Incongruent skills and experiences in online labor market

Abstract

In online labor markets (OLMs), employers evaluate job applicants based on the published information in their profiles. In this study, we empirically investigate how applicants’ skill information as well as their previous experiences affect employers’ hiring decisions when considering the heterogeneity of online labor markets. We find that (1) employers prefer job applicants with more incongruent skills in low-skill industries, but not in high-skill industries; (2) job applicants’ experiences attenuate employers’ price sensitivity on making hiring decisions in both high-skill industries and low-skill industries. Our findings provide a different perspective from the prior literature on OLMs by considering the labor market heterogeneity, i.e. high-skill industries and low-skill industries. We also provide new insights into employers’ assessment on job applicants’ skill and experience information.

Introduction

Online labor markets (OLMs) such as Upwork, Freelancer, oDesk and Guru, provide employers with a large pool of job candidates, provide applicants with a wide spectrum of work opportunities (Brynjolfsson et al., 2003, Chen and Horton, 2016, Hong et al., 2015), and have emerged to substitute traditional employment practices of contracting with conventional employees or outsourcing companies (Malone and Laubacher, 1999). Enterprises and individual employers are increasingly inclined to seek, hire, supervise, evaluate, and pay workers using these online platforms. With so many participants across different geographic boundaries (by February 2019, more than 31,920,532 users have registered with Freelancer.com and 15,333,579 jobs have been posted¹), job applicants face more fierce competition, and employers may not make optimal hiring decisions facing so many choices (Chan and Wang, 2017).

Given the growing importance of employers’ hiring decisions in OLMs, the existing literature has primarily focused on studying factors or mechanisms that may affect employers’ decisions (Chan and Wang, 2017, Hong and Pavlou, 2017, Hong et al., 2015, Mill, 2011, Scholz and Haas, 2011). However, to our knowledge, the findings of these studies are mostly based on one certain industry, without considering the labor market heterogeneity. Traditional labor market literature has demonstrated the heterogeneity between high-skill industries (e.g., engineering) and low-skill industries (e.g., janitor) (Belo et al., 2017, Gonzaga and Guanziroli, 2019, Looney and Manoli, 2016). For example, Belo et al. (2017) showed that firm’s annual return on long-short portfolio is higher in high-skill industries than in low-skill industries. Also, workers in high-skill industries receive high returns to experience, while workers in low-skill industries receive low returns to experience². After examining the jobs in Freelancer.com, we find that similar separations in online labor markets also exist. The high-skill industries in Freelancer.com include projects such as software design or engineering, which require higher education and specialized skills. The low-skill industries include projects like data entry or administration support, which require only basic computer literacy and numeracy abilities. How does such labor market heterogeneity affect the impact of factors/mechanisms that may affect employers’ decisions?

In addition, compared to the traditional labor markets where jobs are mostly local, online labor markets feature jobs that are open to the public without geographic boundaries. With such spatial and temporal separations between employers and applicants, applicants’ information (e.g., skills and experience) becomes crucial for employers to make hiring decisions. This paper seeks to answer the following question: How does such published information affect employers’ hiring decisions? Does such impact vary in different labor markets (i.e., high-skill industries and low-skill industries) and how?

We are specifically interested in two aspects of published information for each applicant. First, we are interested in examining the impact of applicants’ listed skills. Such skill information helps employers to evaluate job applicants’ qualification for a particular project. For example, previous literature has found that the “matching skills” (those that match the set of required skills of a specific job) have a strong effect on employers’ hiring decisions (Kokkodis and Ipeirotis, 2014). In order to increase the matching probability of jobs in online labor markets, job applicants tend to publish as many skills as possible in their profiles³. In addition, the emphasis on multi-skilled workers in organization management may lead applicants to believe that disclosing more skills help them to portray a talented image (Fitzpatrick and Askin, 2005, Liu et al., 2013). However, on the other hand, many research has demonstrated that companies or producers who attempt to span multiple categories have inferior market performance because they are difficult for audience to situate, understand or evaluate (Negro and Leung, 2013, Rao et al., 2005, Zuckerman, 1999). More skills in applicants’ profiles increases the proportion of skills that are not required by the target project. These incongruent skills may increase employers’ uncertainty about applicants’ expertise for a certain task and reduce their chances of being selected. In this paper, we aim to answer the question that how do applicants’ incongruent skills affect employers’ hiring decisions. Does the impact of such incongruent skills vary in high-skill industries and low-skill industries?

