资讯性激励补偿和缓冲库存的选择外文翻译.docx

1、资讯性激励补偿和缓冲库存的选择外文翻译外文文献翻译译文一、 外文原文原文：Informativeness, Incentive Compensation, and the Choice of Inventory BufferBaiman, Stanley; Netessine, Serguei; Saouma, Richard Most of the agency theory work in managerial accounting has studied the relation between performance metric properties e.g., informativ

2、eness, precision, congruity_ and optimal incentive compensation, holding fixed the firms other organizational design decisions_e.g., hierarchical structure, job assignment, production technology_. However, by holding fixed these other organizational design choices, one cannot study whether and how t

3、hey affect the performance metric properties. As Hemmer _1998, 321322_ notes, “the value of a performance measure is determined not simply by its congruity and precision but by its influence on the optimal organizational design Much of the recent theoretical accounting literature has largely ignored

4、 complementarities between performance measures and organizational design.” In this study, we expand on Hemmers _1998_ observation by studying how the informativenessand incentive properties of a performance metric are influenced by one of the firms organizational design choicesthe size of its inven

5、tory buffers. The introduction of just-in-time _JIT_and, more generally, lean manufacturing has led to an increased emphasis on controlling and reducing inventory levels. Managerial accounting has responded to this reduction in inventory with new costing techniques such as back flush costing _Horngr

6、en et al. 2008_. However, we show that the choice of inventory buffers has a more subtle effect on the design of the managerial accounting system, in that it affects the informativeness of performance metrics produced by the managerial accounting system. Similar to our model, Hemmer _1995, 1998_ and

7、 Gietzmann and Hemmer _2002_ examine how different workflow arrangements between agents affect the information available for contracting, and the incentives facing agents. Our work is distinct from these models that do not consider buffer size as one of the principals choice variables. Nagar et al.

8、_2009_ examine the role of inventory buffers in agency problems, although unlike our model, in their setting buffers are filled by agents in order to signal private information and buffer size is not a choice variable. Alles et al. _1995_ do focus on the effect that the choice of buffer size can hav

9、e on the informativeness of performance metrics. The major difference between our work and theirs is that they do not formally model how buffers affect performance metrics, but instead assume a monotonic relation. In contrast, we formally model the inventory process and derive a non-monotonic relati

10、on between buffer size and the informativeness of the agents performance metric. The model consists of a single-period and a firm comprised of a risk-neutral principal and agent. Contracting takes place at the start of the period when the agent is hired to set up a workstation to process incoming in

11、termediate units _e.g., test completed computers or weld automotive chassis_. The intermediate units arrive stochastically into the workstations incoming inventory buffer at a commonly known, mean arrival rate of _. The buffer has capacity b _ N, allowing a maximum of b _ 1 units to be held in front

12、 of the workstation while the workstation processes one unit. If an intermediate unit arrives and the buffer is not full, then the unit is added to the inventory. However, if the buffer is full, then blocking occurs and intermediate units cease arriving until there is space in the buffer, whereupon

13、the intermediate units begin arriving at the same stochastic rate as before. Our assumptions regarding arrival and processing rates follow the standard M/M/1/b queuing model with finite buffers used in the operations management literature.7 Note that while we are modeling a single-period in that the

14、 agent chooses his setup effort only once at the start of the period, we analyze the stochastic evolution of the workstations throughput over the entire period. The transient behavior of throughput in queuing systems with finite buffers cannot be described analytically, although the steady-state beh

15、avior can be described in closed form.8 As a result, the operations management literature typically uses simulations to analyze the transient behavior of queuing models with finite buffers and closed-form analysis to examine steady-state behavior. In contrast, agency models have emphasized rewarding

16、 transient behavior, but simplify the production process _e.g., using the LEN model_ to achieve closed-form solutions.9 In order to incorporate stochastic intermediate unit arrival rates, stochastic processing rates, and finite inventory buffers in our analysis, we focus on the problems steady-state

17、. That is, we assume that the principal is interested in maximizing her steady-state expected profit and that the agent selects his effort to maximize his steady-state expected utility.10 This is a reasonable approximation if, as we assume, the production process reaches steady-state fairly quickly.

