Quản trị kinh doanh - Chapter 10: Rewarding performance

Chapter 10: Rewarding Performance4/8/2013Chapter 10: Rewarding PerformanceHow Strong Should Incentives Be? IntuitionSelling the JobTaxicab DriversSecurities TradersWait Staff in RestaurantsOutsources SalesMiddle ManagersImperfect Evaluations and Optimal IncentivesUncontrollable RiskControllable RiskManipulationDistortions and Multitask IncentivesSummary: How Strong Should Incentives Be?Value of Employee EffortImportance of SortingControllable and Uncontrollable RiskRisk AversionTrust and SubjectivityDistortion and Multitask IncentivesPotential Manipulation Paying for Performance: Common ExamplesRewards or Penalties?The Ratchet EffectLump Sums, Demotions, or PromotionsCaps on RewardsApplicationsProfit Sharing and ESOPsPerformance EvaluationPay-Performance RelationshipCounterargumentsOrganizational Form and ContractingFranchisingCost Plus versus Fixed FeeMotivating Creativity Summary4/8/2013IntroductionWe now think about how to tie that evaluation to rewardseasier in practice than evaluationLet’s first note what we don’t focus on: the level of paylevel of pay = expected value of the whole package, not just salarywhat does the level depend on?market supply & demand for this type of skillsextent of uncontrollable risk in the pay packagedesire to attract better talent through more at-risk payWe examine 5 common types of reward structures4/8/2013Chapter 10 – Rewarding PerformanceAfter completing this chapter, you will be able to address the following questions:What does economics say about paying for performance?How do real world incentive plans differ from the economic ideal?What sort of results could one expect from a properly designed incentive scheme? Which settings are most appropriate for such schemes?What are the potential pitfalls to avoid in implementing a pay-for-performance system?What is the difference between gains from sorting and gains from incentives?How can an incentive scheme help with sorting out talented employees?4/8/2013How Intense Should Incentives Be?“Slope” of pay-performance relationship drives incentives. Stronger if Incentives reinforce strategy moreintrinsic motivation weak, effort is more costly to employeeeffort has more impact on firm valuemore “controllable risk”Evaluation is more effectiveless “uncontrollable risk”employee less risk aversemeasure has low distortionmeasure harder to manipulatesubjective evaluations are done well, & trusted moreSorting for talent is more important4/8/20132. Other Pay-Performance ShapesGeneralizations of slopeThese shapes often encountered in practicewhy / when use each?what problems might you encounter?PayPMBonus With Floor / Quota“Punishment”Lump-Sum, Promotion or DemotionBonus w/ CapLessons: Pay-Performance ShapeRisk & opportunity  reward or punishment shapefloors reduce employee’s downside riskcan facilitate stronger incentiveschange incentives for risk takingex.: employee stock optionscaps may reduce gaming, but can hurt recruitment / retentionnote that pay-performance shape may be implicit when helpfultargets often motivate gaming if performance is near targetSimplicity is a virtueAll jobs have implicit incentivespromotions are most important for many firms, middle managers4/8/2013The Ratchet EffectIf your employee earns high pay from an incentive plan, you may be tempted to change the plan to lower paywhy?Be careful! Doing so may reduce future incentivesgood performance today Þ lower rewards in future Þ “Quota Restriction” this is often called the “Ratchet Effect”application: growth-based targets & measures often have this problemTwo earlier ideas help us understand the Ratchet Effectthe Hold-Up problemImplicit Contracting4/8/2013ImplementationSweat the details on evaluationa fair process is essential for successPay for good performance (& luck)avoid Ratchet EffectsTransition graduallyexpect turnover if a new plan is introducedconvert raises to bonusesprovide initial insurance against downside risk, then phase outstart with test cases & role modelscommunicate & listenReserve the right to adjust over timeFocus incentives where they matter mostUse a 3-stage process to design the planKISS (Keep It Simple, Smarty)No plan is perfect but you can make yours better4/8/2013Marginal IncentiveAs introduced in Lazear & Gibbs Ch9 Marginal incentive for an employee must consider the cause-effect relationship betweenIncremental effortChange in performance measureResponse of compensation systemMarginal benefit to the company Question: Do you think that the incentives connection can be further complicated or diluted by