Re-examining the effect of wage delegation: a replication study of Charness et al. (2012)

Niehoff, Hendrik; Schreck, Philipp Article — Published Version Re-examining the effect of wage delegation: a replication study of Charness et al. (2012) Journal of Business Economics Provided in Cooperation with: Springer Nature Suggested Citation: Niehoff, Hendrik; Schreck, Philipp (2024) : Re-examining the effect of wage delegation: a replication study of Charness et al. (2012), Journal of Business Economics, ISSN 1861-8928, Springer, Berlin, Heidelberg, Vol. 95, Iss. 2, pp. 237-256, https://doi.org/10.1007/s11573-024-01214-7 This Version is available at: https://hdl.handle.net/10419/323469 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) Journal of Business Economics (2025) 95:237–256 https://doi.org/10.1007/s11573-024-01214-7 ORIGINAL PAPER Re‑examining theeffect ofwage delegation: areplication study ofCharness etal. (2012) HendrikNiehoff1 · PhilippSchreck1 Accepted: 31 October 2024 / Published online: 4 December 2024 © The Author(s) 2024 Abstract Experimental research has explored the effects of wage delegation on employee performance, with the pioneering attempt by (Charness etal., Am Econ Rev 102:2358– 2379, 2012) in “The Hidden Advantage of Delegation: Pareto Improvements in a Gift Exchange Game”, published in the American Economic Review. We conducted a replication of their experiment for two main reasons: first, to validate the original findings that have influenced subsequent research, and second, to examine whether these results hold in contexts with reduced social proximity between workers and employers, a relevant issue due to the rise of remote work. Our online experiment, involving 410 participants, followed the original study’s design but used a different sample. We successfully replicated the main finding that wage delegation increases employee effort, though the effect was smaller and largely due to employees granting themselves higher wages. These results support the notion that with decreasing social proximity, formal controls increase in importance. Keywords Wage delegation· Replication study· Performance incentives· Employee empowerment· Intrinsic motivation JEL Classification C91· J31· J33· J41· M52 1 Introduction A growing body of evidence originating from the field of employee empowerment suggests that increasing the degree of autonomy for workers can foster a greater sense of responsibility and job satisfaction (Maynard etal. 2012; Yin etal. 2019). * Hendrik Niehoff [email protected] Philipp Schreck philipp.schrec[email protected] 1 School ofLaw andEconomics, Martin-Luther-University Halle-Wittenberg, Große Steinstr. 73, 06108Halle(Saale), Germany 238 H.Niehoff, P.Schreck By allowing employees more freedom, their motivation can naturally increase, leading to improved productivity. In particular, it has been argued that the democratization of organizations may increase employee satisfaction, effort and performance on multiple dimensions (Harrison and Freeman 2004). Importantly, employee empowerment may imply the delegation of wage decisions to employees themselves with the goal of increasing employee autonomy and satisfaction (Mellizo et al. 2014). In line with this notion a wide range of businesses are practicing this autonomyencouraging approach already, including companies such as Semco from Brazil, Skyline from the United States, or the Spanish company Claravision (Semler 2007; Tuna 2008). Over the years, various experimental studies have investigated the effects of wage delegation on employee performance, with the pioneering attempt to study this relationship by Charness etal. (2012) in “The Hidden Advantage of Delegation: Pareto Improvements in a Gift Exchange Game”, published in the American Economic Review. This landmark study aimed to analyze the outcomes of wage delegation in a principal-agent setting via an incentivized lab experiment. The results of this experiment provide strong evidence of an increase in both wages and worker effort as a result of delegating wage determination. The authors attribute this effect primarily to employees’ increased sense of responsibility, rather than to positive reciprocity toward their employers (Charness etal. 2012). In the current paper, we report the results of a study replicating the experiment of Charness et al. (2012). There are two main reasons for our replication study. The first is that the insights from the study by Charness etal. (2012) have served as a foundation for numerous follow-up studies, each varying in design and empirical approach. For instance, Jeworrek and Mertins (2019) conducted a natural field experiment that validated the positive effect of wage delegation on employee performance in the field. Also other studies, such as those by Mellizo etal. (2014), Charness etal. (2016), Franke etal. (2016), and Faillo and Piovanelli (2017), have built upon the insights of Charness etal. (2012) utilizing different experimental designs to further explore the effects of wage delegation. More recently, Brück etal. (2021) shifted the focus away from the relationship between wage