Factors influencing leadership effects on team performance: A case study on Chinese Manufacturing Plants
Chinese Manufacturing : A Case Study By studying the data of the detailed daily production data at each individual worker level along with each worker’s personnel records at a manufacturing firm in China. Evidence exists that under contingent and non-contingent rewards to team leaders, team leaders are motivated to make changes and make efforts to boost team performance. This is accomplished hrough the channel of encouraging workers to work longer hours. In addition, spending more time on administration, which involves meetings, training, workshop, and cooperation within team members. No evidence has yet to show that productivity is substantially improved under contingent and noncontingent rewards. Potentially due to the limited length of the treatment periods.
Managing productivity of workers largely determines the success of a firm. With similar set-up, environment, and capital input, what helps firms stand out is the way they manage their employees. An efficient and effective way of management could bring huge benefits to the firms without putting in extra effort. Team leaders play a critical role in the management process.
Good team leaders are great assets to both employer and their peer employees.
They could be inspirational and motivational to other workers. In addition, they could allocate work to different individuals in an appropriate way to boost the overall productivity of the team. Furthermore, they can even be the bridge between employers and employees and facilitate the communication between workers in different positions.
Considering the important roles that team leaders play, it is important to figure out the best way to motivate employees to take up the responsibility as team leaders. Non-contingent reward and contingent reward are two critical ways to motivate the team leaders. A lot of the studies have shown that under such reward, team leaders tend to be more hardworking in their exclusive work as being team leaders. In this paper, I try to find different factors that could affect the leadership effects on team performance. Specifically I try to work on the dynamics of team leaders and team members. In addition, how their performance and behavior changes under contingent and noncontingent rewards to team leaders.
Literature Review
Firstly, it is critical to recognize the function of team leaders. Lazear (2011) has talked about the important role team leaders play on overall team performances. They are decision makers, who lead teams to right directions; they communicate with workers and motivate them
to improve productivity; they convey knowledge and share wisdom with other employees; and they have followers who believe in their strategies and leading skills. Most importantly, they are generalists rather than specialists who are good at confronting a wide variety of choices which span many fields. The effects of leadership on team performance is argued to be very significant in many studies. Jacobs & Singell (1993) identified the effects of team leaders on overall team performances by examining managers of professional baseball—holding other players’ contributions constant, better managers win more often with the same players’ performances and their tactical skills is significant in determining team success. In facilitating the positive role of team leaders in raising team performance, the literature has been paying particular attention to incentive for leaders.
There are several channels through which team leaders can boost overall performance among teams. Brandiera et al. (2007) argued that a team leader can affect his/her team performance via two channels: taking actions that influence productivity of existing team members by targeting their effort towards the most able team members, named “target effect”. In addition, affecting the composition of his/her team and raising the proportion of high-ability team members, which is labeled “selection effect”. They conducted a case study of fruit pickers and provided evidence on both target and selection effects. Burke et al. (2006) found out that compared to task-focused behaviors, person-focused behaviors are more related to perceived team effectiveness and productivity during his analysis to examine the relationship between leadership behavior in teams and behaviorally-based team performance outcomes.
A lot of research has been focused on how leadership affects change according to rewards and gift-exchange. Podsakoff et al. (1985) discovered that workers respond to leadership differently, depending on the leader’s reward system: leader contingent reward behavior is likely related to group drive, cohesiveness and group productivity while non contingent reward behavior is negatively related to these group criteria. Deci (1972) reported that when a person receives contingent monetary payments. His or her intrinsic motivation to perform an activity decreases. So does it decrease when the person receives threats of punishment for poor performance. Or negative feedback of his or her performances.
Non-contingent monetary payments, on the other hand, don’t change a person’s intrinsic motivation, and verbal reinforcements seem to enhance intrinsic motivation. Padsakoff et al (1982) have had a contrary opinion on contingent rewards. They believe that contingent rewards have the most pronounced relationship with subordinate performance. It’s also related to subordinates’ expressions of satisfaction with their work, supervision, and advancement opportunities. Similarly, Hunt and Osborn (1981) state that contingent rewards produce more desirable subordinate behavior than non-contingent rewards. Cherrington (1971)’s paper supports the hypothesis that the nature and magnitude of the relationship between satisfaction and performances rely on performance-reward contingencies arranged.