18、 Our assumption that the agents effort is expended in setting up the workstation before production begins, rather than managing the production process as it evolves, is also consistent with our emphasis on the steady-state behavior of the workstation. An example of our modeled setting is a robotic l

19、ine that welds automotive chassis. The agents main task is to program the spot-welding robots. This task involves identifying the optimal position of the chassis, identifying the optimal position of the robotic arms, and making sure that robot paths do not cross each other or the chassis itself _so

20、that nothing is damaged in the process_ while ensuring that welding is done in the least possible number of steps. All of this planning and programming setup is done once by the agent at the beginning of the shift and thereafter, as long as the workstation is in control, the agent does not need to i

21、ntervene. Clearly, the actions of the agent in programming the robots affect the time it takes to weld the chassis _on average_, but the agent does not reprogram the robot once production has begun.Most of the time they also have a high degree of education or expertise. They include anywhere from a

22、quarter to a third of the workforce, but not everyone who uses knowledge. If you are digging ditches, you may have some knowledge on the job, but its not the primary purpose of what you do.Are companies doing a good job of managing and improving the performance of knowledge workers? Theyre not. What

23、 most organizations do is HSPALTA: Hire smart people and leave them alone. Weve spent a lot of effort recruiting knowledge workers and assessing how capable they might be before we hire them. But once theyre hired we dont do a lot to improve their performance. Process improvement has mostly been for

24、 other workers: transactional workers, manufacturing workers, people in call centers. All the serious approaches to improving work have largely escaped knowledge work. We let knowledge workers get away with saying theres no process to their work, that every day is different. We dont measure much of

25、anything about knowledge work. If we dont measure knowledge work, why do you think theres room to improve knowledge worker productivity and performance?Its a pretty well-informed hunch. People improve processes all the time; they just havent done it with knowledge-work processes as much. Its an extr

26、apolation of the same logic in other work, that processes can be improved. Here is one number that indicates performance and productivity can be improved: IDC found that 1,000 knowledge workers can lose as much as $6 million a year just searching for nonexistent data, or repeating work that has alre

27、ady been done. Is it possible every knowledge worker is working to his or her potential? Its possible, but unlikely. We can get a lot better at improving their performance. Why hasnt knowledge management helped more in the effort to improve knowledge-worker performance and productivity? Knowledge ma

28、nagement was an early attempt to intervene in knowledge work. For the most part, it wasnt particularly successful, because we didnt look closely at how knowledge workers did their work. We tried to be too broad in our focus. Most organizations simply created one big repository for all knowledge and

29、all workers. The only way we can get people to use knowledge on the job is to understand how they do their jobs, and then figure out some way to inject knowledge into the course of their day-to-day work, not make it a separate thing you have to consult when you need knowledge. We have to be much mor

30、e targeted in approaching knowledge management. We have to target a specific job. And the best way is to use technology to bake the knowledge into the job. Note that the mean arrival rate of incoming intermediate units and therefore the maximum mean throughput is _. However, if the principal induced

31、 the agent to choose r _ _/_1 _ h_ so that the workstation was set to process the intermediate units at the same average rate at which they arrived, then the instantaneous probability of starving the workstation would be ps = 11+b _ 0, which in turn would limit the workstations mean steady-state thr

32、oughput to K = _b1+b _. That is, inducing the agent to set the workstation to process the intermediate units at the same average rateat which they arrive would significantly limit steady-state throughput and induce a non-trivial probability of starving. Thus, the principal has a natural incentive to

33、 induce the agent to set the workstation to process units at an average rate that is strictly greater than the average rate at which the intermediate units arrive, as commonly assumed in the operations management literature _Hopp and Spearman 2001_. Knowledge workers have a lot of power, and you dont want to impose th