delegation and employee performance and explored the impact on employee creativity. This array of diverse articles underscores the significance and potential importance of replicating the study by Charness etal. (2012). The second main reason for our effort to replicate Charness etal. (2012) is that we want to explore some of the study’s boundary conditions. The original study was conducted in a behavioral laboratory, which is an apt setting for mimicking situations of social proximity as they prevail in organizations. In such settings, interaction partners can develop a shared understanding of a situation, and social norms such as trust and reciprocity may evolve. The original study argued that participants in the role of employees may have increased effort in response to delegation due to an enhanced feeling of responsibility towards the employee (Charness etal. 2012: 2360). However, experimental research has shown that an increased social distance between interaction partners strongly affects how they think and feel about the interaction (Charness etal. 2007; Schmelz and Ziegelmeyer 2020), and that this 239 Re‑examining theeffect ofwage delegation: areplication… distance might have detrimental effects on workers performance (e.g. Cramton 2001; Cramton and Webber 2005; Burbano and Chiles 2022). These findings raise the question whether the original results also hold in settings where people interact across larger distances, as is the case when they work from home. Given the current trend of growing physical distances between companies and employees working remotely (Bloom etal. 2023), this is a highly relevant question. We thus use the design of the original experiment and seek to replicate it using an online participant pool. Conducting experiments online vs. in the lab captures “meaningful differences between working from home and working at the office arrangements as online subjects enjoy greater anonymity than lab subjects, they interact in a less constrained environment than the laboratory, and there is a larger physically oriented social distance between them.” (Schmelz and Ziegelmeyer 2020: 958). Replication studies play an indispensable role in economics and management accounting, though they represent only a modest proportion of all recent academic publications, with a large number of highly cited papers yet to be replicated (Albers 2014; Hensel 2019; Mueller-Langer etal. 2019; Brüggen etal. 2021). This limited prevalence can be attributed to hurdles such as issues around data and code access and a lack of strong incentives to publish these types of studies (Ditzen and Elhorst 2022). Furthermore, the literature lacks a universally agreed-upon demarcation for the concepts of reproducibility, replicability, and robustness and moreover different varieties of replication forms exist (Duvendack etal. 2017; Köhler and Cortina 2021). Nevertheless, the relevance of replication studies in experimental economics has gained increasing prominence, particularly due to the so-called replication crisis in psychology, which has sparked these discussions (Maniadis etal. 2015; Open Science Collaboration 2015). Past replications have often yielded results that were inconsistent with the original studies (Chang and Li 2022). For instance, the findings of a large-scale replication project conducted by Camerer etal. (2016) demonstrated that only 11 out of 18 replicated studies (61%) published in the American Economic Review and the Quarterly Journal of Economics, covering the period from 2011 to 2014, showed a significant effect in the same direction as the original studies. In light of the importance of the Charness etal. (2012) study and the widelyaccepted need for replications of influential empirical results, our study can be best described as a “scientific replication,” in terms of the classification of Hamermesh (2007): It draws a sample from a different population, but uses the same estimation strategy as the original study. Because we conduct our study online, rather than in an on-site behavioral laboratory, both the population and the sample are different from the original study. However, the specification remained the same, as the original experiment was precisely reprogrammed and not modified. Based on their study results, Charness etal. (2012) concluded that wage delegation significantly enhances workers performance and increased earnings for both firms and workers. Interestingly, Charness et al. (2012) found that the increased effort was not primarily motivated by positive reciprocity; rather, Charness etal. (2012) postulated that it might have reflected a heightened sense of responsibility that workers felt towards their outcomes. 