Field experiments have been conducted to test out the theory that non-contingent reward increases team performance.
Ferh et al (1993) first conducted an experiment to prove the fair wage-effort hypothesis, which stipulates that wage increases raise workers’ effort levels even in the absence of any increase in the penalty for shirking. To mimic the labor market, they designed a two-stage game to observe the behavior of the employers and employees. Although the mimic nature of the experiment has brought about limitations to the study, the results still provide evidence for the theory: paying high wages will, on average, induce workers to reciprocate, which is choosing high effort levels.
Gneezy and List (2006) have further optimized the experiment by focusing on real effort in labor markets using field experiments. According to them, one key missing link in the literature between the laboratory and labor markets is the duration of the task. Interaction in the labor market lasts way longer than the lab experiments. People’s decision making behavior has a “colder” and “hotter” phase, so adaptation is found to be critical when discussing gift-exchange behavior. After conducting two field experiments, they have made the following two conclusions: firstly, employees do reciprocate by greater effort during early hours of the tasks. Secondly, this higher wage level is shown not to be consistent, which is that workers have exerted less effort in the gift treatment period after a few hours.
Rao et al have decided to further pursue the experiments and are working on the same issue, but incorporating the limitations that were previously not discussed. Their design goal is to keep the advantageous features of the gift-exchange experiments in the field, while allowing for variation in the return to the firm and in the worker’s piece rate. They have also incorporated estimation of the social preferences in their models:the value to the employer of the worker’s effort and the unknown cost of effort function, by varying the piece rate.
The results of the study have shown evidence that gift exchange in the workplace is likely to be of more limited impact than initially conceived. They give out two possible interpretations for the results: the first explanation is that gift exchange effects are simply of smaller magnitude than found in some of the earlier papers. The second interpretation is that the social preferences towards the employer are, to a first approximation, set at first contact.
Data and Experimental Setup
Detailed daily production data at each individual worker level along with each worker’s personnel records at a manufacturing firm in China were collected for October 1, 2008-April 30 of 2010. The firm has two comparable plants, and one plant is used as the treatment plant and the other as the control plant. All production workers in both plants are divided into production teams, and each team has a team leader. All production workers including team leaders are paid by piece rate (each worker’s wage depends on his/her individual productivity). In addition to piece-rate compensation, each team leader receives leadership pay.
The first period is control period. For the first nine months, in both the control and the treatment, the amount of leadership pay was fixed at 100 RMB per month for all team leaders. The second period is a noncontingent reward period. On July 1, 2009, CEO of the company announced that all team leaders would receive an additional leadership pay of 200 RMB on top of the original leadership pay of 100 RMB in the treatment plant, while no such change was made in the control plant. CEO of the firm made it clear that the additional leadership pay of 200 RMB in the treatment plant was not contingent on anything.
The third period is a contingent reward period. On January 1, 2010, CEO made another announcement to all workers in the treatment plant that the amount of the additional leadership pay which was fixed at 200RMB would be now tied to team performance. Team leaders would be paid 180RMB, 190RMB, 200RMB, 210RMB, 220 RMB based on the performance of the team each month. In sum, each team leader in the treatment plant received a fixed payment of 200 RMB in addition to the original leadership pay of 100 RMB from July 1 of 2009 through Jan 31 of 2010. The fixed payment of the additional leadership compensation of 200 RMB was changed to performance pay which is tied to team performance on January 31, 2010. The data collection continued for three more months till the end of April of 2010.
