Topic 8; Performance Evaluation (2017)Apunte Inglés
Vista previa del texto
TOPIC 8; PERFORMANCE EVALUTION
OBJECTIVES OF PERFORMANCE EVALUATION
Compensation: performance evaluation is necessary for remunerating according to
productivity (salary, bonus, prizes, dismissal…)
Communicate values: evaluation systems allow for communicating the firm’s values and
objectives to employees.
Legal system: evaluation systems protect firm’s decision on promotions and dismissals.
Feedback in a well designed performance evaluation system allows for: improve task allocation inform employees on areas that need improvement identify training needs provide information for contracting.
RELEVANT DIMENSIONS OF PERFORMANCE EVALUATION Evaluation Unit Individual Team Division Evaluator Close superior Comitee of superiors Peers Time Horizon What's the minimum time span necessary for reliable evaluations? Degree of subjectivity Subjective Objective Both indicators Scale of evaluation 0-5 0-100 Good/Bad OBJECTIVE PERFORMANCE EVALUATION Objective performance evaluation is based on observable and quantifiable indicators (Sales volume, number of clients,…) of the evaluated unit. Its main advantage is that allows firms and employees to clearly agree upon a target. Some other benefits of objective measures (sales, profits) are easy to observe and verify, and there won't be much debate between principals and agents. Also, they are cheap.
When setting a benchmark (target), usually firms estimate how much time a particular task require with the goal of determining the most effective work method. Motion studies involve the systematic analysis of work methods, considering the raw materials, the process, the activity at each step... Time studies employ a wide variety of techniques for determining the duration of a particular activity under certain conditions. Usually these studies can shorten the time required. It is usually done by work sampling, which involves selecting a large number of observations taken at random intervals and observing how long employees take performing various components of the job. However, time and motion studies are expensive and must be redone whenever product changes (also, there can be opportunistic behavior: workers can underperform when doing the sampling of the study). When setting a benchmark, we also can look to the past year performance and take it as the benchmark to exceed. This has a bad effect: since benchmarks just go upwards, if workers know that their performance is going to count for the future, they might underperform so that in the next year the benchmark will be lower, and that is harmful for productivity → Ratchet Effect.
Problems appear when it is costly to observe employee output; costs incurred for observability, verifiability and contractibility. More costs are incurred then for generating performance measures. In our simple principal-agent model, the employees output is used in setting compensation. However, it depends on random factors and it is optimal to expend additional resources to measure the employee’s effort level more precisely and so, reduce the employee exposure to this random factors and at the same time reduce the compensation differential. In some cases, observing and measuring the employee's outcome becomes so expensive that the firm begins to look for alternative variables that capture employee performance.
Also, too simple indicators of the employee's performance may favor unwanted behaviors: Sacrifice quality for more quantity May increase indicator level without increase productivity (e.g. new technology; expanding selling area while lowering sales per square meter…) Negative influence on cooperation (ignored by indicator) Concentrate on quantifiable tasks only.
Making benefits in the short run but might cause problems in the long run On one hand, you want to incentivize the sales and the profits, and make the value of the company grow in the market, making the shares be more profitable and incentivize people to buy them, but you also want to provide a good future to the company.
To do performance evaluation to the high level managers and decrease the risk that they do something in the short run that will benefit the company but will backfire in the future, as an owner, as a shareholder, you have to decide the compensation that doesn’t incentivize that behavior. To design the incentive compensation for CEOs and top managers, you have to think about sort and long term, and the market value → you can pay them not only with cash, but also with actions and participations of the firms, so they have to do good actions to make their shares profitable. This way, you make them owners of the companies (by a small percentage, but owners), and make them part of the residual claimant (also you have to say that they cannot sell them "in the short run", they'll have to wait 6/7 years to sell them). They will want money in the future, so they won't have opportunistic behavior. The problem with that is than in this situation the risk for the manager is much higher if he gets his bonus in the long run → the risk adverse agents will prefer a company w/ short term performance pay, and you might attract risk searching managers.
Another problem is the uncontrollable random variables. What you can do as a solution to shield employees of risk, instead of looking to absolute sales, you look to relative sales.