240 H.Niehoff, P.Schreck The results of our replication study confirm that wage delegation does indeed increase worker performance. However, there are important differences in the results of the original and the replication study. Wage delegation only appears to positively affect worker effort in one-shot interactions (stranger matching protocols). In repeated interactions (partner matching protocols), the increased effort appears to be driven solely by higher wages and the higher efforts desired by employers. Additionally, the results indicate that in the stranger scenario, the deliberate decision by the employer not to engage in wage delegation negatively impacts worker productivity levels. A similar negative reaction to non-delegation was observed by Franke etal. (2016), who investigated varying degrees of workers’ involvement in the wage-setting process using a laboratory gift-exchange game. Finally, in contrast to the original study, our data do not rule out the possibility that higher worker effort levels in response to wage delegation are partly due to positive reciprocity. This result also contrasts with subsequent studies in the field of wage delegation, which, like Charness etal. (2012), were able to rule out the effect of positive reciprocity as an explanation for the higher effort under delegation (Jeworrek and Mertins 2019). 2 Study design 2.1 Research question anddesign oftheoriginal study The fundamental question that Charness etal. (2012) asked is this: How do employers and employees alter their behavior when presented with the option to delegate wage decisions? To answer this question, Charness etal. (2012) conducted a laboratory experiment in Granada, Spain, involving 236 student participants. Participants were assigned fixed roles as either a firm or a worker, and performed a stated effort task in groups of two. Workers could provide costly effort which would increase firm profit; the firm, in turn, paid out a fixed wage to the worker. More formally, firm and worker payoffs were determined by the following profit functions: where F denotes the firm, W the worker, e is the worker’s effort level, and w the wage with w∈[20, 21, …119, 120] . c(e) represents the cost of effort (increasing in e). Table1 lists the feasible effort levels and corresponding costs. In essence, the setting represents a gift-exchange game: The worker needs to provide (costly) effort to increase the firm’s payoff, while the firm needs to provide a (costly) wage to increase the worker’s payoff. To introduce the idea of wage delegation, employers in some treatments could choose whether to set the employee’s wage by themselves or whether to allow the employee to determine his own wage. The experiment consisted of 15 periods and 2 initial test rounds. At the start of each period, participants in the role of the firm (the employer) had to specify an expected, non-binding effort from the worker, and decided whether to set the (1) ΠF=(240 −w)×e (2) ΠW= w − c ( e )−20 241 Re‑examining theeffect ofwage delegation: areplication… worker’s wage or whether to delegate this decision to the worker. The worker was first informed of the expected effort level and the employer’s wage-setting decision. If the employer chose to delegate, the worker could set his own wage. If, in contrast, the employer chose to set the wage by herself, the worker was informed of the wage. Finally, the worker had to state the level at which he was willing to exert effort. The experiment was implemented in a 2×2+1 factorial design, resulting in five treatments. Participants were unique to each treatment and session, with two sessions per treatment and between 22 and 24 individuals per session. The wage setting mechanism (delegation yes/no) and the matching procedure (partner v. stranger) were manipulated at two levels, each. In the Partner treatments, firms and workers were anonymously paired for all periods, while in the Stranger treatments, pairs were randomly rematched each period. In the Control treatments, the firm had to set wages without having the option to delegate this decision to the worker. The fifth treatment, the Dictator treatment, used a stranger matching protocol and implemented delegation as the norm: here, only the worker was able to set the wage. The experimental design allows to measure the effect of delegation on wages, effort levels, and profits, depending on whether delegation was (endogenously) chosen by the firm, or exogenously set in the specific treatment. 2.2 Design ofthereplication study The goal of our study is to determine whether the results of Charness etal. (2012) are replicable which would lend further support to its validity and generalizability. Accordingly, we chose an experimental design that would come as close as possible to the original study. In terms of the classification of Hamermesh (2007), our study can be best described as a “scientific replication:” It draws a sample from a different population, but uses the same estimation strategy as the original study. In particular, the specification remained unchanged, as the original experiment was precisely reprogrammed and not modified. As summarized in Table2, our study differs from the original study in four aspects that are not related to the specification. Table 1 Effort levels and costs of effort Effort e Cost c(e) 0.1 0 0.2 1 0.3 2 0.4 4 0.5 6 0.6 8 0.7 10 0.8 12 0.9 15 1.0 18 242 H.Niehoff, P.Schreck In the original study by Charness etal. (2012), a total of 236 participants were involved; however, the authors provided no rationale for