Most of the variables used are directly extracted from the original dataset. Region is 0 when observations are from the treatment plant and is 1 when observations are from the control plant. Teamleader is 1 when the worker is a team leader, and is 0 when the worker is an average team member. Since different auto parts take different lengths of time to make during the production, each unit is worth a certain amount of time that is determined by managers in production plants. Earnedtime is referred to how much a worker is paid by piece rate every single day. It is measured in minutes. Worktime is the amount of time workers spend working in the manufacturing plants every day.
Other variables are calculated from the existing variables in the dataset. Timeperiod1 is a dummy variable that is evaluated to be 1 when it is the first treatment period and 0 otherwise. Timeperiod2 is a dummy variable that is evaluated to be 1 when it is the second treatment period and 0 otherwise.
Workerproductivity is given by the sum of the planned time of all the parts a worker produces divided by the sum of the actual time a person spends on all the parts that he or she produces every day. It is close to one when the worker is as productive as the managers have expected, in other words, when the worker spends a similar amount of time on each unit’s production that is estimated by the managers of the manufacturing plants. It is larger than one when the worker is more productive than estimation of the managers, which is when the worker spends less time on units’ production than the expected time, vice versa when worker productivity is less than one.
Output is the sum of the planned time of all the parts a worker produces per day. It measures how much a worker produces every day in minutes. Rest is calculated by dividing the difference of work time and the actual amount of time a worker spends on production every day by work time. It measures the proportion of time a worker spends not working on his or her own production in a day. Since workers’ are paid by piece rate, workers are mostly not motivated to shirk during their work time in the manufacturing plant. Therefore, the increase in rest for workers can mostly be attributed to cooperation, administration, and other affairs that are not part of their production.
Hypotheses
Hypothesis 1
Contingent reward and non-contingent reward to team leaders both bring about more benefits to team members than team leaders.
Intuitively, team leaders are the beneficiaries of the reward system. In the first treatment period, they get to be paid extra without bearing extra responsibilities. However, because workers in the plants are paid by piece rate, team leaders’ own output could be compromised due to their behavior of reciprocating by spending more time on allocating jobs to other workers. So overall, the extra reward to the team leaders probably could barely compensate for the time during which the team leaders could earn a piece-rate salary.
On the other hand, workers could benefit from the more productive allocation of jobs. They could earn more since they are able to produce more and they are paid by piece-rate. The same logic works for the second treatment period which is the contingent rewards period. The leaders may feel obligated to make more efforts to keep the team performances’ at a desirable level in order to obtain a higher team performance tied reward. This behavior could lead to the compromise of the team leader’s own production performances.
Thus, I predict that in both the non-contingent reward and contingent reward periods, workers benefit more from this reward system than team leaders.
Hypothesis 2
Output increases for workers during the contingent and non-contingent reward period mainly due to the increase in productivity, and the increase in output and productivity is more significant in the contingent reward period.
In the previous studies, leaders tend to pay more attention to their job of allocating workers in an appropriate way so that the overall team performance will be boosted under contingent and non-contingent reward. Increase in productivity for workers will lead to an increase in overall output. Thus, I hypothesize that both productivity and output will increase for workers. According to Padsakoff (1982), Hunt and Osborn (1981), contingent reward appears to be more effective and powerful than non-contingent reward. Therefore, I hypothesize that the increase in productivity and output will be even more obvious and significant in the second treatment period, which is the contingent reward period.
Hypothesis 3
Both team members and team leaders spend more time on work other than their own production, which includes helping each other out, having meetings, and other administration activities, in the non-contingent reward and the contingent reward period.
Either under contingent or non-contingent reward, team leaders in one way or another are motivated to exert more efforts to try and increase the overall team performances. It could involve holding meetings to inspire team workers in their production, encouraging or allocating workers to help each other out during the process of production, or the leader sparing extra time to instruct the fellow team leaders and sharing techniques in production. No matter which way is used, team members and team leaders may spare time for the extra administration efforts from team leaders stimulated by rewards.
Econometric Models
I try to explore how the amount of time earned by workers and team leaders change under the first and second treatment period. Workers are paid by piece rate: they are paid based on how much they produce every day , which is measured in time. Thus, earned time for every worker per day is a better measure than the actual salary, since it resolves the problem of inflation and it is a good indicator of how much a worker is paid, hence a good measure of whether workers and team leaders benefit from the contingent and non-contingent reward method.