Second, we are also interested in examining how an applicant’s experience moderates the impact of a job applicant’s offering price on employers’ hiring decisions. Previous studies have investigated how applicants’ experiences (Agrawal et al., 2016, Scholz and Haas, 2011) and prices (Kim, 2009, Scholz and Haas, 2011) affect employers’ hiring outcomes in online labor markets. Such literature consistently shows that lower prices or more experiences are usually associated with higher chances of winning the contract (Agrawal et al., 2016, Kim, 2009, Scholz and Haas, 2011). Nonetheless, much less is known about the interplay between applicants’ prices and experiences. Recognizing how employers trade off these two factors leads to a better understanding of how applicants’ experiences affect employers’ price sensitivity in online labor markets.

On one hand, applicants’ experiences may reinforce employers’ price sensitivity, making them susceptible to the price change of applicants. In the absence of face to face interviews and pre-employment training, employers in online labor platforms rely heavily on applicants’ platform experiences (accumulated reviews from previous employers) to reduce uncertainty and ensure ability (Hong and Pavlou, 2017, Scholz and Haas, 2011). Such reliance may make price changes from experienced applicants more noticeable to employers than that from less experienced applicants. In this case, with the same price cut, an experienced applicant may enjoy a higher increase in the winning probability than a less experienced applicant. On the other hand, experiences may attenuate employers’ price sensitivity. Previous literature has shown that quality-related information could lead to lower price sensitivity (Alba et al., 1997). In online labor markets, applicants’ platform experiences can serve as quality information of their service that can alleviate the information asymmetry between employers and applicants. Thus, employers may become less price sensitive towards a more experienced applicant than towards a less experienced applicant, in a way that if the more experienced applicant increases his/her offering price, the reduction in its winning probability is smaller than that for a less experienced applicant. How do prices interplay with experiences? Does this interplay vary in different labor markets (high-skill and low-skill industries)?

We therefore present the following research questions:

• RQ1: How do applicants’ incongruent skills influence employers’ hiring decision (incongruent skill effect)? Are such influences the same across the high-skill industries and low-skill industries in online labor markets?

• RQ2: How do applicants’ experiences interact with prices in influencing employers’ hiring decisions (i.e., reinforce or attenuate)? Does it vary in high-skill industries and low-skill industries?

To address these questions, we utilize a dataset of a leading OLM platform, Freelancer.com. We adopt a conditional logit regression approach to fix the effects of jobs and employers. The results show that (1) employers’ decisions are influenced by the proposed incongruent skills, while such influences differ in the high-skill industry and low-skill industry of online labor markets: employers prefer job applicants with more incongruent skills in the low-skill industry, but not in the high-skill industry; (2) applicants’ experiences and prices exhibit a substitution relationship on employers’ selection decisions, in that the positive effect of price reduction is attenuated by applicants’ experiences.

To our knowledge, this study is the first empirical research that investigates employers’ hiring decisions considering the labor-force heterogeneity in online labor markets, i.e. high-skill industries and low-skill industries. Second, we expand the research on job applicants’ skill information (Kokkodis and Ipeirotis, 2014, Kokkodis et al., 2015). In particular, we investigate from a new perspective by demonstrating how excess skill information influences employers’ hiring decisions in online environment. Finally, our research sheds light on the interplay between applicants’ experiences and prices, namely, how applicants’ pricing strategies differ in terms of their experiences. Our research introduces the concept of price sensitivity to services in online labor markets, explaining employers’ price sensitivity towards applicants with different experiences. These findings provide new insights to our understanding of employers’ hiring decisions in online labor markets. We also offer practical guidelines to job applicants to increase their chances of being selected in online labor platforms.

The rest of this paper is organized as follows. In Section 2, we present the past literature from three dimensions (employers’ selection, surplus and number of bids) and four analysis levels (task, individual, platform and environment). We study the impact of incongruent skills on employers’ hiring decisions, and on the interplay between experiences and prices in Section 3. In Section 4, we present our empirical results. In Section 5, we provide robustness test of main analysis. We discuss the managerial implications in Section 6 and then conclude the paper.