this specific sample size. To determine the necessary sample size for our replication study, we performed an ex-ante power analysis. The first step in this analysis involved determining the effect size for each effect mentioned by Charness etal. (2012). This was achieved by initially calculating the pooled standard deviation for each effect using the original data. This pooled standard deviation, in combination with the group means reported in the paper, enabled the calculation of Cohen’s d for each effect. In the next step Cohen’s d was used to determine the necessary sample size for each effect in G*Power (Faul etal. 2007) with the alpha error set at the conventional threshold of 0.05 and the power set to 95% in order to detect the original effect size at the 5% significance level. The analysis yielded a required sample size of 80 for each treatment. In our effort to reach this minimum size, we ended up with a total of 410 participants. We recruited the participants of our experiment using Prolific, a popular subjects pool for online academic studies, which offers a large and diverse base of subjects (Peer etal. 2017; Palan and Schitter 2018). The original experiment was a computerized laboratory experiment, programmed in z-Tree (Fischbacher 2007). We used the original z-Tree files made available online by Charness etal. (2012) to accurately replicate the experiment using the Lioness platform (Giamattei etal. 2020). This replication ensured that all aspects of the experiment, including the screens and other elements, were identical to the original computerized experiment. The experiment was conducted on four consecutive days with at least two sessions per treatment. In line with the design from Charness etal. (2012), no participant was able to participate in more than one session or treatment. In response to dropout rates of more than 70% in our pilot study, we decided to slightly adjust the number of rounds in our replication study. Attrition in online experiments is a well-known issue (Arechar etal. 2018; Giamattei etal. 2020), but given the specific study design, we had to deal with an unusually high dropout rate. In treatments with a partner matching process, each participant who dropped out in any of the 15 periods, left an unmatched partner who had to leave the experiment prematurely. In treatments with a stranger matching process, participants were matched within cohorts. These had to be large enough to allow for stranger matching, and small enough to avoid excessively long waiting times, which would have resulted in higher dropout rates. Hence, a single participant’s dropout would end the experiment for the entire cohort because others who were intended to be matched with this specific participant in later stages of an Table 2 Differences in the designs of the original and the replication study Categories Charness etal. (2012) Replication study Sample size 236 410 Experimental method Laboratory experiment Online experiment Sample Students from the University of Granada Prolific Pool (UK) Number of periods (without training periods) 15 periods 10 periods 243 Re‑examining theeffect ofwage delegation: areplication… experimental session, could not be matched anymore. To mitigate these problems of attrition, we decided to limit our experiment to 12 rounds, including 2 training rounds, and chose a cohort size of four. We thus implemented an imperfect stranger matching where no participant would be matched with the same partner in two consecutive rounds.1 Participants were informed about being rematched after each round, but were unaware of the size of the cohort. We can hence assume that participants could not identify repeated interaction and reputation building was not possible. Taken together, these measures led to a decrease in dropout rates to 45%. 3 Results oftheoriginal study andthereplication study 3.1 Delegation effects oneffort andearnings Table3 provides mean values of the main variables across the five treatments for the original study and the replication study. For each of the two treatments where delegation was possible, the mean values are presented in two separate columns, depending on whether the wage decision was made by the employer or delegated to the employee in a given period (Partner nondelegation v. Partner delegation; Stranger nondelegation v. Stranger delegation). First we turn to the results of Charness etal. (2012). In the Partner treatment, wages and actual effort levels were higher when firms delegated (Partner delegation) than when they did not delegate (Partner nondelegation).2 The comparison of the Partner nondelegation case with the Control Partners treatment, where no option to delegate existed, showed little difference in wage and effort levels. This lack of difference suggests that nondelegation—where it was possible—had no demotivating effect per se. Under the stranger matching protocol, results were very similar. Delegation resulted in workers selecting higher wages and more effort. That is, positive effects of delegation persist, even if firms and workers cannot act strategically. However, in the Stranger treatment, workers