The models are both fixed effect difference-in-difference models, controlling for individuals, teams and different months of the year. The interaction variables TreatmentPeriod1�*Regioni and TreatmentPeriod2� show the direct treatment effects of contingent and non-contingent reward periods. Because Region is valued to be 1 for the control plant in the original dataset, coefficients on TreatmentPeriod1�*Regioni and TreatmentPeriod2� , β4 and β5 show directly how the earned time changes for workers and team leaders in the control plant. Hence, – β4 and – β5 indicate the direct treatment effects of how the earned time of the treatment group changes under the first and second treatment period.
The next six models work on the second hypothesis. Here I study how the actual performances of team leaders and team members change under the non-contingent and contingent reward. I try to explore their performances by studying their productivity, output and working time in the manufacturing plants. Because I would like to test out the hypothesis whether the output would increase during the treatment periods and whether that could be attributed to increase in productivity, I also need to find out how the length of working time change under contingent and contingent rewards, since increase in output could be attributed to the increase in working length instead of the increase in productivity.
The control variables are the same as in Model 1 and Model 2. Controlling for individuals, teams and different months of the year could decrease the potential endogeneity of the treatment effects in case the change is induced by individuals or characteristics of different months of a year.
The last two models study the last hypothesis. The hypothesis states that team leaders in one way or another are motivated to exert more efforts to try and increase the overall team performances during both treatment periods. Team members and team leaders may spare extra time for increase in administration efforts from team leaders stimulated by rewards. In order to find whether they are spending more time on work that is not focused on their own production, I created a variable rest to explore their behavior under contingent and non-contingent rewards.
It is calculated by dividing the difference of work time and the actual amount of time a worker spends on production every day by work time. It measures the proportion of time a worker spends not working on his or her own production in a day. Since workers’ are paid by piece rate, workers are mostly not motivated to shirk during their work time in the manufacturing plant. Therefore, the increase in rest for workers can mostly be attributed to cooperation, administration, and other work that are not part of their production.
Results and Discussion
The first two models study whether team members benefit more than team leaders under contingent and noncontingent rewards. In Model 1, regression results show that in the first treatment period, the coefficient on the interaction variable Region*Treatment1 is -52.86, which indicates the direct treatment effect for the control plant for an average team member. Hence, this difference-in-difference model shows that an average worker of the team in the treatment plant is likely to earn 52.86 minutes more of piece rate during the first treatment period, which is the non-contingent reward period. In Model 2, under the first treatment period, the coefficient on the interaction variable Region*Treatment1 is -61.82, shows that a team leader in the treatment plant is likely to earn 61.82 minutes more of piece rate during the non-contingent reward period.
On top of that extra earned time, each team leader is earning 200 RMB of bonus for the treatment period, and an 100 RMB bonus as a team leader reward.
Thus, under non-contingent reward, the team leader not only enjoys a similar level of increase in earned time, but also enjoys an extra 200 RMB bonus for the treatment’s purpose.
This shows that team leaders’ own production is not compromised due to the first treatment period.
On the contrary, the team leader’s own production is improved. It could be explained such that increasing the output of the team leader himself or herself can also be considered a way of reciprocating. This can be achieved by either improving productivity or the length of the working time. The increased earned time for each worker has a pretty desirable result, as it is about 10% of the average earned time for a worker. The coefficient is statistically significant as well. Overall, the regression results disprove that team members benefit more than team leaders under non-contingent reward.
In the second treatment period, the coefficient on the interaction variable
Region*Treatment2 is 153.9 in Model 1, which indicates the direct treatment effect for the control plant for an average team member. This coefficient is statistically significant. Hence, this difference-in-difference model shows that an average worker of the team in the treatment plant is likely to earn 153.9 minutes less of piece rate during the first treatment period. The coefficient on the interaction variable Region*Treatment2 for team leaders is 69.68 in Model 2, and not statistically significant.