You need a reference performance, either from outside the firm (you compare to other firms in the industry, so if there is a crisis, everyone notices it) or from inside the firm (you sum the sales of each of the sales managers and divide it for the number of manager, and you know the media → if they are above it, you get a bonus, and if you are under it, you don't. this might generate problems between the group, getting high performers punished by the group, they get more pressure from de group, getting unmotivated, and this is not what the firm wants). The group from which you calculate the reference point should be really similar to your group.
Employees may often behave opportunistically in ways that affect their performance evaluation. We can think about two examples of such opportunism: Gaming the system and Time horizon.
→ Gaming: As output is the measured variable, employees engage in dysfunctional activities that reduce quality to increase personal output, which can be extremely harmful for the firm.
→ Horizon Problem: Output measurement usually focuses on the short term. That might make employees to focus on the short run and not thinking in long run actions that benefit the firm, which may lead in a high surplurus in the short term and a huge deficit in the long-term. (e.g. a salesman near retirement).
RELATIVE (OR SUBJECTIVE) PERFORMANCE EVALUATION Multiple employees performing similar tasks can provide useful signals about the random errors affecting individual employees, so it is useful to use one’s performance relative to the others, looking to the average. This reduces the risk employees bear and so reduces firm’s total costs by lowering differentials. However, many employees will be tag this type of performance as unfair, due to the fact that they think they are better than their partners. We can evaluate performance with respect to other units inside–within- or outside the firm-across- firms.
There are different ways of evaluating relative performances: Standard-Rating-Scale System Goal based system With scales you can rate from 1 to 5 many different things: task competition, decisionmaking, teamwork, response to training….
Goal based are easy things, such as number of clients, and you can easily see how the worker has performed.
The evaluated employees can cause potential problems with supervisors; e.g. employees might not trust the objectivity of the supervisors. Evaluated employees usually believe that they are more productive than others, and they do not want to take into account the opinion of their supervisors. Also, as it is done a group average of performance, the group has incentives to punish the extremely productive employees who raise that mean, or an explicit employee collusion to hold down the benchmark can occur (everyone agrees to produce less). Moreover, employees might try to get themselves classified into a reference group with weak performance so that they seem to be above the average; and they'll have the incentive to recommend hiring less competent new employees (it improves the relative performance of the current employees).
Consider strategic behaviors by: evaluated employees they adapt behavior to supervisors’ preferences or evaluation criteria influence costs – politicking → is rotating supervisors a solution? (it's not a really efficient one, since it's costly) supervisors o may try to avoid conflicts that could rise from explicit evaluation equal points to everybody (does it happen to students?) more points to old or more powerful employees may use only objective criteria in evaluation.
firm: reneging on promises → reputation as safeguard.
o o With relative performance evaluation, the risk of difference between groups is smaller, because everyone got the same conditions, and there are no risks that may harm the performers in general REMEMBER TEAM PRODUCTION The core concept of team production is complementarity. There are also three core items, ownership (residual claimant), incentives (how to incentivize at a team level) and group size (especially related to monitoring) Productivity of one member affects the others → complementarily Paying for team production → incentive to cooperate but also there's a risk of free riders.
If you want to evaluate at team level, you have to think of monitoring, and the cost of controlling free-riders → is it smaller, larger or equal to the benefits of having people together as a team? I the benefits of having people as a team as higher than the costs of free-rides, teams are a really good decision, and vice versa.
If you think about the control of team members, and teams, you have non specialized (mutual or by term, and 360º review → we know x about the agent, but we want to know about other dimensions of his performance. Usually, the boss or the principal evaluates the agent (1 dimension). Or, you teammates can evaluate you, and they might have more information than the supervisor, so they have specific knowledge that goes beyond the knowledge of the supervisor, so they can rate you better. There's a risk: the agent is not only evaluated by the other, but also evaluates the others. There's also external evaluation, control by the business partners that you have relations with, such as suppliers or clients.
Finally, employees can evaluate their supervisors, and we have the same problem that with "peer" evaluation. This is a really good review, because you have lots of different evaluations for only one team member, and you can get a really good idea about the performance of a specific team member. There are not many jobs that uses this evaluation, because it takes a lot of time and high opportunity cost. If it’s a huge group, this is very costly.) ...