exerted less effort compared to the Partner treatment, highlighting the role of enduring employment relationships in influencing worker behavior. These results suggested that workers may increase their effort when they consider long-term strategic factors. Charness etal. (2012) suggested that the increased effort in delegated scenarios might have been due to two distinct factors: First, it could have stem from workers’ positive response to delegation or from their negative reaction to intentional nondelegation. Second, it could have been due to higher wages that workers set for themselves if they have the chance to do so. To disentangle these effects, for each of the treatments where delegation was possible, Charness etal. (2012) regressed effort on 1 The article by Charness etal. (2012) does not provide a definitive clarification on whether the stranger matching process used in the experiment is a perfect or imperfect stranger mechanism. 2 If not stated differently, all differences discussed in this section are significant in a statistical sense ( p≤0.001 ). To test for statistical differences, Charness etal. (2012) used two-tailed Wilcoxon signedranks tests. 244 H.Niehoff, P.Schreck a set of explanatory variables, including wage, desired effort, a delegation dummy (taking value 1 if the firm chose to delegate in a given period, and 0 otherwise), and period.3 As the results reported in Table 4 reveal, even when controlling for wage effects, delegation significantly and positively influenced effort in both matching scenarios. Finally, delegation benefited both firms and workers by leading to higher total earnings. Specifically, in the Partners delegation scenario, total earnings, and hence social efficiency, were 50% higher compared to the Partners nondelegation setting. Table 3 Summary of the five experimental treatments (Charness etal. 2012 and replication results)a a The abbreviations in each column include "P" or "S", denoting the partner and stranger treatments, and "D" or "ND," indicating whether the scenarios with delegation or nondelegation were examined. "Wage" refers to the average wage set by either the firm or the worker across all periods. "Desired effort" denotes the average effort requested by firms, while "actual effort" represents the effort workers are willing to provide. "Firm earnings" and “Worker earnings” indicate the average profit earned by both workers and firms based on a predefined profit function across all periods. "Total earnings" is the cumulative profit achieved on average by both firm and worker over all periods Partners PND Partners PD Partners control Strangers SND Strangers SD Stranger control Dictator Wage Original 76.79 114.70 75.95 45.60 117.85 51.24 113.05 Replication 69.68 106.06 70.86 58.21 104.34 58.29 92.21 Desired effort Original 0.78 0.81 0.72 0.61 0.67 0.59 0.68 Replication 0.69 0.68 0.68 0.71 0.74 0.69 0.69 Actual effort Original 0.43 0.78 0.46 0.15 0.34 0.17 0.36 Replication 0.47 0.65 0.51 0.42 0.63 0.47 0.51 Firm earnings Original 64.29 84.62 68.36 27.53 41.91 30.24 45.93 Replication 72.50 85.77 80.82 74.53 85.88 86.28 78.32 Worker earnings Original 51.09 84.15 50.01 24.93 93.81 30.35 88.96 Replication 43.46 76.30 43.92 32.62 75.36 35.87 73.40 Total earnings Original 115.38 168.77 118.37 52.46 135.72 60.59 134.88 Replication 115.96 162.07 124.74 107.15 161.24 122.15 151.72 3 As we will explain in more detail further below, Charness etal. (2012) estimated Generalized Least Squares (GLS) random effects models. 251 Re‑examining theeffect ofwage delegation: areplication… Table 6 Replication success for main findings Finding Method used Replication sucess RESULT 1 Wage delegation increases worker performance Mean comparison: Effort PND vs PD and SND vs SD (Wilcoxon matched-pairs signed-rank tests) Yes Wage delegation increases earnings Mean comparison: total, firm, and worker earnings PND vs PD and SND vs SD (Wilcoxon matched-pairs signed-rank tests) Total earnings: Yes Firm earnings: Partially (only in Stranger treatment) WorkerEarnings: Yes The wage delegation effect exists even when controlling for wage effects GLS Random effects regression Partner vs Stranger (Explanatory variables: Wage, Desired effort, and a Delegation dummy) Yes (Stranger) No (Partner) RESULT 2 There is no negative reaction to "intentional nondelegation" GLS Random effects regression PND vs Partner Control SND vs Stranger Control (Explanatory variables: Wage, Desired Effort, and a Nondelegation dummy) Yes (Partner) No (Stranger) RESULT 3 Increased effort is not due to positive reciprocity, but seems to reflect the increased responsibility of workers Mean comparison: Effort SD vs Dictator (Wilcoxon-Mann–Whitney-Test) No 252 H.Niehoff, P.Schreck In order to address the question raised by Charness et al. (2012) regarding whether higher worker effort in delegation cases is due to positive reciprocity, we compare the stranger and dictator treatments from our replication study. Table 3 clearly shows that the effort level in the Stranger treatment under delegation (0.63) is higher than in the Dictator treatment (0.51). This difference is statistically significant (Z = 3.318, p = 0.001, two-tailed test). Thus, the results indicate that unlike in Charness etal. (2012), positive reciprocity might indeed be a factor influencing worker behavior in terms of their effort under delegation. Considering how firms chose delegation over the course of 10 periods, it becomes apparent that, in contrast to the findings of Charness etal. (2012), the replication study does not show a significant increase in delegation over time. A Spearman test indicates no significant upward trend in both the Partners (ρ = −0.044, p = 0.371) and Strangers (ρ = 0.007, p = 0.852) scenarios. 