Thus, it is uncertain how the earned time changes under contingent reward for team leaders.
Considering all the past papers that show contingent reward induces more desirable results when it comes to subordinates’ performances, the results are quite surprising. I would expect that during the second treatment period, workers would have an even bigger boost in earned time, but the results turn out to be the contrary. This might be attributed to three reasons.
Firstly, the second treatment period lasts only three months, which could potentially bring about insufficient information.
Secondly, in the second treatment period, workers in the control plant might have an increase in earned time for some unknown reason, so that the coefficient in the different-in-difference model could show a contrary effect than expected.
Thirdly, the number of observations in the control plant is substantially smaller than that in the treatment plant, which could also potentially affect the results. Therefore, overall there isn’t significant evidence that team members benefit more than team leaders under contingent reward.
The next six models discuss the change in performances for team members and team leaders under the two treatment periods. I mainly focus on three variables: productivity, length of working time, and output. In Model 5, coefficient of the interaction variable Region*Treatment1 is -19.57 with statistical significance, showing an average worker of the team in the treatment plant is likely to have a 19.57 increase in output during the first treatment period. Model 6 shows that the coefficient of the interaction variable Region*Treatment1 is -36.86, which is statistically significant, indicating that a team leader in the treatment plant is likely to have a 36.86 unit of boost in output in the first treatment period.
Thus, the evidence provides support for the hypothesis that the output increases for both team leaders and team members under non-contingent reward.
Then, I study how productivity changes for team leaders and team members under non-contingent reward, and try to explore whether an increase in productivity causes the increase in output. According to Model 3, coefficient of the interaction variable Region*Treatment1 is 0.114 with statistical significance, showing an average worker of the team in the treatment plant is likely to have a 0.114 decrease in productivity during the first treatment period. In Model 4, coefficient of the interaction variable Region*Treatment1 is 0.0353 with no statistical significance. Therefore, the results don’t support the hypothesis that the output increases for workers during the non-contingent reward period.
To explore the behavior of team leaders and team members under non-contingent rewards, I also look at their working time’s change. In Model 7, coefficient of the interaction variable Region*Treatment1 is -126.2 with statistical significance, indicating that an average worker of the team in the treatment plant is likely to have a 126.2 minutes increase in length of the working time during the first treatment period. In Model 8, coefficient of the interaction variable Region*Treatment1 is -79.43 with statistical significance, which shows that a team leader in the treatment plant tends to have 79.43 minutes of increase in the working time in the manufacturing plant.
This completes the story for the increase in output for workers.
It is very likely that the increase in output is due to the increase in the working time for workers instead of increase in productivity. It may be the case that under the non-contingent reward period, the team leader reciprocates by not only working longer hours himself or herself, but also encouraging the fellow team members to also work longer hours.
After discussing the behavior of workers in the first treatment, now I am focusing on how performances change for workers in the second treatment period, which is the contingent reward period. Model 5 and 6 show that the coefficients of the interaction variable Region*Treatment2 are 151.3 and 93.44, indicating that the output decreases for both workers in the contingent reward period. A similar result is produced for productivity. Model 3 and 4 show that the coefficients of the interaction variable Region*Treatment2 are 0.264 and 0.0342, indicating that the productivity decreases for both workers in the contingent reward period.
Length of working time, on the hand, increases for both team leaders and team members during the contingent reward period, but not as significantly as in non-contingent reward period. Model 7 and 8 show that an average worker in the treatment plant is likely to have an increase of 49.07 minutes of working time during the second treatment period, and a 26.23 minutes increase for a team leader in the treatment plant during the second treatment period. Therefore, there isn’t enough evidence to bolster the hypothesis that productivity and output increases for workers under the contingent reward.
The last hypothesis tests whether team members and team leaders spend more time on work other than their own production in the non-contingent reward and the contingent reward period.
