/3201064.3201074.

Maffie, M.D. (2020), “The role of digital communities in organizing gig workers”, Industrial Relations: A Journal of Economy and Society, Vol. 59, pp. 123-149.

Meijerink, J. and Keegan, A. (2019), “Conceptualizing human resource management in the gig economy: toward a platform ecosystem perspective”, Journal of Managerial Psychology, Vol. 34, pp. 214-232, doi: 10.1108/JMP-07-2018-0277.

Mizruchi, M.S. and Marquis, C. (2006), “Egocentric, sociocentric, or dyadic?: Identifying the appropriate level of analysis in the study of organizational networks”, Social Networks, Vol. 28 No. 3, pp. 187-208.

MondayMan (2020), “Can we do a ‘no mask’ cancel for people who have valves and vents in their masks?”, available at: https://www.uberpeople.net/threads/can-we-do-a-no-mask-cancel-for-people-who-have-valves-and-vents-in-their-masks.417516/ (accessed 22 September 2021).

Panteli, N., Rapti, A. and Scholarios, D. (2020), “If he just knew who we were’: microworkers' emerging bonds of attachment in fragmented employment relationship”, Work, Employment and Society, Vol. 34, pp. 476-494, doi: 10.1177/0950017019897872.

Petriglieri, G., Ashford, S.J. and Wrzesniewski, A. (2019), “Agony and ecstasy in the gig economy: cultivating holding environments for precarious and personalized work identities”, Administrative Science Quarterly, Vol. 64, pp. 124-170, doi: 10.1177/0001839218759646.

Rothaermel, F.T. and Hess, A.M. (2007), “Building dynamic capabilities: innovation driven by individual-, firm-, and network-level effects”, Organization Science, Vol. 18 No. 6, pp. 898-921.

Schmidt, G.B. and Jettinghoff, W. (2016), “Using Amazon Mechanical Turk and other compensated crowdsourcing sites”, Business Horizons, Vol. 59, pp. 391-400.

Schneider, B., Goldstein, H.W. and Smith, D.B. (1995), “The ASA framework: an update”, Personnel Psychology, Vol. 48, pp. 747-773, doi: 10.1111/j.1744-6570.1995.tb01780.x.

Stevens, C.D. and Szmerekovsky, J.G. (2010), “Attraction to employment advertisements: advertisement wording and personality characteristics”, Journal of Managerial Issues, Vol. 22, pp. 107-126.

Tan, Z.M., Aggarwal, N., Cowls, J., Morley, J., Taddeo, M. and Floridi, L. (2021), “The ethical debate about the gig economy: a review and critical analysis”, Technology in Society, Vol. 65, 101594.

Union, Freelancers (2020), “About freelancers union”, available at: https://www.freelancersunion.org/about/ (accessed 22 September 2021).

Watson, G.P., Kistler, L.D., Graham, B.A. and Sinclair, R.R. (2021), “Looking at the gig picture: defining gig work and explaining profile differences in gig workers' job demands and resources”, Group & Organization Management, Vol. 46, pp. 327-361.

Wells, K.J., Attoh, K. and Cullen, D. (2020), “Just-in-Place” labor: driver organizing in the Uber workplace”, EPA: Economy and Space, Vol. 53, pp. 315-331, doi: 10.1177/0308518X20949266.

Williams, P., McDonald, P. and Mayes, R. (2021), “Recruitment in the gig economy: attraction and selection on digital platforms”, The International Journal of Human Resource Management, Vol. 32, pp. 4136-4162.

Working Washington (n. d.), “We are working Washington”, available at: http://www.workingwa.org/about (accessed 22 September 2021).

Wronska, M.K., Bujacz, A., Gocłowska, M.A., Rietzschel, E.F. and Nijstad, B.A. (2019), “Person-task fit: emotional consequences of performing divergent versus convergent thinking tasks depend on need for cognitive closure”, Personality and Individual Differences, Vol. 142, pp. 172-178.

Corresponding author

Gordon B. Schmidt can be contacted at: gschmidt@ulm.edu

Related articles