4 Discussion & conclusion As summarized in Table6, we use the main results of Charness etal. (2012) to structure our discussion of the degree to which we were able to successfully replicate the original study. RESULT 1: Delegating the wage decision enhances worker performance and increases the earnings of both firms and workers relative with the case where firms do not delegate. We were able to largely confirm the first result of Charness etal. (2012), showing that delegation leads to increased worker effort in both the Strangers and Partners treatments. Consistent with Charness etal. (2012), our GLS regression analysis validated the impact of delegation on effort levels in the Stranger treatment. However, this effect was not observed in the Partner treatment, which is unexpected given that firms and workers could more easily engage in strategic decision making: workers display higher effort levels in response to delegation to increase the likelihood of wage delegation in the next round. Regarding earnings, we were able to confirm all but one of Charness et al. (2012)’s findings. In the case of the Partner treatment, the replication study’s results did not show a significant difference in the average earnings of firms between cases of delegation and nondelegation. RESULT 2: The increment in the effort level with delegation is not due to a negative reaction to nondelegation, but instead seems due to the positive effect of delegation, controlling for the wage receive. Our GLS regression revealed that in the Partners treatment we were able to replicate the results of the original study. That is, there was no negative worker response to intentional nondelegation in this case. In contrast, we did observe a negative effect of nondelegation in the Strangers treatment, although this effect was very small compared to the influence of wage and desired effort. RESULT 3: The higher observed effort when the worker has been given the freedom to choose the wage is not due to positive reciprocity, but seems to reflect the increased responsibility that the worker has for the final outcome. 253 Re‑examining theeffect ofwage delegation: areplication… We failed to replicate this result. When comparing effort levels across the dictator and the stranger delegation treatments, we did find significant differences which Charness etal. (2012) did not observe. So, in contrast to the original study, positive reciprocity may indeed have been an influencing factor in our sample. Taken together, the above described results prompt the question why some of our results are inconsistent with those of the original study. It is particularly striking that in the original study the effect of delegation on effort is so much stronger in the Partners treatment (see Table3 and Figs.1, 2, 3, 4). On one interpretation, either our result is a type II error (false negative), or the result of the original study is a type I error (false positive). To further evaluate the validity of this argument, additional replication studies would be necessary. On an alternative interpretation, the observed inconsistencies are due to the increased social distance between participants which our experimental design caused and which may have prevented a sense of togetherness among workers and employers. While wage disparities remain comparable to those in the original study, the differences in effort levels are substantially smaller in our study. This suggests that with the social distance induced by our online setting, workers were inclined to increase their own compensation when they could set wage levels by themselves, yet they did not strategically choose higher effort levels to the same degree as in the original study. Employers, in turn, could not establish trust and did not increase the rate of delegation over time. We hence did not observe a significant trend in the practice of delegation, which could have served as an indicator of the development of mutual trust. Given these tendencies, wage delegation did not increase firm earnings in our setting. These results contrast with Charness et al. (2012: 2369), who suggest that wage delegation may cause an enhanced sense of responsibility among workers, leading to a kind of virtuous circle: delegation leads to higher efforts and increased earnings for both workers and firms. Our results, on the other hand, rather support the notion that formal controls become more important as the social distance typical of remote work increases (Schmelz and Ziegelmeyer 