Model 9 shows that the the coefficient on the interaction variable Region*Treatment1 is -0.0275, statistically significant, denoting that the proportion of time an average worker in the treatment plant spends on work other than his or her production increases by about 2.75% under non-contingent reward.
Considering the average time a worker spends on work other than his or her production is about 2% of the working time per day, 2.75% increase is quite substantial. Model 10 shows a similar result for team leaders in the treatment plant. The coefficient on the interaction variable Region*Treatment1 is -0.0083, statistically significant, denoting that the proportion of time a team leader in the treatment plant spends on work other than his or her production increases by about 0.83% under non-contingent reward. 0.83% seems to be a small number, but it is nearly 50% increase in the proportion of time compared to the mean of about 2%.
Hence, the results have provided evidence for the hypothesis that team members and team leaders spend more time on work other than their own production in the non-contingent reward period. The increase in such time can largely be attributed to increase in cooperation, administration, and other work that are not part of their production.
Similar results have shown for the contingent reward period. In Model 9, the coefficient on the interaction variable Region*Treatment2 is -0.041, statistically significant, indicating that an average worker in the treatment plant spends about 4.1% more time on work other than his or her production under contingent reward. In Model 10, the coefficient on the interaction variable Region*Treatment2 is -0.0223, statistically significant, indicating that a team leader in the treatment plant spends about 2.23% more time on work other than his or her production under contingent reward. The results bolster the hypothesis that team members and team leaders spend more time on work other than their own production in the contingent reward period.
Conclusion
The three hypotheses focus on different aspects of the changes in both team leaders’ and regular workers’ behavior. The first hypothesis studies whether team leaders and team members gain benefits under non-contingent and contingent rewards, and whether team leaders or team members are the main beneficiaries under such rewards. It turns out that under non-contingent reward, both team leaders and team members earn more hours of piece rate, and team leaders earn an extra bonus on top of the increase in earned hours.
By looking at the results of the second hypothesis which focuses on workers’ performances in production, it can be concluded that the increase in earned hours can be mainly attributed to the increase of length in working hours for both team leaders and team members under non-contingent rewards. The output increases for both team leaders and team members, but there is no evidence to show that such output boost is due to increase in productivity.
According to the third hypothesis which looks at team leaders’ and team members’ time spent on work other than their own production, both team leaders and team members spend a larger proportion of time on work other than their own production, which potentially could involve cooperation, administration, meetings, and other work that are not part of their production, for example, training and workshops held by team leaders. Based on results of the second and third hypothesis, it can be deduced that under non-contingent rewards, team leaders have indeed made changes and made efforts to reciprocate and tried to boost team performances. Unfortunately, the main channel of increasing the output for workers is through working longer hours and not boosting productivity. Ideally, the boost in output is due to the team leaders’ more efficient allocation of jobs to the team members. It doesn’t turn out to be this case.
However, there are indeed significant results that support the reciprocal behavior of team leaders, although not producing the desirable results of boosting productivity. The limited length of the treatment period could potentially contribute to the results. If the team leaders had more time to figure out how to efficiently arrange work for team members, there might be an increase in productivity considering the team leaders’ motivation to improve team performances.
During the second treatment period, which is the contingent reward period, results seem to be less significant, probably due to the short length of this treatment period.
There are no substantial results to bolster the hypothesis that team leaders and team members could benefit from the contingent reward. Neither are there results that show output and productivity increase for team leaders and team members under contingent reward.
However, results of the second and third hypothesis do imply that under contingent rewards, team leaders have tried to make changes and make efforts to boost team performances to secure the team performance-tied bonus, since working time increases for both team leaders and team members, and they have spent more time on work other their own production, attributed to increase of time in cooperation, administration, meetings, training and workshops held by team leaders. Unfortunately, such efforts don’t seem to pay back as well as expected, but it doesn’t mean that given more time, contingent reward will not produce desirable results.
Limitations still apply to the study, such as that the treatment periods don’t necessarily last long enough to provide information to arrive at a solid and robust conclusion. In addition, the contingent reward period is introduced right after the non-contingent reward period. Thus, there is potential risk that the treatment effects are tainted. Lastly, however, the study still provides some intuition behind the dynamics of how team leaders and team members’ performance changes under the contingent and non-contingent reward to team leaders.