extraneous factors?4/8/2013Marginal This – Marginal ThatRandomness in measuring performance results in a risk premium that needs to be included in the compensationSignificant in that it is proportional to the SQUARE of the (std) errorIt is also proportional to the incentive itself. Why? Fig. 10.2a, T must be set high enough to accommodate riskMulti-tasking IncentivesIdentified tasks distort effort away from non-identified tasksBalancing incentives must also account for possible differences in measurement errorTasks can be separated into different jobs ORA composite performance measure can be designedThis can be complicated – what’s the intuition?4/8/2013Performance Pay MethodsRewards vs. PenaltiesPenalty scheme may be appropriate when performance upside is cappedReward scheme may be appropriate when higher performance increases value; downside floor is for employee risk aversionRaising T as effectiveness improves is commonplace, despite suggestions to the contrary (Ratchet Effect)Lump Sum Lazear & Gibbs says this is the most commonplace. Why?What reasons are offered for the use of Pay Caps?4/8/2013PFP and Profit SharingTheory suggests that connection between employee behavior and firm profit is too small in the aggregateFree riders are not individually worth enforcingConnection may be direct enough for small groupsExplanations for actual use of profit-sharing (revisited)Free-rider enforcementPFP plans often implemented at nascent firms4/8/2013PFP and ContractingContracting structure permits incentive schema that may not be possible with employeesFranchisingClose to perfect outcome measureParent company retains strategic decision powerCost PlusUse with well-defined tasks and known inputsFixed FeeUse when outcome is well-defineCan help to encourage creativity in processSpecification is important when quality is a concern4/8/2013Motivating CreativityDoes PFP stifle freedom of thought?Outcome- or output-based objectives can incorporate creativity 4/8/2013What Makes a Good Boss?[Google’s Project Oxygen]Be a good coachspecific, constructive feedback, balancing negative & positiveregular meetings, presenting solutions tailored to employee strengthsBe productive & results-orientedfocus on what employees want to achieve & how to helphelp prioritize & remove roadblocksEmpower & don’t micromanageExpress interest in employee’ success & personal well-beingHelp employees with career development4/8/2013What Makes a Good Boss?[Google’s Project Oxygen]Be a good communicator & listenencourage open dialog & listen to issues & concerns of employeesHave a clear vision & strategyinvolve in setting & evolving vision & making progress toward itHave key technical skills so you can help advise the team4/8/2013What Makes a Bad Boss?[Google’s Project Oxygen]Have trouble making a transition to the teamsometimes fantastic individual contributors are promoted to managers w/out necessary skills to lead peopleoutside hires don’t always understand managing at GoogleLack a consistent approach to performance management & career developmentdon’t help employees understand how these workdoesn’t coach employees on their options to develop & stretchnot proactive, waits for employees to come to themSpend too little time managing & communicating4/8/2013Compensation Design: SummaryIdeally, use a 3-stage process(hardest, most important) Evaluate performance(easier) Tie pay to performance(easiest) Set the overall level of pay4/8/2013“The Value of Bosses”[Lazear, Shaw & Stanton, 2011]Managers have much larger effects on productivity than peerspeer effects nearly disappear after controlling for supervisormanager raises or lowers productivity of all subordinatesManagers vary substantially in qualityreplacing a boss in bottom 10% with one in top 10% raises subordinate productivity 11%Good managers increase productivity of best performers, even more than that of worst performersManagers affect productivity via monitoring; esp. “coaching” Improving 1st-line “management” should be a priority4/8/2013Recap – Chapter 10Economics says the ideal PFP system pays all the commission gained to the employee and charges a fixed fee up frontThere are examples of “Selling the Job,” but more often there is a “draw” (threshold) required before commissions begin At Safelite, an actual case of switching to incentives increased productivity by 36%Dual effect of incentives:Sorts out those who find the incentives attractiveIncreases individual effort compared to flat payFor PFP to work, outcomes should be measurable and the task should be containable (within an individual or a company)4/8/2013