2020). As a final interpretative effort, we consider whether the observed inconsistencies could be attributed to differences in sample composition. Notably, our study diverges from the original study by recruiting participants from the UK via Prolific, while the participants in Charness etal. (2012) were students in Spain. Expanding the participant pool to include a general population via an online participation platform offers the advantage of enhanced external validity. Also university-based participation pools have been characterized as "WEIRD" – Western, Educated, Industrialized, Rich, and Democratic (Henrich etal. 2010). On the other hand the educational qualifications between our sample and that of the original study are remarkably similar, with over 70% of our participants possessing a tertiary education degree. Also, previous experimental research indicates that the behavior observed in student samples in economic experiments is broadly in line with that exhibited by other demographic cohorts (Exadaktylos etal. 2013). Thus, the variation in the sample composition may be considered to only have a marginal effect relative to the impact of social proximity. 254 H.Niehoff, P.Schreck Our findings provide important avenues for further research. It would be interesting to conduct the experimental procedure in both an online environment and a physical laboratory setting, concurrently procuring analogous samples from an identical population (Hergueux and Jacquemet 2015). In the online experiment, the potential for high attrition rates could possibly be mitigated by conducting the experiment simultaneous for all participants, thereby reducing variations in participant engagement and ensuring a consistent application of experimental conditions (Mason and Suri 2012). Moreover, further qualitative research could facilitate a more profound understanding of the underlying reasons behind the observed phenomenon of employers delegating less and employees demonstrating less effort. In conclusion, wage delegation seemingly requires social proximity to be effective. These differences warrant further investigation in subsequent experiments to determine why exactly remote work seems to complicate trust-based controls. Author contributions This paper is co-authored by Hendrik Niehoff (HN) and Philipp Schreck (PS). All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by HN and PS. The first draft of the manuscript was written by HN and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding Open Access funding enabled and organized by Projekt DEAL. We acknowledge financial support by the Friede Springer Stiftung. Data availability The data that support the findings of this study are available at openICPSR: https:// doi. org/https:// doi. org/ 10. 3886/ E2092 06V1. Declarations Conflict of interest The authors have no relevant financial or non-financial interests to disclose. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. References Albers S (2014) Preventing unethical publication behavior of quantitative empirical research by changing editorial policies. J Bus Econ 84:1151–1165. https:// doi. org/ 10. 1007/ s115730130675-3 Arechar AA, Gächter S, Molleman L (2018) Conducting interactive experiments online. Exp Econ 21:99–131. https:// doi. org/ 10. 1007/ s106830179527-2 Bloom N, Barrero JM, Davis S, etal (2023) Survey: Remote Work Isn’t Going Away–And Executives Know It. Harvard Business Review. https:// hbr. org/ 2023/ 08/ surveyremoteworkisntgoingawayandexecu tivesknowit 255 Re‑examining theeffect ofwage delegation: areplication… Brück C, Knauer T, Meier H, Schwering A (2021) Self-set salaries and creativity. J Bus Econ 91:91–121. https:// doi. org/ 10. 1007/ s1157302000985-z Brüggen A, Maas VS, Van Den Abbeele A (2021) Replication in experimental management accounting research - an editorial. J Manag Control 32:155–160. https:// doi. org/ 10. 1007/ s0018702100324-5 Burbano VC, Chiles B (2022) Mitigating gig and remote worker misconduct: evidence from a real effort experiment. Organ Sci 33:1273–1299. https:// doi. org/ 10. 1287/ orsc. 2021. 1488 Camerer CF, Dreber A, Forsell E etal (2016) Evaluating replicability of laboratory experiments in economics. Science 351:1433–1436. https:// doi. org/ 10. 1126/ scien ce. aaf09 18 Chang A, Li P (2022) Is economics research replicable? sixty published papers from thirteen journals say “often not.” CFR 11:185–206. https:// doi. org/ 10. 1561/ 104. 00000 053 Charness G, Haruvy E, Sonsino D (2007) Social distance and reciprocity: an Internet experiment. J Econ Behav Organ 63:88–103. https:// doi. org/ 10. 1016/j. jebo. 2005. 04. 021 Charness G, Cobo-Reyes R, Jiménez N etal (2012) The hidden advantage of delegation: pareto improvements in a gift exchange game. Am Econ Rev 102:2358–2379. https:// doi. org/ 10. 1257/ aer. 102.5. 2358 Charness G, Cobo-Reyes R, Lacomba JA etal (2016) Social comparisons in wage delegation: experimental evidence. Exp Econ 19:433–459. https:// doi. org/ 10. 1007/ s106830159448-x Cramton CD (2001) The mutual knowledge problem and its consequences for dispersed collaboration. Organ Sci 12:346–371. https:// doi. org/ 10. 