Chinese Manufacturing : A Case Study Written by Yuxuan Zhang
References : Chinese Manufacturing : A Case Study Written by Yuxuan Zhang
Bandiera, O., I. Barankay, and I. Rasul. “Incentives for Managers and Inequality among Workers: Evidence from a Firm-Level Experiment.” The Quarterly Journal of Economics 122, no. 2 (January 2007): 729–73. https://doi.org/10.1162/qjec.122.2.729.
Burke, C. Shawn, Kevin C. Stagl, Cameron Klein, Gerald F. Goodwin, Eduardo Salas, and Stanley M. Halpin. “What Type of Leadership Behaviors Are Functional in Teams? A Meta- Analysis.” The Leadership Quarterly17, no. 3 (2006): 288–307.
https://doi.org/10.1016/j.leaqua. 2006.02.007.
Cherrington, David J., et al. “Effects of Contingent and Noncontingent Reward on the Relationship between Satisfaction and Task Performance.” Journal of Applied Psychology, vol. 55, no. 6, 1971, pp. 531–536., doi:10.1037/h0032041.
Connelly, Brian L., Laszlo Tihanyi, T. Russell Crook, and K. Ashley Gangloff. “Tournament
Theory.” Journal of Management 40, no. 1 (2013): 16–47.
https://doi.org/10.1177/0149206313498902.
Deci, Edward L. “The Effects of Contingent and Noncontingent Rewards and Controls on Intrinsic Motivation.” Organizational Behavior and Human Performance, vol. 8, no. 2, 1972, pp. 217–229., doi:10.1016/0030-5073(72)90047-5.
Fehr, E., G. Kirchsteiger, and A. Riedl. “Does Fairness Prevent Market Clearing? An
Experimental Investigation.” The Quarterly Journal of Economics 108, no. 2 (1993): 437–59. https://doi.org/10.2307/2118338.
Gneezy, Uri, and John List. “Putting Behavioral Economics to Work: Testing for Gift Exchange in Labor Markets Using Field Experiments,” 2006.
https://doi.org/10.3386/w12063.
Hunt, J. G., et al. “A Multiple Influence Model of Leadership.” 1981, doi:10.21236/ada128679.
Jacobs, D., and L. Singell. “Leadership and Organizational Performance: Isolating Links between Managers and Collective Success.” Social Science Research 22, no. 2 (1993): 165–89. https://doi.org/10.1006/ssre.1993.1008.
Lazear, Edward P. “Leadership: A Personnel Economics Approach.” Labour Economics 19, no. 1 (2012): 92–101. https://doi.org/10.1016/j.labeco.2011.08.005.
Lazear, Edward P., and Sherwin Rosen. “Rank-Order Tournaments as Optimum Labor Contracts.” Journal of Political Economy 89, no. 5 (1981): 841–64.
https://doi.org/10.1086/261010.
Podsakoff, Philip M., et al. “Effects of Leader Contingent and Noncontingent Reward and Punishment Behaviors on Subordinate Performance and Satisfaction.” Academy of Management Journal, vol. 25, no. 4, 1982, pp. 810–821., doi:10.5465/256100.
Podsakoff, P. M., & Todor, W. D. (1985). “Relationships Between Leader Reward and Punishment Behavior and Group Processes and Productivity.” Journal of Management, 11(1), 55–73. https://doi.org/10.1177/014920638501100106
Rao G, DellaVigna S, List J, Malmendier U. Estimating Social Preferences and Gift Exchange at Work. Working Paper.
Chinese Manufacturing : A Case Study Written by Yuxuan Zhang
https://scholar.harvard.edu/files/rao/files/socialpreferencestowrdsemployerfeb16.pdf
Leading Artificial Intelligence and Financial Advisor – Rebellion Research
Chinese Manufacturing : A Case Study