1287/ orsc. 12.3. 346. 10098 Cramton CD, Webber SS (2005) Relationships among geographic dispersion, team processes, and effectiveness in software development work teams. J Bus Res 58:758–765. https:// doi. org/ 10. 1016/j. jbusr es. 2003. 10. 006 Ditzen J, Elhorst JP (2022) Introducing the Replication Studies section. Spat Econ Anal 17:7–9. https:// doi. org/ 10. 1080/ 17421 772. 2022. 20181 69 Duvendack M, Palmer-Jones R, Reed WR (2017) What is meant by “replication” and why does it encounter resistance in economics? Am Econ Rev 107:46–51. https:// doi. org/ 10. 1257/ aer. p2017 1031 Exadaktylos F, Espín AM, Brañas-Garza P (2013) Experimental subjects are not different. Sci Rep 3:1213. https:// doi. org/ 10. 1038/ srep0 1213 Faillo M, Piovanelli C (2017) Wage delegation and intrinsic motivation: an experimental study. CEEL Working Papers Faul F, Erdfelder E, Lang A-G, Buchner A (2007) G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 39:175–191. https:// doi. org/ 10. 3758/ BF031 93146 Fischbacher U (2007) z-Tree: Zurich toolbox for ready-made economic experiments. Exp Econ 10:171– 178. https:// doi. org/ 10. 1007/ s106830069159-4 Franke J, Gurtoviy R, Mertins V (2016) Workers’ participation in wage setting: a gift-exchange experiment. J Econ Psychol 56:151–162. https:// doi. org/ 10. 1016/j. joep. 2016. 07. 002 Giamattei M, Yahosseini KS, Gächter S, Molleman L (2020) LIONESS Lab: a free web-based platform for conducting interactive experiments online. J Econ Sci Assoc 6:95–111. https:// doi. org/ 10. 1007/ s4088102000087-0 Hamermesh DS (2007) Viewpoint: Replication in economics. Can J Econ/revue Can D’économique 40:715–733. https:// doi. org/ 10. 1111/j. 13652966. 2007. 00428.x Harrison JS, Freeman RE (2004) Is organizational democracy worth the effort? Acad Manag Exec 1993–2005(18):49–53 Henrich J, Heine SJ, Norenzayan A (2010) The weirdest people in the world? Behav Brain Sci 33:61–83. https:// doi. org/ 10. 1017/ S0140 525X0 99915 2X Hensel PG (2019) Supporting replication research in management journals: Qualitative analysis of editorials published between 1970 and 2015. Eur Manag J 37:45–57. https:// doi. org/ 10. 1016/j. emj. 2018. 03. 004 Hergueux J, Jacquemet N (2015) Social preferences in the online laboratory: a randomized experiment. Exp Econ 18:251–283. https:// doi. org/ 10. 1007/ s106830149400-5 Jeworrek S, Mertins V (2019) Wage delegation in the field. J Econ & Manage Strategy 28:656–669. https:// doi. org/ 10. 1111/ jems. 12313 Köhler T, Cortina JM (2021) Play it again, Sam! an analysis of constructive replication in the organizational sciences. J Manag 47:488–518. https:// doi. org/ 10. 1177/ 01492 06319 843985 Maniadis Z, Tufano F, List JA (2015) How to make experimental economics research more reproducible: lessons from other disciplines and a new proposal. Replication in experimental economics. Emerald Group Publishing Limited, pp 215–230 256 H.Niehoff, P.Schreck Mason W, Suri S (2012) Conducting behavioral research on Amazon’s Mechanical Turk. Behav Res 44:1–23. https:// doi. org/ 10. 3758/ s134280110124-6 Maynard MT, Gilson LL, Mathieu JE (2012) Empowerment—Fad or fab? A multilevel review of the past two decades of research. J Manag 38:1231–1281. https:// doi. org/ 10. 1177/ 01492 06312 438773 Mellizo P, Carpenter J, Matthews PH (2014) Workplace democracy in the lab: workplace democracy in the lab. Ind Relat J 45:313–328. https:// doi. org/ 10. 1111/ irj. 12054 Mueller-Langer F, Fecher B, Harhoff D, Wagner GG (2019) Replication studies in economics—How many and which papers are chosen for replication, and why? Res Policy 48:62–83. https:// doi. org/ 10. 1016/j. respol. 2018. 07. 019 Open Science Collaboration (2015) Estimating the reproducibility of psychological science. Science 349:aac4716. https:// doi. org/ 10. 1126/ scien ce. aac47 16 Palan S, Schitter C (2018) Prolific.ac—A subject pool for online experiments. J Behav Exp Financ 17:22–27. https:// doi. org/ 10. 1016/j. jbef. 2017. 12. 004 Peer E, Brandimarte L, Samat S, Acquisti A (2017) Beyond the Turk: Alternative platforms for crowdsourcing behavioral research. J Exp Soc Psychol 70:153–163. https:// doi. org/ 10. 1016/j. jesp. 2017. 01. 006 Rammstedt B, John OP (2007) Measuring personality in one minute or less: a 10-item short version of the Big Five Inventory in English and German. J Res Pers 41:203–212. https:// doi. org/ 10. 1016/j. jrp. 2006. 02. 001 Schmelz K, Ziegelmeyer A (2020) Reactions to (the absence of) control and workplace arrangements: experimental evidence from the internet and the laboratory. Exp Econ 23:933–960. https:// doi. org/ 10. 1007/ s1068302009666-8 Semler R (2007) Out of this world: doing things the Semco way. Glob Bus Org Exc 26:13–21. https:// doi. org/ 10. 1002/ joe. 20161 Tuna C (2008) Pay, your own way: firm lets workers pick salary. Wall Street J 7:86 Yin Y, Wang Y, Lu Y (2019) Antecedents and outcomes of employee empowerment practices: a theoretical extension with empirical evidence. Hum Resour Manag J 29:564–584. https:// doi. org/ 10. 1111/ 17488583. 12243 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.